University of Southern California Dissertations and Theses

Economizer performance and verification: effect of human behavior on economizer efficacy and thermal comfort in southern California

(USC Thesis Other)

Economizer performance and verification: effect of human behavior on economizer efficacy and thermal comfort in southern California

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ECONOMIZER PERFORMANCE AND VERIFICATION:
EFFECT OF HUMAN BEHAVIOR ON ECONOMIZER EFFICACY AND THERMAL
COMFORT IN SOUTHERN CALIFORNIA

By

Tighe Glennon Lanning

______________________________________________________________________________

A Thesis Presented to the
FACULTY OF THE USC SCHOOL OF ARCHITECTURE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF BUILDING SCIENCE

August 2013

Copyright 2013 Tighe Glennon Lanning

II

Dedication

This thesis is dedicated unconditionally to my family. The findings and associated
insights stemming from this research were realized only through their unwavering support and
encouragement.
Thank you Mom, and thank you Dad.

III

Acknowledgements

I would like first and foremost to acknowledge and give thanks to Professor Karen
Kensek for her involvement in this research. Karen’s guidance, creativity of thought, and
boundless technical advice can be found on every page herein. Karen, your continued assistance
breathed life into this endeavor every step of the way.
I would also like to acknowledge Tim Kohut for the many roles he played, each integral
to the success of this thesis. Thank you for believing in me, and in the fierce pursuit of a better
built environment. If the architectural profession was as ahead of its time as you are, it would not
be playing catch up.
I would like to thank Professor Joon-Ho Choi both for his technical support and for his
guidance in maintaining the structural clarity of this document. I would also like to thank
professor Murray Milne for providing the original inspiration upon which this thesis was
developed.
Finally I would like to thank the remaining USC faculty members who helped bring this
thesis to fruition. Particular appreciation goes to Professor Douglas Noble, Ilaria Mazzoleni, and
Jeff Landreth for keeping me on track. In addition I would like to thank the Los Angeles
Community Development Commission for their support, and the tenants who opened their
homes to me over the previous year.

IV

Table of Contents
Dedication ..................................................................................................................................... II
Acknowledgements ..................................................................................................................... III
List of Figures ........................................................................................................................... VIII
List of Tables ............................................................................................................................... XI
Terms and Abbreviations ......................................................................................................... XII
Abstract .......................................................................................................................................... 1
Hypothesis ...................................................................................................................................... 2
Chapter 1: Introduction ............................................................................................................... 3
1.1 - Problem ............................................................................................................................... 4
1.2 – Passive Design .................................................................................................................... 7
1.3 – Economizers ..................................................................................................................... 11
1.4 – Case Study Building ......................................................................................................... 13
1.5 – Objective ........................................................................................................................... 15
Chapter 2: Background Research ............................................................................................. 16
2.1 - HEED and Economizers ................................................................................................... 16
2.1.1 – Previous HEED Studies Using Economizers ............................................................ 16
2.1.2 – Considerations when Using Economizers ................................................................. 17
2.2 – Case Study Comparison, Economizers and Air Conditioners .......................................... 19
2.3 - Data Loggers ..................................................................................................................... 21
2.4 - Human Behavior ............................................................................................................... 21
2.5 - Weather data ...................................................................................................................... 22
2.6 - Conclusions from background research ............................................................................ 23
Chapter 3: Methodologies .......................................................................................................... 24
3.1 - Scope of Work ................................................................................................................... 24
3.2 - Phase 1: Site Documentation and Experiment Setup ........................................................ 27
3.2.1 - Construction Verification ........................................................................................... 31
3.2.2 - HVAC Verification .................................................................................................... 32
3.3 - Phase 2: Climatic Calibration ............................................................................................ 42
3.3.1 - Typical Meteorological Year Data ............................................................................. 44
3.3.2 - Weather Station Data .................................................................................................. 45
3.3.3 - Collected Outdoor Data .............................................................................................. 46
3.4 - Phase 3: Occupant Behavior ............................................................................................. 48

V

3.4.1 - Tenant Education ........................................................................................................ 50
3.4.2 - Locked Thermostat ..................................................................................................... 51
3.4.2 - Vacant Unit Test ......................................................................................................... 52
3.5 - Overview ........................................................................................................................... 52
Chapter 4: Results....................................................................................................................... 54
4.1 - Scope of Work ................................................................................................................... 54
4.2 - Phase 1: Site Documentation and Experiment Setup Results ........................................... 61
First Floor Units:.................................................................................................................... 62
Second Floor Units: ............................................................................................................... 65
Third Floor Units: .................................................................................................................. 67
Outdoor Units: ....................................................................................................................... 70
4.2.1 - Construction Verification ........................................................................................... 72
4.2.2 - HVAC Verification .................................................................................................... 73
4.3 - Phase 2: Climatic Calibration Results ............................................................................... 84
4.3.1 - Typical Meteorological Year Data ............................................................................. 84
First Floor Units:.................................................................................................................... 84
Second Floor Units: ............................................................................................................... 87
Third Floor Units: .................................................................................................................. 88
4.3.2 - Weather Station Data .................................................................................................. 89
First Floor Units:.................................................................................................................... 89
Second Floor Units: ............................................................................................................... 91
Third Floor Units: .................................................................................................................. 92
4.4 - Phase 3: Occupant Behavior Results ................................................................................. 93
4.4.1 - Tenant Education ........................................................................................................ 93
4.4.2 - Locked Thermostat ..................................................................................................... 96
4.4.3 - Vacant Unit Test ......................................................................................................... 97
4.5 – Results Overview .............................................................................................................. 98
Chapter 5: Analysis and Discussion .......................................................................................... 99
5.1 - Overview ........................................................................................................................... 99
5.2 - Phase 1: Site Documentation and Experiment Setup Analysis ......................................... 99
First Floor Units:.................................................................................................................. 102
Second Floor Units: ............................................................................................................. 105
Third Floor Units: ................................................................................................................ 107
5.2.1 - Construction Verification ......................................................................................... 112

VI

5.2.2 - HVAC Results and Analysis .................................................................................... 112
5.3 - Phase 2: Climatic Calibration Analysis ........................................................................... 122
5.3.1 - Typical Meteorological Year Data ........................................................................... 124
5.3.2 - Weather Station Data ................................................................................................ 125
5.3.3 - Collected Outdoor Data ............................................................................................ 127
5.4 - Phase 3: Occupant Behavior Analysis ............................................................................ 128
5.4.1 - Tenant Education ...................................................................................................... 128
5.4.2 - Locked Thermostat ................................................................................................... 129
5.4.2 - Vacant Unit Test ....................................................................................................... 130
Chapter 6: Conclusions and Recommendations .................................................................... 131
6.1 – Overview ........................................................................................................................ 131
6.2 - California Energy Code Recommendations .................................................................... 134
6.2.1 - Incentives .................................................................................................................. 134
6.2.2 - Combined Air Conditioner/Economizer .................................................................. 135
6.3 - LEED Recommendations ................................................................................................ 138
6.3.1 - Commissioning ......................................................................................................... 139
6.3.2 – LEED-Specific Education Program ......................................................................... 140
6.4 - Education-Based Recommendations ............................................................................... 140
6.4.1 - Building Management Education ............................................................................. 141
6.4.2 - Tenant Education ...................................................................................................... 142
6.5 - Software Recommendations ............................................................................................ 144
6.5.1 – Occupant Behavior .................................................................................................. 144
6.5.2 - Weather Data Review ............................................................................................... 145
6.6 - Hardware Recommendations .......................................................................................... 146
6.7 - Thermostat Design Recommendations: .......................................................................... 146
6.7.1 - Feedback ................................................................................................................... 147
6.7.2 - Simplicity ................................................................................................................. 148
6.7.3 – Specificity ................................................................................................................ 149
Chapter 7: Future Work .......................................................................................................... 151
7.1 – Thermostat Design, User Interface ................................................................................. 151
7.1.1 – Feedback .................................................................................................................. 153
7.1.2 – Constraints ............................................................................................................... 156
7.1.3 – Monitoring ............................................................................................................... 157
7.2 – Software Development ................................................................................................... 157

VII

7.2.1 – Weather Analysis Tool ............................................................................................ 159
7.3 – Additional ....................................................................................................................... 159
Bibliography .............................................................................................................................. 160

VIII

List of Figures
Figure 1: Hourly Cooling Loads per Area, Load per Square Area ................................................. 6
Figure 2: Climate- Specific Vernacular Architecture, Charleston Single Style ............................. 8
Figure 3: Climate- Specific Vernacular Architecture, Adobe Pueblo Housing .............................. 9
Figure 4: Evaporative Cooler, Left, and Schematic Diagram, Right ............................................ 10
Figure 5: Temperature Patterns, Economizer, Left, and High/Low Mass, Right. Edited. ............ 12
Figure 6: Typical Air Economizer Diagram ................................................................................. 13
Figure 7: California’s Climate Zones ........................................................................................... 14
Figure 8: Simulated EUI Values, Economizers and Air Conditioners (HEED Simulation) ........ 20
Figure 9: Graphic Abstract............................................................................................................ 24
Figure 10: Photo of Courtyard, Looking North ............................................................................ 26
Figure 11: Split-Core AC Current Sensor, Left, HOBO Data Logger, Right ............................... 29
Figure 12: Plan of Typical Apartment Unit Used for Testing ...................................................... 30
Figure 13: Photo of Apartment Unit Used for Testing ................................................................. 30
Figure 14: Typical Floor Plan (Preceding Number Corresponds to Respective Subchapter) ...... 32
Figure 15: Typical Pressure Relief Damper at Living Room, Opened for Photograph ................ 34
Figure 16: Flow Hood Measuring Air Flow at Supply Dampers ................................................. 34
Figure 17: Smoke Gun Warming up for Testing .......................................................................... 36
Figure 18: Economizer Diagram, Edited to Reflect Case Study Building Conditions ................. 39
Figure 19: ASHRAE Figure 5.2.3 “Air Speed Required to Offset Increased Temperature” ....... 40
Figure 20: Perspective View of Example HEED Model .............................................................. 42
Figure 21: TMY (CZ 8) Temperatures and WSD (Pink Overlay), June – September 2012 ........ 44
Figure 22: Typical Meteorological Year (TMY) Data, June – September 2012 .......................... 45
Figure 23: Weather Station Data (WSD), June – September 2012............................................... 46
Figure 24: Outdoor Data Logger Temperatures, Various Locations, June – September 2012 ..... 47
Figure 25: Data Logger at Roof, Left, and at Courtyard Vegetation, Right ................................. 48
Figure 26: Excerpt from Management-Provided “Green Living Guide” ..................................... 50
Figure 27: Locked Thermostat, Left, Economizer’s “Override” Switch, Right ........................... 52
Figure 28: 1
st
Floor Plan, Not to Scale .......................................................................................... 55
Figure 29: 2
nd
Floor Plan, Not to Scale ......................................................................................... 56
Figure 30: 3
rd
Floor Plan, Not to Scale ......................................................................................... 57

IX

Figure 31: Roof Plan, Not to Scale ............................................................................................... 58
Figure 32: North-Facing Unit B115 Floor Plan, 850 SF, Not to Scale ......................................... 59
Figure 33: West-Facing Unit (A109, A209, A305, A309) Floor Plan, 1,040 SF, Not to Scale ... 60
Figure 34: South-Facing Unit (B100, B200) Floor Plan, 1,315 SF, Not to Scale ........................ 61
Figure 35: Unit A109 (1
st
Floor, West-Facing) ............................................................................ 62
Figure 36: Unit B100 (1
st
Floor, South-Facing)............................................................................ 63
Figure 37: Unit A209 (2
nd
Floor, West-Facing)............................................................................ 65
Figure 38: Unit B200 (2
nd
Floor, South-Facing) ........................................................................... 66
Figure 39: Unit A305 (3
rd
Floor, West-Facing) ............................................................................ 67
Figure 40: Unplugged HOBO Data Logger .................................................................................. 68
Figure 41: Unit A309 (3
rd
Floor, West-Facing) ............................................................................ 69
Figure 42: Outdoor Temperatures, Outdoor Data Loggers, WSD File......................................... 71
Figure 43: Outdoor Data Loggers, WSD File, TMY File, June – September 2012 ..................... 72
Figure 44: Typical Exhaust Panel, Closed Position ...................................................................... 74
Figure 45: Barometric Dampers within Exhaust Panels Screwed Shut and Inoperable ............... 74
Figure 46: HVAC Duct Layout, Unit AX09 (All Floors) 1,040 SF, Not to Scale ....................... 75
Figure 47: HVAC Duct Layout, Unit A302, 663 SF, Not to Scale .............................................. 77
Figure 48: GLAD Press’n Seal™ Plastic Wrap Covering Return Duct ....................................... 80
Figure 49: Smoke Gun Testing Air Leakage at Economizer Ductwork ....................................... 81
Figure 50: Outside-Air Damper Opening in Response to Lowering Set Point............................. 82
Figure 51: Kestrel Anemometer, Typical Thermostat in Case Study Apartments ....................... 82
Figure 52: Outdoor-Air Damper Being Tested for Airtightness ................................................... 83
Figure 53: Predicted Temperatures, TMY File, Unit A109 (1
st
Floor, West-Facing) .................. 85
Figure 54: Predicted Temperatures, TMY File, Unit B100 (1
st
Floor, South-Facing) ................. 86
Figure 55: Predicted Temperatures, TMY File, Unit A209 (2
nd
Floor, West-Facing) ................. 87
Figure 56: Predicted Temperatures, TMY File, Unit A305/A309 (3
rd
Floor, West-Facing) ........ 88
Figure 57: Predicted Temperatures, WSD File, Unit A109 (1
st
Floor, West-Facing) .................. 89
Figure 58: Predicted Temperatures, WSD File, Unit B100 (1
st
Floor, South-Facing) ................. 90
Figure 59: Predicted Temperatures, WSD File, Unit A209 (2
nd
Floor, West-Facing) ................. 91
Figure 60: Predicted Temperatures, WSD File, Unit A305/A309 (3
rd
Floor, West-Facing) ........ 92
Figure 61: Tenant Education- Temperature and Economizer Amperage, Unit A109 .................. 94

X

Figure 62: Tenant Education- Temperature and Economizer Amperage, Unit A309 .................. 95
Figure 63: Locked Thermostat- Temperature and Economizer Amperage, Unit A309 ............... 96
Figure 64: Locked Thermostat (Example) .................................................................................... 96
Figure 65: Existing Conditions- Temperature and Economizer Amperage, Unit A309............... 97
Figure 66: Results- Vacant Unit Test, Unit A109......................................................................... 98
Figure 67: Average High and Low Temperatures, Units A109, A209, A309, Outdoor (WSD) 111
Figure 68: Schematic Diagram of Duct Leakage Behavior ........................................................ 116
Figure 69: Schematic Diagram of Sealed Ducts ......................................................................... 116
Figure 70: Supply and Return Air Temperatures, Unit A302 ..................................................... 118
Figure 71: Adaptive Thermal Comfort Model ............................................................................ 133
Figure 72: Typical Daily Load Curve ......................................................................................... 136
Figure 73: Peak Shaver and Valley Deepener ............................................................................ 137
Figure 74: Set Point Errors Seen at Thermostats ........................................................................ 143
Figure 75: Toyota Prius Consumption Display .......................................................................... 149
Figure 76: Carrier Debonair's Display Functions ....................................................................... 152
Figure 77: Conceptual Thermostat, Comfort Range 65 F - 78 F, Predicted Efficacy: 81% ....... 155
Figure 78: Conceptual Thermostat, Comfort Range 62 F - 83 F, Predicted Efficacy: 62% ....... 155

XI

List of Tables
Table 1: 09/13/12 Flow Test, Corresponding to Plan in Figure 46 .............................................. 76
Table 2: Air Changes per Hour (ACH), Based on Flow Test Results in Table 1 ......................... 76
Table 3: 09/19/12 Flow Test, Corresponding to Plan in Figure 47 .............................................. 78
Table 4: Air Changes per Hour (ACH), Based on Flow Test Results in Table 3 ......................... 78
Table 5: Air Temperatures at Supply and Return Ducts ............................................................... 78
Table 6: 09/19/12 Flow Test, Return Damper Covered (Corresponds to Figure 47) ................... 79
Table 7: Supply / Return Ratios, All Units ................................................................................... 81
Table 8: Baseline A109 Temperatures, (ASHRAE Effective Comfort Range Shaded) ............. 102
Table 9: Baseline B100 Temperatures, (ASHRAE Effective Comfort Range Shaded) ............. 103
Table 10: Baseline A209 Temperatures, (ASHRAE Effective Comfort Range Shaded) ........... 105
Table 11: Baseline B200 Temperatures, (ASHRAE Effective Comfort Range Shaded) ........... 106
Table 12: Baseline A305 Temperatures, (ASHRAE Effective Comfort Range Shaded) ........... 107
Table 13: Baseline A309 Temperatures, (ASHRAE Effective Comfort Range Shaded) ........... 108
Table 14: Baseline Outdoor Temperatures, (ASHRAE Effective Comfort Range Shaded) ...... 109
Table 15: Average Baseline Temperatures, (ASHRAE Effective Comfort Range Shaded) ...... 110
Table 16: Summarized Results, Air Changes / Hour Tests ........................................................ 114
Table 17: Supply and Return Air Temperatures, Unit A302 ...................................................... 119
Table 18: Temperature Ranges, TMY Data, Weather Station Data (WSD) ............................... 123
Table 19: TMY-Based Temperature Ranges .............................................................................. 124
Table 20: WSD-Based Temperature Ranges .............................................................................. 125
Table 21: A209 Recorded and Predicted Temperature Ranges .................................................. 127
Table 22: Outdoor Temperature Ranges, Outdoor Data Loggers ............................................... 128

XII

Terms and Abbreviations

• Air Economizer - A ducting arrangement, including dampers, linkages, and an automatic
control system that allows a cooling supply fan system to supply outside air to reduce or
eliminate the need for mechanical cooling (California Energy Commission 2010).

• BTU – British Thermal Unit
o The amount of energy needed to raise one pound of water one degree Fahrenheit.

• Climate Consultant
o A climate analyses software (The Regents of the University of California 2012).

• CZ – Climate Zone
o Refers to one of sixteen regional climatic zones throughout California, designated
by the California Energy Commission, and based on weather conditions within.

• EPW - EnergyPlus Weather (weather data file)
o File format used by HEED software to model weather data.

• EUI - Energy Use Intensity
o Used to measure a building’s energy use relative to its size.

• HEED - Home Energy Efficient Design
o A residential energy modeling software (UCLA Department of Architecture and
Urban Design 2012)

• kBTU - Kilo-British Thermal Units
o A measure of energy equivalent to 1,000 British Thermal Units.

• R-Value - Resistance Value
o A measurement of thermal resistance, the inverse of the U-value.

• SHGC - Solar Heat Gain Coefficient
o Energy transmittance factor of a window or door.

• TMY - Typical Meteorological Year
o An EPW file comprised of data averaged over multiple years.

• U-Value - Overall heat transfer coefficient
o A measurement of thermal resistance, the inverse of the R-value.

• WSD - Weather Station Data
o Weather data file specific to the case study building in this thesis.

1

Abstract
California has set a zero net-energy conservation goal for the residential sector that is to be
achieved by 2020 (California Energy Commission 2011).

To reduce energy consumption in the
building sector, modern buildings should fundamentally incorporate sustainable performance
standards, involving renewable systems, climate-specific strategies, and consideration of a
variety of users. Building occupants must operate in concert with the buildings they inhabit in
order to maximize the potential of the building, its systems, and their own comfort. In climates
with significant diurnal temperature swings, environmental controls designed to capitalize on this
should be considered to reduce cooling-related loads. One specific strategy is the air-side
economizer, which uses daily outdoor temperature swings to reduce indoor temperature swings.
Traditionally a similar effect could be achieved by using thermal mass to buffer indoor
temperature swings through thermal lag.
Economizers reduce the amount of thermal mass typically required by naturally
ventilated buildings. Fans are used to force cool nighttime air deep into the building, allowing
lower mass buildings to take advantage of nighttime cooling. Economizers connect to a
thermostat, and when the outdoor temperature dips below a programmed set-point the
economizer draws cool air from outside, flushing out the warmed interior air. This type of
system can be simulated with reasonable accuracy by energy modeling programs; however,
because the system is occupant-driven (as opposed to a truly passive mass-driven system) any
unpredictable occupant behavior can reduce its effectiveness and create misleading simulation
results. This unpredictably has helped prevent the spread of economizers in the residential
market.

2

This study investigated to what extent human behavior affected the performance of
economizer-based HVAC systems, based on physical observations, environmental data
collections, and energy simulations of a residential building in Los Angeles, California. Tangible
measures for alleviating problems, such as user-friendly interface design and the incorporation of
human behavior into energy models are recommended based on these observations.

Hypothesis
Economizers have a successful history of usage in commercial projects; however, their track
record in residential projects has not developed to the same degree. Residential applications
involve additional constraints beyond those typically found in commercial applications, often
centered on the user and their behavior. While many building codes throughout the United States
require large-scale commercial cooling systems to incorporate economizers to some degree, no
codes promote the incorporation economizers into residential buildings. Occupant behavior is
hypothesized to be a major factor of the performance of residential economizers.

3

Chapter 1: Introduction

California has set a zero net-energy conservation goal for the residential sector that is to be
achieved by 2020 (California Energy Commission 2011).

A significant amount of discussion and
focus has been brought to the designing and building of a tighter and more thermally robust
building envelope, coupled with increased efficiencies for heating, cooling, and domestic hot
water systems. In most cases, however, the final push to achieving widespread net-zero buildings
will involve renewable systems and climate-specific passive strategies. One environmental asset
available in some climate zones (CZ) are the diurnal temperature swings. Fan economizers
(essentially whole house fans connected to a smart thermostat) make use of these temperature
differentials for cooling and can minimize or eliminate the need for mechanical air conditioning.
The occupant sets a thermostat, and when the outside air is below the cooling set point the
economizer is triggered; it pulls air in from outside and flushes out the warm interior air with
cooler exterior air. This usually occurs during the nighttime as the temperature falls. If there is
sufficient mass and insulation in the residence it will cool down during the nighttime and remain
cool during the following day. If combined with daytime closing of blinds and the use of
overhead ceiling fans to reduce the apparent temperature, it is feasible that in some climate zones
air conditioning may not be necessary to maintain thermal comfort per ASHRAE’s Standard 55
(ASHRAE Standard 55 2010).

4

1.1 - Problem
Air conditioners account for the majority of all residential HVAC loads in the United States (Cox
2010) and have contributed to the evolution of houses increasingly reliant upon these systems to
maintain comfortable temperatures. The assumption that a project will feature air conditioning,
almost regardless of its location, is generally a safe one. Air conditioners have become
ubiquitous throughout the country as we strive for comfort and improved indoor environmental
quality; proposing a home without one would undoubtedly be met with hesitation from clients,
developers, and any parties contractually liable. The result of this has been the development of
buildings that disregard climate-specific passive strategies once inherent in the design process
(Rosen 2011)
As buildings became less individualized and more formulaic the approach to climatic
parameters and indoor conditioning standardized with it. Many modern homes use the same
basic HVAC systems regardless of size and location, and many office buildings lack operable
windows. This standardization of design is the result of a design sensibility that promotes full
removal of the occupant (as an active participant) from the surrounding environment- indoor
conditions should remain comfortable regardless of outdoor conditions. This notion evolved and
spread as air conditioners went from novelty items (Plumer 2012) to standard equipment. As a
result, buildings today not only have air conditioners by default, they often have oversized air
conditioners. Oversizing of equipment is common practice for several reasons:
• It reduces the contractor’s (or engineer’s) liability, as they’re often contractually bound to
provide thermal comfort for the occupants.
• Providing apartment-specific sizing requires more work on the engineer’s part, while the
use of rule-of-thumb standards is still commonplace and frequently considered sufficient.

5

• It’s perceived as the conservative approach, and therefore it’s an easy sell to the client.
Although it adds expense, oversizing reduces the risk of overheating during the hottest
times of the year- a risk that’s easy for a client to imagine and thus carries a lot of selling
power, particularly in warm climates.
• It was less of an issue when energy was inexpensive and climate change was not
receiving the attention it is today, and it takes time for any industry to change,
particularly when faced with resistance from within the industry.
It is worth scrutinizing these reasons because there are equally substantial drawbacks
associated with oversizing any HVAC equipment. Air conditioners (like boilers and water
heaters) operate at peak efficiency when in steady-state operation. Oversizing makes it difficult
for an air conditioner to operate steadily because the amount of conditioning required is less than
the unit’s conditioning capacity. This causes the unit to regularly turn on and off, alternating
between full power and no power, which can severely decrease overall efficiency (Yeo and
Wang 2010). An analogous effect is the decreased fuel economy in cars when in stop and go
traffic compared to steady driving. HVAC systems with variable speed motors can provide
higher efficiencies by varying the fan’s speed, thus reducing the number of on/off cycles (Yeo
and Wang 2010). These units are relatively new and are not widespread, although they do
provide potential for future development.
Figure 1 shows the relationship between cooling energy needed as the conditioned area
increases and the cooling energy needed per square foot for the same area. It’s clear that as the
space increases fewer BTU/SF are required to keep it cool. Freestanding houses, apartments, and
other individually conditioned dwelling types typically fall into the small end of the building
type/size scale, which limits any possible efficiency of scale (Figure 1). Correct HVAC sizing is
then especially important in small buildings.

6

Figure 1: Hourly Cooling Loads per Area, Load per Square Area
(http://www.energystar.gov/ n.d.)
Air conditioners have proven so successful at providing thermal comfort especially for
hot and humid climates they’ve become a crutch for building designers to lean on. Increasing
their efficiency will help curb this country’s per capita electrical consumption, but enabling
designers to safely explore other means of cooling can reduce the number of air conditioners and
potentially have a larger impact. Introducing alternate methods of cooling into mainstream
building design is a more sustainable path for future developments. Shifting the perception of air
conditioners away from the default HVAC equipment and towards equipment that’s included if
needed could encourage the reintroduction of passive design back into mainstream residential

7

architecture. The US uses roughly the same amount of electricity to power its air conditioners as
the entire continent of Africa uses for all its electrical needs (Cox 2010). Reducing our AC-based
consumption can have a very real impact on the effort to reduce consumption globally. Many of
California’s climate zones offer conditions suitable for passive design strategies, which must be
capitalized on to meet its net-zero efforts.

1.2 – Passive Design
Passive cooling strategies were once an inherent part of building design. Before air conditioning
became commonplace, considerable care went into providing thermal comfort through building
design. Cross ventilation through large and opposing windows, night flushing, using mass for
thermal storage purposes, large eaves, wrap-around balconies, etc. were all part of the common
building language. Their inclusion in residential buildings diminished as the ability to provide
thermal comfort through mechanical means increased. Figure 2 is an example of how
architectural design was employed in response to climate particulars. The Charleston “Single”
Style, from the mid 1700’s, developed a long, narrow floor plan, oriented perpendicular to the
prevailing wind. At only one room deep, the breeze could pass through the double hung windows
unobstructed by walls or doors. Shutters and large eaves or porches reduce solar gain, and could
be used as living areas in warmer months. From the layout to the details, much of the design is
based around the climatic particulars of Charleston, SC (Herman 1997). An example of
vernacular architecture in a climate closer to CZ 8 is the traditional adobe architecture of the
Pueblo peoples in the south western United States, from approximately 1200 A.D. Figure 3 shows
a multi-family housing complex with small windows and high mass walls. Maximizing the
enclosed volume while minimizing the exposed façade helped reduce solar gain, and resulted in

8

comparatively low ceiling heights. The Charleston “Single” used high ceilings to help induce
cross ventilation, as was necessary for comfort in the humid climate. The Pueblos were not faced
with high humidity in their arid climate, and their buildings reflect this. Very small windows
reduced solar gain and thick adobe walls buffered the diurnal temperature swings. The Pueblos
also frequently constructed buildings against, and sometimes within exposed stone hillsides. This
tactic further buffered the indoor temperatures and served other social functions within their
societies (Nabokov and Easton 1989). Both of these architectural styles were heavily informed
by the need to provide thermal comfort in thermally uncomfortable climates. With no mechanical
cooling a poorly designed building could be inhabitable, and architects were forced to
incorporate passive strategies.

Figure 2: Climate- Specific Vernacular Architecture, Charleston Single Style
(Manter n.d.)

9

Figure 3: Climate- Specific Vernacular Architecture, Adobe Pueblo Housing
(Galuzzi n.d.)
A more recent example of a hybrid passive and active system is the swamp cooler,
commonly used in the south west United States, Figure 4: Evaporative Cooler, Left, and
Schematic Diagram, Right. Like economizers, swamp coolers combined a traditional cooling
method (evaporative cooling) with mechanical assistance to accommodate the strategy into
homes. Fans blow air over or through a medium saturated with water, evaporating the water
while cooling and humidifying the air, which is then distributed throughout the building. The
swamp cooler is a present-day interpretation of a traditional concept, seen in early architectural
form in the evaporative wind catchers in Egypt around 1300 A.D. (Attia and de Herde 2009).

10

Figure 4: Evaporative Cooler, Left, and Schematic Diagram, Right
(Seeley International Pty Ltd. 2006)
These strategies all require a degree foresight by the designer to ensure they’re correct
choices for their particular condition. This degree of individuality comes at a price - it’s
generally less expensive to specify an air conditioner rather than investigate more creative
methods of providing cooling. Air conditioners allow flexibility in space planning, material
specification, wall-to-window ratios, etc., which removes the need to factor climate and location
into a building’s design. In multifamily projects removing the need for climate-specific design
makes it cheaper to specify air conditions in all dwelling units, regardless of orientation, height,
etc. The result is generally a thermally comfortable building with low upfront costs; this is an
attractive combination for developers looking to maximize sales and for consumers looking to
maximize floor area. The tenant, however, is left to deal with the building’s inherent
inefficiencies, i.e., unnecessarily high operating costs. The practice of designing and constructing
buildings en masse has enabled the average house sizes to increase, despite an overall decrease in
the size of the average household (number of occupants). In 1940, for example, the average US
house provided about 375 SF per/occupant. In 2005, US houses averaged almost 1,225
SF/occupant. Even when removing the number of occupants from this calculation the average

11

house today is still considerably larger; about 1,500 SF (1940) and almost 2,500 SF currently
(U.S. Census Bureau 2003)Today’s houses are more comfortable, more expensive to maintain,
and much, much larger than ever before.

1.3 – Economizers
Economizers do not represent a truly passive design strategy; they require electricity (although
considerably less than air conditioners (Kensek and Lanning, Building Analytical Modeling:
Analysis of Various Cooling Strategies as Modeled in HEED 2012)) and connection to a
thermostat to function. Additionally there must be a user to control the thermostat’s set points.
Economizers are used to exaggerate the daily cycle of warming/cooling outdoor temperatures by
mechanically bringing outdoor air into a building through a series of dampers tied to an air
handler and controlled with a thermostat, Figure 6: Typical Air Economizer Diagram. Depending
on the thermostat’s set points, the outdoor air warms or cools the building materials deep within
the space, reaching materials which otherwise would be unaffected by outdoor conditions. This
serves two functions: it increases both the surface area of the affected materials, and it increases
the temperature fluctuations they experience. In theory, the result is a daily indoor air
temperature swing similar to that found in a high mass building (see Figure 5). By contrast, a
high mass building designed to use thermal storage to its benefit is truly passive, because it does
not require energy or an occupant to operate. The exterior mass will slowly heat and cool
naturally, and will buffer indoor temperatures accordingly, whether or not occupants are present.
A low mass building without an economizer or similar HVAC system will respond very quickly
to outdoor conditions; indoor temperatures will then respond accordingly. A low mass building,

12

then, if left unchecked, can quickly become too hot or cold for comfort depending on the outdoor
temperature.

Figure 5: Temperature Patterns, Economizer, Left, and High/Low Mass, Right. Edited.
(European Commission 2008)
Residential building standards and material cost generally dictate low mass construction
typologies. Stick-frame and light-gauge metal framing construction types constitute the majority
of houses in the US today. These construction types are considered low mass because they’re
designed to perform to code with minimal material. High mass construction types include
materials like concrete, stone, brick, and adobe within their structure and are generally more
expensive and/or impractical to build and transport. These construction types can be blended
together to achieve the economic benefits of low mass construction with the thermal-storage
benefits of high mass construction, as will be discussed in Section, 1.4 – Case Study Building.
Economizers help low mass construction behave like high mass construction (from a thermal
storage perspective), enabling the continued use of standard building methods and materials.

13

Economizers represent a novel bridging of traditional concepts with current industry
requirements, and their incorporation into mainstream building design provides an opportunity
for significant reductions in energy use.

Figure 6: Typical Air Economizer Diagram
(Liescheidt Accessed 23 September 2012)

1.4 – Case Study Building
In previous studies, it was shown that it is possible in half of California’s sixteen climate zones
to design homes that can be comfortable without air conditioners (Milne, Morton and Kohut
2006). During the early architectural design phase of a progressive, energy efficient low-income
housing project in southern California, a careful prediction was made of indoor air temperatures.
Using the software program HEED (Home Energy Efficient Design), predicted results showed

14

that indoor temperatures would always fall within the comfort range as defined by the American
Society of Heating, Refrigeration, and Air-Conditioning Engineers (ASHRAE Standard 55 2010).
Because the apartments were designed with high mass first floors and supplemented with
additional interior thermal mass on upper floors, it was predicted that they would be comfortable
without air conditioning. Critical to achieving these comfort levels were the economizer fans that
would bring up to ten air changes per hour during the evenings when outdoor temperatures were
below the comfort range.
Throughout the building the apartments proved prone to overheating, even when there
was available outdoor air for cooling. By investigating the cause of this overheating evidence
supporting the use of economizers can be attained, while at the same time restoring thermal
comfort in the building. Demystifying their operation and highlighting potential pitfalls will
provide evidence for residential applications, and bolster support their use in future projects.

Figure 7: California’s Climate Zones
(California Energy Commission 2013)

15

1.5 – Objective
The objective of this thesis is to use the previously described case study as a testing ground to
explore the use of economizers in residential applications. Their incorporation into commercial
buildings is well known and has been accepted into US building codes (California Energy
Commission 2011) but their use in residential projects remains limited. By better understanding
the risks associated with economizers, their design and use in residential projects can be
appropriately altered. The potential they possess to reduce energy consumption necessitate
thorough consideration, however, like any unproven technology faced with initial setbacks
residential economizers require a thorough vetting to make them work as desired.
To accomplish this goal the issues seen in the case study building will be explored to
determine the major contributing factors behind them, with particular emphases given to
understanding the cause of the persistent overheating. Correcting these issues can help promote
the use of economizers in future projects, or, alternatively provide evidence that they should be
restricted to commercial applications.

16

Chapter 2: Background Research
This chapter reviews previous studies and other sources relevant to the use of economizers.
Reviewing the potential benefits and problems associated with economizers will help determine
the appropriateness of their usage in the case study. Background information pertinent to the
main areas of this thesis was studied to help understand the decisions made during the designing
of case study building and the overall direction of the thesis. These areas included the use of
economizers in HEED, performance and operational costs comparisons with air conditioners,
data loggers, human behavior, and weather data files.

2.1 - HEED and Economizers
2.1.1 – Previous HEED Studies Using Economizers
HEED was developed at the UCLA Department of Architecture and Urban Design, by the
Energy Design Tools. The engine that performs calculations for simulations designed within
HEED is Solar-5. It has been validated per ANSI/ASHRAE Standard 140-2001- Standard
Method of Test for the Evaluation of Building Energy Analysis Computer Programs (ASHRAE
2004). These validation studies have shown that HEED accurately predicts the energy
performance of all types of residential buildings, including those that meet California’s Title 24
Prescriptive Compliance Approach (UCLA Department of Architecture and Urban Design 2012).
HEED has been used in studies similar to this thesis. “Eliminating Air Conditioners in
New Southern California Housing” (Kohut and Milne 2010) showed that a residential building
deliberately designed to be naturally ventilated could work in CA Climate Zone 9. Climate Zone
9 is comparable to Climate Zone 8, with almost identical numbers of heating and cooling degree

17

days, recorded and design high and low temperatures, and similar design strategies per Pacific
Gas and Electric’s Title 24 recommendations (Pacific Gas and Electric n.d.). The design and
construction of the building is similar to the case study building, providing a suitable opportunity
for comparison. The building was designed to be cooled without air conditioners, and HEED
simulations showed that economizers would be sufficient to keep temperatures within the limits
set by ASHRAE Standard 55. Air conditioners were eventually installed in the apartments
instead of economizers; however, the as-built conditions still reflected a design around natural
ventilation strategies. The designers felt the building could still be cooled only through
ventilation and performed experiments to test that. The apartment unit tested remained
comfortable during the hottest part of the year, using only natural and fan-forced ventilation.
This confirmed the suitability of pairing natural and fan-forced ventilation with increased mass
in this climate. The maximum temperatures recorded in the ventilated apartment were within 1 F
of the original predictions made by the HEED model using economizers. While this experiment
does not provide an exact comparison to the case study, it does demonstrate that options other
than air conditioners can be successful if appropriately employed.
Another study, “Energy Efficient Affordable Housing; Validating HEED’s Predictions of
Indoor Comfort” (Milne, Morton and Kohut 2006) found similar results. Both the accuracy of
HEED’s predictions and the ability to use natural ventilation (along with the corresponding
architectural considerations) are further confirmed.
2.1.2 – Considerations when Using Economizers
Economizers are not commonly used in residential applications, and they remain generally
unfamiliar to users accustomed to air conditioners. Air conditioners by comparison are relatively
straightforward to understand and operate. The concepts behind the successful use of

18

economizers, particularly providing cooling through mass and thermal lag, are not commonly
used and remain foreign to general users. This unfamiliarity must be accounted for through the
design of the interface, tenant education, or by other methods deemed appropriate. Economizers
can also experience reduced performance should any of their components fail. Proper
maintenance and periodic assessments of their components should be done to prevent mechanical
failures, and users should be able to tell on their own if any failures occur.
In the analysis “Economizers in Air Handling Systems,” Steven Liescheidt reviews the
risks associated with malfunctions within economizers that commonly go unnoticed. Simulations
showed that a malfunctioning outside-air damper could increase air conditioning loads by as
much as 84 percent (Liescheidt Accessed 23 September 2012). The case study building does not
have air conditioning so an open damper would not increase loads to the same degree. It is safe
to assume, however, that indoor conditions would deteriorate accordingly. Liescheidt further
examines this problem by reviewing several studies revealing widespread malfunctioning among
economizers in existing buildings. Another study, “"I Always Turn it on Super": User Decisions
About When and How to Operate Room Air Conditioners” reviews the wide range of usage
patterns across physically similar conditions. Although not a study of economizers, it
nevertheless demonstrates that occupant usage patterns play a large role in thermal comfort and
shows these patterns are as numerous and varied as the individual occupants. The study was
conducted across several similar apartments, and usage patterns showed some users operated
their air conditioners 3 times as much as others. The occupants’ schedules were decided “on the
basis of many non-economic factors, including: daily schedule, folk theories about how air
conditioners function and the body's heat tolerance, personal strategies for dealing with all
machines, and beliefs and preferences concerning health, thermal comfort, and alternative

19

cooling strategies” (Kempton, Feuermann and McGarity 1992). These potential problems need to
be considered when deciding if economizers are the right choice for a particular project.

2.2 – Case Study Comparison, Economizers and Air Conditioners
In a previous and currently unpublished study, one of the apartments in the case study building
was modeled in HEED, and the differences in thermal comfort and energy consumption were
recorded. In the studies discussed in Section 2.1.1 – Previous HEED Studies Using Economizers,
HEED was able to predict indoor temperatures to within +/- 1.8 F of the observed temperatures.
In this simulation, which was run from June 6 – September 4, the unit showed good results for
the use of the economizer; the economizer was able to maintain temperatures within the
prescribed thermal comfort range (between 70 F and 80 F) for 83.24% of the time, with the
majority of the remaining hours spent under 70 F, not above. When using actual weather data the
dwelling exceeded 80 F only 1.84% of the time, and the highest recorded indoor temperature was
81.8 F for the testing period (Kensek, Lanning and Kohut, et al. 2013). The electrical energy
consumption was considerably less when simulated with an economizer than with an air
conditioner, even with the best available SEER rating as shown in Figure 8: Simulated EUI
Values, Economizers and Air Conditioners.

20

Figure 8: Simulated EUI Values, Economizers and Air Conditioners (HEED Simulation)
A comparable study prepared for Southern California Edison reviewed various cooling
strategies and found similar results. A model similar to apartment A309 in the case study
building was simulated over the same period, with the following variations: bare (no cooling), an
economizer, an economizer plus fins and overhangs at the windows, and with an air conditioner.
Each variation was simulated with the model facing east, west, and south. The results
overwhelmingly showed that the economizer alone was capable of keeping the apartment cool,
while consuming a small fraction of the energy used by the air conditioned model (Kensek and
Lanning, Building Analytical Modeling: Analysis of Various Cooling Strategies as Modeled in
HEED 2012).

21

2.3 - Data Loggers
Data loggers (pictured in Figure 11) have been used successfully in many studies for monitoring
and recording interior temperature and equipment amperage. Onset Computer Corporation, the
manufacturer of the HOBO data loggers used in this thesis, periodically releases its own “White
Papers,” technical reports written by industry leaders. Two of these papers, “Analyzing Air
Handling Unit Efficiency”, and “Addressing Comfort Complaints with Data Loggers” were
regularly referenced during the various studies conducted. These papers helped inform how data
loggers could most effectively report meaningful data about how the economizers were being
operated (Rosenberg 2012) (Onset Computer Corp. 2012). Data loggers were also used in a
technical report published by the California Energy Commission titled “Small HVAC Problems
and Potential Savings Reports.” This report explained in detail which components of the
economizers should be reviewed during the preliminary verification portions within Section
3.2.2 - HVAC Verification (California Energy Commission 2003).

2.4 - Human Behavior
Human (occupant) behavior was originally considered to be one of the largest factors affecting
the indoor environment in the case study. Numerous sources were referenced to gain a better
understanding of how to glean information from observed usage patterns, their impact on indoor
conditions, and methods of getting the occupants to change. “Effects of Variations of Occupant
Behavior on Residential Building Net Zero Energy Performance,” discusses which energy-
intensive components in residential projects are most affected by occupant behavior. Not
surprisingly HVAC operation (specifically the heating and cooling set points) ranks highly in
most of the simulations and case studies. The study shows that usage patterns relative to HVAC

22

systems had less of an impact in energy-efficient residential buildings than comparable,
conventionally designed projects (Field and Brandemuehl 2011).
Despite the occupants’ particular interest in reducing energy consumption, tenant
education plans combined with instantaneous-feedback thermostats showed little long term
positive effects (Parker, Hoak and Cummings 2010). After three years an exit survey was
conducted of the occupants and their usage patterns: the average energy reduction was 1%,
whereas it averaged closer to 10% at the beginning of the study. The results are important to
consider but are rather discouraging. The study corroborates the need for inherent energy saving
strategies as opposed to optional strategies, as even the most dedicated users in the study seemed
to lose interest relatively quickly.

2.5 - Weather data
Drury Crawley provides a glimpse of the myriad climatic factors that must be accommodated
into an accurate energy model (Crawley, Which Weather Data Should You Use For Energy
Simulations of Commercial Buildings? 1998). Despite the developments in energy modeling
software and weather data files since the paper’s publishing in 1998, many of the issues
discussed are still applicable today. The lack of weighting of each climatic factor remains
particularly important when creating a custom weather file. This paper, along with the more
recent “Improving the Weather Information Available to Simulation Programs” (Crawley, Hand
and Lawrie, Improving the Weather Information Available to Simulation Programs 1999) helped
inform which climatic factors from the TMY weather files could be combined with the weather
station data to achieve the best possible results in the case study simulations.

23

Although not directly related to weather data, the cooling effects of air movement were
shown to be quantifiable (Steadman 1979). Steadman discusses how the apparent temperature,
i.e., the temperature perceived by an occupant, can be affected by factors not typically
considered. Among these factors are body mass, amount of clothing, physical activity, exposure
to solar radiation, and wind speed. Steadman’s research was pivotal in the creation of the
commonly used Heat Index. The case study building relied on reduction in apparent temperature
through its overhead ceiling fans, and the reduction of direct solar exposure through blinds to
achieve comfort. Steadman’s work allowed these factors to be quantified and incorporated into
the energy model, where they informed the design process and the strategies employed.

2.6 - Conclusions from background research
This background research informed the strategies employed by the architect when designing the
case study building, as well as the methods used to perform the tests within this thesis. The
precedents set saved immeasurable hours performing trial and error type tests. In particular the
research on human behavior and weather data directly impacted the proposed recommendations,
while the studies on the use of economizers in HEED helped shape the recommendations given
for rectifying the issues in the case study building. The human behavior precedents reiterated
how unpredictable and variable occupants can be, necessitating a subjective teaching approach
specific to each occupant.

24

Chapter 3: Methodologies
3.1 - Scope of Work

Figure 9: Graphic Abstract
Of the many variables that impact indoor environments, temperature is potentially the largest
contributor to occupant comfort and the indoor environmental quality. Other contributing factors
include humidity, carbon dioxide, carbon monoxide, and toxins released from recently installed
building materials. This study focused primarily on thermal comfort, with indoor temperature
being the main metric as it was the chief complaint among residents in the case study building.

25

Humidity was not a major concern as inside the building because outdoor levels in this climate
remain relatively low throughout the year.
Located approximately 20 miles southeast of downtown Los Angeles, on South Atlantic
Avenue in East Rancho Dominguez, CA, the project represents a forward-thinking effort on the
part of the developer to introduce sustainable building measures into low income housing.
Completed in 2009, it was the first multifamily development in Los Angeles County to achieve
Platinum certification for Leadership in Energy and Environment Design (LEED for Homes)
from the U.S. Green Building Council - the highest possible ranking. The building is a 3 story
rectangular structure oriented north-south with a central courtyard and contains 70 apartment
units, a childcare center, and a healthcare clinic. The total floor area is 98,776 square feet. There
are many sustainability measures incorporated into the building’s design, including an array of
photovoltaic panels on the roof, a gray water irrigation system fed from discharged laundry water,
and high efficiency lighting fixtures throughout to name a few.
Immediately upon completion of the building the tenants complained of the “oven effect,”
the experience of walking through the front door into their apartment and into perceptibly hotter
air. The general consensus was that interior temperatures typically ranged between 5 and 15
degrees above exterior temperatures, regardless of the time of day (Case Study Tenant 2012).
Nighttime was described as offering some relief; however lowering the temperature was
typically induced through operating the windows and using ceiling and floor fans to move air, as
opposed to operating the economizers. There was a general lack of understanding by tenants and
building staff of what the economizers do and how they should be operated. It became clear early
in the study that observing their behavior and usage of the economizers could potentially
illuminate a strategy to increase the thermal comfort during the warmest months of the year.

