Comparing Modeling Tools Visual MINTEQ and OLI Studio to Evaluate Scaling Tendency of Aerated Anaerobically Digested Solids: A Pilot Study

Purpose: Common phosphate minerals in wastewater treatment plants, such as struvite/magnesium ammonium phosphate (MAP, MgNH4PO4- 6H2O), calcium phosphates such as amorphous calcium phosphate (ACP, Ca3(PO4)2), and iron salt precipitates such as vivianite (Fe3(PO4)2 8H2O), are known for their tendency to scale and deposit within anaerobic digesters, pipe bends, and dewatering equipment downstream of anaerobic digesters. These deposits pose maintenance and operational challenges. Precipitation of these minerals occurs when the concentrations of ionic constituents exceed the solubility product (Ksp) of the solid, measured as scaling tendency (ST) or saturation index (SI). The solubility of these precipitates depends on factors such as pH, temperature, competing ions, and nucleation sites. Chemical thermodynamic modeling can evaluate the parameters contributing to scaling and, thereby, safeguard process equipment from scaling-induced damage, improve treatment efficiency, and regulate phosphorus sequestration to produce nutrient-rich biosolids. This modeling work could benefit the wastewater treatment industry by providing rapid optioneering for solutions limiting and targeting scaling mineral deposition and assessing various process optimizations for new pilot technologies. In this study, two chemical thermodynamic models were evaluated: (i) Visual MINTEQ Version 4.0 and (ii) OLI Studio. Model inputs were gathered from a pilot-scale experimental setup at Hampton Roads Sanitation District's (HRSD) Atlantic Treatment Plant (ATP), in Virginia Beach, VA. The pilot evaluated the impact of aeration, mixing, and chemical addition on the scaling tendency of thermally hydrolyzed, pretreated, anaerobically digested solids. Critical parameters such as pH, temperature, alkalinity, orthophosphate (OP), ammonia (NH3), and major cations and anions concentrations were measured as model inputs. The solids formed during these evaluations were characterized by: X-Ray Diffraction (XRD), Scanning Electron Microscopy (SEM)/Energy Dispersive X-Ray Spectroscopy (EDS), and Raman Spectroscopy analysis to validate predictions of the two models. Methods: The pilot setup at ATP consisted of four tanks (Figure 1), each 11 feet tall and capable of holding about 45 gallons of digestate. These tanks operated as daily batch fed continuously stirred tank reactors, maintained by pump recirculation at a 3-day solids retention time (SRT). Aeration was provided using fine bubble diffuser membranes, operated either at a constant airflow rate or via a dissolved oxygen (DO) set point. The effects of chemical injection of magnesium hydroxide [Mg(OH)2] and calcium hydroxide [Ca(OH)2] were also studied at various Ca+Mg:OP ratios of 0 (control), 0.25:1, 0.5:1, and 1:1. The modeling process was based on pilot operational conditions: anaerobic digestate (pilot influent), aerated anaerobic digestate constantly at 5 LPM, and aerated anaerobic digestate on a DO setpoint of 0.2 mg/L. Inputs for these operational conditions were collected during pilot tests and averaged for model insertion. For each model run, pH and temperature sensitivity analyses were conducted from pH 7-9, and temperatures 20 - 40°C to determine effects on precipitate formation according to Visual MINTEQ and OLI Studio. Experiments with Mg(OH)2 and Ca(OH)2 were conducted in both models for the aerated cases to simulate chemical injection and evaluate the effectiveness of each chemical in OP removal, as well as the accuracy of each model in predicting mineral formation. Results: Both Visual MINTEQ and OLI Studio effectively predicted mineral formation in anaerobic digestate under varying conditions. Initially, both models predicted the formation of metal phosphates (Mg, Ca, Fe) in anaerobic digestate. Upon aeration at 5 LPM, both models identified struvite as the dominant mineral formed in the pH sensitivity analysis, though they differed in predicting other compounds. Visual MINTEQ predicted the formation of calcite and vivianite, while OLI Studio predicted tricalcium phosphates, calcite, and siderite. In the temperature sensitivity analysis, both models yielded consistent results for calcite and struvite formation, suggesting that temperature has a less significant impact on mineral precipitation than pH. In the aerated run targeting a DO setpoint of 0.2 mg/L with Mg(OH)2, both models accurately predicted struvite as the dominant mineral, with its formation increasing alongside the chemical dose, as depicted in Figures 2 and 3. Table 1 compares the OP removal predicted by each model to the measured pilot results for the same test. Both models closely aligned with the measured OP removal via struvite formation, and solid-phase characterization confirmed struvite as the major compound, as indicated by XRD pattern peaks. The presentation will showcase the outcomes derived from each model, offering insights into their respective abilities to predict the potential for the precipitation of controlling solids. The precipitate characterization analysis will also be compared to determine the validity of results and to further optimize both modeling software. The results of this study will enhance our understanding regarding scaling minerals likely to form within anaerobic digestion pretreated with a thermal hydrolysis process, post anaerobic digestion, and on downstream equipment, as well as in a potential post aerobic digestion setup. A validated model could then be used to predict scaling potential and characterize minerals formed at other WRRFs to better treat these nuisance scaling minerals and determine their potential for harvesting/sequestration as a nutrient rich product.