Files
Abstract
Sanitary sewers often see an increased flow rate from wet weather, caused by inflow and infiltration [I/I]. There are many modeling techniques used to simulate I/I that are used by industry professionals; however, many of these models fail to account for the differences between inflow and infiltration. Both react to moisture already in the environment in unique ways. This antecedent moisture is a result of both precipitation and temperature and is not readily modeled in the methods commonly used, yet it has critical effects on the increased flow from I/I.
System identification is a method of building a mathematical model of a dynamic system. This method is capable of modeling an adapted unit hydrograph called a Linear Transfer Function for multiple rain events that occur close together, as well as modeling the difference in I/I generation caused by seasonal changes. Unlike other methods, system identification requires few parameters and just two time series, temperature and precipitation, to accurately model the increased flowrate, as well as determine the effects that antecedent moisture has on I/I, and how it changes seasonally.
This paper looks at the use of system identification to model the flowrate into two different types of sewer systems, a combined sewer, and a separate storm and sanitary sewer. The two systems were located within the same area and managed by the Milwaukee Metropolitan Sewerage District [MMSD]. The systems experience the same weather and rain events yet react in vastly different ways. This paper touches on the difference between the two systems.