In junction traffic operations, vehicle delay is one of the most essential performance measures of effectiveness. It allows traffic engineers to assess the performance of a traffic system component or the efficacy of a system-wide control plan. Real-time applications such as adaptive signal control, congestion management, and dynamic traffic assignment often use this technology. Obtaining real-time data on intersection performance, such as control delay, may be time-consuming and labor-intensive. This study presents a new approach for estimating network-level real-time delay from passive traffic counting. Total Travel Delay Estimation Technique (TTD) is proposed for signalized intersection delays that can be computed by examining real-time data from arrival and departure detectors upstream and downstream of a junction. The proposed estimation method mathematically manipulated equations that relate the input-output model and vehicle O–D data acquired from the Automatic Turning Movement Identification System (ATMIS). The developed methods utilize the obtained real-time traffic detection system as input data. The proposed methods are applied for three cases: simple, semi-generalized, and generalized networks, where any of them can be used as a building TTD estimation block for the whole actual network. Results from the TTD were compared to VISSIM output, and a statistical test was conducted under varying traffic conditions (low, medium, high, and saturated). The findings show that the proposed methodology can yield stable and reliable results in various traffic volumes and turning movement conditions. Future field implementation studies for the suggested methods are recommended to evaluate the model’s reliability and efficacy in real-time traffic scenarios.