Graduation Year


Document Type




Degree Name

MS in Civil Engineering (M.S.C.E.)

Degree Granting Department

Civil and Environmental Engineering

Major Professor

Mahmood H. Nachabe, Ph.D.

Committee Member

Sarina Ergas, Ph.D.

Committee Member

Jeffrey Cunningham, Ph.D.


Sanitary Sewer Overflow, Sewer management, Sewer Rehabilitation, RDII, Infrastructure Improvement


Following large rain events, excess flow in sanitary sewers from inflow and infiltration (I/I) cause sanitary sewer overflows (SSO), resulting in significant problems for Pinellas County and the Tampa Bay area. Stormwater enters the sanitary sewers as inflow from improper or illegal surface connections, and groundwater enters the system as infiltration through cracks in subsurface infrastructure. This pilot study was designed to develop methods to separate and quantify the components of I/I and to build a predictive model using flowmeter and rainfall data.

To identify surface inflow, daily wastewater production and groundwater infiltration patterns were filtered from the flow data, leaving a residual signal of random variation and possible inflow. The groundwater infiltration (as base infiltration, BI) was calculated using the Stevens-Schutzbach method, and daily wastewater flow curves were generated from dry weather flow (DWF) data. Filtered DWF values were used to construct a range of expected residuals, encompassing 95% of the variability inherent in the system. Filtered wet weather flows were compared to this range, and values above the range were considered significant, indicating the presence of surface inflow.

At all 3 flow meters in the pilot study site, no surface inflow was detected, and the I/I was attributed to groundwater infiltration (as BI). Flow data from 2 smaller sub-sewersheds within the greater sewershed allowed analysis of the spatial variability in BI and provided a method to focus in on the most problematic areas. In the sub-sewershed with the shallowest water table and most submerged sanitary sewer infrastructure, an average of 56% of the average daily flow consisted of groundwater, compared to 44% for the entire study site.

Cross-correlation analysis suggests that rain impacts the water table for up to 9 days, with the highest impact 1 to 3 days after rain events, and the water table, in turn, impacts infiltration for up to 6 days. The highest correlation between rainfall and infiltration occurs 3 to 5 days after a rain event, which corroborates observations from Pinellas County that severe flows to the reclamation facility continue for 3 to 5 days after severe storms. These results were used to build a linear regression model to predict base infiltration (per mile of pipeline) during the wet season using the previous 7 days of daily rainfall depths. The model tended to under-predict infiltration response to large storm events with a R2 value of 0.52 and standard error of regression of 5.3.

The results of the study show that inflow can be detected using simple time series analysis instead of traditional smoke and dye testing. In this study site, however, groundwater infiltration is the only significant source of I/I. Additionally, water table and sewer invert elevations serve as useful indicators of potential sites of groundwater infiltration. Infiltration can be modeled as a function of the previous 7 days of rainfall, however simple linear regression cannot fully capture the complexity of the system response.