University of Southern California Dissertations and Theses

Solar thermal cooling and heating: a year-round thermal comfort strategy using a hybrid solar absorption chiller and hydronic heating scheme

(USC Thesis Other)

Solar thermal cooling and heating: a year-round thermal comfort strategy using a hybrid solar absorption chiller and hydronic heating scheme

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content SOLAR THERMAL COOLING AND HEATING:

A YEAR-ROUND THERMAL COMFORT STRATEGY USING A HYBRID SOLAR
ABSORPTION CHILLER AND HYDRONIC HEATING SCHEME

by

Jason P. Kirchhoff

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 2010

Copyright 2010 Jason P. Kirchhoff

ii

Acknowledgements

I offer a sincere thank you to my thesis committee, without whom this work
would not have been possible. Professors Karen Kensek, Ed Woll, and Marc Schiler, all
of the USC School of Architecture, were instrumental in reviewing the many versions of
this work and guiding it to completion.
An additional thank you goes to Peter Simmonds of IBE Consulting Engineers.
Peter graciously provided access to design documents for the Audubon Center at Debs
Park. These documents were vital to performing this study. Without Peter’s contribution,
this work could not have gotten off the ground.
A final thank you goes to Jeff Chapman of the Audubon Center at Debs Park. Jeff
very graciously allowed full access to the energy management system at the Audubon
Center. These data were the absolute core of this study, and without it there would have
been no study at all. All too often, owners of highly touted buildings are reluctant to
release performance data, fearful that their building might not be performing to the level
that is claimed. Jeff and the Audubon Society are true progressives, not only comfortable
allowing their building to be evaluated, but actively involved in the study and eager for
the results. This attitude should be a model for all green building owners.

iii

Table of Contents

Acknowledgements ............................................................................................................. ii

List of Tables .......................................................................................................................v

List of Figures .................................................................................................................... vi

Abstract ............................................................................................................................... ix

Chapter One: The Thermal Comfort Challenge...................................................................1

Chapter Two: Solar Thermal Cooling and Heating ...........................................................10
Solar Thermal Energy ..................................................................................................11
Harvesting the Energy: Solar Thermal Collectors .......................................................16
Thermal Comfort .........................................................................................................26
Hydronic Solar Thermal Heating .................................................................................28
Solar Thermal Cooling .................................................................................................31
History and Development ............................................................................................39

Chapter Three: The Audubon Center at Debs Park ...........................................................42

Chapter Four: Data Acquisition .........................................................................................60
Data Parameters ...........................................................................................................61
Data Collection and Organization ................................................................................73

Chapter Five: Performance Analysis .................................................................................76
Climate and Site ...........................................................................................................76
Thermal Comfort .........................................................................................................82
Adaptive Comfort ........................................................................................................90
System Components.....................................................................................................92

Chapter Six: Economic Analysis .....................................................................................117
Economics ..................................................................................................................117
Environment ...............................................................................................................123
Architectural Implications .........................................................................................124
Efficiency ...................................................................................................................125

Chapter Seven: Conclusions and Design Guidelines .......................................................127
A Note on Climate .....................................................................................................127
System Performance ..................................................................................................130
Adaptive Comfort ......................................................................................................132

iv

System Components...................................................................................................133
Economic Conclusions...............................................................................................139

Chapter Eight:Future Work..............................................................................................144

Bibliography: ...................................................................................................................150

Appendices
Appendix A: Floor Plans ............................................................................................152
Appendix B: Building Loads ......................................................................................153

v

List of Tables

TABLE 5-1: Heating and cooling degree days in California climate zone 9 ....................77

TABLE 5-2: Load calculations for the Audubon Center at Debs Park .............................81

TABLE 5-3: Cooling season thermal comfort statistics ....................................................85

TABLE 5-4: Heating season thermal comfort statistics ....................................................88

TABLE 5-5: Summary of data for collector efficiency, method 1 ....................................97

TABLE 5-6: Summary of data for collector efficiency, method 2 ....................................98

TABLE 5-7: Summary of data for collector efficiency, all methods ..............................100

TABLE 6-1: Motor and pump specifications ..................................................................119

TABLE 7-1: Payback calculation for solar thermal system ............................................141

TABLE B-1: Load breakdown for the library/conference room .....................................153

TABLE B-2: Load breakdown for the offices .................................................................154

TABLE B-3: Load breakdown for the Discovery room ..................................................155

vi

List of Figures

Figure 1.1 – 2008 United States energy use by sector .........................................................2

Figure 1.2 – Building energy use breakdown ......................................................................3

Figure 1.3 – Yazaki absorption chillers ...............................................................................6

Figure 1.4 – The Audubon Center at Debs Park ..................................................................7

Figure 1.5 – Renewable energy sources ..............................................................................8

Figure 2.1 – Pyranometer ...................................................................................................13

Figure 2.2 – World map of peak sun..................................................................................15

Figure 2.3 – Solar radiation spectrum ................................................................................17

Figure 2.4 – Flat plate collector .........................................................................................19

Figure 2.5 – Parabolic trough concetrating solar thermal collectors .................................21

Figure 2.6 – Heat pipe evacuated tube collector schematic ...............................................24

Figure 2.7 – Evacuated tube collector with a cylidrical absorber ......................................25

Figure 2.8 – Evacuated tube collector unit ........................................................................26

Figure 2.9 – ASHRAE Standard-55 thermal comfort diagram..........................................28

Figure 2.10 – Solar thermal hydronic heating system ......................................................31

Figure 2.11 – Vapor compression refrigeration .................................................................34

Figure 2.12 – Single stage absorption chiller schematic....................................................37

Figure 2.13 – Yazaki WFC-10 10-ton chiller performance curve .....................................38

Figure 3.1 – Entrance to the Audubon Center ...................................................................42

Figure 3.2 – South and west facing roof ............................................................................46

Figure 3.3 – Detail of individual evacuated tube collector ................................................47

vii

Figure 3.4 – Header pipe and exchange manifold schematic ............................................48

Figure 3.5 – Solar thermal collector with inlet/outlet detail ..............................................49

Figure 3.6 – Solar thermal collector layout .......................................................................50

Figure 3.7 – Building control and monitoring schematic ..................................................50

Figure 3.8 – Hot water storage tank ...................................................................................51

Figure 3.9 – Heating mode water flows .............................................................................53

Figure 3.10 – Mechanical equipment.................................................................................55

Figure 3.11 – Pipes and conduit .........................................................................................56

Figure 3.12 – Cooling mode water flows...........................................................................58

Figure 3.13 – Valve V2 ......................................................................................................59

Figure 4.1 – In-line sensors ................................................................................................65

Figure 4.2 – Storage tank temperature sensor ....................................................................68

Figure 4.3 – Valve V1 ........................................................................................................71

Figure 4.4 – Master datasheet sample image .....................................................................75

Figure 5.1 – Heating and cooling degree days in Burbank, 2008 ......................................78

Figure 5.2 – Heating and cooling degree days in Riyadh, 2008 ........................................79

Figure 5.3 – Psychrometric chart for California climate zone 9 ........................................80

Figure 5.4 – Interior comfort filtered datasheet screenshot ...............................................83

Figure 5.5 – Cooling season interior temperatures ............................................................85

Figure 5.6 – Heating season interior temperatures ............................................................89

Figure 5.7 – Adaptive comfort ...........................................................................................91

Figure 5.8 – Temperature rise in solar collectors, cooling season .....................................94

viii

Figure 5.9 – Temperature rise in solar collectors, heating season .....................................94

Figure 5.10 – Performance curves for Yazaki WFC-10 absorption chiller .....................102

Figure 5.11 – Overheating condition ...............................................................................105

Figure 5.12 – Insolation and storage tank temperature on September 24 ........................108

Figure 5.13 – Insolation and storage tank temperature on October 15 ............................109

Figure 5.14 – Effect of overcast weather on storage tank temperature ...........................111

Figure 5.15 – Storage tank temperature, rainy winter days .............................................112

Figure 5.16 – Office air temperature versus building supply temperature ......................113

Figure 5.17 – Delivered cooling ......................................................................................115

Figure A.1 – Audubon Center floor plan with HVAC zones .........................................152

ix

Abstract

With over 40% of total energy use in the United States coming from buildings, it
is clear that future conservation strategies must dedicate a significant focus towards the
built environment. Space heating and cooling represent two of the three largest building
energy loads and thus provide an excellent springboard for alternative technologies that
use renewable energy as the primary energy source. One such system uses solar thermal
energy as the primary driver for thermal comfort. Heat energy from the sun is absorbed
and stored in water. In the winter months, the hot water is pumped directly into the
building to provide space heating. In the summer, the hot water runs an absorption chiller
to provide cooling to the space.
The Audubon Center at Debs Park in Los Angeles, CA uses just such a system.
The off-grid, LEED platinum building is designed to first utilize passive strategies to
minimize building loads. The remaining peaks are then met with the solar thermally
driven system. The building uses no backup energy sources, relying on its unique system
for year-round thermal comfort. An extensive network of sensors monitors the various
system components in the building, providing the data used for the in depth system
analysis presented in this paper.
The system performed well in the cooling season, meeting the cooling loads in the
fully occupied office spaces for nearly the complete duration of the study. The heating
season was drastically different, with the space too cool for comfort 40% of the time.
This highlights one of the main paradoxes of a solar thermal system. Solar cooling relies
on the synchronization of peak loads and peak energy availability. On the other hand,

x

solar heating peak loads rarely coincide with peak energy availability. Various
explanations of the successes and failures of the system are presented in this thesis, along
with design ideas for future systems based on those conclusions. To do this, a number of
different comparisons between time-matched data sets were performed to quantify the
system performance under a wide variety of conditions.
The final bit of analysis concerns alternative strategies for the building that could
also use renewable energy for thermal comfort. These are compared against the existing
system in terms of expected performance, as well as cost and aesthetics.

1

Chapter One: The Thermal Comfort Challenge

According to the 2007 DOE Building Energy Data Book, buildings currently
account for over 40% of total energy use in the United States (DOE 1). With the world
entering an uncertain energy future, statistics like this cannot be ignored. It makes sense
that much of the building design community is shifting their focus to building energy,
realizing that a truly remarkable building can gain as much acclaim for its performance as
for its architectural expression.
Buildings that respond to nature are nothing new. In fact, the zero-energy building
was the norm for thousands of years until technological advances gave architects and
engineers the ability to manufacture indoor environments. The age of cheap energy, fed
though the rapacious use of fossil fuels, created an architectural revolution of sorts. There
was no need for a building to harmonize with nature, to ebb and flow with its available
energy. This wasn’t necessarily a bad thing; it led to some of the most creative,
astonishing structures ever built. Today however, these structures often serve to remind
us how far we strayed from the idea of building performance. Fossil fuels are now
understood to be far from limitless. In fact, the argument against burning fossil fuels has
two fronts. Many push for energy reform in the name of the environment, backing their
argument with data that points to changing climates and dangerous levels of greenhouse
gas concentrations. Others may argue that a shift is necessary to guarantee a supply of
energy for future generations, as uncertain but alarming data regarding fossil fuel
reserves and peak oil scenarios continue to appear. In the built environment,

2

simultaneous efforts are underway to both reduce the amount of energy buildings use and
harness the energy that is used from renewable sources. There is perhaps no better
resource for both these endeavors than the sun. The most reliable of all the renewable
resources has been continuously showering the planet in free energy for over four billion
years. The energy conscious building designer sees the sun as neither friend nor enemy,
but rather a force to guide the ultimate design of the building.

Figure 1.1: 2008 energy use by sector. The built environment is the combination of commercial and
residential use. Source: 2008 DOE Energy Information Administration (http://www.eia.doe.gov/)

3

Figure 1.2: Building energy use break down. Space heating and space cooling can be nearly eliminated
with a solar thermal combined system. Source: United States Department of Energy 2008 Buildings Energy
Data Book

A breakdown of building energy use shows that space heating and space cooling
account for two of the three largest building loads, combining to represent 42% of the
annual energy consumed. Space cooling is generally driven by electrical chillers, while
space heating generally requires the burning of a fossil fuel (though it too is sometimes
electrically driven). These two energy users are what create a comfortable environment in
our buildings. Even in climates that are ideal for naturally conditioned buildings, these
systems are usually installed as a guarantee of occupant comfort. For this reason, it’s
often not practical to talk of eliminating these systems in any case beyond a private
residence with a particularly motivated owner. Rather, we should investigate new ways of
delivering thermal comfort, especially those that utilize renewables as the main energy
source. Again the sun, the driver of environmentally sensitive architecture, is the top
candidate. Space heating makes the most logical sense. Even on cold days, the sun
provides free energy in the form of light. Converting this light to heat is simple in theory;
any dark surface will absorb the light and heat up. Building on this principle, we can

4

transfer the heat to a storage medium and tap it whenever necessary. A common storage
choice is water. A glance at a table of specific heat values will show that water has a very
high thermal storage capacity relative to other commonly available materials. Capturing
solar radiation in water, when combined with a well insulated storage tank, has the
potential to meet both the space heating and domestic hot water demands for many
buildings. Less obvious is how to use the sun for cooling. An emerging technology
known as absorption cooling is capable of doing this. The absorption process ultimately
relies on the same thermodynamic principles as the traditional chiller. However, different
choices of refrigerants plus a rearrangement of the cycle processes creates a system that is
driven by heat, rather than by an electrical compressor. The core technology is not new –
for many years absorption chillers were installed in buildings with large supplies of waste
heat, often from industrial processes. Propane fired refrigerators are another example of
the absorption process using heat to deliver space cooling. The emerging technology
differs in that operates on significantly lower temperatures, in the range of those
achievable from standard solar thermal collectors. Newer, smaller absorption chillers
capable of operation at water temperatures under 100° C are being developed by a small
number of companies spread throughout the world, meant to meet the space cooling
demands of a building via an infinitely renewable and an easily storable energy source.
The common thread for the systems mentioned above is the use of solar thermally
heated water as the energy source. For space heating, the hot water is pumped into the
building and passed through fan-coil units or radiators, similar to a boiler-driven hydronic
heating scheme that is seen in many buildings. In absorption chiller space cooling, the hot

5

water is an input into the chiller in a closed loop. The output is a separate loop of chilled
water. This chilled water is fed into the building and run through fan-coil units similar to
any chilled water setup. In many buildings, the systems can be combined to meet the
year-round thermal comfort demands of the occupants. This elegant solution relies on
infinitely renewable solar thermal energy as its main energy source. A combined system
(sometimes shortened to “combisystem”) such as this can drastically reduce the use of
electrical and fossil fuel energy sources without sacrificing thermal comfort. There exists
tremendous potential for this technology, and research is ongoing throughout the world to
evaluate its benefits.
The architect and engineer who decide on a solar thermal comfort strategy face a
number of challenges. Building loads depend largely on construction, occupancy
schedules, and of course, climate. Because the system demands a relatively high hot
water temperature, the challenge arises of exactly how to produce and store the hot water
necessary to guarantee thermal comfort in any climatic condition the building might
experience. Evacuated tube solar thermal collectors must be in place for harnessing the
sun’s energy and directing it to the storage medium. The orientation and sizing of these
collectors is vital - and comes with significant architectural implications. The collector
array must be sized in concert with the size of the storage tank, which itself must be sized
based on expected climatic conditions. In the United States, few HVAC engineers are
even familiar with systems of this type, let alone experienced enough to fully design one.

6

Figure 1.3: Three Yazaki absorption chillers. Source: www.yazakienergy.com

This paper explores a case study building, along with existing research, to develop
a design guide for an interior comfort scheme that uses solar driven absorption cooling
and hydronic heating. The case study is the Audubon Center at Debs Park in Los
Angeles, California. This award winning, LEED platinum rated building has over a half
mile of solar thermal collectors on its roof fueling its combisystem. A single pipe loop
enters the building from the HVAC shed. In the cooling season, this pipe is fed with
chilled water from a 10-ton Yazaki single-stage solar absorption chiller. In heating
season, the loop is fed with hot water directly from the 1200 gallon water storage tank. A
network of valves controls flows and an extensive monitoring network provides system
data. In this paper, that system data are compared against on-site climate data to provide
insight into the successes and failures of the system, ultimately using this information as
the basis for the design guide. It is a goal of this paper to avoid the use of advanced

7

computer simulations of systems, instead relying on empirical site data in concert with
basic calculations for heat flow and thermal comfort. Ideally this approach will be more
appropriate for early stage development and design of future systems.

Figure 1.4: The Audubon Center at Debs Park in Los Angeles, CA uses a solar thermal interior comfort
strategy. The evacuated tube solar thermal collectors are visible on the roof. Source: United States Green
Building Council LEED case study (http://leedcasestudies.usgbc.org/overview.cfm?ProjectID=234).

8

Figure 1.5: Renewable energy represents only 7% of energy generation in the United States. A
further breakdown shows that only 1% of renewables are solar energy. Source: 2008 DOE Energy
Information Administration (http://www.eia.doe.gov/)

A year-round solar thermal interior comfort strategy should be considered for
buildings with climate conditions meeting appropriate design parameters. As with all new
technology, the systems remain costly. However, by utilizing renewable energy, the cost
of the system essentially represents a prepayment for energy. It seems that the only sure
thing regarding energy is that the cost for the end user will continue to increase. Because
of this, the amount of time required to earn back the investment in renewables continues

9

to fall. As the push for high-performance buildings gains momentum, it’s important for
designers to be aware of low-energy alternatives for building design. A year-round
thermal comfort strategy using a hybrid solar absorption chiller and hydronic heating
scheme is one such system. Its early integration into the design of buildings could prove
to be a tremendous step towards reducing the carbon footprint of the built environment
and ensuring a safe and stable energy future.

10

Chapter Two: Solar Thermal Cooling and Heating

This chapter provides the background information needed for the basic
understanding of a year-round thermal comfort strategy that relies on solar thermal
energy as its main energy source. Perhaps the most fundamental step is a thorough grasp
of exactly what solar thermal energy is. It remains the ultimate driver of a thermal
comfort system of this type, and as such it is a perfect springboard for this chapter.
Following the explanation of solar thermal energy is a description of the various types of
thermal collectors in use throughout the world for solar thermal applications, including
evacuated tube solar thermal collectors, which are the type used at the Audubon Center at
Debs Park. Collectors of this type have proven to be a very effective choice for HVAC
applications, and are thus described in detail in this chapter.
It is relatively easy to understand how solar thermal energy can be put to use for
space heating, as anyone who has ever stepped into a hot car can attest. A radiant hot
water system such as the one at the Audubon Center is a more complex system, but
operates on the same heat transfer principles that make that car interior so unbearable.
Such a system is described next in this chapter.
Using solar thermal energy to produce space cooling is a far more complex
proposition, and is understandably an idea that seems counterintuitive. For this study, the
device used to produce the cooling is a solar absorption chiller. Using a unique, multi-
step application of thermodynamic principles, the machine is fed with hot water (roughly
180° F) and outputs chilled water (roughly 50° F) suitable for space cooling. It uses a

11

remarkably small number of moving parts and needs only small amounts of electricity to
power the pumps necessary in the system. Though absorption chiller technology is not
new, systems capable of operating at relatively low solar hot water temperatures
(compared to industrial waste heat, often in excess of 250F) are a recent innovation. The
basic principles of absorption chillers are the next part of this chapter. A detailed
description of the system in place at the Audubon Center at Debs Park is presented in
Chapter 3.
Hybrid systems like the one at the Audubon Center are rare but do exist in some
diverse locations, with a noticeably higher concentration in Europe. The end of this
chapter will explore some case studies throughout the world and summarize some of the
existing technical research in the area of solar thermal heating and cooling.

