Solute mass balance

A discussion forum for Hydrus-1D users.
Post Reply
janite
Posts: 11
Joined: Wed Jul 07, 2010 7:20 am

Solute mass balance

Post by janite » Wed Aug 18, 2010 9:03 am

Hello,
I am modeling solute transport through unsaturated soil with the Unsatchem module in Hydrus 1D. The results are okay, apart from my Hydrogencarbonate. I get very big errors, but only for this solute. All other solutes are fine. I enclosed part of my results. It shows where my error comes from. My lower boundary condition is free drainage. What you can see is, that I have a lower flux, into my soil with HCO3. This is striking because, I do not have a water flux into that direction. So the the HCO3-flux should be the same as the K-flux: there should be no flux at that time.

--------------------------------------------------------
modelled solutes K HCO3
--------------------------------------------------------
Cum. upper flux [M/L2] -0.705E+00 -0.452E+02
Cum. lower flux [M/L2] 0.114E-24 <b> 0.226E+00 </b>
Mass in solution [M/L2] 0.705E+00 0.739E+02
Absolute error [M/L2] -0.225E-04 -0.335E+02
Relative error [%] 0.003 <b>45.360</b>
-----------------------------------------------------------

Any ideas or suggestions what might cause this?
I am not modelling root water uptake, no evatransporation, no co2-transport, no solids (calcite..) in the ground. Just a simple infiltration experiment.

Thanks in advance,
Jana

PS: Sorry for the bad formating of the table.

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Post by Jirka » Wed Aug 18, 2010 2:12 pm

I can not say. I would need to see the entire project (you can zip it (botht he folder and *.h1d file) and send to me). However, I'm leaving tonight for a two-week trip to Prague, and thus do not expect answers from me in the meantime.

Jirka

janite
Posts: 11
Joined: Wed Jul 07, 2010 7:20 am

Post by janite » Thu Aug 26, 2010 1:06 pm

I sent you the project.

I have one more question. I defined my mass unit as mg and my distance unit as mm. Now when I look at my concentration profiles the unit of concentration is meq/L. Is there a way for me to change the meq into mg?

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Post by Jirka » Fri Aug 27, 2010 10:41 am

I have briefly checked your project.
a) I believe that your input it rather unrealistic. You simply can not have soil solution (in the soil) that would have concentrations of all major ions (Ca, Mg, Na, K, HCO3, SO4, and Cl) equal to zero. UNSATCHEM is not written for some ideal solutions, but was written using real constants for real solutions. I am not at all surprised that it does not work well for your inputs.
b) You need to strictly use units (i.e., meq/L or mmol_c/L), for which the UNSATCHEM code was written. That is because a lot of chemical constants (e.g., for calculating activity coefficients do depend on these units). If you want to use other units than these, then you need to use the HP1 program (HYDRUS coupled with PHREEQC), which accept pretty much any units, since it has built in internal conversions.

Jirka

janite
Posts: 11
Joined: Wed Jul 07, 2010 7:20 am

Post by janite » Wed Sep 29, 2010 10:08 am

Thanks for your efforts.
As for a) Concentrations are neglectable in my test soil. I will give it a try with minor concentrations though. Otherwise I should probably not be using Unsatchem then.
b) My bad. I was relying on the Help-Section of Hydrus instead of having a close look into the manual. [:I]

I found that the Equil.Out file has an EC Value in it. THis time I checked the manual and the help but couldnt find any information. I am assuming this is the electrical conductivity? If so, what is its unit and how is it calculated?

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Post by Jirka » Wed Sep 29, 2010 2:21 pm

H1D Manual, p240:
Table 13.11. EQUIL.OUT - chemical information.
EC - Electric conductivity of the soil solution (dS/m).

(calculated using the method 3 of McNeal et al., 1970).
McNeal, B. L., J. D. Oster, and J. T. Hatcher, Calculation of electrical conductivity from solution composition data as an aid to in-situ estimation of soil salinity, Soil Sci., 110, 405-414, 1970.

J.

kuhlalex@msu.edu
Posts: 3
Joined: Wed Oct 05, 2016 6:57 pm
Location: USA

Re: Solute mass balance

Post by kuhlalex@msu.edu » Thu Sep 19, 2019 5:41 pm

Hello,

I am using HYDRUS1D with UNSATCHEM to model calcite dissolution for a multi-year hourly model. I am adding fairly dilute precipitation at the surface, with precipitated calcite below 1.5m depth and CO2 production turned on. I'm initializing the profile with the precipitation solution composition. I am not implementing any CEC. All concentrations are much less than 10 meq/L.

I was trying to calculate the EC myself using the method described in McNeal et al. 1970, using the output ion concentrations in Conc.out to compare to the EC output from Equil.out. Strangely, I get identical values below the calcite interface, but they are very different above the interface (see attached image). The shallow EC values output from Equil.out are much higher (~0.341 dS/m) than the values summed from the ions (~ 0.1 dS/m). Is it possible there is something else that is effecting the shallow EC besides the ions (CO2, VWC)?

