By Hydrosimulatics INC  

System-based Water Supply Feasibility Analysis and Design

Background & Constraints

A university is planning to build an extension campus near a city. Capitalizing on their prestigious academic programs and worldwide reputation, and satisfying the educational needs for a rapidly growing urban and suburban population, the campus is designed for 45,000 students, professors, and supporting staff.

Planners have identified a potential site approximately 1 by 2 miles in size with very useable land that works well in terms of stormwater drainage, geotechnical stability, and transportation and infrastructure. But there may be a “drinking water supply issue”. Although there is a river going through the site area, water quality is a concern because of known pollution sites on/along the river. Also, the river freezes over in the winter.

Planners identified groundwater as the best or most cost-effective source because a very permeable aquifer exists beneath the campus site.  There are, however, a number of delicate, potentially limiting constraints:

To the northeast, about 3.5 miles away, is a special, highly valuable groundwater-fed wetland: the prairie fen. These types of wetlands harbor an extreme amount of biodiversity, including endangered species like the Mitchel Satyr butterfly and many other rare or valuable animals, insects and plants. No influence is allowed on the wetland and on the groundwater hydrology in the fen’s source water area. The change in the groundwater flow into the fen caused by pumping must be limited to less than 1% of the flux under natural condition (before the well field is installed).

To the north of the site, a short stretch of a trout stream gets quite close (2.5 miles) to the proposed site. The state Department of Environmental Quality (DEQ) will deny any application for pumping more which would cause significant streamflow depletion or adverse impacts. The change in the flux into the trout stream caused by pumping must be limited to less than 3% of the flux under natural condition (before the well field is installed). 

Farm lands of assorted crops exist to the west of the site about 3.5 miles away. These areas are known nitrate contamination hotspots.

South and southeast of the farm, about 2 miles away, is a residential area. Homeowners rely on their private water wells for drinking water supply. Any well interference (lowering of water levels) due to new large-supply wells is prohibited.

Groundwater in the aquifer – which is part of the master discharge area of the entire basin – sits on a pool of brine that is naturally inching upwards.  To avoid “pulling up” the salty/mineralized groundwater, the available drawdown needs to be limited to less than 20m based on a prior study in the area.

To control costs of the water intake system, the wells must be placed together in a "well field" on or close to campus.  The well field should be contained within a circular area with a diameter no larger than 4 miles. 



Objectives and Deliverables:

Perform a system-based water supply feasibility analysis to present to the campus planners. As part of your overall analysis, you should:

  • Under natural conditions:
    • Identify the source water area for the prairie fen
    • Delineate potential impact areas downstream from the farms (nitrate sources)
  • Design an optimal, sustainable pumping pattern within the given limiting factors
    • Assess changes in source water areas / potential impact areas in response to pumping;
    • Specify the change in flux to the Praire fens and trout streams (if any) due to pumping
    •  Estimate sustainable yield (maximum pumping rate that can be sustained without violating the ecological, hydrological, and legal constraints).
  • Delineate the 10-year wellhead protection area (WHPA) for the water supply wells; consider using a more accurate, nested submodel for WHPA delineation(s) 

Prepare a 1-2 page report that summarizes your approach and findings. You should discuss your findings with regards to responsibility for the contamination. Include any detailed model results / graphics in support of your conclusions in an appendix.


Given Information

The aquifer consists of relatively uniform fine-grained sands extending from the bedrock to the land surface, resulting in unconfined conditions. Both Lost Lake and Long Lake are deep and assumed to be strongly connected to the aquifer. The wetland functions as a “drain”, with the head at or above the land elevation year-round. The Trout Stream and River and partially connected to the aquifer and will exchange water to/from the aquifer depending on the local hydrogeologic conditions.

Both the proposed campus site and the residential areas have significantly reduced (or even negligible) infiltration of precipitation due to extensive impervious surfaces and infrastructure.

Field Data

The following information/data are available from a preliminary study:

  • Average Trout Stream stage: 39.3 ft
  • Trout Stream leakance: 5 m/day
  • Average Trout Stream depth: 1m
  • Average River stage: 10 ft
  • River leakance: 50 m/day
  • Average River bottom elevation: 0ft
  • Average water level from monitoring well on east edge of prairie fens: 40.5 ft
  • Leakance of wetland sediments: 1 day-1
  • Average Long Lake water level: 41.5ft
  • Average Long Lake depth: 38 ft
  • Long Lake bed leakance: 150 day-1
  • Average Lost Lake water level: 41ft
  • Average Lost Lake depth: 34ft
  • Lost Lake bed leakance: 150 day-1
  • Average aquifer thickness: 150 ft
  • Average annual recharge in non-residential areas and outside campus: 6.8 in./yr.
  • Average annual recharge in residential areas and inside campus: 1 in./yr.
  • Aquifer hydraulic conductivity: 100 ft/day
  • Aquifer effective porosity: 0.16
  • Typical per capita water use in the region: 165 gallons per day
  • Typical pumping rate for high-capacity drinking water supply well: 1000 GPM


Land surface elevations were measured at the points shown in the graphic below. Also included in the graphic are the elevation measurements at each location. 


MAGNET/Modeling hints

  • Use 'Synthetic mode' in MAGNET to create a model domain with the same dimensions as described above.  
    • Go to: 'Other Tools' > 'Utilities' > and click "Go to Synthetic Case Area' to access Synthetic mode. (Click OK to prompts that appear)
    • Once synthetic model domain appears, go to 'Utilities' > and click 'Geometry Locked' and then 'Geometry unlocked'. Then click anywhere inside the model domain. After answering OK to the prompts that appear, you will be able to click-drag any of the vertices to see the distance between vertices. NOTE: vertices are numbered and distances are indicated by d##, e.g., d21 is the distance from vertex 1 to vertex 2.
    • Once you have the correct dimensions, you can click 'Geometry Locked' once more to lock-in the shape. 
  • Overlay the provided SiteMap image file included at the top of the problem description page. 
    • Go to: 'Other Tools' > 'Utilities' > 'Overlay myImage' and follow the instructions in the Help Page ('?' button)
    • Click the 'Use Domain Extent' button to fix the image to the established domain size. (This should be after choosing the image file but before clicking 'Upload'.)
  • Conceptualize the model as 1-layer, unconfined aquifer.
  • Conceptualize both the River and the Trout stream as a two-way head dependent line features. Note that stages, depths/bed elevations, and leakances are provided above. 
  • Conceptualize the lakes as a two way head dependent flux using Zone features added to the domain.
  • Conceptualize the Prairie Fens as a one-way head-dependent zone feature. Leakance and seepage (drain) elevation are given above.
  • Use Zone features to assign zone-specific recharge values in the residential areas
  • Use a Zone feature to input the land elevation data (scatter points) to be spatially interpolated as the aquifer Top Elevation.
    • First, overlay the provided image file of the measurement locations (see top of problem description), following the same steps as you did when you overlaid the site map.
    • Then add a zone that stretches across the entire domain (make sure it is just inside the domain boundaries).
    • Go to the Elevation tab, check the box next to ‘Top Elevation’, select ‘Scattered Points’, and click the ‘…’ options button. A new submenu will appear.
    • Use the ‘Click to Add Data’ button to interactively add scatter points to the model domain. (You need to click this button each time you want to add a data point)
      • Each time you add a point, a row entry is added to the table. The second to last value in that (comma-separated) row is the observed elevation.
      • By default, the observed value is 0.0.Change this (edit the text) to the appropriate value for each entry (units: meters) using the information in the graphic provided above.
      • You can see where your measurement locations are on the map by clicking the 'Add Markers' button after adding data to the table. The 'Show Points' / 'Hide Points' button allows you to toggle back and forth between displaying and not displaying the points.

  • Use a large grid size (NX=80 or 100) to better capture the head dynamics at the well and to improve the water balance analysis.
  • Note that the lateral and bottom domain boundaries are treated as ‘no-flow’ boundaries.
  • Use particle tracking applications on your simulated flow patterns to track the movement of groundwater flow or potential contaminants.
    • For Welllhead Protection Area (WHPA) delineation ONLY, consider using a nested submodel containing your designed well field and the 10 year wellhead protection area.
    • Make sure the lateral boundaries of your submodel are sufficiently far from the wells so that you can perform 10 years of reverse particle tracking before the particles "hit" the submodel boundary
    • You can use a zone feature to specify the submodel boundary:  'Flow Property' tab > ' Zone Type' section, then check the box next to 'Submodel domain'.
    • Then go to Domain Attributes > 'Simulation Settings' tab > and click the box next to 'Boundary Condition from Parent Model.
    • Change your simulation length to 3650 days or 10 years (note: just change the simulation length, don't check the modeling transient flow box, your flow is steady state)
  • To analyze the head at a particular location in the model: 
    • Go to: 'Analysis Tools' > 'Analysis' > and click 'NodalValues'.  Then, when you use the cursor to click anywhere on the map, a "flag" marker will appear. 
    • When you click on the flag, model results for that location will be shown in a pop-up window, including the simulated head.