Solar Thermal Energy
With the exception of atomic and geothermal energy, the sun is often viewed as
the only energy source on Earth. Other types of energy are just byproducts of the sun’s
sweeping effect on the planet – wind power is driven by thermal swings in the air that
result from the sun’s heat; hydro power is a result of the solar-driven water cycle; even
fossil fuels are collections of organic matter that millions of years ago thrived on
sunlight. As fossil fuel resources dwindle, and their environmental impacts mount, direct
use of solar energy stands out as a major player in the world’s energy future. It is the

12

perfect resource, well understood and infinitely renewable. In fact, in one hour, the planet
absorbs enough solar energy to meet human energy needs for that year (Lewis 2006). Of
course we can’t collect and convert all of that energy, but it is clear how harnessing even
a minute portion is extremely valuable for world’s energy future.
Solar energy available on the earth’s surface is measured as solar insolation. The
word is a blend of incident solar radiation (USDOE 2), and therefore does not appear in
other, non-solar types of energy calculations. The redundancy “solar insolation” is
commonly used and will appear interchangeably with “insolation” in this text and other
works in the area. By definition, insolation is the amount of solar energy incident on a
given surface in a certain time. The most common unit is Watts per square meter (W/m
2
),
though there are other ways to express the quantity. With the Watts per square meter unit,
it appears there is no “per time” term, but remember that a Watt is defined as one Joule
per second, so the unit could be expressed as
𝐽 𝑠 ∙ 𝑚 2

which is the energy (J) per area (m
2
) per time (s) as described above. An alternate
measure is Btu/hr·ft
2
. For the remainder of this text, Watts per square meter will be used.
Since insolation is measured on a given surface, it will be maximized when the
surface is perpendicular to the sun’s rays. For example, consider a one square meter plate
flat on the surface of the earth, with the sun directly overhead. On a clear day, this plate
will receive a peak of roughly 1000 W/ m
2
of total insolation, a standard amount known
as peak sun. If the same plate is rotated 90 degrees so that is perpendicular to the ground

13

(and parallel to the sun’s rays), it will receive zero W/m
2
of direct insolation (though it
will still pick up significant diffuse radiation from the half-sky). For angles in between,
the amount of insolation varies with the cosine of the angle. Insolation values are
commonly given for an imaginary surface assumed to be perfectly facing the sun at the
point of measurement. This is known as direct radiation. A device that measures direct
radiation is assumed to point at and follow the sun throughout the sun’s track across the
daytime sky. The other component of insolation is diffuse radiation. This is energy that
has been scattered by the atmosphere and thus arrives at the sensor from anywhere in the
daylit hemisphere. The combination of direct and diffuse radiation is known as global
horizontal radiation, which is usually the value reported by less sophisticated
measurement devices with a single output. A visual representation is shown in Figure 2.1.
The device used to measure insolation is known as a pyranometer.

Figure 2.1: Sophisticated pyranometers (and pyroheliometers) report insolation as individual components
like those shown above. The sensor at the Audubon reports global horizontal irradiance, like the leftmost
image. Source: http://www.seco.cpa.state.tx.us/publications/renewenergy/images/exhibit3-02.jpg

14

At the top edge of the Earth’s atmosphere, the average direct normal solar
insolation is 1367 W/ m
2
, known as the solar constant. This value represents the incident
energy on an ideal absorber plate that remains perpendicular to the sun’s rays at all times.
It is an average, accounting for the slight variation in incident energy due to the distance
between the earth and the sun and other solar phenomena such as sunspot activity (NASA
1). The insolation at the surface of the earth varies greatly, depending on atmospheric and
weather conditions. Peak sun is used as an approximation of the conditions on a clear day
with the collector pointed directly at the sun. For many solar energy applications, the
design parameter used is average hours of peak sun, which allows for a rough assumption
of available solar energy without using highly sensitive insolation values. In Figure 2.2, a
map of the world is shown with color-coded zones representing average hours of peak
sun per day. Not surprisingly, desert regions experience considerably more peak sun
hours than northern latitudes. Maps like this are valuable for determining the size of a
proposed solar energy system.

15

Figure 2.2: Worldwide average hours of peak sun. Source: SunWize Technologies

The energy emitted from the sun represents a range of frequencies, with
wavelengths in the infrared, visible, and ultraviolet part of the electromagnetic spectrum.
Ultraviolet rays are largely filtered out by atmospheric ozone, leaving only visible light
and infrared radiation available for solar thermal energy. Visible light is the part of the
spectrum we can see; it is what lights up objects and gives them color. Infrared radiation
is heat, most of what you feel when you stand in direct sunlight. Much of the infrared
radiation from the sun is absorbed in the lower atmosphere, largely by water vapor. Still,
much of it moves on to the Earth’s surface. Solar thermal energy systems collect both
types of radiation, but are designed to respond more to visible light than to infrared
radiation.

16

Harvesting the Energy: Solar Thermal Collectors
A solar thermal collector is a device that converts sunlight into heat. The simplest
example might be large pot of water that has been painted black. The pot will absorb
energy in the form of sunlight and heat up as a result. The water in the pot will absorb
this heat and, at the end of the day, will be significantly warmer than it was in the
morning. The pot can also absorb infrared radiation from the sun, though the long
wavelength portion is at a much lower energy than visible light, and suffers absorption
from atmospheric gases. The color of the pot is very important. Color is an indication of
reflected light in the visible spectrum. For example, green leaves are reflecting the green
part of the visible spectrum while absorbing the other colors. A clear surface, like glass,
is allowing the entire visible spectrum through
1

1
While all the colors of the spectrum pass through the glass (making it clear), not all of the energy passes
through. In clear float glass, 83.4% of the visible light energy is transmitted through the glass. The
remainder is reflected or absorbed by the material.
. It reflects no light back into your eyes,
making it “clear.” A white surface reflects all colors of light, the colors mixing in your
eye to be interpreted as “white.” A black surface absorbs the entire spectrum, reflecting
no light back into your eyes. If you were to look at a theoretically perfect black surface,
you would see literally nothing at all, as no light is reflected to your eyes. You would
know where the surface is because you could deduce from the black “hole” in the image.
From this, one can see why a dark surface is preferred for solar energy absorption – it
absorbs all incident energy without reflecting any back.

17

Figure 2.3: The solar radiation spectrum. The effect of the atmosphere is seen in the reduction in solar
irradiance between the yellow and red sections. Notice how visible light represents the most continuous and
energetic part of the solar spectrum. This is why solar thermal collectors are designed for this type of
radiation. The purple lettering identifies airborne substances that absorb certain radiation bands. The effect
of water vapor (H
2
O) absorption on infrared radiation is evident.
(http://en.wikipedia.org/wiki/File:Solar_Spectrum.png)

Today’s solar hot water collectors are basically technologically advanced versions
of the black pot. For thermal comfort hot water applications, there are three basic types of
collectors:

• Flat plate collectors are the simplest types, consisting of a network of fluid pipes
in contact with an absorbent background, typically a dark colored metal plate. The
absorbent plate is insulated on the back side to prevent heat loss to the

18

environment. The pipe and absorber system is enclosed in a thin box with the sun-
facing side made of glass. The glass top surface is important for a few reasons. As
mentioned earlier, glass lets nearly all incident visible and shorter wavelength
light through. It can be treated with a special anti-reflection coating to get this
transmission even higher, with transmission coefficients over 90% commonly
available. The dark absorber plate absorbs this solar energy and begins to transfer
it to the embedded pipes and the flowing thermal fluid (usually water or glycol-
water solution). The glass sits a couple of inches above the absorber plate, with an
air-filled cavity space between the two surfaces. This cavity is important for
keeping the heat contained in the system. The glass and air help to minimize
conductive losses, much like in a double glazed window. The glass also traps in
the air space heat that would generally move away via convection into
surrounding environment. The final effect of the glass is to minimize radiation
losses. Based on the laws of blackbody radiation, the absorber plate will reradiate
much of the energy it has absorbed. The temperature of the plate means that most
of this radiation will be long-wave infrared energy. Glass happens to be highly
opaque to energy in this range. The energy radiating up from the plate is reflected
by the glass back downward, where it is reabsorbed and can again potentially
conduct in the heat transfer fluid. This property of transmitting visible light but
reflecting longwave infrared radiation is routinely used in greenhouses to keep the
space considerably warmer than the surroundings, with the phenomenon known
as the “greenhouse effect.” Collectors of this type are usually fixed and installed

19

at an angle determined to be optimal for solar thermal collection. Temperatures
from panels like this may not get high enough for cooling, so these systems are
often used for domestic hot water.

Figure 2.4: Schematic of a flat plate collector. Source: Buderus (http://www.buderussolar.com)

• Evacuated tube collectors work on a similar principle as flat plate
collectors, but are more efficient devices that produce higher water
temperatures because there is no air to conduct or convect heat away from
the absorber. They have proven to be the best device for the type of system
described in this paper and thus are described in further detail in the next
section.

20

• Parabolic trough collectors work by using a curved, mirror-like surface to
focus light on an absorber. The focused light creates very high
temperatures in the absorber, and this heat is rapidly moved into heat
transfer fluid. These systems provide the highest efficiency but are the
most difficult and costly to install. The geometry of a parabola ensures that
all incident radiation parallel to the parabola axis will be reflected to the
focal point. Conversely, if the parabola does not line up exactly with the
sun, all of the heat completely misses the absorber and the device becomes
worthless. Because of this, parabolic trough collectors must follow the sun
throughout the day, pointing their imaginary axis right at the sun to guide
the reflected radiation to the collector. This adds to the complexity and
cost of the system.

21

Figure 2.5: Parabolic trough collectors. These collectors heat a glycol solution to provide domestic hot
water to a jail. Source: NREL (http://www.nrel.gov/)

Evacuated Tube Collectors

Solar thermally driven chillers or heat pumps require a relatively narrow
temperature range to operate properly. Flat plate collectors deliver the least heat of the
three types of units, so very large areas are needed to produce the 160F+ temperatures
required to operate a single effect chiller. There is often insufficient roof area for such an
installation. Parabolic trough collectors deliver concentrated energy and can heat water to
very high temperatures well above the boiling point of pure water. This can allow for
high efficiency double-effect systems, but there are serious drawbacks. Because of the
geometry of the collectors, they rely on strong, direct sunlight. Because of this, overcast

22

conditions reduce their performance drastically. In addition, the cost and complexity of
the troughs and trackers can be prohibitive. Evacuated tube collectors achieve the right
temperatures and are available at reasonable cost. They are usually installed in a fixed
position, eliminating the need for solar tracking mechanisms like in a parabolic trough
system.
An evacuated tube collector system is made up of a series of parallel heat tubes,
each connected at the top to a header pipe or manifold. Each heat tube consists of a metal
heat pipe connected to a black metal absorber plate. Both components are usually made
of copper, as it is a very conductive metal. The heat pipe/absorber setup is surrounded by
a glass tube, with the space between the tube and the absorber evacuated. As in the flat
plate collector, the glass and air help keep heat in the tube. The evacuated space prevents
convective heat loss out of the system; the thermal properties are similar to a double
glazed window. However, in a window the space between the glazing layers cannot be
evacuated without the glass layers being sucked toward each other, deforming the
window shape. With the cylinder shaped glass in the collectors, the force is equally
distributed and no deformation occurs.
The black absorber is coated with a special material designed to maximize
absorption of light. Sunda, a Chinese company that produced the tubes used in the
Audubon Center, claims their absorber absorbs 92% of the energy incident on it. As
mentioned above, the glass and vacuum structure is vital to keep this heat in the system.
Transfer of the heat into the header tube is done through an interesting thermodynamic
process. Inside of the metal heat pipe is a small amount of fluid (either water or a

23

specialized thermodynamic solution) sealed in at below atmospheric pressure. As the
fluid heats up in the reduced pressure environment, it vaporizes. The tubes are angled at a
minimum of 15 degrees above the horizontal to allow the natural convection shown in
Figure 2.6 (Sunda 1). The hot gas rises up the heat pipe by convection. At the top, the
heat pipe is joined to a manifold containing a steady flow of liquid water. At this
intersection, the heat pipe has an integrated condenser space, allowing the gaseous fluid
to condense and transfer its latent heat to the water flow in the manifold. This is done
indirectly via a heat exchange process; the fluid in the heat pipe and the water in the
manifold never actually mix. The now condensed fluid falls back down the heat pipe due
to gravity, and the cycle repeats. Because of this setup, a broken or malfunctioning tube
can be removed from the system without having to stop the water flow or drain the
system. The water in the manifold continues to heat up as it sweeps by the remaining
evacuated collectors. Evacuated tube hot water systems are capable of heating water to
over 190° F.

24

Figure 2.6: Schematic of an evacuated tube collector heat tube. Notice that the tube must be angled
to allow for natural convection of the hot vapor. Source: www.sunmaxxsolar.com

In some evacuated tubes, the circular cross section of the collector tubes means
that the tube is directly facing the sun for a long period (Figure 2.7), which represents an
improvement over fixed flat plate collectors. However, finned tubes like those at the
Audubon reduce this “tracking” effect. In all cases, the space between adjacent tubes is
not able to absorb heat, lowering the efficiency of the overall collector area. One solution
is to install specially designed curved reflectors below the tubes to redirect stray light
back into the collectors. The tubes work best in cold climates, where there is no danger of
overheating the heat transfer fluid. In some tubes, a special heat sensitive metallic strip
can halt the evaporation/condensation process to prevent overheating.

25

Figure 2.7: In some evacuated tubes, a cylindrical absorbing layer means that evacuated tube collectors are
facing the sun for a longer period of time than a flat plate. A reflector below the tube can redirect energy
that passed between two adjacent tubes. Source: Regulus (www.regulus.eu)

In nearly all solar thermal installations, the hot water is stored in an insulated
storage tank. This allows for continued use of the thermal energy even after the sun has
set. Heat energy, unlike light energy, can be directly stored. This represents one of the
main differences between solar thermal and solar photovoltaic technology. In a
photovoltaic energy system, light energy is directly converted to electricity by use of a
special panel. This electricity must be used either immediately or stored in an array of

26

batteries. In solar thermal applications, the heat can be stored in a number of materials,
including concrete, earth, and water. This is particularly useful for the heating season,
where the coldest temperatures usually occur at times of minimal solar availability
(overcast or after sunset).

Figure 2.8: An evacuated tube solar thermal collector. Source: www.greenterrafirma.com

Thermal Comfort
Defining what exactly constitutes thermal comfort is a challenge. Conditions that
are perfectly comfortable for one person maybe entirely uncomfortable for another. The
most common guideline for the thermal designer is ASHRAE Standard 55 – Thermal
Environmental Conditions for Human Occupancy. The American Society of Heating,
Refrigeration, and Air-Conditioning Engineers (ASHRAE) publishes a variety of

27

guidelines that are helpful standards in the architecture and engineering profession.
Standard 55 breaks down thermal comfort into six factors: air temperature, relative
humidity, mean radiant temperature, activity level, air movement, and clothing level. An
acceptable level is determined from a predicted mean vote (PMV), which estimates the
number of people who would consider the condition “comfortable.”
In all cases, the main comfort factors are air temperature, humidity, and clothing
level. An assumption for thermal comfort is that the occupant is wearing a certain amount
of insulative clothing, measured using the clo scale
2

2
The scale was invented by interior comfort expert and ASHRAE hall of fame member P. Ole Fanger.
. Winter conditions allow for an
assumption of 1.0 clo, which is equivalent to a business suit or sweater and trousers
(Engineering 1). For space heating, at 1 clo, the acceptable range is from roughly 68 to 78
degrees, depending on relative humidity. For cooling, the assumption is 0.5 clo,
equivalent to a light blouse or shirt and light trousers. The comfortable temperature range
is roughly 76 to 81 degrees, depending on relative humidity. Figure 2.9 provides a
graphical representation of the thermal “comfort zone” based on air temperature,
humidity, and clo.

28

Figure 2.9: ASHRAE-55 thermal comfort zone based on air temperature, relative humidity, and clothing
insulation

Hydronic Solar Thermal Heating
A hydronic heating system is one that uses water as the heat transfer medium.
Traditional systems employ a boiler, usually gas or oil fired, to heat large quantities of
water. The hot water is then pumped into radiators in the heating zones. The earliest
systems used steam to transfer the heat, so no pump was required and only one pipe was
necessary. Hot steam would rise from the basement boiler to the radiators. Steam that had
given off its heat would condense into liquid and return via gravity to the boiler, passing
the rising steam on the way. The inability to control individual radiators led to the

29

development of systems with two pipes - one for hot supply water and one for return
water. Separate supply and return circuits allowed for individual control of any radiator
with the same temperature water delivered to each.
A solar thermal hot water system operates on essentially the same principles as
traditional hydronic systems, except that the boiler is replaced with an array of solar
thermal collectors and a hot water storage tank. The water storage tank is essential, as the
coldest temperatures, and thus the highest heating demands, usually occur when the sun
has set or is hidden by clouds. The hot water storage tanks for solar thermal systems are
highly insulated. A backup heat system is usually employed for situations where a
prolonged lack of solar thermal energy leaves the system unable to keep up with heating
demand. This backup system may be a traditional gas or oil fired boiler, an electric hot
water heater, or a tankless water heater.
As mentioned earlier, traditional hydronic systems use radiators to transfer heat
into the space. If the heating system piping is also used for cooling in the summer, the
same radiators cannot be used effectively. Heating radiators are usually placed low in a
space. They heat the air around them and this air rises, setting up a convection cell. A
cooling radiator would have to be placed high in a space to set up a similar convective
cell. The same problem occurs when attempting to use a radiant floor or ceiling; heating
is best embedded in a floor while cooling tends to be achieved with radiant ceiling panels.
An additional drawback to these systems is that they only provide sensible heating - there
is no provision for removal or addition of humidity. Also, there is the need for an
independent ventilation system to bring fresh air into the space.

30

To solve these issues, most combined heating and cooling systems rely on forced
convection via fan coil units (FCUs) or air handling units (AHUs). In hydronic systems
of this type, the space is heated or cooled by pumping the appropriate temperature water
through the coil. In heating mode, a fan forces cool space air over the coil, with the air
absorbing the heat of the water through the conductive piping and then blowing into the
space. In cooling mode, warm space air passing over the coil gives up its heat to the
chilled water before being returned to the space. An FCU usually blows air directly into
the space, with an individual FCU for each air supply. The AHU blows the conditioned
air into a series of ducts which distribute it to supply vents a distance away from the
AHU. An advantage of these types of units is their ability to control the moisture content
of the air and integrate ventilation. The Audubon Center uses three air handler units, one
for each HVAC zone in the building.

31

Figure 2.10: Schematic of a solar hydronic heating system with a ducted air handler unit. Source:
www.solarpanelsplus.com.

Solar Thermal Cooling
Using the heat of the sun to ultimately produce cooling can be a difficult idea to
grasp at first. In most applications, a device known as an absorption chiller is used.
While the heating system described earlier can be termed “direct,” solar cooling would be
called indirect. In the absorption cycle, heat drives a system of thermodynamic processes,
one of which, the evaporation of a coolant, ultimately provides the cooling effect. Before

32

understanding the absorption cycle, it is helpful to understand how a traditional
compressor driven chiller works.
A standard air conditioner works by changing the state of a coolant, usually a
synthetic chemical, in a cyclical process. Because of the cyclical nature, one can describe
the system from any starting point. In this instance, let’s consider that the coolant starts as
a cool liquid. This liquid is fed into an expansion valve. The reduction in pressure causes
the coolant to boil, or flash off (the coolant’s ability to boil at this pressure and
temperature are precisely why it was chosen). A core concept of thermodynamics
explains that as matter changes phase, it either releases or gains heat, but does not
increase in temperature. In a liquid to gas transition, the amount of heat is known as the
latent heat of the phase change. In the case of our coolant, the change is from liquid to
gas. This phase change requires heat, equal to the latent heat of the coolant. In many
instances, the latent heat is very large compared with sensible heat. Water uses five times
as much heat in evaporation as it takes to heat the water from freezing to boiling
temperatures. The coolant flashes off in a region of the air conditioner aptly named the
evaporator. The coolant itself is flowing through a network of pipes in the evaporator.
The pipes are surrounded by what needs to be cooled – either water or air. The latent heat
driving the evaporation of the coolant is removed from the coil surroundings. The now
cooler air or water is then moved to the zone that needs cooling.
With the cooling effect having taken place, the air conditioner must now
manipulate the coolant so that the cycle can repeat. Upon leaving the evaporator, the
coolant is a low pressure warm gas. It is moved to a region of the air conditioner known

33

as the compressor. In the compressor, an electric (or otherwise fueled) motor is used to
compress the warm, low pressure vapor into a superheated, high pressure vapor. The hot
vapor moves into a condenser. In the condenser, the hot coolant vapor is exposed through
a heat exchanger to air or water at the outdoor temperature, considerably cooler than the
very hot gas. The coolant gas gives up its heat to the outside air or water, which rejects
the heat to the outside. The reduction in heat of the coolant causes it to condense into a
liquid. Continuing to expose the liquid to outside air or water cools it further, finally
resulting in a cool liquid. The cycle can then repeat. In effect this cycle has absorbed heat
at a low temperature (from the cooled space) and rejected the same heat at a higher
temperature (to the outside world).

34

Figure 2.11: The traditional cooling cycle relies on an electrically driven compressor.

