Air Toxics Emissions Inventory Protocol for the Great Lakes States

7.0 Automated Tools Available for Developing the Regional Inventory

This section briefly describes automated tools that are available to assist individual Great Lakes States in developing their component of the regional inventory, and for compiling and maintaining this regional database of emissions data and estimates. As stated in Section 1.0, the instructions contained in this Protocol are not dependent on the use of any of these tools. A given state can follow the procedures described in the Protocol and prepare their portion of the regional inventory without using any of these tools. However, the use of the tools summarized below, particularly the Regional Air Pollutant Inventory Development System (RAPIDS), will facilitate the development of a complete, quality assured air toxics inventory. In addition, the regional database (see Section 5.0) will be compiled using RAPIDS, and is defined in Section 5.0. Non-direct RAPIDS users will be required to import their emissions data and estimates into RAPIDS and use the Great Lakes National Program Office (GLNPO) Upload and Data Converter client applications.

7.1 RAPIDS

The development of RAPIDS was a multi-year effort involving the preparation of several major versions of the system. Versions 1.0 and 2.0 are described below. Version 1.0 is referred to as the initial version of RAPIDS, while version 2.0 is considered to be the full RAPIDS implementation. The following discussion describes both of these versions of RAPIDS.

RAPIDS is an ORACLE®-based (ORACLE® Server 7) client/server system developed using ORACLE's Computer Aided System Engineering (CASE) tool. The user interface and most applications use Power Builders 3.oa. SAS® is also used for costing client applications. The software is a combined air toxic and criteria pollutant emissions inventory system that uses a new multi-media, generalized, device-based data model. The structure of the RAPIDS database is depicted in Figure 7.1. RAPIDS was developed to assist the participating states in preparing their respective portions of the regional air toxics emissions inventory in a consistent and rigorous manner. RAPIDS is designed such that the same core data structure can be used by both private industry and public agencies, facilitating the electronic transfer of information among interested parties. In the future, industry may be responsible for compiling the desired emissions data, calculating emission estimates for the target compounds, and electronically transferring this information to the participating states. However, in the near term, each participating state will have the option of having RAPIDS installed locally and using this system to prepare their portion of the regional inventory.

Since it is anticipated that several states will be direct RAPIDS users, this section also includes certain procedures that states should follow when using RAPIDS to compile their portion of the regional inventory. Several of the tools summarized below are incorporated into RAPIDS version 2.0.

7.1.1 Initial Version of RAPIDS (Version 1.0)

The initial version of the RAPIDS system (Version 1.0), depicted in Figure 7-1, included the following components:

An ORACLE® back-end database consisting of various ORACLE data tables of facility and area source emissions data and estimates located on a separate (i.e., separate from the front-end client applications) file server at each participating state;

A GLNPO Upload client application that uploads each state's set of ORACLE tables to the regional repository located at GLNPO;

A Data Import client application that facilitates the import of emissions data and estimates maintained by the states external to RAPIDS into the back-end database;

A Data Entry client application consisting of various screens that are used to enter various types of emissions data, and emission estimates derived external to RAPIDS;

A Source Emission Estimator consisting of various client applications that support interactive and batch applications, and that allow the user to compute emission estimates for facility and area sources using certain emission estimation techniques (EETs);

A QC Checker client application that performs various statistical checks on the emissions data and estimates contained in the ORACLE® back-end database;

A Report Generator consisting of various client/applications that generates summary reports of the emissions data and estimates contained in the ORACLE® back-end database; and

A Data Converter client application that converts the emissions data and estimates into the U.S. EPA's Aerometric Information System (AIRS) Facility Subsystem (AFS), and Area and Mobile Source Subsystem (AMS) transaction records.

The initial version of RAPIDS included the following emission estimation techniques:

Product of emission factors (derived primarily from Factor Information Retrieval System/Great Lakes Commission FIRE/GLC), supplemented by approved source-specific factors, and emission factors derived from a combination of AP-42 criteria pollutant emission factors and speciation profiles) and activity data;

Mass balance by either taking the difference in input and output flows as entered or the difference between calculated inputs and outputs; and

Speciation of either total organic gas (TOG) or particulate matter (PM) emission estimates (computed by the user external to RAPIDS), including the algorithms needed to adjust volatile organic compound (VOC) emissions to TOG emissions and factor the TOG/PM emissions into individual air toxic pollutant emissions.

In addition, emission estimates prepared by industry or the state can be directly entered into RAPIDS. In general, these entries will be used in lieu of any automated emission calculations.

