May 9-10 Meeting, Madison
Wisconsin
Participants
Illinois: Buzz
Asselmeier
Indiana:
Michigan: Dennis McGeen
and Allan Ostrander (phone)
Minnesota: Chun Yi Wu
New York: Carlos
Mancilla
Ohio: Tom Velalis
Ontario: Cong Doan
Pennsylvania: Karen Gee
(phone)
Wisconsin: Orlando Cabrera-Rivera
GLC: Jon Dettling and Pete Giencke
EPA: Suzanne King (phone)
Other: Mark Young (phone)
Project Staffing
Changes
Jon will be leaving the GLC this
summer, likely around the end of June. There is a good chance he will remain
with the project under some sort of arrangement for some period of time beyond
June, but how long is uncertain and will depend both on the GLC's progress in
replacing him and on his own job-hunting progress.
Orlando noted
that he will likely be leaving the WDNR for a 2-year position with the Commission for
Environmental Cooperation in Montreal.
He will be the project manager for the CEC North American pollutant release
inventory.
WDNR Emissions
Inventory System Demo
Orlando gave
the group a demonstration of the AEMS system Wisconsin uses to compile the emissions
inventory. The basic structure is similar to RAPIDS,
with a few exceptions. Some of the screen navigation features in AEMS could be considered in
the new RAPIDS re-design. There is an ability to change
the emissions data year by selecting from a menu.
There is a "select" feature allowing the user to select sources based
on a wide variety of criteria. Orlando
gave an example where all sources emitting arsenic were selected and this then
becomes the working group of sources. A source can be selected from the list.
>From that point, in addition to the usual RAPIDS functions, there is an ability
to look at all emissions from that source. This can be done from the source
level without needing to drill down to the stream.
Orlando showed that
within the process screen, the SCC description is shown, including the 4
description levels. Cong asked whether changing the SCC will trigger the
calculator to recalculate emissions with the new SCC. Orlando responded that this depends on
whether the emissions are reported or calculated.
Orlando showed the
emission factor table. This includes a column for the reporting threshold for
each pollutant. This can be used to check the calculations that are done by
WDNR to determine whether emissions were properly reported. This screen also allows
for the selection of the emission factor to be used in the calculations. Buzz recalled that previously
there was an ability to see what emission factors are being used by other
sources with the same SCC elsewhere in the state.
Orlando sent to the
group previously a link to a video on the WDNR website that shows the web
reporting feature. Chun Yi asked whether the web interface requires manually
entering the numbers or if a data file can be uploaded. Orlando was unsure and will try to find
someone during the break who can better answer questions regarding how the web
tool works.
Within AEMS, the query screens
have a "save file" button that allows a text file of the list of selected sources to be
exported.
Orlando also showed
that within the source selector, there is a field for "region," which
are groups of counties. You can therefore select sources a run queries by
regions. WDNR used this to allow looking at sources based on the WDNR
management regions (e.g., SE Wisconsin).
Orlando showed that he
uses a set of queries in Access to produce RAPIDS import files from AEMS. He
has done this also for the NIF format for mobile sources.
Orlando showed that
there is a standard function in AEMS that shows two consecutive years of
emissions reported for a given source. This is useful for quality assurance
purposes.
Project Wiki
Website
Jon demonstrated to project Wiki
website (http://www.glos.us/wiki/display/RAPIDS/Home).
This website allows users to make edits to the content through their web
browser. Potentially, this can replace many of the other project file sharing
and storage tools as a place to keep track of project documents. Because of the
editing and change-tracking features, it will be an advantage to have many
project documents moved to Wiki pages so that there is a single version and
when edits are made, they are made to this common page. Jon has already begun
moving various items to the website as a starting point. This includes the area
source methodologies, project pollutant list, the project protocol and much of
the content of the original project website. The Wiki allows a high level of
access control, so that some items can be viewable by the public and others
only by those within the group. Similar control settings are available for
editing content.
The site allows for attachments to
be posted to web-pages and links to documents or other webpages to be inserted.
There is a search function that will search all the pages in the site as well
as the text of all attachments (e.g., .xls, .doc, .pdf files). This is a
powerful tool for locating documents within the site. Very large documents
(100MB or bigger) can be attached, so the site has the potential to replace the
FTP site as a means of exchanging and storing project documents.
Jon briefly demonstrated how to
log into the site and view and edit pages. All group members are able to make
edits to the web pages and it should be a group responsibility to keep things
up to date. If there are any questions on how to access or edit the site,
people should contact Jon.
