The Detroit Water and Sewerage Department (DWSD) has undertaken the development of a
large water quality model for the entire city of Detroit. This model is estimated to be
approximately 25,000 pipes to simulate nearly 3,000 miles of water mains.
For the purpose of collecting data for the model calibration, an extensive field program was
implemented by the DWSD. The program used an innovative approach to conduct a series of
35 chlorine loss tests that were completed in conjunction with head loss tests. Using this
approach facilitated the direct calculation of the pipes' chlorine wall decay coefficients.
These tests were performed in isolated 500 to 2,000 foot pipe segments under varying flow
conditions. The majority of the pipes are 70 to 135 years old unlined cast iron pipes ranging
from 6-inch to 16-inch in diameter.
In completing the field program, the project team found that the chlorine data was exhibiting a
wall demand relationship that was different than the current formulation for water quality
modeling. Excessive loss of chlorine was measured in these pipes and analysis of the data
shows that the chlorine loss is directly related to the flow rate, i.e. the higher the flow rate, the
higher the chlorine loss in the pipe. This phenomena is counter to the mass transfer-limited
chlorine wall decay formulation which is widely used in most water distribution system models
such as EPANET.
The findings from the field tests suggest that an alternative chlorine wall demand formulation
may be needed to calibrate the water quality model for the Detroit distribution system. This
paper describes the procedures used in the field and the results of the field program.
Includes 9 references, tables, figures.