In the absorption cycle, the cooling effect is achieved in a similar way, via the
evaporation of a coolant in the presence of air or water as a heat transfer fluid. How the
coolant is regenerated is what separates the technology. In general, an absorption cycle is
defined as one in which the gaseous coolant is absorbed by a secondary substance known
as the absorbent. The now mixed solution is heated to separate the coolant and absorbent,
and the cycle repeats. A common system in solar fired chillers, and the system in place at
the Audubon center, is a lithium-bromide absorption chiller. Lithium-bromide is the
absorbent; the coolant is water. The cycle consists of four steps described next.
The Lithium-Bromide Absorption Cycle

35

Step 1: Evaporation
Again, consider the starting point to be the coolant as a cool liquid. It must be evaporated
for the cooling effect to take place. In a similar fashion to the compressor cycle, the liquid
water is passed through an expansion valve into the low pressure evaporator. The low
pressure causes the water to flash off. The evaporator coil is part of the building’s
primary chilled water loop. Return water flows into the evaporator having just given off
its “coolth” inside the building. The coolant evaporates in the presence of the coil,
absorbing the heat of the return water and cooling it significantly. The now cool return
water moves back into the building, now labeled supply water, where it will cool the
space. The coolant has done its job, but it is now a hot vapor (having evaporated and
absorbed significant heat). The coolant vapor moves on to the next step, absorption.

Step 2: Absorption
This step represents the drastic difference between absorption chillers and electrically
driven devices. In the traditional system, electrical energy is used to heavily compress the
coolant vapor. In an absorption chiller, the coolant vapor is absorbed into a solution. As
mentioned earlier, the absorbent is a concentrated solution of lithium bromide. Lithium
bromide is a non-toxic salt and an extremely powerful desiccant. Because of its desiccant
nature, the solution readily absorbs the water vapor. The vapor condenses as it is
absorbed, giving off the latent heat it absorbed in the previous step. This heat is removed
by a cooling tower that is necessary in all absorption chiller installations. The coolant

36

vapor and the concentrated desiccant have mixed, creating a now dilute mixture. This
solution is pumped into the chiller generator.

Step 3: Generator
In order to be used again, the coolant must be separated from the dilute solution. Hot
water from the solar collectors, via the storage tank, is fed into the generator, which
contains the dilute solution. The heat causes the water in the solution to boil out of the
mixture (the system is depressurized, so the water boils at well under 212F). At the end of
the process, what remains are pure water vapor and a concentrated lithium bromide
solution. The process is analogous to solar desalinization, where heat from the sun is used
to distill sea water, creating fresh water vapor and leaving behind a far more concentrated
saltwater as the byproduct (a repeat of this cycle over thousands of years creates the high
salt content of water bodies such as the Dead Sea and the Great Salt Lake). The
concentrated lithium bromide is moved via gravity to the absorber, which was discussed
in Step 2. The hot water vapor is moved into the next stage, the condenser.

Step 4: Condenser
In the traditional chiller, the superheated vapor had to be cooled and condensed before
the cycle could repeat. This was done by exposing it to outside air that was at a much
lower temperature than the vapor. The absorption chiller, with its coolant now as a hot
vapor, must also condense the coolant for it to be of any use. Though air cooling is
possible, the most efficient technique is to again utilize the cooling tower mentioned in

37

step 2. The hot coolant gas is exposed to a loop of cooling water, causing it to condense
into a cool liquid. The heat of this condensation is rejected by the cooling tower. With the
coolant now a cool liquid, the process can repeat. The cycle is illustrated in Figure 2.12.
The dual function of the cooling tower is shown in yellow.

Figure 2.12: A simplified view of a single stage absorption chiller capable of running on hot water
produces by an array of evacuated tube solar thermal collectors. Source: Yazaki Energy
www.yazakienergy.com

The cycle just described is a single-stage absorption cycle. Higher efficiency double and
triple stage cycle technology exists, but requires much higher input temperatures that
cannot be produced by non-concentrating solar collectors. For that reason, this study will
focus exclusively on single effect chillers.

38

The driving factor of cooling capacity for this system is the temperature of the
inlet water. The Yazaki chiller used at the Audubon will operate when the heat medium is
between 158F and 203F degrees, though cooling capacity grows as the inlet temperature
increases, as shown in Figure 2.13.

Figure 2.13: This chart from Yazaki shows the conditions necessary for the chiller to reach its nominal
output. This chart can be used as a baseline for design.

Though the chiller is a heat pump, it does not offer near the efficiency of electrically
driven heat pumps (where coefficients of performance between 3 and 4 are common). In
the book “Solar Assisted Air Conditioning in Buildings,” produced by the International
Energy Agency, the COP of the average single stage solar fired absorption chiller was
listed as 0.65. This makes it necessary to supply roughly 1.5 times as much energy as the

39

system can ultimately deliver. While this would be very inefficient for a traditional
system, in the case of solar fired system it is less bothersome because the energy source is
free and plentiful. Still, this factor must be considered when designing the system.

History and development
As mentioned earlier, solar thermal heating is a very natural process, and as such
there is little historical tracking of its development. Solar cooling has a much more recent
and traceable timeline of development. The first demonstration of a solar assisted cooling
system occurred at the Paris World Exhibition in 1878 (Henning 2007). The next century
saw little development in active solar cooling, though in this time traditional electric
chilling developed rapidly and changed people’s perception of thermal comfort. It wasn’t
until the 1970s and 1980s that interest in solar fueled absorption started again (Henning
2007). An oil embargo during this time created one of the first pushes for renewable
energy; even the White House was outfitted with new solar thermal panels for domestic
hot water. However, like much environmentally focused research at the time, the end of
the 80s saw a significant reduction in progress in the area. It wasn’t until the 1990s that
the technology really gained momentum. Renewed environmental consciousness,
combined with concerns about dangerous refrigerants, led to a surge in research.
Additionally, concerns about peak power and brownouts led engineers and researchers to
look harder at the problems posed by traditional cooling schemes. The decade also saw

40

mass production and lower cost for highly reliable solar components (Henning). Lastly,
government incentives for renewable energy incentives spurred installations of solar
thermal systems.
A small number of solar thermally driven HVAC systems exist in the world, with
a significant majority in Europe. A 2006 study in the ASHRAE journal identified 40
absorption cooling systems in Europe ranging in size from 3 to 72 tons of cooling power.
Most of them are in Germany and Spain and provide cooling primarily to office space.
The units were installed between 1979 and 2004. Thirteen of fifteen Spanish installations
occurred after 1999, showing the novelty of the technology (Balaras 2006). There appears
to be a very small number of installations in the United States, with the Audubon as the
most well known example. Small bits of information regarding two systems in the
Sacramento area, one in Texas, and one in Florida were uncovered, but no verifiable
studies or in-depth research were found for any of the installations.
The worldwide leader in research into these systems is the International Energy
Agency. This multi-national body was founded in 1974 and serves to promote and carry
out research and development into “improved energy technology.” One of the first
programs established by the agency was the Solar Heating and Cooling Programme. This
34-year old program is dedicated to using solar energy to “heat, cool, light and power
buildings.” The United States is a member of the program.
The Solar Heating and Cooling Programme works on specific tasks in the area.
Task 25, established in 1999, is Solar-Assisted Air-Conditioning of Buildings. The
objectives for this task are described as:

41

“to improve conditions for the market introduction of solar-assisted cooling
systems in order to promote a reduction of primary energy consumption and
electric peak loads due to air-conditioning and thereby to develop an
environmentally friendly method of air-conditioning of buildings.”
The operating agent for Task 25, Hans-Martin Henning of the Fraunhofer Institute, edited
a book in 2004 (and a second edition in 2007) called “Solar Assisted Air-Conditioning in
Buildings: A Handbook for Planners.” It is a tremendous resource for this study.
Further research into solar cooling systems is underway at Carnegie-Mellon
University in Pittsburgh, PA, where they have installed a parabolic trough and double-
stage absorption chiller to cool an academic building. Professor D.S. Kim of the Delft
University of Technology has also authored a number of papers on solar driven
absorption chillers and the thermodynamic processes involved in their operation.

42

Chapter Three: The Audubon Center at Debs Park

Upon construction, the Audubon Center at Debs Park set a new standard for
environmentally sensitive structures in the United States. Shortly after completion, the
building was awarded LEED-platinum certification (the highest level of certification
from the United States Green Building Council), the first building in the United States to
earn the distinction. This chapter will describe the building and the unique design and
construction approach that made it a benchmark for energy use. Particular attention will
be given to the steps used to moderate heating and cooling loads, followed by a detailed
description of the solar thermal HVAC system.

Figure 3.1: The entrance to Audubon Center at Debs Park.
Source: http://raydlett.com/albums/audubon/pages/AUCD6B_jpg.htm

43

The decision to pursue LEED platinum status for the building was made early in
the design process and was motivated by two important considerations: the cost of
connection to the public electrical grid and the securing of a private donation. The
building, although in an urban metropolis, actually sits in a somewhat remote section of a
county park. The nearest utility connections are over a quarter mile away, and there
would have been a significant upfront cost to bring the public network to the site. The
cost premium associated with going off-grid was only slightly higher than the price of
grid interconnection, and going off-grid fit perfectly with an aggressive green building
strategy. The second event was financing via a private donation that was earmarked for
an energy efficient design. Because of these circumstances, the building was designed as
an eco-friendly structure from the ground up, the first Audubon Center in California to do
so (USGBC 1). The architecture firm was EHDD of San Francisco. Collaboration and
support came from a variety of engineers and consultants, including IBE Consulting
Engineers as the mechanical and plumbing consultant and CTG Energetics as the LEED
consultant. Many green features were not related to the HVAC system, but still
contributed valuable LEED points. A large amount of natural, local, and recycled
building materials were used. A water reclamation system and native landscaping reduces
the water need to 30% of a comparable campus. A recycling scheme during construction
funneled 97% of construction waste to recycling plants (USGBC 1).
The 5,032 square foot building is broken into 4 primary spaces. The discovery
room is a large space that serves as an indoor educational center and hands on laboratory
for the many local children who visit the site through their schools. There is an employee-

44

only library and meeting room on the south end of the building connected to a small
catering kitchen. At the main entrance is a staffed reception lobby. Connected to this area
via a hallway are the building offices. The offices are the only areas with consistent daily
occupancy and contribute the most to the internal building loads. Within the office space
is a separately ventilated copy room, as well as a number of computers and other office
loads.
Specifically important to this study are the features designed to directly reduce the
heating and cooling loads. With only 7 or 8 full-time occupants, and no heavy internal
loads such as a server room, the building loads are envelope dominated. Because of this,
properly evaluating the climate and designing in response to it was vital. The temperate
and generally pleasant climate of Southern California makes for much smaller base loads
than would be seen if the building was in Miami or Detroit. By working to reduce these
loads, the size of the HVAC system could be reduced, very important for the new and
pricey technology implemented at the Audubon. Climatic data for the site is detailed in
Chapter 5.
Exposed thermal mass was used throughout the building. Poured concrete floors
and CMU walls are visible in the discovery room and the reception area. The materials
were chosen for their well understood thermal properties, absorbing and storing heat
during the day, then releasing it at night. In relatively dry, warm climates such as
Southern California, this serves to regulate interior temperatures and minimize large
temperature swings and fluctuations. High spaces with operable clerestory help to flush
out heat that has risen to the top of the room due to convection. The windows also assist

45

in ventilating the space, providing 100% natural ventilation when there is a sufficient
breeze. When this is not the case, energy-efficient fans handle the ventilation needs.
Fluorescent lights are used in the building, which adds much less sensible heat to
the space than incandescent lighting. However, the ample glazing on the building reduces
the lighting load to near zero. With no night-time occupancy, lights are only needed for a
very small period of time around the winter solstice. The high amount of glazing presents
other thermal challenges that were addressed. All south facing glass is shaded by a
decorative trellis covered with vines. Direct penetration of sunlight into the windows is
minimized in the summer, reducing heat gain and easing the load on the cooling system.
All windows feature high performance low-e double glazing (USGBC 1).
All these systems combine to reduce the building HVAC loads to a level well
below what might be expected in a standard building. The next challenge was to meet the
heating and cooling demands. An innovative solar thermal system, one of only a small
handful world, was used. Since the system is driven by solar thermal energy, the natural
starting point for explaining the system is the solar thermal collectors. There are 408 total
tubes, each 100mm in diameter and 2000mm in length. The tubes are grouped in modules
of 8. There are 51 modules: a 3 by 11 grid on a south facing roof and a 2 by 9 grid on a
west facing roof. Both roofs have a slope of 14 degrees. The modules are produced by the
Chinese company Sunda and are the heat pipe variety described in Chapter 2 of this paper
(Wright 1). The array occupies 1080 ft
2
of roof space. The actual absorber plate within
the tube does not extend completely to the edges of the tube. This, in combination with a

46

small space between adjacent tubes, drops the actual area of solar absorptive material to
722 ft
2
(Bergquam 1). Each evacuated tube collector weighs 4.6 kg.

Figure 3.2: The south facing roof (above) and the west facing roof (below)

47

The heat pipe is an 8mm diameter copper tube. The absorber is a fin shaped strip
of metal connected to the central tube, as shown in Figure 3.3. The absorber material is
copper-aluminum, with a thickness of 0.47mm. The fin has an aluminum nitride coating,
which is spectrally selective to increase solar absorption. Sunda claims an absorption
coefficient of 0.92 for the collector. Surrounding the heat pipe and fin is a borosilicate
glass tube 2.5mm in thickness. The glass transmittance is 91%. The space inside the tube
is evacuated to less than 10mbar (Sunda 1). This vacuum causes an inward force on the
glass that is naturally countered by the cylindrical shape of the tube and prevents
substantial convective exchange between the absorber and the glass.

Figure 3.3: A close-up of an individual evacuated tube solar collector. The ridged absorption fin is clearly
visible. Source: www.seasolarstore.com

Each tube plugs into a manifold atop the module. The top of the heat pipe, which
is a condenser chamber, exchanges the acquired solar heat with a water flow moving

48

through the manifold. The manifold header pipe, shown in Figure 3.4, is constructed from
a 1mm thick copper pipe. The header pipe is inserted into an aluminum casing and
capped, leaving only the inlet and outlet exposed. This casing is shown in Figure 3.5.

Figure 3.4: Left - The header pipe within the heat exchange manifold of a Sunda solar collector module.
This header pipe is actually for a unit with 4 collectors. Right - The top of the heat pipe goes into the port
and through to the other side of the manifold, ensuring a large amount of surface area for heat transfer.
Source: http://www.sundasolar.com/product_collectors.html

49

Figure 3.5: The completed unit. With the manifold cover in place, the unit has two openings, a water inlet
and outlet.
Source: http://www.sundasolar.com/product_collectors.html

On the south facing roof, each line of 11 modules is strung in series. The 3 strings
are then connected in parallel. Similarly, the west roof has 2 strings of 9 panels. In this
configuration, each roof plane has two pipes leading from it, one supply and one return.
The two supply lines are joined with a T-fitting just beyond the solar collector pump,
labeled as P1 in Figure 3.7. The inlet for the pump is in the bottom of the hot water
storage tank. The two return lines each feed into the hot water storage tank individually.
A schematic of the solar collectors is shown below in Figure 3.6.

50

Figure 3.6: The layout of the solar collector array at the Audubon Center

Figure 3.7: The control schematic for the HVAC system at the Audubon Center.
Source: Audubon Center at Debs Park control system software.

51

The next major component is the storage tank. In all solar cooling installations, a
storage tank is recommended to provide the ability to supply cooling when solar
insolation is not high enough to provide adequate energy for the chiller. It is even more
important in the winter, when peak loads are completely out of phase with the availability
of solar energy. At the Audubon Center, there is a 1200 gallon glass lined steel container
surrounded with polyurethane foam insulation. The tank is built to supply water to the
system at a maximum 195F for cooling and 140F for heating (Bergquam 1). The tank has
manual and automatic pressure relief valves to prevent issues arising from pressurized
steam or overfill. A photo of the collector tank is shown in Figure 3.8.

Figure 3.8: The hot water storage tank at the Audubon Center

52

Beyond the tank, the system relies on a well designed, albeit somewhat confusing
piping scheme to meet the thermal demands of the building. The heating mode is very
easy to understand. The building pump, labeled as P3 in Figure 3.7, pulls hot water from
the storage tank, passes it through a check valve, and pumps it into the building through
the supply line of the “primary loop.” The supply line diverts water into three “secondary
loops,” one for each HVAC zone in the building. Each secondary supply leads to a fan
coil unit, which blows conditioning air over the piping, picking up heat from the water
and delivering it to the space. The removal of heat from the water cools it down; this
water is now considered secondary return water. At this point, the secondary return loop
moves the cooler water to the return side of the primary loop. This return line brings the
water back to the storage tank. This is nearly identical to a standard hydronic heating
system, except the traditional boiler is replaced with the solar thermally driven storage
tank. If environmental conditions allow, both the building pump and the solar pump
operate at the same time. Water will flow into the building, give up its heat to the space,
and return to the bottom of the storage tank. Here it will be drawn into the solar array,
enter one of the two solar arrays and heat up again before returning to the storage tank,
ready to be redistributed into the building. The highest heating demand, however,
happens in the evening when no solar radiation is available. Operating the heater in the
evening would generally deplete the heat in the storage tank until heating loads could no
longer be met. This type of depletion is avoided at the Audubon Center because the
building is usually unoccupied in the evening, thus requiring no heating of the space. The

53

temperature of the storage tank drops during the evening only due to conduction through
the walls of the tank.

Figure 3.9: Water flows when the system is in heating mode.

The cooling system is more complex, but can be followed reasonably well in the
simplified schematic shown in Figure 3.12. Most importantly, the Audubon uses the same
primary and secondary loops for heating and cooling, with the mode of operation
determined by the season. The mode switch is handled by two valves, labeled V1 and V2
in Figure 3.7.
To operate as a chiller, hot water is again pulled from the storage tank. Instead of
being routed into the primary loop, it is diverted into the absorption chiller by valve V1,
which is electronically controlled. The chiller uses the hot water to fuel the absorption

54

process. The hot water gives off its heat in the generator, then returns to the storage tank
where it can be heated again if insolation is sufficient. This loop is driven by the
generator pump, labeled as P2 in Figure 3.7. The chiller in turn produces chilled water for
meeting the cooling loads of the building. This chilled water is sent through a check valve
into the supply line of the primary loop
3

. After moving into the building and into the
secondary loops, the chilled water absorbs heat in the building spaces, providing the
space cooling. The return portions of the secondary loops recombine in the return of the
primary loop. This return terminates back at the chiller, delivering the return water back
to the evaporator to be cooled once again. A valve, labeled as V2 in Figure 3.7, diverts
the return water back into the chiller, rather than into the storage tank as in heating
season.

3
The check valve, like the one in the heating line, is a one-way valve. When cooling, the heating line
check valve prevents the flow of cooling water back into the storage tank. When heating, the check valve in
the cooling line prevents the flow of heating water into the absorption chiller. Both valves are shown as “Z”
shaped symbols in the Figure 6.

55

Figure 3.10: The metal structure to the left is the absorption chiller. In the background is the storage tank.
The pipe in the foreground come from the chiller and contains chilled water to supply the primary loop of
the building.

56

Figure 3.11: A shot of the HVAC area shows the maze of pipes and conduit necessary for operation and
control of the system. The hot water line check valve is not shown.

As discussed in Chapter 2, the thermodynamic processes involved in absorption
cooling produce a significant amount of waste heat that must be rejected in order for the
unit to run efficiently. The heat is removed by a cooling tower placed next to the chiller.
The two condensation cycles in the absorption chiller deliver their latent heat to an
independent water loop leading to the cooling tower. This loop is driven by an
independent pump, labeled as P4 in Figure 3.7. The cooling tower, made by Marley, is
capable of rejecting up to 290,000 btu/h. It receives water from the chiller at 94F and
feeds it back into the system at 85F, ready to absorb more heat and continue the cycle
(Bergquam 1).

57

Notice that in cooling mode there are between 2 and 4 four water loops active at any
time. These four are the generator, building, cooling tower, and solar collector loops.
1) The generator loop – hot water is sent into the chiller generator to start the
absorption cooling process. It is output back to the storage tank at a lower
temperature.
2) The building loop – the primary hydronic HV AC loop, brings chilled water into
the building where it is fed to the secondary loops. After absorbing the heat in the
space, it is returned to the chiller warmer than at the start. It is re-cooled and
cycled into the building again.
3) The cooling tower loop – this loop does not function all the time. It is off when
the cooling tower is operating at a low enough capacity that there is not much
need for heat rejection. This generally happens only very early in a duty cycle for
the chiller. Once the chiller is functioning near capacity, the cooling tower loop
runs concurrently.
4) The solar collector loop – since cooling demands and solar energy availability
generally coincide, the system often runs the collector loop to maintain the
temperature of the water in the storage tank.

