Negative sorption coefficient?

Hello all,

I have modeled a contaminant transport from a spill in feflow and got some results. My first setup was implementing no sorption. In second simulation, I applied k=0.2 as Henry (linear) sorption coefficient and surprisingly the concentrations were higher than initial case with no sorption, which is opposite to the expectation.
I tried k=-0.2 and the results make sense, but I could not find a reference to this. Anyone know if we actually should put the sorption coefficient in negative?

It's hard to judge why you would get higher concentrations in the fluid, but the sorption coefficient should not be negative.

Please make sure, however, that the quantity of the sorption coefficient is defined correctly. FEFLOW defines the Henry sorption using the sorptivity coefficient, which is the distribution coefficient kd multiplied with rhos (the density of the solid phase). You can find details in the FEFLOW book on page 131.