Shaun

You will need to run FEFLOW in solute transport mode that incorporates advection
and dispersion. You will then need to set the density factor to a value that
accurately reflects density of the brines in the lake. You should then set the
lake as a Type I flow boundary (ie specified head) and a Type I solute boundary
(ie constant salinity).

You are going to be disappointed with the operation of FEFLOW in this instance.

We have found that where you have sharp salinity gradients, as is the case here,
that your model will become unstable and you will get massive over-estimation
and under-estimation of salinity around the interface. This problem is
associated with numerical dispersion and can only be solved by using an
extremely fine mesh. In this regard there is a commonly used rule for cell size
based on the Peclet Number. I think the rule effectively indicates that the
size of your mesh cells must be less than half the dispersion coefficient that
you are using. In this type of model your dispersion coefficient should be less
than ten and hence your mesh size needs to be around 5m or smaller. We have
found that this effectively precludes running a large three dimensional model as
you propose. We have had success with running two dimensional models with very
small cells but this may not suit the geometry of your particular problem.

We have developed models of this nature in arid parts of Australia where
groundwater salinity can exceed that of seawater. Generally we have been
interested in the interaction of this groundwater with rivers that carry fresh
water.

Good luck

Brian


________________________________

From: feflow@... on behalf of shaunglorie
Sent: Thu 1/31/2008 6:59 PM
To: feflow@...
Subject: [feflow] Re: salt lake modelling



In the first instance we are trying to construct an idealized model.
The situation is we have a lake that is hyper saline and was formed
some time ago through evaporation. The groundwater flows towards the
lake and this prevents the salinity from moving far from the lake.
However we are looking at extracting groundwater from the region
adjacent to the lake and are interested in seeing what effect this
abstraction will have on the salt water interface.

The initial simplified model is a 3D block model looking at modelling
salt lake and two hydrological layers. These are a shale layer on top
of a sand aquifer. The extent of the model is 50km by 50km. the shale
layer is 150m thick(top) and sand is 50m thick. The salt lake was
modeled using a rectangular geometry 7.5km in width by 50km by 25m
depth.

At this stage we are not interested in any thermal flow. We have some
data on salinity gradients to calibrate the model. Our primary
interest is on the salinity gradients and their changes with
abstraction of groundwater. We are not modelling the evoporite
formations and we are considering the lake to be a constant source of
hypersaline water.

Any information on how to model and parameterize this type of problem
would be appreciated. We would especially appreciate receiving
similar examples.

Shaun

--- In feflow@... <mailto:feflow%40yahoogroups.co.in> ,
"hydromodeler" <hydromodeler@...>
wrote:
>
> Shaun,
>
> Some questions.
>
> Is it a rift basin and if YES, then are you concerned about the
> thermohaline flow? If you take out the heat, it becomes simpler.
> How have you conceptualized your problem in terms of 2-D or 3-D,
> cross-section, or what? What is the scale of the problem?
> Is this a scientific model where you'll be exploring model behavior
or
> are you looking to calibrate to heads or some other data?
> Do you have data on the salinity gradient?
>
> Is the top of you model flat or are you including advective forcing
> due to gravity, such as flow off of the flanks of a rift basin?
> Are you modeling the evaporite formation?
>
>
> --- In feflow@... <mailto:feflow%40yahoogroups.co.in> ,
"shaunglorie" <shaunglorie@> wrote:
> >
> > hey every one
> >
> > i was wondering if any one had any examples on how to model salt
lakes.
> > Preferably within Australia but not necessary. I am modeling an
aquifer
> > where the salt lake is an evaporation point and has become hyper
> > saline.
> >
> > Any information would be greatly appreciated.
> >
> > Thanks
> > Shaun
> >
>





SKM is committed to working with its clients to deliver a sustainable future for
all. Please consider the environment before printing this e-mail.
Notice - This message contains confidential information intended only for the
exclusive use of the addressee named above. No confidentiality is waived or lost
by any mistaken transmission to you. If you have received this message in error
please delete the document and notify us immediately.
Any opinion, text, documentation or attachment received is valid as at the date
of issue only. The recipient is responsible for reviewing the status of the
transferred information and should advise us immediately upon receipt of any
discrepancy.
All email sent to SKM will be intercepted, screened and filtered by SKM or its
approved Service Providers.

[Non-text portions of this message have been removed]