Source rock deduction

[rudolff16]

Dear all,

it is a well known that using of ions ratios can help to predict the potential groundwater origin or the type of hosting rocks, but those methods are rather old. The question is how to predict the source rocks of a groundwater sample from its chemical constituent in PHREEQC?
I my case the ratio Na/Cl in groundwater samples suggest another sources for Na+ in addition to the halite dissolution such as basalt weathering (Plagioclase), so how to model it?
I would appreciate any help from you

[dlparkhurst]

I think inverse modeling is a useful exercise to try to match mineral/gas sources and sinks to the changes in elemental concentrations in solution. If you assume rain is the starting solution, it can be approximated with pure water. Then the exercise is to try to account for the change in composition between pure water and your analyzed water composition by dissolving and precipitating minerals and gases that are plausible for your geologic terrane.

You can look at the documentation fro NETPATH and PHREEQC to find examples of inverse modeling. PHREEQC version 3 examples 16, 17, and 18 use inverse modeling. Perhaps the classic Garrels and Mckenzie example of silicate weathering would be appropriate for you (example 16).

[rudolff16]

Hi dlparkhurst,

thank you very much for your suggestion,
I will try to apply the mentioned script and tell you about the results

thanks

[rudolff16]

I am going to use the exp18 of Madison Aquifer to simulate the dedolomitization in my study area using rainfall water as first solution and groundwater as a second one. Can I apply the script in the case where information about isotopes composition are available?
thanks

[dlparkhurst]

Example 18 demonstrates the use of isotopes, but if you do not have isotope data, simply remove
the -isotopes block from INVERSE_MODELING.

[rudolff16]

Dear dlparkhurst,
I tried to run the script several time, but nor models found in the output
when I set the -tolerance 1e-2 and uncertainty to 0.05, PHREEQC gives 150 models but without any results phase mole transfer.
the model based on the assumption that rainwater is first evaporated and only 10% of residual water is in interaction with groundwater. After that, de-dolomitization process takes place by dissolution of gypsum, calcite precipitation and dolomite dissolution. A cation exchange process is also expected. (units are in ppm)
solution 1   Rainwater:
Ca       Mg      Na             K         S(6)        S(-2)     Cl          Fe(2)   Fe(3)  C(4)   N(5)   N(-3)
10.4   1.22     5.98          0.01       29.62    0.001     7.09      2.14    0.72   12.2    1.33   0.67
pH 6.4 and Temp 12.9 °C
P(CO2)= -3.5
solution 2 Groundwater:
Ca          Mg         Na        K       S(6)      S(-2)     Cl             Fe(2)     C(4)       N(5)
125.9     107.8     446.5    6       665.3     0.01      484.67       0.1       179.98   37.58
pH 7.39, Temp 23 °C
P(CO2)=-1.5

If you like I can provide you the script file via Email
please any suggestions?
thanks

[dlparkhurst]

Attach the script to a post.

[rudolff16]

here is the script:
Title simulate Dedolomitization, ion exchnage in the Neogene aquifer
SOLUTION 1 Rainwater
units ppm
ph  6.4
temp  12.9
Ca  10.42
Mg  1.22
Na  5.98
K   0.01
S(6) 29.62
S(-2) 0.001
Cl  7.09
Fe(2) 2.14
Fe(3) 0.72
C(4) 12.2
N(5) 1.33 as NO3
N(-3) 0.67  as NH4
EQUILIBRIUM_PHASES
CO2(g) -3.5
REACTION 1 # Evaporation
H2O -1
49.959
save solution 2
end
MIX 1
    2 10
save solution 3 Recharge water
SOLUTION 4 Groundwater
units ppm
temp  23
pH   7.39
redox S(6)/S(-2)
Ca  125.9
Mg  107.8
Na  446.5
K   6
Fe(2) 0.1
Cl 484.67
S(6) 665.3   
S(-2) 0.01
C(4) 179.98
N(5) 37.58 as NO3
EQUILIBRIUM_PHASES
CO2(g) -1.5
INVERSE_MODELING 1
-solutions 3 4
-tolerance   1e-2
-uncertainty 0.05
-range
-balances
N
-PHASES
Dolomite        dis
Calcite         pre
Gypsum          dis
Sylvite
Fe(OH)3(a)
Halite          dis
CaX2            pre
MgX2            pre
NaX             pre
PHASES
 Halite
   NaCl = Na+ + Cl-
   log_k      1.582
   delta_h 0.918 kcal
Sylvite
        KCl = K+ + Cl-
        -log_k  0.0
 EXCHANGE_SPECIES
Na+ + X- = NaX
   log_k      0.0
   -gamma   4.0   0.075
Ca+2 + 2X- = CaX2
   log_k      0.8
   -gamma   5.0   0.165
Mg+2 + 2X- = MgX2
   log_k      0.6
   -gamma   5.5   0.2
END

[dlparkhurst]

You have both NH3 and NO3 in your initial water, final water has only NO3, which is not consistent with having some sulfide. You expect the nitrate to be reduced before sulfide is generated. However, you can either add O2(g) in the evaporation step or add O2(g) in the inverse modeling.

Here is an input file that generates some models. You can change phases and dis/pre constraints to see if you find any models that fit your understanding of the system better.

[rudolff16]

Dear dlparkhurst,
thank you very much for your support, really appreciated.