Emission estimation techniques other than those listed above can be accommodated in RAPIDS. The RAPIDS version 1.0 documentation specifies how users could define their own emission estimation techniques for source/device/process combinations and incorporate these algorithms into the system. An example of such an algorithm is U.S. EPA's tanks model (see Section 7.3.3). To accommodate these user-defined emission estimation techniques, the user needs to develop and add the appropriate metrics required by each algorithm.

7.1.2 Full RAPIDS Implementation (Version 2.0)

The conceptual design of the full RAPIDS system (Version 2.0), depicted in a Figure 7-2, will contain the following additional features/capabilities:

An upgraded ORACLE® back-end database to include motor vehicle emissions data and estimates, and a separate treatment of fugitive emissions. The addition of a separate fugitive emissions data model provided the added capability to store and manage air toxics-related fugitive emissions data and estimates from refinery and chemical manufacturing operations at a level of detail that goes well beyond the capability currently provided by most systems. Sources covered by the Hazardous Organic NESHAPS (HON) rule collect information at the level of detail included in the fugitive emissions model.

Various Source-Specific Emission Estimators consisting of several client applications that incorporate source-specific models and algorithms, allowing the user to calculate emissions from storage tanks, landfills, surface impoundments, wastewater collection devices, and other devices.

A Motor Vehicle Source Emission Estimator client application that calculates link-specific and area-wide on-road motor vehicle emission estimates.

A Biogenics Source Emission Estimator client application that calculates biogenic emissions (accounting for the effect of the leaf canopy) for various vegetative species.

An upgraded Data Converter client application that allows users to both download emissions data and estimates into AIRS and upload emissions data and estimates contained in AIRS to RAPIDS.

A Graphics Generator client application that provides ARC/INFO® access to the ORACLE back-end database using ARC/INFO's ORACLE DATABASE INTEGRATOR. The DATABASE INTEGRATOR allows ARC/INFO software to access tabular data stored in an ORACLE database and associate it with spatial features stored in ARC/INFO. This allows states to develop various graphical applications (e.g., emission density plots, by county, for each target compound, depict major air toxics point sources relative to air toxic monitoring sites, etc.).

An Emissions Model client application that temporally resolves emissions data and estimates into hour-by-hour values, and spatially allocates this information into grid cells (uniform and non-uniform) as required by air toxic deposition and photochemical models. This application speciates VOC and nitrogen oxides (NOx) for several different chemical mechanisms (e.g., Carbon Bond IV, Statewide Air Pollution Research Center [SAPRC]) into the lumped chemical species required by various photochemical models (e.g., Urban Airshed Model). This application formats the data for input into these air quality models.

7.1.3 Structure and Contents of RAPIDS

In it's basic applications, RAPIDS can be used for data entry, to display data, and to perform simple queries. However, an understanding of the structure and contents of RAPIDS is necessary to be able to fully use its capabilities, including complex queries and import and export of data from/to other systems. Figure 7-3 is an overview of RAPIDS showing the major data items contained therein. These items include information needed to describe the sources, devices, and processes of interest and information available from other databases (e.g., emission factors from FIRE/GLC and speciation factors from SPECIATE).

The following discussion and figures present the basic entities in RAPIDS. An entity in RAPIDS is a repository for certain types of information that defines and describes a unique occurrence of that entity. For example a Device Entity in RAPIDS contains information that defines a specific device (e.g., ID number and name) and provides data that describe something about that device (e.g., date that the device was installed).

A figure shows the relation between the RAPIDS core entities: Geographic, Source, Device, Process, and Stream. The figure shows a hierarchical relation among these entities. The Geographic Entity can be a nation, state, county, city, or region. A Source Entity is a location within an area defined in a Geographic Entity; a source can be a facility source or an area source. The Device Entity is a location within a Source; a device is the location of a process. The Process Entity identifies a process where material is created, used, or otherwise changed. The Stream Entity identifies a specific flow of material into or out of a process.

Another figure shows the various codes that are used in RAPIDS to describe one of the Core Entities described above; each code is defined by an entity. The Geographic Type Entity, Source Code Entity, Device Code Entity, Process Code Entity, Stream Type Entity, Material Code Entity, SIC Code Entity, SCC/AMS Code Entity and Tier Code Entity each list the valid values of those codes. The SCC/AMS Relation Entity identifies logical relations between SCC codes and AMS codes.

