recharge from pit

A discussion forum for HYDRUS users. HYDRUS is a software package for simulating water, heat and solute movement in one-, two- and three-dimensional variably saturated media. Happy Posting!
Post Reply
devappa
Posts: 11
Joined: Thu Jan 18, 2024 4:21 am
Location: India

recharge from pit

Post by devappa » Sat Mar 30, 2024 11:57 am

sir,
i used hydrus 3d version 3.04 (Licensed version)to simulate infiltration from the small pit having dimensions of 85 cm deep and 65 cm diameter in 2D Domain . initially pit was filled upto 82 cm and water in the pit was noted for different time period. i run the model to optimize the hydraulic properties For sandy clay loam soil at the site. for this i considered axis symmetric domain ( for inverse solution : x column: time interval; Y column: depth of water in pit at that time; Type: time- Head, position 2, weight=1 for all data values) to run model but the fitted and observed values are not matching as attached in the photo file.

for more details i attached input data and project file, anyone please rectify the error and suggest to proceed


thanking you
Devappa
TNAU (India)
Attachments
inverse solution out put.JPG
inverse solution out put.JPG (97.3 KiB) Viewed 2142 times
inverse solutiom pit one.zip
(1.31 MiB) Downloaded 120 times
demo pit trial data.xlsx
(12.59 KiB) Downloaded 123 times

Jirka
Posts: 6079
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: recharge from pit

Post by Jirka » Sat Mar 30, 2024 9:14 pm

For a problem like this, you should use the reservoir BC (see the attached file). If you measured the location of the water level in the pit (reservoir), that’s location 1 (the first observation node, not the second one). Without any fitting, this is what I got.

J.
Attachments
Untitled.png
Untitled.png (19.99 KiB) Viewed 2132 times

Jirka
Posts: 6079
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: recharge from pit

Post by Jirka » Sat Mar 30, 2024 9:14 pm

The project. J.
Attachments
inverse solutiom pit one.rar
(1.11 MiB) Downloaded 114 times

devappa
Posts: 11
Joined: Thu Jan 18, 2024 4:21 am
Location: India

Re: recharge from pit

Post by devappa » Sun Mar 31, 2024 7:11 am

sir,
thanks a lot for clarifying errors in the boundary condition selection in previous.
in the image attached you have given diameter instead of radius of pit i think (Reservoir) and further i run the model with radius of pit 32.5 cm as per my case dimensions.

my another question is that,
1) Now i fill this pit (well type or reservoir) with small 6m stone chips around the hallow pipe at the center of pit having enough slots to infiltrate from sides and bottom. for this case is it necessary to assign boundary condition (BC) for slotted pipe ? if yes what may be the right BC?
i am thinking that no need to apply BC for pipe inserted to observe water level because due to enough open slots made water within pipe and stone chips as same water level.

2) Shoud i proceed with reservoir BC with two sub regions ( stone chips and soil region)
Attachments
well radius or diamter error.JPG
well radius or diamter error.JPG (59.28 KiB) Viewed 2128 times
23273cac-7e3f-4464-bbce-a92df69f968b.jpg
23273cac-7e3f-4464-bbce-a92df69f968b.jpg (234.31 KiB) Viewed 2128 times
2355ffc6-db3f-4b50-9790-6d78822f9750.jpg
2355ffc6-db3f-4b50-9790-6d78822f9750.jpg (134.14 KiB) Viewed 2128 times

Jirka
Posts: 6079
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: recharge from pit

Post by Jirka » Sun Mar 31, 2024 4:38 pm

These are the decisions that you have to make based on your understanding of the system. J.

devappa
Posts: 11
Joined: Thu Jan 18, 2024 4:21 am
Location: India

Re: recharge from pit

Post by devappa » Sat Apr 06, 2024 1:53 pm

sir, i run model for inverse solution with two material (1 sand to compensate gravel porosity & sandy clay loam) . after running the model why there is no correlation matrix and why its showing observed cumulative water loss depth only but not fitted cumulative water loss (instead showing depth of water in well for various time period recorded).in time variable B C window i have given water depth in well for different time . and in inverse solution window (X= time, y= cumulate water depth loss from well, type= time -head, position=1 i.e., observation node number as i measured water depth in well from bottom center).

i want to know
1) optimize the parameter,
2) observed vs fitted infiltration from well
Attachments
mini well 2M infil.zip
(1.11 MiB) Downloaded 117 times
observed infiltration depth losss and pressureead.JPG
observed infiltration depth losss and pressureead.JPG (116.37 KiB) Viewed 2017 times
no correlation matrixapril.JPG
no correlation matrixapril.JPG (111.32 KiB) Viewed 2017 times

Jirka
Posts: 6079
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: recharge from pit

Post by Jirka » Sat Apr 06, 2024 5:21 pm

In the project that I sent you previously, the project was correctly set up for parameter optimization. Here, I can see from your figures, that it is all wrong again. You need to learn from what I give you, otherwise I cannot help you.

