This paper presents a combined economic dispatch and demand response optimisation model under renewable obligation. Real data from a large-scale demand response programme are used in the proposed model. The model uses renewable obligation policy and direct load control to find an optimal energy and reserve strategy that minimises generation costs and maximising renewable penetration. The real data from the South African large-scale demand response programme are used in which the system operator can directly control the participation of electric water heaters at the substation level. The actual load before and after demand reduction are used to assist the system operator in making optimal decisions on whether a substation should participate on the demand response programme. The application of real data to the proposed model avoids the traditional approaches which assume arbitrary controllability of flexible demand. The effectiveness of the model is tested using the modified IEEE 30 and 118-bus systems. The results show that the proposed model can achieve significant demand reduction as high as 830 MW for the IEEE 30-bus system and shift up to 401 MW. Moreover, a total cost reduction of 21.56% can be achieved using the IEEE 118-bus system.