This work spotlights Economic load dispatch under Availability Based Tariff (ABT) using Artificial Neural Network (ANN) based forecasted load and frequency. The ABT has three main ingredients i.e., Fixed charge, Capacity charge and Unscheduled Interchange (UI) charge. The conventional Economic load dispatch (ELD) is a day advance movement, aspiring minimization of the functioning cost subject to system as well as device constraints. The probable load, generation and network availability are the key factors to ELD problem. In this dissertation work, frequency element is added to conventional ELD problem. Under ABT, frequency plays significant part in choosing the real time generation and transmission. The conventional ELD algorithm is accustomed to integrate frequency dependent part of tariff. Thus, former knowledge of frequency along with load will be contributing in the structuring of more realistic schedule of generation. Evaluation of hourly frequency and load using conventional statistical investigation has been reported in previous works. But these methods have certain issues with variation in constraint values. This can be answered using artificial intelligence techniques.