Leakages from pipelines are low-probability but high impact events for distributors of crude/refined oil as well as drinking water. Several techniques including geological surveys, fiber optics, acoustics, pressure drop analysis and mass imbalance are in use to detect the leakages. However, all the techniques have their own drawbacks such as low sensitivity, high false alarm rate and process parameter dependencies. Pipe joints, being the most prone to leakages, have been targeted in one of our R&D projects for real-time leakage monitoring. Instead of relying on indirect indicators such as vibration and pressure drop, we are proposing a new kind of microwave leak detector which can easily be clamped on a joint. The sensor utilizes a uniquely designed microwave resonator to detect the change in dielectric properties (caused by leakage) of a 3D printed porous medium embedded inside. Protective clamp-style cover not only protects the sensor from environmental effects such as rain but also helps avoiding false alarms. The proposed system can not only detect the leak event but can also estimate its intensity.