Blast furnace slag is well known as by-product of hot metal production in integrated steel and iron mills. About 300 million tons of this slag is produced, based on a worldwide production of approximately 1.100 million tons of hot metal per year. State of the art systems cool down the liquid blast furnace slag rapidly by using lots of water. Hence a fine and glassy product is gained, called slag sand, which can be used in the cement industry as clinker substitute due to its latent hydraulic properties. Unfortunately the high energy content of the liquid slag cannot be used if the slag is cooled by water. Additionally the wet slag sand has to be dried for further use requiring even more energy. This thesis describes the development of a dry process which granulates the blast furnace slag by air. This way it is also possible to get a glassy product of high quality. Additionally hot air is gained which can be used for heat recovery applications. Downstream sand drying is not necessary due to the dry method. The development of the experimental plant as well as the accomplishment of granulation tests is shown. The results gained are analyzed which helps design a site based pilot plant in the future.