D R A F T TABULATIONS OF OZONE REACTIVITY SCALES FOR VOLATILE ORGANIC COMPOUNDS Documentation of Software and Data Files as of November 28, 1990 Prepared by William P. L. Carter Statewide Air Pollution Research Center University of California Riverside, California 92521 Prepared for Air Force Engineering and Services Center Tyndall Air Force Base, Florida 32403-6001 Dr. Daniel A. Stone Project Supervisor PREFACE This document, and the software it describes, was prepared at the Statewide Air Pollution Research Center at the University of California, Riverside (SAPRC/UCR). The author was funded by an assignment agreement with the Air Force Engineering and Services Center, Air Force Engineering and Services Laboratory (AFESC/RDVS), Tyndall Air Force Base, Florida 32403, and this work was carried out under the supervision of Dr. Daniel A. Stone of AFESC/RDVS. However, the word processing for this documented was funded by the California Air Resources Board through contract no. A932-094 to SAPRC/UCR, and the reactivity scales in the data files discussed in this document were prepared under separate funding and are described in a separate draft report which has been submitted to the EPA. This document is a draft which has not been reviewed by the Air Force, the California Air Resources Board, or the EPA. Mention of trade names does not constitute endorsement or recommendation for use. The software and reactivity factors described in this document were developed for scientific research purposes only. Those who use this software or these reactivity factors for other purposes do so at their own risk. The author, the University of California, or the agencies funding this work may not be held liable for any damages due to use of software or the reactivity scales discussed herein. SUMMARY This document describes computer data files and associated software giving ozone reactivity factors for volatile organic compounds (VOCs). These reactivity scales are described in the draft report entitled "Development of Ozone Reactivity Scales for Volatile Organic Compounds" (Carter, 1990a), which has been submitted to the EPA for evaluation. The scales give estimates of effects on ozone formation of adding small amounts of individual VOCs to the existing emissions in various ozone pollution episodes or "scenarios". These "incremental reactivities" were calculated using "floating box" type airshed models employing a detailed chemical mechanism for atmospheric reactions of organic compounds, and employing "EKMA" models of ozone pollution episodes in 12 urban areas in the United States. The data files give the incremental reactivities of the VOCs in the six general or multi-scenario scales documented in the draft EPA report, including the "maximum reactivity" (MaxRct) and the "maximum ozone" (MaxO3) scales, and the four multi-scenario scales derived using various averaging methods and methods of quantifying ozone impacts. In addition, they give the reactivities of the VOCs in the 93 "single-scenario" scales used to develop the general scales. Reactivity factors are given for 288 types of VOCs, which account for most of the VOC emissions in present inventories. However, because of uncertainties in chemical mechanisms and limited data, reactivity factors for most are highly uncertain. Of the 223 VOCs which have been classified as to their estimated level of mechanistic uncertainty, only 37 have chemical mechanisms which have been tested at least to some extent using environmental chamber data. The mechanisms and reactivity assignments for 65 VOCs are undocumented and preliminary, and their reactivity factors should only be used for assessing reactivities of complex mixtures. Software is included with these data files to produce tabulations of reactivity data from selected scales for selected VOCs or mixtures of VOCs. Data files giving compositions of several VOC mixtures of interest are also included. This software can be used to evaluate how ozone reactivities vary among the 99 general or single-scenario scales, and to compare ozone reactivities of compounds or mixtures. This is being provided to the AFESC so they can conduct an initial assessment of the utility of ozone reactivity scales in their efforts to assess ozone pollution problems caused by emissions of VOCs resulting from various Air Force operations. The data files and software are provided on a 1.2 MB 5 1/2" computer diskette which can be read and run on IBM-compatible PC's using the MS-DOS operating systems (version 3 or higher). AT-class computers with a hard disk and a math co-processor are required to use this software. The software is provided both in the form of executable (.EXE) files and as FORTRAN-77 source files which can be compiled by the Lahey F77L compiler (version 4.0 or higher). CONTENTS I. INTRODUCTION II. SUMMARY OF DATA FILES AND SOFTWARE III. REACTIVITY SCALES A. Single Scenario Scales 1. Base Case Scenarios 2. Maximum Reactivity and Maximum Ozone Scenarios 3. "Averaged Conditions" Scenarios 4. Tabulation and Nomenclature of Scenarios B. General and Multi-Scenario Scales 1. The MAXRCT and MAXO3 Scales 2. Base Case Relative Reactivity Scales C. Integrated Ozone Reactivity Scales D. Kinetic and Mechanistic Reactivities E. Format of Reactivity Data (.RCT) Files 1. Single-Scenario .RCT Files 2. MAXRCT.RCT and MAXO3.RCT Files 3. Multi-Scenario .RCT files (BASE.RCT and BASEI.RCT) IV. VOCs AND MIXTURES A. Individual VOCs and the VOC.PRM File 1. Classes of VOCs on Reactivity Data Base 2. Format of the VOC.PRM File B. Mixtures and Composition Data (.CMP) Files C. Examples of Composition Files V. DESCRIPTION OF SOFTWARE A. Running REACTTAB 1. Interactive Mode 2. Batch Mode 3. REACTTAB Options in MODELING.PRM B. Running RENORCMP C. Installation of Software and Data Files VI. REFERENCES I. INTRODUCTION The formation of photochemical ozone continues to be an intractable problem in many areas. Ozone is not emitted directly into the atmosphere; it is formed by the interactions of volatile organic compounds (VOCs) with oxides of nitrogen (NOx) in sunlight, and effective ozone control strategies must focus on both of these precursors. VOC controls can be particularly effective in reducing ozone in urban centers where NOx sources are abundant. VOCs can differ significantly in their effects of their emissions on ozone formation (their "ozone reactivities"), and the development of effective strategies to reduce ozone requires an ability to quantify these differences. Although there are a number of ways to do this, among the most useful is that of "incremental reactivity". This is defined as the change in ozone formation caused by adding a small amount of a VOC to the existing emissions in an ozone pollution episode, divided by the amount of VOC added. This cannot be measured experimentally, but can be estimated using computer airshed models, provided that the models have an adequate representation both of the conditions of the episode and of the kinetics and mechanisms of the reactions of the VOC which affect ozone formation. Incremental reactivities of VOCs depend both on the nature of the VOCs atmospheric chemical reactions and on the environment in which the VOC is emitted. The most important environmental factor is the ratio of total emissions of reactive organics to NOx in the airshed (the ROG/NOx ratio), but other aspects also have non-negligible effects. The fact that VOC reactivities (i.e., incremental reactivities) depend on environmental conditions complicates the development and use of VOC reactivity scales for regulatory and control strategy assessment purposes. One approach which can be employed is not to attempt to develop a reactivity scale which is valid for all conditions (since this is impossible), but to determine what reactivity scale would give the best overall results when it is used. To investigate this, the author calculated incremental reactivities for a wide variety of VOCs for a set of scenarios representing a variety of airshed conditions, and used them to develop various "generalized" or "multi-scenario" reactivity scales, which were then compared. A detailed report describing this study has been submitted to the EPA (Carter, 1990a). While not conclusive (since the degree to which the set of scenarios employed represents a realistic distribution of airshed conditions is highly uncertain), the results of this study were highly suggestive that a maximum reactivity scale -- i.e., a reactivity scale for NOx conditions where VOCs have their greatest effect on ozone formation -- may give a good approximation to "optimum" scale for general use. The predictions of the maximum reactivity scale were found to correspond reasonably closely to those in the multi-scenario scale derived to minimize total predictions in absolute ozone changes caused by adding the VOCs to all the scenarios, and also tended to correspond reasonably well to predictions of effects of VOCs on integrated ozone in a variety of scenarios, including those far from maximum reactivity conditions. Because of this, the California Air Resources Board has recently proposed using a maximum reactivity scale to derive "reactivity adjustment factors" for emissions from alternatively-fueled vehicles in its "clean fuels/low emissions vehicle" regulations (CARB, 1990). However, the relative reactivities of the individual VOCs where highly variable from scenario to scenario, and the average of ratios of reactivities (relative to peak ozone concentrations) was not well predicted by the maximum reactivity scale. The latter was found to correspond better to the scale developed for NOx conditions where maximum ozone concentrations are formed (the "maximum ozone" scale). These results, and the conclusions arising from them, are described in more detail in the draft EPA report (Carter, 1990a). The work described in the draft EPA report represent in many respects a preliminary investigation of the development of general ozone reactivity scales for VOCs. The chemical mechanisms used to calculate the reactivities of many of the VOCs are highly uncertain, and a major effort is underway to improve them and provide data needed to test them. The representativeness of the scenarios employed to develop the scales is probably even more uncertain; the scenarios employed are not only based on "floating box" type models which are obviously oversimplifications of the real world physical situations, and are also based on highly limited data bases. No effort was made in this project to evaluate the appropriateness or accuracy of the scenarios used; they were chosen because they were used in previously published studies (Gery et al. 1987; Whitten, 1988) and were made available to the author in computer readable form, and because they represented a represented a sufficiently varied set of conditions for the purpose of evaluating how to derive general or multi-scenario reactivity scales. Work on developing an improved set of scenarios for reactivity assessment is being carried out by the staff of the California Air Resources Board (Croes, private communications, 1990) and others. Nevertheless, the reactivity scales developed in the work described in the draft EPA report (Carter, 1990a) represent the current state of the science in assessments of incremental reactivities of individual VOCs for a wide variety of airshed conditions. As such, those investigating the use of ozone reactivity scales in developing VOC control strategies, or those who wish to assess the relative ozone impacts of different types of emissions, may find these scales useful. For example, they are being widely used for assessing differences of ozone impacts of exhausts from vehicles using alternative fuels (e.g., see Lowi and Carter, 1990). In addition, the various operations of the United States Air Force results in emissions of many types of VOCs which contribute to local ozone formation, and these reactivity scales might prove to be useful tools to the Air Force in assessing the relative impacts of these emissions. Therefore, as part of an assignment agreement for the AFESC, the author is making available data files giving all the reactivity scales developed in that work, and software and data files designed to assist in using these data files for assessing reactivities of individual VOCs and mixtures. These are described in this document. II. SUMMARY OF DATA FILES AND SOFTWARE The files and software described in this document are distributed on a 1.2 MB 1/2" MB computer diskettes which can be read by IBM-AT-compatible computers. A listing of the files on the diskette is given in Table 1. These include: (1) scenario reactivity data (.RCT) files, where each file gives the reactivity data for all VOCs for an individual scenario or for a general or multi-scenario reactivity scale; (2) the "VOC.PRM" file giving a complete listing of each type of VOC for which reactivity data are available, and other information related to the VOC such as its molecular weight and mechanism uncertainty code; (3) "composition" (.CMP) files giving detailed composition data for selected mixtures of VOCs of interest, including the mixture used to represent the base case ROG emissions in when the reactivity scales were calculated; (4) FORTRAN source and executable files for the REACTTAB program, which can be used to tabulate reactivities of selected VOCs and mixtures for selected scenarios or reactivity scales; (5) FORTRAN source and program files for a RENORCMP program, which can be used to assist in the preparation of composition files needed for assessing the reactivities of mixtures; (6) a "MODELING.PRM" file which is used by the programs to obtain control options and to locate the files which it needs (and which the user may need to edit if the recommended options or directory structure is not used -- see Section V); and (7) various command (.BAT) files which might be useful for installing the software and compiling the programs. Table 1. Listing of Files on Distribution Diskettes. Recommended directories for installation of files are also shown. ** NOTE ADDED 6/21/91 ** THIS IS SUPERCEDED BY REACT\FILES.DOC AND SOURCE\PGMS.DOC ------------------------------------------------------------------------- File Name Description ------------------------------------------------------------------------- Main Directory READ ME Installation instructions and disclaimer. INSTALL BAT Copies files to hard disk and creates sub-directories. FILES DOC Description of files (this table plus other information useful for installation). FILES1 LIS List of all files on diskette #1. REACTTAB DOC Draft of manual documenting the software and data files. SCENARIO DOC Listing of scenarios used to derive reactivity scales. CMPFILES DOC Listing of distributed composition files. REACTTAB EXE Executable programs (math co-processor required). Source RENORCMP EXE files in SOURCE sub-directory. VOC PRM Parameters for VOCs. MODELING PRM Program control parameters. The following can be used as REACTTAB input. BASESCEN LIS All distributed "base case" scenarios. MAXRSCEN LIS All distributed "maximum reactivity" scenarios. MAXOSCEN LIS All distributed "maximum ozone" scenarios. ALLVOC LIS All VOCs on VOC.PRM DOCVOC LIS All VOCs tabulated in EPA report (Carter, 1990a). ALLCMP LIS All distributed composition files. CMPFILES Subdirectory. * CMP Composition files. See ALLCMP.LIS for list of files. RCTFILES Subdirectory. General reactivity scales. MAXRCT RCT Maximum reactivity (MaxRct) scale. MAXO3 RCT Maximum ozone (MaxO3) scale. Multi-scenario relative reactivity scales. BASE RCT Reactivities relative to peak ozone. BASEI RCT Reactivities relative to integrated ozone. Single-scenario scales. AVGCONMR RCT "Averaged condition scenario" maximum reactivity scale. AVGCONMO RCT "Averaged condition scenario" maximum ozone scale. ?-???BS RCT "Base case" scenario scales. See BASESCEN.LIS. 21 Files. ?-???BSI RCT "Base case" scenario integrated ozone scales. Append "I" to names in BASESCEN.LIS. 21 Files. ?-???MR RCT "Maximum reactivity" scenario scales. See MAXRSCEN.LIS. 28 Files. ?-???MO RCT "Maximum Ozone scenario scales. See MAXOSCEN.LIS. 21 Files. SOURCE Sub-directory "Batch" files to compile and link programs. Requires Lahey's F77L (version 4) and OPTLINK programs. BLDALL BAT Compile and link both REACTTAB and RENORCMP. RTBLD BAT Compile and link REACTTAB. RNCBLD BAT Compile and link RENORCMP. RTLINK BAT Link REACTTAB. RNCLINK BAT Link RENORCMP. REACTTAB FOR Main module for REACTTAB. LDDMS FOR Used by REACTTAB to read VOC.PRM. LDRCT FOR Used by REACTTAB to read .RCT files. FMTOUT FOR Used by REACTTAB to format tabulated reactivity values. RENORCMP FOR Main module for RENORCMP. PARSE0 FOR Various utility subroutines used by both programs. INTV FOR REALV FOR ------------------------------------------------------------------------- Although its use is not necessary, the REACTTAB program is provided to serve as the interface between the user and the reactivity data files. The operation of this program is described in Section V. Briefly, the program will ask the user for the type of reactivity data to be output, then prompt for lists of scenarios and VOCs or mixtures for which reactivity data are desired. The names of the scenarios and of the VOCs or mixtures can be given directly to the program in response to the response, or (if there are many scenarios, VOCs, or mixtures) can be listed in files which the user can specify in response to the prompts. The program can also be run in "batch" mode, with all the information it needs given on the command line. The resulting reactivity tabulations can either be output to the screen or in files named by the user. If only two scenarios or VOCs are listed, the program will also give percent differences between them. The scenarios or reactivity scales are identified by a 5-8 character code which identifies the .RCT file giving the reactivity data. The various scenarios or scales which are included in this distribution are discussed in Section III, and their development is documented in more detail in the draft EPA report (Carter, 1990a). If the user is interested in reactivities for only a subset of these scenarios or scales, the unneeded .RCT files can be deleted to save disk space. All the individual types of VOCs for which reactivity estimates are made are listed in the file VOC.PRM, and