NISTIR 5758
i III lull iili I lilil
AlllDM TDlbSb
A Workplan to Analyze the Energy Impacts of
Envelope Airtightness in Office Buildings
Steven J. Emmerich
Andrew K. Persily
David A. VanBronkhorst
Building and Fire Research Laboratory
Gaithersburg, MD 20899
NIST
United States Department of Commerce
Technology Administration
dtute of Standards and Technology
QC
100
.056
NO. 5758
1995
I
NISTIR 5758
A Workplan to Analyze the Energy Impacts of
Envelope Airtightness in Office Buildings
Steven J. Emmerich
Andrew K. Persily
David A. VanBronkhorst
December 1995
Building and Fire Research Laboratory
National Institute of Standards and Technology
Gaithersburg, MD 20899
U. S. Department of Commerce
Ronald H. Brown, Secretary
Mary L. Good, Under Secretary for Technology
National Institute of Standards and Technology
Arati Prabhakar, Director
Prepared for:
U.S. Department of Energy
Office of Energy Efficiency and Renewable Energy
Office of Building Technologies
Washington, DC 20585
Abstract
US. office buildings consume approximately 1.2 EJ (1.1 Quadrillion BTUs or Quads) of energy,
0.72 EJ (0.68 Quads) of which is associated with space heating, cooling, and ventilation. These
estimates, and other analyses of energy consumption in office buildings, are based on building
energy analysis programs such as DOE-2. These analyses have been helpful in identifying
opportunities for energy efficiency, developing building energy efficiency standards and
predicting future energy consumption levels. Although these programs contain sophisticated
models of heat transfer and HVAC system performance in buildings, they are acknowledged to
have shortcomings in accounting for the energy associated with building airflows, particularly
infiltration of outdoor air through leaks in the building envelope. These airflows, and their
dependence on weather and ventilation system operation, are more complex than the models used
in these programs. The simple models of infiltration, ventilation and interzone airflows that are
used in these programs do not enable the analysis of the energy consumption associated with
building airflow or the impact of options that may reduce this energy consumption, such as
increased envelope airtightness or better control of ventilation system airflow rates. This report
describes the impact of building airflows on energy consumption in multi-zone buildings and the
analysis approaches that can be used to account for the energy associated with these airflows.
Plans to link a multi-zone network airflow analysis program with a building energy analysis
program are discussed. An initial estimate of the energy associated with infiltration in US. office
buildings, based on a simplified analysis approach, is presented. This estimate reveals that
infiltration in U.S. office buildings accounts for 0.074 EJ (0.07 Quads) of space heating energy
use, which is 18% of the total heating energy use, and 0.0025 EJ (0.0024 Quads) for cooling,
which is 2% of the total.
Key Words: airflow modeling, building energy simulation, building technology, commercial
buildings, computer simulation, HVAC systems, infiltration, ventilation
iii
Acknowledgements
This work was sponsored by the US. Department of Energy, Office of Building Technologies
under Interagency Agreement No. DE-A101-9CE21042. The authors wish to acknowledge the
efforts of John Talbott in support of this project.
Table of Contents
Abstract Hi
Acknowledgements v
Introduction 1
Airflow and Energy Use in Office Buildings 3
Airflow in office buildings 3
Interaction of building airflow and thermal loads 4
Consideration of airflow in building energy use studies 5
Approaches to Combined Thermal and Airflow Analysis 7
Modeling 7
Simulation 9
Project Plan 11
Candidate Buildings 11
Project Phases 12
Other Applications of Simulation Tool 13
Initial Estimate Based on PNL Infiltration Rates 14
Results 15
Summary 17
References 18
Appendix A - Description of Method for Initial Estimate 23
vii
Introduction
US office buildings consume a large amount of energy each year, a substantial portion of which
is used by building heating, ventilating, and air-conditioning systems. A study prepared by
Pacific Northwest Laboratories (PNL) for the Gas Research Institute (Briggs et al. 1992 and
Crawley and Schliesing 1992) estimates that US office buildings consume approximately 1.2 x
1018J, or 1.2 EJ, (1.1 Quadrillion BTUs or Quads) of energy, 0.72 EJ (0.68 Quads) or 60% of
which is associated with space heating, cooling, and ventilation. The PNL study was conducted
to characterize in detail the energy requirements of the office building sector for use in targeting
research of new gas-fueled technologies. PNL performed a statistical categorization of the
existing and future national office building stock which resulted in a total of 30 prototype
buildings that are intended to represent construction through 1995. Energy simulations were then
performed on these buildings using the DOE-2 program (Curtis et al. 1984) to calculate annual
energy use. This estimate is supported by a report published by the U.S. Department of Energy
which estimates that the actual energy consumed in US office buildings in 1989 was 1.3 EJ (1.2
Quads) of which 0.62 EJ (0.58 Quads) was associated with space heating, cooling, and
ventilation (ElA 1994). The DOE estimate is based on a combination of an extensive commercial
building energy consumption survey (EIA 1 992) and energy simulations.
Many other analyses of energy consumption in commercial buildings are based on DOE-2 or
similar building energy analysis programs. These analyses have been helpful in identifying
opportunities for energy efficiency, developing building energy efficiency standards and
predicting future energy consumption levels. Although these programs contain sophisticated
models for the heat transfer in buildings, they are acknowledged to have shortcomings in
accounting for the energy associated with building airflows, particularly infiltration of outdoor air
through leaks in the building envelope. Several researchers have discussed the need for improved
consideration of the combined problem of building airflow and heat transfer (Axley and Grot
1989, Clarke and Hensen 1990, Kendrick 1993, Klobut 1991, Lomas 1991, Pelletret and Khodr
1990, Schneider et al. 1995 and Tuomaala and Rahola 1995). The interactions of airflow into and
■within buildings with heat transfer and storage in buildings and with building HVAC system
operation are complex. Building thermal analysis programs and multizone airflow programs have
been extensively used to examine building energy use and airflow separately. However,
calculation of the combined problem has received much less attention. To date, most of the
efforts at analysis of the combined problem have been focussed on the simpler case of
single-family residential buildings (Melo 1986, Pelletret 1987, Fischer 1993).
Better information on the actual energy impacts of building envelope leakage and poorly
controlled ventilation system airflow rates is needed to determine the cost-effectiveness of
improved airtightness and system control. Mechanically ventilated commercial buildings
currently experience significant amounts of envelope air leakage and poor control of ventilation
system airflows. The simple models of infiltration, ventilation and interzone airflows that are
used in most existing building energy analysis programs do not enable the analysis of the energy
consumption associated with building airflow or the impact of options that may reduce this
energy consumption, such as increased envelope airtightness or better control of ventilation
system airflow rates. Research is being performed with the objective of improving building
1
airtightness and optimizing ventilation system control in order to reduce building energy use and
improve indoor air quality. However, construction of tighter buildings and implementation of
better controls involves increased costs for state-of-the-art technologies and materials, and an
understanding of the energy impact of these impacts is needed to justify their additional first
costs.
The National Institute of Standards and Technology (NIST) is conducting a study for the US.
Department of Energy (DOE) to assess the energy impacts of building airtightness and
ventilation system control. In this project, NIST will perform whole building airflow and thermal
analysis simulations to quantify the energy costs of airflows associated with building leakage and
poor ventilation system control in large, commercial buildings. The project has three primary
objectives. The first objective is to determine the percent of total energy consumption in
commercial buildings due to envelope infiltration and non-design ventilation system airflow
rates. The second is to determine the energy savings potential of improving building airtightness
and ventilation system control in existing buildings. The third objective is to determine the
energy savings potential in new construction.
This report summarizes the efforts to develop the project and the method and results of the first
phase of the project. It consists of four main sections titled: Airflow and Energy Use in Office
Buildings, Approaches to Combined Thermal and Airflow Analysis, Project Plan, and Initial
Estimate Based on PNL Infiltration Rates. The first section discusses the building physics
involved in the interaction of building airflow dynamics and thermal loads and describes some
shortcomings in the treatment of airflow in building energy use studies. The second section
describes the selection of the analytical tool to be used in the project, and includes a review of
available analytical tools. The next section outlines the research plan including a discussion of
candidate buildings for simulation and a description of the project phases. This section also
includes a discussion of other applications for the simulation tool that are not part of the current
project plan. The final section presents an initial estimate of the national energy impact of
infiltration in office buildings.
2
Airflow and Energy Use in Office Buildings
The impact of building airflows on energy use in office and other commercial buildings has
received only limited attention. As discussed later in this report, the consideration of the energy
impacts of airflow has generally been restricted to intentional outdoor air intake through
mechanical ventilation systems. Little attention has been given to unintentional infiltration
through envelope leaks or to detailed representations of mechanical ventilation system airflows.
The analysis of the energ>’ impacts of building airflow has been limited, in part, because of the
complexities of airflow in office buildings, which are usually mechanically ventilated and almost
always must be considered as multizone systems. This section presents some background
information on airflow in office buildings, how these airflows impact thermal loads, and how the
interaction of airflow and energy has been addressed in previous studies of energy use in office
buildings.
Airflow in office buildings
Airflow into, w ithin, and out of office buildings is a complex phenomena. These buildings
almost always beha\ e as multizone airflow systems, meaning that the resistances to airflow
between different building zones are significant relative to the resistance across the exterior
envelope and that significant airflows exist between different portions of a building. In addition,
most office buildings have mechanical ventilation systems that supply and remove air from these
zones, adding complexiw to the problem. Airflow rates in multizone systems are determined by
weather conditions (air temperature, wind speed and wind direction), effects of surroundings on
the exposure of the building to the wind, the temperature distribution within the building, the
airflow rates to and from each zone from mechanical ventilation systems, and the airtightness of
the exterior envelope and of the interior partitions between zones. The physical mechanisms that
determine these airflow rates are well understood and include temperature and wind-induced
pressure differences as well as the operation of mechanical ventilation equipment, however the
complexity of the airflow patterns in any given building can be overwhelming.
It is generally assumed that the envelopes of modem office buildings are relatively airtight. In
addition, the ventilation systems of these buildings are generally designed to operate with an
excess of outdoor air intake over exhaust airflow out of the building such that the building is at a
higher pressure than outdoors and the infiltration of outdoor air into the building through leaks in
the building envelope is minimized or eliminated. When performing analyses of energy use in
office buildings, these assumptions often translate into infiltration rates of zero. However, field
studies of office buildings have shown that these assumptions are not necessarily valid. While the
exterior envelopes of office buildings have generally been assumed to be fairly airtight, the
results of pressurization testing of the exterior envelopes have shown that the envelope
airtightness of these buildings is similar to that of leaky residential buildings in terms of leakage
area per unit wall area (Persily and Grot 1986, Shaw and Reardon 1995, Tamura and Shaw
1976). In addition, ventilation system airflow rates do not necessarily correspond to their design
values, which can alter the intended pressurization of these buildings (Persily and Norford 1987).
Schliesing et al. (1993) summarizes literature reports that describe problems with HVAC system
maintenance and operation such as stuck dampers, blocked return vents, and disconnected
3
controls. Leaky buildings and imperfect control of ventilation system airflow rates can combine
to yield significant air infiltration rates, on the order of one-half to one air change per hour as
opposed to rates of about one-tenth as are often assumed to exist in the buildings. In addition to
increased infiltration rates, ventilation system airflow rates that differ sigmficantly firom their
design values can result in unintended airflows between different zones of a building. These
airflows can be undesirable from the perspectives of energy, thermal comfort and indoor air
quality.
Tracer gas measurements of air infiltration rates in office buildings have shown that the
assumption of low infiltration rates, upon which the design, operation, and energy analysis of
such buildings are based, are often false. Grot and Persily (1986) reported air infiltration rates in
eight federal office buildings constructed since 1976 that varied from about 0.2 to 0.7 ach. Persily
and Norford (1987) reported measurements of total air change, infiltration and outdoor air intake
under a range of weather conditions in a three-story office building constructed in 1984. The
infiltration rates in this building were generally on the same order of magnitude as the intentional
outdoor air intake rates and were strongly dependent on the mode of outdoor air intake control.
The existence of significant air infiltration rates in office buildings can have a number of
negative consequences. These include increased energy consumption, thermal comfort problems,
the degradation of indoor air quality because the infiltrating air in not filtered, the increased
potential for moisture damage, and the degradation of envelope materials and interior
furnishings. While an assumption that infiltration does not exist, or the rates are very low, in
office buildings is not necessarily correct, it is not clear whether the associated assumption of
negligible energy consumption due to infiltration is appropriate or not. In order to answer this
question, the energy use associated with heating or cooling this infiltrating air must be
determined.
Interaction of building airflow and thermal loads
Building heating and cooling loads depend on the heat gains and losses through the building
envelope, heat and moisture gains due to internal sources such as people and equipment, storage
of heat within the building, and the transfer of heat between the zones of the building. The
physical phenomena involved include transient conduction, radiative exchange (both longwave
and shortwave), convection from surfaces (both internal and external), and bulk convection (due
to infiltration and interzone airflows in a multizone building). Most of these phenomena can be
modeled with available simulation packages such as DOE-2 (Curtis 1984), TRNSYS (Klein
1992), and HVACSIM+ (Park 1986).
Assessing the energy impacts of building airflows requires consideration of the mechanisms and
factors that induce these airflows and of the interactions of building airflows with other heat
transfer processes. First, there is the energy required to heat or cool the infiltrating air. However,
the energy liability associated ■with infiltration is not always straightforward; it can depend
strongly on the type of HVAC system and the strategy employed to control this System. For very
low and very high outdoor air temperatures, infiltrating outdoor air ■will add to the space
conditioning load by an amount proportional to the indoor-outdoor air temperature difference and
4
to the air infiltration rate. During mild weather conditions, infiltration can cool the building,
thereby decreasing the cooling load. If the infiltrating airflow is induced by building
depressurization caused by inadequate outdoor air intake, then the infiltration air is simply
replacing ventilation air and does not necessarily entail an energy liability. However, ventilating
a building through infiltration as opposed to mechanical ventilation may involve negative
consequences in terms of indoor air quality.
Additional interactions between airflow and thermal loads are associated with the indoor air
temperature. Indoor air temperatures are important factors in determining air infiltration rates due
to the stack effect. In addition, the indoor temperature impacts the operation of the HVAC
system, for example through the modulation of supply airflow rates in VAV systems, which in
turn affect building pressures and therefore infiltration rates. The resultant infiltration rates are
important determinants of thermal loads, and therefore impact the interior air temperatures
themselves.
Consideration of airflow in building energy use studies
Past efforts analyzing the annual energy use in large buildings have suffered from two main
shortcomings regarding the impacts of building airflows. The first is the use of simplified
mathematical models of airflow in buildings. For example, an analysis of the energy impact of a
demand-controlled ventilation system used a multizone building but assumed values for
infiltration and interzone flows that depended only on operation of the HVAC system (Emmerich
1993). Several studies on the energy impacts of increasing outdoor air intake in commercial
buildings have used a simulation program which, at best, uses a simple single-zone infiltration
model considering wind-dependence (Eto 1990, Eto and Meyer 1988, Steele and Brown 1990,
and Zmeureanu et al. 1 992). A detailed analysis of the energy use in a modem office building
assumed constant values for building infiltration that depended only on whether the building was
occupied (Norford 1984). These are just a few examples of the consideration of airflow in
building energy use studies in the past. Numerous examples of building energy use studies with
simplified airflow models may be found in the literature.
Another shortcoming has been in the assumed values of infiltration and ventilation rates in large
modem office buildings. Past analyses of energy use in such buildings have generally assumed
that only small amoimts of air infiltrate the building through the envelope due to a combination
of tight constmction and a positive pressure maintained inside the building by the mechanical
ventilation system(s). Since low infiltration rates are assumed, the total building air exchange rate
is assumed to be dominated by outdoor air intake through the mechanical ventilation system. The
mechanical ventilation systems are assumed to provide a minimum level of outdoor air intake,
keep the supply fan airflow rate sufficiently above the return airflow rate to pressurize the
building, and properly adjust the positions of system dampers for economizer cycle operation.
