Steel is made in various ways. The Acid Bessemer process consists in forcing compressed air in numerous small streams through molten cast iron, in iron vessels (converters) which are lined with ganister, a silicious sandstone. These can be rotated on trunnions. Basic Bessemer steel is made in similar converters by the Thomas-Gilchrist or basic process, which can be applied to pig irons containing phosphorus. The latter is removed by giving the converter a basic lining of calcined magnesium limestone mixed with tar.
In the Martin process steel is obtained by melting together pig iron with steel scrap, wrought iron scrap, &c., on the hearth of a Siemens regenerative furnace with a silicious lining.
In iron smelting the most important danger is from blast furnace gas rich in carbonic oxide. Sulphur dioxide, hydrocyanic acid, and arseniuretted hydrogen gas may possibly be present.
When work was carried out in blast furnaces with open tops the workers engaged in charging ran considerable risk. But as the blast furnace gas is rich in carbonic oxide and has high heating capacity these gases are now always led off and utilised; the charging point is closed by a cup (Parry’s cup and cone charger) and only opened from time to time mechanically, when the workers retire so far from the opening as to be unaffected by the escaping gas. The gas is led away ([fig. 29]) through a side opening into special gas mains, is subjected to a purifying process in order to rid it of flue dust, and then used to heat the blast, fire the boilers, or drive gas engines.
Severe blast furnace gas poisoning, however, does occur in entering the mains for cleaning purposes. Numerous cases of the kind are quoted in the section on Carbonic oxide poisoning.
The gases evolved on tapping and slag running can also act injuriously, and unpleasant emanations be given off in granulating the slag (by receiving the fluid slag in water).
In the puddling process much carbonic oxide is present. Other processes, however, can scarcely give rise to poisoning.
The basic slag produced in the Thomas-Gilchrist process is a valuable manure on account of the phosphorus it contains; it is ground in edge runners, and then reduced to a very fine dust in mills and disintegrators. This dust has a corrosive action already referred to in the chapter on Phosphorus and Artificial Manures.
The poisoning caused by ferro-silicon is of interest. Iron with high proportion of silicon has been made in recent years on a large scale for production of steel. Some 4000 tons of ferro-silicon are annually exported to Great Britain from France and Germany. It is made by melting together iron ore, quartz, coke, and lime (as flux) at very high temperature in electrical furnaces. The coke reduces the quartz and ore to silicon and metal with the production of ferro-silicon. Certain grades, namely those with about 50 per cent. silicon, have the property of decomposing or disintegrating into powder on exposure for any length of time to the air, with production of very poisonous gases containing phosphoretted and arseniuretted hydrogen. The iron and quartz often contain phosphates, which in presence of carbon and at the high temperature of the electrical furnace would no doubt be converted into phosphides combining with the lime to form calcium phosphide; similarly any arsenic present would yield calcium arsenide. These would be decomposed in presence of water and evolve phosphoretted and arseniuretted hydrogen gas. In addition to its poisonous properties it has also given rise to explosions.
[In January 1905 fifty steerage passengers were made seriously ill and eleven of them died. In 1907 five passengers died on a Swedish steamer as the result of poisonous gases given off from ferro-silicon, and more recently five lives were lost on the steamer Aston carrying the material from Antwerp to Grimsby.[C] This accident led to full investigation of the subject by Dr. Copeman, F.R.S., one of the Medical Inspectors of the Local Government Board, Mr. S. R. Bennett, one of H.M. Inspectors of Factories, and Dr. Wilson Hake, Ph.D., F.I.C., in which the conclusions arrived at are summarised as follows: