Date of Award

Spring 1-1-2015

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Civil, Environmental & Architectural Engineering

First Advisor

John Zhai

Second Advisor

Michael Brandemuehl

Third Advisor

Moncef Krarti

Abstract

Although high efficiency filter is one critical component in the Air Handler Unit (AHU), HVAC system is potential contaminant emission source. Released contaminants can be transported through HVAC system and impacts the indoor air quality (IAQ). Effective control and improvement measures are required to remove the contaminant source located in HVAC systems in order to eliminate its influence on the IAQ. Accurate and fast identification of contaminant sources in HVAC systems makes it. This thesis studies the application of adjoint backward probability model in identification of contaminant source in Building HVAC system. The adjoint backward probability model was mostly applied to identify contaminant source information in groundwater and inside building. According to the similar properties between water and air, and same contaminant transport fate in water and air, the adjoint probability model is applied to study the contaminant source identification in HVAC systems.

Sensors are used to detect contaminant concentration change in certain sampling locations of HVAC ductwork. Using sensor detection information, we can trace back and find the source information. In this research CONTAM is used to provide a steady state airflow field. A simple building model with three zones and detailed duct work is built. This model is applied into later research in identification of contaminant source in HVAC system.

Four cases are analyzed in the research to study the application of adjoint backward probability method. The first case is identifying an instantaneous contaminant source location with known source release time and source release mass. The second case is identifying the location of a dynamic contaminant source with known release time and known release mass. The third case is identifying source release time and release location simultaneously for a decaying contaminant source with known source release mass. The fourth case is identifying the location of a dynamic contaminant source in a two-floor building with known release time and known release mass. The conclusions come to that a sensor network with two sensors reading historical concentrations can identify source information accurately. Further, in future research, contaminant source information will be recovered without knowing any source information in advance.

Share

COinS