This is the calculation I am using to estimate EC that is resulting in a discrepancy with the output EC, could anyone confirm if this is the appropriate implementation of method 3 in McNeal?

tot_Cat = Ca+Mg+Na+K [meq/L]
tot_An = HCO3+SO4+CL [meq/L]
EC = 0.055*Ca./tot_Cat+0.0414*Ca + 0.06*Mg./tot_Cat+0.0356*Mg + 0.023*Na./tot_Cat+0.0452*Na + 0.03*K./tot_Cat+0.066*K + 0.029*HCO3./tot_An+0.0348*HCO3 + 0.077*SO4./tot_An+0.0507*SO4 + 0.03*CL./tot_An+0.066*CL [dS/m]

As an aside, I also am wondering about the distinction between meq and mmol in the model - the UNSATCHEM section under the Solute Transport Parameters section of the application help says "Solution concentrations must be given in units of meq/L (i.e., mmol_c per L of soil solution).". Is the model accounting for the relationship between meq and mmol for the divalent ions Ca, Mg and SO4, since the McNeal equation uses meq/L?

I appreciate any and all insight into this frustrating issue!
EC.JPG
EC.JPG (24.4 KiB) Viewed 3011 times

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Solute mass balance

Post by Jirka » Thu Sep 19, 2019 6:24 pm

I do not have time to look at your calculations due to my current travel in Israel. Let me simply copy here the routine that is used in UnsatChem to calculate EC. J.

************************************************************************

* Electrical conductivity determination
* (see McNeal and Oster, S.S., 1970. refer to method 3)

subroutine ecal(eccal,TCa,TMg,TNa,TK,TSO4,TCO3,THCO3,TCl)
implicit real(a-h,k-z)
integer min1
dimension Ecions(9),Coeff(10,9)
data Coeff/
1 .0414,.055,.0355,.260,.0350,.400,.0350,.400,.0323,.940,
2 .0356,.060,.0269,.440,.0329,-.17,.0329,-.17,.0210,2.20,
3 .0452,.023,.0402,.270,.0373,.540,.0373,.540,.0306,1.85,
4 .0660,.030,.0620,.230,.0597,.460,.0597,.460,.0563,1.14,
5 .0507,.077,.0407,.585,.0332,1.22,.0332,1.22,.0268,2.50,
6 .0470,.070,.0382,.510,.0307,1.26,.0307,1.26,.0238,2.70,
7 .0348,.029,.0291,.315,.0291,.315,.0291,.315,.0291,.315,
8 .0660,.030,.0620,.230,.0597,.460,.0597,.460,.0563,1.14,
9 .0629,.183,.0492,.865,.0417,1.62,.0417,1.62,.0348,3.00/

CaT=TCa
MgT=TMg
SO4T=TSO4
MgSO4I=0.
SCaton=TCa+TMg+TNa+TK
SAnion=THCO3+TCO3+TSO4+TCl
if(TSO4-TCa.gt.0.) then
CaSO4I=TCa
TSO4=TSO4-TCa
if(TSO4-TMg.ge.0.) then
MgSO4I=TMg
else
MgSO4I=TSO4
end if
else
CaSO4I=TSO4
end if
Ecions(1)=TCa-CaSO4I
Ecions(2)=TMg-MgSO4I
Ecions(3)=TNa
Ecions(4)=TK
Ecions(5)=SO4T-CaSO4I-MgSO4I
Ecions(6)=TCO3
Ecions(7)=THCO3
Ecions(8)=TCl
Ecions(9)=CaSO4I+MgSO4I
eccal=0.
do 11 I=1,9
TempI=Ecions(I)
if(TempI.lt.0.) TempI=0.
Index=2*min1(5.,(1.+TempI/50.))
if(SCaton.gt.1.e-10)
! eccal=eccal+((Coeff(Index-1,I)*TempI+Coeff(Index,I)*TempI/
! SCaton))
11 continue
TCa=CaT
TMg=MgT
TSO4=SO4T
return
end

************************************************************************

kuhlalex@msu.edu
Posts: 3
Joined: Wed Oct 05, 2016 6:57 pm
Location: USA

Re: Solute mass balance

Post by kuhlalex@msu.edu » Fri Sep 20, 2019 8:04 pm

Jirka -

Thank you for the rapid response and the routine details. I was not accounting for the pairing of CaSO4 and MgSO4 and once I did I was able to better replicate the modeled EC curve. There is still some mismatch that may be due to the fact that CO3 is not output to Conc.out and therefore I'm not adding that in, but it does appear to be used in the subroutine.

I do have one quick follow up question if you have any chances to answer -

I see that SAnion is calculated but is not used to fractionate the anion intercepts in the do-loop. Is this for simplicity since SAnion and SCation are generally the same? I find I'm getting quite a different EC value near the surface when I distinguish between them.

Thanks!

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Solute mass balance

Post by Jirka » Fri Sep 20, 2019 10:02 pm

Yes. SAnion and SCation should be the same (to have the charge balance in the solution). J.

Post Reply