TESTIMONY OF STEPHEN HARPER, INTEL CORPORATION
BEFORE JOINT MEETING OF
SENATE JUDICIARY COMMITTEE AND
ENVIRONMENT AND PUBLIC WORKS COMMITTEE
JULY 16, 2002
Thank you, Chairman Jeffords
and Chairman Leahy, for the opportunity to address this joint hearing regarding
New Source Review policy issues. My
name is Stephen Harper. I serve as the
Director of Environment, Health, Safety, and Energy Policy for the Intel
Corporation. I am here to address the
committees today about one specific aspect of New Source Review (NSR), namely
Plantwide Applicability Limit (PAL) permitting approaches. Intel has been part of an informal coalition
of companies from the pharmaceutical, chemical, automotive, and electronics
industries that have been advocating promulgation of a PAL rule by the U.S.
Environmental Protection Agency (EPA) for several years now. Many of our coalition members have
experience with PAL-type permits at their facilities and believe strongly that
EPA should promulgate a PAL rule as a logical next step in a long process of
piloting, perfecting, and proliferating flexible permitting approaches that
protect the environment and provide operational flexibility to facilities.
Much effort has been
expended over the last ten years by industry, states, EPA, and the public –
under both Democrat and Republican administrations – to “reinvent” or innovate
new approaches to environmental protection.
Intel has participated in many of these efforts and is intimately
familiar with the mixed result of successes and failures from these
endeavors. We feel strongly that
PAL-type permits are one of the most successful innovations to emerge from
these many reinvention efforts. The
time has come to build on this success and take PALs into the mainstream of NSR
permitting.
Why does Intel care about
PALs and other forms of flexible permitting under the Clean Air Act? In simplest terms because of the importance
of operational flexibility in being able to innovate new products and processes
and quickly respond to market conditions.
As in many other industries, there are only two types of semiconductor
companies – “the quick and the dead.”
We feel strongly, therefore, about being quick.
Intel operates ten
semiconductor “fabs” or fabrication facilities in the US, producing PentiumÒ processors
and other semiconductor products. These
facilities employ many thousands of highly-skilled US workers. The capital investment required to bring a
new fab into full production is in the $2-3 billion range. The life-cycle of a semiconductor fab
involves numerous upgrades and innovations in production technology, chemicals,
and processes. A “typical” Intel fab,
for example, experiences two or more technology generations over a five-year
period; as many as 75 upgrades and innovations each year in process steps,
methods, and chemicals; and the installation of between 175 and 500 new process
tools over a two-year technology transition.
Once a fab has commenced
production, profitability depends upon reaching and maintaining high levels of
production as quickly as possible. Traditional air quality permitting
approaches, under NSR and other EPA and state programs, would require
potentially hundreds of permit revisions to implement the upgrades and
innovations that are critical to successful start-up and ramp-up of a fab. The potential delays attendant to such
revisions are – simply put – incompatible with the profitable operation of
U.S.-based semiconductor fab that must compete in a global marketplace where
success hinges upon being quick-to-market.
Traditional permitting approaches would require numerous permit
modifications and threaten significant delays for companies like Intel as we
install new manufacturing tools, convert to new manufacturing processes, change
chemicals, and expand production capacity to respond to market conditions.
Driven by the
incompatibility of traditional permitting approaches with semiconductor
manufacturing requirements, Intel has long pursued an objective of minimizing
our permitting burden. We have done
this in two ways. The first is to
reduce our emissions of all pollutants as much as we can so as to achieve
“minor source” status under the Clean Air Act. The second priority has been to work with EPA and the states to
pilot and prove new, innovative, and more flexible permitting approaches.
A PAL permit provides an
emissions cap or caps for an industrial facility. The cap provides a clear method for determining whether changes
at a PAL-covered facility trigger NSR permitting requirements. The need to obtain an NSR permit revision
only applies when a facility’s emissions increase beyond the PAL cap. In addition to the cap, a PAL or PAL-type
permit typically specifies certain kinds of facility changes that are
“pre-approved.” A facility with a PAL
can undertake a pre-approved change without becoming subject to NSR as long as
the facility’s emissions remain below the cap(s).
It is important to clarify
the difference between a PAL permit under the NSR program and what I am terming
a “PAL-type” permit. PALs per se
relate only to facilities that qualify as “major” under the Clean Air Act by
virtue of the magnitude of their emissions.
I am using the term “PAL-type” permits to refer to minor source permits
involving both an emissions cap and pre-approval of certain operational
changes. As I will make clear shortly,
Intel has experience with both types of permit.
There are three categories
of benefits provided by PAL and PAL-type permits. Most importantly, PALs provide significant environmental
benefits. PAL emissions caps provide
certainty regarding the emissions impact of a facility. Moreover, since these emissions caps are set
at levels that reflect the air quality improvement needs of an airshed, PAL
caps typically entail emission reductions compared to traditional permitting
approaches. Emissions caps, moreover,
provide a very powerful incentive for pollution prevention. The only way a facility can increase its
production and still stay under its cap is to reduce its emissions per unit of
production. PALs allow facility
environmental engineers to spend less time dealing with the burdens of
permitting paperwork and free them up to concentrate on reducing emissions
through pollution prevention.
