meet the 1975 standards. The applicants further argue that ruinous legal liabilities could be imposed on them under provisions of the Clean Air Act that force the manufacturer to warrant the catalyst and provide for the recall of models of vehicles when a significant number are found to exceed standards. Some have also sought to raise a fear that the catalyst will pose a danger to the vehicle and its occupants.

Such arguments deserve careful consideration.

The

It is clear to begin with that a catalyst "failure" will neither harm the driver nor damage the vehicle. term is used to describe a situation in which the catalyst for some reason deteriorates and therefore fails to burn the pollutants passing through it. The catalyst then sits inert on the tail pipe of a vehicle which performs in all other respects exactly the same way it did before.

Ford, when questioned on this point, said that the danger it feared from the nationwide installation of catalysts was simply that they would not control pollution as they should, and that Ford Motor Company would be exposed to legal liability in consequence. Tr. 2191-93. General Motors was

even more emphatic. Tr. 2431-2437. Similarly, the National Academy of Sciences testified that in expressing reservations about the use of catalysts it did not mean in any way to imply chat vehicles in which the catalyst failed would not be safe and would not operate properly. Tr. 1605-06. .

The only form of catalyst failure that any manufacturer suggested might be dangerous was melting. This can occur when the catalyst is supplied with an overdose of unburned hydrocarbons or carbon monoxide (caused, for example, by a failed spark plug) which overheats the catalyst due to higher

temperature of combustion going on inside it. However, the only manufacturer of catalyst containers who testified stated that his company was willing to warrant that any such melting failure would not burn through the outside can if his company had supplied it, and that the outside of the can would not ever get dangerously hot. Tr. 1541, 1550-51 (Walker Manufacturing Co.) Similarly, Ford testified that their catalyst containers had an adequate margin of safety against such failures. Tr. 286-87.

In my view such a record is clearly enough to outweigh a few recitals of testing mishaps, Tr. 384, 875, an asserted

94-492 0-73-pt. 1-2

lack of sufficient knowledge by American Motors, Tr. 2363-64, and the perpetual fears of Chrysler. Tr. 2289-93. (Chrysler's expressed fears are contradicted by its own submission, which states: "When ['catastrophic failure'] occurs, there is no indication to the driver of the failure, except that in some cases the vehicle actually drives better and fuel economy may improve." C. App. p. I-34.)

It is difficult if not impossible to determine now what frequency of catalyst failure should be anticipated when catalysts are put into mass production and installed on cars for regular use. A substantial incidence of catalyst "failure" has been experienced by auto manufacturers in various testing programs. Claimed failure rates in the range of 10 to 20 percent have been made and Chrysler says it experienced failure rates "up to 40 percent."

In many cases, however, it appears that the auto companies have attempted to represent any physical damage to the catalyst as a "failure." In fact, a comparison of five melted or cracked catalysts from Riverside West (all claimed as "failures" by Ford, see F. App. Table 4-6) with fourteen unfailed catalysts that is made in the "Failure Analysis" section of the Technical Appendix indicates that the physical damage had no statistically significant effect on catalytic activity.

Chrysler data was not sufficient for such a comparison. It may be noted, however, that the dramatically "failed" catalyst portrayed in C. App. Appendix G. pp. 19-20 was tested after the extensive melting depicted had occurred and found to have conversion efficiencies of 70% for HC and 90% for CO.

The Technical Appendix also gives a detailed breakdown of the number of emissions failures due to engine malfunctions of a type that can be expected not to occur in production cars, and of the number of catalytic failures that appear to have been cured by technical advances (for example the "clamshell" mounting Chrysler has developed) or to have resulted from failure to use the most advanced system (for example, the lack of heat resistant ignition wires in Ford's Riverside West program).

It appears that the test cycles on which Chrysler ran catalysts are designed to overstress engine components so they

will show their weak points quickly, and that in the past vehicles run on these cycles have had component failure rates about ten times higher than in the field. Tr. 368, 418-19, 229, 2301, 2306-07. Many of these failures, moreover, have been associated with engine malfunctions of a type which the manufacturers generally admit will not occur often in normal use. Tr. 76-77, 416. See also Tr. 2959. Future experience with catalyst failure is also likely to be reduced as a result of improvements in heat resistance properties of catalysts, and progress in developing overtemperature protection devices. Consequently, it now appears probable that the overall effectiveness of catalysts installed in production vehicles will be reduced only within relatively narrow limits as a result of catalyst failure.

