Extension and application of the three-phase weak acid/base kinetic model to the aeration treatment of anaerobic digester liquors

The kinetic model for the carbonate (inorganic carbon) system in three phas es and the phosphate, short chain fatty acid and ammonia. mixed weak acid/b ase systems in single phase, based on modelling the kinetics of the forward and reverse dissociation processes of the weak acid/base species, is exten ded to model the three-phase mixed weak acid/base chemical reactions in ana erobic digester liquor aeration. The kinetic reactions for ion pairing, Pre cipitation of struvite, newberyite, amorphous calcium phosphate, calcium an d magnesium carbonate and stripping of NH3 gas are added to the model. A pr eliminary validation of the mode was done by comparing model predictions wi th an equilibrium based struvite precipitation model; and equilibrium exper imental results from the literature on simultaneous precipitation of calciu m and magnesium carbonate and phosphate minerals. To collect data to valida te the non-steady state behaviour of the model, aeration batch tests were c onducted on liquor from a spent wine upflow anaerobic; sludge bed (UASB) di gester and a sewage sludge (SS) anaerobic digester. In the batch tests pH, Ca, Mg; Total PO4-P (P-T), free and saline ammonia (FSA, N-T) and H2CO3* al kalinity (from which inorganic carbon C-T is calculated) with time were mea sured over 24 to 56 h. To simulate the batch test results, solubility produ ct and ion pairing constants were obtained from the literature End the mine ral precipitation and CO2 and NH3 gas stripping rates calibrated by minimis ing the error between the predicted and measured results. A very good corre lation was obtained for all the measured parameters in six out of the seven batch tests, four on SS and. three on UASB anaerobic digester liquors (ADL ). From the model results the masses of the different calcium and magnesium carbonate and phosphate minerals that precipitated and the CO2 and NH3 str ipped via the gas phase could be calculated. Comparing the results from the two liquors it was found that: the minerals that precipitated were very si milar in both, viz. in decreasing proportion of precipitate mass formed, st ruvite (MgNH4PO4) (82 to 89%), amorphous calcium phosphate (ACP) (5 to 15%) , calcium carbonate (CaCO3)(0-6%), magnesium; carbonate (MgCO3) (0 to 58) a nd newberyite (MgHPO4) (0.1 to 0.3%); the solubility product values for the se minerals were the same in both liquors and within the range:of literatur e values, but the specific precipitation rates were different; the rates of struvite and ACP precipitation were 9 and 2 times faster in the UASBDL tha n in the SSADL respectively; in contrast, the rate of CaCO3 precipitation w as 140 times faster in the SSADL than in the USASBDL; the rates of MgCO3 an d MSHPO4 precipitation were approximately the same in both liquors; and the stripping/dissolution rates of oxygen (O-2), carbon dioxide (CO2) and ammo nia (NH3) increased-with increased aeration rate in the batch tests. The ki netic modelling approach allows the determination of the specific precipita tion rates for a number of minerals Simultaneously in an integrated manner from a single batch test.