Loops 1 and 2 are always active if the system is actively cooling. Loops 3 and 4 are not
necessarily active, but generally will prove to be on when the chiller is operational.

58

Figure 3.12: The system flows in cooling mode. The primary loop is fed with chilled water produced by the
absorption chiller, shown in blue. Hot water from the storage tank, shown in red, fuels the absorption
process. Two independent loops keep up the supply of hot water and reject waste heat.

Operation of various loops is run by a sophisticated control strategy, operated via a
software interface on a dedicated computer. The mode is set to either heating or cooling
by a schedule within the software. Valves V1 and V2 route the water flow into the
appropriate channels depending on the operation mode. Valve V2 is manually operated,
which differs from the electronically driven operation of valve V1. On the day of the
switch from heating season to cooling season, or vice versa, someone familiar with the
system must walk out to the HVAC area to rotate the manual knob on valve V2. This
requirement can lead to problems. On October 14, 2009, the computer followed the set
schedule and switched the system in heating mode, turning valve 1 to the 0% position.
The staff at the Audubon Center, dealing with very warm weather at the time, did not turn
the manual switch knowing that cooling was still very obviously needed. The system was

59

non-functional for the next 36 hours until Audubon’s outside HVAC consultant reset the
seasonal schedule to cool for a few more weeks.

Figure 3.13: Valve V2 is operated by manually turning the disc on the top of the housing.

The HVAC system at the Audubon Center shows the need for experienced
mechanical, plumbing, electrical, and control engineers to design the system, as well as
an experienced contractor to install it. For the purposes of this study, a proper
understanding of the setup is vital to understanding the potential of this technology and
for the development of a design guide for future systems.

60

Chapter Four: Data Acquisition

As mentioned in Chapter 3, the designers of the HVAC system at the Audubon
Center included a robust network of sensors integrated into a customized energy
management system (EMS). It is likely that the design team recognized the uniqueness of
the project and took an aggressive monitoring strategy to both control system
performance and help diagnose problems for the high tech system. The energy
management system polls the sensors periodically and stores the data in logs whose
storage capacity varies by parameter.
The sensors are installed throughout the site, with some in the building, some
outside, and some integrated into the HVAC equipment. Included are multiple flow
sensors, temperature sensors, Boolean status indicators, and an on-site weather station.
This paper used data from 24 individual sensors, which represent roughly 75% of the
sensors on-site. The unused sensors were not relevant to the project goals. In addition to
the internal network installed with the building, two external HOBO
temperature/humidity sensors were installed in the building to confirm the interior
conditions reported by the EMS. In all, 29 parameters were monitored and combined to
create the data set necessary for the analysis portion of the project.
This chapter will first describe each of the data parameters and its importance to
the study. It will then describe the data acquisition and organization process. Analysis of
the data occurs in Chapter 5.

61

Data parameters
General Parameters
1) Time and date: This is the only parameter that is not coming from a sensor, but
rather from the central EMS computer. The time and date stamp occurs in the
output of every individual sensor and is vital for aligning the data into a coherent
set. Data are output as YYYY-MM-DDTHH:MM:SS, where Y stands for year, M
for month, D for day, H for hour, M for minute, and S for second. For the EMS
sensors, the sensing interval was every 15 minutes, with measurements at 0, 15,
30, and 45 minutes past the hour. All time stamps were saved in 24 hour format,
eliminating confusion between AM and PM.

Weather Station Parameters: A Davis Vantage Pro2 Plus weather station is installed on
the roof of the Audubon Center. It collects the data parameters described below, plus a
few others that were not relevant to the study. The weather station collected data once
every 5 minutes, after which it was uploaded to an internet site and archived (currently
data exists back to June 2009). The data can be retrieved from:
http://www.wunderground.com/weatherstation/WXDailyHistory.asp?ID=KCALOSAN42
In order to match the weather data to match the time-stamped EMS data, it had to be
reduced to 15 minute intervals. This was done in Microsoft Excel by isolating the time
stamp and filtering it to leave only times that matched up with those being monitored by
the EMS. For example, the weather station reported full data for 14:10, 14:15, and 14:20.

62

Since the EMS collects only at quarter hour increments, the weather data filter eliminated
the 14:10 and 14:20 data points. These extra data were not deleted, but rather hidden
from view. This kept the data available for advanced analysis if necessary. It was very
important to get this data from the building site. In many cases, climate data from a few
miles away would suffice. However, this isn’t the case in Southern California. The
unique geography of the area creates many diverse microclimates ranging from coastal
conditions, mountain areas, and hot valleys. The Audubon Center is roughly 18 miles
inland. Temperature information from Los Angeles International Airport, which is on the
coast, would not have worked for this study.

2) Outdoor Temperature: the exterior temperature a few feet outside of the Audubon
Center, measured in degrees Fahrenheit.

3) Outdoor Humidity: the moisture content of the air just outside the Audubon
Center, measured as a relative humidity. The exterior temperature and the exterior
humidity largely determine the need for air conditioning in this building,
particularly in the exhibition rooms that have limited internal loads.

4) Dewpoint: another indicator of the moisture content of the outside air, measured
in degrees Fahrenheit.

63

5) Hourly Precipitation: an estimate of the amount of rain falling in one hour,
measured in inches. Tipping bucket tips after collecting 0.01” of rain.

6) Daily Precipitation: a measurement of the total rain falling on the site in one day,
measured in inches. This always resets to 0 at the first measurement of a new day.
The daily and hourly rain data are important for identifying periods of inclement
weather and the intensity and duration of the weather event. Precipitation
generally implies a significant lack of solar energy, and it is one goal of this study
to identify the system response during and after weather events.

7) Solar Radiation: a measure of the global horizontal insolation, measured in Watts
per square meter. This represents the main energy source for the system. As such,
it is one of the most important data parameters. The sensor is reporting insolation
on a horizontal plate, but the solar thermal collectors are angled to match the roof
pitch. This leads to a slight difference between the reported insolation and that
which is actually incident on the collectors. However, this difference is small.
Additionally, because insolation is commonly reported as global horizontal, this
paper will use that quantity in its analysis and recommendations to remain
consistent with the most general data availability. The angular modifier to
insolation will be ignored.

64

System Parameters: a network of flow, temperature, and operation status sensors are
installed throughout the HVAC system. They are usually visible and well marked.
Temperature and flow sensors are installed directly into the system piping. The sensor is
connected to an electrical junction box, from which a communication line is run to the
control unit. Measurements were taken every 15 minutes and time stamped in the format
shown in parameter 1.

8) F1 Building Water Supply Flow: a flow meter installed just downstream from the
pump that moves water into the building primary loop, measuring in gallons per
minute. This measurement is vital for understanding the cooling or heating energy
delivered to the building. It is also an indicator of when the system was running. It
shows, for example, if the system improperly ran during unoccupied hours or if it
failed to run when conditioning was needed. The former would indicate a controls
problem while the latter would indicate a system problem related to design or
operation.

65

Figure 4.1: Flowmeter F1 (left) and temperature sensor S6 (right) on the building supply line.

9) F2 Solar Collector Hot Water Flow: a flow meter in the solar loop, measuring the
flow of water through the collector network in gallons per minute. This number is
necessary to calculate the amount of heat transferred by the collector network,
which is necessary for calculating their thermal efficiency. It is also an indicator
of when the pump was running and is valuable to determine if solar collection is
being optimized.

10) FC1 Discovery Room Temp: interior temperature in the Discovery Room,
measured in degrees Fahrenheit. This room is a freestanding exhibit and education
room that is generally unoccupied and free of internal loads. When it is used,
occupancy jumps, sometimes to 50 or more people. The room is considered a

66

separate zone and has its own air handler unit supplying conditioned air to 3
supply vents. The room is connected to an unconditioned storage, mechanical, and
garage space.

11) FC2 Library Room Temp: interior temperature in the Library Room, measured in
degrees Fahrenheit. This room is a library and meeting space for employees only.
It is generally unoccupied and has some internal loads from kitchen equipment in
a prep space connected to the room. When in use, occupancy jumps to
approximately 8 to 10 people. The room is considered a separate zone and has its
own air handler unit.
12) FC3 Director’s Offices Room Temp: interior temperature in the offices, measured
in degrees Fahrenheit. The offices are a separate zone with their own air handler
unit. They are the only permanently occupied spaces with approximately 6 full
time employees. There is a range of office equipment including computers,
datacom equipment and panels, copy machines, and kitchen appliances. Knowing
the interior temperature of the spaces is the main indicator for determining if
thermal comfort demands are being met. Additionally, we can use the data to
determine the performance of individual zones against one another and what
might be the effects of the particular zone’s load profile.

13) S02 Solar Array Entering Temperature: the temperature of water as it flows out
from the bottom of the storage tank and into the collector array. Knowing this

67

temperature is required for calculating the thermal energy delivered by the
evacuated tube collectors.

14) S03 Solar Array Leaving Temperature: the temperature of the water after it leaves
the collector array and before it renters the storage tank, in degrees Fahrenheit.
This value, along with the array entering temperature and the flow rate, gives an
approximation of the heat delivered by the array to the storage tank.

15) S04 Storage Tank Temperature: the temperature, in degrees Fahrenheit, of the
water in the storage tank. Photos indicate this temperature probe is in the top of
the tank, which is important considering the stratification effect present in the tank
when water is being drawn from and added to it. The water at the top of the tank
is the hottest, meaning it is less dense then the cooler water at the bottom (which
causes the stratification). The chiller will not operate until the hot water supplied
by the tank reaches 158F. Still, the chiller works well below its nominal capacity
at this temperature. Temperatures above 170F are desired, because at this range
the chiller can operate at 0.75 times its nominal value, meeting the 7.5 ton
modeled peak loads (broken down in Chapter 5).

68

Figure 4.2: A Picture showing S4, the Hot Water Storage Tank Temperature

16) S06 Building Water Supply Temperature: the temperature, in degrees Fahrenheit,
of the water entering the supply side of the primary building loop. In the winter,
this water is drawn directly from the hot water tank. In summer, the line is fed
with chilled water from the absorption chiller, with a valve blocking the line to the
hot water tank.

17) S07 Building Water Return Temperature: the temperature, in degrees Fahrenheit,
of the water in the return line of the primary loop. In the heating season, this water
is routed back to the storage tank. In the cooling season it is sent to the return line
of the chiller to be re-cooled. The supply and return water temperatures, combined

69

with the flow rate in the loop, give an idea of the heating or cooling energy
delivered to the space.

18) Solar Collector Hot Water Delta: the difference, in degrees Fahrenheit, between
the thermal collector supply temperature and the return temperature. In the control
scheme, this difference is used to determine when to start and discontinue water
flow through the collectors. In the original design, a delta of 8 degrees (meaning
water returning from the collectors was 8 degree warmer than water entering) was
used to trigger the collector pump.

19) Building BTU Cooling/Heating: the system reports a number listed as Bldg BTU
Cooling or Bldg BTU heating depending on the season. This number is
continually lower (by 1000s of Btus) than would be expected or is indicated by
other measures. As such, it is very suspect and not frequently used in the study.

20) Cool Mode/Heat Mode: a simple true/false indicator, one for each season, which
reports the operational status of the system. While there are other ways of
determining this (like valve position), the measurement is a valuable resource for
determining when system parameters were changed and how the occupants
pushed back scheduled dates for system changeovers.

70

21) P1 Solar Pump: a true/false indicator, reported as “Active” or “Inactive,”
describing the status of the solar thermal collector pump. The parameter is
valuable for quickly filtering down information to perform calculations regarding
heat transfer in the collector array.

22) P2 Chiller Generator Pump: true/false indicator, reported as “Active” or
“Inactive,” describing the status of the pump moving hot water into the chiller
generator. This parameter allows for rapid isolation of times in which the chiller
was operational and producing cooling.

23) P3 Building Pump: a true/false indicator, reported as “Active” or “Inactive,”
describing the status of the pump supplying the buildings primary hydronic loop.
In the cooling season, the status of the pump is nearly identical to P2. In the
heating season, when the chiller has been shut down completely, the pump kicks
on when there is a heating demand. The parameter is valuable to quickly filter out
data points produced when no HV AC was necessary.

24) Valve V1 Position: a percentage indicator, from 0 to 100, describing the position
of Valve V1. This valve controls what type of water flows into the primary
hydronic loop; chilled water in cooling season, hot water in heating season. A
value of zero indicates heating position, while 100% indicates cooling mode.
These are the only two values seen in the data set, meaning the parameter behaves

71

more like true/false operator. This valve is electronically controlled by the heating
and cooling schedule.

Figure 4.3: Valve V1. Hot water from the storage tank enters from the left. In heating season it is send
downward to the building supply line. In cooling mode is follows the pipe on the right into the chiller.

25) Valve V2 Position: a status indicator, reporting “cooling mode” or “heating
mode,” describing the position of Valve V2. This valve routes the return water
from the primary building loop. In cooling season, the return water (now warmer
than supply) is sent back in the chiller to be cooled again. In heating season
(return water cooler than supply) it is sent back to the storage tank where it might
go through the collector array again if insolation is sufficient.

72

External Backup Sensors
26) Temperature Data Logger – Discovery Room: a small sensor installed by the
author to independently measure temperature, in degrees Fahrenheit, in the
Discovery Room. The data were used as both a backup and a confirmation of FC1
Discovery Room Temp

27) Humidity Data Logger – Discovery Room: the same HOBO sensor measuring
temperature also has a relative humidity sensor. These data are valuable for
analysis of interior thermal comfort, which is a factor of both temperature and
humidity. Interior humidity reporting was not done by the preinstalled sensing and
control system.

28) Temperature Data Logger – Offices: a small sensor installed by the author to
independently measure temperature, in degrees Fahrenheit, in the Offices. The
data were used as both a backup and a confirmation of FC3 Director’s Offices
Room Temp.

29) Humidity Data Logger – Offices: independent measurement of relative humidity
in the building offices. These data are valuable for determining thermal comfort in
this space, which sees higher latent heat gains due to continual occupancy.

73

Data Collection and Organization
Data were collected over a period of 5 months; this represented two distinct
thermal comfort challenges. Collection began on September 11, 2009, in the middle of
the cooling season. This start date was chosen because, upon the first research visit, it
was the first date for which full data were available in the control software log.
Approximately every 10 days, a visit was made to the Audubon Center to download the
available data. The last dataset was collected February 2, 2010. This was chosen as an
end date as it was determined there was enough heating season data to proceed to the
analysis phase of the project.
Cooling season data were collected from September 11, 2009 until the end of the
cooling season as determined by the control schedule. The listed end date for cooling
season was October 1, 2009. This differs from the original design parameter of
November 1 and is a curious choice considering that September is historically the second
hottest month of the year at the building site. Not surprisingly, the schedule was adjusted
on October 2, with the changeover date pushed back to October 16. On that date it was
again still far too warm to disable the cooling equipment. The changeover was pushed
back again, this time to the original design parameter of November 1. The system was
adjusted a 3
rd
and final time to a changeover date of November 7, which ultimately
became the last date of the cooling season master spreadsheet. All system adjustments
were made by a contracted energy consultant who works with the Audubon center to help
operate and optimize the system.

74

Construction of the spreadsheet was relatively straightforward. Each data
parameter was logged with a timestamp followed by a value for the parameter. Once the
full set of cooling season data were in hand, the first step was to make a continuous,
chronological comma delimited file for each parameter. This consisted of a data reported
every 15 minutes from the start to finish date. For example, for storage tank temperature
the CSV file would contain on each line two values, a time stamp and the temperature,
separated by a comma and followed by a carriage return. Once this had been done for
every parameter, each individual parameter’s dataset was imported into a master
spreadsheet in Microsoft Excel. The timestamps were left in place to properly align all
the data, such that each row in the spreadsheet contained the value of each parameter at a
given time. Once this was confirmed, the extra timestamp columns were removed. The
final cooling master data grid was 5520 rows by 24 columns. Not every parameter was
placed in the master sheet. For example, the position of Valve V2 remained in cooling
mode for the duration of the cooling season. There was no need to put it in the analysis
spreadsheet.
Construction of the heating season master sheet was done in the same manner. It
runs from November 8, 2009 until February 2, 2010 and measures 8296 rows by 22
columns. It has more rows than the cooling sheet because of the longer total time period.
It has fewer columns due to the removal of the “P2 Chiller Generator Pump” parameter
(not used in heating) and the “Heat/Cool Mode” parameter (no change in value).

75

Figure 4.4: A sample of the Cooling Season Master Spreadsheet including some data missing due to
equipment malfunction.

A small number of data points are blank. This was caused either by a malfunction
in the reporting equipment or the time between visits to the center being longer than the
log’s storage buffer. In the cooling season master sheet, 1152 of the over 132,000 data
points are missing, representing less than 1%. In the heating season, 6.2% of the data are
missing. This comes almost exclusively from entire dates being empty due to extended
time between visits to the Audubon around the holidays. By simply eliminating these
days from the analysis, the missing data become inconsequential.
The large amount of data available provides a robust platform for the analysis and
evaluation of the performance of the HVAC system. This analysis is described in the next
chapter.

76

Chapter Five: Performance Analysis

This chapter presents a statistical analysis of the data collected from the Audubon
Center over the course of five months. It begins with a breakdown of the site climate and
building loads, which establishes performance criteria for the HVAC system. Next, the
actual performance data from the building site is presented. This ranges from simple
comparisons, such as the internal temperature versus the outdoor temperature, to more
complex estimates of mass flows, efficiencies, and heat transfer in the system.
Throughout this analysis, care was taken to identify secondary and tertiary factors that
help to explain the observed data. While this chapter is heavy on data analysis, it is
purposely light on definitive conclusions from the information. That is reserved for
Chapter 7, where the conclusions are presented as design guidelines for future systems.

Climate and Site
The first consideration for the performance of the HVAC system is an
understanding of the climatic conditions at the building site. The Audubon Center at Debs
Park is firmly within California Climate Zone 9, described as a Southern Californian
inland valley climate (PEC 1). Reference cities for this climate are usually listed as either
Los Angeles - Civic Center or Pasadena. The Audubon sits in between these reference
points. The Guide to California Climate Zones, published by the Pacific Energy Center (a

77

division of Pacific Gas and Energy), contains information about the historic climate
within Zone 9. The document shows that the highest recorded temperature was 110F in
1955. The all-time low of 28F occurred six years earlier and is representative of the
essentially frost-proof locale. The 99% design temperatures are 37F in the winter and 93F
in the summer
4

. Climate Zone 9 is influenced by both inland air, which tends to be warm
and dry, and coastal air that is usually cool and moist (PEC 1). This creates a relatively
comfortable climate, albeit one with warmer summers and cooler winters than nearby
coastal cities. The rainy season coincides with the heating season.

TABLE 5-1: Heating and Cooling Degree Days for major cities in Climate Zone 9.
Source: Pacific Energy Center

Table 5-1 shows the heating and cooling degree days for major cities in Climate
Zone 9. The base temperature for the calculations was 65F. It shows the area is neither
heating nor cooling dominated, with similar values for both. Generally, the heating
demand increases as one moves inland, presumably due to the stabilizing effect of the

4
Engineers and architects generally do not design a building to meet all-time temperature extremes, instead
using temperature window that represents 99% of the expected temperatures.

78

coastal climate becoming less influential than the desert climate present in the inland
regions of Southern California.

Figure 5.1: The distribution of heating and cooling degree-days in Burbank, CA for 2009 climate data.
Data Source: Degreedays.net

79

Figure 5.2: The one-sided degree-day distribution in Riyadh, Saudi Arabia helps to illustrate the mildness
of Climate Zone 9. Data Source: Degreedays.net

80

Figure 5.3: Psychrometric chart for Climate Zone 9.
Source: Software – Climate Consultant 4, Data: US DOE

Figure 5.3 shows an annual psychrometric chart for Climate Zone 9. Winter
conditions are generally cool and somewhat moist, with 13 inches of rain a year
occurring mostly in the winter. Summer conditions are warm with comfortable humidity
levels. Based solely on the data points (there is one for each hour of the year), it is clear
there isn’t radical deviation from the comfort zone, making passive comfort strategies an
effective energy efficiency strategy. Climate Consultant 4 verifies this with its Design
Strategies box (in yellow), which quantifies the predicted effect of various energy

81

efficient measures. Because the Audubon was designed as an energy efficient structure
from the earliest stages of the design process, it takes advantage of passive strategies as
the primary comfort mechanism in the building, putting high thermal mass, natural
ventilation cooling, and internal heat gains to use in providing year-round comfort. The
solar thermal HVAC system was intended to handle peaks extending beyond what
passive strategies could handle.
Load sizing estimates were performed by the project mechanical engineer, IBE
Consulting Engineers of Sherman Oaks, CA. After dividing the building into three
HVAC zones, the building loads were calculated as shown in Table 5-2. The full load
calculation showing the breakdown of each load source is shown in Appendix B.

