Wireless sensor networks (WSN) consist of a large group of sensors which are deployed in the field in order to monitor the physical and environmental conditions. Some areas where WSNs are used include Environmental applications like forest fire detection, flood detection, Military applications like target tracking, battlefield surveillance, Health applications like disease prevention etc. The major problem with this type of system is the overall reliability in the presence of some faulty sensors (also referred to as nodes). To achieve reliability, nodes in the network are required to communicate with each other and have an agreement among them to achieve a common task. This research is concerned with the agreement problem among nodes where they need to collectively agree on a common value where the network may contain some faulty nodes. This common value does not need to be exact but can be approximate. Network Convergence is achieved when all the non-faulty nodes in the network have that approximate common value. To achieve convergence, the nodes perform an agreement protocol which involves the collection of values from neighbors, apply convergent voting algorithms and broadcast voted values to their neighbors in a repeated pattern. Each such pattern is called a round. This may need multiple rounds to reach convergence. The voting algorithm is applied to the received values of nodes in order to reduce the differences of opinions among the nodes. A voting algorithm is convergent if the common value stays within the range of the values held by non-faulty nodes. The main purpose of this study is to evaluate the performance of different convergent voting algorithms in terms of the number of rounds needed to reach convergence in the network and the rate of convergence. This study is done for partially connected networks, in the presence of hybrid faults and restricted to only synchronous communication systems. In partially connected networks, nodes are able to communicate only with their immediate neighbor nodes. Hybrid faults involve nodes with different severity of behavior. In synchronous systems, each round of voting has a finite bound on the processing and communication delays. This study will further be analyzed based on asynchronous communication systems because the most realistic applications are asynchronous in nature. The performance evaluation is done through simulation and data are collected for various user input. The simulator is designed in such a way that the user can analyze the network on runtime by adding, removing or moving the sensor from one point to another point in the field.