This is the result from the project that I sent you that compares measured and fitted heads. You can obviously also see it from the figure in my previous posting.

J.
Attachments
Untitled.png
Untitled.png (25.15 KiB) Viewed 2011 times

devappa
Posts: 11
Joined: Thu Jan 18, 2024 4:21 am
Location: India

Re: recharge from pit

Post by devappa » Mon Apr 08, 2024 12:23 pm

sir,
sorry for the trouble,

initially i practiced the project without considering stone chips (Gravel) filled around the slotted pvc pipe (The project that u sent back after correcting Reservoir BC). after that i run the model for considering the stone chips filled around the pvc pipe (30 cm Diameter). Now i applied Reservoir BC (well radius=15 cm, depth to well bottom=-85 cm and initial water level in the well -3 cm) as u applied accordingly. now in Domain I created two surfaces with 2 different material (one as sand material to compensate stone chips high porosity (17.5 cm width) and second material as sandy clay loam soil). though i practiced as u corrected in the project and i couldn't able to get proper results for this.

first i run the model for forward run then there is no proper results (First project attached below) and second case with inverse run ( by giving no constraints for minimum and maximum value to get final optimized results). here there is no correlation matrix , no final optimized results, huge difference in observed and fitted values.

Note1: sir, there is no final optimized results i. e., non linear least square final results (alpha, theta r, theta s and Ks , n) in the project that u sent back to me earlier, though its showing good observed and fitted values

here with i am attaching forward run project and inverse project sedately.
1) what i did wrong here

after applying Reservoir BC even though it considers variable pressure head and seepage face BC internally i tried below BC conditions


when i did not get the results by considering seepage face at well bottom and side (Reservoir boundary condition), then in another two projects i run the model for seepage face at the well side wall and variable head 1 (Dirchlet BC) at bottom of well along with reservoir boundary condition (well radius=15 cm, depth to well bottom=-85 cm and initial water level in the well -3 cm). then i got the results somewhat similar as in the project that u sent me back. but i felt like result is wrong bcs at the end of the time (394 min) the water level in the well is 62.5 cm (observed data) the pressure head the seepage face just slightly above bottom is (-27.957 in pressure distribution). even though there is good observed and fitted values but why there is no correlation matrix. here with attach another to projects with bottom as variable pressure head boundary.

is this B C conditions wrong here?

Note2: in yesterdays project i tried to run hydrus by giving cumulative loss of water
Note3: in below attached files( miniwell 2M is forward run and miniwell 2M infil is inverse solution for Reservoir BC (seepage facet well bottom and side)) and another two projects are Reservoir BC with variable pressure head 1 at the bottom. i need to compare forward and inverse solution results
Attachments
mini well inverse.zip
(958.67 KiB) Downloaded 117 times
mini well 2M.zip
(850.01 KiB) Downloaded 119 times
mini well 2M infil.zip
(864.64 KiB) Downloaded 113 times
mini well .zip
(888.29 KiB) Downloaded 112 times

Jirka
Posts: 6079
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: recharge from pit

Post by Jirka » Mon Apr 08, 2024 5:46 pm

I will not have time to look at your projects (now four of them) again (if anyone else wants to do that, go ahead). I’m trying to help HYDRUS users with their initial steps (so that they are on the right path; and I have already done that in this case). I do not have time to spend hours on projects of individual HYDRUS users. If you need such help, you can request a paid service from PC-Progress. Jirka

devappa
Posts: 11
Joined: Thu Jan 18, 2024 4:21 am
Location: India

Re: recharge from pit

Post by devappa » Sat Apr 13, 2024 7:37 pm

sir ,
note: i am using 3D profesional licensed version 3.xx
i am applying seepage face Boundary condition first and then after in reservoir boundary condition window i am applying initial water level in the well and well bottom level (i mean Z coordinates respectively), radius of the well.

1) my question is , is this what reservoir boundary condition means actually?

or is there any optipn like Reserviour BC in the boundry condition options window (in list of BCS, No flux, seepage face, variable head 1, variable head 2, 3, 4, free drainage......... list)

2) confirm, the option variable head 1 is a Dirichlet BC.
3) axis symmetric domain in simulating well infiltration is 3D?

4) applying the bottom boundary as variable head 1 and well side as seepage face (also applying reservoir BC inputs in a window) provided some results but my confusion is whether it compulsary that bottom and side of well wall in 2d axisymmetric domain should be a seepage face while applying reservoir BC.

Jirka
Posts: 6079
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: recharge from pit

Post by Jirka » Sat Apr 13, 2024 11:43 pm

I have posted multiple test examples on a reservoir BC on the HYDRUS website (https://www.pc-progress.com/en/Default. ... eservoirBC). Have you looked at these examples?

I also provide a detailed description of Reservoir BCs in the Technical Manual (Section 2.7.2.10. Reservoir Boundary Conditions) and User Manual (Section 6.3.4. Reservoir Boundary Condition). Have you looked at these texts?

It seems to me that you can find answers to all your questions in these resources. J.

devappa
Posts: 11
Joined: Thu Jan 18, 2024 4:21 am
Location: India

Re: recharge from pit

Post by devappa » Mon May 27, 2024 2:50 am

thank you sir,

1) i used initial parameters from neural network prediction by known sand , silt , clay % values and average bulk density ((well top + well Bottom)/2), , and i want to optimize these values with inverse solution for the falling head, the problem with these initial value is that after running the model the estimated depth of water in the well is very less than the observed value. i adjusted Ks value to smaller value and i got par results of water levels to observed data but optimized value of theta s (saturated) is more than the porosity of soil.
Note : maximum value is 0 , i kept this to run as unconstrained and if i keep this porosity value as maximum i am getting same value in the optimized results.

2) sir, once after running the model i am getting water distribution in vertical and horizontal direction from the well but i am unable to find how much length it is distributed. i can draw dimensions of the geometry.
Attachments
Screenshot (3).png
Screenshot (3).png (426.46 KiB) Viewed 1602 times

Jirka
Posts: 6079
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: recharge from pit

Post by Jirka » Mon May 27, 2024 7:27 pm

Ad 1. I have no comments.
Ad 2. For this purpose, we have implemented in Version 5 of HYDRUS the tool “Measure Distance” between two selected positions:
Untitled.png
Untitled.png (383.36 KiB) Viewed 1564 times
However, since this tool was not available in Version 3, you will need to calculate the distance based on the coordinates of finite element nodes.
J.

devappa
Posts: 11
Joined: Thu Jan 18, 2024 4:21 am
Location: India

Re: recharge from pit

Post by devappa » Fri Jun 21, 2024 11:52 am

Sir ,
i have simulated water loss from well for the axis-symmetric domain but i need to design geometry in a 3D Soil block installed with well.

1) is it possible to design 3d well in soil block and can we apply reserviour BC for this? if yes pls give me steps to design a small well of diameter 100 cm and depth of 50 cm in a 500 cm square soil block
2) i run a simulation for a axis symmetrical domain and i have one observation point at the bottom of the well to know the depth of water in well for different times and as water infiltrates from well bottom and sides there will wetting front around the well so i also installed 3 tensiometers around the same periphery of well in 3 different depths ( records data for every 5 min), i want compare simulated results at the points with my tensiometer data so where to enter these tensiometer data with time in hydrus as input? or whether we have to compare these tensiometer data values with simulated resultes by creating observation nodes at those depths?
3) in axis symetric domain simulation i injected the well with a discharge of 0.0075 m3/sec for 10 min and i entered (- 0.0075 L3/T) in VAR.flux1 column for this 10 min, is it right way? once after filling the well i used diffrent depth of water ( after loss with time ) in well for inverse simulation and in this while applying Reservoir BC is it z coordinate of the initial water level is Z coordinate of the WELL bottom?
my observation values recorded at well bottom is diffrent time and tensiometer records in different time so is there chance togive all these tensiometer data as input as observed data?

Jirka
Posts: 6079
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: recharge from pit

Post by Jirka » Fri Jun 21, 2024 7:15 pm

I will not have time to answer your questions as I’m currently traveling in Africa.

However, I would want to stress one thing. This discussion forum is meant to help you start your project, it is not meant to solve all your problems. You need to do your own job as well. For example, the reservoir BCs are fully described in the technical manual.

J.

Post Reply