Organics modelling

[arsenist]

Hi,

I am trying to model monsoonal influxes of organic matter (particularly humic/fulvic/protein-like fractions, identified by fluorescence properties) from the Ganges river to groundwater, using PHT3D. My colleague has developed a groundwater flow model using MODFLOW already and I would like to add the geochemical aspect to this. I am new to PHT3D/PHREEQC and would like to know if this is possible and what parameters need constraining/defining to make this possible.

I have found the Tipping & Hurley database which contains data on metal binding to humate and fulvate, but I cannot find any data on humic, fulvic or amino acid binding to ferrihydrite for the surface chemistry. I assume sorption to the sediment would be one of the predominant processes affecting organic matter influx? Would this have to be experimentally defined in this case? Or perhaps use a Kd value in the MT3DMS part?

I would assume kinetics of the degradation of organic matter would also be important to the modelling, especially for different sub-components of organic matter (as protein-like organic matter is thought to be more reactive). How can I incorporate different rates of degradation for different fractions? Is there any data on these rates available?

I’m just generally a bit stuck on this. Is there anything else I’m not considering? Perhaps I will consider just a humate fraction on its own to reduce the complexity of the problem.

Cheers

George

[Leo]

Hi George,

a bit late to your question and probably not a good answer, because I haven't worked with PHT3D before. For reactive transport modeling, I am using PHAST, which however, doesn't build on MODFLOW. You can define your own rate definitions and phases quite easily, exactly like you would do in PHREEQC. There, it is also quite straightforward to look at different degradation rates for several organic substances.

I also cannot help you with the missing data, but if you have questions "translating" your MODFLOW model to PHAST (which can be very frustrating), I might be able to answer some of your questions.

You can also find some information on PHAST on this forum.

Leo

[MichaelZ20]

This is Richard Winston's reply to the question:

If you import a MODFLOW model into ModelMuse you can change the model selection to PHAST but PHAST is very different from MODFLOW and you will need to do considerable work to generate an equivalent PHAST model to the MODFLOW model.

[dlparkhurst]

The biggest difference is that the grid for PHAST is point distributed (the domain starts and ends at nodes), compared to cell centered for MODFLOW (the domain starts at the edge of a cell surrounding a node).

There is also a GUI Phast4Windows by Scott Charlton (which I think is excellent) to define model properties and run simulations (in parallel). So, it depends on how complicated the MODFLOW model is. If your model uses relatively simple boundary conditions and property distributions, you can make an equivalent model with P4W relatively easily. Most people make crazy complicated MODFLOW models, and it could be difficult to make an equivalent model for PHAST. Usually, I think it is better to stick with PHT3D. However, I can only provide help explaining how PHREEQC handles geochemistry, because I don't know the details of how PHREEQC is implemented in PHT3D. I don't think it is a complete implementation of PHREEQC capabilities as it is in PHAST.