Inverse modelling of nitrogen species in soils irrigated with wastewater

[joserevueltas94]

Hello everyone, In my PhD I am working on reactive transport of major ions and nitrogen species. Recently I am using Phreeqc to make an inverse model to quantify some chemical reactions. Two of the samples were taken at 80 cm depth and the other is a wastewater sample. I wrote the reactions of nitrogen species in the "PHASES" command, however, when I run the model not all the proposed chemical reactions are taken into account. Furthermore, I obtain negative mole tranfers for N(-3) in "Redox mole transfers" in all the models and I am not sure what it means.

I think that negative values of redox species indicate reduction and positive oxidation, in this case, negative values for N(-3) are not possible in this well aerated system. It is not clear to me the meaning of negative or positive values in "Redox mole transfers" for the modelled chemical species.

With this post I seek help to understand how to interpretate the obtained results and how to indicate to the model some resctrictions related to the chemical species in "SOLUTION_SPECIES".



Here is my code

SOLUTION 1   80 cm depth - 20/12/2019   
   units mg/l   
pH   8.20   
Na   253.81   
N(-3)   0.29   as NH4
K   24.17   
Ca   62.49   
Mg   26.72   
Alkalinity   597.90   as HCO3
Cl   166.77   
N(3)   0.00   as NO2
N(5)   11.41   as NO3
P   11.06   
S(6)   130.93   
      
      
      
SOLUTION 2   Wastewater 16/01/2020   
   units mg/l   
pH   8.14   
Na   189.42   
N(-3)   25.85   as NH4
K   30.71   
Ca   39.47   
Mg   26.29   
Alkalinity   496.36   as HCO3
Cl   156.11   
N(3)   32.82   as NO2
N(5)   4.32   as NO3
P   4.74   
S(6)   87.33   
      
      
SOLUTION 3   80 cm depth - 17/01/2020   
   units mg/l   
pH   7.85   
Na   236.01   
N(-3)   0.00   as NH4
K   20.20   
Ca   57.33   
Mg   22.89   
Alkalinity   443.51   as HCO3
Cl   186.58   
N(3)   2.73   as NO2
N(5)   36.32   as NO3
P   13.75   
S(6)   136.46      




INVERSE_MODELING
        -solutions 1 2 3
        -uncertainty 0.05
        -phases
               Calcite
               Ca3(PO4)2(Amorfo)
               H2O(g) preci
               NO3_1  force diss
               NO3_2  force diss
               KX
               CaX2
               NaX 
               Cysteine diss
               Denit  diss
               NH4X force prec
               MgX2
               NO2_1  diss
               N2(g) prec
           
        -range
        -balances
                Na
                S(6)
                Mg
                Cl
        #-force_solutions true true


PHASES
 
Denit
      CH2O + 0.8NO3-  = 0.4N2 + HCO3- +0.2H+ + 0.4H2O
NO3_1
      NH4+ + 2O2 = NO3-   + H2O + 2H+
NO3_2
      NO2- + 0.5O2 = NO3-
NO2_1
     NH4+ + 1.5O2 = NO2- + 2H+ +H2O

Cysteine
      C3H7NO2S + 4.5O2 = NH4+ + SO4-2 + 3CO2 + H+ + H2O

Ca3(PO4)2(Amorfo)
      Ca3(PO4)2 = 3Ca+2 + 2PO4-3

[dlparkhurst]

You should not use phases to define the nitrogen reactions, only sources that introduce N to the water and sinks that remove N from the water. Transitions among redox states should be handled by the inverse modeling equations.

Balances are set up automatically for all elements in the phases that are used. If you have an element that is not in a phase, then you need to use -balances to include that element. That element would be either conservative, or determined by mixing of the initial solutions, or by evaporation of water. Here Cl is included in this way.

As a rule, you need a separate source or sink for each element in the mole balance calculation. I have added Gypsum as a source/sink for sulfate. You can consider what processes are reasonable for sulfate.

Initially, I usually use -minimal to reduce the number of models. You can remove the option after you have the basics working, I could argue that the additional models found are not significant within the uncertainties that you have specified.

The signs of the redox reactions are tough to interpret, but your final water has only N(3) and N(5), so any N(-3) contributed from the initial waters must be oxidized to these species or removed as N2(g).

Code: [Select]

PHASES
Cysteine
      C3H7NO2S + 4.5O2 = NH4+ + SO4-2 + 3CO2 + H+ + H2O

Ca3(PO4)2(Amorfo)
      Ca3(PO4)2 = 3Ca+2 + 2PO4-3
END

SOLUTION 1   80 cm depth - 20/12/2019   
   units mg/l   
pH   8.20   
Na   253.81   
N(-3)   0.29   as NH4
K   24.17   
Ca   62.49   
Mg   26.72   
Alkalinity   597.90   as HCO3
Cl   166.77   
N(3)   0.00   as NO2
N(5)   11.41   as NO3
P   11.06   
S(6)   130.93   
     
SOLUTION 2   Wastewater 16/01/2020   
   units mg/l   
pH   8.14   
Na   189.42   
N(-3)   25.85   as NH4
K   30.71   
Ca   39.47   
Mg   26.29   
Alkalinity   496.36   as HCO3
Cl   156.11   
N(3)   32.82   as NO2
N(5)   4.32   as NO3
P   4.74   
S(6)   87.33   

SOLUTION 3   80 cm depth - 17/01/2020   
   units mg/l   
pH   7.85   
Na   236.01   
N(-3)   0.00   as NH4
K   20.20   
Ca   57.33   
Mg   22.89   
Alkalinity   443.51   as HCO3
Cl   186.58   
N(3)   2.73   as NO2
N(5)   36.32   as NO3
P   13.75   
S(6)   136.46     

END

INVERSE_MODELING
        -solutions 1 2 3
        -uncertainty 0.05
        -phases
               Calcite
               Ca3(PO4)2(Amorfo)
               Cysteine diss
               H2O(g) preci
               KX
               CaX2
               NaX
               NH4X force prec
               MgX2
               N2(g) prec
   Gypsum
        -range
END