Due to its unique advantages, unmanned aerial vehicle (UAV) communication systems are promising candidates for next generation communication. Although, even with its potential, UAVs suffer from practical limitations such as energy consumption. To mitigate this aspect, we propose a UAV energy consumption minimization technique through the assist of multiple terrestrial intelligent reflecting surfaces (IRSs). By generalizing the probabilistic line-of-sight model, we introduce a channel model capturing the characteristics of the IRSs and user equipments (UEs). Then, by deriving the closed form expression of the UE achievable rate, we formulate an optimization problem to minimize the UAV energy consumption for specific UE data requirements. With problem transformations for tractability and through the successive convex approximation technique, we achieve a local optimum solution. Through the results, we confirm that exploiting the IRSs is indeed extremely beneficial to minimize the UAV energy consumption.