BEFORE THE ADMINISTRATOR
              ENVIRONMENTAL PROTECTION AGENCY
                      Washington,  D.C.
        In re:   APPLICATIONS FOR SUSPENSION OF 1977

          MOTOR VEHICLE EXHAUST EMISSION STANDARDS
         Chrysler Corporation, Ford Motor Company,
        and General Motors Corporation, Applicants.
               DECISION OF THE ADMINISTRATOR
March 5, 1975

-------
               DECISION OF THE ADMINISTRATOR
I.   Introduction

     Section 202 of the Clean Air Act, 42 U.S.C. 1857f-l,
as amended by Section 5 of the Energy Supply and Environ-
mental Cpordination Act of 1974, P.L. 93-319, 88 Stat. 258,
requires that emissions of carbon monoxide and hydrocarbons
from automobiles sold in this country during the 1977 model
year be reduced by at least ninety percent from their 1970
levels.  Under the standard EPA test  procedure, the emis-
sion levels needed to comply with this requirement are a
maximum of .41 grams per mile (g/mi) of hydrocarbons  (HC)
and 3.4 grams per mile  (g/mi) of carbon monoxide (CO).  In
addition, that section authorizes the Administrator of EPA
to suspend the effective date of these reductions for one
year only - until the 1978 model year - if he finds after
public hearings that the following conditions have been met:

          "The Administrator shall grant such
          suspension only if he determines that
          (i) such suspension is essential to
          the public interest or the public
          health and welfare of the United States:
          (ii) all good faith efforts have been
          made to meet the standards established
          by this subsection; (iii) the applicant
          has established that effective control
          technology, processes, operating methods,
          or other alternatives are not available
          or have not been available for a suf-
          ficient period of time to achieve
          compliance prior to the effective date
          of such standards, and  (iv) the study
          and investigation of the National
          Academy of Sciences conducted pursuant
          to subsection  (c) and other information
          available to him has not indicated that
          technology, processes, or other alter-
          natives are available to meet such
          standards."

-------
                         - 2 -
     Before the 1974 amendments, the Clean Air Act required
the 90% HC and CO reductions to be achieved by the 1975
model year, subject to a one-year extension which the
Administrator of EPA could grant if he found that the auto
companies had satisfied the statutory requirements quoted
above.  Although the 1974 amendments changed the law to
defer the required reductions and related suspension pro-
vision for two years, they did not change either the level
of the required HC and CO reductions or the terms on.which
a suspension of their effect could be granted.*  Proceedings
under the prior version of the statute are therefore rele-
vant to my decision here.

     The first application for a suspension of what,were
then the 1975 standards was filed with EPA on March-. 13,
1972 by A.B. Volvo Ltd,, of Sweden.  .Shortly thereafter-,-.
applications were also received from Chrysler, Ford,. General
Motors,, and International Harvester.  Former Administrator
Ruckelshaus denied all five applications in a decisipn,
issued May 12, 1972.

     The four American applicants appealed this decision:to
the courts, and on February 10, 1973, the United States
Court of Appeals for the District of. Columbia. Circuit
remanded the applications to EPA for further consideration.
International Harvester Co. v. Ruckelshaus, 478 F.2d 615
(D.C. Cir. 1973).  In its opinion the court defined a wide
range of both technical and public policy issues which EPA
had to consider and make findings on before a decision to
deny suspension could be sustained under the law.

     Public hearings were held under the remand order in
March of 1973, and on April 11, Mr. Ruckelshaus issued his
decision upon remand,  38 Fed. Reg... 10317  (April 26, 1973).
He found that although the catalytic converters needed on
     *Congress did, however, change the oxides of nitrogen
 (NOX) emission standard that would accompany the statutory
HC and CO standards.  For 1975, the NOX standard - established
administratively by EPA.- was  3.1 g/mi; a level of 2.0
g/mi has been legislatively, established for 1977, with a
level of 0.4 g/mi to be achieved in the 1978 and subsequent
model years.

-------
                        - 3 -
nearly all vehicles in order to achieve the statutory stan-
dards had been developed to the point where they were an
available and effective means of emission control, there
had been so little experience in manufacturing and instal-
ling them that it was not prudent from a mass production
standpoint to force their use on all cars sold in a single
model year.  In addition, a rigorous technical analysis
had shown that models representing only 66% of domestic
auto sales could be predicted with high confidence to
meet the statutory standards in the 1975 model year, and
this was regarded as too low a number to ensure that the
"basic demand" for automobiles would be satisfied if the
suspension were denied.

     Accordingly, the Administrator granted a suspension
and established two sets of interim emission standards for
the 1975 model year.  Cars sold in California, which is
the only state permitted by the Clean Air Act to have its
own auto emission standards, were required to meet levels
of .9 g/mi HC, 9.0 g/mi CO, and 2.0 g/mi NOX.  These HC and
CO limits were set at levels thought to require the use of
catalytic converters on most model lines.  Cars sold in
the other 49 states were required to meet levels of 1.5
g/mi HC, 15 g/mi CO, and 3.1 g/mi NOX.  These HC and CO
limits were set at levels thought not to require the use
of catalytic converters on most model lines.  These stan-
dards have since been adopted by Congress for the 1976 model
year.  A full range of automobiles has been certified and
is being produced for 1975 at both emission levels.*  About
85% of 49 state cars and almost all California cars will
be equipped with catalytic converters in the 1975 model
year.
     *The 1970 revision of the Clean Air Act also
authorized a one-year suspension of the 1976 standard
for NOX emissions of .40 g/mi on the same terms that
governed a suspension of the 197.5 standards.  On July 30,
1973 Acting Administrator Fri found that NOX control
technology at that time was too undeveloped to support
a prediction that the standard could be met by 1976 and
granted a suspension.

-------
                         - 4 -
     Under the 1974 amendments to the statute, a request
for suspension of the 1977 emission standards may be filed
any time after January 1, 1975.  Knowing that the major
domestic auto companies intended to file suspension appli-
cations, I urged them to file promptly, so that the required
hearing could be held early enough, and could be expanded
in scope by use of EPA's authority under §202(b) (4),* to
provide a useful record in connection with proposals for
further legislative revision of the auto emission standards.
Both Ford and Chrysler filed thc;ir applications on Monday,
January 2, while General Motors filed about a week later.
Well over three weeks of public hearings have been held on
these three applications, and testimony has been received
from a wide range of witnesses, including domestic and
foreign auto companies, manufacturers of catalysts, carburetors,
and fuel injection systems, elected officials, and repre-
sentatives of other Federal agencies and of environmental
interest groups.

II.  Summary of Decision

     I have decided to grant the request of the applicants
for suspension of the effectiveness of the statutory emission
standards for HC and CO in the 1977 model year, and am
establishing as interim standards for that model year the
current Federal interim standards for those pollutants,
i.e. 1.5 g/mi HC and 15 g/mi CO, together with the statutory
emission standard of 2.0 g/mi NOX for the 1977 model year.
Furthermore, it is my view that consideration should be
given to  (a) extending through the 1979 model year this
same set of emission standards, and  (b) establishing emis-
sion standards of 0.9 g/mi HC and 9.0 g/mi CO for the 1980
     *Section 202(b)(4) of the Clean Air Act requires the
Administrator of EPA to report annually to the Congress
on the state of progress of auto emission control, and
authorizes him to hold hearings and subpoena information
and witnesses for this purpose.

-------
                         - 5 -
and 1981 model years, retaining the 2.0 g/mi NOX standard.
For the model years 1982 and beyond, it is my view that
the original statutory HC and CO standards of 0.41 g/mi
HC and 3.4 g/mi CO should remain our national goal, together
with an emission standard of 2.0 g/mi for NOy or such more
stringent NOX standard as may be warranted by the conclusions
drawn from our ongoing review of the need for a new, short-
term ambient standard for oxides of nitrogen.

     Three major issues have dominated these proceedings.
They are:

     (1)  The progress made in developing, and the status
of, technology to control auto emissions to the levels
called for by the Clean Air Act;

     (2)  The impact on fuel consumption and on the general
state of the economy of increasingly tighter levels of
auto emission controls; and

     (3)  The impact on the public health of automobile
emissions of carbon monoxide, hydrocarbons, and sulfuric
acid.

     These are the first EPA suspension hearings to be held
since the widespread introduction of the catalytic converter
for emission control began in the fall of 1974.  Much of
the inquiry has naturally centered on the past and potential
performance of that particular device, which many of the
auto companies  (all of the domestics) have relied upon
heavily to achieve the 1975 and to approach the statutory
levels of EC/CO control.

     In many ways, catalysts have performed far better than
some predicted when the 1975 interim standards were first
established two years ago.  Contrary to many predictions,
both the production of catalysts and their installation
on automobiles is proceeding without difficulty.  The
President of the National Academy of Sciences has stated
that, as of November 1974 "significant advances have made
catalytic emissions-control systems much more satisfactory
for the control of automotive emissions than we anticipated
in 1973".

-------
                         - 6 -
NAS Kept. p. V.*
     *In this Decision, the following abbreviated citations
      are used:

     NAS Rept.  The Report by the Committee on Motor Vehicle
     Emissions of the National Academy of Sciences dated
     November 1974.

     FE Rept.  "Potential for Motor Vehicle Fuel Economy
     Improvement-Report to the Congress" prepared by EPA
     and the Department of Transportation and'dated
     October 24, 1974.

     Status Rept.  "Automobile Emission Control: The
     Technical Status and Outlook as of December 1974"
     prepared by EPA's Emission Control Technology  .
     Division and dated January 1975.

     C. App.  The Request for Suspension of Chrysler
     Corporation (5 Vols., dated January and February
     1975).

     F. App.  The Application for Suspension of Ford
     Motor Company dated January 1975.

     GM App.  The request for Suspension of General
     Motors Corporation dated January 1975.

     Tr.  The transcript of the hearings held on this
     matter from January 21, 1975 to February 6, 1975.

     Sulfate Tr.  The transcript of the hearings held
     from February 18, 1975 to February 12, 1975 to
     consider sulfate emissions from catalyst-equipped
     vehicles.

     Other submissions are cited by the name or initials
     of the submitting company and the date  submitted,
     e.g., C. 4/18/75 p. 2.

-------
                         - 7 -
     It now appears from preliminary data that the durability
of catalytic emission control installed on production cars
is at least as good as for pre-catalyst emission control
systems.  Both Ford and General Motors have run extensive
vehicle fleets equipped with catalysts under normal highway
driving conditions in California, and have reported highly
satisfactory durability and emissions control results.  NAS
Kept. 122-127; F. App. IV-A.  Indeed, GM testified with some
emphasis that their expectation was that catalyst cars in
the hands of the consumer would show less of an emissions
increase with increasing mileage than their prototypes had
evidenced while passing the EPA certification test.  Tr.
411-17, 512-14.

     The function of the catalyst is to accelerate the rate
at which the exhaust HC and CO gases coming out of the
engine react with oxygen in air to form harmless C02 and
water.  The catalytic material speeds up these reactions,
and allows them to take place at lower temperatures than
would otherwise be required.  Catalysts in use on today's
cars can reduce the hydrocarbons and carbon monoxide emit-
ted from the vehicle tailpipe by 50 to 60% at the end of
50,000 miles of EPA durability testing.  By the 1977 model
year, this reduction may approach 70%.

     This high conversion efficiency has two consequences.
First, it allows the engine to be retuned for better fuel
economy at the expense of emissions coming out of the
engine, since much of the task of cleaning them up can be
left to the catalyst.  This has helped make possible the
average 13.5% fuel economy improvement realized by 1975
cars over comparable 1974 models.  As the NAS has stated:

           [SJince exhaust treatment tends to decouple
          emissions control from the traditional
          engine-design constraints, there may be an
          inherent advantage to this approach,
          especially in the control of emissions from
          engines whose design has evolved without
          emissions constraints.
          NAS Rept. p. 31.

     Second, it appears that the high-efficiency after-
treatment provided by catalytic technology will be needed
in the next few years, and may be needed even in the
longer term, if the statutory emission standards of  .41
g/mi HC and 3.4 g/mi CO are to be attained by a vehicle
that is also capable of providing the improved fuel
economy the nation is now demanding.

-------
                         - 8 -
     At less stringent levels, approaching the California
interim standards of .9 g/mi HC and 9 g/mi CO, the evidence,
though conflicting, indicates that compliance can be
achieved by about 1980 without the use of catalysts and
with equivalent fuel economy to what catalysts would
provide.

     The rapid advance of emission control technology is
more fully described below.  To summarize, work on such
significant modifications of the present engine as
stratified charge or rotary engines has moved during the
past two years at a somewhat slower pace than might have
been anticipated, though industry interest remains high.
This is due both to the failure of some of these systems
to completely bear out their first promise upon detailed
examination, and to the unwillingness of the industry to
commit resources to significant changes in engine con-
figuration before long term emission standards  (particularly
for NOX) are known with greater certainty than at present.
By contrast, a whole range of modifications to the present
engine in the fields of fuel preparation, exhaust treatment,
and electronic control of engine operation are in more or
less advanced stages of development.  These modifications
have the promise of improving both emission control and
fuel economy, although in some cases at some additional
expense, and their rapid progress has decreased the relative
appeal of more drastic engine changes.

     The report of the National Academy of Sciences, given
a fair reading, indicates that the technology to meet the
statutory HC and CO emission standards is available.  In
addition, a detailed Technical Appendix prepared by my own
staff reaches the same result in concluding that models
representing 90-100% of 1975 production will be able to
certify at the statutory standards in 1977.  Since the
NAS testified that it relied largely on somewhat outdated
information that did not take account of recent improve-
ments, Tr. 2380-81, and since the predictive methodology
used in the Technical Appendix is deliberately conservative,
a high degree of confidence can be assigned to these
projections.  Attainment of the statutory emission standards
in the 1977 model year would provide the maximum protection
of the public health from HC and CO emissions from those
vehicles that is achievable with present technology.

     I thus find that catalyst technology exists and could
be applied to meet the statutory HC and CO emission levels
on a very large proportion of automobiles by 1977.  I also

-------
                         - 9 -
find that this would carry with it a reduction in fuel
economy of five to ten percent in 1977 as well as an
increase in first cost which is estimated at $100 to $150.
Although the fuel economy reduction and price increase
would in theory lead to some decrease in sales, the extent
of such adverse sales impact is too slight and uncertain
to be reliably quantified.

     Based on the foregoing factors alone -- which together
with the issue of "good faith" are the only factors con-
sidered in prior suspension decisions -- and considering
the matter in the context of the court's opinion in
International Harvester v. Ruckelshaus, I would deny the
suspension of the statutory emission standards.

