In the process of oil drilling, well leakage poses a significant threat, leading to both economic losses and environmental pollution. Therefore, a novel approach for detecting well leakage is proposed, utilizing cepstrum analysis of transient pressure waves. This method aims to identify the location and amount of drilling fluid leakage by analyzing the propagation characteristics of transient pressure waves within the wellbore. By monitoring the time-dependent and amplitude variations of characteristic peaks in the pressure wave signal, valuable information regarding the leakage location and volume can be obtained. Experimental results demonstrate the effectiveness of the CEEMDAN-WT-CCF-HJS algorithm in reducing noise interference. In the presence of leakage, cepstrum analysis exhibits distinct characteristic peaks in the reflected wave, facilitating leakage location determination and estimation of the leakage volume. The experimental error is observed within the range of 2.25% to 9.10%, thereby offering theoretical support and technical guidance for real-time well leakage detection in practical applications. This research thus contributes towards safeguarding well integrity and minimizing the adverse consequences associated with well leakage.