Loads (btu/hr)
Zone
Sensible
Cooling
Latent
Cooling
Sensible
Heating
Latent
Heating
Discovery Room 26682 2322 10816 0
Conference/Library 15993 634 9061 0
Offices/Gift Shop 22483 2298 14889 0
65158 5254 34766 0

Total Cooling 70412
Total Heating 34766

TABLE 5-2: Load calculations for the Audubon Center at Debs Park. Courtesy of IBE Consulting
Engineers.

These calculations suggest a 6-ton (72 kbtu/hr) unit would suffice for the
building. However, the operating procedure manual, which was written by the system

82

designer, describes a 7.5 ton cooling load. The building loads from the mechanical
engineer are dated August 29, 2003. However, the claims of 7.5 tons are not dated,
making it unclear which load profile was ultimately settled upon. This paper will use the
7.5 ton estimate since it was used by the solar thermal system designer in his design
scheme. Notice that cooling loads outweigh the heating loads in all cases. This is
primarily a factor of the building’s design that takes advantage of available “free” heat.
Solar heat gain is significant in the heating profile, as is internal loading from electric
equipment and occupants. In the offices, which are the only continually occupied spaces
in the building, these sources combine to provide over 7,000 btu/h of heat energy into the
space. The largest heating load in all three spaces is conductive heat loss through the
glazing, though this is minimized by the use of energy efficient double glazed low-e
windows. Infiltration is only significant in the offices, where occupants moving in and
out introduce outside air in to the space.

Thermal Comfort
A natural starting point for the analysis of the performance of the HVAC system
is to determine if thermal comfort conditions were met. This same technique was used for
both the cooling and heating season data, though with slight changes to the parameters.
The first step was to filter the data to show only data points falling within occupied hours.
For the Audubon center, this was from 8:00 AM to 6:00 PM. The building is rarely

83

occupied outside normal operating hours. Filtering the data was accomplished by first
splitting the time stamp into two columns, one with the date, and one with the time. The
time column was then filtered with a simple “between” filter in Microsoft Excel, limiting
it to times between 8:00 and 18:00 (the time stamp was in 24-hour time to avoid
confusion). The data were then trimmed so that only the time stamp and temperatures of
interest showed. In this case, those data fields were exterior temperature and the interior
temperature in each of the three spaces. Figure 5.4 shows a screen shot of the completed
data profile. A plot was made showing each of the temperatures, shown in the following
pages.

Figure 5.4: A sample of the data for determining interior comfort

To confirm the functionality of the building integrated temperature sensors, the
temperatures in the discovery room and the offices were then compared against data from
the HOBO sensors installed by the research team. In both cases, the HOBO and building

84

integrated temperature data plots had an identical shape, though with a slight offset in
readings. In the discovery room, the HOBO sensor consistently reported a higher
temperature than the building integrated sensor, with an average difference of 1.84
degrees. However, the matching shape of the plots and consistency of the difference
between the readings (standard deviation was only 0.48) likely indicates that both
temperature sensors are fully functional and reliable, with the small temperature
difference due to different calibration. The HOBOs were not calibrated before the test
due to the desire to quickly get them in place before the cooling season ended. Therefore,
it is not known which sensor is closer to the correct temperature. However the
temperature difference is small enough to have minimal effect on the study, so the
temperature from the building integrated sensor is used in this chapter. In the offices, the
situation was similar, though the average difference between the HOBO and the building
sensor was only 0.44 degrees, a very strong agreement.

85

Cooling Season

Exterior
Temp
FC1
Discovery
Room
Temp
FC2
Library
Room
Temp
FC3
Director's
Offices
Room
Temp
average 76.2 72.1 73.2 74.9
standard dev. 10.9 3.4 3.0 2.7
max 103.4 87.1 82.5 83.4
min 44.0 64.2 65.5 67.5

TABLE 5-3: Statistics for the cooling season thermal comfort data

Figure 5.5: Plot showing the cooling season internal temperatures in the three HVAC zones in the
Audubon. The comfort zone is shown in light blue.

86

Figure 5.5 shows the internal temperatures in the building during the cooling
season. The exterior temperature is also shown for reference. Evident from the exterior
temperature line is the large diurnal swing present at the building site. This is a good
condition for nighttime flushing of the building, though this is not a strategy the building
uses. However, the cool nighttime temperatures do allow for some natural cooling
building via heat loss to the outside. Also apparent from the outside temperature plot is
the need for mechanical cooling. With temperatures often exceeding 90F and sometimes
100F, passive strategies alone are unable to adequately cool the building. The absorption
chiller is specifically installed for these situations. The highlighted comfort zone is 71F to
80F. The upper limit was chosen to match that of the typical range suggested by the
PEC’s profile of Climate Zone 9. The lower limit comes from the thermostat setpoint,
which for all three zones was set to 71F.
The offices were occupied throughout the time period shown. The Library and
Discovery Room saw occasional occupancy spikes but were entirely unoccupied for a
majority of the time. This likely explains why both the Library and Discovery Room are a
few degrees colder than the offices; there are rarely internal loads from people and
electrical equipment. The offices experience sensible and latent heat from the employees
in addition to sensible heat gain from five or six computers and miscellaneous office
equipment.
Table 5-3 shows the average temperature and standard deviation for each of the
data sets and includes the observed minimum and maximum. During occupied hours in
the cooling season, the offices averaged 74.9F. A standard deviation of only 2.7 degrees

87

indicated that temperature was generally steady. The office space exceeded the 80F upper
comfort limit in 2.9% of the data points. It fell below the 71F lower threshold 7.1% of the
time. The maximum temperature experienced in the space was 83.4F which, while
outside the comfort zone, is reasonable considering the small amount of time in which the
space was too hot. In Figure 5.5, it can be seen that this maximum occurred after a period
of multiple days in which the exterior temperature topped 100F. The minimum
temperature experienced in the office space was 67.5F. This occurred on November 7,
2009, the last day of the cooling season. It is not so much an indictment of the HVAC
system as it is an indicator that the switchover to heating mode was well timed. The other
two spaces were slightly cooler, on average, than the offices. They also experienced
slightly more variation in temperature than the occupied spaces. Temperatures exceeded
the comfort zone in 1% of the data points for both spaces. Uncomfortably cold conditions
were present in 41.1% and 26.1% of the data for the Discovery Room and Library Room,
respectively. A glance at Figure 5.5 shows that over-cooled conditions in the two spaces
seems to follow cooler overnight lows than days when conditions were comfortable. It is
possible that exterior conditions overcooled the spaces, with the coolth retained by the
exposed masonry walls and exposed concrete floor. Because of the building design and
occupancy profile, there was no method for reheating. The Discovery Room’s large south
facing glass doors are in the cooling season fully shaded by a large trellis topped with
translucent plastic roofing and thick vines, all but eliminating solar gains through the
glazing. Exterior insulation on the walls stops solar and ambient energy from penetrating
into the thermal mass. With virtually no plug loads and infrequent occupancy, internal

88

gains did not exist in the spaces. The two-pipe hydronic HVAC system also was not
capable of supplying heating water while in the cooling mode. The average temperature
in the Discovery Room was 72.1F, while in the Library it was 73.2F. This is not too far
off from the 71F setpoint, suggesting that the HVAC system was effectively cooling the
spaces. The data were reviewed to determine the amount of time the chiller was
operational in order to differentiate between envelope driven thermal comfort and that
due to the mechanical system. The data parameter P2 Chiller Generator Pump was used
for this purpose. This parameter was added to the dataset, which was still filtered to show
only occupied hours. The analysis showed the chiller pump was active 50.3% of the time.
The cooling season thermal comfort is the combination of the natural climate, the
envelope and the mechanical system.

Heating Season

Exterior
Temper
FC1
Discovery
Room
Temp
FC2
Library
Room
Temp
FC3
Director's
Offices
Room
Temp
average 62.7 64.5 65.5 68.4
standard
dev.
10.0 4.4 4.9 3.2
max 88.7 75.6 76.0 75.7
min 37.4 54.3 53.9 59.2

TABLE 5-4: Statistics for the heating season thermal comfort data

89

Figure 5.6: Plot showing the heating season internal temperatures in the three HVAC zones in the
Audubon. The comfort zone is shown in light blue.

Figure 5.6 shows the internal temperatures in the building during the heating
season, with the exterior temperature again provided for reference. The diurnal swing is
again evident in the data. Also, the Library Room and Discovery Room are again
consistently cooler than the Director’s Offices. The highlighted comfort zone is 68F to
77F. In this case, the lower limit of the Climate Zone 9 profile matched the building wide
setpoint of 68F. The upper limit was chosen based on the ASHRAE-55 interior comfort
chart, using an assumption of 50% relative humidity and 1.0 clo.
Immediately obvious from Figure 5.6 is the large amount of data falling outside of
the lower end of comfort zone. The average interior temperatures for the heating season
were 64.5F, 65.5F, and 68.4F for the Discovery Room, Library Room, and offices,

90

respectively. Only the offices had an average within the comfort zone, though only by 0.4
degrees. A standard deviation of 3.2 degrees indicates that the space was too cool a
significant portion of the time. This is confirmed by the data. The offices were below 68F
in 44.5% of the data. Performance was worse for the other two spaces, which for the
Library Room saw 66.3% of the data below 68F and for the Discovery Room 73.5% of
the data below the threshold.

Adaptive Comfort
Because it is an active system, the Audubon’s HVAC setup is generally subject to
the recommendations of ASHRAE Standard-55 for determining human comfort. Another
option is to view the system performance in terms of the adaptive comfort model, which
is generally reserved for naturally ventilated buildings. ASHRAE studies have shown
occupants have relaxed standards for comfort when they are aware the building is
naturally ventilated (Santamouris 1). In discussions with the Audubon staff, it became
clear that they were well aware of their system’s design and limitations and appear to
share the feelings regarding comfort for a building that is “naturally” conditioned.
The ASHRAE adaptive comfort chart, shown in Figure 5.7 relates interior
temperatures to the mean monthly exterior temperature. To perform the adaptive comfort
analysis for the Audubon Center at Debs Park, the mean outdoor temperature for
September through February was calculated. The interior temperatures for the month

91

were then paired with their corresponding monthly average. The focus was on the offices,
which were the only real representative of adaptive comfort since the measure is
inherently tied to the opinions of the occupants. The results are shown in Figure 5.7.
Cooling season data are shown in blue, while data points for the heating season are in red.

Figure 5.7: Audubon climate data plotted on the ASHRAE adaptive comfort model

The figure indicates that the absorption chiller worked well, as nearly all of the
cooling season data fits with the 80% acceptability limits of the adaptive comfort model.

92

However, heating season shows the same problems as when the data were evaluated with
a static thermal comfort model. A significant portion of the data falls well below the
lower limits considered comfortable. The chilly conditions are also supported by
anecdotal evidence. On many winter visits to the site to collect data, employees were
wearing hats and jackets at their desks.

System Components
Evacuated Tube Collectors
The evacuated tube collectors at the Audubon Center are the most visible part of
the HVAC system. While all the other components are hidden in an HVAC shed next to
the building, the solar collectors are on the roof, highly visible from the courtyard of the
building. While the aesthetics of such a system are up for debate, the performance of the
tubes can be analyzed from the available data. The factors that were calculated were the
amount of heat captured and delivered by the array, as well as the efficiency of the array.
One parameter needed to calculate the heat transfer in the array is the total
temperature increase between the inlet and outlet. Both of these temperatures are
available in the dataset; the difference will be referred to as “delta.” Knowing that a
higher delta value would indicate a higher heat transfer into the storage tank, it was worth
investigating factors leading to a high delta. Intuition would suggest that the temperature
increase in the array would rise with increasing insolation, as increased energy

93

availability ought to translate to increased energy absorption. This was tested by plotting
the delta against the insolation at that moment, shown in Figures 5.8 and 5.9. To produce
the graph, the data were first filtered to show only timestamps where the solar collector
pump was active. Delta was calculated by subtracting the inlet temperature from the
outlet temperature, and insolation was an independent parameter available from the
weather data portion of the dataset. An efficiency formula for the collectors considers the
ambient temperature as a variable, so the process was done for both the cooling season
and the heating season, rather than combining the two. Negative values of delta were
sometimes present at the very beginning of a pumping cycle. These were likely due to
stagnant water in the system. The inlet temperature was equal to the temperature at the
bottom of the storage tank. If water had been sitting in the collector array for a while, it
likely lost some its heat via conduction to the outside. It was assumed that these values
were occurring before enough water had cycled through the array to pick up any heat,
leading to the negative values. The negative values were eliminated before the chart was
finalized.

94

Figure 5.8: The temperature rise in the solar collectors for changing insolation values during the cooling
season. The black line is a best fit regression line with a forced zero-intercept.

Figure 5.9: The temperature rise in the solar collectors for heating insolation values during the cooling
season. The black line is a best fit regression line with a forced zero-intercept.

0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
0 100 200 300 400 500 600 700 800 900 1000
Delta T (F)
Insolation (W/m
2
)
Insolation vs Delta T - Cooling Season
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
0 100 200 300 400 500 600 700 800
Delta T (F)
Insolation (W/m
2
)
Insolation vs Delta T - Heating Season

95

In both figures, the data do not show an obvious correlation, especially at lower
insolation values. Some points are clearly outliers, such as the small number of points
above a delta of 25 degrees. The cause of these high values is likely a stagnation
condition, much like the one described that caused negative delta values. The difference
here is that the pump was off under strong sun, so the stagnant water in the tubes heated
up significantly without flowing anywhere. When the pump kicked on, the initial reading
for the output temperature reflected this temperature rise. Once the flowrate was
established, these large numbers disappeared. A best-fit line, forced to pass through the
origin, does reflect a direct relationship between the two variables. At higher insolation
levels, the data show stronger correlation. At these higher levels of insolation, it can be
inferred that the sky was relatively clear and there was ample sun. The consistency of the
incident energy would imply a more consistent response from the system. In the cooling
season chart, there is a cluster of data points above 600 W/m
2
. A similar cluster appears
in the heating season chart above 450 W/m
2
. In both cases the cluster indicates a
consistent temperature rise of between 3 and 7 degrees.
Unfortunately no data were available for the temperature rise in the individual
solar thermal arrays. Rather the data reflect the combined effect of both the south facing
and west facing arrays. The consequence of this is that the results cannot be extrapolated
to predict the temperature rise in an array, say, twice as large. If each array had been
individually monitored, their different sizes would likely produce two different values,
which would help towards making this conclusion.

96

The heat collection efficiency of the tubes was calculated in three different ways:
1) by a manufacturer’s published formula, 2) by average calculated heat transfer values
and 3) by instantaneous heat transfer values. The methods are discussed below.
Within the building files for the Audubon Center was a specification sheet for the
evacuated tube collectors. This sheet contained a formula for calculating the thermal
efficiency of the tubes:

EFF=0.77 – 2.88 (T
i
– T
a
)/I (1)

where T
i
is the inlet temperature in C, T
a
is the ambient temperature in C, and I is the
solar insolation in W/m2. The reliability of the equation was initially questionable,
especially at low insolation levels where the result is often a negative number. To
perform the calculation, the data were first filtered to isolate timestamps where the solar
pump was active. Blank temperature and insolation readings were then removed. The
needed parameters, all available in the master dataset, were then plugged in. Using this
method, the average efficiency in the cooling season was 0.398, though with a standard
deviation of 0.74. In the heating season the average efficiency was 0.273 with a standard
deviation of 0.80. The variability in the result cast some doubt on the equation.

97

Cooling
Season
Heating
Season
Average Array Entering Temp (C) 66 46
Average Ambient Temp (C) 28 20

Average Collector Efficiency 0.398 0.273

TABLE 5-5: Summary of data for method one, manufacturer’s formula

To mathematically calculate the heat transferred into the solar loop required the
use of a common formula in HVAC engineering. It is usually used to quantify the heat
transfer in hydronic heating and cooling loops, which is analogous to the heat gain in the
tubes. The equation is

Q=F*W
H2O
*60 min/hr*(T
o
-T
i
) (2)

where F is the flow rate in gallons per minute, W
H2O
is 8.35 lbs per gallon of water, T
o
is
the outlet temperature in Fahrenheit, and T
i
is the inlet temperature in Fahrenheit.
Multiplying 8.35 * 60 produces the formula Q = 500 * F * ΔT. From this, efficiency was
calculated as:

EFF = Q/(I*A*3.41btu/W·h) (3)

where I is insolation and A is area. An important consideration must be made regarding
the area. At the Audubon, the collectors occupy 100.33 m
2
of roof space. However the

98

actual area of solar absorber is 67.08 m
2
. This is because of the space between adjacent
tubes, the space occupied by the manifold, and the small space between the absorber fin
and the inside edge of the glass tube. The equation for Q is constant, but the efficiency is
different depending on which area one chooses. Using the absorber area will always give
a higher efficiency that is likely a more accurate description of the technology. However,
knowing the efficiency based on total roof area is more desirable as a design tool, as it
represents the true space required for an installation once the desired heat input has been
calculated.
The averages method simply substituted in average observed values for the
variables in Equations 2 and 3. The average values are shown in Table 5-6.

Cooling
Season
Averages
Heating
Season
Averages
Insolation 538 338
Solar Collector Flow Rate (gpm) 11.86 11.84
Array Entering Temp (F) 151 115
Array Leaving Temp (F) 159 122
Heat transfer (Btu/h) 47600 41300

Collector Efficiency 0.389 0.531
Whole Roof Efficiency 0.260 0.355

TABLE 5-6: Summary for method two, which uses Eqs. 2 and 3

Much of the data in Table 5-6 follows intuition. Average insolation is lower in the
heating season probably due to both the overcast weather often present in the winter
months in Southern California, and the lower sun angle (insolation values are global

99

horizontal). Lower ambient temperatures lead to lower temperatures throughout the
system, as reflected in the entering and leaving temperatures. However, there is one
significant difference between the first and second method. Both divide by the insolation;
however Equation 1 subtracts this value from a constant. Under this condition, lower
insolation creates lower efficiencies. In the averages method, no constant modifies the
fraction. Therefore a small insolation means a small denominator, which makes the
efficiency increase. This difference is evident in the results. The manufacturer’s formula
showed a lower efficiency in the winter than in the summer, while the equations from
method two said the exact opposite. In fact, the heating season efficiency from method
two is twice that of method one. However, the cooling season efficiency is nearly
identical in both methods (0.40 versus 0.39). The cooling season represented the period
of higher insolation. As was seen earlier in this section, the temperature rise in the tubes
was also more consistent under high insolation values. This begins to suggest that the
behavior of the tubes at low insolation values is unreliable and should not be used for
design calculations.
The third and final calculation method again used Equations 2 and 3, but this time
calculated the instantaneous efficiency for every applicable time stamp. Initial results
were inconsistent with the previous two methods. The cooling season data showed an
average collector efficiency of 0.64 with a complete roof efficiency of 0.43. The heating
season was even more extreme, with an average collector efficiency of 1.19 and a rooftop
efficiency of 0.79. Based on the first two methods, it was thought that maybe low
insolation values were causing the large discrepancies. To test this, the data for method

100

three were filtered so that only insolation values greater than 300 W/m
2
were considered.
This brought the results down considerably and put them right in line with the previous
calculations. Cooling season efficiencies were 0.38 and 0.25 for the collectors and
complete roof, respectively. For the heating season, values were also 0.38 and 0.25. The
results are summarized in Table 5-7.

Method
Insolation
Conditions
Cooling Season Heating Season
Collector
Efficiency
Whole Roof
Efficiency
Collector
Efficiency
Whole Roof
Efficiency
1 All 0.398 N/A 0.273 N/A
2 All 0.389 0.260 0.531 0.355
2 Greater than 300 0.340 0.234 0.350 0.234
3 All 0.637 0.426 1.186 0.793
3 Greater than 300 0.375 0.250 0.375 0.251

TABLE 5-7: A summary of the data for all three methods of calculating the collector efficiency.
Highlighted numbers are those that show the low-insolation vulnerability of the chosen method.