     However, the foregoing factors are not the only factors
that must govern my decision in this matter.  This year
there is an unprecedented concern uniquely important to
this Agency that alters the decision I would otherwise
make.  This concern is that the use of the only technology
now available to the auto manufacturers to achieve reduc-
tions in HC/CO emissions to the statutory levels in 1977,
i.e., the catalytic converter, would increase significantly
emissions of another pollutant that can be harmful to
health, namely sulfuric acid.  Though substantial uncertainty
exists as to the exact magnitude of the impact of such an
increase in the sulfuric acid emissions on air quality, the
weight of scientific opinion indicates a legitimate cause
for concern that such emissions represent a risk to public
health.

     My concern  with emissions of unregulated pollutants
from catalysts, and specifically with sulfuric acid, is
not new.  In testimony before the Senate in November of
1973, I stated that EPA had intensified its research
program to define and quantify what at that time was
preliminary data concerning sulfuric acid emissions.  I
committed the Agency to a number of actions including
a) development of a test procedure for "sulfuric acid,"
b) consideration of means of controlling such emissions
and c) estimating the impact of these emissions on air
quality and the public health.  As discussed below, much
work has been successfully done on the first two of these
programs and while the public health question remains
unquantified, certain conclusions can now be reached.

-------
                         - 10 -
     More specifically,  data accumulated in the past year
and a half leave no doubt that catalyst-equipped cars
emit greater quantities  of sulfates - primarily small drop-
lets or particles of sulfuric acid and some other sulfur
compounds - than do non-catalyst equipped cars.  Wide
ranges of uncertainty exist both as to the actual health
effects of sulfuric acid emissions and the degree to
which catalyst equipped  cars do in fact contribute to
increased atmospheric sulfate loadings, even on a local
basis.  There is, however, a reasonable consensus that
sulfuric acid emissions  from cars pose a risk to public
health even though the extent of that risk cannot yet be
quantified.  An EPA staff paper, reproduced at Tr. 2256-96,
suggests that as additional model years of catalyst-
equipped cars are introduced into service, there is an
increasing risk that adverse health effects from sulfuric
acid will be observed, particularly in sensitive popula-
tions.  At some point, catalysts might begin to do more
harm by creating sulfuric acid than good through additional
control of hydrocarbons  and carbon monoxide.

     Though the EPA staff paper may tend to overstate
the imminence of the problem, the witnesses at our hearing
generally agreed that it provides as good an estimate
of the problem as can be made at this time given the
tremendous uncertainties which are inevitable in any
estimate of this matter.  Future studies of this issue
are necessary and are being undertaken.  However, their
results cannot be expected to provide definitive informa-
tion on the health questions involved for at least
two more years.

     This suspension decision will not wait for those
studies.  The concrete dilemma it poses is this:


     On the one hand, by moving to very stringent
emission standards such as the statutory or California
interim levels, we can ensure that proven technology will
be applied to achieve the maximum degree of public health
protection against damage from HC and CO emissions that is
currently achievable in the 1977 model year.  Howeverr at

-------
                         - 11 -
these low levels of HC and CO emissions, catalysts are
sure to be used on almost all vehicles, and used
together with air pumps; air pumps significally increase
sulfuric acid emissions from a catalyst-equipped vehicle,
in the range of 50 to 100% over a catalyst operating
without an air pump.  A vehicle air pump is used to in-
ject extra air from the atmosphere directly into the
exhaust system ahead of the catalyst.  The extra oxygen
so provided allows the catalyst to do a more complete
job of converting HC and CO to C02 and water.  However,
the extra oxygen also encourages the conversion of
sulfur dioxide (S02) in the exhaust gas to the sulfate
compound S03, which then reacts with water available in
the exhaust to form sulfuric acid.

     On the other hand, by staying at the national interim
levels of 1.5 g/mi of HC and 15 g/mi CO, we would accept
emission reductions for these two pollutants that are
well short of the best that technology can achieve, but
which are still stringent enough to result in continued
reduction in total HC and CO emissions.  The potential
of a sulfuric acid problem would be substantially
lessened by such a step.  Many vehicles will be capable
of meeting those standards in ]977 without catalysts,
and most of those that will use catalysts will be able
to do so without the use of an air pump.


     With reluctance, and with full awareness that I
may be erring on the side of caution, I have for these
reasons decided to choose the second alternative, and
to continue the 1975 national interim standards for
the 1977 model year.

-------
                         - 12 -

     I believe that this decision is fully consistent with
the statutory mandate and with the opinion of the Court of
Appeals in the International Harvester case.  The Clean Air
Act requires me to find that "effective technology" to
achieve the standards is lacking before I grant a suspension.
I do not believe that if the evidence were plain that sulfate
emissions were an imminent danger, anyone would argue that I
could not consider that fact in making my judgment here.
Certainly Congress could not have intended that "effective"
would describe a technology that did more harm to public
health in one aspect than it prevented in another.

     Of course, the evidence is not clear.  Nevertheless, in
giving high weight to an uncertain risk of severe adverse
consequences, whether those might take the form of immediate
danger to health or of commitment to a regulatory course that
might later have to be changed precipitously, I believe I
am responsive to the court's concern that any decision on
the statutory criteria "take into account the nature and
consequences of risk of error", 478 F. 2d 643.

     Nor do I believe that a contrary decision is compelled
by the NAS Report.  Though I do not question its conclusion
that the technology to meet the current certification
standards for HC and CO and to acheive substantial compliance
in use is in fact available, the Committee on Motor Vehicle
Emissions which produced the report did not consider the
question of sulfuric acid in any depth.  Tr. 2388-90.  The
Court of Appeals has made clear that I am not bound by the
NAS conclusions "as to matters interlaced with policy and
legal aspects", or to the extent that I may rely on "later-
acquired research and experience".  478 F. 2d 649.

     My decision does not rest on a judgment that the problem
of sulfates is more severe than the problem of currently
regulated pollutants.  The data are much too fragmentary to
support any such prediction.  Rather, this decision rests on
a judgment about risks and the weight to be assigned them.
Since most persons addressing the point agreed that there is
a potential sulfuric acid problem and that its extent is
surrounded by uncertainties of all sorts, the possibility
that it may prove worse than anticipated cannot be dismissed.

     It is almost a rule of life that problems are easier to
prevent than to cure, and easier to cure the earlier they are
detected.  The decision I have reached will significantly slow
down the growth of the potential sulfuric acid problem and
give all involved some needed time to assess its  true magnitude
and settle on steps to deal with it.  All these steps involve
substantial lead time before they can become effective.  In

-------
                          - 13 -


particular, I expect the automobile companies to devote a
considerable part of their engineering resources to work
on characterizing and coping with sulfuric acid emissio'ns.

     The same considerations that govern my suspension
decision affect my conclusions as to options I believe
should be considered for the longer term.  Here, I have
attempted to reconcile two objectives.  They are preventing
an increase in automotive sulfuric acid emissions to current-
ly projected levels and preserving as much as possible the
momentum we have built up toward further reductions of HC and
CO emissions from the automobile.  Where conflicts between
these two objectives have occurred, I have tended to emphasize
the former because of the concerns that I have expressed
earlier in this document.

     Accordingly, I would suggest that consideration be given
to retaining for the 1978 and 1979 model years the 1975
national interim standards for HC and CO, and to establish-
ment of emission standards at the interim California levels
for these two pollutants for 1980 and 1981.

     As an integral part of this approach, I intend to establish
an emission standard for sulfuric acid emissions from motor
vehicles.  Though a Notice of Proposed Rulemaking will be issued
within two months, both the industry lead times involved and
the inherent complexity of the matter preclude making such a
standard applicable earlier than the 1979 model year.

     The decision as to the level for such a sulfuric acid
standard will be a very difficult one.  Unfortunately, data
are not now available, and probably will still be unavailable
at the time that an emission limit must be established for
sulfuric acid for 1979 model year vehicles, to determine what
level of sulfuric acid emissions could confidently be considered
acceptable from a public health standpoint.  However, the
catalyst has both proven and potential benefits for control of
HC, CO, and NOX together with fuel economy.  To a degree,
these benefits are unequalled by other technologies now known
or foreseen.  Clearly it would not be responsible to stifle
this technology for insubstantial reasons.  But, if catalysts
cannot be used safely despite their benefits, this hard
decision will have to be made.

     The level most completely and certainly protective of
public health from any sulfuric acid risks, in the absence of
health effects data that would permit establishment of an
adverse effects threshold for sulfuric acid, would be at or
near the level of sulfuric acid emitted  from non-catalyst

-------
                          - 14 -
equipped cars, which is estimated at about .001 g/mi.  Such
a level almost certainly could not be met by catalyst-equipped
cars operating on gasoline with currently anticipated sulfur
levels, even with maximum feasible sulfuric acid controls on
the vehicle.

     Maximum control of sulfuric acid emitted from catalyst-
equipped vehicles may permit achievement of sulfuric acid
emission levels in the range of -01 g/mi.  With the addition
of gasoline desulfurization, which could be implemented over
a 3-6 year period, sulfuric acid emissions from catalyst-
equipped vehicles may be further controllable to levels on the
order of .005 g/mi.

     Unless reductions to the range of such levels are
acceptable from a public health standpoint, the most conserva-
tive emission standard that could be set would be at a level
essentially equivalent to that of non-catalyst cars.  This
would almost certainly mean the demise of catalyst technolbgy
for the forseeable future which, in turn, would cast serious
doubt on the ability of the industry to achieve the statutory
HC and CO standards together with improved fuel economy or to
achieve NOX emission levels much below 2.0 g/mi.

     On the other hand, a sulfuric acid standard set at the
lowest level likely to be achievable with catalysts, probably
in the vicinity of .005-.01 g/mi. (depending upon whether
gasoline is desulfurized) clearly will permit at least somewhat
greater sulfuric acid emissions and, therefore, presumptively
a somewhat greater health risk, than the first alternative.

     By remaining at the national interim standards in trie'1977
and 1978 model year, we can ensure than manufacturers will not
have any significant incentive to make changes that would'in-
crease sulfuric acid emissions before a sulfuric acid emission
standard can be established.  Indeed sulfuric acid emissions
should decrease as new types of emission controls are phased in.

     Imposition of a sulfuric acid emission standard in the
range being considered may make it difficult for some manufac-
turers to simultaneously meet stricter standards for HC and
CO.  Therefore, I believe that HC and CO emission standards
probably should not be tightened in the first year that a
sulfuric acid emission standard becomes effective.  This will
help ensure that the projected sulfuric acid standard can be
met on schedule.

     For this reason, I am recommending that the current
national interim standards be extended through the 1979 model
year as well.

-------
                         - 15 -
     By 1980, it should be possible to move to tighter HC and
CO emission levels with no compromise of sulfuric acid control.
The many new developments in emission control technology dis-
cussed below should be coming into general use by then.  I
believe that they should make it possible to meet the California
interim levels without the use of catalysts should that be
necessary, or to substantially decrease sulfuric acid emissions
from catalyst-equipped cars by reducing the amount of work the
catalyst must do or by some other means such as a sulfate trap
if a suitable one can be developed in time.

     I do not recommend that the current statutory HC and CO
standards be deferred beyond 1981.  Those standards, we now
believe, will probably require the use of an oxidizing catalyst
if acceptable fuel economy is to be maintained.  Though future
developments in catalyst technology, study of the sulfuric
acid problem, or study of sulfuric acid control may well change
our conclusions, we currently believe that desulfurization of
gasoline may be required if the statutory HC and CO emission
levels are to be achieved concurrently with acceptably low
emissions of sulfates.  However, as noted earlier in this
discussion, even substantial gasoline desulfurization, combined
with catalyst modifications, may not produce acceptably low
sulfuric acid emission levels.
     Where emissions of NOx are concerned, I recommend that the
current legislatively established level of 2.0 g/mi. be continued
through 1981.  Preliminary health effects data suggest the need
by the early 1980 's for a nitrate standard or a more stringent
ambient N02 standard than now exists.  Auto emission controls
of NOX beyond 2.0 g/mi., together with tighter stationary source
controls, may well be needed to help meet such an ambient
standard.

     Recent studies have shown, contrary to what was widely
believed in the past, that emission of NOx can be reduced to
2.0 g/mi. from 3.1 g/mi. without significant adverse impact
on fuel economy.  Accordingly, I believe that since a reduction
of greater than one third in emissions of this pollutant can be

-------
                          - 16 -
achieved at little cost, and by the use of current technology,
national policy should require it.*

     I do not believe that any NOX emission standard for the
period beyond 1981 should be established at this time.  The
health data is in a state of flux, and in two years or so we
should have a clearer picture.  However, the possibility that
a much stricter standard  for NOx than is currently in force
may need to be established is a major reason for my desire to
avoid steps that might discourage the further development of
catalytic technology.  There is no currently feasible technology
except the catalyst that is capable of permitting an NOX standard
much below two grains per mile to be met on a full range of
automobiles.

     I believe that this decision and my associated longer-term
conclusions represent a responsible approach to the problems of
auto emission control.  I believe that it is consistent with con-
tinued progress toward cleaner air.

     Nevertheless, it is critically important that the circum-
stances which have led me to conclude that the most prudent
choice is to continue the Federal interim standards for the
near term not be interpreted as signalling any slackening of
our commitment to ongoing efforts to cope with our chronic
     *I should point out, however, that there is a chance that
emissions of sulfuric acid from automobiles may increase if
this alternative is chosen over what would have happened were
the current national standards of 3.1 g/mi. to be extended.
Control of NOX may cause HC emissions to rise slightly.  This
in turn may tempt a manufacturer to obtain needed additional
HC control by feeding it with air from an air pump.  Use of an
air pump could cause sulfuric acid emissions from the vehicle
to increase by as much as 100%.

      However, technology is available to control NOX to the
2 g/mi. level, and to control any slight increase in HC with-
out the use of an air pump.  I expect manufacturers to act
responsibly and to avoid use of an air pump, even if that
should mean that the car gets 2 or 3% less fuel economy than
it could get with an air pump; however, we are satisfied that
it is possible to achieve adequate control of HC at 2 g/mi.
NOX without fuel penalty.

-------
                          - 17 -
problems of oxidant and CO.*  While some upper bound is set
to this risk by the fact that even the national interim
standards by themselves will result in improving air quality
for several more years, that improvement by itself will not
be nearly enough in many areas to attain the standards.

     In testimony before me, Mayor Goldschmidt of Portland,
Oregon said:

          There is a talent in this country for
          accommodating ourselves to adversity; and
          pretty soon, what we [may] forget is that
          we ever had clean air or that we ever had
          a goal of getting to it.  Tr. 1991.

     We must see to it that such a possibility does not become
a reality, and I will do what I can to minimize the possibility,
In particular, although I believe a pause in the tightening
of auto emission standards is necessary for the reasons I have
outlined, I want to emphasize strongly the need for the auto-
mobile industry to continue, and in fact increase, its efforts
to develop safe methods of achieving further reductions in
auto emissions.

     In addition, we must redouble our efforts to control
other sources of the auto-related pollutants, including both
stationary sources, and vehicles other than passenger cars,
and to move forward with effective programs that will help to
clean up the air, conserve energy, and stimulate mass transit
by reducing our overreliance on the private automobile.

     EPA remains committed to all controls on automotive
pollutants needed to meet air quality standards, including
transportation control measures.  I do not believe that States
and local governments should be penalized by the necessity of
temporarily altering our auto emission control program.  But
there should be no doubt that all reasonable and available
     *Another danger inherent in this approach is that the
auto companies will cease to pursue the new technologies
that now appear so promising once they are no longer needed
for emission control purposes.  The past record of the
industry does not give cause for optimism in this regard.

-------
                          - 18 -
measures that are within their control and needed to meet our
clean air goals should be implemented.

     The actions I plan to take include the following:

     (1) Emissions from a large number of vehicles — those
currently classified as heavy duty — have to date been con-
trolled only slightly.  These vehicles contribute significantly
to urban air pollution, and the proportion of their contribution
has increased as passenger cars have been controlled to current
levels.  I expect very shortly to propose more stringent
emission controls for heavy duty vehicles.

     (2) EPA studies have clearly indicated that hydrocarbon
emissions caused by evaporation from the fuel system of an
automobile (as distinguished from incomplete combustion in the
engine) are far greater than had been previously estimated and
may amount to the equivalent of tailpipe emissions of nearly
two grams per mile.  Both the NAS and my own technical staff
have strongly recommended that a new evaporative emission
standard be imposed.  I shall make every effort to impose such
a new standard by the 1979 model year.

     (3) At current levels of auto emission control, motorcycles
emit substantially more pollution than a new car does.  An
Advance Notice of Proposed Rulemaking to correct that situation
has been issued, 39 Fed. Reg. 2108 (Jan. 17, 1974), and the
final regulations.should be in effect for the 1978 model year.

     (4) The escape of vaporized gasoline when vehicles are
refueled are another major source of hydrocarbons, in amounts
equivalent to about .4 grams per mile.  The NAS has stated that
"reduction in emission from these sources  [evaporation and
refueling losses] must be achieved before reduction of exhaust
HC emissions below the present standard of 1.5 g/mi. will have
a significant effect on total HC emissions from light-duty
vehicles".  NAS Rept. p. 18.  EPA has already issued regulations
on this point, requiring a 90% reduction in such emissions.  I
plan to insure that the standard for vehicle refueling will
require the use of the most effective control devices which are
available or can be developed.

     (5) Finally, EPA will re-examine current regulations pro-
viding  for control of hydrocarbon emission from such things as
paints, solvents, dry-cleaning liquids, and refineries with a
view to tightening them.  Many of these regulations were drafted
some time ago and do not reflect the state of the technology
which can be achieved with additional effort.

-------
                         - 19 -
III.  Discussion

     1.   Technology

     The record of these suspension proceedings suggests that
the technology of automobile emission control has passed its
initial stages and entered on the period of rapid growth and
development that most new technologies enjoy.  More different
approaches to problems are being pursued than in the past.
Witnesses talked more in theoretical terms, relating what
they were doing to the characteristics of the engine, and
seemed to be more confident of their ground.  There seem to
be possibilities of combining the new approaches that are
being worked on in many different ways that have not yet
been explored.  And there was a significantly increased area
of agreement between the engineers on the hearing panel and
their counterparts in industry.

     For these reasons, and also because of the continuing
high degree of public interest at this time in the technology
of emission control and what can be expected for the future,
a general survey of the field is appropriate here.

          a.  The Basics of Emission Control

          The conventional automobile engine, like most other
heat engines in use today, works by burning fuel in air to
release heat.  The energy of combustion causes the burning
mixture of fuel and air to expand, and this expansion is
used to produce mechanical work.

     More precisely, in today's automobile engine air is
taken from the outside, and fuel is metered into it as the
air passes through the carburetor.  This creates a mixture
of air and fuel.  That mixture passes into the intake mani-
fold, which serves as a holding and distribution chamber
from which the individual cylinders can draw it.

     This mixture is then drawn into each individual cylinder
by a down-stroke of the piston, which creates a vacuum, and
is compressed on the next up-stroke.  When the point of
maximum compression is approached, the spark plug is fired
and the mixture ignited.  The expanding gases push the
piston down again, and are then discharged into the exhaust
system by the next up-stroke.

     Since automobile air pollution is the direct product
of combustion in the cylinders, efforts to describe and
control it have looked first to the combustion process and
how it might be modified.

-------
                          FIGURE  I
                 Stoichiometric
                                        Conventional Engine
                                     — Lean Burn Engine
                         AIR/FUEL RATIO
FIGURE III-l  The  Relationship of Typical Engine Emissions and
Performance to Air/Fuel  Ratio.  The Vertical Scale  is Linear
and Shows Relative Rather than Absolute Values  for  Each Param-
eter.

-------
                         - 20 -
     The relationship of primary importance has been the ratio
of air to fuel in the mixture fed to the combustion chambers.
If this mixture contains just the amount of air that is
theoretically needed to completely burn all of the fuel, on
the assumption that all chemical reactions proceed as far as
the materials present allow, the engine is said to be
running "at stoichiometric".  Since chemical reactions in the
real world are not this perfect, there will be incomplete
combustion even in an engine running at stoichiometric.  If
the air-to-fuel ratio is greater than stoichiometric, the
engine is said to be running "lean"; if it is less, the engine
is running "rich".

     In the conventional automobile engine, the relationships
between air/fuel ratios, fuel economy, power, and emissions
of HC, CO, and NOX are well known.  They are set forth in
Figure 1, a reproduction from the NAS Report.

     The first step auto makers generally took to comply with
emission control requirements was to "lean out" the air fuel
ratio by moving from richer mixtures that optimized driveability
to the lean side so as to provide more air than theoretically
required for complete burning.  This generally improved fuel
economy as well as reducing emissions.  Tr. 245, F. App. II-B-
7.

     There is a limit, however, to how far to the lean side
conventional engines can be calibrated and still have enough
gasoline in the mixture to ignite predictably and burn
smoothly.  If such performance cannot be achieved, not only
does driveability suffer; the incomplete combustion causes
emission of hydrocarbons to rise.

     Accordingly, as emission control requirements became
tighter, the manufacturers began to adjust the point in the
combustion cycle at which the spark plug fires in order to
provide some additional emission control.  Ideally, for
maximum efficiency the spark should be fired somewhere before
the piston reaches the top of the compression stroke.  This
ensures that as much of the burn as possible will take place
while the piston is still high in the chamber, so that the
heat energy released by the expanding gases will be available
to the maximum extent for driving the piston on its down
stroke.

     However, having the combustion energy released so early
in the tightest confines of the piston cylinder also means
that combustion will be less complete because of the higher
surface to volume ratio that occurs when the piston is near
the top of its stroke.  The high surface to volume ratio
results in a greater fraction of the hydrocarbon compounds

-------
                         - 21 -
being in contact with the surfaces of the engine during the
combustion process, where they are not burned because of
flame quenching.  Combustion near top dead center also
maximizes expansion of the mixture that occurs thereby
lowering exhaust temperatures and reducing the continued oxi-
dation reaction that occurs in the exhaust system.

     To combat this, the spark can be "retarded" or fired
at a later point in the cycle.  This delay in starting com-
bustion means that less expansion will have occurred when
the piston reaches the bottom of its stroke and the exhaust
gas will be hotter, which in turn allows the reaction by
which hydrocarbons are burned to harmless substances to
continue in the exhaust system.  This results in lower
hydrocarbon emissions from the tailpipe.*

     As far as we know today, spark retard always and unam-
biguously results in a reduction in fuel economy.  For this
reason, the auto companies have begun to eliminate or reduce
its use whenever they can.  However, to the extent that they
are unable or unwilling to apply other methods to meet
a given emission standard  spark retard will be used, and
fuel economy will suffer.  This point has been stressed over
and over again.  G.M. App. I-b-2; F. App. II-B pp. 13-14;
C. App. IV-A-15; Tr. 71, 299.

          b.  The State of the Art

              i.  Non-Catalyst Technology

              The auto companies are pursuing two basic
approaches in their attempt to reduce emissions from the
engine itself.  The first is to control and vary certain engine
functions more precisely under different driving conditions;
the second is to change some of the basic fixed components
of the engine so as to make its baseline emissions lower.
     *Carbon monoxide is not nearly as sensitive to spark
position as hydrocarbons being influenced mainly by the
air/fuel ratio.  G.M. App., Appendix 7, Attachment 1, Fig.
9.  This is because CO burns to harmless substances at a
significantly higher temperature than hydrocarbons, which
means that the extra heat supplied to the exhaust system
by spark retard alone is less effective in controlling it.
See Tr. 748, 1250.

-------
                         - 22 -
     The premise of the first approach is that there is a
wide gap between the variety of different speed, acceleration,
load, temperature, and altitude conditions which an engine
is likely to encounter and the ability of such engine
functions as spark timing and fuel metering to adjust to
them.  In part, this has been the result of ignorance - the
auto companies have just not known what exact engine adjust-
ments would be optimum under any given circumstances.  In
recent years, studies of this matter have picked up consid-
erably.*  G.M. App. Appendix 7, Attachment 1, Ford letter
1/14/75.  Second, even where the proper adjustments are
known, the technology in use on automobiles today is often
just not flexible enough to achieve them.  Efforts to control
engine functions more closely have yielded one major success
so far.  It involves the use of modulated exhaust gas recir-
culation (EGR) to reduce NOX emissions.

     The amount of NOX formed in an engine is dependent on
high temperature, length of combustion, and the amount of
oxygen present, all of which combine to maximize NOX emissions
when the air/fuel ratio is slightly lean.  Accordingly, the
first steps taken to reduce emissions of hydrocarbons and
carbon monoxide also increased NOX emissions.

     To control this, the auto companies began to recirculate
some exhaust gas back to the engine air intake, where it
was inserted and passed through the combustion process once
again.  The use in the combustion chamber of this exhaust
gas rather than air tends to lower the oxygen concentration
and the combustion temperature, and hence to lower NOX emis-
sions.
     *A year and a half ago, at our hearings on suspension
of what was then the 1976 oxides of nitrogen emission
standard, Dr. Hutcheson of the National Academy of Sciences
testified that under the pressure of the emission standards
in the Clean Air Act

          Tne automotive industry, in my opinion,
          has in the last three years or so
          learned more about the engine in the
          automobile that they make than they
          ever knew before.  1976 Suspension
          Hearings transcript, p. 1299.

This knowledge appears to be growing as the pressure is
kept up.

-------
                         - 23 -
     For several years it was thought that the addition of
an inert substance to the combustion chamber would inevitably
deteriorate the quality of the combustion and harm fuel
economy.  Work by General Motors has now demonstrated that
this is not the case, and that in fact moderate amounts of
EGR can improve fuel economy.  Gumbleton, Bolton and Lang,
"Optimizing Engine Parameters with Exhaust Gas Recirculation",
SAE* Paper 740104, reproduced at G.M. App.,  Appendix 7,
Attachment 1.  Besides changing the ratio of specific heats
of the mixture, EGR tends to improve fuel economy for two
other reasons:

     (i)  Because the EGR dilutes the "fresh" mixture,
          less throttling is required at a given speed
          and load point.  This reduces the pumping
          losses that make the conventional engine
          inherently less efficient than the diesel.

     (ii) EGR provides "mechanical octane" allowing
          more spark advance without knock than would
          otherwise be the case.

See Tr. 836; G.M. App. Appendix 7, p. 7.

     The GM work also indicates that a higher percentage of
EGR in the air stream can be tolerated without adverse
effects at higher engine loadings than at low ones.  Since
NOX emissions also increase with engine loadings, this
increasing EGR tolerance would allow more EGR to be provided
just when it was needed.  See Tr. 870  (Ford).

     However, the EGR systems initially in general use did
just the opposite — they provided the maximum percent of
EGR at low loads, and decreased it as the loads got higher.
For some time the defects of this system in causing decreased
fuel economy and driveability were attributed to EGR generally.

     Since then, EGR systems that provide a more constant
percentage of EGR with increasing engine load have been
developed and are in use on some 1975 models.  They are a
major reason, along with catalysts, for the  increased fuel
economy of the 1975 models, though not all 1975 cars use
them.  To this extent, then, the theoretical studies of
engine performance initiated by the need to control emissions
have paid off.  Tr. 458-59, 461  (GM).
     *The Society of Automotive Engineers  (SAE) is the
established forum of the auto industry for the presentation
and exchange of technical papers.

-------
                          - 24 -


     But the final step to providing an increased proportion
of EGR at higher loadings cannot be taken until new hardware
is developed.  Chrysler and Ford are both developing a system
that would accomplish this by controlling EGR metering
electronically.  Ford hopes to have this system ready for the
1978 model year.  C. App. III-A-11.  Tr. 168; F. App. VI-J
pp. 2-3, Tr. 869-70.  Each of the big three auto companies
is also working on electronic control of major engine
variables.  GM App. Appendix 15, pp. 11-12, F. App. VI-E
p. 2, Tr. 39-40 (Chrysler).  See also Tr. 3138, 3146-48,
(Bendix testimony that such a system can be ready by the 1978
model year).

     The two engine functions aside from exhaust gas recir-
culation that would probably be controlled by an electronic
system are spark timing and air/fuel ratio.  Though ideally
all three should be controlled at once so that their total
operation can be put together in the most efficient way,
control of any single function by itself would also be a
significant advance.

     It appears that electronic control of spark timing could
definitely be installed on a large number of 1977 vehicles.
GM plans to introduce it by then, while Chrysler is aiming
at 1976.  Tr. 75-76; 186-87; 576-84; GM App. 3-a-5.

     The time in the piston cycle at which current spark
systems fire is generally governed by simple mechanical or
electrical linkages to such other engine functions as engine
speed and manifold vacuum.  Electronic control would allow a
wider range of variable to be sensed, and would allow the
timing of the spark response to be more flexibly programmed.
Status Report p. 3-15; GM App. Appendix 8; C. App. II-C & D-6;
Tr. 576-78  (GM).

     Electronic control of the air-fuel ratio seems farther
off, even though the principle to be used is well understood.
Volvo plans to introduce a form of such control on its 1978
models.  Tr. 1337-8.

     To control air/fuel ratio electronically, a sensor
would be placed in the exhaust stream to monitor its com-
position and feed any necessary adjustment signals back to
the fuel metering system.  Tr. 3289  (Chrysler),  ("ultimate
solution" to control problems); Tr.  3518  (Ford),  ("we
desperately would like to have a control system that would
allow us to peg air/fuel ratio").  Bosch has developed a
sensor that is capable of "pegging" the air/fuel ratio at
stoichiometric and will last for 15,000 miles, NAS Rpt. p.
61; Tr. 2988-90, and it is this sensor that Volvo plans to
use.  Tr. 1379  (See also Tr. 1054-55  (VW), 1843  (Engelhard).