are discussed in Section IV, below. Each type VOC has a 2-8 character name which must be used when identifying it to the REACTTAB program, and in files specifying compositions of mixtures. Section IV (and the VOC.PRM file) gives a tabulation of these VOC names, with a description of the types of VOCs they represent, and their mechanism uncertainty classification. In addition to tabulating reactivities of individual types of VOCs, the REACTTAB program will also output reactivities of mixtures of many VOCs. If a VOC name given to REACTTAB is not on the list of VOC names in VOC.PRM, the program assumes it refers to a mixture of VOCs, and will attempt to read a "composition" (.CMP) file for the mixture. These files give the VOC name and the relative molar amount of each VOC in the mixture. Each VOC in the file must be identified by a recognized VOC name which is listed in VOC.PRM. For the REACTTAB program to correctly calculate its per-gram or per-carbon reactivity, the relative molar amounts of the VOCs in the .CMP file must be normalized to one carbon total. This can be done using the RENORCMP program as described in Section V. The distribution includes composition files for several mixtures of interest, including those representing base case ROG emissions and aloft VOCs in the reactivity calculations (Carter, 1990a), several representative vehicle exhaust mixtures from an analysis recently carried out by the California Air Resources Board (CARB, 1989), two representative mineral spirit mixtures (Weir et al., 1988), and several representative surrogate jet fuels employed in a previous studies we carried out for the Air Force (Carter et al. 1984, 1987). These are tabulated in Section IV. As also discussed in Section IV, the user can create other composition files using any text editor, and then running RENORCMP program to normalize the mixture. If an improper VOC name unreadable data exist in the file, RENORCMP or REACTTAB will output error messages. In sections III and IV, below, the scenarios, reactivity scales, and the VOCs and VOC mixtures represented in the current set of files are summarized and briefly discussed. The REACTTAB and RENORCMP programs, and the installation of the programs and files, are discussed in Section V. III. REACTIVITY SCALES The data files documented by this report can be used to obtain VOC reactivity tabulations for 99 different reactivity scales. These include two general reactivity scales, four multi-scenario scales, and 93 single-scenario scales. The single-scenario scales represent reactivities in two "averaged conditions" episodes and in various episodes for 12 urban areas in the United States. The latter include 21 "base case" scenarios, 28 "maximum reactivity" scenarios and 21 "maximum ozone" scenarios. Each of the 21 "base case" scenarios has two scales, one based in impacts of the VOCs on peak ozone, and one based on integrated ozone. Each reactivity scale has a 5-8 character code which is used to identify the file containing the reactivity data. For the "multi-scenario" scales, an additional code is used to identify how the scale was derived. The various types of scales and scenarios are summarized below. A. Single Scenario Scales 1. Base Case Scenarios The reactivity scales discussed here were derived using single cell "box" or EKMA models for various episodes for 12 urban areas in the United States (Carter, 1990a). These were taken from previous studies of Gery et al. (1987) and Whitten (1988) without modification beyond the minimum necessary for compatibility with the software and mechanism used in this study (see Carter, 1990a). They were in turn taken from various EKMA models developed for EPA state implementation plans. Although they represent the modelers' best estimates at the time they were developed for the conditions of these episodes, they are based on extremely limited data and probably should not be considered to accurately represent any real episode. However, they represent a wide variety of conditions, and differences in reactivities among them are useful for assessing the extent to which reactivities vary from scenario to scenario. Note, however, that all scenarios are based on using the same mixture mixture to represent base case ROG emissions (i.e., the mixture in ALLCITY5.CMP), so they are not useful for assessing effects on reactivity of variabilities or uncertainties in this aspect of scenario conditions. In addition, they all represent single day episodes, so are not useful for investigating multi-day effects. 2. Maximum Reactivity and Maximum Ozone Scenarios As discussed elsewhere (Carter, 1990a), the maximum reactivity (MAXRCT) and the maximum ozone (MAXO3) general reactivity scales were developed by averaging reactivities calculated for the maximum reactivity and the maximum ozone scenarios, respectively. The maximum reactivity scenarios were derived from the base case scenarios by adjusting total NOx inputs to yield the maximum incremental reactivity of the base case ROG mixture. The maximum ozone scenarios were derived from the base case scenarios by adjusting total NOx inputs to yield the maximum peak ozone concentrations. The ROG/NOx ratios are always higher in the maximum ozone scenarios than in the maximum reactivity scenarios. 3. "Averaged Conditions" Scenarios For the purpose of conducting sensitivity studies, one set of scenarios was derived to represent an approximate "average" of the conditions of the other scenarios. Reactivity scales are given for maximum reactivity and maximum ozone NOx inputs for this set of "averaged" conditions. The maximum reactivity "averaged conditions" scenario is designated "AVGCONMR", and the maximum ozone scenario is designated "AVGCONMO". The reactivities for these two scenarios correspond quite closely to those in the MAXRCT and MAXO3 scenarios, discussed below. 4. Tabulation and Nomenclature of Scenarios Table 2 lists the various episodes used to derive the base case, maximum reactivity, and maximum ozone scenarios. [For a more complete summary of the conditions of these episodes, see Table 1 of Carter (1990a).] The code name for the scenarios is given by the ID code shown on the left-hand column of Table 2, followed by the suffix "BS" for base case scenarios, "MR" for maximum reactivity scenarios, or "MO" for maximum ozone scenarios. [For example, W-LA1BS is the code for the base case scenario for the W-LA1 episode from Whitten (1988), and G-PX4MO refers to the maximum ozone scenario derived from the G-PX4 episode from Gery et al. (1887). The reactivity data for these two scenarios are in the files W-LA1BS.RCT and G-PX4MO.RCT, respectively.] Note that complete reactivity calculations for base case, maximum reactivity, or maximum ozone conditions were carried out for only a subset of all the scenarios listed in Table 1 of Carter (1990a). Scenarios for which a base case, maximum reactivity, or maximum ozone .RCT file exists are those where an ROG/NOx ratio is given for that episodes. Episodes for which no .RCT file exists are not listed in Table 2. Table 2. Individual Scenarios for which Reactivity Scales were Derived. THIS IS FILE = REACT\SCENARIO.DOC The files BASESCEN.LIS, MAXRSCEN.LIS, and MAXOSCEN.LIS contain the names of all the base case, maximum reactivity, and maximum ozone scenarios, respectively, for which .RCT files exist. (They do not include the "averaged conditions" scenarios. The user can add these to the MAXRSCEN.LIS and MAXOSCEN.LIS if desired.) These can be used as inputs to the REACTTAB program if it is desired to have a tabulation of the reactivities of a given VOC or mixture for all these scenarios. B. General and Multi-Scenario Scales 1. The MAXRCT and MAXO3 Scales The MAXRCT and MAXO3 scales are "general" reactivity scales derived by from incremental reactivities calculated for the individual maximum reactivity or maximum ozone scenarios, respectively. The reactivity data for these scales are in the files MAXRCT.RCT and MAXO3.RCT. The reactivity data from these scales for the VOCs with documented mechanisms are also given in Table 3 in Carter (1990a). Note that the author believes that the MAXRCT scale is most appropriate for use in general reactivity assessment applications, and is the scale proposed for use in the California "clean fuels/low emissions vehicle" regulations (CARB, 1990). On the other hand, others believe that the MAXO3 scale is more appropriate, since it represents maximum ozone conditions, and since it represents NOx conditions which are believed to be closer to the average for many urban areas. In any case, it can be argued that together they represent respectively the highest and lowest conditions of NOx availability which are appropriate for developing reactivity scales (Carter, 1990a). Therefore, if only a limited number of scales are used, these would give the best indication of the variability of reactivities with conditions. 