Comparative study between numerical and analytical groundwater modelling

Abstract (Swedish)

This study compares analytical and numerical models for groundwater modelling. The focus of
comparison is to investigate the change of the analytical results when there is only a partial
penetration of the aquifer. Two methods of comparison are used, the first one consisting of
calculating a radial distance from a shaft to the point where the drawdown is 0.3 meters with both
models. The second method used was by comparing the radius of influence, which was
determined through the numerical model as the radius inside which all groundwater recharge was
transported to the shaft. The second method gave the best results as the models become more
similar when examining the radius of influence through the groundwater recharge flow direction.
There was only a slight difference in result, between the numerical and analytical method when
there was a partial penetration with this method. The difference was deemed to be inside the
margin of insecurity. However, this study has not verified that the analytical model can be used
in all cases of partial penetration. It does however give an indication that it might be possible to
use it for soils with high hydraulic conductivity, such as sand.

Popular Abstract

Can pen and paper groundwater models compete with computer
programs?
Could something as simple as a pen and paper equation on the back of an
envelope compete with advanced and expensive computer programs?
Surprisingly, the answer could be yes for sandy soils.
When modelling how a construction project will affect the groundwater surface,
it’s easy to assume that the more advanced the model is, the better the result will
be. As it turns out, that’s not always the case. Advanced models are made from
many small parts, each of which has to be assigned accurate data for the model to
represent reality. This means that in order to create an advanced groundwater
model, highly detailed investigations are needed to determine how easily water
flows through each layer of soil.
However, if the conditions are simple, a one-dimensional model, basically an
equation you can solve with a pen and paper, can provide around the same
accuracy as the advanced computer model. These models are very easy and fast
to make, and they don’t require the same level of detail in the investigations as
they generalise the conditions for the area. Whereas making an advanced
computer model requires detailed studies, expensive computer programs and
many hours on the computer.
You might be wondering what the catch is. Well, the obvious answer is that
simpler models rely on more simplifications. For example, they always assume
that when water is pumped from a shaft, the bottom of the shaft is at the same
level as the bottom of the groundwater reservoir. This means that an impermeable
layer begins at the exact level as the shaft bottom, blocking water from flowing
vertically. In reality, this isn’t always the case.
This study investigated how much that simplification affects the results. Meaning
what would happen if the groundwater can travel below the bottom of the shaft.
This was investigated by comparing the results of the simpler and more advanced
models in sandy soils to see how their results would change when water travelled
below the shaft bottom. As it turns out, both types of models gave similar results.
If these simpler models can be used more than what they currently are, large
savings can be made in construction projects as less resources are needed, and
potential problems caused by a lowered groundwater level could be detected
earlier.
However, the only soil type used in this study was sand. It is possible that simpler
models could be used in other types of soil such as clay or till, but further studies
are needed to confirm or deny this.

• author: Lettevall, Joel
• supervisor: Daniel Erdal; Jan-Erik Rosberg
• organization: Engineering Geology; Civil Engineering (M.Sc.Eng.)
• alternative title: Jämförelsestudie mellan numerisk och analytisk grundvattenmodellering
• course: VTGM05 20242
• year: 2025
• type: H3 - Professional qualifications (4 Years - )
• subject: Technology and Engineering
• keywords: Numerical modelling, Analytical modelling, groundwater modelling, FEFLOW, Dupuit, drawdown, radius of influence
• publication/series: ISRN
• report number: ISRN LUTVDG(TVTG-5193)/1–47/(2025)
• language: English
• additional info: Examiner: Gerhard Barmen
• id: 9211405
• date added to LUP: 2025-09-02 13:04:33
• date last changed: 2025-09-02 13:04:33