Ensuring a post-disaster water supply is crucial to supporting human life and various social activities, and consists of a quick restoration of damaged water facilities and providing an emergency water supply through tank trucks. In our previous study, we developed a recovery process simulation of a post-disaster water supply system, considering both restorations and the emergency water supply based on the modeling framework of multiple interdependencies among lifelines, enterprise activities, and civil life activities. We optimized repair plans using this simulation, incorporating the genetic algorithm (GA). The genetic algorithm, however, requires a large amount of computational time to achieve the optimization, which suggests that the GA is not practical for real-time repair planning. An alternative planning method, offering both less computational time and higher explainability, is heuristics-based planning using empirical rules regarding restoration and the emergency water supply. There are several common empirical rules regarding the restoration prioritization of damaged pipelines and resource allocation. In this context, we implemented seven empirical rules, while also running the recovery process simulation with a plan generated using these heuristics. We then compared the results with the optimal plan generated by the genetic algorithm. The results showed that the computational time was reduced enough for real-time use, while the obtained repair plans exhibited quasi-optimality compared to those obtained by the GA.