A second benefit PALs
provide is enhanced public participation. Under traditional approaches, air quality permitting authorities
notify the public of numerous changes, big or small, at facilities, providing
opportunities for public input into whether or not permit modifications should
be granted. At best, what the public
sees in the traditional case is a series of incremental changes and piecemeal
information about facility operations that provide little understanding
regarding the overall impact of a facility on local air quality. Under a PAL, however, the public has the
opportunity to be involved in the initial process of establishing the PAL
permit and emissions caps. In this
setting the public can gain a much better sense of the overall operations of a
facility, the kinds of operational changes that are contemplated, and the
likely air quality impacts of the facility over the term of the permit. The public has a much enhanced opportunity
to view the facility holistically, rather than in a fragmented way.
A third type of PAL benefit
accrues to the permitted facility in the form of operational flexibility. For major sources concerned about NSR
applicability, PALs provide a “bright line” that eliminates ambiguity about
whether or not operational changes trigger NSR requirements. PAL-type permits provide minor sources the
same type of flexibility regarding state minor source NSR requirements.
A major part of our
corporate commitment to innovating new permitting approaches has involved
partnership with EPA, the states, and members of the public to pilot the basic
concepts underlying the PAL rule that EPA currently is finalizing. The first of these partnership commitments
came in the 1992-1995 timeframe where Intel, EPA, and Oregon developed a PAL
permit for Intel’s Aloha, Oregon fab as part of EPA’s “Pollution Prevention in
Permitting Program” (P4). The second
major partnership involved Intel, EPA, and Maricopa County, Arizona jointly
undertaking one of the first pilot projects under EPA’s “Project XL” program at
its Ocotillo campus in Chandler, Arizona.
Intel’s P4 permit was a PAL
permit under the Federal NSR program because our Aloha fab was a major source at
the time the permit was issued. Our XL
permit for the Ocotillo fab is not, strictly speaking, a PAL, because that
facility is a minor source under the Clean Air Act and, thus, no NSR
“applicability” issues arose.
Nonetheless, our Ocotillo permit functionally is the same as the Aloha
permit and has provided another valid test of the emissions cap and
pre-approved changes features of a PAL.
I previously described the
environmental benefits of PAL permits.
Let me now show how those benefits were realized in practice in our
Oregon and Arizona pilot projects. The
environmental benefits at our Aloha, Oregon fab are very dramatic. The attached exhibit provides a
graphic demonstration of the powerful incentive PALs provide for aggressive
pollution prevention programs. This
chart shows facility VOC emissions per production unit and total production
units. Motivated by the need to find
room for growth under our PAL cap, our Aloha fab reduced emissions of VOCs by
over 90 percent per unit of production since 1990. Some of this reduction occurred prior to 1995 under an Oregon
PAL-like permitting program. Even more
dramatic reductions occurred after our NSR PAL came into effect in 1995.
The combination of the
pressure of an emissions cap and the operational flexibility under our Aloha
PAL fueled an aggressive pollution prevention program. The success of that program allowed Intel to
add an additional fab at our Aloha campus without the need to increase our cap. Indeed, we reduced overall VOC emissions and
voluntarily lowered our VOC cap from 160 tons per year to 130 tons per
year. This was done to support the
successful efforts of Oregon and the Portland region to reduce overall regional
emissions and qualify Portland for re-designation as an Ozone Attainment area
in 1997. Intel’s consistent reductions
over time, combined with this area redesignation, allowed our Aloha fab to
itself achieve minor source status under the Clean Air Act in 1999.
The environmental results
under our PAL-type permit at our Ocotillo campus have been equally
dramatic. Through our aggressive
pollution prevention program, the Ocotillo facility – which sits on 720 acres,
employs approximately 5,000 people, and produces a high volume of semiconductor
devices – emits approximately 25 tons of VOCs annually. This emissions level is in the neighborhood
of what several large gas stations would produce. Our emissions reductions at Ocotillo have been so dramatic that
we have constructed and are now operating a second fab on this campus – all
under the XL cap.
As I have shown, Intel’s
experience piloting PALs and PAL-type permits with EPA and state and local
permitting authorities has been dramatically successful. Other companies that have worked with EPA
and the states to test the PAL approach also can tell similar success
stories. Several of the other companies
in our informal “PAL coalition” -- including DaimlerChrysler, DuPont, and Merck
– have successfully piloted the PAL approach at one or more of their
facilities. Other companies in other
industries are applying the PAL approach as we meet today, including BMW,
GM/Saturn, and several oil refineries.
At this point, PALs have been demonstrated successfully in a number of
very different industrial sectors.
Intel believes, as I
mentioned at the outset of my testimony, that PALs are one of the most
significant regulatory innovations to emerge from the last ten years of
regulatory reinvention activities at the Federal and state level. Indeed, PALs are an example of the right way
for a regulatory agency like EPA to innovate.
First you try some pilot projects.
You evaluate your experience and, where success has been demonstrated,
you build on that success by mainstreaming the innovation in your regulatory
program.
“Mainstreaming” the PAL
success story will be aided greatly by EPA promulgation of practical PAL
provisions either discretely or as part of a larger NSR rule. Regulatory action is not necessary as a
legal matter. The successful PAL pilot
projects at Intel and other companies have utilized existing legal
authority. On the other hand,
regulatory action by EPA will promote the PAL concept by providing greater
guidance to permitting authorities and sources regarding the benefits of PALs and
PAL-type approaches. With the
promulgation of clear “rules of the road,” sources and states will be better
able to craft PAL permits that realize the environmental, public participation,
and operational flexibility benefits I have cited.
In sum, Congress should be
encouraging flexible permitting approaches like PALs. I will be glad to answer any questions the committee members
might have.