Urban flooding in dense institutional areas is often caused by the combined effects of limited drainage conveyance, downstream outfall restrictions, and complex surface-flow interactions. This study evaluates the flood vulnerability of the Bangladesh University of Engineering and Technology (BUET) campus in Dhaka, Bangladesh, using a coupled 1D–2D PCSWMM modeling framework. A 10-year, 6-hour design hyetograph for Dhaka was used to assess the performance of the existing stormwater drainage system. Under baseline conditions, the campus experiences severe localized waterlogging, with a maximum inundation depth of 0.88 m, a flooded area of 267,900 m², and a flood duration of 18.85 hours.

Initial improvement scenarios were developed to evaluate pump-based interventions, including existing temporary pumps within the BUET campus and a proposed 6.6 m³/s pump at the Islambag outfall. These scenarios produced only limited reductions in inundation depth, flooded area, and flood duration, indicating that pumping alone cannot fully mitigate flooding when upstream pipe conveyance constraints and internal hydraulic bottlenecks remain unresolved. This occurs because pumps can remove water only after it reaches the pumping locations, while undersized conduits and restricted hydraulic connectivity continue to delay stormwater movement from flood-prone areas toward the outfall.

To identify more effective mitigation options, 25 combined pipe–pump improvement scenarios were then simulated by pairing upgraded conduit diameters from 5 ft to 9 ft with outfall pump capacities from 6.6 to 14.5 m³/s. The results show that combined conveyance and pumping upgrades substantially improve drainage performance compared with pump-only interventions.

A multi-criteria sensitivity analysis was conducted by comparing hydraulic performance with estimated infrastructure cost under varying priority weights. From a performance-oriented perspective, the 9 ft pipe diameter with a 10.5 m³/s pump produced the best hydraulic outcome, reducing the maximum inundation depth to 0.63 m, flooded area to 101,300 m² and flood duration to 7.21 hours. For more budget-constrained implementation, the 5 ft or 6 ft pipe configurations with a 14.5 m³/s pump provide practical short-term alternatives with meaningful flood reduction benefits.

Considering subsurface utility congestion and construction feasibility within an established campus, this study also suggests buildable alternatives to a single large-diameter pipe, such as a 12 ft × 5.5 ft rectangular box culvert or twin parallel 6 ft pipes. Overall, the study demonstrates how PCSWMM can support scenario-based, cost-conscious and phased drainage modernization for flood-prone urban campuses and similar dense urban environments.

Figure 1: BUET area and Drainage Model

Figure 2: Integrated 1D–2D PCSWMM model developed for simulating drainage network flow and surface inundation in BUET campus.

Figure 3: Baseline inundation condition under the existing drainage system, showing severe localized flooding in low-lying campus areas.

Figure 4: Sensitivity analysis showing how the optimal pipe–pump combination changes under different hydraulic-performance and cost priorities

Figure 5: 2D inundation maps showing optimized short-term and long-term drainage configurations: short-term option with 6 ft conduit and 14.5 m³/s pump and long-term option with 9 ft conduit and 10.5 m³/s pump