Emissions Factor
Tables
Jon sent a file of emission
factors for area sources prior to the meeting. This file is based on tables of
emission factors that many of the states/province sent to Jon several weeks
prior to the meeting. For each area source SCC, a table is given showing the
emission factors contained in each state's system, including the pollutant, the
numeric factor and the denominator unit.
There is also a table showing for
which SCCs each state has emission factors in their system. This is one means
of comparing the emission factors used by each state/province for each area
source category. Jon made a recommendation that a mapping of SCCs for area
sources (and maybe also mobile and nonroad sources) be made for reporting of
data and storage in the regional repository. This should help solve
inconsistencies among states/provinces regarding which SCCs are used and should
eliminate some confusion that can arise when inexperienced people are viewing
the data, such as assuming the lack of emissions for a state within a certain
SCC means that that state didn't estimate that source category. For mobile
sources, this might result in some significant reductions in the data file
size, as it's usually the mobile sources data that dominates the data volume
due to the many divisions of SCCs. Each state province could still estimate and
report emissions how they chose. These changes would be imposed at the level of
the regional repository. After some discussion, the group agreed to this
scheme.
For the emission factor tables,
Chun Yi noticed that many of her factors were missing. She will look into this
and send Jon an update. It was also noted that it would be useful to have the
NEI methodology emission factors included for comparison. In many cases, it
will be likely that the factors used by most states are the NEI numbers. For
others, there is a need to document the source of the factors. Including the
NEI numbers will help pinpoint those factors where information on the source is
needed.
Jon noted that he eliminated from
the list those pollutants that are not on the group's master list of material
codes for the project. This included omitting PM and VOC factors. In some
cases, the "emission factors" included by some states appear to
actually be speciation factors in that the denominator is in units of PM or
VOC. If possible, it may be good to restore these PM or VOC factors, at least
for SCCs where this is the case.
Jon will work on continuing to
update these tables. In the meantime, they can serve as a reference to people
as they compile their 2005 area source files. If there are emission factors
being used by others that they are not including, they can check with that
state to get more information about that factor. For 2005, an attempt will be
made to run a QC check on combinations of pollutants and SCCs to ensure that if
emission factors are available, people are using them.
2005 Inventory
Each state/province gave an update
on its progress in preparing the 2005 inventory:
IL: The 2005 inventory is complete
MI: Point sources are done and
QA/QC'ed. Area sources are about 2/3 done. Will be completed shortly.
MN: Point sources are in progress.
A few area source categories have been completed. Chun Yi will be on vacation
until June 19th. She should be able to complete the inventory by early
September.
NY: Point sources are done and
QA/QC'ed. There were some errors in attempting to run this data through the
RAPIDS import. The area source calculations are in progress. Mobile sources
have been done. Non-road has been delayed due to staff being devoted to working
on GHG emissions
OH: Currently working on the CAP
inventory. Toxics will be completed by September.
ON: Most area source sectors are
done. Pesticides and autobody refinishing have been added. Point sources is
still waiting on some additional changes. Mobile
and Nonroad have not yet been done.
WI: Point sources are done. Area
sources are waiting on the EF discussion from this meeting to reach some
resolution. Grant will be doing the nonroad data for toxics. The mobile source
inventory should be completed by the end of June.
The group agreed that all data
compilation could be completed and submitted for QA/QC by September 15th. It is
expected that the RAPIDS3 imports will be working by that time and that data
can be submitted in either RAPIDS2 format or NIF3 for import into the new
system.
RAPIDS Design
Mark Young distributed some
documents for discussion prior to the meeting. Based on the notes from the
prior discussion, Mark has changed some of the description field to have longer
lengths. The geographic table is not shown, but has been drafted. This will
include the PostGIS format for processing the geographic data. In many cases,
the numeric fields have been changed to double-precision. In addition, many
fields have also been given a companion unit field to allow for more
flexibility in which units are used. In some cases, certain fields may also
need a denominator units code. There was a question about the "maximum
nameplate capacity" field. This appears to be only for EGUs and it isn't
used at all in the typical RAPIDS estimations. It is a part of the NIF format.
Chun Yi asked about the
possibility to include multiple years of data. Mark responded that this would
be up to the user; the design can support it if the user would like to do so.
Chun Yi mentioned that some emission units in her state change their emission
unit during the course of the year. She's not sure how this would be handled
with start dates and end dates. Mark noted that operation start and end dates
could be added to the system if this is needed to represent certain situations.
The geographic coordinates can
correspond to any level, including both release points and source-level.
In addition to the coordinates,
the regions is a component that is not yet in the ERDs. Mark will be able to
send this later in the day.
