Clustering techniques for wireless sensor networks (WSNs) have been extensively studied and proven to improve the network lifetime, a primary metric, used for performance evaluation of sensor networks. Although introduction of clustering techniques has the potential to reduce energy consumption and extend the lifetime of the network by decreasing the contention through either power control or node scheduling, scalability remains an issue. Therefore, the optimality of the cluster size still needs to be thoroughly investigated. In this paper, a single cluster head (CH) queuing model is presented. Using an event based simulation tool (Castalia), key issues that affect the practical deployment of clustering techniques in wireless sensor networks are analysed. These include identifying the bottlenecks in terms of cluster scalability and predicting the nature of data packets arrival distribution at the CH. Results presented show that this analysis can be used to specify the size of a cluster, when a specific flow of data is expected from the sensing nodes based on a particular application and also the distribution of the inter-arrival times of data packets at the CH follows exponential distribution.