2. Base Case Relative Reactivity Scales The base case relative reactivity scales are multi-scenario scales derived from reactivities calculated from the base case scenarios. Since reactivities vary widely in magnitude from scenario to scenario, they are given not as absolute incremental reactivities, but as "relative reactivities". These are defined as ratios of incremental reactivities of the VOCs to incremental reactivities of the base case ROG mixture (i.e., reactivities relative to that of the ALLCITY5 mixture). These are derived using two different methods, designated the "average ratio" (AR) or the "least squares fit" (LF) methods. The average ratio relative reactivities are simply the average of the relative reactivities in the individual base case scenarios. The least squares fit relative reactivities minimize the total error in ozone predictions throughout the base case scenarios, and can be derived by plotting the VOC reactivities vs. the base ROG reactivities, and using the slope of the least squares line, forced through zero, as the relative reactivity (Carter, 1990a). Note that the least squares fit method weighs more heavily reactivities in scenarios where adding VOCs have greater effects on ozone, and thus tend to correspond closer to the maximum reactivity scale. The data for the base case relative reactivity scale derived from the incremental reactivities in the individual scenarios are in the file BASE.RCT. This file gives both the average ratio and the least squares fit scales. To specify the average ratio scale, the code "BASE/AR" should be used to identify the scenario or scale, while to specify the least squares fit scale, the code "BASE/LF" should be used. C. Integrated Ozone Reactivity Scales The reactivity scales discussed above are all derived based on effects of the VOCs on peak ozone concentrations. Reactivities for the base case scenarios were also calculated based on effects of the VOCs on ozone concentrations integrated over time (Carter, 1990a). There are referred to as "integrated ozone reactivities". For each base case scenario for which there is a BS.RCT file, there is also a BSI.RCT file containing containing the integrated ozone reactivities. The suffix "I" is appended to the base case scenario designation to indicate that integrated ozone reactivities are used. For example, to obtain a tabulation of integrated ozone reactivities for the W-LA1BS scenario, the code "W-LA1BSI" should be used as the scenario designation. The integrated ozone reactivity data for this scenario is in the file W-LA1BSI.RCT. Likewise, the multi-scenario base case relative integrated ozone reactivity scales are referenced by the codes "BASEI/AR" or "BASEI/LF", depending on whether the average ratio or least squares fit integrated ozone fit scale is desired. The data for these scales are in the file BASEI.RCT. D. Kinetic and Mechanistic Reactivities In general, the user will be most interested in using these scales to obtain incremental reactivities of VOCs or mixtures, since incremental reactivities estimate the actual effects of the VOCs on ozone. However, for those who are also interested in a more fundamental chemical analysis of the factors affecting reactivity, or those who are interested in estimates of how much the VOCs are reacting, the files can also be used to obtain estimates of kinetic and mechanistic reactivities. Kinetic reactivities are defined as the fraction of the VOC which reacts in the scenario, and mechanistic reactivities are defined as the amount of ozone formed caused by adding the VOC to the emissions, divided by the amount of VOC which reacts (Carter, 1990a). The incremental reactivity (the amount of ozone formed divided by the amount of VOC emitted) is then the product of the kinetic and mechanistic reactivity. These two components of reactivity are more straightforward to estimate than incremental reactivities because they have a more direct fundamental chemical interpretation. The reactivity data (.RCT) files for all the individual scenarios and for the MAXRCT and MAXO3 scales include the kinetic and mechanistic reactivity information for the VOCs. The REACTTAB program can output these reactivities for these scales if the appropriate option is input when the program is run (see Section V). However, the multi-scenario relative reactivity data files BASE.RCT and BASEI.RCT do not contain information concerning kinetic and mechanistic reactivities, and therefore these options cannot be used for these scales (i.e., for the BASE/AR, BASE/LF, BASEI/AR, BASEI/LF scenario designations). This is because these scales give ratios of reactivities (i.e., relative reactivities) only, which cannot be broken down into kinetic and mechanistic reactivities. REACTTAB can also be used to obtain kinetic reactivities of mixtures. These may be of interest because they indicate the total amount of VOCs which react in the mixture. On the other hand, REACTTAB will not output mechanistic reactivities of mixtures. These require additional computations and are not particularly meaningful chemically. E. Format of Reactivity Data (.RCT) Files The .RCT data files included in this distribution are ASCII files containing reactivity data for the various individual scenarios or scales. There are three formats for these .RCT files, one used for all the single-scenario scales, one used for the MAXRCT and MAXO3 scales, and one used for the multi-scenario relative reactivity scales (i.e., for BASE.RCT and BASEI.RCT). From the point of view of the REACTTAB program, the single-scenario and the MAXRCT and MAXO3 formats are the same, since the program does not use the data fields where they differ. The main difference between these two formats (for REACTTAB) is that the single-scenario and MAXRCT and MAXO3 files contain kinetic and mechanistic reactivity information, while the multi-scenario files contain no such information, but contain two reactivity scales, one derived using the average ratio method, and one derived using the line fit method. 1. Single-Scenario .RCT Files The single scenario .RCT files contain a number of "header" or "comment" records which are ignored by the REACTTAB program. (Note that the ScenID notation given at the start of the file is that used internally at SAPRC, and will be slightly be different from that used in this distribution.) Included in these "comments" are mechanistic reactivities of the various "pure mechanism" species which can be used to estimate mechanistic reactivities of VOCs which react only with OH radicals, as discussed in Appendix B of Carter (1990a). The code designations are consistent with the notation used in Table 2 of Carter (1990b). The "Typ" of "R" means it is a radical species, and "P" means it is a product species. They are given for OH radical rate constants of 3 x 10^3, 3 x 10^4, and 1 x 10^5 ppm^-1 min^-1, respectively.] The record containing a single "," indicates the end of the comments and the beginning of the data read by REACTTAB. The data following the comments consist of one record for each VOC for which reactivity data are given. These consist of: the VOC name as used in VOC.PRM (labeled "DMSname" in the files -- where "DMS" stands for "detailed model species"); the kinetic reactivity (KR); the mechanistic reactivity (MR) in units of moles ozone per mole C VOC reacted; the incremental reactivity in units of moles ozone per mole C VOC (IR/C) and the incremental reactivity in units of grams ozone per gram VOC (IR/G). In addition, for VOCs which react only with OH radicals (except for CO), the files include the data fields labeled "MRparm" "Error", which are ignored by REACTTAB. The "MRparm" field contains the mechanistic reactivity estimated from mechanistic reactivities of the "pure mechanism" species in its mechanism (see Appendix B of Carter, 1990a). If the mechanistic reactivity of the VOC was not calculated explicitly, it is the same as the mechanistic reactivity given in the "MR" field, and the "Error" field is blank. If the mechanistic reactivity was calculated explicitly, the explicitly-calculated mechanistic reactivity is in the "MR" field, and the difference between this and the value estimated from "pure mechanism" species is given in the "Error" field. (The values in the "Error" field should be small compared to those in the "MR" field, except for VOCs with relatively low mechanistic reactivities. Except for integrated ozone (BSI) .RCT files where mechanistic reactivities have higher magnitudes, their magnitude should be no greater than approximately 0.05. If they are not, the reactivities of VOCs with data in the "MRparm" field and not in the "Error" field will be unreliable.) 2. MAXRCT.RCT and MAXO3.RCT Files These files are similar to the single-scenario .RCT files in that they contain comments which are ignored by the program in records up to one with a single ",", and in that the first five fields of the reactivity data records contain the VOC name, the kinetic reactivity, the mechanistic reactivity in moles ozone per mole C, the incremental reactivity in moles ozone per mole C, and the incremental reactivity in grams ozone per gram C, respectively. These are the data which are used by the REACTTAB program. The comments in these files differ in that instead of giving mechanistic reactivities of parameters, they give names of .RCT files used to calculate the average kinetic and mechanistic reactivities which are tabulated. These files are derived by averaging kinetic and mechanistic reactivities of the maximum reactivity or maximum ozone scenarios for the various urban areas, and are not included in this distribution. (They can be obtained from the author if desired.) In addition, instead of containing the "MRparm" and "Error" fields, the files contain data fields giving the (one sigma) standard deviations of the averages of the kinetic and mechanistic reactivities. These are ignored by REACTTAB. 