Great Lakes Regional Toxic
Air Emissions Inventory
Steering
Committee Conference Call Minutes
January 10, 2006
Participants
Illinois: Buzz
Asselmeier
Indiana:
Michigan: Dennis McGeen
and Allan Ostrander
Minnesota: Chun Yi Wu
New York: Carlos
Mancilla
Ohio: Tom Velalis
Ontario: Cong Doan and
Peter Wong
Pennsylvania: John Hulsberg
and Karen Gee
Wisconsin: Orlando Cabrera-Rivera
GLC: Jon Dettling
EPA:
Other: Mark Young
Programming language
Mark has been reorganizing some of
the data elements in the prior version of the design. He's been paying
particular attention to computation issues and is working on finding an ideal
location for some of the elements that are not in NIF, but are needed for
computation. He's also looked into the issue of defining time periods as
entities rather than using just start and end dates. Mark will show this during
the Chicago
meeting and would like to get feedback as to whether people think the approach
will work. Mark has also begun looking at the prior set of reference tables
from RAPIDS 2.X for inclusion in the new data model.
Orlando asked whether
is would make sense to have some data, such as for particularly complex
sources, to see how they would be represented in the current data model. Mark
has been using some of the RAPIDS test data, but this is somewhat old and may
not be ideal. If people have other data that might be useful for purposes of
mapping how it fits into the framework.
Tom asked whether the NEI
reference tables would be a sensible starting point. Mark stated that these
will be used in addition to the current RAPIDS reference tables. A combination
of these will likely be used, depending on the particular type of data and how
it is handled in the new data model. In some cases a combined table might be
made to assist in conversions. For example, the pollutant code table could
contain the original RAPIDS code, the NEI code, the CAS number and other
information that might be used elsewhere.
The group may need to make a decision
whether they would like to begin using the NEI numeric pollutant codes as a
primary key or whether they would like to keep the RAPIDS codes, with the
capability of mapping to the NEI codes.
Mark also suggested that there are
other types of data elements, such as control efficiencies and rule
effectiveness that it would be helpful if he had some "real world"
data to use to examine the mappings and relationships of these data types.
Carlos suggested that the CAS
codes could be used for the pollutant codes. John Hulsberg added that the EPA
has developed a "decimal CAS" system or something similar where some
issues around the potential for duplication due to the hyphens are resolved.
PDEP uses this system in many of their databases. Orlando pointed out that one issue with using
CAS is that there are some cases where CAS numbers don't exist for pollutants
or groups. This would need to be addressed. Mark pointed out that the NEI codes
are a hodge-podge of different things; some are CAS numbers (minus the
hyphens), some are based on something entirely different.
Tom said that the OEPA has just
adopted the NEI table as it is for reasons of not having to maintain the table
themselves, but just adopting EPA updates when they are available.
Orlando suggested
adding a field to the RAPIDS' Material Table for the NEI numeric pollutant
code. This field is already included in the Rap_Conversions Table.
Cong asked whether the Web-FIRE
database should be incorporated directly in some way. It may be worth including
this information directly. The group will also likely need to do periodic
updates of its own to that data. Tom noted that OEPA is developing an approach
to include Web-FIRE. He can distribute to the group any documentation they have
on this. This will probably include some kind of a reviewed update, where the
system detects new records and asks the user if they want to make the additions
and replacements.
Design issues to discuss at Chicago meeting
In Chicago, Mark is intending to show the data
model as it stands. Mark would also like to discuss data storage issues and
determine the most efficient way to handle certain types of information,
particularly the items that are currently handled in the activities table. Orlando asked Mark to
prepare and send in advance a list of items he plans to address. Mark will do
so. Jon noted that Pete has been working on the issue of integrating the
database design with PostGIS standards for handling of geographic information
and can present this to the group. Jon will check on the room location and will
send it to the group.
RAPIDS Programming Language
Following the last call, Jon and
Pete compiled a listing of what they saw as the advantages and disadvantages of
both Java and .Net as a language for the RAPIDS application. This was sent to
the group in December. They made an effort to include reference to the issues
and concerns that others have raised over this issue. Following this, Chun Yi
has sent some email correspondence from within their agency discussing the
implications of the language choice on their companion software, MIRROR. After
examining the issue, it appears that the only case in which use of Java would
present a constraint on the MPCA (who's IT personnel program in .Net
exclusively) would be if MPCA desired to make direct modifications to the
RAPIDS code. Programming stand-alone companion software (such as MIRROR) in
another language will not be a problem. Other states were asked whether that
had other outstanding issues or concerns and what there preference was for the
programming language.
IL: No preference
MI:No preference
MN:No preference
NY:No preference
OH:No preference
ON: prefer Java
PA: prefer Java
WI: No preference
GLC: prefer Java
Based on this, it was agreed that
Java would be chosen as the language to use.
Server Issue
Jon summarized the progress in
recovering the server that crashed over Thanksgiving weekend. It had been sent
to an outside contractor who specializes in Sun servers. The contractor
returned the server saying they had gotten the error fixed. However, the GLC
was able to access only 2 of 6 partitions when the server was re-mounted. The
contractor has been contacted again to reexamine the problem and restore full
functionality. GLC is waiting for their response.
Chun Yi expressed a concern over
the date of the last backup and would like to see a system in place for more
frequent backing-up of the repository databases. Jon concurred that this was
needed and committed to addressing it in the near future.
Benzo(a)pyrene
Jon will send a revised version of
the b(a)p report by next week so that the group has adequate time to review it
prior to the meeting. He has been working with Jon Bates to update the
information of Coke Oven Gasses and has been including some other changes the
group discussed previously and/or that have been sent by email. At the meeting,
he would like to devote 30-60 minutes to discuss developing a "qualitative"
assessment of a 10-15 year trend in b(a)p emissions for the region by examining
what we know about certain source categories. Other comments on the report can
also be discussed in Chicago,
with the intention of having a finalized version sometime in February. A final
dataset will also be compiled for release with the report and a downloadable
file.
Chun Yi, Jon and Orlando have
prepared and submitted abstracts regarding the b(a)p reassessment for both the
June AWMA Annual conference and the May EPA Emissions Inventory conference. For
each, a shortened version of the b(a)p report will be prepared and submitted
for the conference proceedings.