Abstract: ［Objective］ By establishing a numerical seepage analysis model that aligns with real drainage systems and introducing the concept of a ′virtual permeability coefficient′ for secondary lining, the objective is to delve into the correlation between numerical methods and theoretical formulas, with expectation to leverage the efficiency and practicality of theoretical formulas in predicting external water pressure. ［Method］ Based on the principle of equivalent stable drainage volume in underwater tunnels, the concept of a ′virtual permeability coefficient′ for the secondary lining is introduced. On this basis, key factors, including the spacing of circumferential drainage blind pipes, the thickness of geotextiles, and their permeability coefficients, are selected as primary research factors. By adjusting these factors, multiple numerical seepage analysis models consistent with real drainage systems are established. ［Result & Conclusion］ The actual external water pressure acting on the secondary lining exhibits significant spatial distribution characteristics. Longitudinally, the variation in external water pressure displays periodic fluctuations corresponding to the spacing of circumferential drainage blind pipes. Circumferentially, the closer the position is to the longitudinal drainage blind pipe, the lower the external water pressure, with maximum circumferential water pressure occurring at the arch vault, followed by the inverted arch, and the smallest pressure on sidewalls. The reduction coefficients of external water pressure calculated with theoretical formulas are generally smaller than those derived from numerical methods. The stronger the drainage capacity of the design parameters, the smaller the difference between the two calculation results. The reduction coefficient consistently follows a decreasing trend from the vault to the invert to the sidewalls. When applying theoretical formulas directly in quantitative engineering design, it is necessary to introduce a comprehensive correction factor greater than 1.0 to ensure engineering safety. The value of comprehensive correction factor should be determined based on the specific structural location, with zones divided by the sidewalls. For the upper structure, a range of 1.48-1.97 is recommended, while a proper range of 1.21-1.39 for the lower structure.