Intel
Final Project Agreement -- Air
Screening Analysis
Screening Analysis
For Air Emissions At the Ocotillo Site
In cooperation with Maricopa County Environmental
Services Department ("MCESD") and the stakeholders involved
in Project XL, Intel agreed to analyze the potential effects associated
with the scenario of any single hazardous air pollutant ("HAP")1
being emitted during routine operations from the Ocotillo Site to the
full limit of the relevant plant site emission limit ("PSEL").
This is accomplished utilizing a SCREEN3 dispersion model to determine
the property line and maximum onsite air concentration of the chemical
which is compared to the Arizona Ambient Air Quality Guidance (AAAQG)
level2. In addition, as set out in the Final Project Agreement
("FPA"), Intel and MCESD conducted this screening analysis
for a number of non-HAP chemicals which have been assigned an AAAQG.
Any new chemicals introduced at the Ocotillo Site which produce air
emissions will also undergo this screening analysis. This analysis provides
additional evidence that the PSELs set out in this FPA are protective
of human health within the limitations and uncertainties associated
with the analytical techniques employed. This attachment to the FPA
describes, in general terms, the screening analysis for air emissions
at the Ocotillo Site. The specific model protocols and their site-based
parameters are maintained by MCESD and Intel.1 The term Hazardous
Air Pollutant refers to the list of 189 chemicals and chemical categories
set out in Section 112(b) of the Clean Air Act, as amended. 2
The Arizona Ambient Air Quality Guidelines (AAAQG) referred to herein,
and in the air permit, are the guidelines dated July 15, 1992.
1. Dispersion Modeling
Intel, in cooperation with MCESD, the United States Environmental
Protection Agency ("EPA") and the Arizona Department of Environmental
Quality ("ADEQ"), has performed dispersion modeling as a screening
analysis for the Ocotillo Site. The screening analysis was conducted using
EPA's approved SCREEN3 dispersion model (1991-1996, Trinity Consultants
Inc. Version 1.0). The parameters used in the screening analysis are listed
at the end of this attachment. The screening analysis assumed an emission
rate of 10 tons per year for any HAP. This emission rate was based on
the maximum PSELs for aggregate organic and aggregate inorganic HAPs set
out in the FPA and on the assumption that any single HAP could be emitted
to the full extent of the relevant PSEL. Under the screening analysis
the dispersion model predicted that the emissions rate of 10 tons per
year would result in maximum properly line concentrations of 14.9 micrograms
per cubic meter (ug/m3) averaged over a 1 hour. period, and 5.96 ug/m3
averaged over a 24 hour period.
2. The AAAQGs
The Arizona Department of Health Services (ADHS) has developed
a list of AAAQG levels. These levels were derived by making an adjustment
for the differences in the averaging times for exposure and applying a
safety factor to limits originally established to protect individuals
exposed in occupational settings. For example, occupational exposure levels
are intended to be safe for individuals exposed to those levels for 8
hours per day, 7 days per week for a working lifetime. They generally
assume an 8-hour average exposure time; however, longer average times
(e.g., 24 hours) are more appropriate for establishing community health
guidelines. In addition, a safety factor is applied to provide adequate
protection for the general public, which includes people who may be more
sensitive than workers (e.g., children and the elderly). The AAAQGs established
by ADHS are for guidance purposes only and are not intended for use in
deriving regulatory limits. Notwithstanding the non-binding nature of
the AAAQGs, Intel, in cooperation with the stakeholders involved in the
XL Project, employed these levels to provide additional assurance that
the PSELs in the FPA and air permit are protective of public health. In
addition, to address concerns raised by stakeholders regarding non-HAP
chemicals, Intel also analyzed chemicals emitted from the Ocotillo Site
that, while not HAPs, have established AAAQGs. Intel also has agreed in
the FPA to work cooperatively with the ADHS and MCESD to evaluate the
public health implications and establish limits if necessary for any chemicals
it may introduce to the air in the future that are associated with potential
health concerns, regardless of whether such chemicals are HAPs or have
an established AAAQG. Finally, Intel has agreed under the FPA to apply
the screening analysis to maximum onsite ambient air concentrations of
chemicals modeled under the FPA. The AAAQGs are well suited for the screening
analysis described herein because they are likely to be conservative (i.e.,
err on the side of public health by setting levels well below what would
be likely to cause adverse effects in the general population). This conservative
approach, combined with the conservative assumption that screened chemicals
will be emitted at 10 tons per year in the case of HAPs (or at the predicted
emissions rate for any non-HAP if such rate is greater than 10 tons per
year), make this screening analysis a useful tool in providing an additional
assurance that air emissions do not pose a special risk to employees,
onsite visitors or the community.
