Among Time-Sensitive Networking (TSN) mechanisms, the Credit-Based Shaper (CBS) ensures bounded latency, while Frame Replication and Elimination for Reliability (FRER) enhances fault tolerance through redundancy. However, efficiently combining CBS deployment with FRER while minimizing infrastructure costs remains a challenging problem. This paper presents an enhanced framework for CBS deployment in TSN networks incorporating FRER, improving the cost efficiency and feasibility guarantee of a state-of-the-art (SoA) FRER-aware partial CBS deployment framework. We first conduct an extensive evaluation of the SoA framework across a wide range of synthetic networks. The results reveal a low success rate in identifying feasible configurations and a high optimality gap for minimizing the TSN-enabled devices, with respect to a brute-force baseline. To address these issues, we introduced an improved framework based on a feasibility-driven CBS configuration heuristic optimizing bandwidth reservation and preserving schedulability for non-shaped traffic classes. Experimental results demonstrate that the proposed framework achieves a full success rate whenever a feasible solution exists, while significantly improving cost-efficiency. We further analyze the proposed solution in terms of sensitivity to network sizes and loads, scalability, and applicability to automotive use-case. The results show that the proposed framework mitigates the blocking effects for high-priority traffic, improves delay bounds for medium-priority traffic, and remains computationally efficient for a broad range of networks, compared to brute-force method.