Mathematical modelling is reaching a high acceptance level within the glass industry. Today most new furnaces are being modelled before the final design is decided. It is clear that the modelling helps to optimise the furnace in respect to glass quality, energy efficiency and furnace life time. The extra effort for modelling is leading for sure to a quick payback of this extra investment and increased profit over the furnace life time. Even the furnace life time can be extended with better insight on temperature distribution and glass speeds that corrode the refractory. Many glass produces are always asking us; "what is the optimal glass depth?" There is not just one answer to this, but the paper demonstrates how mathematical modelling can help to find the optimal furnace depth for a certain furnace designing.