Lecture 5
Lecture 5
Lecture 5
Fault detected
2
Air Handling Unit (AHU)
3
Shein and Bushby (2005) 4
Can you list sensors and actuators in an AHU?
5
AHU sensor
• Supply air temperature
• Return air temperature
• Mixed air temperature
• Outdoor air temperature
AHU actuator
• Heating / cooling coil valve
• Outdoor air damper
• Exhaust air damper
• Return air damper
AHU controller logic
• Supply air temperature control
• Supply airflow rate control
• Economizer control
• Heat recovery control
6
VAV
7
VAV control
Schein, Jeffrey & House, John. (2003). Application of Control Charts for Detecting Faults in Variable-Air-Volume
Boxes. ASHRAE Transactions. 109.
8
Sensors and actuators in a VAV
Source: H. Shahnazari, P. Mhaskar, J. House, T. Salsbury. “Heating, ventilation and air conditioning systems: Fault detection and isolation
and safe parking”. Comput. Chem. Eng. 2018 9
• Let’s imagine that AHU outdoor
air intake damper stuck open
• Changes in operation?
• Which sensors / actuators will start
behaving differently?
10
• Let’s imagine that a VAV terminal unit
reheat coil valve stuck
Changes in operation?
• Which sensors / actuators will start
behaving differently?
12
Point-by-point survey of HVAC
systems in Canal Building
13
14
16
17
18
19
• Many sensors, actuators, meters
• 2754 in Canal Building
• 6676 in River Building
• 9443 in Main Stats Building Tunney’s Pasture
• Facility managers do not have time to point by point search
for faulty sensors, actuators, meters
Need to analyze the building data and find anomalies
20
Faults are improper operating conditions
Symptoms are the observable changes in the operation in
sensors, actuators, and meters due to faults
FDD: Algorithms to flag anomalies and triage them based on
their importance
Diagnose
21
Gunay, Burak, Weiming Shen, and Chunsheng
Yang. "Characterization of a building's
operation using automation data: A review and
case study." Building and Environment 118
(2017): 196-210.
22
Visual inspection of anomalies
Hard faults
23
VAV 1 – Cooling season
*Loop error is the difference between the setpoint and the measured value
25
• Zone 6 does not seem to receive enough air?
• What could be the potential reasons?
1. Damper stuck closed
2. Airflow sensor is broken
3. Temperature sensor bias forcing damper to remain closed
26
Damper motor feedback signal
suggests that it is always open
27
The room temperature is persistently over 2°C over the setpoint.
The room does not seem to receive enough air.
Likely root-cause is a damper stuck closed fault
28
VAV 2 – Cooling season
29
30
• Zone 3 does not seem to receive enough air
• What could be the potential reasons?
1. Damper stuck closed
2. Airflow sensor is broken
3. Temperature sensor bias forcing damper to remain closed
31
• Zone 12 seem to receive lots of air
• What could be the potential reasons?
1. Damper stuck open
2. Airflow sensor is broken
3. Temperature sensor bias forcing damper to remain open
32
33
34
Anomalies ≠ Faults
36
Visual inspection
37
Time-series visualization
• Visual inspection of hourly/daily energy use can
help detect anomalies
38
Visual anomaly detection
39
Systematic approach
Prioritization
Isolation of
based on Anomaly
the root
impact on detection
cause
KPIs
40
Inverse modelling and FDD
41
42
43
1. Inverse modeling of AHU
44
Inputs:
1. Difference between return and outdoor temperature
2. Operating schedule of the AHU
3. Cooling-coil valve position
4. Supply air pressure temperature drop across the cooling coil
Output:
1. Difference between supply and return temperatures
46
48
49
2. Inverse modeling of VAV zones
50
Inputs:
1. Difference between indoor and outdoor temperatures
2. Horizontal solar irradiance
3. Work hour index
4. Thermal load from VAV
5. State of the perimeter radiators (on/off)
6. Average temperature of the surrounding core spaces
Output:
1. Rate of change in temperature
51
52
53
VAV pressure sensor fault Radiator Valve Stuck Off
57
• Detecting actuator faults in
presence of other systems
to help maintain the same
manipulated variable
• Reheat coil valve is broken
but temperatures are kept
at comfortable levels
because of the radiators
58
Soft faults
• Hard faults
• Failed sensors, actuators, meters, etc.
• Soft faults
• Mistakes in controls program
59
Supply Air Damper Position
Economizer for
free cooling
60
61
Bibliography
Gunay, Burak, Weiming Shen, and Chunsheng Yang. "Characterization of a
building's operation using automation data: A review and case study." Building and
Environment 118 (2017): 196-210.
Schein, Jeffrey, et al. "A rule-based fault detection method for air handling
units." Energy and buildings 38.12 (2006): 1485-1492.
Schein, Jeffrey, and Steven T. Bushby. "Fault detection & diagnostics for AHUs and
VAV boxes." ASHRAE Journal 47.7 (2005): 58.
Kim, Woohyun, and Srinivas Katipamula. "A review of fault detection and
diagnostics methods for building systems." Science and Technology for the Built
Environment 24.1 (2018): 3-21.
73