-------
                         - 25 -


     The domestic manufacturers,  however,  want to calibrate
at a leaner ratio for better fuel economy, and sensors that
will work in that range are still in the early development
stages.  F. App. VI-M; Tr. 3133-36;  3303,  3307, 3511.  No
witness would predict when they might be ready for general
use. *

     The industry, in addition to these efforts to control
engine functions more closely, is attempting to reduce the
baseline emission characteristics of the present engine by
changing some of its components.   Efforts here are centered
in two areas:  Reducing emissions during the first minute or
so of th'e Federal emissions test, and enabling the engine
to run farther in the lean range without adverse consequences.
     *Air/fuel ratio could probably not be controlled as
tightly by such a system in a car equipped with a conven-
tional' carburetor as in one using fuel injection, in which
the fuel is sprayed directly into each intake port through
a nozzle.

     In a carbureted vehicle the air/fuel mixture must
pass through the carburetor and be distributed to the intake
ports after the fuel has been metered in.  During this
period the air/fuel ratio may be disturbed by settling of
fuel on the manifold walls and similar phenomena.  In
addition, the time that the mixture takes to make this
journey delays the response of the total system to feed-
back signals.  Tr. 1504-05; 3014-15.

     With current control systems, these differences are
probably not important, NAS Rpt. p. 48, G.M. App. 3-a-6; Tr.
887, but with more sophisticated approaches such as a "three-
way" catalyst, a switch to fuel injection or some other
alternative to the conventional carburetor would most
likely be necessary.  G.M. App. Appendix 15 p. 3.  Though
fuel injection technology is fully developed, no domestic
manufacturer has any plans to make this switch in quantity,
and it probably could not be completed until several years
into the 1980's.  Tr. 882-8; 3024-28, 3031.

-------
                         - 26 -
     When an engine starts "cold" after having been shut off
for some time, the low temperature makes it hard to mix the
fuel with the air so as to create a partly vaporized combus-
tible mixture.  To counteract problems in getting the ignition
to catch the choke is used to increase the fuel/air ratio
during this period.  However, the excess fuel is not burned
completely even when ignition is achieved, and so HC and CO
emissions during cold starts tend to be very high.  On
catalyst-equipped cars this problem is magnified, since
these emissions go essentially uncontrolled by the catalyst
which is still at too low a temperature to have begun working.

     For these reasons, the auto companies have devoted an
increasing proportion of their development efforts to the
"cold start" with which the official EPA certification test
begins.  Tr. 569-71; 777.  The aim has been to reduce the
need for excess fuel metering, and the time during which it is
applied.  To this end, devices such as chokes which automati-
cally and quickly turn themselves off and pipes to divert
exhaust gases past the wall of the intake system to heat it
up quickly and thus warm the fuel during cold starts have
been developed and installed on many current production
vehicles.

     Two new developments in this area appear to have
immediate promise.  The first is a small electric resistance
heater  (like a hot-plate) which would be installed in or
near the outlet of the carburetor and, when turned on, would
vaporize a small amount of fuel whatever the outside tem-
perature was.  By use of this device, reported by Chrysler,
C. App. IV-A-21, the substantial excess amount of fuel pro-
vided by the choke would be replaced by a smaller amount of
fuel heated and vaporized so as to be more readily burnable.
Chrysler, however, declined to set a target date for intro-
ducing this device, though they did indicate that it would
not happen in the near future.  Tr. 207-10.

     The second development - the Dresser carburetor - is
far more sweeping and promising.  It would also require
extensive new investment, which may be one reason the auto
companies have been markedly reluctant to explore its
potential.  There is no reason to believe, however, that the
final cost would be dramatically increased over present
carburetors.  Tr. 1459-60; 1558-59; 1509-10.

     In all carburetors there is a narrow passage between
the fuel metering device and the intake manifold through
which the air/fuel mixture passes.  The speed of the mixture

-------
                         - 27 -
at that point is a function of the pressure differential
across the passage, which varies due to changes in the throttle
position.  The Dresser carburetor simply varies the size of
that opening and no throttle is used.  The size of the open-
ing is controlled to a range that ensures that the speed of
the mixture at the throat of the passage is almost always
the speed of sound.  Tr. 1433-36.  The shock wave created
downstream of the throat by the transition from supersonic
to subsonic flow atomizes the fuel droplets and creates a
very fine and even mixture of air and fuel.

     Since the amount of air flowing through the carburetor
can be precisely calculated from the fact that the speed of
sound is a constant, control of fuel metering should be all
that would be necessary to achieve very precise control of
the air/fuel ratio.  Tr. 788, 1437.

     The evidence at the hearing was virtually unanimous that
the Dresser carburetor represents a significant advance over
prior systems, although the auto companies were noticeably
less enthusiastic in this regard than other witnesses.  NAS
Kept. p. 48, Status Kept. pp. 3-11-12, 7-41-45, F. App. VI-
C-10  ("test fixture results of an encouraging nature");
C. App. Vol. IV, p. 152  ("considerable promise" but "unproven");
G.M. App. Appendix 19 p. 9  ("no significant improvement in
exhaust emission control or fuel economy"); Tr. 214-24
(Chrysler)  ("excellent potential"); Tr. 787 (Ford) ("we are
encouraged"); Tr. 1496-97 (Carter Carburetor)   ("a definite
improvement over the present day carburetors"); Tr. 1554-
55  (Holly Carburetor)  ("convinced" it "represents a real
and significant advantage over the types of carburetors that
are employed on cars today"), 2414 (NAS).

     The principle of the Dresser carburetor has been known
since 1970.  Tr. 1451-52.  Yet even Ford, the company which
has worked the most with this device, testified that even
on an optimistic view it would not be possible to put it on
production vehicles before the 1979 model year.  Tr.  786,
795.  Other witnesses concurred.  Tr. 1502  (Carter Carburetor);
1556  (Holly Carburetor).

     The very even mixture which the Dresser carburetor pro-
duces means that ignition during most cold starts should be
achievable without choking and warm-up devices.  In addition,
the improved combustibility of such a mixture should make it
possible to operate further in the lean range without misfires.

     Operating further  in the lean range in this manner would
amount to a change in the characteristics of the engine
described in Figure 1.  With better fuel preparation, operating
lean need no longer cause misfires and an increase in emissions.
Instead, the leaner mixture can make possible more complete
combustion of the fuel, thus reducing HC and CO emissions,
while the reduced combustion temperature also lowers the
formation of NO  .  The  "lean burn engine" much discussed
at the hearings is simply a more or less conventional engine

-------
                         - 28 -
with various new components and adjustments that allow it
to operate leaner.

     Many improvements in use or under development to improve
operation in the lean range affect the combustion process
itself.  High-energy ignition which provides a longer and
hotter spark and eliminates moving parts in the ignition
system is standard equipment on some vehicles today.  Minor
modifications to the piston chamber and head are also being
investigated.*  By far the most public interest here, however,
has centered on stratified charge engines such as the CVCC
now being produced by Honda.

     These engines all work by dividing the fuel/air mixture
in the combustion chamber into two portions, one fairly rich,
and the other quite lean.  The rich charge is ignited first,
and it in turn provides enough energy to ignite even a mix-
ture too lean to be touched off by an ordinary spark plug.
Mixtures which are thus very lean on the average can be
ignited this way.**
     *Yamaha has developed a system of engine modifications
that it claims can achieve the statutory emission standards
through a "lean burn" approach without the use of a catalyst
and at a total cost of fifty dollars.   Though the details of
the system are still proprietary and will not be discussed,
cars equipped with the Yamaha system have been tested at the
EPA Ann Arbor facility and represent a significant engineering
advance in the opinion of my technical staff.  Status Report
7-46-47; Tr. 1401-02; 1426-27.

    **As discussed below, Ford and Chrysler have cancelled
their rotary engine  (Wankel engine) development programs,
and GM has postponed the introduction of its rotary engine
indefinitely, in each case because of inability to achieve
emission standards with anything like acceptable fuel
economy.

     Toyo Kogyo  ("Mazda") which introduced the rotary,
however, seems confident of making the rotary competitive
again by adapting the stratified charge concept to it.
Since the rotary engine works on the same principle as the
piston engine, with the rotor taking the place of the
pistons and certain specified areas along the rotor chamber
wall taking the place of the cylinders, there is no
theoretical reason why this cannot be done.

-------
                           29 -
     Finally, the industry is also working on modifications
to the exhaust system.  The most dramatic of these - the
catalytic converter - is discussed in the next section.
Other modifications short of this, however, have also made
great progress.

     One other approach is simply to insulate the exhaust
system so that heat will be retained in it longer.  This
will make it possible for HC and CO emissions to burn them-
selves to harmless substances for a longer period even
after they leave the engine.  Ford has achieved substantial
results with this approach, but claims that most of the new
developments will not be ready for production by the 1977
model year.  F. App. VI-A-1-3; VI-F-3-5; VI-0; Tr. 782-84.
See also Status Rept. 3-7; C. App. IV-A-22-24; Tr. 212-13.

     As noted on page 11 above, many cars use an air pump
to accelerate the combustion process in the exhaust system
either with or without a catalyst.  It is generally agreed
that in theory maintaining a constant proportion of air
in the exhaust system is the best way to promote combustion,
a condition that air pumps currently in use do not provide.
Both GM and Ford have experimented with "modulated air" to
correct this fault, Ford with encouraging results, GM with
results it labels indifferent.  F. App. VI-L; G.M. App.
Appendix 6 pp. 4-6; see also C. App. Vol. IV p. 154.  Though
the modification to the air pump needed to accomplish this
is simple and cheap, Status Report p. 3-11, no auto company
has announced plans to install it on their 1977 models.

          ii.  Catalyst Technology

          The most significant facts about catalyst technology
have already been set forth - that it has proved highly
effective in controlling emissions, fully capable of mass
production, durable in certification  testing and field
trials, and that it gives promise of being durable in use.

     The National Academy of Sciences has stated that:

          What is really needed is a better understanding
          of the engine characteristics that lead to
          catalyst deterioration.  As these are better
          understood through experience and this infor-
          mation is reflected in better engine-control
          designs, it should be possible to meet much
          lower standards easily with catalyst-equipped
          vehicles.  NAS Rept. p. 41.

-------
                         - 30 -
     Despite this, and even though considerable work was
reported on tighter control of other engine functions,
virtually no work along these lines was reported by the auto
companies.

     Indeed, the only major new development disclosed by auto
companies was aimed at the problem of "cold start" emissions
discussed in the preceding section.  This work involves the
addition of a small second catalyst to the exhaust system
very close to the engine and ahead of the main catalyst.
The small size of the catalyst, combined with its position
close to the source of heat in the engine, means that it
will reach working temperature and begin converting HC and
CO to harmless substances sooner than the main catalyst can.
Once the main catalyst begins functioning, the small catalyst
can be switched out of the exhaust system to preserve its
durability.  Status Report 3-9-10; Tr. 1927-29.  See also
Tr. 170-71; 538-43, 558.

     Both GM and Chrysler have indicated that use of a
"start catalyst" is part of their first choice system for
meeting the 1977 standards.  C. App. IV-A-1; GM App. 4-a-l-
2-3.  Though not much testing has been done, and though the
auto companies were very cautious, it appears that this
device has the potential for both improved emission control
and improved fuel economy at the same time.  Tr. 298, 556-57.

     The catalyst companies appear to be concentrating their
efforts on new catalyst formulations that will retain a
high efficiency for controlling HC and CO for longer periods.
Substantial progress appears to be being made.  Tr. 562-66;
767-69; 1829; 1933; 2103-04; 2212; 2219-20; 2239-40; 2304.

     In at least one area of analysis more general agreement
between EPA and the auto companies on how to assess catalyst
performance seems to have been reached.  Though certain
areas can be pointed to where the functioning of the catalyst
and the engine to which it is attached may influence each
other, Tr. 182, 584-85, 2084-85, the three major auto
companies appear now to have adopted EPA's long-standing
position that these are minor enough to be overlooked, and
that for analytical purposes emissions from any conventional
automobile engine operating without a catalyst can be adjusted
to account for the addition of a catalyst simply by adjusting
for the percentage reduction in emissions which that particular
catalyst has been shown to produce.  G.M. App. 4-a-2, p. 6,
9 F. App. III-D; Tr. 329-31  (Chrysler); Tr. 808  (Ford).
In other words, the catalyst functions as a percentage
reduction device by which the emissions from an engine are
reduced to a fraction of their former value without otherwise
affecting the operation of the engine in any way.

-------
                         - 31 -
     What is more, oxidation catalysts can also be added to
any variety of "alternate engine" as long as it uses unleaded
gasoline, and the same emission reductions will result.
This takes on importance as it begins to appear (see Section
III-1-d below) that some of the most attractive "alternate
engines" from a fuel economy standpoint may need a catalyst
to meet low emission numbers.

          c.  Ability to Achieve the Standards in 1977

          The central question raised by these applications
is whether "effective" control technology is available to
achieve compliance with the statutory HC and CO emission
standards in the 1977 model year.  As I indicated above,
the question of sulfuric acid emissions to my mind prevents
such a determination.

     If it were not for this issue, the determining factors
here would be:

     (i)   Can enough models of vehicles to meet 1977
"basic demand" be certified prior to commencement of 1977
model year production?;

     (ii)  Will these vehicles comply with the other
emission control requirements of the Clean Air Act, such
as assembly-line testing and in-use compliance?; and

     (iii) Can these vehicles be mass-produced in quantity?

     No witness seriously denied that the actual hardware
needed for 1977 compliance could be produced and installed
on vehicles in the time remaining.  Such an argument would
be hard to accept in any event, since the industry has
already demonstrated its capacity to smoothly make the far
greater production shift attendant on widespread introduc-
tion of the catalytic converter during the 1975 model year.

     As for certification, all three applicants admit that
they can certify a number of model lines at the statutory
standards in 1977.  Chrysler and General Motors express
some uncertainty as to whether they can certify all their
models, and only suggest that they probably could not, Tr.
35, 49, 70, 156  (Chrysler); 300, 394,  (GM), while Ford
makes a determined effort to demonstrate by quantitative
analysis that only a small percentage of its vehicles could
qualify.  F. App. III-D, Letter of 1/22/75; Tr. 628-29.

-------
                         - 32 -
     Assessing the validity of these claims is considerably
more difficult than it was at the time of our two hearings
held in 1973.  The level of vehicle testing aimed at
achieving the statutory emission levels has dropped off
considerably since that period, an interesting fact which
argues that either the auto companies are already confident
of their ability to achieve the standards, or that they do
not believe that the standards will be enforced.

     Nevertheless, my technical staff has again estimated
the ability of the industry to certify by use of a quanti-
tative methodoloby.

     The characteristics of this methodology have been the
subject of far-reaching disclosure and comment in the past,
and no detailed recapitulation is necessary here.  However,
two points about it should be made.  The first is that
every promising test result on a given car is adjusted to
take account of the possibility that the same car, tested
again, might not perform so well, and only those readings
which are so low as to give high confidence that the results
of any retest would also be below the standards are used to
predict an ability to certify.  The second is that although
test results obtained from cars that did not use the full
range of emission control systems available were adjusted
to indicate what the result would be if the missing pieces
had been used, this was only done to the extent that the
portions omitted will be available for production in 1977.
Accordingly, no credit was taken for most of the promising
new approaches discussed in the preceding section.

     The results of this approach are that models represent-
ing between ninety and one hundred percent of the sales of
each of the applicants, and of their total sales, are
predicted with high confidence to certify at the statutory
levels in 1977.

     The auto companies also argued, with varying degrees of
specificity, that whatever their ability to achieve certifi-
cation might be, their ability to comply with the recently
proposed EPA assembly-line test regulation, see 39 Fed.
Reg. 45360  (Dec. 31, 1974), would be substantially less/ at
least if any more than half the cars were required to meet
standards.

     That argument has been considered, and an analysis
of it is set forth in the Technical Appendix.  Briefly,
it indicates that assembly line testing may have some
impact on the ability of the manufacturer to certify, but
that tighter quality control in actual production can
tend to offset any such effect.

-------
                         - 33 -
     Predicting the ability of 1977 systems to comply with
those sections of the Clean Air Act aimed at ensuring that
vehicles meet the emission standards in customer use is
more difficult, both because the actual methods of adminis-
trative implementation of these provisions have not been
settled by EPA and because there has been little experience
with catalyst cars in the hands of consumers.  However,
what information there is suggests as noted above that
catalysts are durable in use, and that cars meeting the
1977 emission standards in 1977 could achieve substantial
compliance even with fairly stringent approaches to imple-
menting these provisions.

          d.  Emissions, Fuel Economy, and Alternative Engines

              i.  Emissions and Fuel Economy - The Effect on
                  the Engine

          Every automobile engine, when tuned for maximum
fuel economy, emits pollutants at a certain rate.  For some
engines, this rate can be quite low.  Some diesel engines
meet the 1977 standards without any adjustment for emission
control at all, Tr. 755, and Honda testified that its CVCC
engine delivers the best fuel economy when tuned to the
California standards, not to some higher number.  Tr. 1228.

     The same relationship holds for more conventional engines
and their associated after-treatment devices.  To every
combination of carburetor, intake manifold, combustion
chamber design, ignition system, catalyst, and the like,
there corresponds a given emission level when the engine
is tuned for maximum fuel economy.

     If that system is now required to reduce emissions
below that level, this can be done in two ways.  The engine
can be recalibrated, probably through the use of spark
retard, and fuel economy will suffer.  Alternatively, the
system design can be changed so that its basic emissions are
lower.  There is no inherent reason to believe that such
adjustments will be bad for fuel economy, and in the past
many such adjustments have been good for fuel economy.  Status
Report 3-3, 6-1-2; Tr. 245.

     This point was extensively discussed at the hearings,
and accepted by the auto companies, albeit at times with some
reluctance.  Tr. 75, 169-70, 187, 234, 330  (Chrysler); Tr.
464-65  (GM); Tr. 752-57, 783-84, 814, 819-20  (Ford); Tr.
1038  (VW).

     It is for this reason that the National Academy of
Sciences and my technical staff are in agreement that:

-------
                         - 34 -
          The reduction of emissions insures neither a
          reduction nor an increase in engine efficiency,
          with the citing of examples of both cases
          possible.  NAS Kept.  31; Status Report 2-1;
          2-3; 3-3; Appendix B.

     The testimony at the hearings indicated that the engine-
catalyst systems being used to meet the current national
interim standards could not deliver very much more fuel
economy than they do now even if all emission standards were
completely removed.  Tr. 453 (GM)  ("I think we are probably
near the flat part of the curve"); Tr. 245  (Chrysler)
("[I]f NOX were not a consideration, we have got 4 or 5
percent in fuel economy we could pick up ... [o]ffhand,
I don't think [a relaxation of HC and CO controls from that
level] really would help much").

     At this point, two questions arise.  The first is
whether new advances in technology might change this relation-
ship, and make it possible to gain dramatic increases in
fuel economy through the installation of some new device
that would also increase emissions.

     The answer here is simple.  Absolutely no testimony
concerning any such device was introduced in these proceedings
and in the opinion of my technical staff, there is no reason
to believe that any such device exists.

     The second question is whether technology is being
developed that has the potential for reducing emissions and
improving fuel economy at the same time.  The answer is that
virtually every new approach under investigation has this
potential.

     Electronic control of engine  functions should have the
effect of making greater fuel economy possible at any given
level of emission control, and vice versa.  It would replace
engine adjustments made for no reasons other than the short-
comings of the current control system with adjustments that
have been made for a purpose.  That purpose could be either
fuel economy or emission control or any combination of the
two.

     Indeed, though the necessary  detailed work on engine
relationships has not yet been done, the example of EGR
control discussed above makes it reasonable to expect that
in other areas potential trade-offs can be minimized, and
one factor maximized, at essentially no sacrifice in terms
of the other.

     Efforts to reduce  "cold start" emissions are almost all
good for fuel economy,  since they  generally aim at reducing
the use of excess fuel at this point in the driving cycle.

-------
                        - 35 -
     Engines using the Dresser carburetor and other "lean
burn" systems also appear to have inherently lower emissions
when tuned for maximum fuel economy than do current engines.
This tendency is to be expected, since operating leaner
(i.e. with less fuel in the air) without misfire should make
the same amount of fuel go farther and also reduce pollution
as more complete combustion is achieved.

     Finally, improved catalysts and exhaust treatment devices
will reduce emissions at essentially no cost in fuel economy.

-------
                         - 36 -
          ii.  Fuel Economy and "Alternative Engines"

          Much discussion in recent years has focused on the
ability of engines that differ from the current one in some
respects to achieve low emission levels, and it has often been
said that the ultimate answer to the auto emissions problem
lies in the general adoption of one or more of these "alternative
engines".  With increasing concern about fuel economy, the dis-
cussion has broadened to include the fuel economy benefits that
might be realized at low emission levels through general use in
passenger cars of either the diesel, the Wankel, or the strat-
ified charge engine.

          There can be no doubt that diesel-powered automobiles
would have substantially greater fuel economy than any alterna-
tive foreseeable in the next ten years, or that their HC and
CO emissions are naturally so low that meeting the 1977 emission
standards is no problem at all.  NAS Kept. p. 70, Tr. 898
(Mercedes-Benz).

          However, the domestic auto companies have no firm plans
to introduce diesel automobiles into this country, even though
General Motors, at least, does produce a diesel for sale in
Europe.  Tr. 590-91.  Mercedes-Benz, by contrast, is stepping up
its diesel sales in the United States.  Tr. 901-02; 937.

          Though the domestics all give the alleged inability of
the diesel engine to meet the ultimate statutory NOX emission
standard of  .4 g/mi as a major factor in their decision, it
appears that such other characteristics of the present diesel
engine as low power-to-weight ratio resulting in sluggish per-
formance, starting problems at low temperature, noise, odor and
particulate* emissions, and the modifications that would have to
be made to current engine blocks to enable them to accommodate
the increased stresses of diesel operation are at least of
equal importance.  GM App. 4-a-7, pp. 8-10, F. App. V-c-1-12;
C. App. IV-G-14; Tr. 589-91 (GM); 1142-48  (Nissan).  Though in
all probability many of these problems could be eliminated or
     *There is some concern that the diesel engine, even
though it does not use a catalyst, may have a problem of sulfate
emissions due to the high sulfur content of diesel fuel.
NAS Kept. 144, Tr. 2386-87.

-------
                          - 37 -


reduced with work, very little along these lines was reported.
NAS Kept- 74; Tr. 946-47.

          The Wankel engine is at the opposite extreme.  General
Motors was planning to introduce it in the 1975 model year,
but those plans have been shelved indefinitely due to poor
emission control  and fuel economy performance.  GM App. 4-a-7
pp. 13-14, Attachment 2.  The other two major auto companies
have dropped their rotary engine programs outright.  F. App.
V-l, V-O16; C. App. IV-G-9-10.

          However, Toyo Kogyo, which first introduced the
rotary engine into commercial use in the Mazda and which has
demonstrated that the statutory emission standards can be
achieved with it, plans to improve fuel economy by 50% over
1974 levels and achieve the statutory emission levels at the
same time.  Tr. 1240-41.

          As discussed above, Toyo Kogyo plans to do this by
adapting to the rotary engine the principle of stratified
charge, which probably has received more attention than any
other "alternative engine"approach.

          The bulk of that publicity has focused on the use of
stratified charge in what is otherwise essentially a conventional
piston engine.  One such modified engine - the Honda CVCC - has
been in production in Japan for some time and has just gone on
sale in this country.

          The weight of opinion seems at present to agree that
the Honda engine may not be the most attractive stratified
charge engine from a fuel economy standpoint.  An engine of a.
somewhat different design, of which the Ford PROCO is one
example, seems to surpass it.  However, those other engines will
require a catalyst to meet the 1977 statutory standards.  NAS
Kept. p. 9, FE Rept. pp. 47-48, Tr. 685, 3534-37 (Ford), 2357
(NAS), F. App. V. p. 2.  Honda believes that its system has
good fuel economy potential and based on its past performance
may well be proved right in the end.  Tr. 1212.  See also NAS
Rept. p. 68.

          Whatever alternate engine, if any, might ultimately
be chosen, one thing clear from the hearing record is that
introduction of such engines in quantity into domestic produc-
tion cannot be expected any time soon.  The NAS has estimated
that if the decision to convert to diesels were taken today,
and a high percentage of the domestic machine-tool industry
made available to modify the engine lines so that diesel
engines rather than gasoline engines could be produced on them,
only 12% of American production could be converted to diesels

-------
                           -  38  -
by 1980, and only 17% by 1983.  HAS Kept. p. 114.  There is
no reason to believe, however, that the domestic industry is
close to making a decision to convert any significant amount
of its capacity to diesels.

          Conversion to the Wankel engine would take even
longer, since it has a completely different shape from the
current engine and much more extensive retooling would be
necessary.  As noted, the domestic industry has cut back or
eliminated its Wankel programs.

          As for the stratified charge, the NAS has estimated
that up to 27% of domestic production could be converted to
the use of CVCC engines by 1980 if the decision to convert
were taken today, with the percentage reaching 41% by 1983.
NAS Kept. p. 114.  See also Tr. 1236 (Honda estimates less
than 4 years lead time to begin CVCC production).

          There is no prospect, however, that anything like
that time-table will be met.  In the first place, the domestic
manufacturers have all stated explicitly that they will not
even seriously consider conversion unless the statutory NOX
standard of .4 g/mi. is changed.  They insist that engines
which must operate in the lean range, like the stratified
charge, cannot be used with a NOx catalyst, which requires
rich or stoichiometric conditions, and that use of a NOX
catalyst will be needed to achieve the statutory NOX standard.*

          But even if the NOx emission standard were relaxed,
there is no assurance that the change-over would be made.  Since
the auto companies can achieve both the statutory emission
standards and - eventually - very significant improvements in
fuel economy with the conventional engine plus a catalyst, the
     *There is at least some theoretical doubt about this last
point.  EGR alone can control NOX emissions to very low levels
with more or less constant fuel economy, but after a certain
point only at the price of a steep rise in hydrocarbon emissions

      Improvements in hydrocarbon control techniques could
therefore make possible increased control of NOx wj-thout the
use of a NOX catalyst.  F. App. VI-E-3; Tr. 457-61 (GM); Tr.
3316-21 (Chrysler).  At present, however, it appears that the
required degree of HC control could only be provided by an
oxidizing catalyst.

-------
                         - 39 -


relative attractiveness of the stratified charge approach has
been diminished.  The substantial investment that would be
required to make the change-over is another strike against it.

          Chrysler and General Motors discussed their stratified
charge programs only in the most general and conclusory terms,
and resisted any implication that they were eager to move away
from the present basic engine configuration.  GM App. 4-a-7,
pp. 1-8; Appendix 21, p. 3; C. App. IV-G-10-13, Tr. 53, 319.

          Even Ford, which has the most advanced stratified
charge program among the applicants, declined to say definitely
that it would introduce a stratified charge engine if relief
from the NOX standard were obtained, Tr. 682-86, and estimated
that complete change-over, even if the decision to go ahead
were made, could not be accomplished much before 1990, Tr. 3521-
22.  Chrysler has estimated the first domestically produced
stratified charge engine could not be on the market until
1980.  C. App. IV-E-6.

          iii. The Impact of Engine Characteristics on Total
               Fuel Consumption by the Automobile

          In evaluating all the publicity that has surrounded
the alleged "trade-off" between emission control and fuel
economy, it is important to recognize that the actual operation
of the engine is a relatively small factor in determining the
total amount of our fuel supply consumed by automobiles.

          Historically, growth in the number of automobiles and
the amount each automobile is driven has had far greater impact
here than any decrease in average miles per gallon.

          Page 20 of the EPA-DOT 120 day Report on Fuel Economy
contains a circle graph showing that of the total increase in
automobile fuel consumption between 1950 and 1972, 85% could
be traced to growth in the number of automobiles, 8% to in-
creased use of each individual automobile, and only 7% to a
decrease in miles per gallon achieved by the average car.

          When we focus on that last 7%, it is clear that the
overwhelming influence on miles per gallon is vehicle weight.
NAS Kept. p. 20.  This was generally conceded to be true at the
suspension hearings, Tr. 77 (Chrysler), 424-25  (GM), 2028 (UOP),
and is self-evident from EPA certification test results, which
show that in 1975 vehicles of 3000 pounds and under had fuel
economy far better than vehicles of 5000 pounds and over.

          It is interesting, then, to examine the extent to
which the domestic manufacturers are planning to rely on re-
ductions in the average weight of their model line to achieve

-------
                          - 40 -


the 40% fuel economy improvement goal to which they have
committed.

          Reduction in the weight of the average car sold by
a given manufacturer can be achieved either (i) by selling a
greater proportion of smaller models, such as more compacts
as opposed to intermediates, or (ii) by redesigning a given
model  (such as the intermediate)  to reduce its weight.

          No manufacturer anticipated any future shift of its
production to smaller models in excess of the trend of two to
three percent a year that has been established for some time.
C. App. Vol. IV, p. 160, GM 2/4 letter, Attachment 1; Tr. 79-
87, 248-50; 318; 421, 438, 650-51.  Indeed Chrysler expected
the proportion of cars over 4000 pounds that it would sell to
be markedly greater in 1980 than in 1974.*

          When questioned on this point, all three applicants
replied that they only can sell what the market wants, and
that they could not enforce a shift to smaller cars.  Tr. 260-
61, 270, 273 (Chrysler); 318, 428 (GM); 690-91 (Ford).

          The economic forecasting models on which these sales
predictions were based did not take account of any governmental
action that might increase the price of gasoline.  Tr. 436 (GM);
723 (Ford).

          Two of the three applicants, however, are embarked
on programs to redesign cars in a given category to reduce
their weight.**  GM said it will spend three billion dollars
fdr this purpose between now and 1980.  GM 1/16 submission,
esp. Attachment 4; Tr. 318, 421-30.
     *A lighter car will also have an easier time meeting any
given set of emission standards.  NAS Rept. p. 105, F. App.
Appendix 3-C, Tr. 696-97, 1488.

    **The 40% improvement goal has been distributed unequally
among different manufacturers, with those whose 1974 fuel
economy performance was the worst undertaking to make the
most significant gains.  Under this formula GM has said it will
try to achieve a mileage gain by 1980 of 53%, Chrysler of 35%,
and Ford of 30%.

-------
                         - 41 -
          Ford, though not giving investment predictions, did
indicate that it, too, would be active in this field.  Tr.
647-48.  Chrysler, however, gave no similar indication, and
in fact strongly suggested it had no such intention.  C. App.
Vol. IV, pp. 208-10; Tr. 259, 262.

          Finally, even when weight is ignored, changes in
engine efficiency are only one of the modifications that can
be made to a vehicle to increase its fuel economy.

          The EPA-DOT report lists a number of other measures
such as radial tires, better transmissions, smaller engines
for big cars and better streamlining that could also be
employed.  The relative impact of these measures  (excluding
weight changes) for different car sizes is estimated as
roughly equal to the gains that could be realized by increases
in engine efficiency.  FE Rept. pp. 40, 42.  Since that report
has been criticized for being over-optimistic about engine
efficiency improvements, while its estimates of non-engine
improvements have been endorsed by General Motors, GM 1/16
Letter p. 2, these percentages probably underestimate the
relative importance of non-engine improvements.  But see C.
App. Vol. IV, pp. 145-51.

     e.   Lead in Gasoline

     All gasoline stations over a certain size are currently
required by EPA regulation to have available at least one
grade of unleaded gasoline of at least 91 octane for use by
catalyst equipped cars.  These regulations have been judicially
upheld.  Although unleaded gasoline will be increasingly re-
quired in future years as the number of catalyst-equipped
cars on the road increases, and will be required at at least
the present levels as long as the number of catalyst-equipped
cars on the road does not decline, several parties to the
suspension hearings, including Chrysler, suggested that future
emission control systems ought to be made compatible with the
use of lead in gasoline.*  Tr. 179, 1446, 2833.
     *Lead has to be removed from the gasoline used by catalyst-
equipped cars because otherwise it will poison the catalyst,
reducing or eliminating the catalytic activity.

     Chrysler has suggested that this may not happen if lead
alone is used in the gasoline, instead of in combination with
certain other chemicals as is presently the case.  However,
they appeared to have some doubts as to whether this really
was the case, and practically every other witness disagreed
with the suggestion.  C. App. I-B-34, IV C&D-17-18, Tr. 2847-
51  (Du Pont); But see Tr. 172, 191-2, 367-8 (Chrysler); 479
(GM); 862-3  (Ford); 1665-80  (Exxon Research & Engineering);
1943-44 (Matthey-Bishop).

-------
                          - 42 -
     To understand their argument, some background is neces-
sary.  The ratio between the volume created in the piston
chamber of an engine when the piston is at the end of the
power stroke and the volume created when the piston is at
the point of maximum compression is called the "compression
ratio".  Ideally this ratio should be as high as possible,
since the more tightly the fuel-air mixture is compressed
before the power stroke begins, and the greater the distance
through which that given quantity of fuel/air mixture moves
the piston during the power stroke, the more work is obtained
from the same combustion.

     However, when an air/fuel mixture is compressed in a
cylinder, the compression heats it.  At high compression
ratios, this can cause premature partial or complete detonation
of the mixture, causing "knock".  The measure of a given fuel's
resistance to "knock" is called its "octane".*

     Traditionally, refiners have increased the octane of their
gasoline by adding lead to it, and the use of lead additives
increased greatly with the increase in compression ratios and
octane requirements in the 1950*s and 1960's.  The lead in-
hibits the tendency of the gasoline to spontaneous ignition,
and its mere addition changes a low-octane fuel into a higher-
octane one.  The same octane levels that are reached by the
addition of lead can only be reached without it by use of more
high octane hydrocarbon components which are produced by
additional refining.  More intense refining processes in turn
are expensive and involve some energy loss.
     *Some proportion of production cars generally show an
increase in the average octane of gasoline required at a given
spark setting after they have been in use for some time.  If
higher octane gasoline is not available, the spark must then
be retarded to avoid "knock" with a consequent further loss
in fuel economy.

      Concern has been expressed that 1975 automobiles may be
more susceptible to this "octane requirement increase" than
cars sold in prior model years.  This question has been exten-
sively explored, and, though the data is not all in yet, the
majority of witnesses felt there was no reason to believe
this would be the case.  Tr. 282-289 (Chrysler) ,-604-13. (GM) ;
838-44  (Ford); 972-73  (Mercedes-Benz); 1016-17 (AMC); 1096-99
(VW); 1130-31  (Nissan); 1196-97  (Toyota); 1365-69  (Volvo);
1658-63  (Exxon Research and Engineering); 2835, 2852-61  (Du Pont)

-------
                          - 43 -
     Auto manufacturers anticipated that due to this,  high-
octane unleaded gasoline would not be available when the
catalyst came into general use, and since the 1971 model year
they have designed their cars with relatively low compression
ratios to enable them to run on low-octane unleaded gasoline.
This has entailed a fuel economy penalty estimated by EPA at
5%.  Status Bept. p. 3-5.  See also NAS Rept. p. 15 (6%).

     Those who urge putting lead back in gasoline of course
admit that it entails removing catalysts from cars  (or at
least from new cars) and at least a temporary relaxation in
emission standards.  Their argument is either that the
standards are more stringent than necessary, or that new
engine developments such as the stratified charge will enable
the standards to be met by cars that use leaded fuels.

     What the first argument misses is that even at more
relaxed emission standards, catalysts may allow more fuel
economy to be gained by retuning the engine than is lost by
lowered compression ratio.  This happened in the 1975 model
year, when use of catalysts to meet the national interim
standards enabled the auto industry to achieve levels of fuel
economy even better than pre-controlled cars.  Status Rept.
p. 3-6.  Ford testified that in the 1976 model year it would
be adding more catalysts to its cars aimed at the national
standards in order to improve fuel economy.*  Tr. 837, See
also NAS Rept. pp. 15, 135.

     It is true that many of the emission control developments
discussed above have the potential for achieving emission levels
down to the level of the California standards without catalysts.
But given the sluggish pace at which the industry usually moves,
it would be a mistake to believe that these devices will be
used across the board before the 1980's.  And even then, it is
virtually certain that control with good fuel economy at the
level of the statutory s tandards will still depend on the use
of the catalyst.
     *In fact, it is not at all certain that the full 5% of
lost fuel economy could be regained even if compression ratios
on the current engine could be increased.  An increase in
compression ratio tends to increase HC and NOX emissions, and
the adjustments that would have to be made to cope with this
might well offset at least partially any fuel economy that
might be gained.  NAS Rept. pp. 53, 135, Status Rept. p. 3-3;
Tr. 559-61  (GM); 847-48 (Ford).

-------
                         - 44 -
     What is more,  there is no certainty that the use of
lead in gasoline will be of any benefit after some years
have passed.

     One way to increase the compression ratio possible with
a given engine is to raise the fuel octane.  But another way
to raise compression ratio is to make changes to the engine
itself to alter its combustion characteristics so that in-
creased compression ratios are possible without a change in
octane.  The auto companies are focussing now on items like
changes in the shape of combustion chambers and the measured
use of EGR that can provide such "mechanical octane".  Tr.
836, 838 (Ford); Tr. 1485-86 (Dresser carburetor provides
same "mechanical octane"), F. App. III-B-22.

     To the extent that this work succeeds, the potential
advantage to adding lead will be reduced, since there is a
limit to the increase in compression ratio that will still
yield fuel economy benefits.  Little benefit is achieved
beyond a compression ratio of about 10:1 because of higher
rates of heat loss that occur as the compressed mixture be-
comes hotter and hotter.  The heat loss on compression and
during combustion eventually exceeds the benefits of increased
expansion.

     And if in five years or so, the shift is on to an
alternate engine such as the stratified charge, there will
be no benefit to adding lead to gasoline at all.  The Honda
CVCC, the Ford PROCO, the rotary engine, and the gas turbine
all can give their best performance on fuels of low octane.
WAS Rept. pp. 15, 136; Tr. 1235  (Honda recommends "unleaded
or low lead" gasoline for use with CVCC).

     Quite apart from fuel economy considerations, there are
other merits to keeping the lead out of gasoline.*

     Lead in gasoline tends to settle on engine parts such
as valves,  spark plugs, and cylinder heads, and impede their
operation.   Though other chemicals are added to the gasoline
to counteract this tendency, they are not completely success-
ful, and periodic maintenance to remove these deposits is
necessary for a car running on leaded gasoline.
     *Possible public health benefits from reducing lead
emissions from automobiles are not touched on here because
that matter is before the courts.  EPA has applied for a
rehearing in the case of Ethyl Corporation v. EPA, D.C.
Cir. No. 73-2205, and further comment on the matters at
issue there would be inappropriate in this context.

-------
                          - 45 -
     It is generally accepted that the savings an owner of
a car that runs on unleaded gasoline will realize from not
having to perform this maintenance are about equal to the
extra cost due to reduced gasoline mileage.  NAS Rept. pp.
15, 135.

     There are also emission control benefits from removing
the lead from gasoline, since lead deposits when they build
up inhibit the functioning of the emission control system.
It has been estimated that the reduction in emissions for
a given non-catalyst automobile between running on leaded
gasoline and on unleaded gasoline amount to .75 g/mi of HC
and .75 g/mi of NOX-  NAS Rept. p. 128.  See also pp. 14,
120, 127-28; GM App. 4-a-2-3-4; Tr. 514 (GM).  It is reason-
able to expect that this difference will increase as future
emission control systems become more sophisticated and are
built to tighter tolerances.

-------
                         - 46 -



          f.   The Question of Sulfates

              i.  The Dimensions of the Problem

         "Sulfates" is the chemical name generally given to
a class of sulfur compounds consisting of a sulfur-oxygen
group (S04) bonded together with an atom or molecular group
of varying composition.  Examples are sulfuric acid (H
804), ammonium sulfate (NH3 804), and sodium sulfate,
     A large portion of the sulfur emitted into the atmosphere
as sulfur dioxide (S02) eventually reacts with oxygen to form
one of these sulfate compounds.  Gasoline contains small
quantities of sulfur, which is normally oxidized to sulfur
dioxide (802) in the engine and, after expulsion through
the exhaust, becomes part of the general pool of atmospheric
sulfur being converted to sulfates.

     It is now generally accepted that catalysts in current
use on automobiles speed up this oxidation process (speeding
up oxidation of HC and CO is of course the catalyst's
intended function) and thus cause  the 802 Pr°duced in
the engine to be converted to 803.  The 803, when mixed with
the ample water vapor always present in vehicle exhaust,
becomes sulfuric acid  (f^SO*) an<^ is emitted from the tail-
pipe as fine particles that would readily be inhaled deeply
into the lungs.  Although this chemical reaction of 802 i-n
the catalyst does not appreciably increase the contribution
of the automobile to total atmospheric sulfate loadings,
it can substantially increase local ambient air sulfuric
acid concentrations around roadways.  On non-catalyst
cars, almost all if not all of the 802 produced in the engine
is emitted from the tailpipe as 802» and represents at most
one percent of all 802 emi-tted into the atmosphere from all
sources.

     In assessing the extent of any such potential problem,
three questions must be addressed.  These are

     (i)    What are the sulfate emissions from catalyst-
equipped vehicles?

     (ii)   What is the effect on public health of given
atmospheric sulfate or sulfuric acid levels? and

     (iii)  If given levels of atmospheric sulfates may
have adverse health effects, is it possible that emissions
from catalyst cars may result in atmospheric levels above
that "threshold"?

     Virtually every aspect of the sulfates question is
replete with uncertainty.  There is uncertainty about the
health effects of "sulfates" generically, together with a

-------
                         - 47 -
distinct possibility that health effects may vary according
to the size, chemical composition and acidity of the particu-
lar "sulfate" involved, and whether or not other pollutants
are present.  There is uncertainty as to what total emissions
of sulfates will be from a population of catalyst cars,
since actual real-world vehicle driving patterns (which can
have marked effects on the rates and levels of sulfate
emissions) are not known precisely enough.  Finally, there
is uncertainty about what atmospheric concentrations will
actually result from a given level of sulfate emissions,
and there seems no way to answer this question short of
actual measurements in the field.

     However, there is little question that sulfuric acid
is emitted from catalyst equipped motor vehicles in a
particle   size range that can be inhaled deeply into the
lungs.  There is also no question that the amount of such
emissions is substantially greater than such emissions
from non-catalyst cars.

     The estimates given below should be read in the con-
text of the knowns and unknowns just stated.

     Emissions of ^SO. from catalyst cars vary widely
under different driving conditions.  At low speeds, when
air pressure and air flows in the catalyst are also low,
much of the 803 produced may remain within the catalyst.
When the car accelerates, the 803 is released and high
sulfuric acid readings may be recorded.  Finally at high
sustained speeds this storage and release phenomenon ceases
to be important, and 20% to 35% of the sulfur in the gasoline
will be converted to sulfates.

     This makes estimating the H2S04 emissions of a catalyst
car under any given driving conditions very difficult, since,
under the influence of storage and release, those emissions
in a great many cases will depend in part on how the car has
been driven in the past.  A great deal of EPA's work on the
sulfuric acid problem in the past year has gone into deter-
mining how reliable emission rate estimates can be made in
these circumstances.

     As a result, though the difficulties have by no means
been eliminated, EPA has developed a set of emission factors
to describe the emissions of various types of catalyst cars
in current use, Tr. 2269, which, though concededly subject
to some uncertainty, have been generally accepted.  Sulfates
Tr. 7, 20-25, 172-75, 181, 295-96.

     Concern about the effects of sulfuric acid in the
atmosphere is not new.  However, the concern for sulfates in
general increased when epidemiological studies conducted by
EPA under the acronym CHESS  (Community Health and Environmental

-------
                         - 48 -
Surveillance System)  observed correlations between increased
levels of atmospheric sulfates and several forms of irrita-
tion of the respiratory system thus suggesting cause and
effect relationships.