Another figure shows the Activity Entity, its relation to the Core Entities described above, and its various Code and Type Entities. The Activity Entity is the entity where almost all data on the Core Entities is stored. Activity is any information that describes something about a given type of geographic area, source, device, process, stream, or material. Examples include a temperature, the amount of material that flows into or out of a process, the size or condition of a device, the population of a geographic area, and an ID number assigned to a facility. In addition to the item being described, other information includes the time period over which the activity occurred, the appropriate units (defined in a Unit Code Entity), and any material involved. Activity data uses Metrics (defined by the Metric Code Entity) to identify information included in an Activity Entity; examples of Metrics include HEIGHT, MASS FLOW, USAGE, SEAL TYPE. The Value Type Entity defines information characterizing the Metric, such as MAX, MIN, AVG, etc. The Method Type Entity defines the methods that are used to describe the method of emission calculation. The Reference Type and Data Code entities provide other information for handling special data situations.

Another figure shows the Metric Map Entity and its Relation to the Metric Entity and the Core Entities. The Metric Map Entity identifies the Metrics that apply to a specific Geographic, Source, Device, Process or Stream Entity. For example, the Metric HEIGHT applies to a Device Entity with a device code of STACK.

The next figure shows the Group entities, Group Type and Group Member, that provide the capability to group Geographic, Source, Device Process and Stream Entities. The Group Type Entity defines the valid type of groups; the Group Member Entity links each member of a group to its group entity.

Figure 7-9 shows the Location Coordinate Group Entity, which identifies a group of coordinates and information on their type and accuracy (as defined in the Accuracy Type and Coordinate Calculation Type Entities), used to geographically locate a Geographic Entity, a Source Entity and a Device Entity. The Location Coordinate Entity which contains the actual coordinates.

The next figure shows the four entities used to identify the operating schedule of a process during a typical day of the week and typical hour of the day using a relatively simple coding scheme. The Process-Specific Schedule Entity contains information on more detailed operating schedule information, either typical or for a specific time period.

Another figure shows the Legal Entities, Legal and Legal Relationship. The Legal Entity identifies each individual or organization of interest. The Legal Relationship Entity relates each Legal Entity to a Source Entity or a Device Entity.

Another figure shows the general logic and tables used to drive the emission calculation client application. The steps involved in doing an emission calculation and the major reference tables used are illustrated. These steps are as follows:

Determine information on a source/device/process for which an emission calculation can be performed (this can be a device that creates emissions, controls emissions or releases emissions to the atmosphere).

Search the Calculation Protocol reference table for the highest priority calculation method for the selected source/device/process and emittant. If a higher priority method was unsuccessful, use the next highest method.

Using the Method/Metrics reference table, find the various data items needed to perform the calculation (e.g., emission factors, activity data, speciation factors, conversion factors) for the method identified in the Protocol reference table.

Perform the calculation using the data found. If the calculation was successful, write the results bach to the database as an activity record. If the calculation was unsuccessful (i.e., emissions calculate to zero, the method specified in the Protocol reference table could not be found, or one or more of the required data items could not be found), try the next method (if any).

7.2 Emission Factor Tools

7.2.1 FIRE and FIRE/GLC

A part of the U.S. EPA's air strategy is to assist the federal, state, and local agencies in assessing criteria and air toxic pollutants from various sources. Prior to the development of the FIRE, the Crosswalk/Air Toxics Emission Factor Database (XATEF), the VOC/PM Chemical Speciation Database Management System (SPECIATE), and the AIRS AFS Emission Factors (AFSEF), they all contained emission factor information. Other sources of emission factors included the Compilation of Air Pollutant Emission Factors (AP-42), and Locating and Estimating (L & E) documents. These tools are the basic sources of emission factors used to prepare air toxic inventories, State Implementation Plan (SIP) inventories, economic analyses, review of Prevention of Significant Deterioration (PSD) applications, New Source Review Permit applications, and other federal, state, and local agency assessments of air pollution sources.

A single database system containing quality-rated emission factors for criteria and air toxic pollutants was needed to consolidate emissions data from the various databases, and to meet the requirements of the 1990 Clean Air Act Amendments (CAAA). U.S. EPA developed FIRE, with a personal computer-based system written in C++, to meet this need. FIRE contains both a repository system, which is a collection of all the emission factors U.S. EPA has identified, and a distribution system, which represents U.S. EPA's recommended factors for specific pollutant/source combinations.

Emission factor data for inclusion in FIRE were obtained from a computerized literature search, other in-house projects, existing emission factor databases, recently developed L&E documents, the revised sections (Supplements E and F) of AP-42, and the SPECIATE database. The emission factors were extracted from the various sources and entered into one of seven repository modules described below:

XATEF - Made up of data found in XATEF database, which preceded FIRE. XATEF is a collection of air toxic emission factor data that are linked to potential emission sources;

AP-42 - Made up of toxic and criteria data extracted from the most recently released version of AP-42, including Supplements E and F;

AFSEF - Made up of criteria data from the AFSEF database;

CRB93 - Made up of toxic data extracted from source test reports from the California Air Resources Board (CARB) pooled source testing program ("Hot Spots");

L&E - made up of toxic data extracted from L&E documents developed or revised in 1993;

LITRE - Made up of toxic data extracted from references identified from a computerized literature search; and

CRBII - Made up CARB "Hot Spots" toxic and criteria data extracted from non-pooled source test reports obtained for another Radian work effort for U.S. EPA's Air and Energy Engineering Research Laboratory (AEERL).

Quality ratings were assigned, as necessary, based on the rating system found in U.S. EPA procedures (U.S. EPA, 1992). The FIRE repository system has two quality indicators:

A rating for the emission factor quality (factor quality); and

A rating for the quality of the source test from which the emission factor was developed (data quality).

The data quality rating is used only in U.S. EPA's FIRE repository system and indicates the quality of the source test or data in the reference. The factor quality rating is used in both the repository and distribution systems and indicates the quality of the emission factor derived from the source test or other reference data.

The rating for the data quality has a range from A to D. Factor quality has a range from A to E and U. A "U" rating for unratable factor quality was specified by U.S. EPA to be used for factors based on engineering judgment, source tests with certain deficiencies, lack of supporting documentation, and other reasons. A "U" does not necessarily imply poor quality factors, but may only mean that those data have not been reviewed to determine their actual quality rating.

Once the repository system was populated, it was necessary to select the best rated emission factors for the selected pollutants available in the repository system to be included in the distribution system. To determine factors for the distribution system, all of the emission factors in the repository system for the selected pollutants were first grouped by pollutant, SCC, control device, and process. Sources with an SCC less than six digits were not considered for the distribution system. Other data were not considered because of inconsistent units, age of the data, and reliability of the data. Quality ratings and variability factors were considered. Identical SCCs/pollutants were then evaluated to determine the best emission factor. In many cases, only a single emission factor existed in the repository system, and that emission factor was selected for the distribution system. In some cases, emission factors were averaged or compiled (using weighted factors) to produce the selected emission factor. If an AP-42 factor was available, that factor took precedence over other factors, followed by an L&E factor.

The description of the source is coded in the SCC, so the FIRE distribution system is designed to present one emission factor for each pollutant for each SCC with the same process and control status. However, there may be more than one emission factor for each SCC because the process may differ slightly, as indicated in the details of test methods or in the process or source fields, or the source may be controlled or uncontrolled, or controlled with a different device. Therefore, it may be necessary to look at more than just the emission factor field when using the distribution system.

The previous description of the FIRE repository and distribution systems and development are applicable to what is being called the FIRE/GLC system. The FIRE/GLCsystem was created by electronically extracting the target compounds listed in Table 3-1 from both the FIRE repository and distribution systems, and importing that data into new modules. No emission factor data were located for the following pollutants:

• N-Octyl phthalate;
• Hexachlorobutadiene;
• Hexachloroethane;
• 2,4,5-trichlorophenol;
• Atrazine;
• Heptachlor;
• Methoxychlor;
• Parathion;
• Pentachlorophenol;
• Trifluralin;
• Pentachloronitrobenzene; and
• Alkylated lead compounds.

No emission factor data were found under the term "polycyclic aromatic hydrocarbons (PAHs)," which encompasses a number of individual compounds. However, emission factors for specific PAHs (e.g., fluoranthene, chrysene, naphthalene) were located. One reason for the lack of emission factors for the above 13 pollutants is that nine of the compounds are used as pesticides/herbicides/fungicides, and source testing for pesticide compounds is limited. In addition, the use of many of these pesticides is being phased out or is significantly restricted.

The FIRE/GLC database management system contains the repository and distribution systems that include data for the pollutants listed in Table 3-1. A "tutor" module, included in the distribution system, is a module to be used with the FIRE user's manual for instructional purposes. The FIRE user's manual in WordPerfect® format and an on-line "HELP" section are also included in the system.

7.2.2 Mobile5a

Mobile5a is an integrated set of FORTRAN routines for use in the analysis of the air pollution impact of gasoline-fueled and diesel highway mobile sources. The program provides the user with a flexible analytical tool that can be applied in a wide variety of air quality planning functions. Mobile5a calculates emission factors for various types of light-duty and heavy-duty vehicles.