3. Multi-Scenario .RCT files (BASE.RCT and BASEI.RCT) As indicated above, these files have a substantially different data format than used in the single-scenario and the MAXRCT or MAXO3 .RCT files. But like the other types of .RCT files, the initial records consist of comments which are ignored by REACTTAB. The comments give the .RCT file names of the scenarios used to derive the multi-scenario scales. Note that the scenario names are those used internally at SAPRC, and differ somewhat from those used in this distribution. The comments also give the name of VOC or the composition (.CMP) file for the mixture which the tabulated reactivities are relative to. In the cases of the files in this distribution, this will always be "ALLCITY5", the mixture used to represent base case ROG emissions. The comment records are terminated by a record containing a single "." character. (The use of a "." record instead of a "," tell the REACTTAB program what format to expect for the reactivity data to follow.) Each record following the "." in the file contains reactivity data for a single VOC. The first field consists of the VOC name; the second field contains the relative reactivity derived using the average ratio method; the third field contains the (one sigma) standard deviation of the average ratio and is ignored by REACTTAB; the fourth field contains the relative reactivity derived using the least squares fit method; and the fifth field contains the (one sigma) standard deviation of the least squares fit (the standard deviation of the slope of the line, forced through zero, fit to plots of the VOC reactivity vs. the base case ROG reactivity), which is also ignored by the program. The standard deviations are given as percentages (and indicated by a "%") unless they are high relative to the magnitude of the average or slope. Although the .RCT files in the this distribution do not contain them, the REACTTAB program will allow comment records to be included among the VOC reactivity data records. The program ignores any records where the first character is a "!", no matter where it occurs in the .RCT file. IV. VOCs AND MIXTURES A. Individual VOCs and the VOC.PRM File 1. Classes of VOCs on Reactivity Data Base The classes of VOCs for which data files in this distribution give reactivity estimates are listed in the VOC.PRM file, and Table 3 gives a subset of the information contained in this file. That table gives the name which must be used to identify the VOC when running this software, the description of the VOC, its molecular weight and number of carbons, the general method used to represent it in the mechanism or to estimate its reactivity, and its mechanism documentation or uncertainty code, which is discussed below. The VOCs listed in Table 3 can be considered to fall into two categories: those whose mechanisms are documented either by Carter (1990b) or in Appendix A of Carter, 1990a), and those whose mechanisms are preliminary and not documented. The former are indicated by having been assigned a non-zero mechanism uncertainty code, while the latter have documentation codes of "0". (Some of these undocumented VOCs have not been assigned their final name, and temporary names derived from their SAROAD classification number are used.) It is important to recognize that the reactivities of the VOCs with codes of "0" must be considered to be preliminary and subject to change. Updates of the mechanisms and their documentation are currently underway as part of ongoing programs at our laboratory funded by the U.S. EPA and the California ARB. Table 3. Individual VOCs and VOC Classes whose Reactivities are Estimated. Mechanism representation methods and uncertainty and documentation codes are also shown. THIS INFORMATION IS IN FILE = MECH\DMS.PRM As indicated by a footnote to Table 3, the uncertainty codes for the VOCs with documented mechanisms range from 1 through 9, with the higher numbers indicating the more uncertain mechanisms. The uncertainty classification is based on the availability of data to test the mechanism, combined with the author's subjective assessment of the approximate level of uncertainty in our knowledge of their atmospheric photooxidation mechanisms as they relate to ozone formation (Carter, 1990a). The "undocumented" VOCs with a code of zero may not necessarily have the most uncertain mechanisms (though almost all would probably be given a 7-9 classification), but since their mechanisms are preliminary and undocumented, the user should use their reactivity estimates with caution. In particular, it is recommended that reactivities for these undocumented VOCs not be used except for analysis of complex mixtures where they make only relatively minor contributions. As discussed in Section V, an option can be specified in the MODELING.PRM file giving the highest uncertainty code which can accepted by the program (with a "0" being treated as a high number). The default contained in the distributed file is for the program to ignore these undocumented VOCs. If the user wishes to use them, he must edit the MODELING.PRM file, and change it as indicated in the comments therein. 2. Format of the VOC.PRM File The VOC.PRM file contains the information given in Table 3 and also the OH radical rate constant assigned to it in the current mechanism. The initial records of the file are comments which describe the data contained in the files (including the meanings of the codes), and which are ignored by the programs. The comments are terminated by a record containing a single "." character. The following records are either comments which are ignored by the program (indicated by a "!" in column 1), labels for classes of VOCs to follow (indicated by a space or tab character in column 1), or records giving data for a VOC. The latter contain the following information: the VOC name; its number of carbons; its molecular weight, its OH radical rate constant in units of ppm-1 min-1; the mechanism representation code; 3); the uncertainty and documentation code; and the VOC description. B. Mixtures and Composition Data (.CMP) Files One of the more useful characteristics of incremental reactivities is that incremental reactivities of mixtures can obtained by linear summation of incremental reactivities of its components. The REACTTAB program can take advantage of this to estimate incremental reactivities of mixtures of the VOCs listed in Table 3. To use this feature, one must create a composition (.CMP) file for each mixture. The mixture must be given a name which is different from the name of any VOC listed in VOC.PRM (or Table 3), and that name is used to identify the .CMP file to the REACTTAB program. (Any name which is can be used for a legal DOS file is acceptable.) The user can input mixture names to the REACTTAB program in any place where a VOC name is a legal input -- the program "knows" it is a mixture because it is not on the list of VOC names read from VOC.PRM, and because it can find a .CMP file for it. If no .CMP file exists, the program outputs an "Unrecognized VOC or CMP" error message. Mixture composition (.CMP) files are ASCII files which contain at least one record for each VOC, and optional "comments" which are ignored by the program. Any record with a "!" character in the first column is treated as a comment and is ignored. The records for VOCs consist of the VOC name, followed by one or more spaces or "tab" characters, and then its relative molar amount. If the same VOC appears more than once in the file, the amounts given in the records are added together. Normally, the molar concentrations of the VOCs in the composition file should be "normalized" so that the total number of moles carbon of the VOCs in the file sum to unity. This can be easily done by running the "RENORCMP" program. This is discussed in Section V. C. Examples of Composition Files The computer files in this distribution include 30 composition (.CMP) files specifying various interests of potential interest. These include the composition files used in the derivation of these reactivity scales, and in the associated calculations investigating the effects on reactivity on changing the composition of the base case ROG emissions (Carter 1990a). The most important of these is the ALLCITY5 mixture, which was used to represent the base case ROG emissions in the derivation of all the reactivity scales in this distribution, and which was used as the standard when defining the "relative reactivities" in the multi-scenario scales. Composition files representing total ROG emissions taken from California and NAPAP emissions inventories are also included in this distribution. They can be seen to be quite different from the ALLCITY5 mixture, in general having