Simulating the BS 1377-3 pH test - some thoughts

[Tom]

The current British Standard for determining pH of a soil extract is as follows:

1. Allow a soil sample to air-dry
2. Crush the sample to < 2mm grain size
3. Obtain a sample of mass approx. 30-35 g
4. Weigh out 30 g of soil and place in a 100 ml beaker
5. Add 75 ml of DI to beaker, stir for a few mins, cover and stand for > 8 hours.
6. Calibrate and clean pH meter electrodes
7. Immerse electrodes into soil suspension
8. Take 2-3 readings of pH value, accept if within 0.05 pH units

This test can be simulated in PHREEQC:

The hypothetical sandy sulfate-bearing soil has the following properties:

a) 30 g dry mass
b) 90 % quartz, 5 % gypsum, 5 % kaolinite

-------------

SOLUTION 1 #DI
temp 18 #assumed laboratory temperature
water 0.075 #75 ml used for test

EQUILIBRIUM_PHASES 1 #total mass 30 g, assumes 8 hours is sufficient for equilibrium
gypsum 0 8.71e-3 #1.5 g gypsum
quartz 0 0.449 #27 g quartz
kaolinite 0 5.81e-3 #1.5 g kaolinite

END


----------

The output file estimates a solution pH of 6.9 using the phreeqc.dat database.

How comparable is this simulation to the likely pH determined in the lab for a similar soil?
Is equilibrium a reasonable assumption?

Discussion of the suitability of this simulation (and for other soil compositions) encouraged...

[dlparkhurst]

Gypsum could react in that time frame, but it has little effect on the pH. Kaolinite might react, but I doubt it would reach equilibrium, and it is fairly insoluble anyway. Quartz reacts very slowly. Calcite reacts relatively fast, and in many cases would be an important factor, but it is not present in your case.

However, my guess is that surface reactions would be important in determining what the pH will be. In turn you have to ask what sets the pH of the surfaces. It could could be any number of solutions or reactions that occur over longer periods of time, including rainwater, organic decomposition, oxidation of pyrite, soil additives, or other mineral reactions. Probably a difficult thing to model without making a whole study of each soil and hydrologic environment. 

[Tom]

Thanks for your comments David

The lack of any carbonate mineral in my ideal soil certainly is unrealistic.

Perhaps a comparison of some laboratory measured pH values (on artificial soils) and PHREEQC simulated values would go some way to indicating the level of disparity....