Applications and Case Studies > Civil engineering

Simulation of acid-soluble sulfate testing: extraction stage


In the UK, estimates of "total sulfate" in a soil sample are made by quantifying the sulfate in a 100 ml dilute HCl extract. The current standard is BS 1377-3:1990, soils for civil engineering.

Quantification is undertaken by gravimetry with a barium sulfate precipitate (quantification stage, see upcoming post, link here once created).

The extraction stage is as follows:

1. Weigh out a 2g crushed sample into a beaker
2. Add 100 ml of 10 % (v/v) HCl
3. Boil and simmer gently for 15 minutes

If the sample contains sulfides, the sample is instead sprinkled onto an already boiling 100 ml solution of 10 % HCl. This aims to remove sulfides by conversion to H2S (g). The unpleasant smell is a good indicator this has taken place!

A possible simulation of this process can be made using PHREEQC. In the code below, 2 g of pure gypsum is tested.


SOLUTION 1 #10 % v/v HCl
temp 100 #assumed for boiling
pH -0.079 #see notes
Cl 1200 #see notes
water 0.1 #100 ml

EQUILIBRIUM_PHASES 1 #2 g of sulfate mineral gypsum, with a molecular mass of 172.17 g/mol
Gypsum 0 0.0116



The output suggests all 2g of the sample dissolves, i.e., gypsum is completely soluble in this test.



Concentrated HCl is typically 37 % with a molarity of around 12 M.

Assuming full dissociation of this strong acid, the concentrated HCl has 12 M H+ and 12 M Cl-.

The extract is made by diluting 10 ml of conc HCl to 100 ml with deionised water, i.e., 10 % volume conc HCl/volume water (v/v). The extract has a volume of 100 ml or assumed 0.1 kg.

In PHREEQC, molalities are calculated from input concentrations whilst moles of a phase are determined according to the mass of water in solution. In the extract, the molality of H and Cl is 1.2 mol/kgw, or 1200 mmol/kgw (default PHREEQC units), due to the dilution. The actual moles of each in the extract is 0.12.


Any thoughts and discussion appreciated.

A simulation using a 2 g sample containing sulfides is given below:

Soil, 2 g total:

10 % pyrite
10 % gypsum
40 % calcite
40 % quartz


SOLUTION 1 #10 % v/v HCl
temp 100 #assumed for boiling
pH -0.079 #see notes
Cl 1200 #see notes
water 0.1 #100 ml

EQUILIBRIUM_PHASES 1 #2 g soil with above composition
Pyrite 0 1.67e-3 #0.2 g at mr = 119.98 g/mol
Gypsum 0 1.16e-3 #0.2 g at mr = 172.17 g/mol
Calcite 0 7.99e-3 #0.8 g at mr = 100.09 g/mol
Quartz 0 0.0133 #0.8 g at mr = 60.08 g/mol



For the above simulation, the output suggests all calcite and gypsum dissolves. 1.12e-4 moles of pyrite (0.01 g) dissolves whilst 8.77e-5 moles of quartz (0.005 g) dissolves.

The molality of sulfate S(6) in the final solution is 1.186e-2 mol/kgw, equivalent to 0.11393 g SO4 in the 100 ml extract. This equates to 5.69 % of the 2 g sample by mass.

Without pyrite, assuming all gypsum dissolves, the sulfate in the sample equates to 0.111587 g SO4 in the 100 ml extract. This equates to 5.58 % of the 2 g sample by mass.

This shows that the dissolution of pyrite contributes approximately 0.12 % SO4 (by mass of the sample). This is highly significant when it is considered that the upper tolerance for acid-soluble sulfate in many aggregates is set at 0.2 % SO4.


The test method advises sprinkling of the sample onto the boiling dilute acid to remove sulfides as H2S.

Q) If PHREEQC considers a closed system, can this loss by simulated in PHREEQC?

The output file suggests H2S in the dissolved species has a molality of 1.955e-03 mol/kgw.

Any thoughts gratefully received....

The difference

You can simulate the loss of H2S several ways. REACTION allows you simply to remove H2S from solution in specified amounts. KINETICS would allow removing H2S by using a rate expression. You can also add H2S(g) as an EQUILIBRIUM_PHASE and set the target partial pressure to a value, presumably a relatively small value.

I do not work with these standard methods, but seems like in weathered soils you may have Al-Fe-OH-SO4 minerals as well, such as Alunite and Jarosite. Do these need to be considered?

Thanks David, your help is greatly appreciated.

The approaches for dealing with H2S are interesting and I will give these a go.

Lots of other sulfates do indeed need considering, there are a whole host of exotic sulfate minerals of which jarosite is definitely important.

Unfortunately there is a lot of misunderstanding surrounding this test and the geochemical processes which occur as it is carried out. The naming of the test as "total sulfate" itself is misleading and in some cases completely wrong, especially when insoluble sulfates such as jarosite, barite and celestite are in abundance. Even a acid-leach cannot quantify these appropriately. As part of my PhD research I have been predicting this behaviour with PHREEQC and testing it in the lab. So far the results are generally in agreement which is promising.


[0] Message Index

Go to full version