- Hi Frederic,

First, you need information about existing pumping in the area. That would

give you an idea and the range of pumping rate in the unsaturated aquifer. If

you are not able to find this kind of information, you could try to find some

about trial pumping in the area. The trial pumping will give you data about

drawdown, exploration radius, permeability, and porosity. which would allow

you to estimate the rate. Yet, the rate will depend on your pumping well

geometry, casing, and also kind of pump.

You need to take care of existing pumping close to your area, in order to

keep the waterlevel high enough for the other wells.

In a productive aquifer, you can go up to 700 m3/h and perhaps a bit more

(this is just my experience), but you can also go down to 50 m3/h. this why

you need information about other pumpings in this aquifer.

ALexis

Alexis Valenza

SOL SYSTÈMES

Éden Parc "Les Cèdres" - Square Vilmorin

06160 Juan les Pins - France

alexisvalenza@...

[Non-text portions of this message have been removed] - The first step is to estimate aquifer hydraulic properties and external

stress conditions (as negatively suggested by dod).

Second, develop a calibrated flow model.

The third step is to formulate the 'hydraulic optimization problem'. This

involves three steps:

1.) define decision variables - in your case well pumping rate (and

maybe location) and water table elevation.

2.) define constraints - e.g., the water table cannot be lowered more

than some level at a number of predefined points (nodes of the mesh). In

addition, you can define a number of 'candidate' extraction well locations.

3.) define objective - e.g., what is the highest rate that can I pump a

well such that I do not lower the water table below the value defined at the

constraint points?

The forth step is to get an optimization package that can solve the problem

defined above. Basically, it systematically searches for the stress

configuration that meets the objective by sequentially setting well location

and rate, calling the flow model to solve for the resulting flow field, and

assessing whether the constraints are met. The algorithm stops when either

the objectives are met (feasible solution), or cannot be met for any

configuration (infeasible solution).

I know of two free optimization packages that can solve your problem. They

come with complete documentation that efficiently walk you through theory

and practice. The only issue is that they have been written to work with

MODFLOW (they have internal calls to MODFLOW to solve the flow problem for

each stress configuration). So you will have to develop a calibrated MODFLOW

model to represent the physical problem. You can find them at:

http://www.ecs.umass.edu/modofc/

http://www.geotransinc.com/modman.html

Joe Guarnaccia

-----Original Message-----

From: Daniel O'Donnell [mailto:dodonnell@...]

Sent: Saturday, June 29, 2002 8:55 AM

To: gwmodel@yahoogroups.com

Subject: Re: [gwmodel] Quantification of water in an unsaturated aquifer

Not much to go on Fred. Have you done the hydrogeologic work up of the

area? Have you done the aquifer testing? If not, hire a professional

with experience, http://home.att.net/~dodonnell/odonnell.html

If you have done the above and still don't know what to do, best of luck

to your client...dod

Frederic Ghogomu wrote:>

To unsubscribe from this group, send an email to:

> Hi everyone,

> I am working on a project where I have to quantify

> water that can be pumped from an unsaturated aquifer.

> I am using FRAC3DVS. This model gives an option where

> you can place pumping well with specified rate. But

> the aim of our study is to determine this rate. I will

> apreciate any advice on the procedure to use or how to

> handle this problem.

> Thanks

>

> Fred Ghogomu

>

> __________________________________________________

> Do You Yahoo!?

> Yahoo! - Official partner of 2002 FIFA World Cup

> http://fifaworldcup.yahoo.com

>

>

> To unsubscribe from this group, send an email to:

> gwmodel-unsubscribe@yahoogroups.com

>

>

>

> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

gwmodel-unsubscribe@yahoogroups.com

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ - I am troubled by Fred's approach and agree with Daniel that a basic

hydrogeologic understanding is essential BEFORE modeling. The wording

of the original question suggested little if any hydrogeologic

information was available.

--- In gwmodel@y..., Frederic Ghogomu <fred_ghogomu@y...> wrote:

> I am working on a project where I have to quantify

> water that can be pumped from an unsaturated aquifer.

I don't know if this statement concerns anyone else, but an aquifer

by definition is saturated. How can an aquifer be unsaturated? If one

were to accept a definition allowing for a portion to be unsaturated,

there would not be one value of hydraulic conductivity; there would

instead be characteristic curves for conductivity determined by

wetting or drying conditions.

> I am using FRAC3DVS. This model gives an option where

> you can place pumping well with specified rate. But

> the aim of our study is to determine this rate. >

Are there any wells in this aquifer? How deep are they? Are they

completed in the same kind of material as the proposed well? Will the

well fully penetrate the aquifer?

If this aquifer is "unsaturated", what portion of it is saturated and

do those conditions change with seasons? Are there barriers to flow

that could affect production? Are there surface water bodies that

could affect or be affected by production?

__________________________________________________ - I agree with Tom. A conceptual model needs to be constructed. FRAC3DVS

is a finite-element variably-saturated flow and transport model. MODFLOW

approach will not work for this model. Unfortunately these days,

groundwater modeling implies applying MODFLOW. There are other types of

models besides MODFLOW. In FRAC3DVS, a well can be screened such that

the screen interesects both saturated and unsaturated zones. Obviously,

the saturated zone will contribute most of the well discharge.

It is essential that prior to model selection, one develops a

conceptual model. A conceptual model will lead to a thorough

understanding of hydrogeology, data gaps, processes taking place in the

flow domain, boundary and initial conditions, etc. Armed with a

conceptual model, one can then select an appropriate model. This

approach will lead to application of appropriate models and reliability

of model results. ASTM has good standards for application of groundwater

models.

If FRAC3DVS is approriate, there is a need to acquire a thorough

understanding of the physical problem and the model.

>>> tcudzilo@... 07/01/02 04:37PM >>>

I am troubled by Fred's approach and agree with Daniel that a basic

hydrogeologic understanding is essential BEFORE modeling. The wording

of the original question suggested little if any hydrogeologic

information was available.

--- In gwmodel@y..., Frederic Ghogomu <fred_ghogomu@y...> wrote:

> I am working on a project where I have to quantify

> water that can be pumped from an unsaturated aquifer.

I don't know if this statement concerns anyone else, but an aquifer

by definition is saturated. How can an aquifer be unsaturated? If one

were to accept a definition allowing for a portion to be unsaturated,

there would not be one value of hydraulic conductivity; there would

instead be characteristic curves for conductivity determined by

wetting or drying conditions.

> I am using FRAC3DVS. This model gives an option where

> you can place pumping well with specified rate. But

> the aim of our study is to determine this rate. >

Are there any wells in this aquifer? How deep are they? Are they

completed in the same kind of material as the proposed well? Will the

well fully penetrate the aquifer?

If this aquifer is "unsaturated", what portion of it is saturated and

do those conditions change with seasons? Are there barriers to flow

that could affect production? Are there surface water bodies that

could affect or be affected by production?

__________________________________________________

To unsubscribe from this group, send an email to:

gwmodel-unsubscribe@yahoogroups.com

Your use of Yahoo! Groups is subject to

http://docs.yahoo.com/info/terms/