Energy Storage System (ESS) has become a vital development for smart buildings and smart grids. Centralized intelligent ESS controller requires intelligent sensing with accurate real-time sensing data, which demands high latency and packet abnormal rate (PAR) performance of the communications between sensors and the control centralizer. Large scale indoor public spaces (IPS), generally characterized by large volumes, require robust connections among deployed sensors for ESS management. Wireless sensor network (WSN) has been proven to be one of the best candidates for data transmission in large scale ESS systems. ZigBee, attributed to its excellent performance on scalability, low power consumption and relatively sufficient data rate, is a crucial wireless protocol for WSNs in building applications. In this paper, a ZigBee based optimal WSN for ESS is proposed. The design considers multiple parameters for ESS which includes power efficiency, throughput, and robustness (i.e. packet loss, packet error). The parameters are optimized by designing a multi-objective optimization formulation on ESS. The optimization achieves a low latency and a low packet abnormal rate (PAR) and yet provides a real-time accurate data transmission. Experimental results reveal that the proposed system achieves a low PAR of 0.08%. The optimal design provides generic guidance for large scale WSNs design for ESS using ZigBee.