3. Comparison of Concentration Estimates to AAAQGs
The concentrations of chemicals listed in Table 1 at the
end of this attachment predicted by the dispersion analysis described
above were compared to the AAAQGs. The conservatively modeled concentrations
did not exceed the applicable AAAQG for any chemical listed in the table
that is emitted to the air from the Ocotillo site, with the exception
of phosphine. Pursuant to the FPA, Intel elected to waive the right to
conduct more sophisticated dispersion analysis and requested that an individual
emission limit for phosphine be established based on the screening analysis.
The screening analysis predicted that an emissions rate of 5 tons per
year would limit the maximum properly line concentration to a level below
the AAAQG established for phosphine. Such a limit is set forth in the
FPA. In cooperation with the stakeholders and the MCESD, Intel also has
conducted a screening analysis (using the procedure described herein)
for other chemicals emitted to the air from the Ocotillo site which are
not HAPs but for which there are established AAAQGs. For each of these
non-HAP chemicals, the screening analyses predicted that the maximum property
line concentrations would be well below (in most cases more than an order
of magnitude) their respective AAAQGs.
4. Introduction of Additional HAPs or Non-HAP Chemicals
with AAAQGs
The FPA sets out the conditions that apply in the case
where Intel introduces new chemicals to the Ocotillo Site which produce
air emissions are not listed in Table 1 of this attachement, and thus
have not been analyzed under the screening analysis described herein.
In general, Intel will evaluate new chemicals which are emitted to the
air through the screening analysis in cooperation with MCESD using either
the SCREEN model or a similar dispersion model approved by EPA. In the
case where the screening analysis indicates that the predicted maximum
concentration of a new chemical exceeds the relevant AAAQG, the FPA sets
out options for conducting more sophisticated dispersion modeling analyses
(using EPA-approved dispersion models) to confirm or refute the results
of the initial screening analysis. If more sophisticated modeling confirms
that the relevant AAAQG is exceeded by anticipated emissions of a new
chemical, MCESD shall establish an air emissions limit for such a chemical
at a level that corresponds to the relevant AAAQG. Intel also may waive
the right to conduct more sophisticated dispersion analysis and elect
to accept an emissions limit for such a chemical established by MCESD
at a level that corresponds to the relevant AAAQG based on the initial
screening analysis.
5. Introduction of Additional Non-HAP Chemicals Without
Established AAAQGs
Section II(A)(3) of the FPA provides a special procedure
that Intel has committed to follow in the event that the Company introduces
new chemicals to the Site in the future which generate air emissions and
have not already been evaluated under the FPA, or under the AAAQG screen
modeling procedure set forth above, and which present potential concerns
to human health or the environment. In such a case, Intel agrees to consult
with MCESD and ADHS to determine if emissions from such a chemical may
pose a health risk based on screen modeling of potential property line
concentrations. As stated in the FPA, if it is determined that an emissions
limit for the chemical is necessary to protect human health, Intel will
limit its annual emissions below the limit which is identified.
6. Inside-the-Property Line Screening Analysis
Under the FPA, Intel also commits to evaluate maximum
onsite (i.e., inside-the-property line) modeled ambient air concentrations
of chemicals that have been modeled under the permit which generate air
emissions, to ensure employee safety and the safety of individuals that
may visit the site. If the screen model analysis indicates a potential
for exceeding the relevant 1-hour AAAQG exposure level for a particular
chemical used on site, Intel may either commit to limit its emissions
of that chemical below the level that would exceed the 1-hour AAAQG at
the point of maximum concentration within the property line or demonstrate
though a more refined EPA-approved analysis that the 1-hour AAAQG will
not be exceeded.
7. Strengths and Uncertainties of the Screening Analysis
The primary strength of the analysis described above is
its conservative approach and assumption that screened chemicals will
be emitted at the full limit of the relevant PSEL or higher in the case
of non-HAPs if the predicted emission rate is greater than 10 tons per
year. The likely over-prediction of concentrations from the SCREEN3 model
and likely under-prediction of safe health risk levels by the AAAQG methodology
make the screening analysis a valuable tool in providing additional assurance
that air emissions at the Ocotillo Site do not pose a special risk to
employees, visitors or the community. The precautionary principle states
that in the face of uncertainty, additional safety factors should be applied
to the analysis. In the analysis described above, several safety factors
have been applied but, as with all simulations of the real world, not
every scenario can be anticipated. For example, synergistic effects (i.e.,
when exposure to a combination of chemicals is more hazardous than exposure
to any single chemical) or antagonistic effects (i.e., when chemical combinations
are less hazardous than the individual chemicals) are not understood by
today's science. In addition, unanticipated dispersion events cannot be
adequately predicted by simulation models. Finally, the potential reaction
of some sensitive subgroups within the population may not always be fully
accounted for by the additional layers of safety factors.