Gas mask and filter adsorber designs are mainly determined by the accurate breakthrough time calculated by fitting descriptive models to breakthrough curves in military chemistry. Currently, in most practical environments, the highly toxic fluoride gas/vapor breakthrough curves with roll-up and/or stepwise breakthrough structures in the C/C0<1.0 range usually occur, which is induced by the existence of pre-adsorbed water vapor. Common and easily modified descriptive breakthrough models have a significant shortcoming in that they are not able to simultaneously fit the above complex breakthrough curves and explain gas desorption, enhanced adsorption again, and stepwise adsorption in the competitive adsorption process. This study adapts three modified descriptive models by introducing a dimensionless breakthrough coefficient to fit roll-up and stepwise 1.1.1.2-Tetrafluoroethane (R134a) breakthrough curves under humid conditions. Results reveal that correlation coefficients between the model and R134a breakthrough data are higher than 0.98, demonstrating that the novel adaptation successfully accounted for R134a desorption, enhanced R134a adsorption again, and R134a stepwise adsorption without significantly increasing the complexity of implementing the models. Further, through relationship curves between fitting parameters and moisture contents, R134a breakthrough curves on activated carbons under a wider moisture content range were successfully predicted. It will greatly reduce the cost of toxic experimentation and improve experimental safety factors.