     Cause-and-effeet relationships, however,  are very hard
to "prove" in epidemiological studies, since their subject
is the population at large in its normal environment, and
there is no possibility of holding other factors constant
as would be done in a laboratory.  Sulfates Tr. 52, 54, 220.
Indeed, where air pollution in particular is concerned the
worst effects may be caused by two or more pollutants acting
together.  Sulfates Tr. 217-18, 230, 241-42.

     At the time these data were generated, the CHESS pro-
gram was new, and the conclusions were accordingly based
primarily on only one year's data.  Sulfates Tr. 45-48.

     In addition, some witnesses pointed to laboratory
studies and to experience with occupational exposures to
sulfuric acid that might be taken as indicating the CHESS
results overstate the potential danger.  Sulfates Tr. 136-
37, 138-39, 560-61, 565-66, 572-74.  These uncertainties
preclude a firm quantitative assessment of the extent of
the sulfates problem at this time.  It will be several
years, at a minimum,  before more definitive data are avail-
able.

     For purposes of judgment in the meanwhile, however, it
is relevant to note that the majority of qualified witnesses
believed that even when all the uncertainties were accounted
for, the CHESS studies gave substantial cause for concern.
Sulfates Tr. 43, 67,  69, 73.

     In many cases this was buttressed by a conclusion that
based on general clinical knowledge and animal experiments*
     *Animal experiments, though they can be done far more
quickly than epidemiologic data on humans can be obtained,
and under far more controlled conditions, by their nature
can rarely be conclusive as to the health impact on
humans.  The possibilities for controlled experiments on
humans are obviously limited.

     In these circumstances judgments as to public health
may often be difficult to "validate" quantitatively,
particularly when the need to take action before a
potential problem grows out of control places some con-
straints on the ability to obtain new data.

-------
                         - 49 -
sulfates and particularly sulfu :ic acid mist in tha particle
sizes generated by automobiles *\?ere the kind of particles
that :T(ight well be expected to cause lung damage.  Sulfates
Tru 41, 48-50, 52, 55-57, 60-61, 64, 223-24, 285, 329-30,
515-16.

     Such witnesses generally did not believe the data were
sufficient to support anything more than a qualitative
judgment about I^SC^ emissions.  In addition, there was a
reasonable consensus that exposure to sulfuric acid in
the atmosphere should be kept as low as possible.

     in this regard the problem of potentially high localized
sulfuric acid levels near roadways from catalyst equipped
motor vehicles is considered to represent a potential health
rJL-ko  This risk could be accentuated by the background
lew:Is of sulfates which are already present from other
sources.

     The final question is when, if ever, sulfuric acid
emissions from catalyst cars would cause localized atmospheric
sulfuric acid readings high enough to reach a level of concern.

     Since there are not enough catalyst equipped cars
currently on the road to provide direct evidence of how tail-
pipe sulfuric acid emissions behave once they have been dis-
charged into the atmosphere, several assumptions are neces-
sary in making any such prediction.  For its model, EPA
assumed that the relationship between tailpipe emissions
and atmospheric concentration and dispersion would be the
same for sulfuric acid as it is for carbon monoxide, a gas.
Since emission levels are known for both CO and sulfates,
and since the atmospheric behavior of CO is also known, the
results in terms of air quality of a given discharge of
sulfuric acid under this model can be easily calculated.

     Criticism was directed at the use of this approach.
Some critics suggested that use of lead, rather than CO, as
the "tracer" material would have been more appropriate.
Sulfates Tr. 6, 253, 272-73.

     Others questioned whether enough attention had been
paid to the observed atmospheric behavior of the "tracer"
pollutants.  Sulfates Tr. 96-100, 124-27.  Many of those
who raised these points, however, conceded they were not
strong ones.  Sulfates Tr. 107, 114-15, 118-122, 190-91,
314-18, 327-28.  Some criticism was also leveled at the use
in the model of emission factors associated with the EPA
fuel economy highway driving cycle, Sulfates Tr. 82-83, 88-
94,. -«uich EPA conceded was a rough assumption, but no
alternative prediction method for ambient concentrations
of sulfuric acid was suggested.

-------
                         - 50 -
     Using this emission model, the EPA staff paper calculates
that, under adverse meteorological conditions, 24-hour
concentrations of ten micrograms per cubic meter might begin
to occur solely as a result of catalyst emissions (i.e.
in addition to preexisting background concentrations) on
and near certain heavily-traveled, multi-lane highways after
four years if the current national interim standards were
retained and no corrective action were taken. Tr. 2259.

     The paper also estimates that such effects might occur
within two years in California where the emissions standards
are more stringent in 1975-76 than in the other 49 states.
That calculation, however, assumed that California's unleaded
gasoline had significantly higher sulfur content than gasoline
elsewhere.  Although this is likely to be true in the future,
it has not been true up to now.  Accordingly, the estimates
stated for California are overly pessimistic in this respect.
Sulfates Tr. 27-32, 78-79, 253, 298-305, 358-59, 607-08, 614.

     Although substantially better quantitative estimates
than those used in the EPA staff paper probably cannot be
developed soon, the wide ranges of uncertainty and the highly
speculative calculations inherent in many parts of the
analysis  (e.g. for dose-response functions) strongly
suggest that only qualitative conclusions can meaningfully
be drawn.  To oversimplify, we cannot define the magnitude
of the H2SC>4 risk created by catalyst vehicles, but we must
conclude that a risk exists.  This risk can be assumed to
increase as increasing numbers of catalyst equipped vehicles
are introduced into service.

-------
                          — 51 —


          ii.  Control Options

          If it is not acceptable over the long term from a
health standpoint to allow any greater levels of H2S04 emissions
than those associated with non-catalyst cars (roughly .001 g/mi),
then it seems almost certain that catalyst cars will fail to
qualify. On the other hand, with vehicle controls and de-
sulfurization of gasoline, catalyst cars may be able to achieve
levels on the order of .005 g/mi.

          The alternative of eventually banning catalysts
(presumably by setting an H2S04 emissions standard that they
could not meet) was endorsed by Chrysler and some others at
the hearing.

          There is no doubt that this would eliminate the
H2S04 problem.  However, attainment of the statutory HC and
CO emissions standards with good fuel economy, and attainment
at all of any NOX emission standard below about one gram per
mile will almost certainly depend on extensive use of the
oxidizing capacity of the catalyst.

          Since the auto companies have been using  (in 1975),
and planning to use (post-1975) the "aftertreatment" capacity
of the catalyst to achieve low pollution levels while retuning
the engine for maximum fuel economy, the question arises as
to how much of a sacrifice in both emission control and fuel
economy improvement would have to be accepted.  Since intro-
duction of any "alternative engine" in any quantity is many
years in the future, roughly 1980, this boils down to the
question of the capacity of the conventional engine, somewhat
modified, to control emissions with high fuel economy and
without the use of a catalyst.

          There was a sharp split of opinion between the three
applicants on this point.  Though all of them conceded that
they could in theory achieve the California interim standards
in the future without the use of a catalyst, Ford and General
Motors testified that several years of lead time would be
needed, and that there would be a substantial sacrifice in
fuel economy in any event.  Tr. 3387-90, 3394  (GM); Tr. 3523-
26 (Ford)  ("We don't want to be committed to catalysts, but
by the same token, we don't want to have catalysts ruled out
as an option"), Tr. 3541-42 (Ford)  ("if we commit to the
President on our fuel economy program, we think the catalyst
route is the way we want to go").  Chrysler, however, claimed
that it could achieve the national interim standards on all
its cars without the use of a catalyst and at no fuel economy
sacrifice by 1978, and that the California standards could be
similarly attained by 1979.  Tr. 3286-88, 3311-12.  See also
Tr. 380.   (Chrysler also said, however, that the catalyst
was a "very good technology" that might be needed to attain
the statutory emission standards with acceptable fuel economy.
Sulfates Tr. p. 202.)

-------
                          - 52 -
          The judgment of my own technical staff is that
Chrysler's position is more representative of the capacity
of the industry as a whole than that of Ford and General
Motors.  The new developments discussed in Section III-l-
b  above seem consistent with this view,,  Chrysler has bo®n
progressive in "engine modifications" and has devoted much
effort to this work.  It is reasonable to expect that once
Ford and GM were to turn their attention to this field, they
would be able to achieve the same results in a shorter time
through the application of their far greater engineering
resources.  (See Tr. 568 (GM) ("borderline" as to whether
equivalent fuel economy to a catalyst could be achieved
without one at emission levels of .9, 9 and 2).

          There is no reason to believe, however, that the
statutory emission levels of .4 g/mi0 HC and 3.4 g/mi. CO
can be achieved any time in the forseeable future with good
fuel economy unless a catalyst is used.

          At present emission levels, catalysts have performed
well.  And, in all likelihood, they will be needed to meet
emission standards more stringent than those in effect today.
I am particularly concerned that we not rule out the use of
the only technology that can provide stringent control of NOy
at a time when new questions are being raised about the health
impact of that pollutant.

          The approach to alleviating the H2S04 problem that
was favored by Ford, General Motors, and several other witnesses,
is to desulfurize gasoline.  All witnesses agreed that the
technology for doing this is fully developed and commercially
available.

          GM, Exxon, and the American Petroleum Institute agree
that this could be done by increasing the price of gasoline
between one and two cents a gallon, and that it would result
in increased refinery energy consumption equal to half of one
percent of the energy in each barrel of crude purchased.
Sulfates Tr. 422-23, 433, 476-77 (2 1/2 cents)? Tr» 1605, 1611-
12, 1644-47, 2129-30, 3384=85.  Other witnesses put the economic
and energy costs significantly higher, and contended thati
dedication of a large fraction of the petroleum industry's
investment capability would be required.  Sulfates Tr,, 356=7,
429-31, 478-79, 594-98.

          Estimates of the lead time required to install de-
sulfurization capacity and bring it on line generally ranged
from three to five years.  Status Kept. p. 5-4, Tr. 1611-12,
2097-98, Sulfates p. 356, 600 (8-10 years).  Even before that,
however, significant reduction in the sulfur content of lead-
free gasoline  (which is the only kind that catalyst cars can
burn) could be achieved by using only low-sulfur feed stocks
for its manufacture.

-------
                          -  53 -


          This stategy defines an approach to the H2SO.
problem which would very significantly reduce it.*  At this
point, however, EPA is not taking administrative action to
bring about this result.  Before desulfurization is required,
we should be more certain than we are now that implementation
will, together with vehicle controls, achieve H2S04 emission
reductions that are acceptable from a public health standpoint.
There are a number of possibilities for adjusting catalyst
systems themselves in order to reduce ^$04 emissions.

          Exxon Research and Engineering has reported that the
amount of sulfuric acid emitted by a catalyst-equipped car
can be cut in half by reducing the amount of excess oxygen in
the exhaust stream, and other witnesses have confirmed this
report, though some disagreement exists over the precise re-
ductions available.  Tr. 1594-95, 1695-96; Sulfates Tr. 159,
297, 349.  Air pumps, which can be used to attain low emission
standards, such as the current California standards, of coarse
supply such excess oxygen in order to enhance the oxidation
of HC and CO.

          My action today in extending the current national
interim standards for 1977 will ensure that use of air pumps
will be minimized in that model year, since those standards
can be achieved without them.** Clearly, I hope by my action
to discourage their use.

          This should hold the line on HoSC^ emissions until
a sulfate emission standard can be established.  Considera-
tions of lead time are such that, as mentioned above, we
cannot project the establishment of such a standard earlier
than the 1979 model year.
     *The technology exists to desulfurize gasoline from its
current level of .03 grams per gallon average to a maximum
of .01 grams per gallon.  Even at these levels, a catalyst
car would emit six to ten times as much H2S04 as an un-
controlled vehicle.

    **There is some possibility that auto makers might
attempt to improve fuel economy even at those levels by
"working the catalyst harder" through addition of an air
pump to gain the additional freedom for engine recalibrations,
Sulfates Tr. 309-10 (Ford) but see Sulfates Tr. 166-67
(Chrysler)  (No incentive).

-------
                          - 54 -


          Sulfuric acid is produced by oxidation in the catalyst
and it follows that the less a given car must rely on the
catalyst to attain emission standards - the less the catalyst is
called upon to oxidize what comes out of the engine - the less
will be the sulfate emissions.  Sulfates Tr. 77, 309.  As
discussed in Section III-1-b above, almost every new improve-
ment in emission control now being developed will reduce
emissions coming out of the engine.  Even if these reductions
are not sufficient to enable a given emission standard to be
met without a catalyst, they will allow reliance on the
catalyst to be decreased, and sulfate emissions will be reduced
correspondingly.  This phenomenon alone, in my judgment, should
make it possible to significantly reduce emissions from catalyst-
equipped vehicles meeting the current California interim
standard by the 1980 model year.

          The prospects are, however, that much more than the
simple effects of engine improvements will be available.  Though
the limited efforts put forth by the industry here make exact
prediction difficult, they give good reason to expect that with
a greater effort much more would be accomplished.

          Three avenues for future work in particular were
explored at the hearing.  These are use of a "three way"
catalyst that operates at or near stoichiometric, use of a
different formula for a conventional catalyst, and use of
"traps" to remove sulfates from the exhaust.

          The catalysts in current use today require an
"oxidizing" (oxygen-rich) atmosphere in which to burn the
hydrocarbons and carbon monoxide.  By contrast, catalysts for
the control of NOX require a "reducing" (oxygen-poor) atmosphere
in which the oxygen atoms can be disassociated from the nitrogen.
For this reason, automobiles targeted at meeting very low
levels of all three pollutants have generally used two catalysts •
one for NOX and then a second for HC and CO.  An air pump to
supply excess oxygen into the exhaust system in between the
two catalysts is used to change the atmosphere from reducing
to oxidizing.

          Some catalysts, however, called "three-way" catalysts,
are capable of reducing all three pollutants to very low levels
when operated at or close to stoichiometric.  Since the air/
fuel range in which these catalysts can achieve this result
is very narrow, they must be operated with a feedback system
to control the air/fuel ratio very precisely.

          Emissions of sulfates tend to depend on the amount of
excess oxygen in the exhaust stream of a catalyst-equipped
vehicle, and since by definition a vehicle operating at
stoichiometric has little excess oxygen in the exhaust, there
is reason to believe that a three-way catalyst operating at

-------
                          - 55 -
stoichiometric would have very low sulfate emissions.  Limited
test data appears to bear this out.  Tr. 1592-94, 1681-84
(Exxon Research & Engineering); 1826-27, see also Tr. 1362.