lower reactivities. In addition, composition files representing examples various types of individual VOC emissions sources, such as vehicle exhausts, and jet fuels and exhausts, and mineral spirits mixtures are also included. These were used by the author in various previous studies. Table 4 lists the composition files included in the distribution, and references previous studies where they were used. The comments in the files themselves in most cases give additional information concerning their derivation. Table 4. Names and Descriptions of Example Composition (.CMP) Files THIS IS FILE = CMPFILES.DOC V. DESCRIPTION OF SOFTWARE The software in this distribution includes two programs: REACTTAB and RENORCMP. As indicated in the previous sections, REACTTAB is used to obtain tabulations of reactivity estimates for selected VOCs and mixtures in selected reactivity scales, and RENORCMP is used to prepare composition files for use by REACTTAB. Both these programs run on IBM-compatible computers, are written in FORTRAN-77, and have been compiled and debugged using the Lahey F77L compiler. The source code for these programs are included in this distribution. Additional programs which may be useful for evaluating VOC reactivities are in preparation, and will be made available at a later date. Sections A and B, below, give a "users manual" for running REACTTAB and RENORCMP, respectively. The installation of this software and the data files it uses is discussed in Section C. A. Running REACTTAB REACTTAB gives tabulations of reactivities for selected VOCs and/or mixtures from selected reactivity scales. The user can specify whether the reactivities are output in "per carbon" or "per mass" units, and whether incremental, kinetic, or mechanistic reactivities are output. The output can go the user's screen or to a file. The program can be run in either interactive or batch mode. In interactive mode all the information it needs is obtained from user prompts, while in batch mode, the information it needs is contained in the command line. This is discussed below. In the following discussion, it is assumed that the program is either in the user's current directory or is accessible through the "PATH", and that the .CMP and .RCT files are either in the current directory or in directories specified in the MODELING.PRM file. The specification of directories and other program control options used by REACTTAB in MODELING.PRM is discussed in A.3, below. 1. Interactive Mode To run REACTTAB interactively, simply give the command "REACTTAB" without any options at the DOS prompt. The program will respond with the a request to give an option code, as follows: Give option code (updated to give options as of 6/2000): M = Output per-gram, O3 formed, reactivities; MOL = Output per-mole, O3 formed, reactivities; MR = Output mechanistic reactivities; K = Output kinetic reactivities; PC = Output per-carbon, O3 max., reactivities; PM = Output per-gram, O3 max., reactivities. +code = Output in CSV format. *code = Output in 4 sig. figs. (default) = Output per-carbon reactivities. Code: _ The user can either enter "M" if he wants the program to output incremental reactivities in units of grams ozone per gram VOC, enter "MR" to output mechanistic reactivities (in units of moles ozone per mole carbon VOC reacted), enter "K" to output kinetic reactivities (fraction VOC reacted, or amount reacted for unnormalized mixtures), or simply enter to output incremental reactivities in units of moles ozone per mole carbon VOC. (In this discussion "" means just hitting the carriage return or "enter" key without entering any other input.) These options remain in effect throughout the run of the program; if he wishes to later output different measures of reactivities or reactivities in different units, he must exit the program and re-run it. The program will then ask for the list of reactivity scales for which reactivities are to be output. The prompt is Give scenario or reactivity scale ( when done). Or "F=filename" to read scenario list from a file. Scenario: _ At the "Scenario: " prompt, the user can either enter the name of a single scale, or characters "F=" followed (without an intervening space) by the name of a file from which to take names of reactivity scales. If a single scale is entered, it is either the name of a single-scale .RCT file -- without the ".RCT" -- or the name of a multi-scenario .RCT file with the characters "/AR" or "/LF" appended to it to indicate whether the average ratio or least-squares fit scale is used. If the .RCT file is not found, an error message will be output, and the program will repeat the prompt. If a valid single scenario or reactivity scale name is entered, the program will read the data for that scenario, and then prompt for another scenario/scale name. The user can either (a) enter the name of another single scale and the process of entering scenario names will be repeated; (b) enter the name of a file from which to read the remaining scenario names ("F=" followed by a file name), or (c) enter a without any input to terminate scenario input. The program also terminates scenario input after scenario names are read from a file. Once scenario input is completed, no additional scenarios or reactivity scales can be read in without exiting and re-running the program. Files containing names of reactivity scenarios or scales consist simply of records giving the names of the scales, one scale per record. No other data should be on the records giving the scenario names. Any records with a blank or a "!" in column one are ignored, and so can be used as comments. The files "BASESCEN.LIS", "MAXRSCEN.LIS", and "MAXOSCEN.LIS", which contain lists of all the "base case", "maximum reactivity" and "maximum ozone" single-episode scales, respectively, can be used if reactivities for these groups of scales are desired. Examples of legal inputs to the "Scenario: " prompt (assuming all the .RCT files have been installed correctly) are: MAXRCT ... to use the "maximum reactivity" general scale; G-LA1BS ... to give reactivities for the base case "Los Angeles" episode. BASEI/AR ... to give base case relative reactivities in the multi-scenario scale derived using the "average ratio" method. F=BASESCEN.LIS ... to give reactivities for all the "base case" single-scenario scales. After the first three of the above examples, the program prompts for more scenarios. After the last example, the program gives messages indicating which scales it is reading, and then terminates scenario input. After scenario input is terminated (either by giving at the "Scenario: " prompt, or having scenario lists input from a file), and also after the program has output a previous reactivity tabulation, the program gives the prompt: If relative reactivities desired, give name of compound or mixture to use as the standard. ( for absolute reactivities.): _ If it is desired to have the reactivities output be relative to the reactivity of some VOC or mixture, the name of that VOC or mixture can be given here. The output reactivities will then be reactivities of the selected VOCs divided by the that of this VOC or mixture. For example, to output relative reactivities (as used in the multi-scenario scales in the BASE.RCT or BASEI.RCT files), one should enter "ALLCITY5" at this prompt. To output absolute incremental reactivities (moles ozone per mole carbon or grams ozone per gram VOC), or kinetic or mechanistic reactivities, enter at this prompt. (Note, however, that the BASE/LF, BASE/AR, BASEI/LF, and BASEI/AR scales are always relative reactivities. To output correct values for these scales in "per mass" units, "ALLCITY5" should always be input at this point when "per mass" units are used. This is not necessary when "per carbon" units are used.) After name of the standard compound or mixture (if any) is entered, the program gives the prompt "Species/Mix: " to ask for names of VOCs or composition files whose reactivities are to be output. At this point, the user can either: (a) give the name of a single VOC listed in VOC.PRM; (b) give the name of a .CMP file defining a mixture (without the ".CMP"); (c) give the input "F=" followed (without any spaces) by the name of a file from which a list of VOCs and/or mixtures can be obtained; or (d) simply enter to terminate this input. Input of species/mixtures is also terminated after input is taken from a file. If an input name is not recognized as a valid species from VOC.PRM the program will look for a composition (.CMP) file with that name. If no such file is found, it will give an error message and ask for more input. If a is entered when no valid species or mixtures have been entered, the program will terminate. The format of files containing lists of VOC or mixture names is the same as that of files containing scenario/scale names. Each record must be a single species or composition file name (without the ".CMP"), or a comment. Comments are records with a blank or a "!" in the first column. Three such files are contained on the distribution: "ALLVOC.LIS", which has the names of all VOCs on VOC.PRM; "DOCVOC.LIS", which has the names of all "documented mechanism" VOCs which are listed