Mineral de Chuquicamata 2-8-2 Locomotives in Chile

Class 61 (Locobase 15404)

Data from Baldwin Locomotive Works Specification for Engines as digitized by the DeGolyer Library of Southern Methodist University, Vol 58, p. 320. See Locobase 14000 for more on the CEC. Works numbers were 58312-58323 in March 1925.

These dozen oil-burning Mikado tanks were bought by the Chile Exploration Company for their well-endowed railway. They apparently repeated the dimensions of six Porter side tanks supplied in 1920 (works numbers 6500-6505 in May 1920). Those engines were in turn developments of the 1918 quartet shown in Locobase 20724), which were delivered with 20" x 22" cylinders. Lengthening the cylinder storkes while reducing cylinder diameters increased cylinder volume. That change and an increase of 15 psi in boiler pressure increased starting tractive effort.

The tracks were laid with 80 lb and 90 lb/yard (40 kg and 45 kg/metre) and offered 2.6% grades and 20 degree curves. Like most Chilean locomotives, the Baldwins were built with an eye on operating altitudes of 10,000 feet (3,048 metres) above sea level.

The engines burned Mexican "Puenca" oil, which heated the superheated steam that passed through 10" (254 mm) piston valves.

Class 87 (Locobase 20724)

Data from "2-8-2 Tank Locomotive, Nitrate Railways, Chili", Locomotive Magazine, Volume XXV [25], No 318 (15 February 1919), pp. 19-20. Works numbers were 6134-6137 in July 1918.

LM's report said that side-tank engines were "not so much favoured by trans-Atlantic builders as by those of Europe." As with most Chilean engines, these Mikado tanks were oil-fired.

NB on build dates and works numbers: Virtually every source of builder information supplied to Locobase by Allen Stanley from his extensive Rail Data Collection shows this foursome with 22" x 30" cylinders and 48" drivers. But Al Weber's individual data card for this class shows a whited-out stroke and correction to 22" x 20". Using the contemporary account sourced above, Locobase believes that Weber's correction was itself an error and has accepted the LM report as the most accurate.