26

Figure 10: Photo of Courtyard, Looking North
The first phase of the study was to measure and record the built conditions in which the
economizers were operating and to then incorporate the data into an energy model created in the
software program Home Energy Efficient Design (HEED). HEED was chosen for use in this
study because of its accuracy in predicting temperatures in several comparable precedents
(Kohut and Milne 2010). Construction details from the building were recorded through site visits
and construction documents, and HVAC details were gained through on-site verification tests.
The resulting data were then incorporated into the energy model to make the simulations match
the actual conditions as closely as possible. Simulations were run of several different versions of
the model to achieve baseline temperatures and Energy Use Intensity (EUI) data and then to
determine the effect the different variables covered in this study had on the occupant’s comfort.

27

The second phase of the study involved examining the weather data used in the energy
model. Simulations were run using Typical Meteorological Year (TMY) data for the region,
California climate zone 8. This is the regional weather data required for compliance with the
California Energy Code, Title 24. The resulting indoor temperatures were from these simulations
to verify compliance with ASHRAE’s thermal comfort requirements. Additional simulations
were then run using up-to-date, recorded weather data obtained on a regular basis from a nearby
weather station. The weather station data differ from the TMY data in that they’re specific to the
site and are for a specific time period, whereas the TMY data represent an amalgamation of data
from multiple years. The TMY data tend to be normalized as a result, without dramatic
variations, while the recorded data include any climatic anomalies measured during the recording
period. By running simulations with both file types it’s possible to view the building’s expected
performance, specifically the indoor temperatures, over the course of a typical year as well as
during weather extremes, e.g. heat waves.
The third phase examined the effect occupant behavior had on the thermal comfort within
the apartment units. By changing the usage patterns of the economizers through education, and
ultimately through removal of the occupant’s control over the economizer, it was possible to
measure the impact different user profiles had on temperatures within the apartments.

3.2 - Phase 1: Site Documentation and Experiment Setup
HOBO data loggers from Onset Computer Corporation were used to record temperature and
relative humidity in five apartments, several locations within the building’s central courtyard,
and on the roof. Recordings were taken hourly, beginning in June 2012 and ending in April 2013.

28

This range accounted for what is typically the warmest part of the year, with outdoor
temperatures over 105 F seen several times. Additionally, data loggers in several apartments
were connected to split-core AC current sensors, also from the Onset Computer Corporation,
enabling them to detect and measure alternating current (AC). The split-core AC current sensors,
hereafter referred to as “split rings,” were placed around the economizer’s AC power source,
measuring and recording when they were receiving power, producing a map of their usage.
Interviews with the tenants during the testing period indicated a lack of understanding on their
part about how the economizers function and how to properly set them. The split rings were
intended as an aid to visualize how the tenants used their economizers, both to understand their
usage pattern and to determine if, when in operation, they’re providing cooling as designed. It
was thought that recording usage would illuminate patterns and potentially aid in educating the
tenants on the proper operation schedules and settings to use.
The data loggers recording temperature are sensitive to direct solar radiation and require
placement away from windows and other openings. When possible they were placed above
doorways where they were out of reach of direct sunlight and any children who might be tempted
to remove them. Their readings were regularly checked using an infrared thermometer to confirm
their calibration and consistency with the unit’s thermostat. Site visits were generally made every
two to four weeks to pick up data from the data loggers and to consult with the tenants and
buildings management as needed. These visits continued consistently throughout the testing
period, totaling over 20, including one extended overnight visit in February 2013 to monitor an
economizer’s functioning throughout the course of a day-night-day cycle to confirm its ability to
cool the apartment.

29

Figure 11: Split-Core AC Current Sensor, Left, HOBO Data Logger, Right
While five units were monitored and observed throughout the duration of the study, one
unit on the top floor was focused on in particular. This apartment was chosen because it
consistently had temperatures among the warmest in the study and because the occupants were
cooperative and interested in learning how to improve their comfort. All of the units tested,
however, had comparable thermal comfort issues, and the problems faced by this dwelling unit
and the resulting measurements were consistent with the issues faced by all of the units tested.
The plan of the dwelling unit chosen for analysis is shown in Figure 12. It is 1,040 square feet,
with 105 square feet of glazing, most of which is westward facing. The unit is on the third (top)
floor, with neighboring units on both sides and below. The economizer can be seen highlighted
in the lower left corner of the living room. There are operable blinds and ceiling mounted fans to
help move air, and like all of the apartments there is no air conditioner.

30

Figure 12: Plan of Typical Apartment Unit Used for Testing

Figure 13: Photo of Apartment Unit Used for Testing

31

3.2.1 - Construction Verification
To ensure maximum accuracy of the energy model’s predictive capacity, site visits were made
early in the study to confirm construction assembly specifications which were originally made
from construction drawings. In addition, dimensions were site-checked, light fixtures were
confirmed to be fluorescent, and the appliances were confirmed to be ENERGY STAR®.
• Glazing: Low-E, double pane, U-Value= 0.40, SHGC= 0.40
• Vinyl frames at all glazing
• Light-colored operable blinds
• Interior walls: R-Value = 13 (adiabatic)
• Exterior walls: R-Value = 21, stucco exterior
• Roof: R-Value = 30, with two layers 5/8” gypsum board at ceiling
• Ground floor: Concrete slab on grade (12” – 18’)
• Upper floors: Lightweight concrete (1.5”) on plywood

32

3.2.2 - HVAC Verification

Figure 14: Typical Floor Plan (Preceding Number Corresponds to Respective Subchapter)
The HVAC systems in each of the units included in the study were inspected and tested
to confirm they were in good condition and were functioning as designed. Confirming the
physical condition and the performance of the HVAC systems could uncover expose issues
relating to thermal comfort and were to be rectified where possible or were incorporated into the
energy model as-is, to better reflect the built conditions.

33

3.2.2.1 - Pressure Relief Damper (Interior Exhaust Panel)
When the economizer system was being designed for this particular project, it was decided to
add an operable panel in each of the living rooms that opened to the outside and could be opened
when the economizer was operating. It was thought this would allow the warm interior air to be
pushed out of the unit as the economizer introduced fresh air into the unit, balancing the indoor
air pressure. Approximately four feet long and eight inches tall, one panel was installed in the
living room of every apartment. One-way barometric dampers within the panel were designed to
remain closed until the pressure inside the unit reached a preset threshold, at which point they
would open and dump air outside. This was designed to help make the overall system as
automated as possible; in the warm months, when the economizer is needed most, the panel
could be left open, and the dampers would respond automatically once the economizer
pressurized the space. This had the added benefit of bypassing the potential security issue of
having to leave windows open, particularly during the nighttime. The barometric dampers,
however, were found to be missing, having never been installed in the first place, and the
dampers behind the panels had all been screwed shut (see Figure 45). Although this nullified the
exhaust panel’s effectiveness, the same effect could be achieved by opening the window, even
slightly, and creating a route for the warm, pressurized air to escape. It did however introduce
another step into the operational procedure the tenant must go through each morning and evening
to achieve thermal comfort.
By design the tenant could keep the blinds and windows shut during the summer, relying
on the dampers to remain closed to prevent infiltration, except at night when the economizer was
operating. For the purpose of this study the panels were kept shut, and it was the requested that
the tenants not use them, which they hadn’t been doing anyway. The tenants were instead

34

instructed to open the windows when using the economizer to balance indoor air pressure. The
exhaust panels were ignored in the energy model.

Figure 15: Typical Pressure Relief Damper at Living Room, Opened for Photograph
3.2.2.2 - Air Changes per Hour

Figure 16: Flow Hood Measuring Air Flow at Supply Dampers

35

For the economizer to effectively cool the apartment, it must provide enough air to cool down
the masses within the unit, specifically the walls, floor, and ceiling. During the design phase of
the building, simulations showed that at the ground floor units the concrete floor provided
enough mass to accomplish this. At the top floors, which have lightweight concrete, a double
layer of drywall added to the ceiling was determined to provide enough mass to accomplish this.
The thermal storage capacities of these building materials enable them to retain a considerable
amount of energy in the form of heat. Blowing cool air across their surfaces forces convection to
occur, transferring some of the stored heat energy into the air, which then must be exhausted.
The limiting factors when using this type of HVAC system are the temperature of the
supply air used for cooling (the outdoor air temperature), the period of time the cool air is
available, and the rate at which the air passes over the mass. The first two factors, the
temperature of the outdoor air and the hours it is available, are products of the surrounding
environment and cannot be controlled. The rate of air, however, can be controlled by the capacity
of the air handler, which sits directly above the economizer. Design-phase tests confirmed that in
this climate, if the air handlers achieved approximately 10 air changes per hour (the number of
times the indoor air is replaced with outdoor air), then the interior masses would be cooled
enough to provide cooling the following day. As the volume of air is different depending on the
number of bedrooms, different sized air handlers were specified per apartment type to meet the
10 air changes per hour (ACH) requirement.
To confirm the units were receiving the designed volume of air from the HVAC system, a
flow hood device was used to measure how many cubic feet per minute (CFM) of air was
passing though each supply register within the apartment. The sum of these quantities was tallied
for each apartment to get the total supply air volume, which was then divided by the volume of

36

the unit to confirm the actual ACH. The flow hood was also used on the return-air dampers in the
living rooms and the outside-air dampers, to verify the equivalent volume of outside air was
being supplied from outside.
3.2.2.3 - Smoke Gun Test
A smoke gun is a device that creates artificial fog through combining water with a glycerin-
based solution. By blowing this “smoke” around areas of suspected leaks it is possible to
visualize the leaks. This method was used to determine the air-tightness of the seals around the
economizer and the air handler. It was not possible to use the smoke gun to test the actual ducts
because they were behind drywall, and as the units were occupied, the conditions under which
the study was conducted did not permit the removal of any drywall. This test therefore was
conducted primarily “upstream” of the economizer and air handler, and not near the interior
supply registers, where it could be used to test leakage from or into the cavity in which the
drywall is installed.

Figure 17: Smoke Gun Warming up for Testing

37

3.2.2.4 - Thermostat Test
The economizers in this building are connected to “smart thermostats” that incorporate data from
multiple inputs to determine how the economizer and air handler should respond. Thermostats in
residential buildings with typical air conditioners are only concerned with the indoor
temperatures – when the indoor temperature is above the cooling set point the thermostat
activates the air conditioner, until the interior temperature falls below the set point. The case
study thermostats require additional data because they must respond to both indoor and outdoor
temperature and relative humidity. Consequently, each HVAC system has an indoor
thermometer and an outdoor thermometer and an enthalpy sensor to measure humidity. Although
humidity was not focused on in this study, the economizers in the case study do measure it, and
factor it into the operation of the dampers to keep indoor conditions comfortable.
The thermostats were tested by first manually measuring the outdoor temperature outside
of each unit, as this data is not displayed on the thermostat readout. The cooling set point on the
thermostat was then set below the outdoor temperature. This prompted the thermostat to trigger
an actuator to open the outside-air damper and the air handler to turn on and begin drawing fresh
air into the unit. This test had the effect of also testing both the indoor and the outdoor
thermometers as well, because both have to be functioning correctly for the thermostat to engage
the economizer. Indoor temperatures were also regularly checked with an electronic handheld
thermometer, which was also compared the thermostat’s readout. Each thermostat in the study
was tested, as well as those in several other units to confirm consistency (see Figure 51: Kestrel
Anemometer, Typical Thermostat in Case Study Apartments).

38

3.2.2.5 - Damper Leakage Test
As with any economizer-based HVAC system there are many parts that must be in good working
condition to realize in actuality the efficiencies seen in simulations. The system in the case study
building has two dampers, one inside that regulates the flow of recirculated air and one on the
building’s façade that regulates the influx of fresh air. Previous studies have shown that faulty
dampers can significantly impair the economizer’s ability to provide cooling (Liescheidt
Accessed 23 September 2012). A damper that cannot close properly can allow air to infiltrate the
HVAC system and change the temperature of the air entering into the space. In the case study
building it is of paramount importance that the outside-air dampers in particular seal properly as
they form the barrier between the interior and the exterior air, which can be as much as 35 F
different. A leaky outside-air damper could potentially allow infiltration of hot outside air when
the economizer is in recirculation mode and at 20 ACH, quickly cause the unit to overheat. With
no air conditioning, there would be no way to reduce the indoor temperature if this happened,
hence the importance to verify the dampers were in good condition.
To test that the dampers were sealing properly the flow hood was again employed. The
set points on the thermostat were adjusted so that the economizers were engaged and
recirculating interior air. As a result, the economizer seals the outside-air dampers and opens the
return-air dampers inside, so no fresh air is let inside. The flow hood then was used to test the
outside-air damper to determine if any air leaked through, and if so, how much.
The dampers are located before the air handler, meaning negative pressure is created at
the outside-air damper. This test must be conducted with the air handler engaged, because it’s
possible that a leaky damper would only become apparent when pressurized. The flow hood does
not create air movement, it only measures it. Additional tests were run under the same conditions,

39

except with the return-air dampers artificially covered. This setup creates more pressure than
would be seen under normal operation and was intended to double check the air-tightness of the
exterior damper.

Figure 18: Economizer Diagram, Edited to Reflect Case Study Building Conditions
(Liescheidt Accessed 23 September 2012), Edited
3.2.2.6 - Overhead Fan Test
As there is no air conditioning in the units, overhead fans were installed to reduce the effective
temperature in the units when needed. On warm days, the fans could be operated to provide an
effective reduction in temperature of approximately 5 F (Figure 19). Although air movement
does not provide actual cooling, it can provide several degrees of effective cooling for the
occupants by increasing the rate of evaporation across the skin. This increase allows the body to
perspire more efficiently and therefore regulate internal body temperatures more effectively. To

40

this effect the tenants were advised to only operate the fans when at home and never to use them
when their unit was empty, as was occasionally observed.

Figure 19: ASHRAE Figure 5.2.3 “Air Speed Required to Offset Increased Temperature”
(ASHRAE Standard 55 2010)
To comply with their ENERGY STAR® designation, the fans were specified to operate
at efficiencies between 155 CFM/watt - 75 CFM/watt depending on the setting (low – high)
(United States Environmental Protection Agency 2013). These volumes would provide enough
apparent cooling to keep temperatures within AHRAE’s standards without causing disruption of
interior conditions, e.g. shuffling papers around or disturbing other lightweight material. The fan
in each room of each apartment was tested using a pocket anemometer (Kestrel 3000 Wind
Meter) to confirm velocity, from which the CFM values were calculated.
3.2.2.7 - HEED Energy Model
The results of these tests were then incorporated into the same energy model used in the design
phase of the building. By incorporating these results into the model it was thought the resulting

41

simulations would be more representative of actual conditions, therefore reducing the possibility
of misleading results. This was done to validate the initial decision to install economizers in the
building, which in theory can work well in climate zone 8 but which require particular care to
work effectively in residential application, particularly when there are no precedents
1
. It was
consequently determined that maximizing the model’s ability to predict indoor temperatures
would reduce the chance of false positives, results which can be misleading and misrepresent
actual conditions despite best efforts toward the contrary.
The next phase of the study furthers the calibration process of the revised energy model
by incorporating climate-specific weather data into the simulation, to better reflect the actual
environmental conditions. As will be discussed, the existence of microclimates within any of
California’s climate zones can impact the results of an energy analysis. As California’s climate
zone 8 covers approximately 800 square miles, the possibility for climatic differences within the
regional zone can be accounted for by incorporating site-specific data, to better tailor the model
to its surroundings. The model was re-run with climate data that incorporate these particulars,
discussed in the next section.

1
While there are precedents for economizers in residential applications, this project represents the first such LEED
certified Platinum multifamily building in Los Angeles County.

42

Figure 20: Perspective View of Example HEED Model

3.3 - Phase 2: Climatic Calibration
California’s climate zone 8, classified as a Mediterranean climate zone covers approximately 800
square miles and is located primarily in Los Angeles and Orange County. It does not meet the
coast, being buffered by climate zone 6, and therefore experiences more significant annual and
diurnal temperature swings (Pacific Gas and Electric Company 2013). These are favorable
conditions for an economizer to operate, with strong potential to reduce the daytime indoor
temperatures. During the warmer months it is critical that the economizer has regular access to
outside air that is at or below the comfort low temperature. During the nighttime the economizer
uses this cool air to flush out the unit, replacing the warm interior air, and cooling the building
materials which draw heat from the indoor air the following day, keeping its temperature down
(La Roche and Milne, Effects of Window Size and Mass on Thermal Comfort using an
Intelligent Ventilation Controller 2004). Access to accurate climatic data is of paramount

43

importance when determining the appropriateness of this type of cooling system, especially if it
is to be the primary (or only) method of cooling, as there is no way to cool down an overheated
apartment once it’s become too hot.
When creating an energy model for a particular location it is typical to use data that
represents the average temperatures to be expected. This way the resulting building will
generally fall within the comfort range most of the time, barring the extremes. This type of data
is present in Typical Meteorological Year (TMY) files which are compiled from historical data
and averaged into an approximation of the climatic conditions for a given year (Crawley, Which
Weather Data Should You Use For Energy Simulations of Commercial Buildings? 1998). These
files are commonly used to inform the design process. The building in this study was tested using
a TMY file and was also tested using actual climatic data recorded by a nearby weather station.
Specifically, temperature, direct normal radiation (kBTU/sf), and diffuse horizontal radiation
(kBTU/sf) were incorporated from the weather station to improve the accuracy of the model.
Figure 21 illustrates the temperature discrepancies between data from the TMY file used
and the site-specific data from the same period, imported from the weather station located near
the case study building. The diagram was created in Climate Consultant and shows the TMY
temperature data plotted over actual weather station data from June through September. The
design highs and lows for these months differ significantly from the TMY data. August was
actually the warmest month in 2012, and all months experienced a wider spread of temperatures,
indicating warmer days and cooler nights. These data will likely affect indoor temperatures in the
monitored dwelling units and were incorporated to increase testing accuracy. An energy model
should take into account the general conditions that can be expected as well as the particulars of
the location, hence the incorporation of data from the local weather station.

44

Figure 21: TMY (CZ 8) Temperatures and WSD (Pink Overlay), June – September 2012
HOBOs were also placed outside in various locations around the building to monitor
temperature; however the data from these data loggers were found to be inaccurate and were
subsequently not used in the study. The exterior data loggers yielded consistently higher daytime
temperatures despite placement to avoid direct solar radiation (Figure 25). They were assumed to
have been placed too near the building and warmed by radiation from nearby building materials.
As a result, the energy model only used temperature data from the TMY file and the weather
station.
3.3.1 - Typical Meteorological Year Data
Typical Meteorological Year weather files are required by Title 24, of the California Energy
Code when creating energy simulations for code compliance. As discussed, CA climate zone 8
covers over 800 square miles, and, like all of the climate zones, contains climatic variables which
are not represented by the TMY file. The more recent and pressing concern of global warming
may also not be accurately represented by TMY files: for example, the current (2013) TMY file
for climate zone 8 consists of data from between 1953 and 1974 (TMY updates are planned for

45

the 2014 CEC revision). Temperature and radiation can be seen plotted in Figure 22: Typical
Meteorological Year (TMY) Data, June – September, to illustrate the lack of variation present in
the file, despite the climatic volatility usually associated with this time of year. Note the
maximum temperature in the TMY data during this time is approximately 80 F.

Figure 22: Typical Meteorological Year (TMY) Data, June – September 2012
3.3.2 - Weather Station Data
To better reflect actual conditions it was necessary to source more specialized weather data. As it
was not possible to accurately record the data necessary to compile a weather data file, data from
a nearby weather station was used. WeatherAnalytics.com compiles and distributed weather data
from stations across the country and provided the WSD for this thesis. By incorporating this
microclimate-specific data into the model it was thought the resulting indoor temperatures would
better match the actual indoor temperatures, as recorded by the interior data loggers. Having
access to current, hourly weather data enabled real-time modeling which proved helpful when

46

going back and forth between the building and the model. The maximum temperature during this
time was almost 95 F, approximately 15 F higher than the maximum in the TMY data. Note how
frequently the temperature in this data exceeds the TMY maximum (80.1 F). Radiation data vary
between the two files, but not to the same degree the temperature data varies.

Figure 23: Weather Station Data (WSD), June – September 2012
3.3.3 - Collected Outdoor Data
Although external temperatures were also recorded, the climate data from the local weather
station was ultimately used for the energy models. The predicted indoor temperatures from the
energy models were compared to the temperatures recorded by the indoor data loggers to
determine if the simulated economizers would provide effective cooling under the actual
measured conditions.

47

Figure 24: Outdoor Data Logger Temperatures, Various Locations, June – September 2012
It was not possible to measure direct and diffuse radiation for these studies. Radiation
data were taken from the TMY file and the weather station, and the energy model results from
each incorporate the respective simulation. There was some discrepancy in the direct radiation
between the TMY file and the weather station data, suggesting variations in solar gains can be
expected when modeling using actual data. Again the TMY file is useful for general modeling
but designers must ensure thermal comfort at all times, not only during average conditions
(Crawley, Which Weather Data Should You Use For Energy Simulations of Commercial
Buildings? 1998).

48

Figure 25: Data Logger at Roof, Left, and at Courtyard Vegetation, Right

3.4 - Phase 3: Occupant Behavior
The third and final phase included the study of occupant behavior and its impacts on thermal
comfort. Occupant behavior was predicted to have significant impacts on thermal comfort. Based
on observed thermostat settings and interviews with the tenants and the management, there
appeared to be a large gap between how the occupants should operate the economizers and how
they were operated. Occupant behavior differed significantly between the simulations and the
actual building. To create a simulation that more accurately reflects the built conditions, either
the simulation needs to incorporate the irregular behavior observed, or the tenants need to adjust
their behavior to match that of the simulated “tenant.” As part of an ongoing effort to improve
comfort, the tenants were met with regularly to discuss how to properly operate their
economizers. It was initially believed that further education would help bridge the gap between
the simulation results and the results of the data loggers in the case study building.
One of the initial requirements by the building’s management program was a tenant
education program, to explain the various energy saving measures employed by the building to

49

the tenants, upon moving in. It was never implemented, however, and it was apparent that neither
the management nor the tenants had a thorough understanding of how to operate the economizers
in particular. The “Green Living Guide” given to the tenants to help them operate their apartment
effectively includes a small portion about the economizer, Figure 26. While the conceptual
information included is helpful, overall it is rather misleading; the photo of the “economizer”
actually shows the exhaust louver, and the economizer itself is not covered. Additionally, the
tenants involved in this study were unaware that they were to receive a guide at all, and most
took it upon themselves to figure out the HVAC system.
The offices within the building are cooled with a ground source heat pump, limiting
thermal discomfort experienced by the staff and potentially reducing the urgency with which the
problem was approached. The management’s initial response to complaints about overheating
apartments, which began shortly after the building’s completion in 2009, was to purchase several
portable air conditioners and lend them to tenants upon request. This solution was short lived;
after the first summer of using air conditioners, those tenants found their electrical bills to be
prohibitively high and the air conditioners were returned. At the time of this study none of the
units tested had air conditioners in their units. Apparently the tenants had decided that it was
preferable to suffer through the summer rather than face higher utility costs. Incidentally, as of
May 2013 the building management is taking bids on the installation of permanent air
conditioners in each of the units. How the tenants will cope with the higher utility costs is
unknown.

50

Figure 26: Excerpt from Management-Provided “Green Living Guide”
(Abode Communities 2009)
3.4.1 - Tenant Education
Tenant education meetings were held after the HOBO data loggers were already in place and
collecting data to determine if additional education impacted thermal comfort. Between June and
July 2012, one-on-one meetings with tenants were conducted in each of the units tested.
Meetings centered on the proper procedures for getting all of the HVAC components to work
together, conceptual information on how the economizer functions including its limitations, and
finally methods of avoiding “short-circuiting” the system. These meetings were intended to
convey a baseline level of information about the HVAC systems, in particular how to keep the
apartments from overheating, which was the major complaint.
It was hoped that the conceptual points covered in the meetings would help the tenants
recognize the effects of a working economizer versus those of a malfunctioning economizer,

51

empowering them to recognize their mistakes, and thus learn through trial and error the settings
they prefer. This level of education was thought to be crucial for successful long-term operation,
as it would empower the tenants to troubleshoot issues based on symptoms they could recognize.
Major points included the following:
• Economizers are not air conditioners, they’re essentially fans.
• Economizers should be used prophylactically to prevent overheating.
• Once an apartment is hot it’s too late to use the economizer to cool it that day.
• The difference between fresh and recirculated air.
In addition, more factual and data-driven information was conveyed to the tenants,
primarily surrounding proper thermostat settings, when to use the economizer, and when not to.
The major points discussed included the following:
• Definition of and proper adjustment of thermostat set points
• How to verify source air, between fresh and recirculated
• How to recognize when a thermostat is/is not properly set
• Typical times when the economizer should be operating
During this period other verification tests were being conducted, as described in Phase 1,
to simultaneously rule out the possibility of mechanical error. This kept the focus on the impact
(if any) a better educated tenant might have on thermal comfort conditions in their apartment.
3.4.2 - Locked Thermostat
The second portion of this phase involved leaving the thermostats in several units locked, hoping
to give the economizer a head-start on cooling the unit. The decision to lock the thermostats was
discussed beforehand with the tenant, each of whom agreed to participate. They were given
directions on how to unlock the thermostats if desired, and each unit also has a master “override”

52

switch that controls the power supply to the HVAC system, shown in Figure 27. The
participating tenants were told the nature of the study; that it was an attempt to decrease the
daytime temperatures in their apartment. It was always made clear that they could withdraw from
the experiment and regain HVAC control at any time if desired.

Figure 27: Locked Thermostat, Left, Economizer’s “Override” Switch, Right
3.4.2 - Vacant Unit Test
The final review of occupant behavior involved waiting for one of the units in the study to
become vacant, and then performing an extended test of the HVAC system with only the author
present and in full control over the settings. This was thought to be the most accurate way to
reproduce the conditions of the energy simulation.

3.5 - Overview
In general, the first half of these tests followed a measurement and verification process, whereby
the details of the built conditions were transferred into the energy models to better understand the
cause of the high indoor temperatures. Data were taken primarily from site visits and then added

53

to the energy models, which were put through increasingly layered simulations. The results of
these simulations were then compared against data recorded on site, to verify the initial decision
to use economizers in the first place. These data were reviewed for inconsistencies, which were
hypothesized to exist mainly within the realm of user error, poor controller design, and the
inability of the software to incorporate a variation of human behavior into its calculations. The
study will determine success by the following measures:
• Reduced indoor temperatures due to mechanical adjustments made to the system.
• Reduced indoor temperatures due to more effective education of the occupants.
• Reduced indoor temperatures due to the system being locked in AUTO, removing the
user entirely from operation of the system.
Before the results were reviewed and analyzed it was assumed that improving the tenants’
understanding of the HVAC systems through education would improve how they interact with
their thermostats. Improving the operation of the thermostats was expected to bridge the gap in
thermal comfort between the case study building and the simulation results. Chapter 4 will focus
on the results from these tests.

54

Chapter 4: Results
4.1 - Scope of Work
The results from the tests described in Chapter 3 are displayed in the following sections. The
data is presented as it was recorded, without comparison or discussion. This follows in Chapter 5.
Results from the three phases are presented in the same chronological order as Chapter 3;
• Phase 1: Site Documentation and Experiment Setup Results
• Phase 2: Climatic Calibration Results
• Phase 3: Occupant Behavior Results
The data loggers described in Chapter 3 were placed as shown in the floor plans that follow (see
Figure 28 - Figure 34) (Abode Communities 2009). Results from the studies include indoor and
outdoor temperature, economizer activity levels, energy use intensity, and radiation data.

55

Figure 28: 1
st
Floor Plan, Not to Scale

56

Figure 29: 2
nd
Floor Plan, Not to Scale

57

Figure 30: 3
rd
Floor Plan, Not to Scale

58

Figure 31: Roof Plan, Not to Scale

59

Figure 32: North-Facing Unit B115 Floor Plan, 850 SF, Not to Scale

60

Figure 33: West-Facing Unit (A109, A209, A305, A309) Floor Plan, 1,040 SF, Not to Scale

61

Figure 34: South-Facing Unit (B100, B200) Floor Plan, 1,315 SF, Not to Scale

4.2 - Phase 1: Site Documentation and Experiment Setup Results
These charts reflect the baseline temperatures and amperages recorded by the data loggers within
each apartment and those located in the courtyard and on roof. The baseline in this study is
defined as the initial recorded temperatures and economizer usage before any measures was
taken to rectify the apparent thermal comfort issues. This baseline period spanned from June 6
th

to September 4
th
, 2012.

62

First Floor Units:
Unit A109
Baseline results from west-facing, first floor Unit A109 are shown in Figure 35: Unit A109.
Indoor data loggers measured temperature (left y-axis) and economizer amperage (right y-axis)
in 5-minute intervals. Blue areas indicate economizer activity, when the economizer was turned
on. Note the lack of economizer activity throughout the duration. The economizer can be seen
“pinging” on an almost daily basis, but this does not indicate engagement of the fan coil or
dampers. The only discernible activity can be seen between 7/3 and 7/5. After 8/19 the
economizer reported no activity at all- the result of the data logger being unplugged by the tenant.
Note also how much lower the temperature at the economizer (green line) was during the test.

Figure 35: Unit A109 (1
st
Floor, West-Facing)

63

Unit B100
Results from south-facing, first floor Unit B100 are shown in Figure 36: Unit B100. Data loggers
measured temperature (left y-axis) in 5-minute intervals. The economizer in this unit was not
monitored. Two data loggers measured temperatures inside the unit; one in the kitchen and one
in the bedroom as shown in Figure 34. The temperature readings were very consistent during the
test, except for a two week period in August when the kitchen was reportedly cooler than the
bedroom.

Figure 36: Unit B100 (1
st
Floor, South-Facing)

64

Unit A115
Both data loggers installed in north-facing, first floor Unit B115 were removed by the tenant
shortly after being installed, and no data from this unit were obtained. This issue is discussed in
detail in Chapter 5, Section: Unit B115. The data loggers were not replaced and this unit was not
included in the tests which followed.

65

Second Floor Units:
Unit A209
Results from west-facing, second floor Unit A209 are shown in Figure 37: Unit A209. Data
loggers measured temperature (left y-axis) in 5-minute intervals. The economizer in this unit was
not monitored. Two data loggers measured temperatures inside the unit; one in the kitchen and
one in the bedroom. Both data loggers reported consistent temperatures within the unit and
clearly show the effects of the August heat wave on indoor temperatures. The temperature in this
unit was near or above 90 F several times during this heat wave.

Figure 37: Unit A209 (2
nd
Floor, West-Facing)

66

Unit B200
Results from south-facing, second floor Unit B200 shown, Figure 38: Unit B200. Two data
loggers were installed to monitor indoor temperatures: the kitchen data logger was discovered
missing during the first site visit so no data were obtained, and the living room data logger went
missing in early August, before the heat wave. Temperature data obtained from the living room
data logger (before it went missing) show consistently higher temperatures in Unit B200 during
June and July than any of the other units.

Figure 38: Unit B200 (2
nd
Floor, South-Facing)

67

Third Floor Units:
Unit A305
Baseline results from west-facing, third floor Unit A305 are shown in Figure 39: Unit A305.
Indoor data loggers measured temperature (left y-axis) and economizer amperage (right y-axis)
in 5-minute intervals. Indoor temperatures during August exceeded 90 F regularly, reaching
almost 95 F. The economizer in this unit was monitored, however the data logger was discovered
unplugged (Figure 40) during the 8/8/12 site visit. It was plugged back in during this visit
although approximately one month of data was lost. Data from the remaining logger suggest the
economizer did not turn on during this time.

Figure 39: Unit A305 (3
rd
Floor, West-Facing)

68

Figure 40: Unplugged HOBO Data Logger

69

Unit A309
Results from west-facing, third floor Unit A309 are shown in Figure 41: Unit A309. Indoor data
loggers measured temperature (left y-axis) and economizer amperage (right y-axis) in 5-minute
intervals. Note the variety of different usage patterns experienced by the economizer. During
June and July the economizer operated on a fairly regular basis, although not on a daily cycle as
the simulations suggested. In late July the capacity of the data logger’s memory was reached and
it stopped recording. This was discovered and rectified during the 8/8/12 site visit. After this date
the economizer was reportedly running continuously, despite outdoor temperatures exceeding
any other period during the study.

Figure 41: Unit A309 (3
rd
Floor, West-Facing)

70

Outdoor Units:
Data loggers were placed on top of and around the building (Figure 28 and Figure 31), and
recorded temperature in 5-minute intervals. Figure 42: Outdoor Temperatures shows the results
of each outdoor data logger as well as the temperature data recorded at the nearby weather
station. Note how the results from each source follow similar patterns, yet how much higher the
data logger temperatures consistently were. The data loggers at the roof exhibited large diurnal
swings in temperature, sometimes exceeding 40 F during a 24 hour period. This data was
determined to be inaccurate and was not used in any of the HEED simulations.

71

Figure 42: Outdoor Temperatures, Outdoor Data Loggers, WSD File
As previously discussed weather data was obtained from a variety of sources; a local
weather station (WSD), the Dept. of Energy (TMY), and the outdoor data loggers whose results
were averaged together. Figure 43 shows the temperatures from each source compared against
each other, to find any anomalies and to confirm the accuracy of the WSD data ultimately used
in the simulations. Polylines produced from each source show the overall temperature trends.
Note the large discrepancies between the various sources, particularly the lack of variability in
the TMY data compared to the other two sources. The data loggers (light blue) and the weather
station (orange, dashed) follow the same trends across the summer, whereas the TMY data (dark
blue) do not. WSD data fell roughly between the TMY and the data logger.

72

Figure 43: Outdoor Data Loggers, WSD File, TMY File, June – September 2012
4.2.1 - Construction Verification
Site visits were made to verify the following as-built conditions:
• Interior fixtures: Confirmed compact fluorescent
• Appliances : Confirmed ENERGY STAR®
• Glazing : Confirmed Low-E, double pane, U-Value= 0.40, SHGC= 0.40
• Glazing frames: Confirmed vinyl, with thermal break
• Treatment : Confirmed white, operable blinds
• Interior Walls : Confirmed R-Value = 13
• Exterior Walls : Confirmed R-Value = 21, stucco exterior
• Ceiling : Confirmed R-Value = 38, with two layers 5/8” gypsum board
• Ground floor : Confirmed slab on grade, depth unconfirmed
• Upper floors : Confirmed R-Value = 19 (adiabatic)

73

4.2.2 - HVAC Verification
The HVAC systems in all the apartment units tested were inspected and confirmed to be in good
working condition. The results from the inspections are discussed in the following sections.
4.2.2.1 – Pressure Relief Damper (Interior Exhaust Panel)
All of the units tested had a pressure relief damper per LA County’s mechanical plan-check
requirements. In all cases the barometric dampers within the panels were found to be screwed
shut. While the panels themselves were still operable, the dampers behind the panels were
inoperable and were without barometric sensors. As the dampers were not working properly in
the first place, the tenants were asked to disregard them entirely and to use their windows to
exhaust any warm indoor air.

74

Figure 44: Typical Exhaust Panel, Closed Position

Figure 45: Barometric Dampers within Exhaust Panels Screwed Shut and Inoperable

75

4.2.2.2 - Air Changes per Hour
The flow hood tests were initially performed in apartment units A109, A209, and A309. The
tests consisted of measuring the volume of air passing through the supply and the return registers,
to confirm the number air changes per hour. Figure 46 shows the duct layout in these units.

Figure 46: HVAC Duct Layout, Unit AX09 (All Floors) 1,040 SF, Not to Scale

76

Table 1: 09/13/12 Flow Test, Corresponding to Plan in Figure 46
Unit # LS 1 LS 2 BS 1 BS 2 BS 3 Supply Total RD Supply / Return Ratio
A109 340 256 284 300 260 1,440 830 1.73 : 1
A209 285 300 210 225 205 1,225 775 1.58 : 1
A309 250 265 290 190 215 1,210 805 1.50 : 1

The results were recorded in cubic feet per minute (CFM) and represent tests performed
with the flow hood device described in Chapter 3. The windows in all of the units were opened
several inches, to mimic the instructions given to the tenants. The economizers were set in
recirculation mode so the outside dampers remained shut.
Table 2: Air Changes per Hour (ACH), Based on Flow Test Results in Table 1
Unit # Area SF Volume CF Supply CFM Supply CFH ACH (CFH/CF)
A109 1,040 7,810 1,440 86,400 11.06
A209 1,040 7,810 1,225 73,500 9.41
A309 1,040 7,810 1,210 72,600 9.29

77

Unit A302
Apartment A302 is not included in other areas of this thesis. During the flow hood testing,
however, it was vacant and served as an experimental unit to conduct the more intrusive test of
determining whether opening or closing windows impacted the volume of air coming from the
economizer. The floor plan and duct layout are visible in Figure 47.

Figure 47: HVAC Duct Layout, Unit A302, 663 SF, Not to Scale

78

Table 3: 09/19/12 Flow Test, Corresponding to Plan in Figure 47
Unit # LS 1 BS 1 Supply Total RD Supply / Return Ratio Test Notes
A302 426 258 684 480 1.42 : 1 Windows open
A302 418 242 660 480 1.37 : 1 Windows closed

The results above were measured and recorded in cubic feet per minute and represent
tests performed with the flow hood device described in Chapter 3. The unit was initially tested
with windows opened several inches and then retested with the windows closed.
Table 4: Air Changes per Hour (ACH), Based on Flow Test Results in Table 3
Unit # Area SF Volume CF Supply CFM Supply CFH ACH (CFH/CF) Test Notes
A302 663 4,425 684 41,040 9.27 Windows open
A302 663 4,425 660 39,600 8.94 Windows closed

In addition to the volume of air being moved, the temperature of the air entering the
return duct and of the air exiting each supply duct was also recorded.
Table 5: Air Temperatures at Supply and Return Ducts
Unit # LS 1 BS 1 RD Temperature Change Test Notes
A302 84.5 F 85.8 F 83.4 F + 2.4 F Windows open
A302 86.1 F 86.0 F 83.5 F + 2.6 F Windows closed

79

An additional test was performed in Unit A302 after repeatedly observing higher supply
volumes than return volumes. It was initially assumed that supply and return volumes would be
similar, but when the discrepancy was found to be recurrent the unit was retested with the return
damper covered. GLAD Press’n Seal™ plastic wrap was used to cover the return damper at the
economizer, which was activated and set in recirculation mode, as had been done in previous air
change tests. The plastic wrap prevented air from entering the return duct, so the only air coming
through the supply ducts in this particular test was concluded to be air entering the ductwork
through leakage at, or before the fan coil.
Table 6: 09/19/12 Flow Test, Return Damper Covered (Corresponds to Figure 47)
Unit # LS 1 BS 1 Supply Total (Duct Leakage) RD Test Notes
A302 97 66 163 0 RD covered, windows open
A302 90 76 166 0 RD covered, windows closed

80

Figure 48: GLAD Press’n Seal™ Plastic Wrap Covering Return Duct
4.2.2.3 - Smoke Gun Test
The smoke gun was used to test air leakage in all HVAC systems in this study. While the results
are not quantifiable per se, all of the units experienced air leakage into the overhead ductwork,
corroborating the surplus of supply air vs. return air seen in the previous. The supply / return
ratios from this section are consolidated below. They consistently show more supply air entering
each unit than return air entering the economizer. Air leaked into the ductwork at different areas
in each unit, though in all cases there was significant leakage around the filter access panel and
any conduit pass-through.

81

Table 7: Supply / Return Ratios, All Units
Unit # Supply / Return Ratio Supply Total – Return Total CFM Test Notes
A109 1.73 : 1 610 Windows open
A209 1.58 : 1 450 Windows open
A309 1.50 : 1 405 Windows open
A302 1.42 : 1 204 Windows open
A302 1.37 : 1 180 Windows closed
A302 - 163 RD covered, windows open
A302 - 166 RD covered, windows closed

Figure 49: Smoke Gun Testing Air Leakage at Economizer Ductwork
4.2.2.4 - Thermostat Test
To test whether proper communication was occurring between the thermostats and the
economizers, the cooling set point of each thermostat was set above the outdoor temperature at
the time of testing. These outdoor temperatures were not recorded although they all fell between

82

50 F and 80 F. The purpose of this test was not to determine the thermostat’s reaction to a
specific temperature but rather to confirm that if the cooling set point was above the outdoor
temperature, the outside-air damper would open automatically.
All of the thermostats successfully opened the outside-air dampers; however it is worth
noting that none did so immediately. In each case it took between 1 – 2 minutes for the dampers
to begin opening after the cooling set point was programmed. Through this testing procedure the
sensitivity of the economizer’s outdoor thermometer was also inadvertently tested. Again, these
comparisons were made on site and not recorded. However, all of the thermostats registered
accurate temperatures which were corroborated by a handheld digital thermometer, Figure 51.

Figure 50: Outside-Air Damper Opening in Response to Lowering Set Point

Figure 51: Kestrel Anemometer, Typical Thermostat in Case Study Apartments

83

4.2.2.5 - Damper Leakage Test
In this test, the outdoor-air dampers were tested for airtightness using the flow hood. All of the
dampers were visually inspected as well for any light penetration, which could indicate areas of
air penetration. The outside-air dampers were tested with the economizers in operation to create
negative pressure within the ductwork, encouraging any potential leaks to become apparent.
There were no leaks in any of the outside-air dampers; all of them sealed completely, registering
0 CFM under load.

Figure 52: Outdoor-Air Damper Being Tested for Airtightness
4.2.2.6 - Overhead Fan Test
Each fan was tested with the anemometer at each of the test units. Air speed was recorded
several times and averaged to confirm consistency. All of the fans moved air between 195 and
325 feet per minute when at the highest setting, and thus all were capable of providing the
apparent cooling required to comply with AHRAE’s limits (see Figure 19).

84

4.2.2.7- HEED Energy Model
The results from the tests in the previous sections were incorporated into energy models created
in HEED. Weather data was reviewed and similarly incorporated into the HEED models,
described in the following section.

4.3 - Phase 2: Climatic Calibration Results
Each apartment unit in the test was simulated in HEED using weather data from California’s CZ
8 (TMY) as well as weather data from the nearby weather station (WSD). Weather data was
recorded with outdoor data loggers, but was found to be inaccurate and was not used in any
simulations. The following sections display the results from these simulations.
4.3.1 - Typical Meteorological Year Data
First Floor Units:
Unit A109
HEED’s predictions are recorded for the same baseline testing period using California’s CZ 8
weather data (TMY). Indoor temperature data (left y-axis) and economizer activity (right y-axis)
were logged hourly from HEED’s results. HEED recorded the economizer’s activity in air
changes per hour, represented by the green line in Figure 53: Predicted Temperatures, TMY File,
Unit A109 (1
st
Floor, West-Facing).

85

Figure 53: Predicted Temperatures, TMY File, Unit A109 (1
st
Floor, West-Facing)
Note the shift in indoor temperature and ACH that occurs at the end of June. The
economizer in HEED is seasonally programmed; during winter and spring months it is
programmed to keep temperatures near the top of the specified comfort zone, and during summer
and autumn months it’s programmed to keep temperatures at the bottom of the comfort zone.
This transition occurred between June and July and results in the shift seen at this time. The
comfort zone was set to 70 F – 80 F for all of the simulations.

86

Unit B100
Indoor temperature (left y-axis) and economizer activity (right y-axis) were logged hourly from
HEED’s results for Unit B100, Figure 54: Predicted Temperatures, TMY File, Unit B100 (1
st

Floor, South-Facing). Like A209, the predicted temperatures in this unit varied considerably on a
daily basis, and the economizer operated throughout the entire test.

Figure 54: Predicted Temperatures, TMY File, Unit B100 (1
st
Floor, South-Facing)

87

Second Floor Units:
Unit A209
Indoor temperature data (left y-axis) and economizer activity (right y-axis) were logged hourly
from HEED’s results for Unit A209, Figure 55: Predicted Temperatures, TMY File, Unit A209
(2
nd
Floor, West-Facing). Note the predicted temperatures in this unit varied considerably more
than the simulations for A109. Additionally the economizer can be seen operating every day,
whereas the economizer in A109 (the apartment directly below) did not turn on until after the
seasonal shift. Unit B200 was not simulated due to the lack of comparative data- the result of
missing data loggers (see section 4.3 - Phase 2: Climatic Calibration Results).

Figure 55: Predicted Temperatures, TMY File, Unit A209 (2
nd
Floor, West-Facing)

88

Third Floor Units:
Unit A305 / A309
Indoor temperature data (left y-axis) and economizer activity (right y-axis) were logged hourly
from HEED’s results for Units A305 and A309, Figure 56: Predicted Temperatures, TMY File,
Unit A305/A309 (3
rd
Floor, West-Facing). The results from these two units were identical and
combined for brevity. The predicted temperatures in these units were similar to those in Unit
A209, although the diurnal temperature swings were slightly greater. Economizer activity is also
similar to Unit A209.

Figure 56: Predicted Temperatures, TMY File, Unit A305/A309 (3
rd
Floor, West-Facing)

89

4.3.2 - Weather Station Data
The HEED model of each unit was then simulated using weather data recorded at the local
weather station and obtained from www.weatheranalytics.com (weatheranalytics.com 2013).
First Floor Units:
Unit A109
Indoor temperature data (left y-axis) and economizer activity (right y-axis) were logged hourly
from HEED’s results for Unit A109, Figure 57: Predicted Temperatures, WSD File, Unit A109
(1
st
Floor, West-Facing). Note how late in the season the economizer activates.

Figure 57: Predicted Temperatures, WSD File, Unit A109 (1
st
Floor, West-Facing)

90

Unit B100
Indoor temperature data (left y-axis) and economizer activity (right y-axis) were logged hourly
from HEED’s results for Unit B100, Figure 58: Predicted Temperatures, WSD File, Unit B100
(1
st
Floor, South-Facing). Note how the indoor diurnal temperature swings are greater than those
in A109. The temperature regularly exceeded 80 F but always remained within the ceiling fan’s
ability to keep temperatures within ASHRAE’s allowable apparent comfort range (ASHRAE
Standard 55 2010). Also note the economizer’s “activation”, which occurs at the end of June,
like the simulations using TMY data.

Figure 58: Predicted Temperatures, WSD File, Unit B100 (1
st
Floor, South-Facing)

91

Second Floor Units:
Unit A209
Indoor temperature data (left y-axis) and economizer activity (right y-axis) were logged hourly
from HEED’s results for Unit A209, Figure 59: Predicted Temperatures, WSD File, Unit A209
(2
nd
Floor, West-Facing). Temperatures were predicted to be greater than those in the Unit A109,
the unit directly below. The economizer operated all summer although the ACH increased
significantly with the economizer’s seasonal shift at the end of June. Unit B200 was not
simulated as the recorded data was insufficient for comparison (see section 4.3 - Phase 2:
Climatic Calibration Results).

Figure 59: Predicted Temperatures, WSD File, Unit A209 (2
nd
Floor, West-Facing)

92

Third Floor Units:
Unit A305 / A309
Indoor temperature data (left y-axis) and economizer activity (right y-axis) were logged hourly
from HEED’s results for Units A305 and A309, Figure 60: Predicted Temperatures, WSD File,
Unit A305/A309 (3
rd
Floor, West-Facing). Results from these two units were identical and
combined for brevity. Predicted temperatures in these units were similar to those in Unit A209,
although the diurnal temperature swings were somewhat greater. Economizer activity was
similar to Unit A209, operating all summer but significantly more after June.

Figure 60: Predicted Temperatures, WSD File, Unit A305/A309 (3
rd
Floor, West-Facing)

93

4.4 - Phase 3: Occupant Behavior Results
This section reviews the results of the tests related to occupant behavior. As discussed in Section
3.4 - Phase 3: Occupant Behavior, several tests were performed that included the tenants. The
results of these tests are presented in the following sections.
4.4.1 - Tenant Education
Tenant education did not begin in earnest until after the summer was over and outdoor
temperatures had begun cooling down. Conversations with the tenants during summer site visits
covered the basic operation of the thermostats and economizers. However, tests specifically
measuring the results of these conversations did not begin until September and October.
None of the units exhibited anything but a temporary change, let alone improvement, in
how their economizers were set. Several examples are presented that describe how any changes
in occupant behavior stemming from educational site visits were quickly disregarded, forgotten,
or went unheeded for some other reason. Possible explanations for this are described in Section
5.4.1 - Tenant Education.
Unit A109
Unit A109 was visited on 9/9/12 as part of a routine site visit to the case study building. It was
observed during this visit that the economizer had been inactive for the preceding weeks. The
tenants were given detailed instructions on how the thermostat should be set to best cool the unit.
It was explained that despite the high outdoor temperatures during the daytime, nighttime
temperatures still dropped enough for the economizer to provide cooling. The thermostat was not
programmed for the tenants; instructions on the correct settings were provided and the tenants

94

agreed to follow the instructions. It was believed that encouraging the tenants to set the
thermostats themselves would enable them to set it correctly in the future.
Figure 61 shows the results of this particular site visit. The economizer was inactive
during the preceding weeks, and remained inactive after the brief period of activity on 9/13. This
period of activity was the only activity during the entire month of September. Note that the
economizer turned on in the afternoon and turned off the same evening, operating only during
the warmest part of the day.

Figure 61: Tenant Education- Temperature and Economizer Amperage, Unit A109

95

Unit A309
Unit A309 was visited on 9/16 during a visit to collect data from Unit A109. Unlike Unit A109,
the thermostat in this unit was found in the “Fan ON” mode. This forced the air handler to run
continuously, though not necessarily to bring in cool air (or changing indoor temperature).
Thermostat settings were explained and instructions were given to the tenants at this time.
Results were briefly successful: the economizer began introducing fresh air on 9/17, visible in
the area highlighted within Figure 62. On the evening of 9/17 the indoor temperature dropped
quickly as the economizer introduced fresh air; indoor temperatures reduced despite rising
outdoor temperatures. On 9/19 the economizer did not operate, and on 9/21 it resumed its
previously seen Fan On operation where the air handler operated but did not bring in fresh air.

Figure 62: Tenant Education- Temperature and Economizer Amperage, Unit A309

96

4.4.2 - Locked Thermostat
The next test was to lock the thermostat, removing the occupant’s control. Unit A309 was again
used. Figure 63 shows the results from the week-long period in November when the test was
conducted. Indoor temperatures can be seen reducing each time the economizer turns on. Figure
64 shows an example of a locked thermostat.

Figure 63: Locked Thermostat- Temperature and Economizer Amperage, Unit A309

Figure 64: Locked Thermostat (Example)

97

Temperature and amperage data from before the tests were performed are presented in
Figure 65. The results are from Unit A309. Note how the economizer consistently turned on late
in the afternoon when outdoor temperatures were at their maximum. It then turned off in the
early evening when outdoor temperatures were near their lowest.

Figure 65: Existing Conditions- Temperature and Economizer Amperage, Unit A309
4.4.3 - Vacant Unit Test
The final occupant-based test was performed in a vacant apartment. This test was conducted to
determine if a knowledgeable user could successfully operate the economizer. Results are
presented in Figure 66: Results- Vacant Unit Test, Unit A109. The economizer can be seen
turning on (shaded blue) at 21:00 on 2/28, and off at 11:00 the following morning. Indoor
temperatures can be seen reducing immediately as outdoor entered the space. Apartments A209
and A309 were monitored during this time and do not show equivalent temperature reductions.
The indoor temperature stopped receiving outdoor at approximate 07:00 on 2/29, at which point
the air handler remained on to recirculate the cool indoor air.

98

Figure 66: Results- Vacant Unit Test, Unit A109

4.5 – Results Overview
The results from the tests reviewed in this chapter show that the temperatures recorded within all
of the units do not match the predicted temperatures, regardless of the choice of weather data.
Indoor temperatures in all units were higher than predicted. All of the HVAC systems were in
good working condition and possessed the ability to cool when operated correctly. Throughout
the testing period the tenants were generally resistant (intentionally or otherwise) to changing
their understanding of the economizers, as well as how they set their thermostats. With a
knowledgeable user the economizers were able to successfully provide cooling, and indoor
temperatures were reduced. Further discussion of these results follows in Chapter 5.

99

Chapter 5: Analysis and Discussion
5.1 - Overview
This chapter analyzes the data from the results in Chapter 4. The data is reviewed to determine
the successfulness of each of the tests performed. The baseline testing period (June – September)
serves as the control against which the tests are measured. Throughout the baseline testing period
the tenants were regularly interviewed to gain a better understanding of how they perceived and
used their HVAC systems. This period functioned as the control group for the tests that followed
afterward, to compare the economizers’ effectiveness before and after implementation of the
tests. The outdoor temperature was lower during the autumn and winter months, reducing the
need for economizers. To perform thermal comfort tests during this time it was necessary to
create artificial comfort zones based on the outdoor temperatures at the time of each study.
Successfulness was then determined by the whether indoor temperatures reached their target
temperature range.
The objective of this analysis is to
• Determine the accuracy of the results and the validity of each test
• Determine the successfulness of each of the tests
• Compare how the results of each test might impact the other tests
• Discuss the potential for success of each method in an uncontrolled environment

5.2 - Phase 1: Site Documentation and Experiment Setup Analysis
Results from Chapter 4 showed that all of the monitored units heated up considerably more than
HEED predicted. All three of the west facing units experienced temperatures above 90 degrees
during the heat wave in August 2012. The largest discrepancy observed between the modeled

100

apartments and the actual apartments was the difference in usage patterns of the economizers.
All of the simulations showed daily operation of the economizer regardless of the weather data
file used. None of the economizers monitored in the case study building exhibited this level of
consistency for more than a few days. When regular economizer schedules were seen in the
building, the indoor temperatures were seen reducing. Misperception of how and when the
economizers should be operated was found to be widespread (Figure 74: Set Point Errors Seen at
Thermostats). Tenants were generally resistant to changing their understanding and usage of the
system, and the tests described in previous chapters were met with fleeting success at best.
According to the simulations performed for this study and others (Kohut and Milne 2010) (Milne,
Morton and Kohut 2006), economizer-based HVAC system can work well in this climate if
operated properly.
It is recommended that for buildings with individual economizers (one economizer per
dwelling unit) that they are simple enough for anyone to operate. Indications about when to turn
the economizer on or off could be displayed on the thermostat to make it obvious to the user if
the settings are incorrect. This could at least reduce the likelihood of prolonged thermal
discomfort by helping the user learn over time which settings are correct.
The tenants in the case study could have received more education regarding correct usage
of the economizers, and it is recommended that an HVAC-specific education program be
included in future projects. Improving the tenants’ basic understanding was shown to have
beneficial effects on indoor conditions, even if only briefly (see Section: 4.4.1 - Tenant
Education).

101

Education could help overcome some of usage problems observed during the studies. The
24 hour delay required between setting the thermostat and attaining thermal comfort was
particularly confusing for the tenants observed. This delay is required by the economizer,
however, and there is no practical technological solution for it. The tenants need to understand
this to effectively cool their apartments.
Another example indicating a lack of education was the usage pattern of the windows,
which must be opened when the economizer is turned on to provide an outlet for the built up
warmth. Observations from site visits indicated that the windows were often left open or closed
for long periods of time; they should have been operated on a daily schedule in accordance with
the economizer’s cooling cycle. If the windows are not opened, the air pressure within the
apartment will eventually reduce the air handler’s ability to bring in fresh air, even if the
thermostat is set correctly.
The degree of specificity required to correctly pair these systems properly requires some
attention be paid by the occupants. The initial HVAC design must consider that not all users will
operate the system perfectly all the time, and in some cases may not be able to operate it
successfully at all (hence the barometric dampers). This could have been corrected in a post-
occupancy commissioning process that includes occupant performance in addition to HVAC
performance- a reasonable recommendation given the risk of high indoor temperatures.
The following charts review the temperatures results from the baseline testing period
from each apartment. Each chart is followed by an analysis of the findings for that unit.

102

First Floor Units:
Unit A109
Table 8: Baseline A109 Temperatures, (ASHRAE Effective Comfort Range Shaded)
A109 (West-Facing) Kitchen Bedroom Econ. Temp. Average WSD
% < 70 F 0.0% 0.0% 0.0% 0.0% 53.0%
70 F < % > 80 F 53.6% 29.5% 85.5% 56.2% 35.5%
80 F < % > 85 F 45.0% 63.9% 14.5% 41.1% 7.4%
% > 85 F 1.4% 6.6% 0.1% 2.7% 4.1%
Total 100.0% 100.0% 100.0% 100.0% 100.0%

Unit A109 experienced average indoor temperatures within the standard 70 F – 80 F comfort
range for the majority of the summer. This can be seen in Table 8; however, even during the
cooler summer months the indoor temperatures hovered at or above this range. To maintain
thermal comfort in warmer conditions like these the relative humidity must be considered and
accounted for if too high. Generally, as the temperature increases the relative humidity must
decrease to maintain comfort; however, in climate zone 8 humidity was not a major issue. As the
economizers do not humidify or dehumidify the fresh air, the tenants are reliant on their ceiling
fans to help maintain comfort.
The average indoor temperature in A109 includes temperatures recorded by the same data
logger used to measure amperage at the economizer. This necessitated locating it near the
economizer, which is located on the exterior wall of the unit. The resulting figures, found under
the heading Econ. Temp., are influenced more heavily by outdoor conditions because of
temperature fluctuations in the façade than the data loggers located deeper within the apartment.

103

For a more concise look at the temperatures it’s best to view the percentage of time the two
monitored rooms were within the comfort zones. The bedroom, for example, was between 80 F
and 85 F for 63.9% of the time tested, and the kitchen was within this range for 45% of the time.
This range is still within the ability of the ceiling fans to provide effective temperatures within or
slightly above the 70 F – 80 F range. There is no doubt that the unit was very warm, but the most
telling data is the difference in hours spent below 70 F inside the unit versus outside. Despite the
heat wave the majority of outdoor hours (53.0%) were actually below 70 F, while the indoor
temperature never dipped below 70 F. This discrepancy, along with the results in Figure 35: Unit
A109 (1
st
Floor, West-Facing) indicate that virtually no relief was provided by the economizer
during the months it was most needed.
Unit B100
Table 9: Baseline B100 Temperatures, (ASHRAE Effective Comfort Range Shaded)
B100 (South Facing) Kitchen Bedroom Average WSD
% < 70 F 0.0% 0.0% 0.0% 53.0%
70 F < % > 80 F 75.0% 66.0% 70.5% 35.5%
80 F < % > 85 F 25.0% 29.6% 27.3% 7.4%
% > 85 F 0.0% 4.4% 2.2% 4.1%
Total 100.0% 100.0% 100.0% 100.0%

Unit B100, a south facing ground floor unit with 4 bedrooms, experienced temperatures more
consistent throughout the unit than the other units. The two data loggers were placed at opposite
ends of the unit but the resulting temperature ranges were very consistent with each other. The
temperature during the early summer months was lower than those in Unit A109. However, it

104

remained between 75 F and 80 F the majority of the time. The unit was never below 70 F despite
outdoor temperatures regularly dropping below that point, suggesting minimal introduction of
fresh air. The economizer was not monitored in this unit, although from the data it can be
concluded that the economizer did not effectively cool the unit.
One area of discrepancy between the recorded temperatures is a two week period in the
middle of August. From the 10
th
through the 21
st
the kitchen temperature remained consistently
below the monitored bedroom despite remaining almost identical to the bedroom temperature
before and after this period. This period coincides with the heat August heat wave. The kitchen is
located deeper within the apartment unit away from windows. The entrance door to the
apartment is also located next to the kitchen and opens to a covered corridor. Temperatures in
this corridor were not recorded; however, tenants were frequently observed propping this door
open to allow cross ventilation through the unit. When discussing this practice with the tenants
they reported that on warm days corridor temperatures were always cooler than apartment
temperatures. This improvised cross ventilation strategy likely cooled down the areas within the
apartments immediately adjacent to the corridor.
Unit B115
Unit B115 was one of two apartments that received data loggers without prior communication
with the tenant. This was not intentional; it was a product of conflicting schedules, and it was
decided that the building management would convey the intention and details of the study. It
became apparent by the first site visit to retrieve data that this was a mistake. Both data loggers
were missing, and the tenant reported no knowledge of them ever having been installed. This
was the only north facing unit included in the test and was predicted to have the mildest indoor

105

temperatures because of the lack of direct solar exposure. It was decided to not replace the data
loggers and continue the study without a north facing test unit.
This experience was not a total loss, however, because it revealed the importance of
earning the tenants’ trust. It was fortunate this setback occurred early in the study because there
was minimal data loss, and much was learned about how to constructively interact with the
tenants. The tenants in the remaining apartments were communicated with on a regular basis, and
a lot of effort went into making the tests as transparent as possible and ensuring that everyone
involved was informed and in agreement with the procedures being done. The data loggers and
the data they collected, the process of uploading to a computer, and the choice of locations of the
data loggers were explained in detail to the remaining tenants. Overall the tenants appreciated
this level of involvement and especially enjoyed seeing how changes made to their HVAC
systems affected their indoor environment.
Second Floor Units:
Unit A209
Table 10: Baseline A209 Temperatures, (ASHRAE Effective Comfort Range Shaded)
A209 (West-Facing) Kitchen Bedroom Average WSD
% < 70 F 0.0% 0.0% 0.0% 53.0%
70 F < % > 80 F 71.9% 70.4% 71.1% 35.5%
80 F < % > 85 F 23.8% 24.1% 23.9% 7.4%
% > 85 F 4.3% 5.5% 4.9% 4.1%
Total 100.0% 100.0% 100.0% 100.0%

106

Unit A209 is directly above and identical in layout to Unit A109. Besides the different floor
height, the only significant difference is the floor construction. Unit A109 is slab on grade
construction whereas Unit A209 has lightweight concrete floors. This reduction in thermal mass
could potentially account for the higher maximum temperatures compared to those in A109.
Without the thermal comfort provided by the additional thermal mass, the indoor temperatures
could be more prone to following fluctuations in outdoor temperatures. The maximum
temperatures recorded by the data loggers in Unit A209 were 91.6 F (kitchen) and 90.5 F
(bedroom). The maximum temperatures seen in A109 were 89.0 F (kitchen), 89.4 F (bedroom)
and 85.1 F (Econ. Temp.). Although the peak temperatures are higher in A209, significantly
more time was spent within the 70 F – 80 F range: Unit A109 averaged 56.2% of hours within
this range (including the aforementioned Econ. Temp. within the average), while A209 averaged
71.1% of hours within this range. Although the peak temperatures are higher, Unit A209 was
significantly more comfortable during the testing period than was A109. The economizer in this
unit was not monitored.
Unit B200
Table 11: Baseline B200 Temperatures, (ASHRAE Effective Comfort Range Shaded)
B200 (South Facing) Bedroom (Missing) Living (Missing 8/8/12) Average WSD
% < 70 F - 2.2% 2.2% 53.0%
70 F < % > 80 F - 79.0% 79.0% 35.5%
80 F < % > 85 F - 18.5% 18.5% 7.4%
% > 85 F - 0.3% 0.3% 4.1%
Total - 100.0% 100.0% 100.0%

107

Unit B200 was the other unit which had HOBOs installed without direct communication with the
tenants. As in B100, communication with the tenant may have resulted in not using B200 as one
of the test units. The tenants in this unit were not disinterested in the study, and there were also
four small children living there. One of the two data loggers went missing immediately, and the
other went missing after about two months. The data collected during this time reflect similar
temperatures seen in the unit below, B100. This is more in line with the expectation that similar
units facing the same direction will experience similar temperatures.
Third Floor Units:
Unit A305
Table 12: Baseline A305 Temperatures, (ASHRAE Effective Comfort Range Shaded)
A305 (West-Facing) Bedroom Econ. Temp. Average WSD
% < 70 F 0.7% 16.7% 8.7% 53.0%
70 F < % > 80 F 71.2% 52.4% 61.8% 35.5%
80 F < % > 85 F 15.1% 19.3% 17.2% 7.4%
% > 85 F 13.0% 11.6% 12.3% 4.1%
Total 100.0% 100.0% 100.0% 100.0%

Unit A305, the first of the top floor units reviewed, experienced peak temperatures even higher
than the units on lower floors yet had the lowest average temperatures of all the units in the test.
Despite the economizer not turning on at all, the majority of hours were spent between 70 F and
80 F. Like Unit A109, this unit had a data logger at the economizer measuring both amperage
and temperature. As discussed, the temperature recordings by this data logger are particularly
susceptible to radiation and conduction from the façade that may not represent actual

108

temperatures with the apartment. The peak temperature in the bedroom was 94.1 F, the highest
recorded temperature in any of the units. This unit also experienced the second coolest
temperature (69.3 F), giving it the largest swing in temperatures of all the units in the test (24.8
F). It is hypothesized that this unit experienced the largest temperature swings because it is
located on the top floor and exposed to both heat loss and heat gain through the roof. This
additional surface area likely caused the indoor temperatures to follow the outdoor temperatures
more closely than the lower, more buffered units. By contrast, the ground floor units experienced
temperature swings over the baseline duration of approximately 12 F – 17 F, likely due to the
thermal buffering provided by both the slab on grade construction, and the mass of earth below.
Predictably, the second floor units experienced overall temperature swings of approximately
17 F – 20 F, situating them directly between the two extremes.
The economizer did not turn on during the baseline testing period, and the data logger
was found unplugged upon a site visit in early August. It was reconnected, but there was
approximately one month of data lost.
Unit A309
Table 13: Baseline A309 Temperatures, (ASHRAE Effective Comfort Range Shaded)
A309 (West-Facing) Kitchen Bedroom Econ. Temp. Average WSD
% < 70 F 0.0% 0.0% 7.1% 2.4% 53.0%
70 F < % > 80 F 62.4% 58.8% 64.6% 61.9% 35.5%
80 F < % > 85 F 27.4% 31.8% 21.5% 26.9% 7.4%
% > 85 F 10.2% 9.4% 6.9% 8.9% 4.1%
Total 100.0% 100.0% 100.0% 100.0% 100.0%

109

Unit A309 is identical in size, layout, and effective location in the building as Unit A305. It is
not surprising then that their temperature range breakdowns are very similar. It is surprising,
however, that even with the large temperature swings seen in the top floor units, neither
experienced temperatures below 70 F for any significant amount of time.
The economizer in A309 turned on regularly, though the data suggests it was operating in
recirculation mode anytime it turned on. Corroborating this is the data in Figure 41: Unit A309,
showing the economizer operating continuously through August. With mechanical error ruled
out, the conclusion is that the set point on the thermostat was set too low for the economizer to
ever actuate the outside-air dampers. Site visits also revealed tenants leaving the thermostats in
FAN ON mode, which means the air handler’s fan would run and air would be recirculates, but
the economizer would not necessarily bring in fresh air.
Outdoor Data Loggers:
Table 14: Baseline Outdoor Temperatures, (ASHRAE Effective Comfort Range Shaded)
Outdoor HOBOs Courtyard (Box) Courtyard (Planter) Roof Average WSD
% < 70 F 34.9% 41.8% 41.2% 39.3% 53.0%
70 F < % > 80 F 38.2% 37.9% 27.6% 34.5% 35.5%
80 F < % > 85 F 14.3% 10.2% 11.9% 12.1% 7.4%
% > 85 F 12.5% 10.1% 19.3% 14.0% 4.1%
Total 100.0% 100.0% 100.0% 100.0% 100.0%

Outdoor data loggers were included in this study to help understand any microclimatic
conditions present around the building. The data collected from these was not intended for use in
the energy simulations but rather as anecdotal feedback about how the building’s surroundings

110

react to the heating and cooling of the building. The two data loggers in the courtyard reported
similar data, while the data from the roof was predictably higher during daytime hours. Overall
the results varied considerably from the weather station data, as shown in Figure 43: Outdoor
Data Loggers, WSD File, TMY File, June – September 2012.
Overview, All Recorded Temperature Data:
The average temperatures from each of these zones are displayed in Table 15 for an overall
comparison of all the temperatures recorded in and around the building. The first number in the
name of each apartment indicates its story, followed by the first letter of the cardinal direction it
faces. For example, A109W is on the ground floor and its glazing faces west.
Table 15: Average Baseline Temperatures, (ASHRAE Effective Comfort Range Shaded)

Apartment Units Outdoor
A109W B100S B200S A209W A305W A309W Ext. HOBOs WSD
% < 70 F 0.0% 0.0% 2.2% 0.0% 8.7% 2.4% 39.3% 53.0%
70 F < % > 80 F 56.2% 70.5% 79.0% 71.1% 61.8% 61.9% 34.5% 35.5%
80 F < % > 85 F 41.1% 27.3% 18.5% 23.9% 17.2% 26.9% 12.1% 7.4%
% > 85 F 2.7% 2.2% 0.3% 4.9% 12.3% 8.9% 14.0% 4.1%
Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Figure 67 graphically displays temperature data recorded in the AX09 line of apartments
during the baseline testing period. Average high and low temperatures within each unit can be
seen increasing from June through July in each unit. Weather station data is shown at the top.
Average high and low temperatures in Unit A109 remain close all summer, and do not fluctuate
dramatically overall. This is likely due to the high thermal mass of the slab on grade floor.
Temperatures in Unit A209 vary more than in Unit A109, though not as much as in Unit A309.

111

Heat transfer through the roof in Unit A309 resulted in greater diurnal temperature swings, as
well as greater fluctuation across the testing period. Outdoor data is displayed at the top and is
most closely paralleled by Unit A309’s temperatures.

Figure 67: Average High and Low Temperatures, Units A109, A209, A309, Outdoor (WSD)

112

5.2.1 - Construction Verification
The process of verifying the built condition was direct. The building was examined visually,
construction documents were referenced, and discussions with the architect provided
confirmation of details not possible to deduce from site visits. The results were found to be
sufficient in the energy models to help maintain thermal comfort.
5.2.2 - HVAC Results and Analysis
5.2.2.1 - Pressure Relief Damper (Interior Exhaust Panel)
The relief dampers were screwed shut sometime before this study. In theory, this adds an extra
step to the procedure the tenants must go through each day to ensure their economizer is working
effectively. In actuality, the tenants appeared not to be following any schedule other than what
they thought best at the time, and the interior exhaust panels were likely doing very little before
they were screwed shut. Additionally, tenants recounted dust and noise coming through the
dampers when they were operational and overall were pleased they were shut permanently
because now they could arrange their furniture more freely.
In a building with tenants more in tune with the building, a change like this would have
been noticed. The tenants in this building though were largely unaware of the role exhaust panel
was designed to play, and none seemed unhappy to shut it and forget about it.
5.2.2.2 - Air Changes per Hour
The results of the air changes tests brought up several questions about how the HVAC system
was functioning in the units. Each HVAC system has one return duct and several supply ducts.
The dampers in the return duct open when the economizer is in recirculation mode, and the air
entering the return duct is then dispersed throughout the unit. The same process occurs when

113

using outdoor air except the return duct closes and the outdoor-air damper opens. The size of the
return duct and the outdoor-air duct must be large enough to allow sufficient CFM to enter,
based on the fan speed. The air handlers in the larger units are designed to move 1,200 CFM, and
800 CFM in the smaller units. The fans are not variable speed; when they’re activated they
operate at full capacity.
The flow hood used to record the volume of air passing through the ducts was an
AccuBalance Air Capture Hood, manufactured by TSI Inc. This flow hood, like all similar units
has an inherent rate of accuracy that must be accounted for when reviewing the results. A
research paper examining flow hoods from popular manufacturers found error rates as high as 30%
to be common (Wray, Walker and Sherman 2002). The report did not mention specific models or
manufacturers, but the data recorded by the flow hood used in the case study was irregular and
warranted review of the instruments used.
All of the units tested reported volumetric airflow rates through the return damper
significantly lower than through the supply dampers. In a perfectly sealed environment the
factors affecting flow rate through ducts are the shape and size of the ducts’ cross section, the
length of and number of bends in the ducts’ run, and the temperature of the air passing through.
These factors need to be considered when testing for duct leakage, and all these factors except
temperature have the effect of reducing flow rates. To clarify, increasing duct length, the number
of bends, or the circumferential length and shape of the duct’s cross section will cause a
reduction in airflow. It can be assumed then, that in typical ducted HVAC systems the volume of
air flowing through the supply register(s) would be less than that flowing through the return duct;
the logic being the surface friction within the ductwork impedes and slows flow. The energy lost
is converted to heat, duct flexion, noise, etc.

114

It was a surprise then, when all of the tested units registered significantly more airflow
through the supply ducts than through the return ducts. Table 16 is a summary of the results from
the air changes/hour tests.
Table 16: Summarized Results, Air Changes / Hour Tests
Unit #
Supply / Return
Ratio
ACH
(CFH/CF)
Supply Total – Return
Total CFM
Test Notes
A109W 1.73 : 1 11.06 610 Windows open
A209W 1.58 : 1 9.41 450 Windows open
A309W 1.50 : 1 9.29 405 Windows open
A302E 1.42 : 1 9.27 204 Windows open
A302E 1.37 : 1 8.94 180 Windows closed
A302E - 163 RD covered, windows open
A302E - 166 RD covered, windows closed

The three apartments with identical layouts (A109, A209, A309) all experienced similar
ratios of supply to return airflow volumes. These ratios ranged from 1.50:1 to 1.73:1, meaning
the sum of their supply ducts delivered between 150% and 173% more air than was pulled
through the return ducts. Unit A302, the east facing one bedroom (the AX09 units have 3
bedrooms) showed similar results. This discrepancy not only contradicts the presumption that
airflow should diminish as it between the return and supply ducts, it does so by such a large
degree that the results required closer examination.
If the flow hood used in the tests was inaccurate, it’s fair to assume any inaccuracy seen
in the return duct results would be paralleled in the supply duct measurements, i.e. the error rate
should apply to all measurements equally and effectively cancel out when reviewing overall

115

ratios, as opposed to volumetric quantities. If, however, the flow hood’s calibration was off by a
flat rate, e.g. + 50 CFM per measurement, than it’s possible that adding together the results from
the multiple supply ducts in each unit could have had a cumulative effect when compared to the
single return duct. If this was the case any cumulative effect would be significantly less in Unit
A302, which has only two supply ducts, compared to AX09’s five supply ducts. The ratio was
less in A302 but not by a significant amount.
Another possible explanation for the higher supply volumes is duct leakage. Depending
on whether a duct is positively or negatively pressurized, leaks result in air expelled from, or
drawn into the duct. Each economizer is comprised of an air handler (fan) to draw in and propel
the air through the ducts, a plenum chamber downstream of this fan containing the return and
outdoor-air dampers, and ductwork upstream of the fan with supply registers to distribute air
throughout the unit. When the air handler is engaged, the plenum is negatively pressurized,
creating a vacuum. Upstream of an operational air handler, the ductwork is positively pressurized.
Figure 68 shows how duct leakage reduces the effective volume of air passing through the
handler, whereas sealed ducts do not unintentionally recirculate air, Figure 69.

116

Figure 68: Schematic Diagram of Duct Leakage Behavior

Figure 69: Schematic Diagram of Sealed Ducts

117

If there is sufficient leakage in the downstream plenum chamber, surrounding air will be
brought into the plenum and mixed with the recirculating or fresh air, depending on which
damper is active. While the smoke gun tests confirmed air leakage into this plenum when the
economizers were engaged, there was no way to measure the volume or confirm whether the
leakage was enough to account for the discrepancies between supply and return volumes. To
measure this quantity, even if anecdotally, the economizer was set to recirculate indoor air and
the return duct was artificially closed with plastic wrap, as discussed in Chapter 4. The plenum
chamber, being negatively pressurized by the air handler, could then only draw air through duct
leakage. The flow hood was then used to confirm the airtightness of the return duct’s damper.
Any air coming from the supply ducts was then assumed to come from leakage at the plenum.
The results of this test could potentially account for the discrepancies in flow rates. As
the test was only performed in A302 (which has the lower volume air handler) the results are
only applicable to the discrepancies seen in this unit. It’s fair to assume, however, that a higher
flowing air handler would create more vacuum in the plenum, drawing more air through any
leaks than would a smaller handler, suggesting results could be extrapolated to other units. The
plenum in Unit A302 reportedly leaked between 163 CFM and 166 CFM with the fan on. The
differences between the supply and return volumes in this unit were between 180 CFM and 204
CFM, depending on the position of the windows. The recorded plenum leakages are close to
supply / return discrepancies and offer a very plausible explanation for the difference.

118

Figure 70: Supply and Return Air Temperatures, Unit A302
The other test performed in this unit was to compare the temperatures at the return duct
and the supply ducts. Both tests were performed during the daytime, and both showed higher
temperatures at the supply ducts that at the return duct by approximately 2.5 F. This would seem
to indicate the air was warming as it moved through the ductwork, but a more likely explanation

119

is that infiltration from duct leakage at the plenum area let warm air into the ducts. As was
previously described in 5.2 - Phase 1: Site Documentation and Experiment Setup Analysis, the
temperatures recorded at or near the building’s exterior walls differed than those deeper within
the units; the temperatures were generally higher during the day and lower during the nights than
the temperatures observed elsewhere in the units. It’s reasonable to assume from these
observations that the air in and around the plenums (which are located adjacent to at least one
exterior wall), would follow the same pattern. This could explain the warmer temperatures at the
supply registers, illustrating the importance of properly taped and sealed ductwork.
Table 17: Supply and Return Air Temperatures, Unit A302
Unit # LS 1 BS 1 RD Temperature Change Test Notes
A302E 84.5 F 85.8 F 83.4 F + 2.4 F Windows open
A302E 86.1 F 86.0 F 83.5 F + 2.6 F Windows closed

5.2.2.3 - Smoke Gun Test
The results of the smoke gun tests are incorporated into and reviewed in Section 5.2.2.2 - Air
Changes per Hour. The tests showed considerable duct leakage near the fan coil in all units. The
volume of the leaked air was not obtainable through this test, although the issue is also discussed
in Section: 5.2.2.2 - Air Changes per Hour.
5.2.2.4 - Thermostat Test
All of the thermostats tested successfully opened the dampers when set accordingly. The main
takeaway from this test was that none of the dampers opened immediately. This is an example of
the subtle differences seen in the HVAC systems in the case study compared to traditional air
conditioned HVAC systems. When using a thermostat to activate an air conditioner in a

120

residential setting, the user often receives some acknowledgement that a change made on the
thermostat has been recognized by the compressor and air handler within the system. It’s
common to hear the compressor turn on shortly after the user sets the thermostat. Even if the
response is delayed, it still allows the user to recognize a change has been registered.
Additionally, the user can go about his or her business and listen for the sound of the compressor
activating. If the compressor cannot be heard, the obvious method of confirming whether the air
conditioner is operating is to manually feel the temperature of the supply air. A supply air
temperature of 55 F has been the standard for air conditioners; it is based on the temperature of
the chilled water loop which is frequently paired together with the HVAC system. The contrast
in ambient indoor temperature compared to 55 F supply air is obvious- even from considerable
distance this temperature differential can be recognized instantly by the hands or face. It is fairly
simple then, to recognize when an air conditioner turns on- so much so it’s often not even
conscious.
The economizers in the case study building do not provide the tenants with indications
like these that they are working. In theory, if an occupant with a basic understanding of their
economizer finds it’s too warm inside, the process of reducing the temperature requires
considerable sleuthing and faith in one’s assumptions. The main difference is that any changes
made on the thermostat will likely not register until the next day. It it’s too warm inside, the
tenant will probably attempt rectification by lowering the cooling set point. If there is no
perceptible change in the temperature of the supply air the tenant may reduce the set point further.
The danger of doing this is that a lower cooling set point will provide lower supply temperatures,
but will provide fewer hours of cooling. Additionally, there is no immediate feedback to the user
that the change will result in lower temperatures. Compounding this is that lowering the set point

121

too far can result in higher indoor temperatures, a concept which understandably proved very
difficult for the tenants in the building to relate to.
The economizers do not have compressors because they do not condition the air, so there
is no audible confirmation of any change. Visual confirmation is also not possible because the
dampers cannot be seen from within the unit. To confirm the thermostats were operating
correctly the return duct was manually removed exposing the return air damper. The outdoor-air
damper can only be seen by removing the air filter and looking through the narrow slit where it
was previously located. There is no effective way to do this because this system was designed to
be a “set-and-forget” system. The theoretical occupant may look for confirmation then, by
feeling for a temperature change in the supply air. The supply air, however, can only be one of
two temperatures; in recirculation mode it’s the same temperature (or slightly different, Section
5.2.2.2 - Air Changes per Hour), and in fresh-air mode it’s the same temperature as the air
immediately outside of the unit. The only instance in which a temperature change would occur at
the supply ducts would be when the economizer switches between these two modes, which this
test showed can take up to two minutes. Without prior knowledge of this delay, and without
indication of the outside air temperature, it’s understandable that the tenant’s patience might
expire before any meaningful change was made. Tenants frequently expressed frustration when
discussing this process in interviews, describing how they gave up fiddling with the thermostat
after not knowing if they were actually making any changes. They would often give up at this
point, manually power-down the economizer altogether, and rely on the ceiling fans for relief.
5.2.2.5 - Damper Leakage Test
None of the dampers reported any leakage. The possibility of warm air leaking through them and
into ductwork was not considered a danger and was not further explored.

122

5.2.2.6 - Overhead Fan Test
The ceiling fans all reported sufficient airflow. Considering the complications associated with
getting the tenants to use the economizers, this was a very important test. Before this study began
several fans were found to be in need of replacement capacitors, and several others were found to
be rotating the wrong way, having been left in winter mode. Both of these issues were rectified
before the study began, and at the time of testing all the fans moved sufficient volumes of air.
5.2.2.7 - HEED Energy Model
The energy models created in HEED predicted both temperature and economizer usage data that
differed significantly from that recorded in the building. These variations are discussed in
Section 5.3 - Phase 2: Climatic Calibration.

5.3 - Phase 2: Climatic Calibration Analysis
The following sections review the results from the HEED energy models when using the data
from the TMY file and the weather data from the nearby weather station. The data from these
two sources are summarized below, and a comparison of their temperature ranges over the test
period is shown.

123

Table 18: Temperature Ranges, TMY Data, Weather Station Data (WSD)
TMY Data WSD Data
% < 70 F 79.9% 53.0%
70 F < % > 80 F 20.0% 35.5%
80 F < % > 85 F 0.1% 7.4%
% > 85 F 0.0% 4.1%
Total 100.0% 100.0%
Minimum 59 F 56 F
Maximum 80.1 F 94.9 F
Average 67.1 F 70.2 F

It is immediately clear from this breakdown that the TMY temperature data is
significantly cooler overall than the WSD temperature data. Worth noting is that although the
TMY data indicates 20.0% of hours were between 70 F and 80 F, the maximum temperature
across the test period was 80.1 F. This means the TMY data never exceeded 81 F. Increasing the
comfort zone to 70 F – 81 F (or simply rounding down instead of up) would result in
temperatures within or below the comfort zone 100% of the time. This comparison can also be
seen graphically in Figure 22: Typical Meteorological Year (TMY) Data, June – September. If
relying solely on TMY temperature data when designing a building, it’s clear that indoor
temperatures could be higher than expected.

124

5.3.1 - Typical Meteorological Year Data
Table 19: TMY-Based Temperature Ranges
TMY-Based Predictions A109 B100 A209 A305 A309
% < 70 F 20.9% 27.1% 29.0% 29.6% 29.6%
70 F < % > 80 F 79.1% 69.2% 62.2% 63.0% 63.0%
80 F < % > 85 F 0.0% 3.7% 8.7% 7.3% 7.3%
% > 85 F 0.0% 0.0% 0.2% 0.1% 0.1%
Total 100.0% 100.0% 100.0% 100.0% 100.0%
Minimum 68.0 F 65.0 F 65.0 F 65.0 F 65.0 F
Maximum 79.0 F 84.0 F 85.0 F 85.0 F 85.0 F
Average 73.2 F 73.7 F 74.2 F 74.1 F 74.1 F

Using the TMY data, HEED predicted indoor temperatures within or below the 70 F - 80 F
comfort for the majority of the time. Given the inaccuracies associated with using this TMY data,
these results carry little weight. The temperature data in the TMY file was so much lower than
that in the WSD file that it was surprising to see any indoor temperatures above 80 F.

125

5.3.2 - Weather Station Data
Table 20: WSD-Based Temperature Ranges
WSD-Based Predictions A109 B100 A209 A305 A309
% < 70 F 8.3% 19.4% 20.8% 21.3% 21.3%
70 F < % > 80 F 88.2% 75.4% 69.6% 70.2% 70.2%
80 F < % > 85 F 3.5% 5.0% 9.0% 7.9% 7.9%
% > 85 F 0.0% 0.2% 0.5% 0.6% 0.6%
Total 100.0% 100.0% 100.0% 100.0% 100.0%
Minimum 69.0 F 66.0 F 66.0 F 66.0 F 66.0 F
Maximum 83.0 F 85.0 F 87.0 F 87.0 F 87.0 F
Average 75.7 F 74.5 F 75.0 F 74.9 F 74.9 F

After examining how much warmer the actual (WSD-based) temperatures were than the TMY
data, it was expected that HEED’s predictions when using this data would be considerably, even
proportionally higher. The predicted temperatures were higher, but the differences were not
substantial. Most important was the percentage of time spent above 85 F, because up to this point
the ceiling fans can provide sufficient effective cooling. Neither of the data sets resulted in 85+ F
predictions more than 1% of the time, in any of the units. The amount of time spent in the 80F –
85 F range increased, but again the increase was marginal. In fact the main difference was that
less time was spent below the comfort zone when the WSD was used. Overall the differences in
predicted temperatures was insignificant, both cases resulted in units that were comfortable the
vast majority of the time, occasionally chilly, and rarely hot. This is due to the economizers in
the HEED model successfully buffering the indoor temperatures. In both cases the economizers
operated in the 5 – 10 ACH range for approximately 30% of the simulated duration. In a given

126

24 hour period this means the economizers operated for over 7 hours. This was crucial to
keeping indoor temperatures comfortable and reflects how long it takes for the mass inside the
units to cool down enough to keep the air temperatures comfortable the following day. Without
the economizer to synchronize indoor and outdoor temperatures appropriately, the units were
free to heat up.
A comparison between the predicted and actual temperatures from Unit A209 is
particularly revealing. A209 was the coolest unit overall based on the recorded data. The
equivalent discrepancies in the other units are even greater. Particularly noteworthy is that the
simulation spent over 20% of the duration below 70 F, while the actual unit never dropped this
low. In retrospect this is because the economizer likely did not operate properly in the actual unit,
while it operated as intended in the simulation. This problem also resulted in the actual unit
spending almost 15% more time within the 80 F – 85 F range than did the simulation. The actual
unit never cooled off, remaining hotter, for a longer period than the simulated unit. For
comparison, the WSD outdoor data is included in the right-hand column.

127

Table 21: A209 Recorded and Predicted Temperature Ranges
A209 (West-Facing) WSD-Predicted HOBO-Recorded Δ WSD
% < 70 F 20.8% 0.0% -20.8% 53.0%
70 F < % > 80 F 69.6% 71.1% +1.50% 35.5%
80 F < % > 85 F 9.0% 23.9% +14.90% 7.4%
% > 85 F 0.5% 4.9% +4.40% 4.1%
Total 100.0% 100.0% - 100.0%
Minimum 66.0 F 71.5 F +5.50 F 56 F
Maximum 87.0 F 91.6 F +4.60 F 94.9 F
Average 75.0 F 78.4 F +3.40 F 70.2 F

5.3.3 - Collected Outdoor Data
The data loggers collecting outdoor temperature data differed significantly from each other and
from the weather station data. They were not used in any further tests. The data is presented here
for review but did not impact the study.

128

Table 22: Outdoor Temperature Ranges, Outdoor Data Loggers
Exterior Temperatures Courtyard (Box) Courtyard (Planter) Roof TMY WSD
% < 70 F 34.9% 41.8% 41.2% 79.9% 53.0%
70 F < % > 80 F 38.2% 37.9% 27.6% 20.0% 35.5%
80 F < % > 85 F 14.3% 10.2% 11.9% 0.1% 7.4%
% > 85 F 12.5% 10.1% 19.3% 0.0% 4.1%
Total 100.0% 100.0% 100.0% 100.0% 100.0%
Minimum 60.2 F 60.5 F 55.2 F 59 F 56 F
Maximum 104.3 F 101.5 F 106.6 F 80.1 F 94.9 F
Average 74.9 F 73.5 F 74.5 F 67.1 F 70.2 F

5.4 - Phase 3: Occupant Behavior Analysis
5.4.1 - Tenant Education
All of the tenant education tests eventually resulted in the tenant reverting back to their previous
methods of interacting with their thermostats. Despite repeatedly explaining the HVAC system
and giving instructions on its use, none of the tenants exhibited any lasting changes in behavior.
One possible reason for this is because these sessions occurred later in the year when outdoor
temperatres were declining. The tenants faced minimal discomfort at this time of year so there
was no immediate need to change. Perhaps if these were conducted at the beginning of the study,
when the repercusions of ignoring the advice could lead to serious thermal discomfort, some
lasting behavioral changes would have been observed. Another possible reason for this outcome
is that it simply required a greater effort. Each apartment unit had multiple tenants, so increasing
the frequency of educational sessions could help inform all tenants.

129

Convincing the tenants to set their thermostats against their intuition proved was
relatively easy during the sessions. However, all of them quickly returned to their previous habits
regardless of any improvements in thermal comfort resulting from the sessions.
5.4.2 - Locked Thermostat
Unit A309 was chosen for this test because of the personal rapport developed between the author
and the tenants. The tenants were agreeable to relinquishing control of their HVAC system for a
period. Other tenants were approached with the idea as well and indicated varying degrees of
willingness. The tenants in A309, however, were agreeable and interested in the study- a
winning combination infrequently seen across the duration of the study. At the time this test was
performed, well after the baseline studies had been performed, the outdoor temperatures were
low enough that the economizer was not needed to maintain thermal comfort. This test was not
intended to make the apartment comfortable; the intention was to confirm that the thermostat and
the economizer were functioning as intended. The thermostat was locked through a particular
key combination, and the tenants were told that if they changed their minds about the study they
could contact the building management to come and unlock the thermostat. The test confirmed
this: the cooling set point was artificially lowered, and the economizer turned on as expected,
generally turning on around 8:00 PM, and turning off early the following morning. This further
confirmed that the HVAC systems in the building could successfully be operated provided the
occupants know how.
This test was attempted at other times in other units. Almost invariably the tenants would
power down the economizer with the override switch, claiming it was too noisy, wasn’t effective,
or perhaps most poignant, was “…operating when it wasn’t needed, so I turned it off.” Relations
with the tenants remained consistently positive and constructive, but there existed an undeniable

130

skepticism that the test was being done to assuage their complaints about overheating to the
building management. It was not uncommon for tenants to agree to participate in a particular test
and then do nothing that was asked of them. In general they were frustrated that site visits did not
result in immediate relief. Given the variety and complexity of the issues present, and the
apparent simplicity of the problem at hand (“It’s too hot!”), it required a lot of patience to get
tenants to dispel any notion of favoritism or association between the studies and the building
management. It was at times better to not perform a test to preserve the relationship. To their
credit no additional data loggers disappeared, and some of the tenants even began taking notes
during site visits.
5.4.2 - Vacant Unit Test
The testing of a vacant unit was intended to reveal at a finer level of detail the effect fresh, cool
air had on the unit’s temperature throughout the day. Like the previous test, the test occurred
when outdoor temperatures were low enough that the economizer was not needed. This was not
the best testing condition, but circumstances necessitated it. It order to get meaningful data, the
cooling set point was set artificially low, which activated the outdoor-air damper, and the
economizer then ran for several hours. The air temperature in the unit cooled by an estimated 7 F
more than it would have been had the economizer not been used. The low indoor temperatures
did not stretch into the next day, possibly because the thermal mass in the unit requires additional
time to synchronize with the incoming fresh air. In retrospect it would have been advantageous
to record surface temperatures of the active thermal masses within the unit. Unfortunately there
was a very small window of time to conduct this test, and at the very least the data provides
additional confirmation that a knowledgeable user can successfully operate the economizer.

131

Chapter 6: Conclusions and Recommendations
6.1 – Overview
The summer testing period served as a control period against which the results of the various
tests could be measured. By better understanding how the tenants used their HVAC systems
during this period, it was hoped any improvement or detriment to thermal comfort resulting from
the post-summer tests would be apparent and measurable. Many of the tests performed in this
study were dependent on cooperation from the tenants; the results of these tests in particular are
specific to these tenants. The results are definitive enough to extrapolate the findings across a
wider subject base, allowing reasonable speculation on the effects of increasing the use of
residential economizers. This chapter reviews the conclusions drawn from the results, ideas for
remedying the issues in the case study building, and methods for avoiding similar situations in
future projects.
All of the monitored units regularly experienced daytime temperatures near (or above)
the top of the thermal comfort range. The comfort zone used in this study was between 70 F and
80 F, noting as well the percentage of time between 80 F and 85 F. When temperatures were
between 80 F and 85 F, thermal comfort could still be obtained through the use of ceiling fans
can drop the apparent temperature to under 80 F. This expansion of the thermal comfort range
confirmed the validity of the use of the economizers during the design phase of the building, and
it is reasonable for this particular application. As shown in previous studies, occupants in
naturally ventilated spaces can remain comfortable across a wider variety of temperatures than
occupants in a sealed building (de Dear and Brager 1998) (La Roche, Carbon-neutral
architectural design 2011). In buildings reliant upon natural ventilation to provide cooling, the
use of an adaptive comfort model is recommended. An adaptive model takes into account factors

132

not relevant in sealed buildings, e.g. the occupant’s clothing level, comfort levels that vary
relative to outdoor conditions, and the ability to control one’s own thermal conditions. Figure 71:
Adaptive Thermal Comfort Model is an example of an adaptive model based on occupants
wearing differing levels of clothing (clo). Other models could be developed that incorporate
different levels of occupant control. For example, if the occupant has control of the natural
ventilation within their space their comfort zone would be different than an occupant without
control.
Obtaining thermal comfort in a naturally ventilated building may demand some flexibility
of the occupants; occupants with a narrow personal comfort range may find themselves
uncomfortable with some frequency. In a building sealed and protected from outdoor conditions
the need to factor in these considerations is diminished, because the outdoor conditions do not
impact the indoor conditions. Accordingly, the indoor environment in a sealed building can be
finely tuned to suit the needs of the occupants, so a static comfort model should be used. As
discussed in Section 1.1 - Problem, the widespread use of residential air conditioning has
encouraged the development of houses resistant to outdoor conditions. The case study building,
with its economizers providing mechanically-aided natural ventilation, is subject to the effects of
outdoor conditions and is an appropriate project to use an adaptive comfort model.

133

Figure 71: Adaptive Thermal Comfort Model
(ASHRAE Standard 55 2010)
Expanding the comfort range to between 70 F and 85 F significantly increases the
number of “comfortable” hours experienced in each apartment. Based on the results it could be
concluded that the apartments were comfortable most of the time, but this would be an
oversimplification of the issue. By comparing the amount of time spent below 70 F inside the
apartments against the outdoor temperatures, it’s immediately clear that the apartments never
cooled down despite the abundance of cool outdoor air. Even though the indoor temperatures
were technically comfortable most of the time, the data suggests they were at the top end of the
comfort zone. If the economizers had been operating properly, the (WSD) simulations suggest
indoor temperatures would regularly dip below 70 F. The following sections review methods for
alleviating the problems seen in the case study building, preventing similar conditions in future

134

projects, and methods for encouraging the use of economizers in residential projects where
applicable. The various recommendations are organized by their relevant sectors in the
architectural and engineering industry, as follows:
• California Energy Code Recommendations
• LEED Recommendations
• Education-Based Recommendations
• Software Recommendations
• Hardware Recommendations
• Thermostat Design Recommendations

6.2 - California Energy Code Recommendations
6.2.1 - Incentives
The California Energy Code is maintained by the California Energy Commission and is the
legislative driving force behind California’s energy reduction goals. One of the measures
included in the updated 2013 code is the incorporation of whole-house fans into residential
projects. To further encourage their use it is recommended that an incentive program be initiated
to award the owner an amount based on the savings demonstrated by an energy model. For
example, if the building’s energy model demonstrates expected savings (resulting from the
specification of economizers over air conditioners) of 50% annually, the incentive offered would
be proportional to those savings. The savings would have to be measured and recorded as proof
of compliance to ensure the economizers are being used correctly and efficiently. This could
have the indirect effect of making sure the building management is knowledgeable and attentive
to the tenants’ operational habits. The incentives would expire after a previously set period of
time, allowing the users to become familiar with the operation of the economizers. Basing the

135

incentive on the proposed savings would encourage projects with high air conditioning-based
loads to opt for economizers instead, while at the same time reducing the likelihood of
specification in an inappropriate project, or the specification solely for the incentive.
An incentive program like this would benefit larger housing projects more readily than
smaller or single family projects, because it requires a budget for special equipment. The
dwelling units would require monitoring devices to track the energy use of the economizers to
confirm their operation. Additional funds would also be required for the review and confirmation
of their success or failure.
6.2.2 - Combined Air Conditioner/Economizer
Another possible revision to the CEC is the extending of opportunities for combined air
conditioner/economizer units to residential applications. Currently this combination is regulated
and in some cases required by the code. Extending these regulations to residential buildings
could serve the same purpose they do in commercial buildings, reducing the load on the air
conditioner by “pre-cooling” the building during the nighttime. This reduces the air conditioner’s
hours of operation and could have the added benefit of shifting its cooling schedule forward a
few hours that would reduce load on the grid at peak times during the day.
Peak loading is a problem because the generators that supply power during peak hours are less
efficient than the generators supplying the grid’s baseline power. Because large scale energy
storage is not common practice, any excess power generated during the peak is typically “thrown
away.” California’s baseline energy requirement, i.e. the minimum load on the grid that is
constant across the day, is generated by steam driven turbines. These turbines are large and can
take several hours or even days to come online or go offline. As discussed in Section 1.1 -

136

Problem, mechanical efficiency is greatest under steady load; the baseline generators are kept
online and are constantly generating power. Any required increase or decrease in output, e.g. for
an incoming heat wave or seasonal shift, is implemented gradually to maintain efficiencies. Peak
load generators are smaller and are typically fueled by gas diesel, enabling them to be brought
online or offline faster than baseline generators. Peak load generators are brought online in the
morning as the population they serve wakes and begin drawing power from the grid. They’re
brought offline at night when the demand decreases to the baseline level, supplied by the more
efficient and constantly running generators. Figure 72: Typical Daily Load Curve shows the
baseline load remaining relatively constant throughout a 48 hours period. The intermediate and
peak loads appear suddenly and then disappear, requiring a generator that can do the same.

Figure 72: Typical Daily Load Curve
(Kerogen Consultants n.d.)
Reductions in the peak loads will in turn reduce reliance on the less efficient generators
that produce this power. Combining an economizer with an air conditioner will result in a
building with a lower “internal” temperature at the beginning of the day. Regardless of its

137

construction type the building will remain cooler for longer, pushing the air conditioner’s
schedule later into the day as well as reducing the overall hours conditioning is required. In a
high mass building this shift will be greater than in a low mass building because the mass will
maintain indoor temperatures for longer.
When used alone, economizers typically function as “valley deepeners” because they do
not operate during peak load times. When used in combination with air conditioners, they
function as “peak shavers” as well because they both reduce and shift the building’s peak
demand away from the grid’s peak loading period. Figure 73 expresses these terms graphically.

Figure 73: Peak Shaver and Valley Deepener
(Liescheidt Accessed 23 September 2012)
The effectiveness of renewable energy sources is also limited by this daily cycle. Wind
driven turbines are most effective when wind speed and direction are constant, which may or
may not align with a grid’s peak loading. Tidal based power generation devices produce
maximum energy in cycles that are aligned with the phases of the moon and have no relationship
to the grid’s daily load cycle. These methods of power generation can reduce the baseline loads,

138

but have the undesirable effect of making the peak load generators schedule even more extreme.
Geothermal based energy production is constant at a daily cycle and changes only minimally
with the seasons. Only solar power follows a schedule roughly aligned with the grid’s peak
loading. Widespread use of economizers can have very positive impacts on the grid’s loading
pattern by reducing overall consumption, and by reducing/shifting its peak loads.
Another benefit of a combined economizer and air conditioner package is the
corresponding reduction in operating capacity required of the air conditioner. As discussed in
Section 1.1 - Problem, oversizing of air conditioners results in higher upfront and operating
costs. When choosing a combined package, the CEC could include a cap on the air conditioner’s
capacity more restrictive than that on a standalone air conditioner. For example, the air
conditioner in a combined package could be restricted to an output capacity of 75% of the
building’s cooling requirements per the Manual J calculations, leaving the economizer to handle
the remainder. This would eliminate the inefficiencies caused by oversizing, and because the air
conditioner would be unable to provide all of the required cooling it would also discourage
reliance solely on the air conditioner.

6.3 - LEED Recommendations
Modern buildings that include passive design strategies are often labeled as such, e.g.
green architecture, eco-friendly, sustainable design, etc. Although the focus on reducing energy
consumption in buildings is step forward for the industry, it remains an optional form of positive
reinforcement at a time when inclusion ought to be compulsory. Leadership in Energy and
Environmental Design (LEED), a building rating system operated by the US Green Building

139

Council, is a popular choice among designers and building owners looking to introduce
sustainable measures into their buildings. But if there were no economic benefits to obtaining
these types of certifications how many building projects would strive for it? Hopefully this focal
shift will continue, and will evolve from an additional (and separate) for-profit design option,
back into the mainstream methods of building design. It’s imperative that all building types
incorporate these measures into their design and not only those that benefit from incentives. It
will be a major milestone then, when the architect and the energy consultant are one, and there is
no need for green rating systems. These views notwithstanding, recommendations for the LEED
rating system are presented.
6.3.1 - Commissioning
There are several additions to the LEED requirements which if implemented could help
encourage the specification of economizers in residential buildings. Interest in revisions to
LEED’s guidelines stemmed from the case study building’s certification as a LEED Platinum
building- the highest available level of certification. While the progressive design and
specification process undergone by the design team is worthy of praise the building itself has
never thermally performed as intended. It is recommended that LEED incorporate a performance
based commissioning process into its requirements to ensure a building is not only built as
designed but is also functioning as designed. After it was completed the case study building went
through a commissioning procedure to ensure the HVAC system installed corresponded with the
specifications and that it was installed correctly. It did not investigate whether the tenants were
using the HVAC system effectively. This change could also be combined with a post-occupancy
evaluation (POE) to better understand from the occupant’s perspective any shortcomings or
failings in the building or its systems.

140

6.3.2 – LEED-Specific Education Program
Inclusion of a required education program into projects using economizers could also help ensure
the HVAC system is operated properly. When applying for LEED points through the
specification of economizers, it’s recommended that a prerequisite be added that incorporates
some level of tenant and building management education into the project. It is critical that any
education stipulation be in the form of a prerequisite so the owner/LEED consultant could not
simply opt out of those points, thus ensuring economizers are not specified for inappropriate
projects.
Alternatively the existing Awareness and Education credits could be modified to
incorporate HVAC-specific training and then paired with points received through specification
of an economizer. AE Credit 1.1 - Basic Operations Training, 1.2 – Enhanced Training, 1.3 –
Public Awareness, and Credit 2 – Education of Building Manager could all be bolstered by
increasing their reach to all types of residential projects and by including HV AC-specific
information in their agendas.

6.4 - Education-Based Recommendations
Education programs are not specific to LEED, of course. Regardless of the framework
surrounding an education program, the importance of the information conveyed cannot be
overstated. As we demand higher performance from buildings while demanding they use
increasingly less energy, the occupants inside will have an ever-increasing role in keeping the
building well-tuned and operating as designed. Buildings reliant on natural ventilation are
particularly susceptible to outdoor conditions and if not operated effectively the metaphorical

141

climate door is left wide open, and indoor conditions can quickly deteriorate. The very nature of
increasing efficiencies means getting more from less and implies the use of smarter, more
evolved methods than previously used. In the case of architecture and building design increased
efficiencies mean getting more performance from less energy consumption and can sometimes
rely heavily on occupant participation.
6.4.1 - Building Management Education
In large multifamily building projects it’s typical for a building management company to handle
the operation and management of the premises. Particularly in buildings with high tenant
turnover rates it’s important that the management is capable of conveying any information
relevant to achieving and maintaining indoor environmental quality. As demonstrated by the case
study building a breakdown in this process can have serious ramifications- everybody
complained of the heat yet nobody could fix it.
The success seen in Section 4.4.2 - Locked Thermostat and Section 4.4.3 - Vacant Unit
Test showed that economizers are capable of cooling the units if used properly. The
comparatively lackluster results from Section 4.4.1 - Tenant Education show the resilience
required when attempting to educate a tenant that is not interested or is otherwise resistant to
learning. Compounding this is the issue of scheduling educational meetings at times when the
potential effects of ignoring them cannot be felt. The education tests performed in the case study
building were not begun in earnest until the autumn and winter months, when outdoor
temperatures were low enough for the building to remain cool regardless of the tenants’ behavior.
The management then must be able to work with the tenants until they’re capable of operating on
their own. This process necessarily repeats anytime a new tenant moves in and may require an
increase in effort on the management’s part during warmer months. The importance of

142

knowledgeable, capable management to serve as the disseminator of information needs to be
considered and appropriately accounted for in the budget.
6.4.2 - Tenant Education
The occupants in this study were repeatedly observed interacting with the economizer as though
it was an air conditioner, and despite explanations about proper set points and when to use the
economizer, the data overwhelmingly displays this error in perception. Because the thermostat
appears to provide cooling, the occupants expect it to. The data suggests that the economizer was
turned on when the occupant wanted instant cooling, unaware that this system, as a thermally
delayed system is required to begin operating long before the effect is to be noticed. Although it
provides around 10 air changes per hour, the cool air needs time to chill the building materials
within the unit to provide the maximum effect. To achieve cooling on a warm summer day, the
economizer needs to have been operating through the preceding evening, and the occupant needs
to then insulate their apartment from the warming exterior. Throughout the study it was difficult
to successfully convey the idea that the economizer is used to provide tomorrow’s cooling, not
today’s cooling.
By forcing the air handler to stay on (FAN ON mode), and setting the cooling set point as
low as possible, the user could hear and feel air moving through the ducts and their apartment
(Figure 74). Furthermore, because it is not easily determined if the economizer is introducing
outdoor air or re-circulating indoor air, the fan could potentially remain on indefinitely, wrongly
suggesting to the user that there is a problem with the system’s ability to cool the air and not a
problem with the settings. This process led the tenants to believe that they were doing everything
in their power to provide cooling and to assume the system was broken. Interviews lead to the
conclusion that at this point that they generally gave up trying to “fix” it, and because the air

143

coming from ducts was as warm as or warmer than the indoor air, they generally turned the
economizer off and relied only on the effective cooling provided by the overhead fans (Case
Study Tenant 2012). Whereas air conditioners are generally clear in what they provide
(immediate cool air), the combination of an economizer attached to a traditional thermostat
leaves room for miscomprehension and confusion of what the individual settings do.

Figure 74: Set Point Errors Seen at Thermostats
The gap between how the tenants were operating the building and how they should have
been operating the building was extreme. Most disturbing, however, was the gap in the
conceptual understanding of how the HVAC systems operated and the resistance to absorbing
new concepts. The effort required convincing an occupant to forget what they know, and relearn
how to set a thermostat proved significant. This illustrates the importance of a building
management capable of repeatedly meeting the demands of the occupants. When choosing an
untraditional HVAC system (like an economizer) it’s recommended the controls provided are
carefully selected to minimize the possibility of misuse on the tenant’s part.

144

6.5 - Software Recommendations
The software program HEED was used to model the various case study apartments. HEED is a
validated software program, meaning it has been successfully tested against ASHRAE Standard
140 for energy modeling and predictive purposes. Several recommendations are proposed to
further increase the predictive capacity of HEED and other comparable software.
6.5.1 – Occupant Behavior
One recommendation stemming directly from the results of this thesis is the incorporation of a
variety of occupant types into the building being modeled. While this can be done circuitously by
purposely skewing certain settings in HEED’s advanced screens, it’s recommended that this
option be given due consideration. If the option to adjust the virtual occupant’s ability to tune the
building were obvious and upfront it might encourage designers to consider this early in the
design phase. The case study building requires the occupants to adjust the thermostat’s heating
and cooling set points, to open and close the windows and the blinds at certain times each day,
and to operate the ceiling fans accordingly. Skipping any of these steps or performing them out
of order can result in decreased thermal comfort. In retrospect it’s possible to mimic the erratic
behavior witnessed in the building, however, there was no reason to suspect during the design
phase that this behavior would be encountered. Incorporation of this consideration as an easy and
obvious option in future software developments could reduce the risk of tenant behavior
impacting thermal comfort levels.
If comfort levels in the model are shown to be susceptible to occupant behavior,
measures could be taken to build in safeguards and the choice of a specific HVAC system itself

145

could be catered to suit particulars of the design. This level of pre-design consideration can be
expected as occupants are called upon to play greater roles in the overall operation and
calibration of buildings and dwelling units.
The software could highlight any areas where thermal comfort could potentially be
compromised by changing the sensitivity of the user. If the sensitivity of the user is lowered the
software would assume the occupants will not behave as assumed, and can alert the designer.
Additionally, recommendations on changes to the HVAC system can be given based on the
chosen occupant. Compound HVAC systems (as seen in the case study building) could then be
designed to be less sensitive to negligent occupants. This idea is discussed further in Section 7.2
– Software Development.
6.5.2 - Weather Data Review
Another recommendation, which is more specific to buildings using economizers or night
flushing systems, is based on reviewing climatic data before the building design commences. A
method of quickly surveying the climatic data used in the energy model for compatibility with
economizer based HVAC systems could inform the HVAC specification process before the
design process begins. By weighing the major factors that dictate whether an economizer should
be considered, such as diurnal temperature swings during warm months, expected maximum and
minimum temperatures, expected humidity levels, etc., a designer can get a quick understanding
of whether an economizer is likely to succeed (comparable to the analyses Climate Consultant
provides). If the location’s climate indicates the potential for success is right on the border, the
designer will know quickly what architectural elements may need to be factored into the design
and can proceed accordingly. Design recommendations similar to those provided by HEED and
Climate Consultant could be preprogrammed and given based on particular climatic data. As

146

previously discussed it’s prudent to use both climate zone data and local climate data to confirm
consistencies or discrepancies. It’s foreseeable that for some locations the TMY data might
indicate economizers are suitable, while the local climate data differs. A discrepancy of this type
would immediately tell the designer particular care is required to make the system work.

6.6 - Hardware Recommendations
Passively cooled buildings are good candidates when considering features like window fins and
overhangs. Because these buildings are designed for a specific location and climate, the addition
of location-specific features is a logical evolution of the design process. For further information
see 2.2 – Case Study Comparison, Economizers and Air Conditioners, as well as the report
Analysis of Various Cooling Strategies as Modeled in HEED (Kensek and Lanning, Building
Analytical Modeling: Analysis of Various Cooling Strategies as Modeled in HEED 2012).

6.7 - Thermostat Design Recommendations:
One of the major factors contributing to the failures seen in the case study building is the
interface of the particular thermostat used. All of the apartments used a variety of the Carrier
Debonair 450- a “smart” thermostat capable of receiving and processing multiple input levels.
This was a necessary requirement of the thermostat specified in the building, as each apartment
provided indoor temperature, outdoor temperature, and enthalpy data to the thermostat, which
then processed the data and instructed the economizer accordingly. The effect of specifying this
particular thermostat, however, was that its interface proved too complicated and confusing for

147

the occupants to use effectively. Several recommendations regarding the design of thermostat
interfaces are posited for consideration in the following sections.
6.7.1 - Feedback
The economizers in the case study building measure both indoor and outdoor temperature,
meaning they “know” the daily temperature swings. If these data were readable, even if only for
the preceding 24 hours, it would be possible to deduce two things: whether or not there were
sufficient outdoor temperatures to provide indoor cooling and whether or not the indoor
temperatures cooled successfully, based on the previous night’s outdoor temperatures. A
feedback loop providing even basic data about the success (or lack of) of the user’s thermostat
settings could at the very least tell the user their current thermostat settings are ineffective. It is
not currently possible for the occupants to know definitively if they set their thermostats
correctly because they do not know the previous evening’s temperatures. If they wake up to a
warm apartment there’s no way to determine if the overheating was due to incorrect settings, or
because previous evening’s temperatures were not cool enough for the economizer to engage the
thermostats.
Perhaps the best method of reducing energy consumption is to create an HVAC system
that can perfectly predict the user’s desires, automatically adjust to fluctuating exterior
conditions, and accommodate multiple thermal comfort ranges simultaneously. In theory, if the
user is able to input detailed personal thermal preference data into the HVAC system, the HVAC
system should then be able to choose most efficiently how it will provide conditioning. To that
effect, the more user types a given system must accommodate (in the case study there was a
minimum of 70 different users, though probably closer to double that) the more automated it
should be (Norman 1988). While this was the original intention of connecting the economizer to

148

the smart thermostat, the details of the pairing resulted in increased complexity that thwarted
correct operation. Short of having a fully automated system, there should be clear signals
indicating when a setting needs to be changed to maintain or achieve thermal comfort. If the
economizer is being used only during the day, and never at night, there should be some
indication to the user that it is not being used as designed. For these residential units, it would
have been helpful to have a readout showing the outdoor temperature, so the indoor temperature
could be compared against it, and the user could determine when to run the economizer.
Additionally, there should be some indication to the user when the system is using fresh air or re-
circulated air, so they may decide to change their set point accordingly. The thermostat might be
able to change the set point based on a few questions to the occupant. Adding a basic layer of
“checks and balances” into the user interface could help improve indoor conditions for some
users.
6.7.2 - Simplicity
Although seemingly contradictory to the recommendations proposed in the previous section, it is
equally important that the thermostats are easy to understand. If the data displayed is sufficient
but intelligible to the “average” user, the type of data displayed is rendered useless. The fields of
graphic, ergonomic, and interface design can be tapped to incorporate some of the principles
behind displaying large amounts of data clearly. The Toyota Prius’ instrument display (Figure 75)
is an example of a readout successfully showing complicated information in a clear and
comprehensible graphic.

149

Figure 75: Toyota Prius Consumption Display
(Gable and Gable 2007)
In projects deemed unsuitable based on budgetary or other considerations, simplified
versions can be designed without the need for a digital display. A series of audio or visual cues
can be incorporated to help an occupant recognize when conditions are appropriate for the
economizer to be used.
6.7.3 – Specificity
It was concluded from the studies that the specification of a thermostat normally used for an air
conditioner based HVAC system proved confusing for the occupants. The thermostats have a
COOL setting, which in a different application would trigger the activation of an air conditioner.
In the case study building this setting instead triggers the economizer; however, the economizer
does not provide cooling in the same sense as an air conditioner. The tenants, having not been

150

briefed on the concepts and proper operation of the economizer regularly interpreted this
thermostat setting as a means of providing cooling on demand.
In projects using combined economizer/air conditioner packages a traditional thermostat
can be used. The occupant does not control when each side operates, so only has to adjust the set
points, and the thermostat will then decide to use either the economizer or the air conditioner or
both. In projects using only economizers, it’s reasonable to recommend the thermostat be
tailored to its respective constraints.

151

Chapter 7: Future Work
This chapter is divided into two main sections reflecting areas where future development could
have a substantial and positive benefit for economizers and naturally ventilated strategies in
general. Section 7.1 – Thermostat Design, User Interface examines the shortcomings of current
thermostat designs, with particular attention paid to the user interface (UI). Section 7.2 –
Software Development reviews how the addition of occupant behavior into energy modeling
software could increase their predictive capacity and where this addition might be placed within
the software’s UI.
7.1 – Thermostat Design, User Interface
Traditional thermostats are designed to work in a variety of applications. The thermostats in the
case study, for example, can technically work in buildings with heating, air conditioning,
humidifiers, economizers (and the multiple temperature/enthalpy inputs they require), heat
pumps, etc., or a combination of systems. In addition they can be programmed to accommodate
which hours of each weekday are typically occupied or unoccupied, what actions should be taken
during vacations, and provide alerts when various components fail or are due for service, etc.
The effect of widening the applicability is, not surprisingly, a UI with many features not
applicable to any one specific building. The Carrier Debonair’s user manual is over 100 pages
long, an excerpt of which is shown in Figure 76: Carrier Debonair's Display Functions. The extra
features can result in confusion when setting the thermostat, leading to thermal discomfort, and
more importantly, a misunderstanding on the user’s part as to its origins. If the user believes the
thermostat is set correctly, they will naturally look elsewhere for potential problems in the
HVAC system, even if there are none.

152

Figure 76: Carrier Debonair's Display Functions
(Carrier Corporation n.d.)
The results of this study, in particular Section 4.4.2 - Locked Thermostat, showed
widespread misunderstanding by the tenants as to the basic functionality of their thermostats.
Some of this confusion was due to the thermostat’s COOL function, which suggested the ability
to providing cooling on demand. In other projects employing the same thermostat this might be
the case, but in the case study building this labeling misnomer caused recurrent belief on the
occupants’ part that their HVAC system was faulty, because it was not responding
“appropriately.”
In applications where the thermostat’s settings are locked in place, primarily commercial
and institutional projects, these considerations are not applicable. In these applications it’s likely
the HV AC system serves many occupants and is controlled by specially trained operators. In
applications like the case study building, where each thermostat controls a small, separate zone,
and where there is limited or no oversight, the variety of potential users is limitless. It is
therefore recommended the thermostats used in these applications account for both the variety of

153

potential users and the specifics of the HV AC system they’re tied to. Both the design and the
functionality of thermostats can be revised to better suit the needs of the occupants. In his book
The Design of Everyday Things, Donald Norman discusses the importance of factoring in
usability and functionality into the design phase of objects and systems. Several points are
touched upon in the following sections which, if incorporated into the design of future
thermostats, could directly improve their usability and therefore indoor conditions they serve.
7.1.1 – Feedback
Most important to the functionality of an object that requires comprehensive user interaction is
the ability to provide feedback to the user. As previously discussed in Section 6.7.1 - Feedback,
the thermostats in the case study building provide little to no feedback to the user. The findings
in that section corroborate Norman’s conclusions. Feedback from the thermostat regarding how
different thermostat setting affect indoor conditions could greatly increase the user’s ability to
learn which settings are appropriate for their specific conditions.
One possible method of conveying such information, for example, how the effectiveness
of the economizer changes as the set points are changed, is to provide the user with visual cues.
If the heating and cooling set points are set outside normal operating conditions or set such that
the economizer’s ability to cool is compromised, the user receives cues in the form of
increasingly contrasting set point colors. A thermostat specifically tailored to pairings with
economizers might also give a prediction of how well it will cool the space based on the current
settings. To provide such a prediction it must factor in previous evening’s temperature range and
then formulate a prediction based on that data combined with the current set points. Basing the
predictions on the previous night’s temperatures may not be as accurate as basing it on actual,

154

predicted weather data, but it enables the thermostat to remain unconnected from the internet and
makes use of data already being measured. Both of these factors will help keep costs down.
As seen in Figure 77 and Figure 78, widening the heating and cooling set points too far
causes the set points to brighten (in this example) as well as a decrease in the predicted capacity.
These sketches provide one possible method of displaying such information and can be refined
for clarity and additional functionality as needed. They are meant only to demonstrate a method
of simultaneously showing how indoor conditions will change based on set points adjustments
and the current indoor and outdoor temperature, which an advanced user can use to more finely
adjust the set points. These changes are circled in red, in Figure 78. By focusing the thermostat’s
display on the results – even if only predicted– the user gains an immediate understanding of the
outcome of their adjustments. A typical thermostat like the Carrier model used in the case study
building gives no indication of the impact a particular adjustment will have. As previously
discussed the thermostat can actually mislead the user through an overabundance of features with
no indication of when the settings are correct or incorrect.

155

Figure 77: Conceptual Thermostat, Comfort Range 65 F - 78 F, Predicted Efficacy: 81%

Figure 78: Conceptual Thermostat, Comfort Range 62 F - 83 F, Predicted Efficacy: 62%

156

7.1.2 – Constraints
Another major point discussed in The Design of Everyday Things is the inclusion of constraints
to prevent unintended misuse. In certain applications it may be deemed best to limit the user’s
control over the thermostat in an attempt to reduce the possibility of user error. While the best
theoretical design would be intuitive and self-explanatory for all users, the reality is that all
products designed for use by the public are influenced by many factors besides usability.
Designing for a specific user is not an option for a product as widespread as a thermostat. This
ubiquity can be accounted for in the design process if it’s given due consideration, but more
often than not the end result is maximum functionality controlled with as few buttons as possible.
The user then must reference the accompanying manual to decipher how best to operate it based
on their application. Like the thermostat itself, the manual has far more content than any one
application or user requires, hence the Carrier manual’s 100+ pages.
Increasing both the variety and specificity of available thermostats will simplify their
design and allow the inclusion of user-oriented design decisions. Built-in constraints are an
example of a user-oriented design decision that requires some specificity. A thermostat
controlling an economizer will have a different set of constraints than one controlling an air
conditioner. Without all the redundant features this type of thermostat could likely be designed
and manufactured for less than current multi-tool type thermostats, while providing a
straightforward and clear UI.
One other possible method of constraining the functions of an economizer’s thermostat is
to separate it entirely from all other HV AC controls. The case study building’s thermostats
control both the heating and the cooling of each dwelling unit. Like most ducted residential
HV AC systems one air handler is used to move air regardless of whether it’s heated, cooled, or

157

naturally ventilated. Although the systems are linked mechanically the controls could be
separated allowing greater specificity to the design.
7.1.3 – Monitoring
Another development relevant to thermostats is an increased ability to track and display data. As
energy consumption goals become stricter, energy monitoring will necessarily become more
widespread. Energy monitoring is a form of feedback that is intended to be reviewed over a
period of time, unlike the immediate feedback stream discussed in Section 7.1.1 – Feedback.
Reviewing consumption over a period allows averaging of data, which can be used to confirm
correct operation of components. It also enables any changes - intended or otherwise – in
consumption to be measured. Reviewing one’s personal energy consumption also provides a
baseline against which energy saving measures can be compared. While the technology required
for this type of monitoring is available and was used in data logger and split ring form, its
integration into thermostats increases costs and is currently limited to high-end building control
dashboards. The additional equipment required within the thermostat, combined with the need
for an interface capable of displaying it can quickly drive up production costs. One potential
inexpensive solution is to provide a USB port at the thermostat, much like that found on a data
logger. Although not a practical solution for all users it could still benefit the overall evolution of
the monitoring and displaying of data. Additionally, apps can be developed for smart phones,
pads, and computers.
7.2 – Software Development
The inclusion of various user types into energy modeling software serves the same purpose as
increasing thermostat usability, but approaches it from the opposite direction. Whereas the
development of user friendly thermostats attempts to make user behavior reflect modeled

158

conditions (by making HVAC control easy and straightforward), the addition of occupant
behavior into energy models attempts to make the model less precise, to better reflect field
conditions. The case study provided a clear example of occupant behavior having a significant
and detrimental impact on indoor conditions. While it’s conceivable to imagine a time when
thermostats are simplified to the point of nullifying any undesirable occupant behavior, the
reality is that thermostat design has a lot of catching up to do. In the meantime confusion like
that seen in the case study will continue to result in uncomfortable indoor conditions and/or
wasted energy.
The development of this feature in software then is not solely intended to increase the
accuracy of the software. The findings from the case study suggested the process of cooling the
apartments was too demanding/and or complex for the occupants to perform on a regular basis.
Adjusting either the overall user type, e.g., Extremely Diligent, Diligent, Competent, Careless,
etc., or adjusting the accuracy with which each HVAC component is operated could help
determine the sensitivity of the design on proper use of the economizer. In the case study
building, and presumably other naturally ventilated buildings requiring occupant participation,
varying the accuracy each HVAC component’s operation could pinpoint potential areas of
concern. For example, the case study requires daily adjustment of windows, blinds, and ceiling
fans, and theoretically only seasonal adjustment of the thermostats. If the designer were able to
easily assign a “user operational accuracy”, e. g., 100% to the windows and blinds, 60% to the
ceiling fans, and 10% to the economizer, it would quickly become apparent which components
impact thermal comfort the most.

159

7.2.1 – Weather Analysis Tool
As discussed in Section 6.5.2 - Weather Data Review, the inclusion of a weather analysis tool
either within or separate from energy modeling software could prevent economizers from being
specified in unsuitable climates. Analysis software like Climate Consultant (The Regents of the
University of California 2012) is useful for this type of analysis, and provides a point of
departure for specific HVAC strategies to be analyzed.

7.3 – Additional
Additional areas of exploration include determining theoretical energy savings at a national scale
if economizers were used widely. The savings would likely be tremendous. Determining the
savings potential in other climates could possibly reveal unforeseen opportunities to use
economizers. Additionally, determining the percentage of existing economizers operating below
their potential may show that economizers only result in low energy consumption if operated
correctly.

160

Bibliography
1. Abode Communities. "Floor Plans, Construction Document." Abode Communities, 2009.
2. —. Green Living Guide. 2009. http://abodecommunities.org/downloads/LEED-resident-
manual.pdf (accessed April 4, 2013).
3. ASHRAE. ASHRAE Standard 140-2001- Standard Method of Test for the Evaluation of
Building Energy Analysis Computer Programs. July 1, 2004.
4. ASHRAE Standard 55. Technical Report, Atlanta: American Society of Heating,
Refrigerating, and Air-Conditioning Engineers, Inc, 2010.
5. Attia, Shady, and Andre de Herde. "Designing the Malqaf for Summer Cooling in Low-
Rise Housing, an Experimental Study." Passive and Low Energy Architecture. Quebec
City: PLEA2009, 2009.
6. California Energy Commission . 2008 Building Energy Efficiency Standards. January 1,
2010. http://www.energy.ca.gov/2008publications/CEC-400-2008-001/CEC-400-2008-
001-CMF.PDF (accessed December 2, 2012).
7. California Energy Commission . Small HVAC Problems and Potential Savings Reports.
Technical Report, 500-03-082-A-25, 2003.
8. California Energy Commission. http://www.energy.ca.gov. July 20, 2011.
http://www.energy.ca.gov/2011_energypolicy/documents/2011-07-
20_workshop/presentations/Revised_Zero_Net_Energy_Definition.pdf (accessed March
10, 12).
9. —. http://www.energy.ca.gov. April 27, 2011.
http://www.energy.ca.gov/title24/2013standards/prerulemaking/documents/2011-04-
27_workshop/presentations/HVAC_Controls_and_Economizing.pdf (accessed May 5,
2013).
10. —. http://www.energy.ca.gov. 2013.
http://www.energy.ca.gov/maps/renewable/building_climate_zones.html (accessed May
13, 2013).
11. —. Revised Zero Net Energy (ZNE) Definition. July 20, 2011.
http://www.energy.ca.gov/2011_energypolicy/documents/2011-07-
20_workshop/presentations/Revised_Zero_Net_Energy_Definition.pdf (accessed
December 10, 2012).
12. Carrier Corporation.
http://www.docs.hvacpartners.com/idc/groups/public/documents/techlit/88-506.pdf. n.d.
13. Case Study Tenant, interview by Tighe Lanning. Case Study Interview (November 05,
2012).

161

14. Cox, Stan. AC: It's not as cool as you think. July 18, 2010.
http://www.guardian.co.uk/environment/2012/jul/10/climate-heat-world-air-conditioning
(accessed February 2, 2013).
15. Crawley, Drury. Which Weather Data Should You Use For Energy Simulations of
Commercial Buildings? Tenchical Report, ASHRAE Transactions 104, 1998.
16. Crawley, Drury, Jon Hand, and Linda Lawrie. "Improving the Weather Information
Available to Simulation Programs." D.B. Crawley, J.W. Hand, L.K. Lawrie, Improving
the weather information available to simulation programs, in: Proceedings of Building
Simulation, IBPSA, Vol. II. Kyoto, 1999. 529–536.
17. de Dear, R, and G. S. Brager. "Developing an adaptive model of thermal comfort and
preference." ASHRAE Transaction, 104(1a), de Dear, R., & Brager, G. S. (1998).
Developing an adaptive model of thermal comfort and preference. ASHRAE Transaction,
104(1a), 145-167 1998: 145-167.
18. European Commission. New4Old.eu. 2008.
http://www.new4old.eu/guidelines/D3_Part2_H2.html (accessed April 08, 2013).
19. Field, Kristin, and Michael Brandemuehl. "Effects of Variations of Occupant Behavior
on Residential Building Net Zero Energy Performance." Proceedings of Building
Simulation 2011. Sydney: 12th Conference of International Building Performance
Simulation Association, 2011.
20. Gable, Christine, and Scott Gable. About.com. 2007.
http://alternativefuels.about.com/od/vehiclereviews/ig/2007-Toyota-Prius.--ch/07-Prius-
full-hybrid-power.htm (accessed April 3, 2013).
21. Galuzzi, Luca. LucaGaluzzi.it. n.d.
http://www.galuzzi.it/photos.aspx?l=IT&aID=3&fID=213&it=1 (accessed June 1, 2013).
22. Herman, Bernard. "The Embedded Landscapes of the Charleston Single House, 1780-
1820." Perspectives in Vernacular Architecture,, 1997: 41 - 57.
23. http://www.energystar.gov/. n.d.
24. Kempton, William, Daniel Feuermann, and Arthu McGarity. "I always Turn it on Super":
User Decisions About When and How to Operate Room Air Conditioners." Energy and
Buildings, 18, 1992: 177 - 191.
25. Kensek, Karen, and Tighe Lanning. Building Analytical Modeling: Analysis of Various
Cooling Strategies as Modeled in HEED . Technical Report, CS11SCE1310.3, 2012.
26. Kensek, Karen, Tighe Lanning, Tim Kohut, and Joon-Ho Choi. "Economizer
Performance and Verification: The Effect of Human Behavior on Economizer Efficiency
and Thermal Comfort in Southern California." Los Angeles: Unpublished, January 03,
2013.
27. Kerogen Consultants, . http://www.natgas.info/html/gasusage.html. n.d.

162

28. Kohut, Tim, and Murray Milne. "Eliminating Air Conditioners in New Southern
California Housing." Proceeding of the American Solar Energy Society Conference,
2010.
29. La Roche, Pablo. Carbon-neutral architectural design. CRC Press, 2011.
30. La Roche, Pablo, and Murray Milne. "Effects of Window Size and Mass on Thermal
Comfort using an Intelligent Ventilation Controller." Solar Energy, 2004: 421 - 434.
31. Liescheidt, Steven. Economizers in air handling systems .
http://www.cedengineering.com/upload/Economizers-In-Air-Handling-Systems.pdf ,
Accessed 23 September 2012.
32. Manter, Greg. The Charleston Single House. n.d.
http://www.buildinghomegarden.com/charleston-house.html (accessed June 1, 2013).
33. Milne, Murray, Jessica Morton, and Tim Kohut. "ENERGY EFFICIENT
AFFORDABLE HOUSING;. Validating HEED's Predictions of Indoor Comfort."
Proceedings of the American Solar Energy Association Conference. Denver, 2006.
34. Nabokov, Peter, and Robert Easton. Native American Architecture. New York: Oxford
University Press, Inc., 1989.
35. Norman, Donald. The Design of Everyday Things. New York: Basic Books, 1988.
36. Onset Computer Corp. Addressing Comfort Complaints with Data Loggers. Technical
Report, Bourne: Onset Computer Corporation, 2012.
37. Pacific Gas and Electric Company. pge.com. 2013.
http://www.pge.com/includes/docs/pdfs/about/edusafety/training/pec/toolbox/arch/climat
e/california_climate_zone_08.pdf (accessed June 01, 2012).
38. Pacific Gas and Electric.
http://www.pge.com/includes/docs/pdfs/about/edusafety/training/pec/toolbox/arch/climat
e/california_climate_zone_09.pdf. Long Beach, n.d.
39. Parker, Danny, David Hoak, and Jamie Cummings. Pilot Evaluation of Energy Savings
and Persistence from Residential Energy. Technical Report, ACEEE Summer Study on
Energy Efficiency in Buildings, 2010.
40. Plumer, Brad. www.thewashingtonpost.com How air conditioning transformed the U.S.
economy. July 7, 2012.
http://www.washingtonpost.com/blogs/wonkblog/wp/2012/07/07/how-air-conditioning-
transformed-the-u-s-economy/ (accessed September 19, 2012).
41. Rosen, Rebecca. The Atlantic - Keepin' It Cool: How the Air Conditioner Made Modern
America. July 14, 2011. http://www.theatlantic.com/national/archive/2011/07/keepin-it-
cool-how-the-air-conditioner-made-modern-america/241892/ (accessed September 12,
2012).

163

42. Rosenberg, Michael. Analyzing Air Handling Unit Efficiency. Technical Report, Bourne:
Onset Computer Corporation, 2012.
43. Seeley International Pty Ltd. http://www.breezairuk.com. 2006.
http://www.breezairuk.com/evaporative_cooling.asp (accessed June 1, 2013).
44. Steadman, Robert. "The Assessment of Sultriness. Part II: Effects of Wind, Extra
Radiation and Barometric Pressure on Apparent Temperature." Journal of Applied
Meteorology, 1979: 874–885.
45. The Regents of the University of California. Climate Consultant. 2012.
46. U.S. Census Bureau. 2003. http://www.census.gov/statab/hist/HS-12.pdf (accessed
January 10, 2013).
47. UCLA Department of Architecture and Urban Design. HEED Validation Reports.
October 2012. http://www.energy-design-
tools.aud.ucla.edu/heed/pdfs/HEED_Summary_Validation_Report.pdf (accessed June 6,
2013).
48. UCLA Department of Architecture and Urban Design. Home Energy Efficient Design.
Los Angeles, January 28, 2012.
49. United States Environmental Protection Agency. Ceiling Fans Key Product Criteria.
2013. http://www.energystar.gov/index.cfm?c=ceiling_fans.pr_crit_ceiling_fans
(accessed April 2013, 13).
50. weatheranalytics.com. Weather Analytics, Precise Global Intelligence. 2013.
www.weatheranalytics.com (accessed June 6, 2012).
51. Wray, Craig, Iain Walker, and Max Sherman. Accuracy of Flow Hoods in Residential
Applications. Technical Report, Berkeley: Lawrence Berkeley National Laboratory,
2002.
52. Yeo, Ye, and Li Wang. "Energy analysis on VAV system with different air-side
economizers in China." Energy and Buildings, 42 (8), 2010: 1220-1230.

Abstract (if available)

Abstract California has set a zero net-energy conservation goal for the residential sector that is to be achieved by 2020 (California Energy Commission 2011). To reduce energy consumption in the building sector, modern buildings should fundamentally incorporate sustainable performance standards, involving renewable systems, climate-specific strategies, and consideration of a variety of users. Building occupants must operate in concert with the buildings they inhabit in order to maximize the potential of the building, its systems, and their own comfort. In climates with significant diurnal temperature swings, environmental controls designed to capitalize on this should be considered to reduce cooling-related loads. One specific strategy is the air-side economizer, which uses daily outdoor temperature swings to reduce indoor temperature swings. Traditionally a similar effect could be achieved by using thermal mass to buffer indoor temperature swings through thermal lag. ❧ Economizers reduce the amount of thermal mass typically required by naturally ventilated buildings. Fans are used to force cool nighttime air deep into the building, allowing lower mass buildings to take advantage of nighttime cooling. Economizers connect to a thermostat, and when the outdoor temperature dips below a programmed set-point the economizer draws cool air from outside, flushing out the warmed interior air. This type of system can be simulated with reasonable accuracy by energy modeling programs

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Enhancing thermal comfort: air temperature control based on human facial skin temperature

PDF

Enhancing thermal comfort: data-driven approach to control air temperature based on facial skin temperature

PDF

Natural ventilation in tall buildings: development of design guidelines based on climate and building height

PDF

Pre-cast concrete envelopes in hot-humid climates: examining envelopes to reduce cooling load and electrical consumption

PDF

Developing environmental controls using a data-driven approach for enhancing environmental comfort and energy performance

PDF

A parametric study of the thermal performance of green roofs in different climates through energy modeling

PDF

Mitigating thermal bridging in ventilated rainscreen envelope construction: Methods to reduce thermal transfer in net-zero envelope optimization

PDF

Double skin façade in hot arid climates: computer simulations to find optimized energy and thermal performance of double skin façades

PDF

Environmental adaptive design: building performance analysis considering change

PDF

Improving thermal comfort in residential spaces in the wet tropical climate zones of India using passive cooling techniques: a study using computational design methods

PDF

Digital tree simulation for residential building energy savings: shading and evapotranspiration

PDF

Visualizing thermal data in a building information model

PDF

Exploration for the prediction of thermal comfort & sensation with application of building HVAC automation

PDF

Impact of occupants in building performance: extracting information from building data

PDF

Streamlining sustainable design in building information modeling: BIM-based PV design and analysis tools

PDF

A high-performance SuperWall: designed for a small residence at Joshua Tree National Park

PDF

Developing a data-driven model of overall thermal sensation based on the use of human physiological information in a built environment

PDF

Energy use intensity estimation method based on building façade features by using regression models

PDF

Energy performance of different building forms: HEED simulations of equivalent massing models in diverse building surface aspect ratios and locations in the US

PDF

Impacts of indoor environmental quality on occupants environmental comfort: a post occupancy evaluation study

Asset Metadata

Creator Lanning, Tighe Glennon (author)

Core Title Economizer performance and verification: effect of human behavior on economizer efficacy and thermal comfort in southern California

School School of Architecture

Degree Master of Building Science

Degree Program Building Science

Publication Date 08/01/2013

Defense Date 08/01/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag economizers,high mass,Human behavior,OAI-PMH Harvest,passive design,thermostats,user interface

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Kensek, Karen M. (committee chair), Choi, Joon-Ho (committee member), Kohut, Timothy (committee member)

Creator Email tighelanning@gmail.com,tlanning@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-312148

Unique identifier UC11294447

Identifier etd-LanningTig-1928.pdf (filename),usctheses-c3-312148 (legacy record id)

Legacy Identifier etd-LanningTig-1928.pdf

Dmrecord 312148

Document Type Thesis

Format application/pdf (imt)

Rights Lanning, Tighe Glennon

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA