Bayesian estimation theory was used in this study to establish a state-space model of the ceramic filter tube degradation process. The model continuously integrates the latest residual pressure drop to update its own parameters, then the change rate of the degradation state, remaining life, and failure probability density distribution of the tube in real time. The residual pressure drop of the Shell Coal Gas Process was analyzed to find that the model results converge to real values as the residual pressure drop increases. The change rate of the ceramic filter tube degradation state calculated by the model gradually decreases over time, which is consistent with the initial rapid increase in residual pressure drop followed by a slower increase in later stages of operation. The amount of particle deposition in the ceramic filter tube wall under different operating times was measured and predicted; the predictions are consistent with the state-space model results. The state-space model also reflects variations in filter tube performance degradation caused by emergent conditions such as leakage or fractures, as it does not make stationarity assumptions for the degradation process.