An artificial neural network-based approach for economic analysis of insulation thickness using heating degree-day values

In Ihis study, two different structures are considered for the exterior walls of buildings. The first structure is sandwich wall structure that consists of 2 cm interior plaster, 2 picces 13.5 cm horizontally perforated brick with insulation material between and 3 cm exterior plaster. The second one is externally insulated wall consisting of 2 cm inner plaster, 20 cm horizontally perforated brick, insulation and 3 cm outer plaster. The artificial neural network is applied to find temperature, then the optimum insulation thickness is obtained for predetermined function using heating degree-day values. For economic analysis, the calculations arc made on two different wall types. The relationship between insulation thicknesses and annual gain is calculatcd for three types of fuel (coal, natural gas, fuel-oil) in four different provinces (Adana, Gaziantcp, Ankara, Mus) in different degree-day regions. It has been found that the return of investment period has dccrcased with the increase of the HDD number and the insulation thickness has increased with the increase of the HDD number. Among the four provinces, the longest return of investment period is found for Adana province.