coiling, applies special fittings, and conducts preventive maintenance programs. The other three primary tank car owner-lessors maintain similar shops at a total of 18 locations.3 Some lessees' plants are equipped to perform light maintenance work, such as tank and safety valve testing, painting, and interior lining, and in many instances lessors arrange to have light repairs performed at these facilities. In addition, some independent "contract shops" are engaged to perform some maintenance and repair work on tank cars.

Table 1 below shows an analysis of empty shop miles operated by GATX tank cars during 4 months of 1966, when approximately 3.5 million shop miles or 3.26 percent of all loaded miles were operated. The selected months were the middle months of each quarter in 1966. Specific data is shown only for railroads showing at least 10,000 empty shop miles. The mileages shown are not the actual mileages traversed, but were drawn from published short line tariff tables.

TABLE 1

EMPTY SHOP MILES AND LOADED MILES FOR RAILROADS PARTICIPATING IN SHOP MOVEMENTS OF GATX TANK CARS

3Texarkana, Ark., El Segundo and Richmond, Cal., Chicago Ridge, Decatur, East St. Louis, and Wood River, Ill., Whiting, Ind., Eldorado, Kans., Baton Rouge, La., North Kansas City, Mo., Laurel, Mont., Portland, Oreg., Milton and Philadelphia, Pa., Longview, Tex., Red House, West Va., and Casper, Wyo.

It should be noted, however, that the "shop miles" category used by GATX includes all empty mileage accumulated by cars moving to and from repair facilities, and thus tends to reflect an inflation of mileage normally attributable to repair movements. For example, if the normal assignment of a tank car had been to move loaded from Chicago to Pittsburg, Pa., and to return empty directly to Chicago for reloading, the statistics of table 1 would count as "shop miles" all empty miles developed by that car both (a) on a movement from Pittsburg (after unloading) to the GATX shop at nearby Sharon, Pa., for routine repairs, and (b) on the return movement to Chicago for reloading, when the repairs were completed.

Protestants suggest that the railroads develop among themselves some form of a pooling arrangement to distribute the burden of empty shop miles more evenly. Table 2 shows

Table 1 demonstrates that those railroads serving General American's eight major repair shops in 1966 (AT&SF, PRR, IC, KCS, SP, MP, LV, NYC, E-L, SCL, SP) accumulated a preponderance of the empty shop miles involved.

4

that some railroads which are major participants in tank car loaded traffic participate in only a minor percentage of empty shop miles. From this comparison, which also shows that these 10 railroads in aggregate reported a substantial excess of loaded over empty tank car miles for their overall 1966 operations, protestants make the point that all respondents should distribute and share the burdens of empty shop miles in a different way from that proposed in the suspended tariff. TABLE 2

RAILROADS WITH LESS THAN AVERAGE PARTICIPATION
IN SHOP MOVEMENTS OF GATX CARS

*Source: Reports by railroads to General American. Reports do not indicate whether empty mileage is actual or shortline. Loaded mileage is presumably short-line.

A number of commodities that move in tank cars are pro duced and marketed primarily within fairly clearly defined peak seasons. To some extent these seasonal movements can be correlated so that a fleet of a particular type of tank car can be used continuously throughout the year for the back-toback seasonal traffic flows. The leading example is that of liquefied petroleum gas and anhydrous ammonia. Both products move in the same kinds of cars--pressure type, heavily contructed with high pressure fillings but no bottom outlet, and with insulation or heat reflective paint. Propane moves in

its greatest volume between September or October and March or April. Anhydrous ammonia moves in the fertilizing season starting in January and extending into May or June. The cars can readily be transferred from one product service to the other, but the production and marketing areas of these products are in different geographical areas. Consequently, the process of transferring the tank cars seasonally between the propane and anhydrous ammonia services generates substantial empty mileage which is not apt to be totally balanced by loaded mileage.

Other comparable combinations of compatible commodities that move in back-to-back seasonal flows include asphalt and road oil on the one hand, and molasses from sugar cane or beets on the other; various types of vegetable oils in combinations with other vegetable oils; and various types of packinghouse products in combination with other packinghouse products.

There are available no precise statistics to show the volume of empty tank car mileage generated each year in transferring cars to meet seasonal traffic demands of the type discussed above. Nor are there any available statistics to show the potential loss of railroad revenues should these seasonal empty car transfers be terminated or curtailed.

General American described several specific examples of the types of seasonal transfers of tank cars that take place. An Illinois shipper of anhydrous ammonia needed 125 tank cars for its shipments between January and May 1968 to the midwest farm belt, especially Illinois, Iowa, Nebraska, North Dakota, South Dakota, Kansas, and Minnesota. General American furnished the cars by moving them empty from California (26 cars), Illinois (42 cars), Iowa (4 cars), Indiana (18 cars), Kansas (15 cars), and Wisconsin (20 cars). Using certain generalized approximations, General American estimates that the empty cars accumulated 85,000 empty miles in reaching the Illinois shipping point. No estimate was made of what proportion of the 85,000 empty miles would otherwise have been accumulated by these cars in returning empty to their last prior points of origin for propane movements. If each of the 125 cars had made one trip a month for 4 months during the January-May season, this block of cars would have accounted for about 500 revenue loads from the Illinois origin. There is, however, no data of record to show that the Illinois origin did originate such a volume of traffic for these cars.

Other GATX examples include anhydrous ammonia cars being forwarded (a) to an Alabama origin from Kansas (21 cars), Texas (3 cars), Louisiana (6 cars), and Tennessee (5 cars); (b) to Bellevue, Iowa, from Kansas (9 cars), Ohio (5 Kentucky (3 cars), Geórgia (2 cars), Michigan (2 cars),

Pennsylvania (2 cars), California (6 cars), and Tennessee
(2 cars); and (c) to a Kentucky origin from California (39
cars); Utah (1 car), Oregon (4 cars), and Texas (1 car).

The transfer of one block of 110 cars accounted for an estimated 83,000 empty miles; another block of 115 cars accounted for an estimated 160,000 empty miles; and a third block of 51 cars accounted for an estimated 38,000 empty miles. As with the example of the Illinois shipper above, there is no way to know how much of this empty mileage would otherwise have been accumulated as empty return mileage to the last prior loading point of the tank cars involved.

The following table shows the geographical distribution of tank cars supplied by General American to meet seasonal shipping requirements for molasses and asphalt during 1968. The molasses season extended generally from late January to March or early April; the asphalt season from March through September.