Using a chemical equilibrium model to predict amendments required to precipitate phosphorus as struvite in liquid swine manure

İpek Çelen Erdem, John R. Buchanan, Robert T. Burns, R. Bruce Robinson, D. Raj RamanWater 2007 · DOI

TL;DR

This paper demonstrates that a chemical equilibrium model, Visual Minteq, can accurately predict the necessary amendments (specifically magnesium and pH adjustment) to precipitate phosphorus as struvite from liquid swine manure. This offers a promising method for reducing phosphorus pollution from agricultural waste and potentially recovering a valuable fertilizer.

What Problem Does This Paper Address?

Liquid swine manure poses an environmental challenge due to its high phosphorus content. Land application, a common disposal method, often leads to over-application of phosphorus relative to crop needs, resulting in soil buildup and potential contamination of surface and subsurface waters. Removing excess phosphorus from manure slurries is crucial. One effective method is precipitating phosphorus as magnesium ammonium phosphate hexahydrate, commonly known as struvite. However, determining the precise amounts of amendments (like magnesium sources and pH adjusters) needed to maximize struvite precipitation can be complex and time-consuming through traditional jar testing, especially given the variable composition of swine manure. This paper aims to address this by validating the use of a chemical equilibrium model to predict these required amendments, thereby streamlining the process of phosphorus removal.

Methods

The core of this research involves using the chemical equilibrium model Visual Minteq (version 2.23). The authors first modified the model's database to include struvite, a mineral not present in the default version. They also removed species with very slow precipitation kinetics, as the intended application involves short retention times in reactors.

The model was then used to simulate four different conditions for phosphorus precipitation from swine manure:

  1. Increasing pH to 8.5 with a base (NaOH).
  2. Adding a magnesium source (MgCl₂·6H₂O) to achieve a Mg²⁺:PO₄³⁻ molar ratio of 1.0:1.0, followed by pH adjustment to 8.5.
  3. Adding a magnesium source to achieve a Mg²⁺:PO₄³⁻ molar ratio of 1.5:1.0, followed by pH adjustment to 8.5.
  4. Adding a magnesium source to achieve a Mg²⁺:PO₄³⁻ molar ratio of 2.0:1.0, followed by pH adjustment to 8.5.

To validate the model's predictions, laboratory experiments were conducted using actual swine manure. The authors performed batch reactions in reactors, adjusted the pH and added magnesium chloride as predicted by the model for each condition. They then analyzed the treated wastewater and the precipitated solids to quantify phosphorus removal and identify the mineral species formed. X-ray diffraction was used to confirm the presence of struvite and other minerals.

Datasets

The primary dataset used in this study was liquid swine manure samples collected from a small East Tennessee swine finishing facility. These samples were analyzed for their chemical composition (OP, NH₄⁺, Mg²⁺, Ca²⁺, K⁺, alkalinity, and pH) to serve as input parameters for the Visual Minteq model. Two samples were collected from the same facility at different times to account for potential variations.

Key Insights and Ideas

  • Visual Minteq is a powerful predictive tool: The study strongly suggests that chemical equilibrium modeling can effectively predict the optimal amendment strategy for struvite precipitation, significantly reducing the need for extensive and costly jar testing.
  • Magnesium is the limiting factor: For swine manure, magnesium availability is identified as the primary constraint for struvite formation. Simply increasing pH is insufficient for substantial phosphorus removal; adding a magnesium source is essential.
  • Struvite formation is highly pH-dependent: The optimal pH range for struvite precipitation in swine manure was confirmed to be around 8.5, aligning with previous research.
  • Beyond struvite: While struvite is the primary target, the model and experiments also identified the potential formation of other minerals like monetite (calcium phosphate) and brushite (calcium phosphate), especially in conditions with less magnesium addition or in the raw wastewater.
  • Ammonia volatilization is a significant challenge: The study highlights that a substantial portion of ammonia can be lost to the atmosphere due to increased pH and aeration, which the model, relying on equilibrium principles, may not fully capture.

Key Results

  • Significant Phosphorus Reduction:
    • Increasing pH alone (Condition 1) resulted in a 55% reduction in OP in laboratory experiments and a predicted 67.4% reduction by the model.
  • Adding MgCl₂·6H₂O to achieve a 1.0:1.0 Mg²⁺:PO₄³⁻ molar ratio (Condition 2) led to a 94% OP reduction in both experiments and model predictions.
  • Higher Mg²⁺:PO₄³⁻ ratios (1.5:1.0 and 2.0:1.0 in Conditions 3 and 4) achieved even higher OP reductions (97% and 98% experimentally, and 99% and 99.5% predicted).
  • Mineral Identification: X-ray diffraction confirmed the presence of struvite in treated samples, particularly in conditions with added magnesium. Brushite was identified in the raw wastewater and in the pH-adjusted-only sample, while monetite was also detected in the latter.
  • Model vs. Experiment Agreement: A statistical comparison (Wilcoxon rank sum test) found no significant difference between the model's predicted and the laboratory's experimental phosphorus removal percentages (P = 0.05), indicating good predictive capability.
  • NaOH Prediction Discrepancy: The model slightly overestimated the required NaOH addition (3-14% higher) compared to experimental results, attributed to assumptions about alkalinity composition.

Limitations and Future Work

  • Ammonia Volatilization: The study acknowledges that ammonia losses due to volatilization are significant and not fully captured by the equilibrium model. This is a key area where the model's predictions may deviate from reality. Future work could explore incorporating mass transport principles for ammonia.
  • Limited Samples: The statistical analysis for model validation was based on a small number of samples, which limited the statistical power. More extensive sampling would strengthen these conclusions.
  • Alkalinity Assumption: The model's assumption that all alkalinity is carbonate-based led to slight overestimations of NaOH requirements. Further refinement of alkalinity characterization in the model inputs could improve accuracy.
  • Other Minerals: While the study focused on struvite, the presence of other minerals like brushite and monetite suggests that a comprehensive understanding of all potential precipitates is important for complete phosphorus management.
  • Future Work: The authors suggest that the chemical equilibrium model can be easily used for iterative analyses to optimize amendment strategies for various swine manure compositions, paving the way for more efficient and cost-effective phosphorus recovery systems.

Summary updated

BibTeX
@inproceedings{erdem2007using,
  author = {İpek Çelen Erdem and John R. Buchanan and Robert T. Burns and R. Bruce Robinson and D. Raj Raman},
  title = {{Using a chemical equilibrium model to predict amendments required to precipitate phosphorus as struvite in liquid swine manure}},
  booktitle = {Water},
  year = {2007},
  doi = {10.1016/j.watres.2007.01.018},
}

Loading related content...

Figure