in Table 3 of Carter (1990a); and "ALLCMP.LIS", which has the names of all composition files listed in Table 4 of this document. If the VOC/mixture input is terminated with at least one valid VOC or mixtures specified, the program will output the prompt "Give output file ( for terminal output): ". If the user enters , the reactivity tabulation will go to the screen, but if the user enters a valid DOS file name, the output will go to that file and not the screen. The latter is obviously more useful if a large amount of data are output, as might be the case if the scenario or VOC/mixture is obtained from files. The output will give the units of the tabulated reactivities, and if ratios of reactivities are output, this is indicated. Note that with unnormalized mixtures, (i.e., where total carbons do not sum to one) the stated units of the output will not be correct. The program will flag (with a "*") names of any mixture which it determines not to be normalized. If the "auto-normalize" option is used (see Section A.3, below) all unnormalized mixtures will be normalized before being processed, and thus this flag will not appear, and the units will be correct. However, the default is for the program not to automatically normalize mixtures, but to give warning messages when unnormalized composition files are read. This option is specified in the MODELING.PRM file, as discussed below. The format of the tabulation will depend on how many scenarios and VOCs have been entered. If six or fewer scenarios or reactivity scales are specified, the program will have one output record for each VOC or mixture, with the reactivities in the various scales given in columns. If exactly two scales are specified, the program will also output the percentage difference in reactivities between the scales. If there are more than six scenarios but fewer than six VOCs, the program will output one record for each scale, and give the reactivities for the various VOCs in columns. If exactly two VOCs are specified in this case, the program will also output the reactivity ratio for the two VOCs for each scale. If more than six scenarios are input, the program will not allow more than six VOCs and/or mixtures to be input. After producing the tabulation (either to the screen or a file), the program will return to the point where it asks whether relative reactivities are desired. The user can then repeat the process with different VOCs (but not different reactivity scales or units or types of reactivity output) if desired. If not, he can either enter a control-Z (which terminates the program at any time input is requested), or two 's in a row to exit the program. 2. Batch Mode To run REACTTAB in the batch mode, the information it needs is given on the command line. When run in this way, it will output only a single tabulation, but otherwise the operation of the program is the same. To run it in this mode, the following format is used (where lower case characters indicate user-specified options, and brackets indicate optional input): REACTTAB vocormixes scales [outputfile] [option] [stdvocormix] or REACTTAB vocormixes,scales,[outputfile],[option],[stdvocormix] The options are as follows: The option "vocormixes" gives the list of VOCs or mixes for which reactivities are desired. It can either be the name of a single VOC or mixture, or in the form of "F=filename" to read the VOC or mixture names from a file, or a list of VOC or mixture names separated by spaces or commas and surrounded by single quote "'" characters. The option "scales" gives the list of scenarios or reactivity scales to sue. As with "vocormixes" it can be the name of a single scenario or scale, or in the form of "F=filename" to read the scenario/scale names from a file, or a list of scenario/scale names surrounded by single quote characters. If the tabulation is to go to a file, the third option gives the file name. If output to a terminal is desired, this is left blank. If the following options need to be specified on the command line, two commas together can be used to indicate that this option is blank. The option "option" can be either a "M" to indicate output of per-mass incremental reactivities, "MR" to output mechanistic reactivities, or a "K" to output kinetic reactivities. If this is left blank (as can be done by not including it or using two commas together), per-carbon incremental reactivities are output. The last option is the name of the standard VOC or mixture if ratios of reactivities are to be output. Several examples of running REACTTAB in batch mode are as follows: REACTTAB 'N-C4 N-C8' 'BASE/LF BASE/AR BASEI/LF BASEI/AR' REACTTAB MS75 F=BASESCEN.LIS COMPARE.MS,,ODORLSMS REACTTAB F=DOCVOC.LIS MAXRCT DOCVOC.MXR M The first example outputs to the user's screen the relative incremental reactivities of n-butane and n-octane in the four base case multi-scenario scales. The second outputs the ratios of reactivities (on a per-carbon basis) of "mineral spirits-75" to "odorless mineral spirits" for all the base case scenarios to the file COMPARE.MS. The third outputs the per-gram MaxRct incremental reactivities of all the VOCs listed in Table 3 of Carter (1990a) to the file DOCVOC.MXR. Note that the data in that file should be the same as given in column 10 in Table 3 of Carter (1990a). If the program cannot properly decode the command line, or if some error occurs, it will output a message indicating the command line format it expects. Other details concerning the operation and options of this program are the same as when the program is run interactively, and are discussed in the previous and the following sections. 3. REACTTAB Options in MODELING.PRM The REACTTAB program reads the file "MODELING.PRM" to obtain special options which control its operation. All these options have defaults, and if the defaults are acceptable, no input in MODELING.PRM is required. (Indeed, the program will run if no MODELING.PRM file exists, but in that case it will create an empty one.) The program will ignore all input in the file which it does not recognize, including options meant for other programs, and comments (which are usually indicated by blanks or "!" characters in the first column.) Note that the author uses this file to give options to many programs (it serves in effect like a local "environment" for each directory), so in general it may have options a particular program may not use.) The following options in MODELING.PRM are recognized by REACTTAB. Note that in the listing of options below, upper case characters and the "=" character are required syntax, lower case characters are variable options, and imbedded blanks are not allowed. CMP=dirname This gives the directory where the program can find .CMP files. The name of the file is appended to the string given for "dirname". If not given, .CMP files must be in the current directory. This option is also used by the RENORCMP program. RCT=dirname This gives the directory where the program can find .RCT files. The name of the file is appended to the string given for "dirname". If not given, .RCT files must be in the current directory. VOC=filename This is the full path name of the file used to read VOC names, molecular weights, etc. The default is VOC.PRM in the current directory. This option is also used by the RENORCMP program. REACTTAB=option This tells REACTTAB how to deal with unnormalized mixtures. Legal options are "NORM", to automatically normalize unnormalized mixtures, or "NONORM", to allow calculation of reactivities of unnormalized mixtures. The default is the same as "NONORM" except that a warning message is given when unnormalized mixtures are input. (With "NONORM" this message is suppressed, and with "NORM" it is irrelevant.) MAXUNC=number If this is given, VOCs with documentation or uncertainty codes greater than the input number (which must be 1-9), or with codes of zero, are ignored, and REACTTAB acts as if they are not in VOC.PRM. The default is use all VOCs regardless of its code number. The distributed files contain recommended values for these options. The recommended directories are discussed in the section describing the installation of the software. It is recommended that the default option be used for "REACTTAB=", and that "MAXUNC=8" be used. However, the latter option will not permit processing of the RADM, SOUCOA, or AFSYNEXH mixtures. B. Running RENORCMP The RENORCMP program can be used to prepare composition files for use by the REACTTAB program. Its primary utility is to "normalize" composition files, i.e., to multiply all concentrations in the file by a factor so that the total number of carbons in the files is unity. It can optionally be used to normalize to give amounts other than one carbon, to normalize to given amounts of mass, or simply to summarize how much mass or carbon is in the file. In addition, it can be used to convert composition files which are in a form not useable by the REACTTAB program into a form which it can use. These features are discussed below. The .CMP files read by the RENORCMP program (as those read by REACTTAB) must be in the directory indicated by the data in MODELING.PRM (see previous section). If no MODELING.PRM