Modeling and Analysis

Because of the large area of the algae farm and relatively shallow water (5 feet or less), a true scale model was not able to be created.  As a result we focused on modeling the overall shape of the algae farm and the input and outputs of the system.  By focusing on this aspect we could observe how the water would flow in the algae farm, and how the system would respond to higher than normal flow rates from the Yahara River and Dorn Creek.  Because we only had peak flow rates for the Yahara River and the Dorn Creek drains a smaller area, we will use the peak flow rate of the Yahara River for Dorn Creek.  This should be a conservative estimate.

For our model we decided to model Design #2.

Actual Design #2 Dimensions


d2

Modeled Design #2 Dimensions

s2 m

Actual Cross Sectional Areas
Yahara River = (275 ft) * (3 ft) = 825 ft²
Dorn Creek = (72 ft) * (3 ft) = 216 ft²
Area of the wall = (6117.3 ft) * (5 ft) = 30,586.5 ft²

Modeled Cross Sectional Areas (using a scaling factor of 247)

Yahara
River
= (1 in) * (2 in) = 0.013889 ft²
Dorn Creek = (0.25 in) *  (2 in) = 0.00694 ft²
Area of the wall = (3 ft) * (2 in) = 0.5 ft²

Actual Discharge Rates
Peak Total Discharge of Yahara River and Six Mile Creek = 1450 ft³/sec

Modeled Flow Rates (using a scaling factor of 247)
Peak Total Discharge of Yahara River and Six Mile Creek = 9.59E-5 ft³/sec or 2.71 cm³/sec

* During our model flow discharge calculations there was an error in our calculation and thus the discharge rate that was performed during the modeling process was incorrect.  We will expand on this issue later in the Modeling and Analysis section. For the time being, below is the corrected calculation of the model discharge rate needed for our model.

Corrected Model Discharge Rate

Instead of using the scaling factor of 247, as was used above to determine the discharge rate, we applied the dynamic property of a Froude number to determine the model discharge rate.

Velocity of Model = (Lenth of Model / Length of Actual)^(0.5)  * (Velocity of Actual)

Actual velocity with 2002 peak discharge of  1450 ft³/sec and cross sectional area of 650 ft² at measurement locality

Vact = (1450 ft³/sec) / (650 ft²) = 2.23 ft/s
Length of Model / Length of Actual = 1/247

Velocity of Model = (1/247)^(0.5) * (2.23) = 0.142 ft/sec

Area of Yahara input of Model = 2 in x 1 in = 0.01389 ft²

Corrected Model Discharge Rate =  Area * Velocity = (0.142 * 0.01389) = 0.00197 ft³/sec or 55.8 cm³/sec

*To be used in future modeling analysis

Modeling Trials

During our modeling, we performed the following three trails:

Video of Trial #1

Our first modeling trial essentially failed due to excessive leaks within the model.  We also didn’t             control the model output discharge rates of the Yahara River and Dorn Creek.  While many things went wrong, trial one was not a complete failure.  It proved that water can be discharged through the wall and we observed the dynamic flow of the system.

Video of Trial #2

Before our second trial we water proofed the model and prepared a system to provide the model peak flow rate of the Yahara River and Dorn Creek.  Our second model was very successful.  There were no major leaks and our design handled the peak flow rate without any build up or flooding occurring in the algae farm.  However it was later determined that the value we used for our model discharge rate was incorrect.  Future modeling will use the corrected model discharge rate.

Video of Trial #3

For the third trial, we attempted to determine the maximum model input flow that would cause overtopping of the wall.  However we were unable to provide a flow that could do this and had to settle with the maximum flow produced by the faucet, which was 309 ml/sec or an actual discharge rate of 8019.99 ft³/sec.

*Calculation of the maximum discharge value above was done using the reverse calculation of the Corrected Model Discharge Rate from above.

1
Model Before Trial # 2

2
Filling up the Model

3
View of the algae barier wall with water near the top of the surface.

4
View loocking down the barrier and algae being contained

5
Modeling the flow of the system

6
Alternative view from picture above.  Notice minimal leaking.  Picture taken 10 minutes after starting.

What we learned

Our model helped us to understand how our design would interact with the inputs from the Yahara River and Dorn Creek.  Using red food coloring we could see how the water would enter the algae farm and advance towards the wall.  Using parsley to represent the algae it could be observed that our system would prevent the floating near surface algae from leaving the farm, thus reducing algae levels in Lake Mendota.  Most importantly the model provided evidence that installing a wall in Lake Mendota would not be the cause of massive flooding in the event of a peak flow event.  With peak flow occurring in the Yahara River and Dorn Creek the wall design allowed water to move freely past it and also held in the algae.