Table 5-7 shows some interesting conclusions. Heating season calculations, which
saw lower average insolation than the late summer months, varied widely when using
Equations 2 and 3. Cooling season data were better, but still showed a lot of variance
when all insolation values were included in the calculation. After focusing on only
insolation values greater than 300 W/m
2
, the data from Equations 2 and 3 were similar to
Equation 1. The results seem to affirm the manufacturer provided formula, making the
initial skepticism appear unfounded. The negative efficiencies at low insolation might
indeed be accurate, as more heat might be lost to the surroundings than is absorbed under
these conditions. The manufacturer equation accounts for all insolation values, while the

101

HVAC formulas used certainly seem to breakdown at low insolation values. Design
conclusions for collector efficiency are presented in the Chapter 7, taking into
consideration all the values from Table 5-7.

Absorption Chiller
Only the cooling season data are relevant for analysis of the absorption chiller.
This section discusses some of the inputs and outputs of the chiller, but does not consider
any of the inner workings, considering the chiller as a black box component that simply
outputs chilled water when supplied with hot water. Manufacturer specification sheets for
the chiller, when combined with the observed data, help to analyze the performance of
the machine.
The documentation for the absorption chiller lists the nominal capacity as 10 tons
(120,000 btu/h). It also lists the heat input necessary to achieve this cooling capacity:
170,000 btu/h. This gives a quick idea of the Coefficient of Performance (COP) for the
chiller.
C.O.P. = 120,000/170,000 = 0.706
However, there are more nameplate conditions for the chiller. The nominal operating
capacity assumes a hot water temperature of 190F with cooling water at 85F. Under these
conditions, chilled water at 48F is output. A performance curve from the chiller
documentation is shown in Figure 5.10. The standard conditions are identified with a
black dot.

102

Figure 5.10: Performance curve for Yazaki WFC-10 absorption chiller.

The data were evaluated to determine exactly what type of performance the chiller
was showing. First, the cooling season master sheet was filtered to show only timestamps
where the generator pump and building pump were active. This would indicate that the
chiller was operational and was pumping chilled water into the building. In the design

103

documents, the system designer specified that the minimum temperature necessary before
activating the chiller was to be 175F. However, during research visits it was discovered
this had been reset to 158F, the absolute minimum at which the chiller could produce
cooling, according to the manufacturer. Audubon staff said this change was made
because the chiller was kicking on too late in the day; a lower start up temperature would
allow cooling during more of the occupied hours. According to the data, the average
temperature of the heat medium inlet during operation was 164F. This value is essentially
represented by the far left vertical line in Figure 5.10. Choosing the correct performance
curve is unfortunately prone to error, as the temperature sensor for cooling water supply
temperatures was not operational during the term of the study. However, the cooling
tower itself was fully functional and was routinely observed in operation during site
visits. Therefore, it was assumed that the cooling tower supplied cooling water at the
nominal 85 degrees required by the chiller. This suggests that the chiller was operating at
a capacity factor near 0.3, or 3 tons, on average. Additionally, the C.O.P. under these
conditions is about 0.35, about half of the predicted value under standard conditions. The
delivered cooling, resulting calculated C.O.P., and inlet temperature profile are discussed
later in the chapter.

Hot Water Storage Tank
Both the cooling and heating systems are driven by hot water; this makes the hot
water storage tank perhaps the most important element in the system. As the energy
storage mechanism, it is analogous to a gas tank, electrical grid, or a battery bank. This

104

section focuses on this simple yet vital element. It primarily analyzes the temperature of
the water in the storage tank while investigating the factors that are controlling it. Factors
relevant to the cooling season are presented first, followed by a look at the system as a
heater.
As mentioned in the last section, the absorption chiller needs 191F water to
operate at a 1.0 capacity factor. Of the 5,519 data points in the cooling season master
sheet, this condition appeared in only 57, just over 1% of the time. A deeper look at the
data revealed that all 57 of these points occurred on two days, October 16 and 17.
October 16 was one of the cooling-to-heating changeover dates that was ultimately
abandoned. However, a review of the valve status showed that an operation mistake was
made that day. The computer, responding to the seasonal schedule, permanently put
Valve V1 into the heating position mode. Valve V2 should also have been switched, but
this valve is manually operated. October 16 saw a high temperature of 101F, so it is
understandable that no employee considered turning on heating mode that day. However,
the combination of one manual valve and one electronic valve put this system into a non-
functional and potentially dangerous mode where pumps would be attempting move
water through closed valves. Data for October 16 and 17 show many odd events, such as
a large increase in the temperature of the storage tank even though the solar collector
pump was off. It is not clear from the data exactly what was happening or where water
flows were taking place. At one point, the water in the storage tank topped 200F,
triggering an alarm that notified staff of the problem. By 4:15 pm on October 17 the
problem was fixed and the schedule reset to put the building in cooling mode. The

105

storage tank temperature never topped 191F again. Though the exact sequence of events
remains unknown, it is clear that 190F+ water is not a regular occurrence in the system,
and the 1% of data points over this value are either erroneous or reflective of a system
malfunction.

Figure 5.11: A valve problem on the scheduled cooling to heating changeover date caused dangerously
high storage tank temperatures. Valve 1 was reset on October 17 at 4:15, after which the temperature in the
storage tank returned to normal levels. The highlighted boxes show some (not all) of the high storage
temperatures.

For the absorption chiller section, the original set point for chiller operation was
storage water greater than or equal to 175F. Analysis of the data showed this condition
existed 5% of the time. However, one-fifth of the data points satisfying this condition are
from the questionable October 16-17 data, and it not known if the water would have
made it to the rated temperature without the identified malfunction. The lack of cooling

106

reported by the Audubon employees seems like it was an accurate complaint. The
building certainly required cooling more than 5% of the time, making the change from
the original design condition appear valid. Of course, the consequence of dropping the
temperature requirement for chiller operation is reduced cooling capacity, as shown in
Figure 5.10. Thermal comfort data presented earlier in this chapter suggests that the
chiller was effective in maintaining comfort during cooling season, so it seems the
cooling capacity, even when reduced, met the loads. The storage tank temperature met
the 158F requirement 37% of the time. A quick analysis of the outdoor temperatures was
performed to determine the percentage of time that cooling was needed. A count was
performed of outdoor temperature data points greater than 71F, the thermostat setpoint in
the building. This condition occurred in 29% of the data points. The chiller was able to
provide some level of cooling more often than it was needed.
Another comparison is to look at the storage tank temperature versus the time of
day. This gives an idea of both the system’s ability to function during operating hours, as
well as the possibility of providing cooling into the evening, which is common in tropical
climates. The data were filtered to show only timestamps where the storage tank
temperature was greater than 158F. There were 2,009 data points meeting this condition.
A count showed that 424 of the points, or 21%, occurred between midnight and 8:00am.
A second count showed that 360 points (18%) occurred between 6:00pm and midnight. In
all, 39% of the data points occurred in unoccupied hours, meaning potential exists for
commercial/industrial buildings with multiple shifts or for residential buildings. One test
looked at if these points were occurring just outside of occupied hours, or well into the

107

night. To do test this, a count was made of data points where the storage temperature was
greater than 158F and the insolation reading was 0 W/m
2
. Of the 784 off-hour data
points, 708 occurred in times of darkness.
The previous step gave the impression that nighttime operation of a solar
absorption chiller was sometimes possible. There were 20 individual days that
experienced a storage tank temperature hot enough to run the chiller after sundown. The
full dataset was filtered to show only these days. Was there a unique set of conditions on
those days that led to the availability of nighttime cooling? In an attempt to identify those
conditions, the data were further filtered to show only daytime climate values. This was
accomplished by simply removing all timestamps with an insolation value of zero.
Analysis showed that the average daytime temperature and insolation on these 20 special
days was 70.7F and 346 W/m
2
. The filter was then reset to show daytime climate data for
all the other days available in the cooling season master set. For these days, the average
temperature was 76.9F and average insolation was 368 W/m
2
. The insolation values are
essentially equal. However, the days where nighttime cooling was available were on
average 6.2 degrees cooler than those without the option. The nighttime cooling option
seems to have arisen on cooler days where the cooling load was reduced, meaning the hot
water tank was likely untapped during the day, and thus able to retain its heat well into
the evening. However, these cool days also tend to precede cool nights, so the nighttime
cooling is unlikely to be needed. It appears that despite the original suggestion of the
data, providing cooling after sundown is a limited option that would work only if the
cooling demand followed a mild, cool day. One situation that might cause this is an

108

evening event that will introduce large occupancy loads into the space. However, with no
energy available to recharge the system, the amount of time that cooling could be
provided would be limited.
A last point of analysis for the storage tank is an evaluation of the time delay
between peak storage temperature and the incident radiation. To do this, the data set was
filtered to show only the insolation and the storage tank temperature. Insolation values of
zero were then removed, so only daytime conditions were present. The first plot, shown
in Figure 5.12, is for a single, very hot and clear day.

Figure 5.12: The insolation and storage tank temperature on September 24, 2009, a hot and clear day. A
cloud the rolled in around 2:30pm is evident.

135.00
140.00
145.00
150.00
155.00
160.00
165.00
170.00
175.00
0
100
200
300
400
500
600
700
800
900
Temperature (F)
Insolation (W/m
2
)
Solar
Radiation

109

In Figure 5.12, there is an obvious time lag between peak insolation and the storage tank
temperature. This is expected, as it takes an amount of time to heat such a volume of
water. Peak insolation occurs at about 1:00. Peak storage temperature appears to occur at
approximately 3:45.
Figure 5.13 shows a similar chart, but for a different extreme. October 15, 2009
was a very hot and clear day; however it followed two rainy, overcast days.

Figure 5.13: The insolation and storage tank temperature on October 15, 2009, a hot and clear day that
followed two rainy, overcast days.

In this case, the storage tank temperature again peaked about 2.5 hours after peak
insolation. However, there is a significant difference in the starting temperature of the
water. Figure 5.14 shows the two temperature plots on the same graph. On September 24,
130.00
135.00
140.00
145.00
150.00
155.00
160.00
165.00
170.00
175.00
180.00
0
100
200
300
400
500
600
700
800
7:30
8:00
8:30
9:00
9:30
10:…
10:…
11:…
11:…
12:…
12:…
13:…
13:…
14:…
14:…
15:…
15:…
16:…
16:…
17:…
17:…
18:…
Temperature (F)
Insolation (W/m
2
)
Solar
Radiation

110

the tank was hot enough to run the chiller at 11:30am. On October 15, no cooling was
possible until approximately 1:45pm. During the overcast days when little energy was
available to replenish the tank, the temperature in the tank dropped to 134F. It is unlikely
that mechanical cooling was needed on those days however, as low temperatures were
present along with the cloudy weather. The following day, exterior temperatures topped
75F by noon, meaning there was about a two hour window where cooling could have
been used but was not available. This effect is small, especially in Southern California
where overcast weather rarely appears in the cooling season. In fact, the October 13-14
period was the only rain on record for the measured cooling season. The effect is much
more significant in the heating season, when peak daytime heating loads occur at the
same time as cold and cloudy weather. This is seen in Figure 5.15, which plots the
storage tank temperature for four consecutive rainy days in the heating season. The
average daytime insolation for these four days was only 48W/m
2
, indicative of extremely
overcast weather. The figure shows that there is significant heat loss in the tank for the
duration of the bad weather. The steep decline was caused by the building pump
circulating the water into the building. The more level areas are heat loss due only to
conduction. This presents a major problem when using the solar thermal system for space
heat without a backup chiller. The average temperature in the office space during this
period was 66F.

111

Figure 5.14: The storage tank temperature on September 24, 2009, the third of three consecutive hot clear
days, and October 15, 2009, a hot and clear day that followed two rainy days. The gray boxes are
representative of the time frame for which mechanical cooling was available.

112

Figure 5.15: The storage tank temperature on January 17-January 20, 2010, four consecutive overcast and
rainy days.

For heating season, the water in the storage tank directly drives the hydronic
system, so its temperature is the only value of importance. The original design documents
specified that the minimum required temperature for heating was 90F. Solar thermal
energy seemed up to the task, as storage tank was above this threshold in 91.6% of the
heating season data. Looking further, the average tank temperature for the entire
measured heating period was 122.7F, well above the 90F threshold, suggesting that the
heating demand should have been met based on the design parameters. However, earlier
in this chapter the data indicated the space was too cool more 40% of the time. The
building pump was active for nearly half of the analyzed hours, indicating that hot water
was indeed flowing to the fan coil units and heating was being delivered. Clearly there
80.0
90.0
100.0
110.0
120.0
130.0
140.0
14:45
1/17/2010
1:00
1/18/2010
16:45
1/18/2010
12:45
1/20/2010
Temperature (F)
Storage Tank Temp - Rainy Winter Days
S04 HW
Storage
Tank
Temp

113

was an issue somewhere in the system. If the storage water was well above the design
point temperature and the building pump was working, why was the space so cold such a
significant portion of the time? To try and identify the problem, the interior temperature
in the office was plotted against the temperature of the supply water into the hydronic
system. A filter was first applied to limit the data to timestamps where the building pump
was active, meaning hot water was flowing to the fan coil units. This is shown in Figure
5.16.

Figure 5.16: Temperatures in the office increase with increased temperature in the storage tank, which
supplies the heating water. The black line is drawn at the room setpoint of 68F.

There is a clear correlation between increasing room temperatures and a higher
temperature in the storage tank. This was further analyzed by performing a count of

114

points falling in the comfort zone for a given threshold temperature in the storage tank.
Of the 1,623 data points in the study, 1,089 fall in the comfort zone (above the black line
in the figure). Since this is data for an active pump only, this tells us that two-thirds of the
time, the space was adequately heated by the solar thermal system. The hydronic heating
system was working to some extent. However, of those 1,089 satisfactory points, only 29
occurred with supply water below 100F. Thus, it is clear that water between 90F (the
minimum to activate the pumps) and 100F was not effective in heating the space. Perhaps
120F is a better setpoint. In Figure 5.16, 52.7% of the data points fall above 120F. For
these points, the corresponding interior temperature is in the comfort zone 91.5% of the
time. It appears that mandating a higher temperature (about 120F) for pump operation
would keep the space adequately warm.

Delivered Cooling
The final element of the performance analysis is an estimate of the delivered
cooling. The calculation was made using Equation 2, where F is the flow rate in the
building primary loop and T
i
and T
o
represent the supply and return water temperature
respectively. The cooling season data were first filtered to show only timestamps where
the building pump was active. The formula was then applied to give a rough estimate of
the cooling delivered. The average was 7.3 tons. This is higher than would be expected
considering the performance characteristics of the chiller described earlier in the chapter.
It is likely that this number is high because the formula includes all heat loss in the water
loop, including conductions losses outside of the building and after it has moved through

115

the fan coil units. Temperature sensors on each of the fan coil units would provide better
data for Equation 2. It was worth investigating if the cooling delivered increased with
increasing storage tank temperature, as would be predicted by the data sheet for the
chiller. The results are shown in Figure 5.17.

Figure 5.17: The amount of cooling delivered, calculated by Equation 2, versus the temperature supplied by
the hot water storage tank.

The figure does show an increasing trend, however the data points are somewhat
scattered. A best fit line, not forced to an intercept seems to confirm the trend. Still, the
data are not tight to the line, indicating that the measurement scheme for delivered
cooling likely needs revised to create a more detailed and consistent data set.
155.00
160.00
165.00
170.00
175.00
180.00
185.00
190.00
195.00
200.00
205.00
0 2 4 6 8 10 12 14
Temeprature (F)
Tons of Cooling
Delivered Cooling versus Heat Medium Temperature

116

The wealth of data available provided for many analysis options. The analysis and
figures shown in this chapter represent only a small portion of what could be examined.
The parameters evaluated above were chosen because they are beneficial for drawing
conclusions about the performance of the system. These conclusions are outlined in
Chapter 7. Much of the future work possible is discussed in Chapter 8. The next chapter
will examine the system from an economic perspective, with special attention paid to
cost, architectural implications, and environmental impact.

117

Chapter Six: Economic Analysis

This chapter evaluates the solar thermal HVAC system from an economic and
environmental standpoint. It also touches on the architectural implications related to
design decisions. The system installed at the Audubon is compared against other
available design options that could also provide thermal comfort while being energy
efficient. The economic analysis primarily involves costs, looking at both the upfront cost
and payback period, as well as incentives meant to facilitate a stronger return on
investment. The environmental analysis concerns the ecological impact of different
design schemes. The architectural implication primarily concerns the footprint required
for different systems. Aesthetic concerns that might arise are also discussed. Lastly, a
discussion on comparing efficiencies for different systems is resented.

Economics
Grid-tied versus off-grid
When considering the cost of a solar thermal HVAC system, one must first
consider the autonomy of the building. One of the driving factors in using the system at
the Audubon was the fact that the building would be off-grid. Ultimately, the building
was connected to city water, but is not connected to the electric, gas, or sewer utility.
Although in one of the most populated cities in the world, the Audubon Center sits in a

118

relatively remote location in an undeveloped park. The nearest utilities are over one-
quarter of a mile away. Design documents indicate the initial cost estimate for an
electrical connection was $75,000. This premium must be considered when comparing
the system at the Audubon against grid-tied options. The decision to go off-grid usually
increases project cost, but it is sometimes a project requirement for leading
environmentally sensitive buildings.

A solar thermal system
In his initial bid, the designer of the HVAC system agreed to supply, install, and
design the system for $94,000. This included the solar thermal collectors, the chiller, the
storage tank, the cooling tower, and the piping and valves connecting those components.
It did not account for the fan coil units or the ducting inside the building
5

5
Since the FCUs and internal ductwork would be common to all active cooling systems discussed in this
chapter, their cost is not considered in the economic comparison.
. The system
also requires electricity to power the pumps that circulate the water through the various
loops as well as a fan for the cooling tower. Quantifying the electrical usage is tricky. In
the winter, two pumps are never active, as discussed in Chapter 3. In the cooling season,
a worst case scenario for electrical use would involve all four pumps and the cooling
tower fan being active simultaneously and at full power. This is possible, if not common,
and is thus considered as the maximum electrical draw for the solar thermal system.
There are multiple sources reporting slightly different power draws. An article in Energy
Pulse (Wright 1) lists the draw as 0.4kW per ton. At 7.5 tons (the design load), this

119

equates to 3kW. The design documentation lists the rated power of the four pumps and
the fan motor:

Item Manufacturer and Model No. Motor HP
M1 Marley 492A 1
P1 Grundfos UPS 32-80 1/3
P2 Grundfos UPS 32-80 1/3
P3 Grundfos UPS 40-80 3/4
P4 Gould 3642 1

TABLE 6-1: Motor and pump power ratings for the solar thermal system at the Audubon Center

The total horsepower is 3.42hp. At 745.7 Watts per hp, this equates to 2.55kW. Also
available was the electrical distribution panel design sheet developed by the project
electrical engineer. This listed the breaker rating for each of the components. The total
was 2kW. This was lower than the rated values of the pumps, though the electrical sheet
had an earlier date stamp than the pump schedule, meaning the pump size might have
changed after the electrical sheet was drafted. Still, there is decent agreement between the
three electrical power calculations. A value of 2.75kW was decided upon for this part of
the study. Were the building grid tied, this value would be handled by the electrical grid.
Because it is completely off grid, this load has to be handled by solar panels and batteries.
Calculating the number of solar panels needed to meet the demands of the solar
thermal system requires a couple of steps. The electricity used by the pumps and fans is

120

AC energy. Solar panels produce DC energy and then rely on an inverter to create a high-
quality AC signal that can power the loads. This process is not 100% efficient –
equipment and wiring losses take away a portion of the power (the exact drop is different
in every installation depending on the specific conditions – 10% will be used here).
Therefore, the 2.75kW load would require about 3kW worth of DC power. In the solar
power industry, battery backed systems are often quoted at about $10-12 dollars per DC
Watt. This price includes the panels, mounting, batteries, and all electrical transfer
equipment such as charge controllers and inverters. Using $11/Watt for this estimate, this
adds an additional upfront cost of $33,000 ($11 x 3kW). Adding this to the original quote
equates to $127,000 for the complete system. This is incredibly expensive for a system of
this size in terms of upfront cost. However, the investment essentially represents
prepaying for energy, as there will be no further energy cost for the life of the system.