file exists, or if it does not contain information about .CMP files, the input file is assumed to be in the current directory. The output file is placed in the same directory as the input file. The RENORCMP program obtains the information it needs to do the normalization (i.e., numbers of carbons and molecular weights) from the file VOC.PRM. If that file is not in the current directory, the "VOC=" input in MODELING.PRM must give the full pathname of a valid VOC.PRM file. RENORCMP will not process any file which contains a VOC which is not listed in VOC.PRM. (Note that, unlike REACTTAB, RENORCMP will recognize VOCs with any documentation/uncertainty code, and ignores the "MAXUNC=" parameter in MODELING.PRM. However, it will abort if it finds any VOC in VOC.PRM with a molecular weight of zero.) To use RENORCMP to normalize composition files to one carbon, give the command: RENORCMP cmpfile or RENORCMP inputcmpfile outputcmpfile If one parameter is given, it is the file name of the .CMP file (without the ".CMP") to be normalized. The old .CMP file is not deleted, but its name is changed to .BAK. (If an old .BAK file exists, it is deleted.) If two parameters are given, the first is the input .CMP file name which is read but not modified, and the second is the output .CMP file which is created. If the two-parameter form is used, the name of the output file must be different from the input file. In either case, the output file has the same relative amounts of the VOCs as the input file, but they are normalized to one carbon total. Any comments found in the input file are copied to the beginning of the output file, and the program also appends comments indicating that the file has been normalized. The RENORCMP program can also be used to "normalize" .CMP files to some total number of carbons other than one or to a specified mass. This is done by giving a nonzero number as the third parameter to the RENORCMP program. To normalize to a given number of carbons, enter the command RENORCMP cmpfile,,carbons or RENORCMP inputcmpfile outputcmpfile carbons where "carbons" is any positive number which is the total number of carbons for the VOCs in the file. This might be useful when comparing reactivities in terms of ozone per some other unit of emissions, such as, for example, computing ozone per mile for vehicle emissions (e.g., see Lowi and Carter, 1990). In the latter case, the "fac" values would be the amount of VOC (in units of mole carbon) per mile. To normalize to a given mass, the code "M" is added as a fifth parameter, e.g., RENORCMP cmpfile,,mass M or RENORCMP inputcmpfile outputcmpfile mass M where "mass" is the number of grams. The number given for each VOC in the output file will be adjusted so the that if each number is assumed to be a number of moles, the total mass of VOCs in the file will be "mass" grams. The program can also be used simply to summarize the number of carbons and the amount of mass in the file. This can be done by giving the code "SUM" as the third option, rather than a number of carbons or mass. If an output file is specified, a summary of the mass or number of carbons will be output to the screen and appended to the end of that file. (This may be useful if a summary file with masses or numbers of carbons for a large number of composition files are desired.) If no output file is given, it will be output to the screen only. For example, the command "RENORCMP ALLCITY5,,SUM" outputs to the screen the number of carbons and mass in the ALLCITY5.CMP file. Since it is normalized, the number of carbons will be 1. Note that the "mass" of a normalized composition file is the same as its average molecular weight per carbon. Finally, the RENORCMP program can be used to convert composition files with non-standard input into a form which the REACTTAB program can use. Probably the most useful feature is to convert composition files with VOC composition data in mass amounts into files with molar amounts as required by REACTTAB. If the first non-comment record in a composition file is the string "#MASS" (with the "#" in column 1), RENORCMP will recognize the file as having mass composition data. Only one "#MASS" record is permitted per file, and no VOC input may preceded it in the file. The program will convert the VOC levels in the file into molar amounts when they are being read in, and the output file will contain molar amounts. Note that the present version of the REACTTAB program will not read composition files with "#MASS" records. Similarly, a "#C" record can be included to convert composition files with mole carbon input into molar input. As with "#MASS", the the "#C" must be the first non-comment record in the file, to indicate that the file has data in the per-carbon format. An additional non-standard input in composition files which is recognized by RENORCMP are "#FAC" records. These records must include the string "#FAC" starting in column 1 and a non-zero number separated from it by spaces, tabs, or a comma. The program will multiply any composition input following it by this number, until another #FAC record is input. These can occur anywhere in the file, and more than one is permitted. If more than one is used, each new one over-rides the previous one -- the factors are not cumulated. Therefore, a "#FAC 1.0" record can be used to cancel the effects of any previous #FAC record. This feature is useful when combining given amounts of several mixtures into one. The "#FAC" records are not copied to the file output by RENORCMP, since the relative amounts of the VOCs are modified appropriately as they are read in. There may be cases when RENORCMP will be run just to convert files with mass compositions or to remove "#FAC" records, with no normalization desired. To do this, the option "NONORM" is specified as the third parameter when running the program. The output file will then give the same absolute amounts of VOCs as the input files, but in molar units and with no "#FAC" records. For example, to convert a composition file called MASSDAT.CMP with mass composition data (indicated by a "#MASS" record as the first non-comment record) to MOLEDAT.CMP with the corresponding molar amounts, use the command "RENORCMP MASSDAT MOLEDAT NONORM". C. Installation of Software and Data Files SEE READ.ME FOR INSTALLATION INSTRUCTIONS. VI. REFERENCES CARB (1989): "Definition of A Low-Emission Motor Vehicle in Compliance with The Mandates of Health and Safety Code Section 39037.05 (Assembly Bill 234, Leonard, 1987)", Report by Mobile Sources Division, California Air Resources Board, El Monte, California. May 19. CARB (1989): "Definition of A Low-Emission Motor Vehicle in Compliance with The Mandates of Health and Safety Code Section 39037.05 (Assembly Bill 234, Leonard, 1987)", Report by Mobile Sources Division, California Air Resources Board, El Monte, California. May 19. CARB (1990): "Low-Emission Vehicles/Clean Fuels -- Technical Support Document," Mobile Source Division, Research Division, Stationary Source Division, and Technical Support Division, California Air Resources Board, Sacramento, CA. Carter, W. P. L., Winer, A. M., Atkinson, R., Dodd, M. C. and Aschmann, S. A. (1984): Atmospheric Photochemical Modeling of Turbine Engine Fuels. Phase I. Experimental studies. Volume I of II. Results and Discussion. Final Report to the U. S. Air Force, ESL-TR-84-32, September. Carter, W. P. L., A. M. Winer, R. Atkinson, S. E. Heffron, M. P. Poe, and M. A. Goodman (1987), "Atmospheric Photochemical Modeling of Turbine Engine Fuels. Phase II. Computer Model Development," Report on USAF Contract no. F08635-83-0278, Engineering and Services Laboratory, Air Force Engineering and Services Center, Tyndall Air Force Base, Florida, August. Carter, W. P. L. (1990a): "Development of Ozone Reactivity Scales for Volatile Organic Compounds," Draft report for EPA Cooperative Agreement CR-814396-01-0, Atmospheric Research and Exposure Assessment Laboratory, Research Triangle Park, NC, November. Carter, W. P. L. (1990b): "A Detailed Mechanism for the Gas-Phase Atmospheric Reactions of Organic Compounds," Atmos. Environ., 24A, 481-518 Gery, M. W., R. D. Edmond, and G. Z. Whitten (1987), "Tropospheric Ultraviolet Radiation. Assessment of Existing Data and Effects on Ozone Formation," Final Report, EPA-600/3-87-047, October. Lowi, A. L., and W. P. L. Carter (1990): "A Method for Evaluating the Atmospheric Ozone Impact of Actual Vehicle Emissions," Presented at the SAE International Congress and Exposition, Detroit, Michigan, February 26 - March 2, 1990. Middleton, P., W. R. Stockwell, and W. P. L. Carter (1990): "Aggregation and Analysis of Volatile Organic Compound Emissions for Regional Modeling," Atmos. Environ., 24A, 1107-1133. Weir, B. R., A. S. Rosenbaum, L. A. Gardner, G. Z. Whitten and W. Carter, (1988): "Architectural Coatings in the South Coast Air Basin: Survey, Reactivity, and Toxicity Evaluation", Final Report to the South Coast Management District, SYSAPP-88/137, Systems Applications, Inc. San Rafael, CA, December. Whitten, G. Z. (1988): "Evaluation of the Impact of Ethanol Gasoline Blends on Urban Ozone Formation," Systems Applications, Inc., San Rafael, CA, SYSAPP-88/029