Conceptual Models > Design of conceptual models

Calculated Total Alkalinity in brine completely different than measured

(1/1)

Idoboro:
Hello there,

I work with reverse osmosis desalination brines originate from the Mediterranean sea and I tried to model my solution with PHREEQC and gave it its chemical composition (pH, Na, K, Mg, Ca, Cl, SO4 and DIC). The solution input and description are attached in the end. Due to the high ionic strength of my solution, I work with Pitzer database.

In the output PHREEQC calculated the water Total Alkalinity to be ~0.57 meq/kg which is much lower than the Alkalinity that I measured which was ~4.7 meq/kg and also much lower than regular seawater alkalinity.

Does anybody know where this huge error can come from and how can it be dealt with?
Is it okay that I inserted the DIC value that I measured as C?

I'll appreciate any help.

EQUILIBRIUM_PHASES 1 Equilibrium with atmosphere
CO2(g)    -3.376 10
O2(g)     -0.6778 10
SOLUTION 1 Sorek Brine-PITZER
temp      25
pH        8.125
pe        4
redox     pe
units     mmol/l
density   1.05
C         4305 uMol/l
Ca        23
Cl        1170
K         44
Mg        116
Na        1004
S         65
water    1 # kg
------------------------Description of solution----------------------------

pH  =   8.125
pe  =   4.000
Activity of water  =   0.959
Ionic strength (mol/kgw)  =   1.547e+00
Mass of water (kg)  =   1.000e+00
Total alkalinity (eq/kg)  =   5.676e-04
Total CO2 (mol/kg)  =   4.403e-03
Temperature (?C)  =  25.00
Electrical balance (eq)  =   2.162e-02
Percent error, 100*(Cat-|An|)/(Cat+|An|)  =   0.81
Iterations  =  17
Gamma iterations  =   4
Osmotic coefficient  =   0.93438
Total H  = 1.110207e+02
Total O  = 5.578753e+01

dlparkhurst:
I can't reproduce your results. Using pitzer.dat, I get an alkalinity of about 5 eq/kgw. Using PRINT; -alk, the following species contribute to the total alkalinity.

--- Code: --- Total alkalinity (eq/kgw)  =   5.036e-03

Species          Alkalinity    Molality   Alk/Mol

HCO3-             3.767e-03   3.767e-03      1.00
MgCO3             6.449e-04   3.224e-04      2.00
CO3-2             6.140e-04   3.070e-04      2.00
MgOH+             6.852e-06   6.852e-06      1.00
OH-               3.081e-06   3.081e-06      1.00
H+               -7.598e-09   7.598e-09     -1.00
HSO4-            -3.472e-09   3.472e-09     -1.00
--- End code ---

Alkalinity is independent of pCO2. In other words, adding or removing CO2 from solution does not change the alkalinity.

Pitzer.dat does not have O2(g) defined, so I am not sure what database you used.

--- Code: ---PRINT
-alk
SOLUTION 1 Sorek Brine-PITZER
temp      25
pH        8.125
pe        4
redox     pe
units     mmol/l
density   1.05
C         4305 uMol/l
Ca        23
Cl        1170
K         44
Mg        116
Na        1004
S         65
water    1 # kg
END
USE solution 1
EQUILIBRIUM_PHASES 1 Equilibrium with atmosphere
CO2(g)    -3.376 10
#      O2(g)     -0.6778 10
END

--- End code ---

Idoboro:
Hi, thanks for the help.

Somehow you got much more realistic value for the Total Alkalinity, you got ~5 meq/kgW when it should be ~4.7, but I got a lower value in an order of magnitude.

I erased the O2(g) equilibrium forcing but I'm getting the same results.
The database that I'm using is:
# data0.ypf.R2
# PITZER THERMODYNAMIC DATABASE INCLUDING ACTINIDES AND TRANSITION METALS
# (09/28/2006)

I got it from a person named Steven Benbow from an organization called "Quintessa". Is it not the correct PITZER database for high ionic strength solutions?

I'm new to PHREEQC and still not so familiar with everything, so I wonder, how could we get such different outputs for the same inputs?

dlparkhurst:
I guess they did not define the alkalinity of aqueous species correctly in SOLUTION_MASTER_SPECIES data block.

I used the database pitzer.dat, which is found in the "database" subdirectory of each PHREEQC distribution.