          Accordingly, the capabilities of three-way catalysts
were extensively explored at the hearing.  The testimony was
almost unanimous that no such catalyst has been developed that
could last more than eight or ten thousand miles, and that no
prediction for improving this performance could be made be-
cause of lack of understanding of the basic chemical mechanisms
involved.  Tr. 2072-75; 2081 (UOP); 2217 (W.R. Grace); 2317
(American Cyanamid); 3399, 3415-16 (GM).  But see Tr. 1841-42,
1906-07  (Engelhard); 2425 (NAS).  In addition, it appears that
at least present versions of the three-way catalyst must con-
tain rare elements such as rhodium which are in very short
supply but catalysts companies believe supplies are adequate..
Tr. 2306.

          Finally, the tight control of air/fuel ratio such a
catalyst requires would mandate the use of exhaust sensors
and fuel injection on any car that used it, since conventional
carburetors do not appear to give the kind of quick response
to the sensor signals that is required.  The testimony was
unanimous that these components, though they represent fully
developed technology, could not be installed on automobiles
in quantity before the 1980 model year.  Tr. 3009-12  (Bosch);
3140-42  (Bendix).

          Some evidence introduced indicated that the durability
problems of the three-way catalyst as far as control of HC
and CO are concerned might be greatly diminished if the engine
were calibrated slightly lean of stoichiometric, a level that
would also result in a considerable sacrifice of NOX control.
Tr. 3402-11.  This evidence is very preliminary at present,
however, and there can be no confidence that such an approach
will provide a solution to even some of the problems of the
three-way catalyst system which are outlined above.  Tr. 3332
(Chrysler).

          It was also suggested that use of a conventional
oxidizing catalyst, rather than a three-way catalyst, at
air/fuel ratios very close to stoichiometric would also
exhibit low sulfate emissions because of the low amount of
excess oxygen present.  Most witnesses agreed, however, that
differences in the ways the two types of catalysts operate
chemically suggested that an oxidation catalyst, unlike a
three-way catalyst, may not be able to provide adequate
control of HC and CO for a long period when operating so close
to stoichiometric.  Tr. 3335-37 (Chrysler); 3396-400, 3438-44
(GM); 3488-92 (Ford).

-------
                         - 56 -
          The difference in the way different catalysts operate
points up another avenue of investigation that in the opinion
of my technical staff holds far greater promise for use on
actual production cars in the next few years.  It may be
possible to formulate a catalyst that would still have high
efficiency for converting HC and CO to harmless materials
while at the same time considerably reducing sulfuric acid
conversion.  Tests of a Matthey Biship platimum-rhodium
catalyst have shown encouraging results of this type.  Thus
some results of this nature have already been demonstrated,
and most witnesses thought it significant and worth more
investigation.  Tr. 1595-98 (Exxon Research & Engineering);
2065 (UOP); 3497-98 (Ford) Sulfates Tr. 251, 259.  One witness
suggested that while the possibility of good results could not
be ruled out, this was not an avenue of research that was likely
to prove helpful in the end.  Tr. 1862-63 (Engelhard).

          Finally, there seems to be no inherent reason why
sulfate traps will not work.  Substances that will remove
sulfates from exhaust gas are well known, and Exxon Research
and Engineering has demonstrated very high control efficiency
with a prototype system.  Tr. 1598-99.  However, much work
remains to be done on making such a device compatible with
the proper running of the rest of the engine system and the
auto companies up to now have hardly touched the problem.  Tr.
344, 466-67, 662, 3341-42.  Given the state of development and
lead times to which the industry generally operates, installa-
tion of traps probably cannot be expected before the 1980
model year.*
     *Use of a trap would not necessarily mean the installation
of a new aftertreatment device on the automobile.  At present,
mufflers employ steel baffles to reduce engine noise.  Since
the substances now under consideration for sulfate traps are
mostly simple chemicals that do not require tight control of
their operating atmospheres, it might well be possible to
package them in-between the steel baffles in the muffler with
likely improvement in acoustical properties.  (Indeed, even
current catalysts have acoustical properties and have made
possible a reduction in muffler size.)

-------
                          - 57 -
     2.   The Report of the National Academy of Sciences

     Under Section 202(b)(5) of the Clean Air Act, I may
only grant a suspension if a study of auto pollution controls
which the statute requires to be made by the National Academy
of Sciences "has not indicated that technology, processes, or
other alternatives are available to meet such standards."

     In its most recent report dated November 22, 1974, the
Academy said:

          The 1977 Federal Emissions Standard of 0.41
          g/mi HC, 3.4 g/mi CO and 2.0 g/mi.  NOX can
          be achieved in certification with present
          technology through improved oxidation catalysts
          and quick warm-up techniques at an average
          increased lifetime cost of about $400 [citations
          omitted].  In spite of inadequate maintenance,
          vehicles on average are expected to meet the
          NOx standard at 50,000 miles and to meet the
          HC and CO standards or exceed them by less than
          50%.  NAS Kept.  p. 1.

     The report also said:

          Cars can be manufactured which, if properly
          maintained,.will meet the 1977 emissions
          standard in actual use for 50,000 miles.
          . . .  It is not yet clear whether the
          performance of specified maintenance will be
          necessary for the fleet average emissions to
          remain below the standards; however, we would
          not expect HC and CO emissions to exceed the
          standards by more than 50% at 50,000 miles
          [even if the maintenance is not performed].
          NAS Kept. p. 13.

     When representatives of the Academy appeared at the
hearing, they explained that the Report had concluded that
a very high percentage of the domestic industry's model
lines would be able to certify at the statutory standards
in 1977, enough to satisfy their conception of "basic
demand."  Tr. 2354; 2394-2404.

     As noted above, however, the Committee did not directly
address the question of sulfates.  Accordingly, though I do
not question its conclusion that the technology to achieve

-------
                          - 58 -
the 1977 standards is currently available/ I have not felt
that conclusion by itself could determine my decision in
this matter.

     3.   Good Faith

     The Act requires that, before I grant an extension of
time to any auto manufacturer, I must find that "all good
faith efforts have been made to meet the  [1977] standards."
Even though industry spending on emission controls has stayed
at a high level, there might have been some difficulty in
making that finding if the current level of control efforts
had not provedsusscesful.

     Industry emission control expenditures, though down
from the peak levels of 1973 and 1974, are projected to re-
main high.  GM testified that its expenditures here in 1975,
1976, and 1977 would average about $200 million per year, down
from the peaks of $310 and $450 million reached in 1973 and
1974 respectively.  Almost all the drop is related to the
high costs of tooling up to produce the catalytic converter
in 1975.  Research and engineering expenses are projected to
remain at the level of $140 million a year throughout this
period.  GM App. Appendix 21.

     Ford projects emissions control research expenditures
of about $175 million per year for 1975 and 1976, down only
slightly from the peak of $200 million in the preceding two
years.  F. App. VIII-B.

     Chrysler historically has spent far  less on auto emis-
sion controls than the other two major companies.  It did
not provide figures beyond 1975, stating  that its uncertain
financial situation made such estimates impossible.  Though
the estimated figures for 1975 show a substantial drop from
1974 levels, Chrysler explained that other budgets had been
even more significantly curtailed.  C. App. Vol. IV p. 205;
Tr. 255-56.

     By the standards of past suspension  decisions, these
expenditures would be taken as sufficient to satisfy the
"good faith" test in its financial aspect.  What is more
disturbing is the significant decline since 1973 in emission
testing of vehicles  (except for testing by Ford) aimed at
meeting the statutory HC and CO emission  standards.  This
drop-off is clear from the face of the Technical Appendix.

-------
                          - 59 -
In the words of the Status Report (p. 6-2):

          [M]ost manufacturers, in the opinion of the
          report team, are working on improved oxida-
          tion catalyst systems at a relatively low
          level currently.  One possible reason for this
          could be the current reduced level of employ-
          ment in the industry in general .... Another
          reason could be that the industry is waiting to
          see what will happen during the upcoming EPA
          Suspension hearings, and also waiting to see
          how their proposals to the Congress and the
          Administration for a moratorium on future
          emission standards are received.

     There might be some difficulty in making a finding of
good faith in the face of such a testing effort but for one
factor.  The industry, both the NAS and my technical staff
agree, has developed the technology to attain the statutory
HC and CO standards in the 1977 model year.   Since there is
no requirement that a company spend more than is needed to
meet the standards, the success of the auto industry here
warrants a finding of "good faith by definition."

     4.   The Public Interest

     The compelling reasons which have caused me to find
that the public interest requires a suspension of the 1977
standards have already been discussed.  Technology that would
increase emission of H2S04, with all the uncertainty and
concern that surrounds its health effects,  cannot be con-
sidered "effective" within the meaning of the Clean Air Act
and, until the exact outlines of the problem are clarified,
it is not in the public interest to compel its application.

     Several other items that might have been discussed
under this heading, such as the potential for use of "alter-
native engines" have been addressed in earlier sections.
The major questions that remain to be considered here are
the impact of achieving the statutory standards in 1977 on
auto cost and fuel consumption and on the economy generally,
and the impact of failing to achieve them on air quality.

          a.   The Impact of Various Standards on Fuel Economy

          Estimates on the cost in fuel economy of achieving
the 1977 statutory emission standards on schedule fell within
a relatively narrow range.

          GM calculated a 16% loss from 1975 levels, but this
was based on comparing fully optimized 1975 production

-------
                          - 60 -
vehicles with 1977 prototypes.   GM App.  4a-l,  Figs.  7 & 8;
I-b-2.  Ford set the loss at 5% from 1975 levels,  or 24%
below what could otherwise be achieved if the HC and CO
standards were not tightened.  F. App. I-c-3,  II-B-2-5,
Appendix 6 page 6.  Chrysler's  figures were a 9% reduction
from 1975.  They estimated that the 1975 levels could be
preserved if the current national standards were extended.
C. App. IV-C&D-5.  Volkswagen testified that achieving the
standards would entail "little  or no" fuel economy loss,
Tr. 1039, 1070-71.

          Given the natural temptation on the manufacturer's
part to err on the conservative side, I believe these
estimates are fully consistent with those of the National
Academy of Sciences that a 5% loss from 1975 levels would
probably result, Tr. 2355, 2381, of the Department of
Transportation that the loss would be 10%, Tr. 2511, and of
my own staff that the loss would be between those two figures
unless new technology is introduced at a faster rate than the
industry Is currently planning for it.

          The impact of meeting interim standards set at the
current California levels would be markedly less.   GM set
the loss at 10% from current 1975 levels if the standards
were attained by "application of current technology used on
1975 California vehicles to all 1977 vehicles", GM App. I-c-
2.  It hardly seems necessary to assume that the 1975 control
technology will remain without significant improvement two
model years later.  Ford estimated that the loss would be 6
to 8 percent over what could have been achieved if the current
national HC and CO standards were extended, or a gain of 14%
over 1975 levels.  F. App. II-B-5, Appendix 6 p. 6.  Chrysler
estimated the loss from 1975 at 5%, C. App. IV-C&D-5, even
though the average industry-wide difference between the two
standards in 1975 was only 6%,  Status Rept. 3-7, and Chrysler
would have had the two intervening years to work on fuel
economy improvements.

          Finally, if the interim California levels were to be
extended  for 1977, but with a NOX standard of 3.1 g/mi  (which
would require legislative action), the loss was estimated by
each of the major auto companies at  3 - 5% over continuation
of present standards.  Tr. 47, 240  (Chrysler); Tr.  314  (GM) ;
Tr. 633-35, 739  (Ford).

          b.   The Economic Impact of Various Standards

          There was  somewhat more variation in estimates of
auto price increases that might accompany attainment of the
statutory emission standards than was the case for  fuel
economy.  Still,  the estimates were all in the same basic

-------
                          - 61 -
area with the auto companies once again tending to the
pessimistic side.

          General Motors estimated that a sticker price in-
crease of $130-50 over present systems would result if a
"warm-up" converter were used, and $35-50 (with possible
upward adjustment for catalyst change) if it were not.  These
figures were in 1975 dollars.  GM App. l-b-2, 4-b-4.

          Ford's estimate is likewise about $150, mostly for
increased size and precious metal loading of the catalytic
converter, though this was inflated by factoring in an
inflation adjustment for 1977 which was estimated at 15%.
F. App. I-C-2; Appendix 6 p. 7; Tr. 831-32.

          Chrysler was significantly higher at $260,  C. App.
IV-K-8, a figure adjusted to account for the effect of an
estimated 20% total inflation by 1977, Tr. 152.

          These figures are at the high end of the range
estimated by my technical staff, see Status Kept. p.  4-4
and significantly more than the estimate of the National
Academy of Sciences, NAS Rept. p. 89.  The National Academy,
however, did not attempt to account for the use of a "start
catalyst."

          There is even more agreement as to the impact of
emission standards set at the current California levels..
GM estimates the sticker price increase at about $23 over
that of the current national system, GM App. I-C-2.  That
increase was set by Ford at $50, (presumably once again with
an adjustment for inflation), F. App. I-C-2, by the NAS at
about $40, see NAS Rept.,p. 89, and by my own technical staff
at about $40, Status Rept. p. 4-4.  Chrysler did not provide
an estimate.

          As for the suggestion that the standards be set at
.9 g/rai. HC, 9 g/mi. CO, and 3.1 g/mi. NOX, the resulting
price increase was described by Chrysler as "nominal, an
additional $10", Tr. 49, and by General Motors as $20 to $25,
Tr. 313.

          Both price increases and reductions in fuel economy
that might result from tighter emission standards could in
theory be expected to affect new car sales adversely.  However,
a study by my staff indicates that even at statutory levels
the impact on sales would be on the order of a 1% reduction,

-------
                          - 62 -
          c.   Air Quality Considerations

          The impact of any relaxation of auto emission
standards on air quality is obviously relevant to any
suspension decision, and vital to consideration of any pro-
posals for legislative action here.  At the same time,
neither the composition of the suspension hearing panel nor
the nature of the witnesses made it possible to assess this
matter in the specific context of the hearing.

          Instead, EPA has performed its own, analysis of this
matter.  That study, together with a summary, is separately
available from EPA and will only be touched on here.  "Air
Quality Impact of Alternative Emission Standards for Light
Duty Vehicles" will be available separately from EPA's Office
of Air and Waste Management and as part of the environmental
impact statement being prepared for the President's proposed
energy program, which includes Clean Air Act Amendments.

          The air quality impact analysis suggests that,
although the relative differences in air quality in 1977-
1981 resulting from various possible combinations of standards,
ranging from the current Federal interims to the statutory
standards, do not appear to be enormous and do not reverse the
current trend toward improvement, they are significant.  These
differences are particularly significant when one considers
the number of areas that may still violate the air quality
standards (especially for photochemical oxidants) by 1985 or
1990 at any of the auto emission control levels and the
difficulties of achieving the remaining emission reductions
needed to attain the standards.

          Accordingly, only a new air quality threat of the
potential scope and seriousness of the sulfuric acid problem
could make the decision temporarily to forego further HC and
CO reductions a relatively clear one.
DATED:
                                            Administrator

-------