Wireless sensor networks (WSNs) find their application in as diverse fields as collection of data for weather forecasting to detection of enemy activities by defense agencies. Considering the diverse and sensitive areas where WSNs are deployed, un-interrupted and timely delivery of data is as important as energy efficient delivery. This necessitates the requirement of a routing protocol that takes into account both the energy consumption and system delays while finding the best route to deliver packet/data from node to sink. In literature, a number of shortest path based algorithms viz. dikshatra, bellman ford, A*, floyd–warshall’s and heuristic search based algorithms viz. Ant colony optimization (ACO), particle swarm optimization, evolutionary algorithms reinforcement learning have been proposed for enhancing the routing efficiency. ACO which is one of the heuristic search algorithms has proven to more efficient for routing methods due to its dynamic and flexible nature. In most of the ACO based routing algorithms total energy consumption and delay incurred by a path have been considered as two main optimization parameters for finding the optimal path between source and sink. However, due to little difference in the respective optimization parameters of different available paths, the convergence time of these algorithms is very high, which results in longer set up delay and higher energy consumption. In this paper, an ACO based routing algorithm has been proposed which considers excess energy (excess energy is that part of communication energy expenditure, which is used to move packet in direction perpendicular to the line of sight direction between source and destination) as one of the optimization parameters. The use of excess energy consumption as route selection parameter leads to faster convergence of the algorithm as well as results in finding more energy and delay efficient path. The proposed method has been simulated and compared with state-of-the-art ACO based routing methods i.e. deflection angle based ACO algorithm and E&D ANTS algorithm. The simulation results indicate that the proposed method has low convergence time, balanced energy consumption, lower time delay, high packet delivery ratio and leads to longer network lifetime.