The modeling of airflow in building energy use studies in the past has neglected important
factors and has ignored the reality of infiltration and ventilation system operation in office
buildings. The current state-of-the-art in whole building airflow modeling requires the use of
multizone airflow models. These models are discussed further in the next section.
5
No reports have been found in the literature of studies analyzing either the energy impacts of
poor building airtightness and ventilation system control or the energy savings potential of
measures to correct these conditions. Therefore, the expenses which can be justified for reducing
infiltration and improving ventilation system control are not known.
6
Approaches to Combined Thermal and Airflow Analysis
This section describes existing approaches for analyzing the energy impact of infiltration and
ventilation system airflows, discusses shortcomings of these approaches, and discusses better
ways to account for the complexity of airflow-related energy in multizone buildings.
The discussion is divided into modeling and simulation tools as the selection of an analysis
approach requires consideration of both. The distinction between modeling and simulation is
described by Jeandel and Palero (1991). Modeling includes the choice, creation, and validation of
models, where the term model is used to describe a mathematical representation of the physical
phenomena involved. The mathematical representation must be adequate to yield results that will
achieve the project objectives, and therefore, the choice of a model follows directly from the
definition of the problem. Simulation includes choice or creation of the simulation tool and the
simulations to be performed. Beyond incorporation of the previously selected model(s), the
selection of the simulation tool will depend on less scientific reasoning. Reasons for choosing a
specific simulation tool may include familiarity or previous experience, ease of use, input/output
processing capabilities or other special features, and availability of support.
Modelling
The objective of the project is to examine, on a national scale, the energy impacts of the airflows
associated with building envelope leakage and poor ventilation system control. The criteria by
which the results will be measured is the annual energy use to operate a building. The energy use
due to these problems must be calculated accurately enough to evaluate the difference between
different levels of building airtightness and ventilation system control.
The intent of this project is to apply available models rather than create new models. The choices
of available models range in level of detail from simple regression equations to detailed room
airflow modelling and include the following;
1 . Regression equation models
2. Room airflow models
3. Building energy balance (with multizone airflow) models
4. Integrated airflow and thermal element models
The first two types of available models can be readily dismissed from consideration for the
analysis approach. Regression equation models are often the results of multiple parametric runs
of a more powerful model which have been reduced to simple relationships (Clarke 1985).
Available regression equation models (such as degree day methods) do not incorporate the level
of detail necessary to provide results capable of accurately determining the difference in building
energy use due to building airflows.
Even as regression equation models are not detailed enough, room airflow models are too
detailed. Room airflow modelling applies the principles of conservation of momentum, mass,
and energy through the use of a computational fluid dynamics (CFD) program. This type of
7
model could provide a detailed analysis of the coupled airflow and heat transfer processes that
occur in a building. A recent review of the application of CFD programs to room airflow
modeling by the International Energy Agency is reported by Moser (1992). However, modeling
an entire building is beyond the capability of current computers for even a steady state condition
and would require a massive effort for data input and results analysis.
In general, building energy balance models provide a level of detail between that of the
regression and CFD models. As the name implies, this type of model involves applying an energy
balance to the building. This category of models can be further broken down depending on the
numerical implementation of the energy balance. Two different implementations include the
response function model and the finite difference model.
Response function models divide a building into large thermal zones. For a given zone, energy
balance equations are written for the entire energy field. The solution of this set of equations
when subjected to a unit excitation function gives the corresponding unit response function
(URF). The complete URF set is determined by repeating for each possible excitation. This
method provides a specific analytical solution to the zonal energy balances. The transfer function
method is the baseline procedure adopted by the American Society of Heating, Refrigerating and
Air-Conditioning Engineers and is considered one of the most accurate methods of calculating
building heating and cooling loads (McQuiston and Spitler 1992). A thorough description of this
type of model may be found in Clarke (1985) and McQuiston and Spitler (1992).
Clarke (1985) also describes the finite difference method of building energy calculation. In this
method, a building is divided into many small volumes and energy balance equations are written
for each volume. Numerical solution techniques are used to solve these equations. Clarke
describes this method as general in concept with better physical insight, however, he states that
the quality of the results depends on the care taken in implementing the model.
These energy balance models do not directly incorporate the consideration of the interactions of
the airflows and thermal processes discussed in the definition of the problem. Accurately
modeling the airflow in large commercial buildings requires use of a multizone airflow model. A
multizone airflow model applies a mass balance to a network of elements describing the flow
paths (HVAC ducts, doors, windows, cracks, etc.) between the zones of a building. The network
nodes represent the zones which are modeled at a uniform temperature. Walton (1989) discusses
multizone airflow modeling in detail.
The fourth method, integrated airflow and thermal element modeling, involves a direct coupling
of the airflow and thermal modeling equations. Axley (1988) develops a modeling approach to
the thermal analysis of buildings that is analogous to the method of multizone airflow modeling
described above. In this method, the building is defined as discrete thermal elements for which
equations describing the thermal transport processes are written. These elemental equations are
then assembled into a matrix which is solved by finite element techniques. Axley (1989)
describes the integrated coupling of this element based thermal analysis model with the airflow
analysis model. This integrated model is described as quasi-dynamic as it combines a dynamic
thermal solution procedure and a steady airflow solution procedure.
8
Three of the modelling options discussed above (multizone airflow modelling combined with
response function, finite difference or element assembly thermal modelling) adequately describe
the physical phenomena involved and could provide results to meet the problem objective. The
specific model and implementation chosen are discussed below.
In general, models need to be validated to determine the quality of the model. Jeandel and Palero
(1991) describes four types of model validation including: numerical validation, analytical
simulation, quantitative validation, and experimental validation. These types of validation apply
to individual models as well as coupled systems of models. No new models will be created in this
project and the individual models to be used have been subjected to various types of validation in
the past. However, new combinations of the models may be used and these combinations should
be subjected to some level of validation.
Simulation
The multizone airflow models discussed above have been implemented in many forms. Walton
(1989) described one of the first widely available multizone airflow simulation programs called
AIRNET. Many other multizone airflow programs have been written since AIRNET, and Feustel
(1991) describes a survey of multizone airflow programs. One particular multizone airflow
program which deserves specific mention is COMIS (Feustel 1990). COMIS was written as part
of an international effort of airflow modelling specialists sponsored by the International Energy
Agency. It is based on the same multizone airflow model as AIRNET and shares many of the
same simulating capabilities. The latest available version of the AIRNET is included in the
multizone airflow and contaminant dispersal program CONTAM94 (Walton and Emmerich
1994). CONTAM94 combines the best available algorithms for modeling the airflow and
contaminant dispersal in multizone buildings with a unique graphic interface for data input and
display.
Response function building energy models have been implemented within many available
simulation tools including TRNSYS (Klein 1992), HVACSIM+ (Park 1986), and DOE-2 (Curtis
1 984). None of these tools incorporates a multizone airflow model at this time. Although these
simulation tools implement the same basic model, there are some important differences between
these tools to consider.
TRNSYS and HVACSIM+ are both modular simulation tools which include a central simulation
'engine' which provides solution routines and performs various output and input data processing
tasks and modules which are implementations of equipment and building energy models. Both of
these tools are very flexible, allowing the user to select the needed modules from the provided
ones and to add needed capabilities by -writing new modules. One significant difference is the
time scale for which they were intended to operate. HVACSIM+ was intended to be used for
detailed and accurate simulation of the control systems of building HVAC equipment which may
require time steps on the order of seconds or minutes. As a result, performing a yearlong
simulation becomes a large computing effort. TRNSYS was originally intended for simulating
solar energy systems, many of which can be accurately simulated -with time steps on the order of
9
hours with year-long simulations being routine. It is anticipated that such time steps will be
appropriate for this project.