7.3 Source-Specific Emission Estimation Tools

The following discussion briefly describes several source-specific emission estimation tools to calculate emissions from water and wastewater treatment operations, storage tanks, and landfills.

7.3.1 Water and Wastewater Treatment Emissions Models

The following discussion describes software that can be used to estimate emissions from water and wastewater treatment operations.

SIMS - In response to requests from state and local air pollution control agencies involved in preparing VOC and toxic air pollutant emissions inventories, the U.S. EPA's Control Technology Center (CTC) and Emissions Inventory Branch (EIB) developed the Surface Impoundment Model System (SIMS). SIMS is a PC-based software package for estimating emissions from surface impoundments and wastewater collection devices. It can be used to estimate emissions from wastewater sources at hazardous waste treatment, storage, and disposal facilities (TSDFs); publicly owned treatment works (POTWs), industrial wastewater treatment facilities and other similar operations.

SIMS also contains models to estimate air emissions from diffused air surface impoundments, junction boxes, lift stations, mechanically aerated surface impoundments, nonaerated surface impoundments, surface impoundments with an oil film, sumps, and weirs. The SIMS emissions estimates are based on mass transfer models developed by U.S. EPA's Emissions Standards Division (ESD). SIMS can estimate a default inlet pollutant profile for the water discharged from any of 29 industry types.

CHEMDAT7 - CHEMDAT7 is a Lotus 123® spreadsheet prepared by U.S. EPA's ESD that includes analytical models for estimating VOCs from TSDF processes. The original models include disposal impoundments, closed landfills, land treatment facilities, and aeration and nonaeration impoundment processes. Predicted emissions can be viewed on the screen or printed. A graphical presentation of the relationships between emission prediction and vapor pressure, and between emissions prediction and the partition coefficient, is also available.

7.3.2 Landfill Air Emissions Estimation Model (LAEEM)

The Landfill Air Emissions and Estimation Model (LAEEM) is software specifically designed for use by state and local regulatory agencies to monitor the emissions of hazardous air pollutants (HAPs) from landfills. The system allows the user to enter specific information regarding the characteristics and capacity of an individual landfill and to project the emissions of methane, carbon monoxide (CO), nonmethane organic compounds, and individual HAPs over time using the Scholl Canyon decay model for landfill gas production estimation. The Scholl Canyon Model is a first-order decay equation that uses site-specific characteristics for estimating the gas generation rate. In the absence of site-specific data, the program provides conservative default values. The user may also tailor decay rate characteristics on an individual basis. An integrated decay rate constant calculator is provided for landfills that may be operating a gas recovery system to allow more accurate assessments of decay attributes. Outputs may be reviewed in either tabular or graphical forms. A help system is also provided with information on the model operation as well as details on assumptions and defaults used by the system.

7.3.3 TANKS

The TANKS program is designed to estimate emissions of organic chemicals from storage tanks. The calculations are performed according to AP-42 equations. The emission estimating equations that form the basis of TANKS were developed by the American Petroleum Institute (API). The user provides specific information concerning the storage tank and its contents (e.g., temperature, tank conditions, Reid vapor pressure of the liquid stored); the TANKS program, then estimates the annual or seasonal emissions and produces a report. The emissions can be separated into breathing and working losses. The TANKS program is written in FoxPro2.5®.

The TANKS program has a chemical database of over 100 liquids and meteorological data for over 250 cities. The user may add new chemicals and cities to their version of the database. The tank styles addressed in the program include underground tanks, vertical and horizontal fixed roof tanks, and internal and external floating roof tanks. The tank contents can consist of single or multiple liquid components. The estimates can be partially or fully speciated based on vapor or liquid composition data.

7.4 Spatial Analyses Tools

The Geographical Information System (GIS) uses modern computer technology to store, retrieve, analyze, update, and display spatially arranged data (e.g., maps). Because the characterization of emissions is enhanced by knowledge of the location and spatial arrangement of all identified sources, a GIS can be a useful tool for emission inventories. Locating each facility, defining the boundaries around each area source, and mapping all road networks can provide valuable information for formulating, evaluating, and implementing emission reduction strategies. Mapping facility and area sources is also important in defining and, subsequently, modifying planning regions of interest (e.g., the Great Lakes Basin). Map features are available in digital formats and from transportation departments, tax offices, planning/zoning offices, and emergency response agencies.