Tom raised the issue of
"volume sources," which are fugitive emissions exiting a building
from an opening at a given height. There may be some additional fields needed
to represent this. However, Tom said that after looking at the release
points table further, he thinks the volume sources in his inventory can be
represented adequately.
Chun Yi also mentioned that for
geographic representation, EPA is planning to accept shapefiles in the future.
Some additional information may be needed to generate these. For example, the
location of tribal emissions could be a concern.
Cong asked whether it is necessary
to have both numerator and denominator for flow rate and velocity units. The
group agreed that it would be sufficient to have just a single unit field for
these items, for example 'cubic feet per minute' all in one field. Generally,
when the denominator is always of the same type (e.g., time or length), then it
is easiest to just have one unit field. Mark noted that a new unit can always
be defined. Mark advised that it will be important to have a standard
conversion unit for each field and then any new unit can be defined as long as
a conversion to that standard is also defined.
On the processes table, Mark asked
whether the heat content, sulfur content and ash content fields are essentially
activity-type data and whether they belong in the processes table. Are these
always associated with processes? Tom suggested that these could perhaps be
placed in the emissions table. These are items that are used in cases where
emissions are based on a calculation rather than just an emission factor. Buzz
asked whether anyone ever has multiple values for these fields for
the same process. If not, then the process table is the most logical place for
this. Cong stated that these are characteristics of the fuel and not of the
throughput. The group agreed to leave these items in the process table. This
will influence how the calculation is set up, but it should be able to be done
either way. Mark agreed that the calculation can go find the data wherever it
is at. The issue to decide is where it makes the most sense to store the data.
The heat content will need a units field (a single field).
It was agreed that for the sulfur
and ash content can be changed to two decimal places for precision. For control
efficiencies, it may be best to leave these as double-precision.
Cong asked whether it makes sense
for dioxins to include TEQ as a unit rather than as part of the material code.
Chun Yi is not sure this will work because some emission factors do not specify
the material, other than that it is a mixture of PCDDs. This will take some
research to determine what is the best way to represent these.
Regarding control devices, Mark
asked whether a sequence number is required to describe the order of the
controls. If the controls are represented in separate records, the sequence
might be needed. Buzz suggested that having data in the primary, secondary or
tertiary control fields could indicate the order. Mark noted that there is a
potential pitfall in the NEI export for doing that because primary control is
required. However, the export may be able to work around this. The NEI only has
the primary control efficiency and total control efficiency. It may come down
to determining whether anyone is tracking their control devices individually.
If so, it would be desirable to have multiple records, perhaps with a sequence
indicator. However, if only total control efficiencies are being used, this may
not be an issue. Tom suggested that in addition to the control efficiencies,
the capture efficiencies would be important for determining the fugitives. Tom
likes the table as it is currently shown. Mark reiterated that if there is any
desire to represent multiple devices with their own control and capture
efficiencies, a sequence field would be needed. There was some discussion on
how to calculate capture and control efficiency and how this would be
represented in the system. In the NEI there is only a single combined total of
capture and control efficiency. This requires the user to do the calculations
on the set-up in question outside of the system and then include the result.
In the activity data table, Mark
asked whether there should be separate fields for maximum, minimum, potential,
etc. These have been carried over into the emissions table. They may not be
needed in other activities. For operating mode, nobody uses multiple values for
these. Chun Yi asked what the value_calc field is relative to the other value
fields. This is used by the calculator to hold the result of the value field
and the exponent field. Several people expressed a preference not to have a
separate exponent field, but to have the main value field have the exponent
with it. The value fields can therefore be changed to include just a single
field.
Chun Yi asked what the
data_source_code is. This is the same as the data_code field in the old system,
which indicates where the values came from (calculated, facility-supplied,
etc.). Chun Yi also asked about the reference codes in the old system. This is
now replaced by the combination of the scenario code and the inventory code.
Mark also noted that the
emission_data_level field used to be a key field in RAPIDS. It's not required
in NEI, but it might be useful to have this required still in RAPIDS. Having
differing codes here, such as site, unit, release point and process, would
allow data to be rolled-up or represented at various levels. Another approach
may be to have a separate table that has roll-up (e.g. source-level) emissions.
Chun Yi asked about the
emission_type field. This is used to specify a typical temporal category, such
as annual or weekday.
Mark discussed a table he included
showing how various data elements are defined, such as generic and state
specific emission factors and rule effectiveness. Each of these may have a
partition of the inventory, and entity level and other characteristics. Mark
will be working on developing this table further. Orlando commented that this will be a little difficult
to grasp. Once this is complete, there may be some additional changes to the
data model.
Next call will be May 23 at 1:30
EST.
The next meeting will be late
august / early September