While employing an implementation of the same model, DOE-2 was not designed as a modular
program. As such, DOE-2 is not easily modified by the user and is generally limited to the
currently available capabilities. No multizone airflow model is incorporated within DOE-2 and
adding one would require significant effort. Modera (1992) reports on a study which considered
the interactions between multizone airflow and building thermal processes by combining DOE-2
with COMIS (Feustel 1990) via a third program called DUCTSIM. This program was written to
pass information between DOE-2 and COMIS but is not a general coupling and would need to be
rewritten for each building simulated.
One available implementation of the finite difference building energy model is the simulation
tool ESP (Clarke 1 988). ESP is a modular simulation tool similar in some respects to TRNSYS
and HVACSIM+. ESPmfs. an implementation of a multizone airflow model based on AIRNET,
was added as an ESP module by Hensen (1990).
An available implementation of the integrated element (or element assembly) models is the
program TFCD (Klobut 1991). TFCD is capable of simultaneously calculating airflows, air
temperatures, and contaminant concentrations in a multizone building. However, this simulation
tool has a limited number of elements and does not yet have the needed capability of modeling
the HVAC system equipment and is not yet intended to be used for annual energy use
calculations. Tuomaala and Rahola (1995) reported the development of a program called BUS
with similar limitations.
One additional simulation program that combines a building thermal analysis model with a
multizone airflow modelling capabilities should be mentioned. The HOUSE-II model (Fischer
1993) incorporates the appropriate models to study the physical phenomena discussed above.
However, this program was designed specifically for residential applications and as such is not
appropriate for the project objective.
The two most promising alternatives for simulation tools for this project are (i) TRNSYS with a
new module based on AIRNET and (ii) ESP. Both of these tools would provide the capabilities
needed to achieve the project objective. ESP has the advantage of not requiring any simulation
tool development but would require learning to use the simulation environment and the finite
difference energy model. Selection of TRNSYS requires developing a new module. While not
trivial, this task is made easier due to familiarity within NIST with the TRNSYS simulation
environment and response function building model and with the AIRNET multizone airflow
model. Selection of this option also has the advantages of making a multizone airflow module
available to TRNSYS users and of taking advantage of the convenient graphic interface of
CONTAM94 for preparing data input to the airflow module. Dorer and Weber (1994) recently
reported the results of a similar effort in which TRNSYS and COMIS were combined to study
passive cooling and natural facade driven ventilation in a school building. Due to the similarities
between AIRNET and COMIS, discussed above, this effort indicates the feasibility of the
proposed option of creating a TRNSYS module based on AIRNET.
10
Project Plan
This section outlines the research plan developed to examine the national energy impacts of
infiltration and ventilation airflows in office buildings. It includes a discussion of candidate
buildings for simulation and a description of the project phases. This section also includes a
discussion of other applications for the simulation tool that are not part of the current project.
Candidate Buildings
The commercial building stock in the United States contains a great variety of buildings in terms
of size, location, design features, and operation. To estimate the energy impacts on this
population requires selecting some representative set of buildings to simulate and then
extrapolating the results to the population as a whole. To save the effort required to define a set
of buildings for the study, the literature was reviewed for reports of sets of prototypical
commercial buildings which could be used.
Briggs et al. (1987) describes a categorization of the office building stock derived from a
statistically valid sample of the nation's office building sector. This effort developed 20 office
building categories based on a statistical technique known as cluster analysis. Categories or
clusters were defined on the basis of physical attributes such as size, age, location, and building
energy loads. Crawley 1 992 added to the categorization of the building stock by specifying 1 0
additional buildings representing recent and future construction. However, the 1 0 buildings
defined to represent recent (circa 1986) and future (circa 1995) construction are essentially only 5
buildings with minor differences between the recent and future prototypes. Therefore, the
building set could be condensed to 25 buildings. These prototype descriptions will be reviewed
further to determine their appropriateness for use in this study. It may be necessary to further
restrict the cases considered based on the available project resources.
Huang et al. (1991) describes a set of 481 prototypical buildings used to develop a building loads
database for assessing cogeneration market potential. The buildings simulated included offices,
hospitals, schools, prisons, hotels, restaurants, supermarkets, apartments, and retail stores
specified to characterize the commercial building stock in 20 urban market areas. This building
set is more elaborate than required for this study, however, some subset of the buildings may be
appropriate. In addition to developing a prototypical building set, Huang conducted an extensive
review of other studies that either defined average building conditions or developed prototype
buildings. Many of these studies developed prototypical buildings specific for one region of the
country and were not intended to be representative of the national building stock.
Another study describing a set of prototypical buildings for a specific region but not included in
Huang's review is reported by UIC (1989). This study develops 10 commercial building
prototypes representative of typical characteristics of the building stock in the Bonneville Power
Administration's service area.
Included in Huang's review but worthy of separate mention is a set of prototypical buildings
developed by Pacific Northwest Laboratories to support commercial building energy standard
11
research (PNL 1983). The building prototypes included offices (small, medium, and large), retail
stores (small and large), an apartment, a hotel, a warehouse, a church, and a school based on real
buildings judged to be typical for their type. Although the buildings have been simulated in many
locations, no statistical representation of the building stock was made.
Friedrich (1994) described another methodology to characterize energy use in the national
commercial building stock. The method uses a three-story building prototype with characteristics
for each US. census region based on Commercial Building Energy Consumption Survey
(CBECS) data (EIA 1992) and using loads specific for the building type (e.g. office). Simulations
are then performed for representative climate zones identified by Hadley and Jamigan (1993) to
find the annual energy use intensity (EUI) for each building thermal zone. The energy use
estimates are then scaled based on CBECS data on building construction, size, and fuel type
distribution. However, there is no simple way to correlate the identified climate zones with the
CBECS data and some type of averaging across census regions is necessary.
Project Phases
Phase 1 . Initial estimate based on PNL infiltration rates
The objective of this phase is to make a rough estimate of the national impact of infiltration
loads. This estimate is not intended to be very reliable but simply to indicate the magnitude of the
energy used, in order to demonstrate the importance of investigating the problem. This phase has
been completed and a summary of the method and the results are described in this report in the
section called Initial Estimate Based on PNL Infiltration Rates. A detailed description of the
initial estimate is included in Appendix A.
Phase 2. Improved estimate based on AIRNET infiltration rates
This project phase will improve on the initial estimate by using airflows from multizone airflow
simulations. Airflow simulations will be performed for each of the prototype buildings under a
range of indoor - outdoor temperature differences and wind speeds to determine whole building
air change rates under these weather conditions. These simulations will be performed both with
the HVAC systems off and on. Then, infiltration heating and cooling loads will be calculated as
in Phase 1 except the calculated air change rates will be used in place of the PNL infiltration
rates. The building set may also be expanded to include other prototypes of other commercial
building market segments (possibly retail, assembly, education, and/or warehouse) .
Phase 3. TRNSYS/AIRNET approach
This phase involves use of the TRNSYS/AIRNET approach described in the previous section.
The first step is developing a TRNSYS module based on the AIRNET program. After the
TRNSYS/AIRNET coupling is complete, a prototype building will be modeled as a trial case to
verify the viability of the simulation method. It will be useful to simulate a building which has
been simulated previously to provide an inter-model comparison of the energy use predictions.
Many questions on the details of the building simulation will be answered at this point.
12
After the trial simulations have been successfully performed, the complete set of prototypical
buildings will be simulated to estimate the total energy consumption due to envelope infiltration
and non-design ventilation system airflow rates. Separate simulations will also be performed to
determine the energy savings potential in both existing buildings and new construction.
Estimating the energy impacts of poor building airtightness and ventilation system control will
also require defining both poor and good building airtightness and ventilation system control for
each of the representative buildings. The published literature on building airtightness and
ventilation system control will be reviewed to assist in making these determinations.
Phase 4. Solution technologies for reducing energy impact
The objective of this project phase is to develop guidance on reducing the energy impacts. This
will begin with identifying technologies for solving problems including envelope design and
construction and ventilation system controls. Experimental work may be performed to evaluate
solution technologies.
Other Applications of Simulation Tool
The simulation tool described above would have many possible applications that take advantage
of the combined building thermal and multizone airflow modeling capabilities. One application
would be analysis of the energy impact of increasing the required ventilation rates in buildings.
The energy impacts of increasing outdoor air intake in commercial buildings have been reported
by several researchers including Eto (1990), Eto and Meyer (1988), Steele and Brown (1990),
Ventresca (1991), and Zmeureanu (1992). As discussed earlier, these studies used simplified
approaches that do not account for the complexities of airflow in multizone buildings.
Consideration of the multizone nature of airflow in large buildings and of the airflow and heat
transfer interactions would result in better estimates of these and other related energy impacts.
Another application would be a parametric study of the interactions of building airflow Avith
various building thermal features. For example, the effectiveness of energy conservation
strategies such as adding insulation or increasing thermal mass may depend on the airtightness of
the building. Although many studies of this type have been performed in the past, the combined
thermal and airflow modelling described here may affect the results of such studies.
A third and very important application would be evaluating indoor air quality. This application
would require further simulation tool development of a TRNSYS module based on the NIST
program CONTAM94 (Walton 1994, Walton and Emmerich 1994) although a less capable
multizone pollutant dispersal module currently exists (Emmerich 1993) and could be used. Such
a tool could be used to evaluate both the indoor air quality and energy impacts of indoor air
quality control strategies.
13
Initial Estimate Based on PNL Infiltration Rates
The first phase of the project involved making an initial estimate of the national impact of
infiltration loads. This estimate is not intended to be very reliable but simply to indicate the
magnitude of the energy use. The initial estimate method uses information from a previous study
by Pacific Northwest Laboratory (PNL) on energy use in office buildings (Briggs et al. 1992,
Crawley and Schliesing 1992). The PNL study was conducted to characterize in detail the energy
requirements of the office building sector for use in targeting research of new gas-fueled
technologies. PNL performed a statistical categorization of the existing and future national
office building stock which resulted in a total of 30 prototype buildings. Energy simulations were
then performed on these buildings using DOE-2 program to calculate annual energy use.
A steady-state method was used to obtain a rough estimate of the annual energy impact of air
infiltration for the PNL building set. For each building, the thermal loads due.to infiltration were
estimated for each hour of a typical year using the equation = q*p*Cp*AT, where is the
sensible load, AT is the indoor-outdoor temperature difference, p is the density of the air, Cp is
the specific heat of the air, and q is the flow rate of infiltrating air. The sensible load is the rate at
which heat must be added to (or removed from) the incoming air in order to raise (or lower) its
temperature by an amount AT. A similar equation describes the latent load due to removal of
moisture during cooling.
Many simplifying assumptions were made in performing these calculations. No transient effects
(i.e. thermal storage) were considered. Balance point temperatures, above which heating loads
were not calculated, were used. Density and specific heat of air were assumed to be constant.
Infiltration flow rates were derived from the air infiltration rates specified for each building by
Briggs et al. (1992), which were adjusted hourly for wind speed but not for temperature. It was
assumed that the infiltrating air would be heated (or cooled) to the current thermostat setpoint of
the building. Briggs et al. (1992) contains detailed operating schedules for each building,
including the thermostat settings for every hour of the day. The assumption of constant
infiltration flow rate and temperature difference during an hour allowed a straightforward
calculation of the total amount of heat added to (or removed from) the building space for that
hour. By neglecting losses in the air ducts, the space heating (or cooling) load becomes
equivalent to the coil load on the HVAC system equipment. Estimates of annual heating and
cooling coil loads were made by summing the hourly loads over the span of a year. The loads
were then converted to energy-use values by applying the overall conversion efficiencies that
were used for the buildings in the PNL study. The details and assumptions involved in this
method are described in Appendix A.
14
Results
Table 1 shows the results of the calculations, in which the heating and cooling loads are
normalized with respect to the floor area that each building represents. For each building, the
associated loads due to air infiltration are shown, along with the total annual heating or cooling
load as predicted by DOE-2, and the percentage of this total accounted for by the infiltration
loads. The last row of the table contains these values for all the buildings together, based on the
individual building \’alues weighted by floor area. Note that these values are the loads on the
heating and cooling coils, not energy consumption, so are independent of the source of energy.
BUILDING
LOCATION
HEATING LOADS
COOLING LOADS
(MJ/m^)
(MJ/mT
Infiltration
Total
% Inf
Infiltration
Total
% Inf
I
Indianapolis. IN
105
656
16%
5
234
2%
2
Cleveland. OH
346
2127
16%
14
355
4%
3
El Paso. TX
26
162
16%
4
429
1%
4
Washington, DC
34
341
10%
4
355
1%
5
Madison. W1
45
313
14%
2
254
1%
6
Lake Charles, LA
20
120
17%
13
621
2%
7
Des Moines. LA
151
1087
14%
7
401
2%
8
St. Louis, MO
104
745
14%
24
764
3%
9
Las Vegas. NV
16
133
12%
4
420
1%
10
Salt Lake City, UT
25
226
11%
2
547
0%
11
Cheyenne, WY
65
382
17%
1
535
0%
12
Portland, OR
70
724
10%
1
199
1%
13
Pittsburgh. PA
78
1357
6%
5
615
1%
14
Amarillo, TX
73
191
38%
6
516
1%
15
Raleigh, NC
15
639
2%
6
1209
0%
16
Dallas, TX
20
185
11%
11
1087
1%
17
Minneapolis, MN
70
651
11%
3
479
1%
18
Boston, MA
46
991
5%
1
990
0%
19
New York, NY
84
233
36%
4
292
2%
20
Los Angeles, CA
6
66
9%
0
1000
0%
21
Raleigh, NC
43
98
44%
9
565
2%
22
Phoenix, AZ
13
49
26%
12
363
3%
23
Pittsburgh, PA
72
155
46%
3
184
1%
24
Pittsburgh, PA
42
49
86%
2
246
1%
25
Charleston, SC
17
64
27%
14
444
3%
All Buildings
59
380
16%
6
494
1%
Table 2: Summary of Annual Heating and Cooling Loads
15
The results indicate that, nationwide, air infiltration is responsible for about 1 6% of the total
annual heating load of the office building stock, but only 1% of the cooling load. One reason for
the disparity between heating and cooling percentages is clear from Equation (a) of Appendix A,
which shows that sensible loads are directly proportional to the inside-outside temperature
difference, AT. Therefore, it is to be expected that heating loads due to infiltration are far greater
than cooling loads, due to the larger values of AT that occur during the heating season.
Furthermore, cooling loads are strongly dependent on the heat generated by internal sources, and
these sources tend to increase the amount of cooling necessary, but decrease the amount of
heating. The end result is that a significantly greater portion of the heating load arises from air
infiltration than of the cooling load.
A closer look at the results for individual building categories reveals that the percentage of the
heating load due to air infiltration varies widely from building to building. Much of the
variability of the percentage due to air infiltration was due to variability of the total heating load
not the infiltration heating load. For example, despite three of the buildings sharing a common
climate (Pittsburgh), the heating load due to infiltration for buildings 13, 23, and 24 varied from
6% to 86% of the total heating load. For these buildings, the infiltration heating load varied only
from 42 to 78 MJ/m^ but the total load varied from 49 to 1357 MJ/m^ The estimated percentage
for ail five of the recent and future building classes (21 through 25) are significantly above the
mean of 16%. In the PNL analysis, these buildings were assumed to meet the building energy
efficiency guidelines of ASHRAE Standard 90.1-1989 (ASHRAE 1989). The more stringent
envelope insulation values prescribed therein decrease conductive losses, making infiltration
loads a higher percentage of the total. In buildings 13 and 15, infiltration is a far smaller
percentage of the heating load than the average, partly because the HVAC systems of these
buildings operate for 24 hours per day. This had the dual effect of eliminating thermostat
setbacks, thus increasing the total heating load, and reducing the infiltration loads because the
building is pressurized day and night.