Overall, catalysts are highly effective pollutioncontrol devices. Even a mediocre catalyst can be expected to destroy 80 percent of the carbon monoxide and about 50 percent of the hydrocarbons that pass through it.

About

NAS

Nor do the costs for the degree of emission control appear excessive. According to estimates in the 1973 NAS Report, with which my staff generally agrees, a 1975 model catalyst equipped car can be expected to cost about $160 more than the emission control system on a 1973 model. $57 of this cost will be accounted for by the catalyst. Report Table 5.2, pps. 90-93. Although additional costs to the consumer will result from the need to use unleaded fuel to avoid catalyst poisoning, unleaded fuel also is expected to create savings in maintenance costs which will be approximately equal to the costs resulting from removal of lead from fuel.

In summary, the development of technology to date, as reflected in the testimony and documents presented in these proceedings, holds promise for meeting the 1975 standards. In particular, catalyst devices now clearly appear to be effective, durable and reasonably inexpensive.

2. Evaluation of Whether Technology is Available to Meet the 1975 Standards.

The initial question raised by these applications is whether "effective" control technology is "available" to achieve compliance with the Federal 1975 standards with respect to 1975 model year vehicles. As previously indicated, a

positive determination of this question must rest upon three separate subsidiary findings, namely:

(a) Enough models of vehicles to meet the 1975 "basic demand" for cars must be certified prior to commencement of production;

(b) It must be feasible to mass produce these cars in sufficient quantity to meet that demand; and

(c) The emissions control systems on these cars must function acceptably in actual use by customers.

(a) Certification

The first question is whether technology has been developed to the point that manufacturers can meet requirements for certification of their 1975 models if tested by the 1975 standards. The certification procedures are based upon tests of prototype and preproduction vehicles. Therefore, examination of the probabilities for certification does not include consideration of any of the problems of mass production. What it does focus upon is the capability of a manufacturer to build a limited number of cars for each model line that it intends to sell which can meet the applicable standards. Since all of the test data is derived from cars which are in essence individually equipped prototypes, the test data bears directly upon this question. Because of the preliminary state of development a year ago, the question of certification was virtually the sole issue seriously discussed at the public hearings last spring.

The methodology used for analysis of test data submitted in these proceedings is discussed in greater detail below. My examination of the fundamental technical issue whether technology is adequate to make it feasible for auto manufacturers to meet the 1975 standards has included extensive analysis of test data utilizing this methodology. It has also included a review of the raw data to evaluate the significance that may properly be attached to test results without making adjustments as required by a system of methodology. It has also included a general review of the overall status of development as reflected in the evaluation of the NAS Report and testimony and other statements of persons having expertise in this field.

On the basis of my examination I find it extremely difficult to predict that enough models of vehicles to meet the 1975 "basic demand" for cars could be certified under the 1975 standards. I find that the 1975 standards can be met by technology utilizing a rotary engine, a stratified charge engine or a light-duty diesel engine. It is clear, however, that a shift over to such technology cannot be accomplished within time to meet more than a fraction of the 1975 basic demand. With respect to conventional internal combustion engines, I find that technology has developed to the point that many models (66 percent of sales) almost certainly would meet certification requirements under the 1975 standards. It is less certain that other models would be able to meet those requirements.

As indicated previously, the Court of Appeals in its decision has directed me to weigh the evidence and make y decision "by taking into account that the risk of an erroneous' denial of suspension outweigh [s] the risk of an 'erroneous' grant of suspension," Dec. p. 58. It cautioned me against holding the "safety valve" of suspension "too rigidly," Dec. p. 44, and advised me that these risk-balancing considerations, though they may seem to speak only to the public interest test," must also be taken into account in determining whether technology is available, Dec. p. 47.

Weighing all of these considerations, I believe that presently available technology is probably effective to achieve compliance with the 1975 standards insofar as the certification requirements are concerned. However, I also believe that there is a significant risk that this determination would prove to be erroneous and that manufacturers would not be able to successfully certify vehicles at the statutory levels in sufficient numbers to meet basic demand for 1975 cars, either in California or throughout the Nation. My decision requiring California cars to meet slightly less stringent standards minimizes these risks without any significant adverse effect on air quality in California and assures that a full line of 1975 cars with catalysts will be certified for California. I believe this decision is in the public interest and is fully consistent with the Court's opinion.

(b) Production

The second basic issue pertinent to my decision in this case is whether it is feasible to produce cars utilizing