The TP removal in SIBPD was explained as follows: (1) phosphorus was assimilated as nutrient for bacteria growing or enzyme composing, researches demonstrated that phosphorus was essential in SNDX275 denitrification. Moon et al. (2008) suggested that nitrate would not be reduced by T. denitrificans without adding KH2PO4. Bruser et al. (2000) indicated that denitrifying bacteria for producing ADP sulfurylase used phosphorus; (2) it was speculated that phosphate accumulating organisms (PAOs) existed in AES for removing a proportion of phosphorus. Previous research suggested that PAOs would uptake phosphorus when the carbon source is sufficient ( Zhu et al., 2011); (3) part of phosphorus was adsorbed by Fe3+ or Fe(OH)3 generated from pyrite in ANS. Several studies reported that Fe3+ or Fe(OH)3 could adsorb a certain amount of phosphorus ( Tang et al., 2014 and Caravelli et al., 2012); (4) another proportion of TP could be adsorbed by filter material at the initial period, especially by volcanic rocks. TP adsorption capacity of volcanic rocks in seminiferous tubules study was 0.92 mg PO43− g−1. However, after adsorption saturation, the microbial removal became the major reason for TP removal.