The estimated annual energy use for heating infiltration air in US. office buildings is 0.074 EJ
(0.070 quadrillion BTU) which is about 18% of the total heating energy use calculated by PNL.
The estimated annual energy use for cooling infiltration air in US. office buildings is 0.0025 EJ
(0.0024 quads) which is 2% of the total cooling energy use calculated by PNL. Among the
buildings representing recent and future construction (between 1980 and 1995), air infiltration
accounted for 45% of the heating energy use, showing the increasing relative cost of air
infiltration in newer, better insulated buildings. This method yielded an estimate of the
infiltration heating and cooling loads for the office building stock only. However, this is one of
the largest commercial building market segments and gives an indication of the magnitude of
energy involved in the issue of building airtightness and ventilation system control.
16
Summary
US office buildings consume a large amount of energy each year, a substantial portion of which
is used by building heating, ventilating, and air-conditioning systems. These energy costs may be
reduced through tighter building envelopes and improved ventilation system control. Better
information on the actual energy impacts of building leakage and poorly controlled ventilation
systems is needed to determine the cost-effectiveness of improved airtightness and system
control. However, commonly used building energy analysis programs exhibit several
shortcomings in modeling the airflow in multizone buildings which limit their usefulness in
studying this problem.
NIST has developed a research plan to quantify, and assess opportunities to reduce, the energy
and indoor air quality impacts of building airtightness and ventilation system control. The energy
impacts of poor building airtightness and ventilation system control will be analyzed with the
TRNSYS simulation program employing the existing building energy balance module and a new
module based on the multizone airflow program AIRNET. The program will be used to model
several prototypical large buildings in representative US climates. The simulation results will be
extrapolated to the commercial building population as a whole based on available statistical
information. The simulation tool will also be used to evaluate potential building envelope design
changes and retrofits in combination with other building design and operation features through
parametric analysis. This simulation tool will also be available to use for indoor air quality
simulation studies including determining both the indoor air quality and energy impacts of
modifications to building design and operation.
An initial estimate indicates that infiltration is responsible for 1 8% of the total heating energy use
and 2% of the total cooling energy use in US office buildings.
17
References
ASHRAE. ASHRAE Standard 90.1-1989 Energy-efficient design of new buildings except
low-rise residential buildings (1989) American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc.
Axley J. "A Discrete Thermal Analysis Method (DTAM) for Building Energy Simulation with
DTAMl Users Manual" (1988) National Institute of Standards and Technology.
Axley J and Grot J. "The Coupled Airflow and Thermal Analysis Problem in Building Airflow
System Simulation" (1989) ASHRAE Transactions Vol. 95, American Society of Heating
Refrigeration and Air-Conditioning Engineers, Inc.
Briggs R, Crawley D, and Belzer D. Analysis and Categorization of the Office Building Stock
GRI-87/0244 (1987) prepared by Battelle, Pacific Northwest Laboratory for Gas Research
Institute.
Briggs R, Crawley D, and Schliesing JS. Energy Requirements for Office Buildings - Volume 1
Existing Buildings (1992) GRI-90/0236.1 prepared by Battelle, Pacific Northwest Laboratory for
Gas Research Institute.
Clarke J. Energy simulation in building design (1985) Adam Hilger Ltd.
Clarke JA and Hensen JLM. "An approach to the simulation of coupled heat and mass flow in
buildings" (1990) Proceedings of the 1 1th AIVC Conference, Air Infiltration and Ventilation
Centre.
Clarke J and McLean. "ESP - A building and plant energy simulation system" (1988) University
of Strathclyde ABACUS report.
Crawley D and Schliesing J. Energy Requirements for Office Buildings - Volume 2 Recent and
Future Buildings (1992) GRI-90/0236.2 prepared by Battelle, Pacific Northwest Laboratory for
Gas Research Institute.
Curtis R, Birdsall B, Buhl W, Erdam E, Eto J, Hirsch J, Olson K, And Winkelmann F. "DOE-2
building energy use analysis program" LBL- 18046. (1984) Lawrence Berkeley Laboratory.
Dorer V and Weber A. Simulation of Passive Cooling and Natural Facade Driven Ventilation
(1994) Proceedings of the 15th AIVC Conference, Air Infiltration and Ventilation Centre.
EIA. Energy End-Use Intensities in Commercial Buildings (1994) Report No.
DOE\EIA-0555(94)/2, Energy Information Administration.
EIA. Commercial Buildings Energy Consumption Survey (1992) Energy Information
Administration.
18
Emmerich SJ. Indoor Air Quality Modeling and Demand Controlled Ventilation MS Thesis
(1993) Solar Energy Laboratory, University of Wisconsin - Madison.
Eto J. "The HVAC Costs of Increased Fresh Air Ventilation Rates in Office Buildings, Part 2"
(1990) Proceedings of Indoor Air 90.
Eto J and Meyer C. "The HVAC Costs of Increased Fresh Air Ventilation Rates in Office
Buildings" (1988) ASHRAE Transactions Vol. 94 Pt. 2, American Society of Heating,
Refrigerating, and Air-Conditioning Engineers, Inc.
Feustel H and Dieris J. A Survey of Air Flow Models for Multizone Structures LBL-30288.
(1991) Lawrence Berkeley Laboratory.
Feustel H and Raynor-Hoosen A (Editors). "Fundamentals of the Multizone Air Flow Model -
COMIS" AIVC Technical Note 29 (1990) Air Infiltration and Ventilation Centre.
Fischer R and Cudnik R. "The HOUSE-H Computer Model for Dynamic and Seasonal
Performance Simulation of Central Forced- Air Systems in Multi-zone Residences" (1993)
ASHRAE Transactions Vol. 99 Pt. 1, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc.
Friedrich M and Messinger M. Method to assess the gross annual energy savings potential of
energy conservation technologies used in commercial buildings (1994) Report No.
PNL-SA-23224, Pacific Northwest Laboratory.
Grot RA and Persily AK. "Measured Air Infiltration and Ventilation Rates in Eight Large Office
Buildings" (1986) Proceedings of Measured Air Leakage of Buildings, ASTM STP 904,
American Society for Testing and Materials.
Hadley DL and Jamagin RE. Selecting Representative Climate Stations for Use in a Building
Energy Model (1993) Report No. PNL-SA-20808, Pacific Northwest Laboratory.
Hensen J. ESPmfs, a building & plant mass flow network solver collaborative FAGO/ESRU
report 90.14.K (1990) Eindhoven University of Technology - Group FAGO and University of
Strathclyde - Energy Simulation Research Unit.
Huang J, Akbari H, Rainer L and Ritschard R. 481 Prototypical Commercial Buildings for 20
Urban Market Areas (1991) Report No. GRI-90/0326 prepared by Lawrence Berkeley
Laboratory for Gas Research Institute.
Jeandel A and Palero I. "Thermal modelling and simulation of buildings at Gaz de France"
(1991) Proceedings of the Third International Conference on System Simulation in Buildings.
19
Kendrick J. An Overview of Combined Modelling of Heat Transport and Air Movement. ATVC
Technical Note 40 (1993) Air Infiltration and Ventilation Centre.
Klein S. TRNSYS - A Transient System Simulation Program Engineering Experiment Station
Report 38-13. (1992) Solar Energy Laboratory, University of Wisconsin - Madison.
Klobut K. Distribution of Contaminants in Buildings by Air Recirculation and Other Airflows
Report B31. (1991) Helsinki University of Technology.
Kdobut K, Tuomaala P, Siren K, and Seppanen O. "Simultaneous Calculation of Airflows,
Temperatures and Contaminant Concentrations in Multi-zone Buildings" (1991) Proceedings of
the 12th ATVC Conference, Air Infiltration and Ventilation Centre.
Lomas KJ. Data sets for validating thermal models of buildings. (1991) BEPAC Technical Note
91/4, Building Environmental Performance Analysis Club.
McQuiston FC and Spitler, JD. Cooling and Heating Load Calculation Manual (1992) American
Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.
Melo C. "Convection and infiltration modelling for the built environment" (1986) Proceedings of
the 1 0th CIB Congress.
Modera M and Jansky R. "Residential Air-Distribution Systems: Interactions with the Building
Envelope" (1992) Proceedings of the Thermal Performance of the Exterior Envelope of
Buildings V.
Moser A. "Numerical Simulation of Room Thermal Convection - Review of lEA Annex-20
Results" (1992) Proceedings of International Symposium on Room Air Convection and
Ventilation Effectiveness.
Norford, LK. An Analysis of Energy Use in Office Buildings: The Case of ENERPLEX (1984)
Ph.D. Thesis, Center for Energy and Environmental Studies, Princeton University.
Park C, Clark D, and Kelly G. HVACSIM+ Building Systems and Equipment Simulation
Program: Building Loads Calculation NBSIR 86-3331. (1986) National Bureau of Standards.
Pelletret R. "Internal Heat Transfers and Heating Needs of Buildings" (1987) Proceedings of the
3rd International Congress on Building Energy Management.
Pelletret R and Khodr H. "Development and Validation of New Aeraulic Model Designed for
Thermal Computation with Particular Attention to the Problems of Air Quality" (1990) Building
Research and Practice, The Journal of CIB.
20
Persily AK and Grot RA. "Pressurization Testing of Federal Buildings" (1986) Proceedings of
Measured Air Leakage of Buildings, ASTM STP 904, American Society for Testing and
Materials.
Persily AK and Norford LK. "Simultaneous Measurements of Infiltration and Intake in an Office
Building" (1987) ASHRAE Transactions Vol. 93 Pt. 2, American Society of Heating,
Refrigerating, and Air-Conditioning Engineers, Inc.
PNL. Recommendations for Energy Conservation Standards and Guidelines for New
Commercial Buildings, Volume III: Description of the Testing Process (1983) Report
DOE/NBB-0051/7 prepared by Battelle, Pacific Northwest Laboratory for US. Department of
Energy.
Schliesing JS, O'Neill PJ, Halverson MA, and Brambley MR. "Building operation: Opportunities
for improvement?" (1993) ASHRAE Journal Vol. 37, No. 11.
Schneider PS, Roux J-J, and Brau J. "Strategies for Solving the Air Flow - Thermal Problem in
Multiroom Buildings" (1995) Building and Environment 30:277-286.
Shaw CY and Reardon JT. "Changes in Airtightness Levels in Six Office Buildings" (1995)
Proceedings of Airflow Performance of Building Envelopes, Components, and Systems, ASTM
STP 1255, American Society for Testing and Materials.
Steele T and Brown M. "ASHRAE Standard 62-1989 Energy, Cost, and Program Implications"
(1990) Bormeville Power Administration.
Tamura GT and Shaw CY. "Studies on Exterior Wall Air Tightness and Air Infiltration of Tall
Buildings" (1976) ASHRAE Transactions Vol. 82, American Society of Heating, Refrigerating,
and Air-Conditioning Engineers, Inc.
Tuomaala P and Rahola J. "Combined Air Flow and Thermal Simulation of Buildings" (1995)
Building and Environment 30:255-265.
UIC. Analysis of Commercial Model Conservation Standards Study (1989) Report No. 8913
prepared by United Industries Corporation for Bonneville Power Administration.
Ventresca J. "Operation and Maintenance for Indoor Air Quality: Implications from Energy
Simulations of Increased Ventilation", (1991) Proceedings of lAQ 91, American Society of
Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
Walton G. AIRNET - A Computer Program for Building Airflow Network Modeling NISTIR
89-4072. (1989) National Institute of Standards and Technology.
Walton G. CONTAM93 - User Manual (1994) NISTIR 5385, National Institute of Standards and
Technology.
21
Walton GN and Emmerich SJ. "CONTAM93: A Multizone Airflow and Contaminant Dispersal
Model with a Graphic User Interface" (1994) Air Infiltration Review 16:6-8, Air Infiltration and
Ventilation Centre.
Zmeureanu R, Fazio P, and Zhaonan M. "Impact of Increase of the Air Ventilation Rate on the
Energy Performance of a Large Office Building in Montreal" (1992) Proceedings of the 5th
International Jacques Cartier Conference.
22
Appendix A: Description of Method for Initial Estimate
This appendix describes the method used to develop the initial estimate of the national energy
use in US office buildings due to infiltration. The results of the initial estimate are included in the
section of the report titled Initial Estimate Based on PNL Infiltration Rates.
Introduction
A simple approach was used to calculate the cumulative annual load due to heating and cooling
of infiltrating air in office buildings nationwide. The leakage characteristics of a given building
were used, in conjunction with hourly weather data from the WYEC tape for an appropriate
climate, to estimate the volume of outdoor air that penetrates the building envelope during a
given hour. The load associated with heating or cooling this air to the thermostat setpoint of the
building was summed over every hour of the year in order to find annual loads for the building.
Infiltration loads were calculated in this manner for a set of 25 buildings, each representing a
certain percentage of the total building stock of the United States. Twenty of these buildings were
developed by Briggs, Crawley, and Belzer (1987) to represent the existing office building stock
as of 1979, and are summarized in the report “Energy Requirements for Office Buildings”
(Briggs, Crawley, and Schliesing 1 992). The other 5 buildings represent construction between
1980 and 1995, and are described in volume 2 of the same report (Crawley and Schliesing 1992).
A summary of features of the 25 representative buildings are shown in Table A-1. The two
volumes of this report include an estimate of the total heating and cooling coil loads experienced
annually in each of the 25 buildings, obtained using the DOE-2 building energy simulation
program. By matching the important parameters in the calculations of energy use due to
infiltration as closely as possible with those used in the PNL analysis, it was possible to compare
these results to the earlier predictions of total loads to estimate the percentage of the total annual
load that is attributable to air infiltration.
23
Bldg.
No.
Floor Area
(m^)
No. of
Floors
Year Built
Location
Floor Area
Represented -
(10‘ m')
Air Change Rate w/
Fans Off
(h-')
1
576
1
1939
Indianapolis, IN
15.6
0.53
2
604
3
1920
Cleveland, OH
24.8
1.00
3
743
1
1954
El Paso, TX
21.5
0.43
4
929
2
1970
Washington, DC
26.5
0.33
5
1486
2
1969
Madison, WI
51.7
0.28
6
2044
2
1953
Lake Charles, LA
31
0.42
7
2601
4
1925
Des Moines, lA
68.2
0.65
8
3716
5
1908
St. Louis, MO
28.3
0.70
9
3902
2
1967
Las Vegas, NV
43.2
0.18
10
4273
3
1967
Salt Lake City, UT
35.5
0.21
11
13935
6
1968
Cheyenne, WY
28.6
0.19
12
16722
6
1918
Portland, OR
27.9
0.54
13
26941
11
1929
Pittsburgh, PA
58.5
0.62
14
26941
6
1948
Amarillo, TX
37.3
0.31
15
27870
12
1966
Raleigh, NC
32.7
0.22
16
28799
10
1964
Dallas, TX
22.9
0.18
17
53882
19
1965
Minneapolis, MN
27.6
0.26
18
67817
10
1957
Boston, MA
16.3
0.16
19
68746
28
1967
New York, NY
43.4
0.32
20
230392
45
1971
Los Angeles, CA
40.8
0.26
21
1022
2
1986
Raleigh, NC
117
0.62
22
1208
2
1986
Phoenix, AZ
92.2
0.58
23
1579
2
1986
Pittsburgh, PA
101
0.53
24
38089
9
1986
Pittsburgh, PA
64.5
0.21
25
46450
14
1986
Charleston, SC
54
0.23
Note: Each of buildings 21-25 represents a mix of building construction in 1986 and 1995.
Table A-1: Summary of Representative Building Set
Method
The algorithm for calculating infiltration loads for a given building consists of the following
steps:
1 . Obtain weather conditions for the current hour: outdoor temperature, humidity, and wind
speed.
2. Determine the appropriate air infiltration rate, based on wind speed and HVAC system status.
3. Determine the appropriate thermostat setpoints of the HVAC system, depending on the
building occupancy schedule.
