Abstract: Recently considerable growth in the development of cluster based routing protocols is not only used to optimize the Wireless Sensor Networks (WSNs) resources but also increase the performance and reliability of the WSNs applications. However, dynamic behavior of sensor node such as energy consumption, geographical distribution and other complex factors makes the design of cluster based routing algorithm a challenging task. In this study, recent research for developing clustering protocols is highlighted and limitations are identified. Based on these limitations, a new granular method for dynamic grid routing protocol based on Quotient Space theory is proposed. The algorithm achieved the efficient clustering by accomplishing the effective comparison and transformation among quotient sets with different granularities. It can comprehensively consider node energy consumption; residual energy balance and geographical distribution balance. Furthermore, the optimal number and shape of cluster is dynamically calculated instead of relying on prior knowledge. The experimental result shows efficient energy and network lifetime as compared to other similar methods in WSNs.