A photovoltaic system
The alternative to solar thermal would be to use a high efficiency electrical
compression chiller. This strategy is becoming more popular in energy efficient designs,
as there are undeniable benefits which are described below. For a grid-tied building, it
requires the smallest initial cost, with the efficient equipment minimizing electric use due
to an impressive coefficient of performance (COP). A quick internet search turns up 4-ton
heat pumps with a seasonal energy efficiency rating (SEER) of 16 for under $4,000.
Consider two of these placed in parallel to meet the peak demand of the Audubon. For
the building to remain autonomous, consider powering the units with photovoltaic panels

121

with battery backup
6

. A SEER of 16 equates to a COP of 4.7. This means to meet the 7.5
ton (90,000 btu/h) demand, 19,149 btu/h of power is required. This converts to 5.6kW.
Again using $11/W, an all electric system would cost approximately $74,000 ($66,000
for the PV infrastructure and $8,000 for the heat pumps). This represents a savings of
over $50,000 versus the solar thermal system while still maintaining the idea of prepaid
energy with no energy bill for the life of the system. In addition, the PV panels still
produce power when cooling and heating are not needed, and that power can go to other
electrical loads such as receptacles. Extraneous hot water produced by the solar thermal
system can be stored, but cannot be used to power building loads other than HVAC (with
the exception of domestic hot water in some systems).
Incentives and adjusted initial cost
Another economic factor relates to incentives, rebates, and buy-downs for the two
technologies. On a federal level, both technologies are eligible for a renewable energy tax
credit equal to 30% of the installed cost. However, solar thermal technology seems to lag
photovoltaics in terms of local government and utility subsidies and incentives. For
example, in Florida the 2006 Florida Energy Act set aside state money for rebates for
renewable energy installations. The residential rebate for a solar thermal domestic hot
water system was $500. For photovoltaics, the program promised $4 per nominal Watt,
up to a maximum of $20,000. Under this scheme, a photovoltaic system has a much

6
This is certainly not the only way to produce energy on-site; wind power or a gas fired microturbine could
be used as well. However, since the existing building has a 26kW photovolatic array for its non-HVAC
loads, solar will be examined.

122

quicker return on investment than a solar thermal installation. Businesses were offered a
per unit rebate for solar thermal of $15/btu with a $5,000 cap. The maximum business
rebate for photovoltaics was $100,000, again at $4 per DC Watt. This discrepancy makes
it difficult for solar thermal technology to catch on. The cause of the discrepancy is very
likely a lack of knowledge about its potential. While photovoltaics installations are
growing rapidly in Florida, only one example of a solar thermally powered air
conditioner was found in the Sunshine State. The difference in the payback period is
significant. Consider the numbers calculated earlier for the Audubon center:

• As a business, the center would be eligible for a rebate on every watt of installed
solar power (as it is well under the $100,000 cap). This equates to an instant
savings of 5600W x $4/W = $22,400. The federal tax credit removes 30% of the
predicted $80,000 up front cost. Therefore the adjusted total is $80,000 (initial
cost) - $24,000 (federal) - $22,400 (state) = $33,600.
• For solar thermal, the Audubon would be eligible for the $5,000 solar thermal
rebate (it quickly tops out), the $4/W rebate for the solar power dedicated to the
pumps, and the 30% federal tax credit. The true total would be $127,000 (initial
cost) – $38,100 (federal) – $12,000 (state solar) – $5,000 (state thermal) =
$71,900.

After incentives and rebates, the solar thermal system costs more than double the
photovoltaic system driving electrical compressors. Until this gap is reduced, solar

123

thermal heating and cooling will remain a “tough sell.” The payback period for the
photovoltaic system would be less than half of that of the thermal system

Environment
From an environmental standpoint, solar thermal may offer the more ecologically
sensitive option. This is primarily due to the reduction in the number of batteries
necessary. Batteries represent a danger to the environment due to the large number of
chemicals they contain. This presents a danger should the batteries leak due to damage
and/or age. In addition, the batteries have an understood life span that varies based on the
level of use and the number of charge/discharge cycles. Beyond a certain threshold, the
batteries ultimately require disposal. Because of the chemical content, batteries should
not be landfilled, though this is often ignored. Photovoltaics also decline over time, and
with current construction methods cannot be recycled, though the estimated life of the
panels does exceed 20 years. On the other hand, for a solar thermal system, the primary
means of energy storage is water, which presents no to harm the environment. The
storage tank can be made of stainless steel which is recyclable. The collectors are made
primarily of copper and glass, both of which can be recycled. While some batteries are
required for the pumps (assuming the system is off-grid), the number is smaller than for
an all electric system. Also, the reduced pump use in the winter will prolong battery life.

124

Architectural Implications
An additional concern is one of aesthetics and architectural implications. Much of
the green sector work in the last few years has focused on better aesthetic integration of
energy efficient technologies into buildings. Chapter 3 of this study showed the solar
collector covered rooftop at the Audubon center. While it is a point of pride for the very
environmentally progressive organization, it would be easy to make the argument that
collectors are an unattractive building feature. The rather crooked installation of the
piping also adds to this point. The architect must also consider the footprint of the various
pieces of equipment necessary to run an off-grid HVAC system. As mentioned earlier,
the solar thermal collectors take up just over 1,000 ft
2
of roof area. However, one must
again consider the photovoltaics necessary to run the pumps and fan. An average solar
panel measures about 64” long and 39” wide (17ft
2
) and outputs roughly 210 W. To meet
the 3kW demand of the solar thermal system, 15 solar panels would be needed, adding an
additional 255 ft
2
of roof area needed to run the system. Using an electric chiller powered
by PV requires 5,600W / 210 = 27 solar panels. This equates to about 460 ft
2
of roof area.
For the architect, the difference is significant in terms of building design. By these
calculations, the roof area needed for an all-electric system is only 35% of what is needed
for a solar thermal system. The equipment footprint on the ground is also bigger when
choosing to go with solar thermal. The thermal system requires a hot water storage tank
and a cooling tower, two pieces of equipment that are not necessary with and air-to-air or
water-to-air heat pump for this size building. With a constrained buildable area, the

125

architect might be swayed to go with an electrical compressor to reduce the space
necessary for unsightly mechanical equipment. The ground based equipment footprint for
an off-grid PV based system is essentially limited to that required for the heat pump(s)
and the battery bank. Because the batteries can be stacked vertically, a battery shed
ultimately takes on a smaller footprint than the cooling tower and storage tank, both of
which require their own concrete mounting pad.

Efficiency
A final point of comparison between a solar thermally driven system and one
powered by photovoltaics is that of the COP. For heat pumps, this number is calculated
by dividing the cooling power delivered, in Watts, by the electrical power required to
produce that cooling. For the high efficiency heat pump discussed earlier, the COP was
determined to be 4.7 (COP = SEER divided by 3.41). In Chapter 5, the COP of the
absorption chiller was shown to be 0.35. However, this is not an equal comparison. The
chiller COP was also a quantity describing energy output over energy input, however this
was entirely related to thermal energy, not electrical energy. As discussed earlier in the
chapter, the maximum electrical energy drawn by the solar thermal system was 2.75kW.
The cooling energy delivered, under this peak condition, is 7.5 tons, which converts to
26.4kW. Diving the cooling delivered by the electricity used, like with the heat pump,
produces a COP of 9.6. This comparison is a more “apples-to-apples” representation of

126

COP for the two systems. On this premise, it can be argued that the solar thermal system
is ultimately twice as efficient as the electric compression system. The basis of this
argument is that electrical energy is what should be counted, as it comes at a price and is
not renewable if coming from a fossil fuel power plant. Solar thermal energy is “free,”
has no carbon footprint, and thus does not need to be considered in a true COP
comparison. However, this assumes the electricity in the equation is coming from the
utility grid. For a system driven entirely by PV, the same argument can be made
regarding “free” energy (this time in the form of sunlight).
It is clear from this chapter that solar thermal HVAC schemes often come up short
in comparison to PV driven schemes. However, as the cost of solar thermal equipment
comes down, the gap is shrinking. When combined with domestic hot water, a solar
thermal “combisystem” sees increased efficiency versus just a hydronic HVAC approach.
With increased incentives and buy-downs, solar thermal could see increased attention in
the energy efficiency market. Further conclusions relating to this are presented in the next
chapter, which combines the economics of the design with the performance conclusions
developed in Chapter 5.

127

Chapter Seven: Conclusions and Design Guidelines

This chapter presents conclusions that would be valuable to an architect or
engineer considering a solar thermal HVAC scheme. The conclusions are in no way
exhaustive, representing only a portion of the design considerations necessary for a fully
operational system. However, they are intended to roughly guide the design of future
systems, making use of the observations from the Audubon Center to provide conclusions
relevant to system sizing, performance expectations, and cost analysis. The analysis
follows the data structure of Chapters 5 and 6, offering conclusions in the same order as
the underlying data were presented.

A note on climate
It is first worth noting that the climate in Southern California does not
traditionally present a challenge for space conditioning. Figures 5.1-5.3 show the
relatively low (and equal) heating and cooling demand predicted by historic data. The
psychrometric chart in Figure 5.3 is especially telling. A large number of points can be
found very close to the comfort zone, indicating that passive strategies would be very
useful for a building in this locale. Indeed, the Audubon Center is designed to take
advantage of these passive strategies, which has a significant effect on the HVAC
strategy of the building. The design idea of the building team was to maximize passive

128

strategies as the primary thermal comfort scheme, then use solar thermal to meet peaks
and extremes which extended beyond the window for passive design. The mild climate
combined with the small, well-built structure means this “peak shaving” region is small
enough to rely on the solar thermal system alone for delivering thermal comfort. For
buildings in heating or cooling dominated climates, or buildings with much larger floor
and facade areas, this strategy would be a dangerous one. At the Audubon, situations
where the thermal system fails are minimized by the limited loads experienced. This is an
exception to the common experience for the HVAC designer. For most situations, a solar
thermal system should be installed with a reliable backup to guarantee the comfort of
occupants. In fact, in the course of the research for this study, no other building was
found that was using solar thermal for 100% of its HVAC loads without a backup system
in place. The solar thermal system is often paired with a traditional system. The two are
either designed to always operate in tandem, thus reducing the load on the traditional
equipment, or with the traditional equipment kicking in only when the building load
extends beyond the capacity of the solar thermal setup. One could of course continue to
increase the size of the chiller and solar collector array until it was large enough to meet
all expected loads. However, this is obviously a limited option, as project budget and roof
area can quickly limit the size and space available for solar thermal equipment. In
addition, while the sun is sure to rise everyday, there is no guarantee of how much
sunlight will fall on the collectors. Cloudy and overcast conditions occurring in
conjunction with large building loads (for example, cold gray weather in the Northeast)
will render a system with no backup useless no matter the system size. A second common

129

backup approach is the use of a backup heat source to directly heat the water. For this
type of installation, the roof area available is generally suitable for enough collectors to
meet the system loads in a majority of cases. In low insolation events where system
temperatures drop, a backup heater (usually gas, but can be electric) kicks in to assist in
getting the water up to temperature
7

. As this chapter progresses it will become clear that a
backup system, either a supplemental heater or a traditional HVAC system, would have
been useful even in the mild climate at the Audubon center. The first conclusion for this
type of HVAC approach is to always install a backup system of some sort. The mild
climate in Southern California can be deceptive, with the design of this particular system
creating the impression that a solar thermal system can always do the job. No matter the
location, the climate is simply not predictable enough to design a system entirely
dependent on naturally available energy. In fact, local codes might prevent the 100%
thermal approach entirely, requiring backup as a guarantee of life safety.

7
One advantage to using a gas heater is that in the summer, peak electricity use can be avoided. Some
utilities charge a premium for electricity use during peak demand periods, and these are usually late
afternoon during a heat event. The same utility might offer gas at a significant discount in the summer just
to have a customer in a time of the year when no one is heating their building. Where to place the backup
heater is another issue that requires careful consideration. See Henning, H. Solar-Assisted Air-Conditioning
in Buildings.

130

System performance
A breakdown of the cooling season data shows that the building stayed quite
comfortable in the cooling season, particularly in the office spaces. The other two spaces
did appear to be overcooled at times, but this was acceptable considering the low set-
point temperature and the lack of internal loads or occupants in these spaces. The same
cannot be said for the heating season. Downright chilly conditions existed in the spaces
for a significant portion of the winter months even though the heating system was active
nearly half the time. This highlights what is essentially the great paradox in designing a
system to provide both heating and cooling from solar thermal energy. In the summer,
peak cooling loads tend to coincide with peak solar thermal energy availability. Put
simply, the hottest days offer the most solar energy while simultaneously creating the
most demand for space cooling. There is an inherent synergy in that relationship; as
outdoor temperatures continue to increase, hotter water is produced, thus increasing the
cooling capacity of the chiller. The loads and the energy source are said to be in phase.
The relationship in the heating season is opposite. The largest heating loads tend to occur
in times of low solar availability. This isn't always the case – many places experience
very clear weather during cold snaps – but it is generally true that in temperate climates
the coldest days are closely tied to those that are overcast or otherwise have limited solar
availability. This is especially true in Southern California. Consider the differences this
leads to in design. If the system was only intended to be a chiller, there is less emphasis
on thermal storage, as the thermal energy is needed rather immediately. In fact, too large

131

a storage reservoir can be detrimental to the system. The high thermal capacity of water
makes it a great heat storage medium, but consequently it takes a significant amount of
time to heat up. Consider that in the cooling season, insolation peaks between noon and
1pm with peak cooling demand generally occurring between 3 and 4pm. If the storage
tank is too large, the tank will take too long to heat up, and the chiller won't be able to run
at maximum capacity during times of peak load. The time lag between energy availability
and putting the energy to use is small. The situation is different for heating. Since times
of high insolation usually don't occur along with a large heating load, it is important to
store the thermal energy for use at another time. Here, the time lag is large, sometimes on
the order of days (consider consecutive cold gray days; the bulk of energy to provide heat
will have come from the last clear day). Generally the choice for the designer will depend
on the climate profile. In Riyadh, certainly the system would be biased towards cooling.
In Minnesota, heating would take precedence. In California Climate Zone 9, there is no
clear winner. What the data show is that the system was designed with a bias towards the
cooling season. It works very well in the mid-to-late summer months, gets through the
relatively mild fall shoulder season, but ultimately breaks down when winter conditions
take hold. The simple solution to solve the seasonal paradox goes back to the installation
of a backup system. In the case of the Audubon, it is understandable that the system is
biased towards cooling. The absorption chiller certainly represents the more cutting edge
and high tech environmental design strategy, and perhaps for this progressive building it
was imperative that the chiller work without too much input from non-renewable sources.
This decision then limits the ability of the system to properly heat the building. A

132

supplementary heater of gas fired boiler would certainly have done the trick. However,
this is an additional issue for an off-grid building such as the Audubon Center. With
connection to the utility gas line not possible, the only options for backup heat are an
electric heat pump, an electric supplemental heater, or a propane fired supplemental
heater. The electric options require a significant addition to the building's photovoltaic
array, which carries with it high additional upfront costs and the need for additional
mounting area. To use propane, a tank would have to be installed and periodically filled,
violating the carbon neutral spirit of an off-grid building. In fact, it could easily be argued
that the addition of a propane tank takes away the “off-grid” status of the building. There
are no easy answers for a building with such an equal distribution of heating and cooling
demand. In a climate with a strong bias in one direction, it is possible the system could
handle the minority situation without backup resources, but ultimately unlikely. It raises
doubt as to whether an autonomous year-round solar thermal comfort strategy is truly
viable in places with distinct heating and cooling seasons. It could perhaps work in a
place like the tropics, where the lack of a heating season alters the definition of “year-
round” thermal comfort; in this location a chiller-only strategy could work.

Adaptive Comfort
Use of the adaptive comfort model seems plausible for a solar thermally driven
system. The employees at the Audubon seemed to possess a relaxed standard for thermal

133

comfort similar to what is described in the theoretical adaptive comfort scenario. Human
testing was not carried out, so this conclusion is entirely anecdotal. However, it was clear
that in the heating season the employees were content to come to work in an extra layer
or two, essentially expanding the comfort zone through an increase in clo value. While
there was acknowledgment that the space was cold, it seemed to be well understood that
this was a result of the cold exterior temperature and limited resources of the “natural”
system which provided heat. There did not appear to be significant complaints despite the
space being at a temperature that would correlate with a high PPD (Percentage of People
Dissatisfied) in a traditional work environment.
The addition of a backup system probably removes the ability to use the adaptive
comfort model. At the Audubon, the occupants essentially considered their system
passive, since it was 100% reliant on solar energy. If they were aware that a backup
system was in place, it can only be assumed that the psychology behind adaptive comfort
would disappear, as the direct correlation between natural forces and interior temperature
would be removed.

System Components
Solar Thermal Collectors
Evacuated tube collectors seem to be the best choice for this type of system, as
outlined in Chapter 2. Figures 5.8 and 5.9 shows that at higher insolation values, which

134

produce more consistent results, the temperature rise is about 4-5 degrees Fahrenheit
between incoming and outgoing water. This conclusion is not easily extrapolated. It is
likely that the size of the array and they way in which the individual units are connected
have significant impact on the temperature rise. It is not possible to simply predict what
effect would have been seen if the entire Audubon array had been hooked up, say, in
series. Further study and computer modeling ideas that might produce a more solid
conclusion are described in Chapter 8.
A more reliable conclusion can be drawn from the collector efficiency findings.
The use of three separate formulas produced relatively consistent results after
adjustments were made for low insolation values. An efficiency of about 37% for the
collector surface appears in the cooling season. This drops to about 33% in the winter.
The collection surface does not occupy the entire area of the solar collector unit; there is
space between evacuated tubes as well as the area occupied by the piping and heat
exchange manifold. When considering the efficiency of the whole collector unit, the
efficiency drops slightly. It appears that the system is about 24% efficient year round
when the total area is considered. A good rule-of-thumb for future designers is 25%, a
number that is easy to remember and is supported by the data. For example, if a designer
has 50 square meters of roof available for collectors and an average insolation of 400
W/m
2
, he or she could expect 50*400*0.25 = 5000 W of thermal power to be transferred
into the storage tank when the sun is strong.
It is important that the solar thermal collectors and hot water storage tank be sited
relatively close together. If an HVAC area were hidden away in the corner of a large

135

property, much of the heat absorbed from the rooftop would be lost during transport to
the distant storage tank. This could be minimized with extensive pipe insulation, but that
adds cost and installation complexity to the project. Keeping the thermal collection and
thermal storage areas close together makes sense for any type of solar thermal
installation.

Absorption Chiller
The most important realization from the Audubon data was that the chiller did not
operate anywhere near its nominal ratings. This is a very important design consideration.
In traditional HVAC design, the chiller is sized as closely as possible to the calculated
maximum cooling load. As described in Chapter 5, the design day cooling load for the
Audubon might not have exceeded 6 tons if you consider the report from the mechanical
engineer. Yet a 10 ton chiller was chosen, a 66% percent increase over the peak load.
While this seems like a gross oversizing, it turned out to be largely appropriate due to the
constraints on chiller operation. In order to meet the nominal cooling power of the chiller,
water at 191 degrees was needed. This rarely occurred in the system. In fact, the chiller
shuts down to prevent crystallization of the desiccant when the water temperature
approaches 212F. Therefore, designing to 191F isn't advised, as a spike in insolation
could cause frequent overheat situations. Only under highly controlled heating situations
(like with a gas heater) could water be heated and held consistently at 191F. To
accommodate the reduced capacity of the chiller, the designer should pick a reasonable
and achievable storage temperature and size the chiller from there. At the Audubon, the

136

average temperature was 164F, leading to a capacity factor of 0.3. In a set up with a bias
towards cooling, solar thermal temperatures between 170 and 180 seem reasonable. The
corresponding capacity factor is between 0.4 and 0.7. Assuming that the highest water
temperatures will be in phase with the highest building loads, use 0.7 to size the chiller.
For example, consider a building with a calculated design day cooling load of 13 tons. To
meet this, we would want a chiller with a nominal capacity of 13/0.7 = 18.6 tons. A 20-
ton chiller would likely be the closest sized model.

Hot Water Storage Tank
The data presented in Chapter 5 show some potential for cooling outside of
traditional working hours, raising the possibility of using a similar system in multi-shift
buildings or in residences in climates where nighttime cooling is needed. However, a
deeper look into the data showed that high nighttime storage temperatures were generally
a result of cool days, meaning no hot water had been drawn from the tank during the
traditional cooling window. This casts significant doubt on the ability of the system to
cool beyond daylight hours. For a sub-tropical climate such as South Florida, a backup
heater would likely be needed to run the system through the night, a common occurrence
(if just for dehumidification purposes) in the muggy Southeast. However, it does show
promise for heating situations, primarily shoulder season conditions with comfortable
days and cool nights. In this scenario, the hot water accumulated during the sunny but
comfortable day is available well into the evening.

