PRODUCER-GAS TESTS. 37 
minimum, it was deemed necessary to make the test on each coal as long as consistent. A 
schedule was therefore adopted at the beginning of this series of tests involving two sixty- 
hour runs per week. The first eight to twelve hours of each test period are used for getting 
the fuel bed into uniform and efficient condition. During these preliminary hours records 
are taken as in the regular tests, but the official test, as reported, includes only the last 
forty-eight or fifty hours of the run, during which time the conditions are maintained as 
uniform as possible. Special attention is being directed to the very important item of 
measurement of the coal actually used and special charts and checking devices have been 
introduced for this purpose. 
Owing to the lack of reliability in the operation of the gas engine, many of the tests 
conducted during the Exposition period were of a few hours' duration only, but since the 
present series of tests began (May, 1905) no difficulty has been experienced in starting the 
engine at 8 a. m. Monday and continuing day and night without a stop until 8 a. m. Saturday. 
During this period two different coals are tested, and the change of gases is made at 8 p. m. 
Wednesday without stopping the engine. 
NUMBER OF TESTS MADE. 
During the period from May 1, 1905, to January l r 1906, sixty-four producer-gas tests 
have been made. One of these tests was run on coke breeze and one on Brazil coal. The 
remaining sixty-two were divided as follows among different States: California, three tests 
on lignite and locomotive cinders; Illinois, eighteen tests on coal; Indiana, eight tests on 
coal; Kansas, one test on coal; Kentucky, four tests on coal; North Dakota, three tests 
on lignite; Ohio, seven tests on coal; Pennsylvania, seven tests on coal; Virginia, four 
tests on coal; West Virginia, five tests on coal; Wyoming, two tests on coal. 
SPECIAL FEATURES OF THE INVESTIGATIONS. 
The results of the majority of the tests have been exceedingly gratifying, official records 
having been made as low as 0.95 pound of dry coal per hour burned in the producer per 
electrical horsepower developed at the switchboard, or 0.80 pound of dry coal per hour 
burned in the producer per brake horsepower, on the basis of an assumed efficiency of 85 
per cent for generator and belt. 
Throughout the tests a constant effort has been made to do away with unnecessary 
appliances. This effort has furnished valuable and interesting information and has centered 
attention on several radical changes in the details of producer-gas plant construction. 
It was found at an early date that more or less sulphur was passing the purifier and enter- 
ing the engine cylinders. Investigations by the chemists showed that purifiers consisting 
of oxidized iron filings and shavings are fairly efficient for coals containing little sulphur — 
1 per cent or less; but it was found that for coals containing larger percentages of sulphur 
the purifier became completely exhausted after about six or eight hours. Mixtures of 
lime and shavings were tried, but with little success. As a result of these investigations 
the purifier has been discarded, and the gas, carrying its full percentage of sulphur, has 
been charged directly into the engine cylinders. This method of operating has been going 
on for many months, and no ill effects have been discovered, though coal has been used 
containing as high as 8.1 per cent of sulphur. 
One feature of the plant as installed was the economizer, used for preheating the air for 
the blast. A series of experiments has shown no effect on the chemical composition of the 
gas or on the efficiency of the plant when air at ordinary atmospheric temperature was 
substituted for preheated air. As a result the economizer, as an economizer, has been 
discarded and the construction of the plant again simplified. 
Other modifications and changes are under investigation at the present time, the most 
important from an economic standpoint relating to the utilization of slack coal in producers. 
The results of the producer-gas tests will be found in the detailed report on each sample 
presented on pages 53-233. 
