Illinois Environmental Protection Agency Crooked Creek TMDLs
Public Review Draft C-2
C.0 Estimating Existing Loads and Flows to the Crooked Creek Watershed Lakes
The ACOE BATHTUB model (Walker, 1987) was set up to simulate nutrient concentrations in the
impaired lakes in the Crooked Creek watershed using the second order nutrient response model. In a
separate application, the model was altered to simulate manganese concentrations using the fixed
sedimentation option.
C.1 Centralia Lake Watershed Loading
Annual flow rates to Centralia Lake were estimated by area weighting flows observed at USGS gage
05593575 on Little Crooked Creek near New Minden, Il. The Centralia Lake drainage area is
approximately 6.28 square miles and the drainage area to the Little Crooked Creek gage is 84.30 square
miles. Daily average flow rates at the gage were scaled down by 0.074 (6.28/84.30) to estimate daily
flows to the lake.
There are no permitted facilities discharging upstream of Centralia Lake (either directly into the lake or
into tributaries of the lake), however the Centralia Community Water Supply withdraws a portion of its
drinking water supply from Centralia Lake. The Centralia Community Water Supply has intakes in both
Centralia Lake and Raccoon Lake, and the average daily pumpage for both lakes combined is 3.9 MGD.
About 0.028 MGD are withdrawn from Centralia Lake, for an annual total of around 10.25 million
gallons (personal communication with Perry White at the Centralia Water Treatment Plant, 11/13/07).
This annual withdrawal was subtracted from the estimated flow rate to Centralia Lake.
C.1.1 Centralia Lake Sedimentation and Internal Loading
The “reverse” BATHTUB model was altered to simulate manganese concentrations in Centralia Lake
using the fixed sedimentation option. For a conservative estimate, the sedimentation rate was set to zero.
The model was then used to back calculate the loads required to simulate the observed concentrations.
The resulting load is equivalent to the external watershed load plus the net load resulting from
sedimentation and release from bottom sediments. Sufficient data to estimate the internal load separately
are not currently available.
C.1.2 Summary of Centralia Lake Inlake Water Quality Data
Typically, watershed loads are input to the BATHTUB model and average inlake concentration is output.
However, watershed and tributary data are not available to estimate loads to the lake. A limited number
of inlake observations of total manganese concentration have been collected across 5 months in 2001 at
one sampling location (ROI-1) and a limited number of inlake observations of total phosphorus
concentrations have been collected during 1998 and 2001 at three stations. A “reverse” BATHTUB
model was therefore applied where average inlake concentrations were used to estimate the load required
given annual flow volume and lake bathymetry data. No adjustment of the calibration factor was needed
with this simulation because the loads were set by year to match average observed concentrations. Table
C-1 summarizes the total manganese data by year, and Table C-2 summarizes the total phosphorus data
by year.
