Visualize Complexity, Discover Solutions, Shatter Limits
|
|
Visualize Complexity, Discover Solutions, Shatter Limits |
|
Case Study: Beaufort TMDL-Beaufort, SC |
|
|
Client: |
Beaufort-Jasper Water and Sewer Authority |
| Project: |
|
|
Issues:
|
South Carolina’s Beaufort and Jasper counties are home to
some of America’s most beautiful coastal areas, including the historic city
of Beaufort and Hilton Head Island. The area has seen booming growth, and to
keep up, Beaufort-Jasper Water and Sewer Authority (BJWSA), the region’s utility, is
constructing an entirely new multi-million dollar wastewater treatment
plant. Like the existing plant, the new plant will discharge into the
Beaufort River estuary, and needs a permit from SCDHEC. Because there are
other treatment plants on the river, SCDHEC required that a TMDL (Total
Maximum Daily Load) be established for the entire water body. A water
quality model must be constructed to determine a TMDL. It must defensibly
correlate the point source discharges to the river’s water quality. For the
Beaufort River, Dissolved Oxygen (DO) is the water quality parameter of
concern. Generally, water quality models of estuaries are the most difficult
kind to develop, often requiring several years to complete. |
| ADMi's Role: | BJWSA and the engineers for the new plant, worked with researchers from the USGS and ADMi to develop a Data Mining-based TMDL model. The USGS had collected two years of hydrodynamic and water quality data from a previously installed network of real-time gauging stations. To this were added ambient weather data, including rainfall, and point source discharges from the treatment plants. In less than a year, the model was completed and delivered to SCDHEC where it is now undergoing mandatory review. |
|
Results:
|
The model revealed some surprising and previously unknown facts about the
Beaufort River. The river is really a tidal slew with a small connector to
another tidal slew at its most inland point. The connector acts like a
siphon, drawing nutrient laden wastewater and rainfall runoff inland. This
greatly increases retention times and diminishes mixing with higher quality
seawater, making the estuary very sensitive to point and non-point sources.
For example, the salinity after one high rainfall event was seen to remain
depressed for two months. A second finding was that the smallest, but most
inland of the treatment plants had a relatively greater impact than the much
larger plants nearer the sea, underscoring the extreme role of the river’s
geometry and unusual hydrodynamics.
This project also showcases
software technology that is new to this type of application. The TMDL model
is made easy to use by a point and click user-interface and supporting
graphics (see below). No typing is required to input data or control the
model’s operation. It also has an integrated optimizer that automatically
computes the maximum allowed discharges for each simulation time step6. This
feature reduces the number of simulations needed to evaluate discharge
scenarios by over 90%. Also under development is an internet-based 3D
visualization program of the real-time data being transmitted by the USGS
gauging network, which is still in operation. This application also
automatically detects missing or erroneous data, and uses machine learning
to replace problem values with highly accurate estimates.
|
|
Home | Products | Services | Training | Technologies | News | Links | Contact |
Send mail to
info@advdataming.com with questions or comments about this web site.
Copyright © 2003 Advanced Data Mining, LLC