24
4. Compare the temperature of the outdoor air with the thermostat setpoints to determine
whether the infiltrating air needs to be heated or cooled.
5. If cooling is necessary, compare the humidity of the outdoor air to the desired humidity to
determine whether latent cooling loads will be present.
6. Calculate the hourly loads using equations (a) and (b) (ASHRAE 1993).
(a) Q, = p*Cp*AT*ACH*V
(b) Q, = p*h,/AW*ACH*V
7. Add the hourly infiltration load to the cumulative total for either the heating or cooling load.
In equations (a) and (b), is the sensible heating or cooling load due to infiltration, Q, is the
latent cooling load, p is the density of the infiltrating air, Cp is the specific heat of the infiltrating
air, AT is the indoor-outdoor temperature difference, AW is the indoor-outdoor humidity ratio
difference, ACH is the infiltration rate in air changes per hour, and V is the total volume of the
building. ACH * V, therefore, represents the volume of outdoor air that enters the building in one
hour.
Application of this algorithm required some assumptions regarding the leakage characteristics of
the building and the HVAC system parameters, most notably the operating schedule and
temperature and humidity setpoints. Whenever possible, the values of these parameters were
taken directly from the input files for the PNL analysis (Briggs et al. 1992). However, in the
cases of indoor humidity levels and HVAC system balance temperatures, no specific information
was available, so additional assumptions were necessary.
Air Infiltration Rates
Air infiltration rates for each of the representative buildings were generated by Briggs et al. for a
wind speed of 10 miles per hour (4.47 m/s), using a model that takes into account building age
and height and an average annual indoor-outdoor temperature difference. For the infiltration load
calculations, these values were scaled linearly with wind speed to generate a table of infiltration
rates for each building for wind speeds between 0 and 20 m/s. Because the PNL analysis did not
account for the dependence of the air infiltration rate on the indoor-outdoor temperature
difference, this dependency was not included in the present analysis. The values in Table 1
represent air infiltration rates that were used when the LTV AC system fans are off. During hours
of fan operation, the resulting pressurization of the building may act to reduce the rate of air
infiltration to some degree. Following Briggs et al. (1992), the amount of this reduction was
based on the height of the building: for buildings of 5 stories or fewer, air infiltration was
reduced to 25% of the fans-off rate, and in taller buildings it was reduced to 50% of the fans-off
rate. Building number 2 has no mechanical ventilation so the infiltration rate was not reduced.
HVAC System Parameters
Due to the assumed effect of building pressurization on the air infiltration rate, it was necessary
to know whether or not the HVAC system fans were running during any given hour of the day.
The PNL descriptions of the representative office buildings include the average number of hours
25
per day that the HVAC systems operate, which ranges between 9.2 and 21 . For each value of this
parameter, a detailed schedule is provided, indicating which hours the fans are considered to be
running. When calculating the loads during such an hour, the infiltration rate was reduced as
detailed earlier to account for building pressurization. Different schedules were utilized for
weekdays and weekends.
The temperature setpoints were designed to reflect the common practice of changing thermostat
settings in order to conserve energy at times when the building is expected to be unoccupied.
Heating setbacks were 2.8 °C (5 °F) below the corresponding occupied-hours heating setpoints,
which ranged from 21.1 °C (70 °F) to 22.2 °C (72 °F). Setpoints for cooling fell between 23.3 °C
(74 °F) and 25.0 °C (77 °F). Cooling setups were fixed at 37 °C (99 °F) for every building,
essentially ensuring that no cooling would occur during unoccupied hours. All of these values
were taken directly from the corresponding DOE-2 input parameters. Schedules similar to those
describing the hours of HVAC system operation were used to determine whether the high or low
setpoint should be used for each hour's calculations. In general, setbacks and setups were in
effect from the time the HVAC system fans cut off in the evening until one hour before they
restarted in the morning. The existing building descriptions do not include a setpoint, per se, for
the humidity of the indoor air. However, the input files for the system subprogram of DOE-2
include a listing for the maximum humidity of the system air. When calculating latent cooling
loads, it was assumed that all infiltrating air that needed to be cooled was also dehumidified to
the maximum level indicated for that building. The maximum level was 70% relative humidity
for the 20 original buildings, and 60% for the 5 buildings representing recent and future
construction.
Balance Points
Another building parameter was introduced to account for the presence of internal heat sources,
such as occupants, lighting, and electrical equipment. At times when the outdoor temperature is
below the thermostat setpoint by a small amount, infiltrating air may not need to be mechanically
heated due to the heat generated by internal sources. The temperature above which this is true is
called the balance temperature, or balance point, of the building. In order to include the ‘free’
heating effect of a building's internal heat sources, a balance temperature was assigned to each of
the representative buildings. If the temperature of infiltrating air fell between the balance
temperature and the heating setpoint, no heating load was assessed during that hour. A balance
temperature was estimated for each building based on properties provided in the PNL input files,
using the following equation (ASHRAE 1993):
t — t _^M!L
^bal - h
^tot
The total rate of heat gain, q^^^, includes internal sources such as occupants, lighting, and
equipment, solar gains through fenestration, and radiative gains through the walls and roof is
the total heat loss coefficient of the building (in W/K) due to infiltration, ventilation, and
conduction. If one assumes that heat transfer among the zones of a building is negligible (PNL
divided the buildings into thermal zones for DOE-2), then each zone will exhibit its own
characteristic balance temperature. Since most heat loss occurs across the building envelope, the
limiting balance temperature (the highest) will be that of the zones having exterior walls. For this
26
reason only the internal heat sources in the perimeter zones were included in the heat gain term
when calculating the balance point for multizone buildings. For each building, a separate balance
point was calculated for unoccupied hours. These estimates assumed no solar or radiative heat
gains since unoccupied hours generally occur at night. Receptacle loads were assumed to be 50%
of their occupied-hours level and lighting loads 25%, while occupancy was at 5% of the
maximum, based on the sample schedules created by PNL. At both times, the interior
temperature t^ was assumed to be equal to the current thermostat setpoint. Balance point
temperatures for the 25 prototypical buildings ranged from -5.5 to 15 °C (22 to 60 °F) during the
day, and from 10 to 17 °C (50 to 62 °F) at night, with averages of 4.5 °C (40 °F) and 14 °C (57
°F), respectively. These temperatures are in the same range as balance points calculated for a
modem office building of 1.1 °C (34 °F) for weekday hours, 2.8 °C (37 °F) for weekend day
hours, and 11.1 °C (52 °F) for night hours (Norford 1984).
Conversion of Coil Loads to Energy Use
The PNL results (Briggs et al. 1 992) include an analysis of cumulative annual energy use,
accounting for conversion efficiencies of HVAC system components and the source of energy
(electricity or gas). By comparing these values of energy use to the corresponding coil loads, a
single number was obtained to represent the system's overall conversion efficiency for heating,
and one for cooling. These energy use-to-coil load ratios were then applied to the infiltration
loads, yielding an estimate of the annual energy cost of air infiltration.
References
ASFIRAE. ASHRAE Standard 90.1-1989 Energy-efficient design of new buildings except
low-rise residential buildings (1989) American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc.
ASHRAE. 1993 ASHRAE Handbook: Fundamentals (1993) American Society of Heating,
Refrigerating, and Air-Conditioning Engineers, Inc.
Briggs R, Crawley D, and Schliesing J. “Energy Requirements for Office Buildings - Volume 1
Existing Buildings” GRI-90/0236.1 (1992) Gas Research Institute.
Briggs R, Crawley D, and Belzer D. “Analysis and Categorization of the Office Building Stock”
GRI-87/0244 (1992) Gas Research Institute.
Crawley D, and Schliesing J. “Energy Requirements for Office Buildings - Volume 2 Recent and
Future Construction” GRI-90/0236.2 (1992) Gas Research Institute.
Norford, LK. An Analysis of Energy Use in Office Buildings: The Case ofENERPLEX (1984)
Ph.D. Thesis, Center for Energy and Environmental Studies, Princeton University.
27
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11