137

A second condition to look at is the time lag between peak insolation and peak
storage tank temperature. In the case of the Audubon Center, this was nearly perfect.
Figures 5.12 and 5.13 show peak storage tank temperature occurring around 3:30pm on a
hot clear day, nearly perfect for meeting peak cooling loads. Ultimately no consensus was
found either among the literature or within the Audubon data for properly sizing the
storage tank. Because the tank sits in between the energy source (the collectors) and the
energy sink (the chiller or building), it disrupts the mass flow between the two making
the heat flow model non-linear and difficult to solve with simple analysis. The answer
probably lies in computer simulation. Modern thermal analysis packages such as
TRNSYS can quickly analyze a system with a range of input variables. Depending on the
climate of the site and the respective heating and cooling needs, a computer simulation
could be performed to optimize storage tank size based on a fixed array size and known
load profile. Some rules-of-thumb do exist: a frequently occurring value is 75 liters of
storage per square meter of collector area. This value may have been the motivation
behind the Audubon Center's design. With 722 square feet of actual absorber, this
suggests a roughly 1400 gallon tank. The 1200 gallon tank in place at the Audubon would
suggest a slight bias towards cooling, also evident in the data. Still further study using
detailed thermal software is the final recommendation for determining optimal storage
tank size.
The effect of long term overcast weather is quite clear from the data. The
temperature in the storage tank can be expected to fall, though with significantly different
rates depending on the season. In the summer, heat loss from the tank will be due to

138

convection, assuming that the overcast days are cool enough to negate the need for
mechanical cooling. Thus the reduction in temperature in the tank is small, maybe 15
degrees more than what would be expected from typical nighttime conduction loss alone.
The effect of this is that on a hot day following an overcast period, the tank will take
longer than usual to heat up, as shown in Figure 5.14. This can create a small window
where the building is calling for cooling but the system cannot deliver as it is not yet up
to temperature. A backup heater or backup HVAC system quickly eliminates this
problem. It might run for just a couple of hours to restore the system to operating
conditions in line with those generally experienced in prolonged periods of clear weather.
The temperature drop is more severe in the heating season. As shown in Figure 5.15, the
tank may drop by upwards of 30 degrees as heat is drawn from the tank to meet building
loads, but no direct sunlight exists to replenish the energy. This can quickly sap the
system of its ability to heat the building, and periods of overcast weather longer than a
day will likely necessitate the need for an external energy source. Again, providing a
backup system is essential in the fight against overcast conditions. In the case of a
prolonged winter overcast scenario, the backup system might be required to meet 100%
of the building heating loads for multiple days.
Figure 5.16 shows a clear direct relationship between the temperature of the water
in the primary heating loop and the air temperature in the office space. This is rather
intuitive – a higher delivery temperature would suggest increased capacity to heat the
space. The usefulness of the figure lies in developing a better standard for the minimum
heating temperature. As was shown in Chapter 5, 120F appears to be a good start. When

139

hot water at or above this threshold was delivered to the office space, the corresponding
air temperature in the space was in the comfort zone more than 90% of the time
8

.

Economic Conclusions
The high upfront cost of a solar thermal heating and cooling system remains as a
significant barrier to entry for most buildings. What is important to understand is that the
cost represents a pre-payment for energy at the current rate. As the cost of grid energy
continues to rise, an ideal solar thermal system does not accrue additional costs. Rather,
the monthly savings chip away at the upfront cost until ultimately the system is
profitable. There are many methods for determining the payback period, and each unique
installation will differ in its economic bottom line. The standard approach is to look at the
annual electric and gas energy consumption eliminated by the installation of the thermal
system, and then value this energy in dollars. A small rate of increase matched to your
local utility’s average is added to the annual value each year. The upfront cost is
generally considered as the price of all equipment, labor, and fees minus any incentives
earned. After a number of years, the value of the savings will eclipse the upfront cost of
the system. This number of years represents the payback period. No full energy model
was performed for the Audubon, so exact payback results cannot be explicitly calculated.

8
A study presented at the 2010 ASES conference prescribed 110-115F water for space heating in
Minnesota, though with finned radiators instead of fan coil units. Because FCUs are less efficient than
hydronic radiators, this second study seems to confirm the chosen threshold for FCUs of 120F.

140

However, some average values can be presented to illustrate the calculation of payback
period. According to the CalARCH California Building Energy Reference Tool, the
average commercial building in California uses 42 kBtu/ft
2
annually
9
. The 2008
Buildings Energy Data Book claims that 30% of commercial building energy use is for
space cooling and heating. Therefore, for the roughly 5,000 ft
2
Audubon Center, we will
assume 63,000 kbtu of annual energy use for thermal comfort
10
. This converts to about
18,400 kWh per year. For a popular Southern California utility, the average rate for
electrical energy in 2010 was 18.6 cents per kWh
11

9
http://poet.lbl.gov/cal-arch/
. Recent rate increases for this utility
averaged 2.75% annually. Using these numbers, the annual energy savings at the current
rate is $3,422. In Chapter 6, the initial system cost was calculated at $72,000. With all the
necessary numbers now known, that payback period can be calculated. The result is
shown in TABLE 7-1. In this particular case, the system will pay for itself during the 17
th

year of operation.

10
42 kBtu/ft
2
·yr

* 30% * 5,000 ft
2
= 63,000 kbtu/yr

11
http://www.pge.com/tariffs/electric.shtm

141

TABLE 7-1: Payback calculation for the Audubon Center based on average thermal comfort energy use for
commercial buildings and California energy rates. The annual rate increase is 2.75%.

A 17-year payback period will certainly be a tough sell to even the most
progressive building owner. For an owner like the Audubon Society, who has a 50-year
lease on the Debs Park center, this might be acceptable, but for others it is worth looking
at other options. As shown in Chapter 6, photovoltaic can cut the payback period in half.
The best economic option appears to be a grid-tied photovoltaic system powering a high-
SEER heat pump. In this approach, the equipment cost is minimized, as an electrical heat
pump is much cheaper and requires a smaller piping infrastructure than an absorption
chiller. In addition, the heat pump requires a much smaller equipment footprint than with
a thermal system since no cooling tower or storage tank is required. By tying the system
to the grid, thermal comfort is guaranteed, as a reliable source of energy is always present
regardless of climate conditions. Because the grid functions as the storage mechanism,
Year Annual Savings Upfront Cost
72000
1 3422 68578
2 3516.11 65061.9
3 3612.8 61449.1
4 3712.15 57736.95
5 3814.23 53922.71
6 3919.13 50003.59
7 4026.9 45976.69
8 4137.64 41839.05
9 4251.43 37587.62
10 4368.34 33219.28
11 4488.47 28730.81
12 4611.9 24118.91
13 4738.73 19380.18
14 4869.05 14511.13
15 5002.94 9508.19
16 5140.52 4367.66
17 5281.89 -914.23

142

the cost of a battery bank is eliminated. Additionally, the strong incentive program for PV
drops the upfront cost. If an off-grid building is necessary due to remoteness or owner
demands, a battery backed PV system is still the cheaper option. Both the upfront cost
and the payback period are more favorable than a solar thermal system.
The economic picture for a solar thermal system is not pretty. If dealing with a
cost-conscious building owner, it will remain difficult to advocate solar thermal unless
equipment costs and rebates catch up to the rapid gains seen in the last decade for
photovoltaic systems. However, there is one strong incentive for solar thermal when
deciding to go off-grid. Thermal energy storage presents a much more environmentally
friendly approach than electrical storage in batteries. Consider what would happen in the
event of a leak. A large water tank spills only hot water, with the environment effects no
different than a hard rain. A leaking battery releases a chemical soup that must be
remediated with extreme caution. In addition, batteries often release hydrogen gas which
can present an explosion risk if not properly vented. Still, these environmental concerns
are a bit of a worst case scenario. Modern battery technology has created sealed batteries
that are largely immune to leaking and dangerous off-gassing. In addition, a certified
solar contractor will design a battery enclosure that properly vents away hydrogen gas
should it be emitted in explosive quantities.
The conclusions in this chapter are meant to guide the future designer through
some basic assumptions regarding the design of solar thermal heating and cooling
systems. However, it's important to consider the physical characteristics in parallel with
the economic impact of design decisions. For the progressive building owner, a solar

143

thermal heating and cooling system represents an advanced green design technology that
relies on renewable energy, eliminating two of the three largest drains on the existing,
largely non-renewable energy infrastructure. By taking into account the performance
characteristics of the Audubon Center at Debs Park, design decisions can be made to both
match the successes of the system and compensate for its shortcomings.

144

Chapter Eight: Future Work

The solar thermal heating and cooling system at the Audubon Center at Debs Park
is highly representative of the recent strides made in energy efficient building systems.
While the system provided a wealth of data for analysis and evaluation, it ultimately
represents only one configuration of the many possibilities for this type of technology.
The future investigator has many options to continue the study of solar thermal interior
comfort systems. These include computer simulations and deeper studies of individual
components.
This study focused entirely on empirical data coming from the building site. This
allowed for careful analysis of this particular system, but is understandably limited in
being extrapolated or considered universally true. Two courses of study would allow for a
more all-encompassing body of conclusions. The first is to repeat the core parts of
Audubon study with other solar thermal installations. Unfortunately there are not many of
these systems in the United States, though a good number are in operation in Europe.
Parameters such as chiller performance, mass flows, and system temperatures could be
collected at various sites and compared against the Audubon data to work to build the
body of knowledge for this type of system. A full understanding of the benefits and
drawbacks of a solar thermal interior comfort system will only come with continued
evaluation of completed projects. A second course of study could focus on simulating a
solar thermal hydronic HVAC system. This would allow for rapid analysis of predicted
system performance in various climates and could produce sweeping guidelines for the

145

ideal climate conditions necessary for high performance. It would also allow for an easy
analysis of different measures and could quickly identify any synergies that might appear.
For example, both the size and type of the collector array and the size and type of storage
tank have an effect on system performance. A complex computer model could run both
variables through a range of values and identify optimum configurations for a given set
up. It would also allow for easier integration of additional components such as back-up
heaters and chilled water storage. Previous studies in this area have used the TRNSYS
software package. The program performs TRaNsient SYtems Simulation, from which the
software takes its name. It has long been used for solar thermal analysis and would likely
be the best choice to do simulation work for a hydronic system. However, it has been
called difficult to learn and no free student version exists.
Additionally, any of the individual components in the Audubon system could
likely support a research paper.

1) Solar thermal collectors. One course of study could be to evaluate different
types of solar thermal collectors. While flat plate collectors are sometimes
assumed to be incapable of efficiently running an absorption chiller, there are
systems that use them. The researcher could evaluate how to use flat plate
collectors with success, for example, by determining the area of panels needed
per ton of cooling or heating for a known insolation profile. Further study of
evacuated tube collectors could focus on how to optimize their performance.
This paper showed that the overall efficiency of the collectors was lower than

146

manufacturer claims. Later research might try to determine optimum tilt
angles and the effects of ambient temperature and shading on heat delivery.
Another interesting experiment might be to test various combinations of
parallel and series panels. For example, with 100 collectors, what heat
delivery and temperature differences would appear with 10 parallel strings of
10 collectors versus 4 parallel strings of 25 collectors? What pump energy
would be required for each configuration? Lastly, there are now working
examples of concentrated solar thermal systems driving absorption chillers.
This type of system can produce water hot enough to run a double-effect
absorption chiller. However, they must track the sun. The researcher could
evaluate the large scale potential for this type of system and attempt to
quantify the losses due to imprecision in tracking of the sun. There is also the
question of the ability of a concentrating system to effectively heat and cool a
building if solar availability is limited but demand still exists.

2) Absorption Chiller. The inner workings of an absorption chiller are a complex
series of thermodynamic processes and, although based on a common core
concept, every company builds their machines with unique properties. One
potential study is an inventory of the available small-scale absorption chillers
available on the worldwide market. The researcher could identify the unique
advantages claimed by each in addition to quantifying the various efficiencies,
power draws, and capacities available on the market. The actual

147

thermodynamics represent another area of research that has received attention
from a small number of professors throughout the world. While this chapter
treated the absorption chiller as a “black box” component, further studies
could look at optimizing other components to match the characteristics of the
refrigerant/absorber in question, or at the potential of different
refrigerant/absorber pairings such as ammonia/water for solar cooling
applications.

3) Storage Tank. As shown in this paper, the hot water storage tank is a simple
yet vital piece of the mechanical system. Further research could focus on this
component, primarily on how to optimally size it for a known chiller. Or,
perhaps there is potential for a series of smaller storage tanks. Small tanks can
be brought to temperature more quickly, though offer less energy storage. Is it
possible to use, say, three 500 gallon tanks in place of a 1500 gallon one, then
selectively route heat to and from them to increase the operating temperature
of the system? For example, if the entire solar array water stream was diverted
to a single small tank, the water would quickly heat up. This tank could then
be tapped to run the absorption chiller or provide heat while the solar array
was diverted to the next tank.

There is also the idea of producing and storing excess chilled water. The
Audubon system does not do this, but many other systems throughout the
world do. The same sizing studies mentioned for the hot storage could be

148

applicable for cold storage. The researcher could evaluate the optimum
temperature and volume for the stored chilled water and how to best use it.
For example, should it be stored at delivery temperature and supplied to the
building when the chiller can’t meet the load? Or should it be used to pre-cool
the return water to reduce the chiller load in times of low thermal energy
availability?

4) Backup Heat Supply. Though the Audubon Center does not utilize a backup
heat source, it showed that one is necessary to guarantee thermal comfort. A
future study could look at how to optimize this component. The two most
important questions would be how large a burner to install and where to install
it. Should it heat water in a separate storage tank that serves as backup hot
water, or should it directly heat the water in the main storage tank, making up
the difference between what solar energy could provide and what is needed by
the system? Or perhaps the most efficient place is in between the hot water
storage tank and the absorption chiller or primarily loop, depending on the
season. It would function like a residential tankless water heater, quickly
bringing the input water from the storage tank up to temperature. A final
calculation could analyze the economics and environmental impact of a
backup heater. An electric heater powered by solar panels would have an
upfront cost for the panels and possibly battery storage. A grid tied electric or

149

gas heater would have both an upfront cost and a quantifiable carbon
footprint.

Lastly, any of the potential future work discussed in this chapter could easily be
combined. An advanced computer simulation could focus on the design parameters of
any or all of the individual pieces of equipment. Likewise, various pieces of equipment
could be closely scrutinized as part of an empirical study of multiple installed systems.
There is also an opportunity to evaluate how the Audubon System could scale up. What
sort of roof area to occupied area ratio is required to keep the system in operation?
As with any advanced building system, there seems to always be another potential
study waiting in the wings. As the technology matures and more case studies pop up, this
will only increase. Hopefully continued interest in the subject leads to greater levels of
understanding, acceptance, and integration of solar thermally driven HVAC systems.

150

Bibliography

Balaras, C, et al. (2006). Solar Cooling: An Overview of European Applications &
Design Guidelines. ASHRAE Journal. 48, 14-21.

Bergquam, J. (2003) Audubon Nature Center: Solar HVAC Owner Training Syllabus.
Retrieved from Audubon Center at Debs Park.

Chauliaguet, C, et al. (1979). Solar Energy in Buildings. Surrey: John Wiley & Sons, Ltd.

Henning, H. (2007). Solar-Assisted Air-Conditioning in Buildings. Stefan: Springer-
Verlag/Wein.

Humphreys, Michael. (2009). The Future of Evacuated Tube Technology. Available:
http://www.altenergymag.com/emagazine.php?issue_number=09.02.01&article=robe
rts. Last accessed March 26, 2010.

Jordan, R and Liu, B (1977). Applications of Solar Energy for Heating and Cooling of
Buildings. 3rd ed. New York: ASHRAE, INC. XI-1 — XIII-7.

Lewis, N. (2007). Toward Cost-Effective Solar Energy Use. Science. 315 (5813), 798 -
801.

NASA. (2003). The Inconstant Sun. Available: http://science.nasa.gov/science-
news/science-at-nasa/2003/17jan_solcon/. Last accessed March 3, 2010.

Santamouris, M.. (2006). Adaptive Thermal Comfort and Ventilation. AIVC VIP. 12, 8
pp.

Sunda Solar. (2010). SEIDO1 - Heat Pipe Solar Vacuum Tube. Available:
http://www.sundasolar.com/product_seido1.html. Last accessed January 8, 2010.

The Engineering Toolbox. (2005). Clo - Clothing and Thermal Insulation. Available:
http://www.engineeringtoolbox.com/clo-clothing-thermal-insulation-d_732.html. Last
accessed February 26, 2010.

United States Department of Energy: Office of Energy Efficiency and Renewable Energy
(2009). 2008 Buildings Energy Data Book. D&R International, Ltd.Available:
http://buildingsdatabook.eren.doe.gov/docs%5CDataBooks%5C2008_BEDB_Update
d.pdf.

151

United States Department of Energy: Office of Energy Efficiency and Renewable
Energy. (2008). Technologies. Available:
http://www1.eere.energy.gov/solar/pv_cell_light.html. Last accessed January 16,
2010.

United States Green Building Council. (2008). Audubon Center at Debs Park. Available:
http://leedcasestudies.usgbc.org/overview.cfm?ProjectID=234. Last accessed April 7,
2010.

Wright, G. (2004). Solar-Thermal HVAC Technology Debuts in Los Angeles at new
Audubon Urban Nature Center . Available:
http://www.energypulse.net/centers/article/article_print.cfm?a_id=610. Last accessed
March 12, 2006.

152

Appendix A: Floor Plan

Figure A.1: Audubon Center at Debs Park floor plan with HVAC zones shaded

N

153

Appendix B: Building Loads

Library/Conference Room

TABLE B-1: Building load breakdown for the library/conference room

154

Offices

TABLE B-2: Building load breakdown for the offices

155

Discovery Room

TABLE B-3: Building load breakdown for the Discovery room

Abstract (if available)

Abstract With over 40% of total energy use in the United States coming from buildings, it is clear that future conservation strategies must dedicate a significant focus towards the built environment. Space heating and cooling represent two of the three largest building energy loads and thus provide an excellent springboard for alternative technologies that use renewable energy as the primary energy source. One such system uses solar thermal energy as the primary driver for thermal comfort. Heat energy from the sun is absorbed and stored in water. In the winter months, the hot water is pumped directly into the building to provide space heating. In the summer, the hot water runs an absorption chiller to provide cooling to the space.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

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

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

PDF

Tiny house in the desert: a study in indoor comfort using moveable insulation and thermal storage

PDF

Night flushing and thermal mass: maximizing natural ventilation for energy conservation through architectural features

PDF

Exterior glare simulation: understanding solar convergence from concave facades using heat maps

PDF

Modular shading: using design to mitigate bus rider thermal heat stress

PDF

Environmentally responsive buildings: multi-objective optimization workflow for daylight and thermal quality

PDF

Thermal performance of a precast roof assembly: achieving comfort using dynamic insulation and photovoltaics in an extreme climate designed for a small residence at Joshua Tree National Park

PDF

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

PDF

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

PDF

Mitigating the urban heat island effect: thermal performance of shade-tree planting in downtown Los Angeles

PDF

Occupant-aware energy management: energy saving and comfort outcomes achievable through application of cooling setpoint adjustments

PDF

Post occupancy energy evaluation of Ronald Tutor Hall using eQUEST; computer based simulation of existing building and comparison of data

PDF

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

PDF

Microclimate and building energy performance

PDF

Harvesting energy with solar panels and adaptive shading for building skins: A case study of an office building in Saudi Arabia

PDF

A simplified building energy simulation tool: material and environmental properties effects on HVAC performance

PDF

Airflow investigation of fabric membrane forms: a fluid dynamic analysis for thermal comfort

PDF

Tilted glazing: angle-dependence of direct solar heat gain and form-refining of complex facades

PDF

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

Asset Metadata

Creator Kirchhoff, Jason Paul (author)

Core Title Solar thermal cooling and heating: a year-round thermal comfort strategy using a hybrid solar absorption chiller and hydronic heating scheme

School School of Architecture

Degree Master of Building Science

Degree Program Building Science

Publication Date 08/04/2010

Defense Date 03/26/2010

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

Tag absorption chiller,Audubon Center at Debs Park,Building Science,Energy,evacuated tube collector,OAI-PMH Harvest,solar,solar thermal,thermal comfort

Place Name California (states), Los Angeles (city or populated place)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Kensek, Karen (committee chair), Noble, Douglas (committee member), Schiler, Marc (committee member), Woll, Edwin (committee member)

Creator Email jkirchho@usc.edu,kirch119@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3276

Unique identifier UC1216542

Identifier etd-Kirchhoff-3899 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-369883 (legacy record id),usctheses-m3276 (legacy record id)

Legacy Identifier etd-Kirchhoff-3899.pdf

Dmrecord 369883

Document Type Thesis

Rights Kirchhoff, Jason Paul

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu