&EPA
   United States
   Environmental Protection
   Agency
                  Airtightness Evaluation of
                  Shelter-in-Place Spaces for
                  Protection Against Airborne
                  Chembio Releases
                  REPORT


0 ' H
Restroom
1 <

Elevator
J
[] ®
Resfroom4
2
H
'H.
A
° 11
1
,
t
                        -«~0-
                         Lobby
                                  \
                                    Stairs
   Office of Research and Development
   National Homeland Security Research

-------

-------
                                                           March 2009 EPA/600/R-09/051

                                                                      NISTIR 7546
                                                                     MIST
                                                                  National Institute of
                                                             Standards and Technology
                                                              U.S. Department of Commerce
        Airtightness  Evaluation  of

        Shelter-in-Place Spaces for Protection

        Against Airborne Chembio  Releases

        Andrew Persily
        Heather Davis
        Steven J. Emmerich
        W. Stuart Dols
        Building and Fire Research Laboratory


        Prepared for:
        U.S. Environmental Protection Agency
        Research Triangle Park, NC


        Funded under IAG DW-13-92178301-0 by the U.S. EPA National Homeland Security Research Center
        (NHSRC), Decontamination and Consequence Management Division (DCMD)
                                                          U.S. Department of Commerce
                                                                 Gary Locke, Secretary

                                                National Institute of Standards and Technology
                                                       Patrick D. Gallagher, Deputy Director
Office of Research and Development
National Homeland Security Research Center, Decontamination and Consequence Management Division

-------
Abstract
Due to concerns about potential airborne chemical and
biological releases in or near buildings, building owners and
managers and other decision makers are faced with a number
of options for increasing their buildings' level of protection
against such events. Among the various technologies and
approaches being proposed and implemented is shelter-
in-place (SIP). SIP strategies involve having the building
occupants stay in the building, generally in a space
designated for such sheltering, until the event is over and
the  outdoor contaminant levels have decreased. While
much guidance is available on the implementation of SIP in
buildings, important technical  issues remain about the degree
of protection provided by a particular space and the factors
that determine the level of protection. In particular, many
recommendations suggest tightening the walls of SIP spaces,
but there has been insufficient  analysis of the relationship
between shelter tightness and the protection provided by
the  SIP space.
In order to address some of these questions, the National
Institute of Standards and Technology (NIST) has undertaken
a project for the U.S. Environmental Protection Agency to
develop and demonstrate evaluation methods to relate shelter
airtightness to the performance of shelter-in-place approaches
for  airborne chemical, biological, and radiological (CBR)
protection of building occupants. The focus of this effort
is on short-term sheltering, on  the order of hours, rather
than longer-term sheltering,  which generally employs
filtration and air cleaning equipment to supply clean air to
the  occupants of the space. This project has consisted of
the  following tasks: a literature review of SIP strategies and
performance issues; development of a study plan for testing
SIP airtightness evaluation methods; implementation of
the  study plan through a combination of experiments and
simulations; and, finally, development of recommendations
on SIP evaluation and possible performance criteria for
candidate SIP spaces.
In actual application, a building owner or manager would
select spaces for use as shelters based on a number of
qualitative considerations identified previously and perhaps
make modifications to increase the degree of protection
offered by the shelter. In the case of unventilated shelters
intended for short-term sheltering, which are the subject of
this study, a key modification is to increase the airtightness
of the shelter through sealing of the boundaries to adjacent
spaces. This project has focused on the relationship of
shelter and building airtightness to the protection provided,
with space pressurization testing examined as an evaluation
method.
Room pressurization testing was seen to be relatively
straightforward as applied in this study. While it is based on
a standard test method for whole building airtightness testing,
its application to individual rooms has yet to be standardized.
Nevertheless, the simulations in this report showed that
interzone pressure effects and system operation in non-test
zones impacted the measurement results by about 10%.
The measured values of airtightness were surprisingly
consistent among shelters tested, as well as the percentage
increase in airtightness through sealing. Under limited sealing
the effective leakage area (ELA) values were in a relatively
narrow range from somewhat above 1 cnvYm2 to about
5 cmVm2. The sealed values were lower, as expected, and
ranged from 0.25 cmVm2 to just under 1 cm2/m2. The percent
reduction due to sealing was surprisingly consistent for the
four spaces, ranging from about 60% to 90%.
The simulation of occupant exposure showed that tighter
buildings and tighter shelters reduce exposure of SIP
occupants, with shelter tightness having a greater impact
than building tightness. Longer-duration sheltering reduced
the protection factor as expected, which highlights the
importance of obtaining and communicating reliable
information on when the outdoor hazard has ended and it
is time to end the sheltering period. Based on these
simulation results, CO2 buildup over time may be more
critical than the reduction in exposure. For the building
and shelter airtightness values considered, the duration
of sheltering was seen to be  more important to the shelter
CO2 level than airtightness. None of the predicted CO2
concentrations were as high as the American Conference
of Governmental Industrial Hygienists (ACGIH) short-term
(1 min) exposure limit of 54,000 mg/m3, but several exceeded
the threshold limit value (TLV), based on an 8-h exposure
over a 40-h work week, of 9000 mg/m3 [42]. The TLV is not
really relevant to a 1-h or 2-h exposure, but the high values
seen in the simulations are of potential concern. Occupant
density, or floor area per shelter occupant, is obviously
an important determinant of these CO2 concentrations.
Many guides recommend 1 m2 per occupant, but this
value produced high CO2 concentrations for many of the
simulations in this study. A lower value of occupant density,
e.g., 2 m2 per occupant, might provide more habitable
conditions for longer periods of time and merits consideration
in future recommendations.
Keywords: building protection, CBR, chembio, shelter-in-
place, SIP

-------
Table of  Contents
1. Introduction	1
2. Literature Review	3
      2.1 Research Reports	3
      2.2 CSEPP and Other FEMA Documents	4
      2.3 Other Guidance Documents	5
      2.4 Summary of Review	5
3. Evaluation Approach and Study Plan	7
      3.1 Theory and Calculation Approaches	7
      3.2 SIP Zone Airtightness Tests	9
      3.3 CONTAM Predictions of Protection Factor	12
      3.4 CONTAM Simulations of SIP Leakage Measurements	13
      3.5 Validation of Predictions from Shelter Airtightness Values	13
4. Study Results	15
      4.1 SIP Airtightness	15
      4.2 CONTAM Predictions of Protection Factor	16
      4.3 CONTAM simulations of SIP Leakage Measurement	30
      4.4 Comparison of Predictions to Measurements	32
5. Discussion	33
6. Acknowledgments	35
7. References	37

-------
 List  of   Figures
Figure 1.  Dose reduction factor (DRF) as a function of time and air change rate	4
Figure 2.  Protection factor as a function of time and air change rate	5
Figure 3.  Schematic of two-zone SIP model	7
Figure 4.  Contaminant concentration values for two-zone example	8
Figure 5.  Protection factor values for two-zone example	9
Figure 6.  SIP pressurization test schematic	9
Figure 7.  Blower door	10
Figure 8.  Sample plot of pressurization test data	11
Figure 9.  Two-story building floor plan	13
Figure 10. Example plot of simulated agent and CO2 concentrations forthe one-zone building	16
Figure 11. Simulated agent concentrations in shelter for multiple shelter leakage values	17
Figure 12. Protection factors, referenced to building, for one-zone model	20
Figure 13. Protection factors, referenced to outdoors, for one-zone model	20
Figure 14. Shelter CO2 concentration for one-zone model - cold/windy weather	23
Figure 15. Protection factors, referenced to building, for office building model	25
Figure 16. Protection factors, referenced to outdoors, for office building model	25
Figure 17. Shelter CO2 concentration for office building model - cold/windy weather	26
Figure 18. Protection factors, referenced to building, for ten-story building model	28
Figure 19. Protection factors, referenced to outdoors, for ten-story building model	28
Figure 20. Shelter CO2 concentration for ten-story model - cold/windy weather	29
Figure 21. CONTAM sketchpad of simulated pressurization tests	30

-------
List  of  Tables
Table 1.  Test site and room dimensions	12
Table 2.  SIP airtightness test results	15
Table 3.  Protection factors for one-zone model (cold and windy weather)	18
Table 4.  Protection factors for one-zone building model (mild and calm weather)	19
Table 5.  Carbon dioxide concentrations  for one-zone model (cold and windy weather)	21
Table 6.  Carbon dioxide concentrations  for one-zone model (mild and calm weather)	22
Table 7.  Air change rates for two-story office building model	24
Table 8.  Air change rates for ten-story building model	27
Table 9.  Results of simulated SIP tests	31
Table 10. Comparison of measured and predicted decay rates	32

-------
 List  of  Acronyms
AC        air conditioning
ACGIH    American Conference of Governmental Industrial Hygienists
ASHRAE   American Society of Heating, Refrigerating and Air-Conditioning Engineers
BFRL      Building and Fire Research Laboratory (part of NIST)
CBR       chemical, biological and radiological
CP        collective protection
CONTAM  NIST-developed multizone airflow and contaminant dispersal simulation program
CSEPP     Chemical Stockpile Emergency Preparedness Program (FEMA program)
DRF       dose reduction factor
ELA       effective leakage area
EPA       U.S. Environmental Protection Agency
FEMA     Federal Emergency Management Agency
HVAC      heating, ventilating, and air-conditioning
IAQ       indoor air quality
MERV     minimum efficiency reporting value, metric of particle filtration efficiency based on ASHRAE Standard 52.2
NIOSH     National Institute of Occupational Safety and Health
NIST      National Institute of Standards and Technology
ORNL     Oak Ridge National Laboratory
PF        protection factor
SIP        shelter-in-place
TLV       threshold limit value

-------
                                                                                                           1.0
                                                                                       Introduction
Due to concerns about potential airborne chemical and
biological (chembio) releases in or near buildings, building
owners and managers and other decision makers are faced
with a number of options for increasing their buildings'
level of protection against such events [1]. A wide range of
technologies and approaches is being proposed with varying
levels of efficacy and cost, as well as varying degrees of
applicability to any particular building. In particular, shelter-
in-place (SIP) has been proposed as a strategy to protect
building occupants from chembio releases, particularly
outdoor releases. SIP strategies involve having the building
occupants stay in the building, generally in a space designated
for such sheltering, until the event is over, the outdoor
contaminant levels have decreased, and it is safe to leave the
building. SIP is often considered as an alternative to building
evacuation under conditions where the outdoor exposure  is
likely to be higher than the exposure in the shelter. While
much guidance is  available on the implementation of SIP in
buildings [2], important technical issues remain about the
degree of protection provided by a particular space and the
factors that determine the level of protection. Additional
questions exist regarding the appropriate duration of
occupancy based on concerns regarding oxygen depletion,
carbon dioxide buildup, and exposure to the chembio agent
over time, as well as the role of nitration and air cleaning in
providing additional protection by pressurizing the SIP space
with clean air. Also, while many recommendations suggest
tightening the partitions to adjacent spaces, there has not been
sufficient analysis of the relationship between the shelter
tightness and the protection provided by the SIP space nor
any recommended quantitative airtightness criteria.
In order to address some of these questions, the National
Institute of Standards and Technology (NIST) has undertaken
a project for the U.S. Environmental Protection Agency to
develop and demonstrate evaluation methods to relate shelter
airtightness to the performance of shelter-in-place approaches
for chemical, biological, and radiological (CBR) protection
of building occupants. The focus of this effort is on short-
term sheltering, on the order of hours, rather than longer-
term sheltering, which generally employs nitration and air
cleaning equipment to supply clean air to the occupants of the
space. However, the results of this effort still have application
to longer-term sheltering as the airtightness of these shelters
determines the amount of clean air supply required to
maintain the shelter at positive pressure.
This project has consisted of the following tasks: conducting
a literature review of SIP strategies and performance
issues; developing a study plan for testing SIP airtightness
evaluation methods; implementing the study plan through
a combination of experiments  and simulations; and, finally,
developing recommendations on SIP evaluation and possible
performance criteria for candidate SIP spaces.
This report is organized by these tasks, with the first section
presenting the results of the literature review. The next
section describes the experimental and other analytical
approaches used in this study,  followed by a section with the
results of those efforts. A series of recommendations for SIP
space evaluation is presented in the final section of the report.

-------

-------
                                                                                                         2.0
                                                                          Literature   Review
In order to support the technical work involved in this
project, a review of the existing literature on shelter-in-
place was conducted. The literature review included several
guidance documents intended for practitioners, to identify
what measures have been proposed to evaluate candidate
SIP spaces for the degree of protection they offer against an
outdoor release. While the review was focused primarily on
finding quantitative evaluation methods to judge candidate
spaces, the review also provided an opportunity to collect
qualitative considerations proposed for SIP spaces.
This review covered a range of documents, which are divided
into three categories: research reports and papers; Chemical
Stockpile Emergency Preparedness Program (CSEPP) and
other Federal Emergency Management Agency (FEMA)
documents; and other guidance documents. The first category
includes the results of several research studies and analyses
of the effectiveness of shelter-in-place protection strategies,
including some experimental studies. The second category
contains a number of reports produced under the FEMA
CSEPP program, which exists to produce information,
guidance, and training material "for formulating and
coordinating emergency plans and the associated emergency
response systems for chemical events that may occur at the
chemical agent stockpile storage locations in the continental
United States." The third category of documents reviewed
consists of SIP guidance material produced by a number of
organizations, both for public education and for building
and system design.

2.1 Research  Reports
SIP research provides an indication of the protection offered
by sheltering within a building, in some cases addressing
the additional protection provided by building tightening or
by filtration and air cleaning. This work includes field tests
[3-6], theoretical analyses [7-9], and simulation studies [10,
11]. The field tests tend to focus on the sheltering provided
by a building, not on the additional sheltering provided by
an SIP space within a building. letter and Whitfield [12]
examined the protection provided by an interior bathroom of
a residence, in which tracer gas tests were used to estimate
airflow rates between the shelter, the rest of the house, and
the outdoors. In a more recent study, these airflows were then
used to estimate protection factors (ratio of outdoor to shelter
exposure) as a function of time. letter and Proffitt conducted
similar tests  in commercial buildings, which are discussed in
more detail later in this report [13].
The theoretical work provides insight into the parameters that
determine the protection provided by the building as a whole
(i.e., airtightness, weather conditions, and particle deposition
rates) but does not focus on SIP spaces. By considering
the mass balance for a single zone, an equation relating
the dose reduction factor (DRF) relative to outdoors was
derived by Engelmann [7, 8] for a step change in the outdoor
concentration from zero to a nonzero value, as follows:
DRF = Indoor exposure/Outdoor exposure =
1 - (1/Rt) + (e-Rt)/Rt
(1)
where t is the time elapsed since the outdoor concentration
increase and R is the building air change rate in units of air
changes per hour, or fr1. Therefore, the lower the value of
DRF, the lower the occupant exposure relative to outdoors.
Figure 1 is a plot of DRF calculated with equation (1) for
a range of values of the building air change rate. Note that
as t increases, the indoor exposure approaches the outdoor
exposure and the DRF approaches 1. Also, lower air change
rates correspond to lower values of DRF, i.e., lower indoor
exposure, but the values of DRF still approach 1 after a
sufficient amount of time elapses. While the plot shows
that DRF approaches 1 regardless of the air change rate,
the  period of interest for most SIP applications discussed in
this report is only on the order of a few hours, in which case
the  different air change rates correspond to very different
DRF values. Similar analysis yields equations accounting
for particle  deposition and a subsequent step decrease in the
outdoor concentration.

-------
                 0            4

           Figure  1  Dose reduction factor (DRF) as a  function of time and air change rate
2.2 CSEPP and Other FEMA Documents
The CSEPP program, sponsored by FEMA, has produced
a significant amount of material relevant to SIP strategies,
much of it produced at Oak Ridge National Laboratory
(ORNL). The resulting publications include guidance
documents, experimental studies, and other materials.
One such publication, Blewett et al. [14], is fairly
comprehensive. This publication includes a literature review,
a discussion of different sheltering approaches, qualitative
criteria for room selection, and the results of sheltering tests
in twelve buildings. The four approaches described in this
reference, as well as many other publications, include the
following:
    Normal sheltering: closing all windows and doors, and
    turning off all mechanical equipment, such as heating,
    ventilating, and air-conditioning (HVAC) systems;
    Expedient sheltering:  applying temporary air sealing
    measures to a shelter  space, such as taping over vents
    or placing plastic sheeting over windows;
    Enhanced sheltering:  applying permanent air sealing
    measures to a shelter  space; and
    Pressurized sheltering: providing filtered/cleaned air to
    the shelter to achieve  an elevated air pressure relative to
    outside the shelter, thereby greatly limiting the entry of
    air and contaminant.
This report and others speak in terms of the protection factor
(PF), which is the outdoor dosage divided by the indoor
dosage at some point in time and therefore equal to the
inverse of the DRF defined above. Figure 2 is a plot of the
protection factor as a function of time, analogous to the DRF
plot in Figure 1. (Note that PF is higher for lower air change
rates and shortly after the release begins, tending towards
a value of 1.0 as time continues. Again, the plot extends for
many hours to show that all the curves tend toward a value
of 1, but the period of interest for most applications is on the
order of only a few hours.)
The experimental portion of the Blewett et al. study [14]
was focused on measurement of the airflow rate between
expediently sealed safe rooms and the outdoors,  as well as
the air change rate of the whole house. The results of these
measurements were used to estimate a range of protection
factors under normal and expedient sheltering associated with
a 10-min and a 1-h outdoor exposure. The 10-min values of
PF range from about 20 to 60 for the whole building, while
the values of PF for expedient sheltering in a bathroom were
almost twice as high. The 1-h values ranged from about
4 to 12 for the whole building and about 5 to 15 for the
bathrooms. The report also summarizes some SIP guidance
available at the time of the report and provides some
qualitative criteria for selecting a SIP space in a building.
The CSEPP has also produced guidance documents on
specific  sheltering approaches [15,  16]. Other reports out
of the CSEPP program are more policy related, discussing
issues such as planning for SIP, deciding between evacuation
and sheltering under a given scenario, managing building
occupants [17-22], and assessing the housing stock near
chemical storage sites  [23]. This latter document looks
primarily at the age of houses near these sites and discusses
airtightness in general but does relate building age to the
airtightness of these houses. Other FEMA documents provide
useful information, such as recommendations on floor area
per person in tornado shelters [24].

-------
/u •
1H
16 •

3 1-^
c 10
u
£ B
1



n .

Air changi
^—
	



\
A
K
" ^ ^~ ~~


: rate (h'1)
4
1
0.5
0 °5







8
                                                                       16
                         20
                                       12
                                 Time (h)
Figure 2  Protection factor as a  function of time and  air change rate
24
More recently, FEMA issued design guidance for shelters
and safe rooms [2]. While much of this report is devoted to
issues of protecting building occupants from the effects of
blast, it also contains important information on protection
from CBR airborne contaminants. FEMA distinguishes
among three levels of CBR protection in shelter, the first
being pressurization of the space combined with particle
filtration and/or gaseous air cleaning. The second class
includes filtration and/or air cleaning but with little or
no pressurization, and class 3 is passive, meaning no
air treatment or efforts to pressurize the shelter space.
The document speaks to airtightening of shelters, either
temporarily when an event occurs or permanently as part
of a preparedness effort. The guide notes that no airtightness
criteria exist for shelters but also notes some key air
leakage sites for sealing efforts and describes the use
of fan pressurization or blower door testing as a means
of quantifying shelter airtightness.

2.3 Other Guidance  Documents
In addition to the CSEPP program, other organizations have
issued guidance on the use of SIP as an exposure reduction
strategy, ranging from general guidance [25-30] to more
detailed design specifications [31, 32]. The latter two
documents from the U.S. Army Corps of Engineers contain
detailed design requirements for collective protection (CP),
identifying four classes of facilities based on their potential to
support integration of CP systems. These classes  range from
those with HVAC systems that are capable of integrating
an overpressure system to those that cannot be pressurized
without extensive sealing. These Army Corps documents
also  speak to floor area per person, air locks, envelope
leakage testing, and filtration systems in a fairly detailed
fashion. While these documents do not contain airtightness
specifications, they do call for a minimum overpressure of
75 Pa for the Class I collective protection and 5 Pa for Class
II. Some SIP guidance is also contained within more general
discussions of building protection strategies [33].

2.4 Summary of Review
While the existing literature and guidance on SIP is well
established and useful, this literature does not provide
quantitative methods for evaluating the degree of protection
that candidate SIP spaces might offer in the event of an
outdoor release. Useful qualitative guidance on selecting
such spaces exists, but there is little quantitative guidance
other than recommendations on the amount of floor area per
person.
The quantitative material that does exist employs the dose
reduction factor and its inverse, the protection factor, as
measures of the degree of protection offered by a shelter.
These parameters are defined in terms of the outdoor
exposure relative to the exposure in the building or SIP zone.
One could make the case that the SIP zone exposure should
instead be compared to exposure  in the rest of the building,
as opposed to the  outdoors, given that the occupants
might stay in the building if no SIP space were available.
Experimental studies have been conducted to evaluate SIP
protection offered by whole buildings or  specific building
spaces, but these studies have tended to be tracer gas
measurements of air change rates of the shelters, which do
not relate directly to exposure reduction.  Also, tracer gas
testing methods are too involved and costly for most
building owners and managers.

-------
As noted, there are many useful qualitative recommendations
for selecting a shelter space, addressing, for example, size,
location, and accessibility, with the recent FEMA guide [2]
providing a very thorough description. Blewett et al. [14]
includes the following attributes of SIP spaces in buildings:
above ground, interior room with few or no windows, no
plumbing fixtures if possible, no window AC units, at least
0.9 m2 (10 ft2) per person, and not rooms with an exhaust fan
linked to a light switch. Price et al. [33] note that the goal for
an SIP space is to create a zone where infiltration is very low,
which usually means being located in the interior portion
of a building, i.e., no windows, and having doors that are
fairly effective at preventing airflow from hallways. They
also note that bathrooms are usually a bad choice because
exhaust ducts can draw air into the room when there are stack
driven airflows (i.e., arising from indoor-outdoor temperature
differences) in a building, even when the fans are off. These
various  qualitative criteria for selecting SIP spaces are
summarized in Section 5 of this report.
While most of the SIP guidance material recommends
minimal air change with the outdoors and perhaps an effort
to increase the airtightness of the shelter space, none of these
documents makes specific recommendations on airtightness
levels and very few address the measurement of shelter air
leakage.

-------
                                                                                                      3.0
                             Evaluation  Approach  and  Study   Plan
While the existing literature acknowledges the importance
of shelter airtightness in protecting occupants against outdoor
releases, this literature does not describe SIP space airtightness
evaluation methods in any detail, nor does it present
airtightness criteria. In order to address this need, the current
project has pursued the evaluation of potential SIP spaces
by measuring shelter airtightness with the fan pressurization
method. The concept behind such an evaluation is that the
shelter airtightness is related to the airflows between the SIP
space, the rest of the building, and the outdoors, and that these
airflows are used to determine the contaminant levels in the
shelter, the exposure of the shelter occupants to an outdoor
release, and ultimately the protection factor. The airflows,
contaminant levels, and occupant exposures for a given event
will depend on the details of the release (i.e., the time profile
and location), building and system  operating conditions
(outdoor air intake and other system airflows), outdoor
weather conditions, and the interzone airflow dynamics within
the building. However, as will be seen, the measured shelter
airtightness can still be used to provide a reasonable indicator
of the level of protection offered by a given shelter.
In order to investigate  the pressurization-testing approach to
SIP space evaluation, the following efforts were pursued under
this project:
  • SIP zone pressurization tests in several spaces
    to determine a range of airtightness values;
  • Computer simulations to relate SIP zone airtightness
    to protection factor;
  • Computer simulations to assess the ability to reliably
    measure SIP zone airtightness  with fan pressurization; and
  • Validation of predictions from shelter airtightness values.
This section describes  the various steps pursued in
investigating the pressurization test approach to SIP space
evaluation, beginning with a discussion of the relevant theory
and calculation methods.
3.1 Theory and Calculation Approaches
In order to relate SIP and building airflows to exposure, a
two-zone mass balance theory is presented. This theory is
employed by letter and Whitfield [12] and others, and is
based on the key assumption that the shelter exchanges air
only with the rest of the building volume and has no direct
airflow connections with the outdoors. Similar theory can
be developed for the more general case in which the shelter
does exchange air directly with the outdoors. Another key
assumption is that both the shelter and the remainder of
the building can each be represented by a single value of
the contaminant concentration. This latter assumption will
often be reasonable for the shelter, particularly for a small
space, but may be less justifiable for the rest of the building,
particularly a large and complex building.
Figure 3 is a schematic diagram of such a two-zone model,
which is described by two mass balance equations of
contaminant, one for each zone:
   dCs _
    ,   — '
                        - QscB - QBc
                                   BB
where
    Vs = volume of the shelter
    VB = volume of the rest of building
    Cs = contaminant concentration in the shelter
    CB = concentration in the rest of the building
    COUT = concentration outside the building
    Qs = airflow between the shelter and the building
    QB = airflow between the building and outside
     t = time
(3)
(4)
                                Figure 3 Schematic of two-zone SIP model

-------
Solving for Cs and CB allows one to calculate exposure in the
building and the shelter for specific volumes, airflows, and
outdoor release profiles. Alternatively, one can use this two-
zone mass balance theory to analyze tracer gas test data to
determine the airflows of interest, as was done by letter and
Whitfield [12] in a low-rise residential building for which an
interior bathroom served as the shelter.
Figure 4 is a sample plot of shelter and building
concentrations for a simple single-zone building with
an SIP space subject to a 1-min elevation of the outdoor
concentration to a dimensionless value of 1.0. The values
in this plot were generated using the multizone airflow
and contaminant dispersal program CONTAM [34] for a
simple two-zone model with weather-driven infiltration.
The building concentration responds relatively quickly,
while the shelter response lags that of the building. After
the building concentration peaks, it then decreases while
the shelter concentration continues to increase. For this
case, the shelter concentration peaks after about 30 h at a
value about one tenth of the building peak. From this point
on, the shelter concentration is higher than the building
concentration. The timing of the two peaks and their relative
magnitudes are a function of the volumes and airflow rates in
the two-zone model, as well as the time profile of the outdoor
concentration. Nevertheless, the shelter peak will always be
lower than the building peak but of longer duration. The fact
that the shelter concentration eventually exceeds the building
concentration demonstrates the need to leave the shelter after
the outdoor threat has passed in order to avoid this longer-
term exposure in the shelter.
Figure 5 shows the protection factor for the shelter as a
function of time for the same case as shown in Figure 4. Two
values of protection factor are presented, one referenced to
the outdoor exposure and the other referenced to the building
exposure. Note that in both cases the shelter provides a high
level of protection relative to the rest of the building early
in the event, but the protection decreases as time progresses.
The protection factors referenced to the building are lower
than those referenced to outdoors due to the contaminant that
enters the building. Eventually, both protection factors will
decrease to a value of one as the indoor exposure eventually
attains a value equal to the outdoor exposure.
            0.002

            o.ooi
            o.ooo
                                     12
                                                          24
                                                       Time (h)
                                                                              36
                                                                                                  48
                    Figure 4 Contaminant concentration values for two-zone example

-------
3.2 SIP Zone Airtightness Tests
While some data exist for interior partitions in buildings [35],
they are limited and not necessarily appropriate for spaces
that have been sealed to provide sheltering. Therefore, this
portion of the study involved using fan pressurization testing
to determine a range of shelter airtightness values with and
without expedient sealing efforts.  The first step in this effort
is to establish a test protocol and apply it to several potential
shelter spaces. The  airtightness tests are based on established
fan pressurization methods, sometimes referred to as blower
door testing. Fan pressurization testing employs a fan to
mechanically pressurize or depressurize a room or building,
while simultaneously measuring the airflow rate through the
fan required to maintain the induced pressure difference.
The tests in this study employed the procedures in ASTM
Standards E779 and E1827 [36, 37], with some modifications
since these standards are intended for testing whole buildings
as opposed to rooms within a building.
Figure 6 shows a schematic of the test configuration for
pressure testing an SIP space. The fan pressurization device
is installed in a door to the SIP space and used to induce
airflow into (out of) the space from (to) an adjacent interior
space. Often a hallway serves as this adjacent space. The
fan airflow raises (lowers) the pressure in the space relative
to the adjoining space(s), with the airflow into (out of) the
space leaving (entering) the SIP space through leakage paths
connecting to adjoining spaces.
           1000000
            100(100
             10000
              1000
               100
               10
           -Referenced to outside


            Referenced to building
                                     12
                                                         24
                                                      Time (h)
                          Figure 5  Protection factor values for two-zone example


/71


Pressure^.
difference
Air
leakage
• — /?

SIP space
*.



YV




V ^ ^
Fan-induced airflow



                                 Figure 6 SIP pressurization test schematic

-------
In fan pressurization testing, the test is conducted at pressure
differences from approximately 10 Pa to 70 Pa. The air
leakage characteristics of the space or building being tested
are calculated from the measured airflow rates and pressure
differentials. A blower door device, shown below in Figure
7, was used for these tests. The airflow rates were determined
from the pressure difference across an orifice plate built
into the blower door, using the manufacturer's calibration,
which was confirmed in a blower door calibration chamber at
NIST, conducted in accordance with ASTM E1258 [38]. The
blower door airflow rate has an associated uncertainty, based
on these calibrations, on the order of 0.01 mVs. The pressure
difference between the SIP zone and the adjacent hallway
was recorded during the tests with a digital manometer that
has a stated uncertainty of +/— 1 Pa.
                                             Figure 7 Blower door

-------
These fan pressurization tests yield a series of pressure
differences and airflow rates, an example of which is plotted
in Figure 8. These data are then fit to a curve of the following
form:
Q = Cp"                                            (5)

where
    Q = airflow rate, mVs
    C = flow coefficient determined by curve fit, mVs«Pa
    p = pressure difference across the SIP zone, Pa
    n = pressure exponent determined from curve fit,
       dimensionless
Once the pressurization test data are fit to a curve of the
form shown in Equation (5), the curve is used to estimate the
effective leakage area (EL A) of the space as a measurement
of airtightness. The EL A is defined as the size of an orifice of
discharge coefficient 1.0 that yields the same airflow rate as
that predicted by the curve fit at some reference pressure, in
this case 4 Pa [39]. The EL A values are then normalized by
the surface area of the SIP space, including the walls, floor,
and ceiling to yield a value in units of cm2/m2.
1500

125U
^ 1000
S. 750

MJU
250


*
•*
.1"
•
**
I
*







0 10 20 30 40 SO 60
Pressure diferencc (Pa)
                             Figure 8 Sample plot of pressurization  test data

-------
SIP airtightness tests were performed in eight spaces in
three buildings, as described in Table 1. All three buildings
were located in suburban office park settings, with the first
(226) building a roughly 40-year-old, 3-story building on the
NIST campus. The second building (NC-Off) is a 21-year-
old, single-story building located in the Research Triangle
Park, NC, area, and the third (RTF) is a fairly new six-story
building on the EPA RTF campus. All of the rooms tested had
no exterior windows or walls, with one exception as noted
below.
Each of the four NIST rooms was tested twice. The first test
was conducted to determine the leakage of the room with all
ventilation systems off and all vents sealed. The second test
was conducted after sealing all visible cracks, outlets, and
penetrations. The four rooms  located in North Carolina were
all involved in a series of SIP tracer gas studies conducted by
letter and Proffitt [13] and were pressure tested only once as
part of the current effort, under the same sealing conditions
as in the referenced study. Note that Room B has one exterior
wall with windows.
                            Table 1 Test site and  room dimensions
Hill
B221
B113
A368
B317
A
B
C
D
| 9j| ffjljf II
226
226
226
226
NC-Off
NC-Off
RTF
RTF
I ill
8.1 x
8.1 x
8.1 x
8.1 x
4.1 x
4.4 x
6.4 x
II
6.5x3.4
3.3x3.4
3.3x3.4
3.3x3.4
2.7x2.7
4.0x2.7
3.8x2.6
8.6x12.1x3.6
                             Space height is the last dimension listed
3.3 CONTAM Predictions of Protection Factor
In order to investigate the relationship between SIP zone
leakage and protection factor, a series of simulations
was performed using CONTAM [34]. Three buildings
were considered in order to develop a more complete
understanding of the impact of various parameters on agent
exposure within the SIP space: a simple one-story building,
a simple ten-story building, and a more realistic two-story
office building.
The simple one- and ten-story buildings were configured as
a single open zone with the SIP space contained within that
volume. The one-story building model has a floor area of
110 m2 and a ceiling height of 3 m, whereas the ten-story
model has a total floor area of 1,010 m2 and a building height
of 30 m (3 m per floor). In both models the shelter floor
area is 10 m2 with a ceiling height of 3 m. These two simple
models provide a first-order sense of the  impacts of shelter
and building tightness.
The two-story office building model, depicted in Figure 9 as
a CONTAM floor plan, includes restrooms, elevators, stairs,
a lobby, and a conference room. (The lines in the figure
depict pressure differences and airflow rates from a set of
CONTAM predictions.) The conference room, located on the
second level, is designated as the SIP zone. Each level has a
floor area of 920 m2 and a ceiling height  of 3 m. The shelter
floor area is 260 m2.
The simulations were performed for two sets of weather
conditions, cold and windy as well as mild and calm.
The former conditions correspond to an indoor-outdoor
temperature difference of 20 °C and a wind speed of 5 m/s,
while the latter is defined by a temperature difference of
2.5 °C and a wind speed of 1 m/s. These two sets of
conditions result in high and low building air change rates,
which bound the results in terms of the amount of outdoor
contaminant that enters the building and therefore is available
to expose the shelter occupants. The building and shelter
leakage values were also varied, including exterior wall
EL A values from 1 cm2/m2 to 20 cm2/m2 and shelter wall
ELA values from 0.1 cm2/m2 to 10 cm2/m2. The exterior wall
leakage values are based on measurements conducted in a
range of commercial buildings [40]. The shelter values are
based on the limited data that exist for interior wall leakage
[35] and the assumption that shelter walls will generally be
tighter than exterior walls. Two values of shelter occupant
density were employed in the simulations, 1 m2 of floor
area per person, based on a minimum recommendation in
several documents [2, 14], and 2 m2 per person. The former
value is consistent with FEMA recommendations for tornado
shelters, where air leakage and contaminant entry are not
issues. Based on potential concerns about CO2 buildup, the
lower occupancy density was also considered.  It happened to
be ten times the default occupant density for office space in
ASHRAE Standard 62.1-2007 [41].

-------

-------

-------
                                                                                                         4.0
                                                                                 Study  Results
This section presents the results of the measurements and
analyses described in the previous section, including the fan
pressurization tests of shelter airtightness, the CONTAM
predictions of protection factor, the CONTAM simulations
to assess potential measurement errors in the field, and
validation of the predictions through comparison with
limited field measurements.

4.1 SIP Airtightness
The results of the zone airtightness tests are presented in
Table 2 for each of the eight rooms tested. As noted earlier,
the NIST spaces were tested once with only limited sealing
and again with extra sealing. Also, ELA values for the
NIST spaces are presented for both zone pressurization
and depressurization conditions. The four spaces in North
Carolina were tested only under depressurization and only
with sealing consistent with the test conditions used in
the earlier study by letter and Proffitt [13]. Under limited
sealing the ELA values are in a relatively narrow range from
somewhat above 1 cnvYm2 to about 5 cmVm2. The sealed
values are lower, as expected, and range from 0.25 cnvYm2
to just under 1 cm2/m2. In general, the pressurization and
depressurization test values are similar, with the exception
of Room 2267A368 with extra sealing in place. The percent
reduction due to sealing is surprisingly consistent for the four
spaces, ranging from about 60% to 90%.
The recent FEMA report on shelters and safe rooms cited
earlier includes fan pressurization test results of a stairwell
that was being  considered as an SIP space and provides
another airtightness data point [2]. The stairwell was
approximately  43 m high and had a cross-section of 10.7 m
by 3.7 m. The airtightness test was done once as is and again
with the doors  sealed. The as is and sealed ELAs at 4 Pa
were 3.1 cm2/m2 and 2.1 cm2/m2,  respectively. These values
are in the range of the measurements presented in Table  2.
This same FEMA report contains a table (3-2) with airflow
values, in units of cfm per ft2 of floor area, required to
pressurize a room to 25 Pa. These values are presented for
four levels of tightness, very tight, tight, typical, and loose.
Converting these values into the units presented in Table 2,
assuming a room size of 5 m x 5 m x 2.5 m, the results are as
follows: 0.47 cm2/m2, 2.36 cm2/m2, 5.89 cm2/m2, and 11.78
cm2/m2. These values from the FEMA report are certainly
consistent with those measured in the eight spaces considered
in this study.
                 Table 2 SIP airtightness test results

Building/Ro
226/B221
226/B113
226/A368
226/A317
NC-Off/A
NC-Off/B
RTP/C
RTP/D
Mean
StdDev
IRlt'SJf


I
om Press
4.80
2.60
1.74
1.76

-
-
-
2.73
1.44
Depress
5.08
2.72
1.86
1.85
6.08
4.67
2.78
3.74
3.60
1.56
} || HI
ISIS
Press
0.94
0.37
0.71
0.52
-
-
-
-
0.64
0.25


Depress
0.98
0.47
0.37
0.61
-
-
-
-
0.61
0.27
1

Press
80
86
59
70
-
-
-
-
74
12
i '*;!'' i^TSji If |V •'"
Depress
81
83
80
67
-
-
-
-
78
7
                 Press and Depress indicate whether the SIP zone was under positive or negative pressure
                 during the test.

-------
4.2 CONTAM Predictions of Protection Factor
This section presents the results of the CONTAM simulations
that were conducted to investigate the relationship between
shelter airtightness and protection factor. These simulations
were performed in three model buildings and predicted the
concentrations of an agent released outdoors and CO2 levels
in the shelter due to the shelter occupants.
Figure 10 shows an example of the simulation results for
the one-zone building model with a building leakage value
of 5 cmVm2 and a shelter leakage of 1 cm2/m2. The plot
shows the agent concentration in the shelter over the 2-day
(2880-min) simulation period, which starts to increase from
0 mg/m3 starting at t = 1 h. The plot also contains the CO2
concentration in the shelter, which increases from an initial
value of 1800 mg/m3. Note that the indoor CO2 increases to
fairly high levels, relative to the ACGIH short-term (1-min)
         exposure limit of 54,000 mg/m3 and the threshold limit value
         (based on an 8-h exposure over a 40-h work week) of 9000
         mg/m3, which strongly suggests limits on the amount of time
         that such a shelter should be occupied [42].
         The simulations for other building and shelter leakage values
         and for the other buildings yield results similar to those seen
         in Figure 10. Tighter buildings and shelters tend to lower the
         peak agent concentration in the shelter but also extend the
         period of time over which the agent remains in the shelter.
         For example,  Figure 11 shows the agent concentrations for
         a building leakage value of 5 cmVm2 and several values of
         the shelter leakage. As the shelter leakage increases, the peak
         concentration also increases  but the duration of the elevated
         shelter concentration decreases. Increased tightness also
         increases the shelter CO2 concentrations due to the decreased
         dilution rates.
                 8E-03
                 6E-03
                 4E-03 •
            "*    2E-03
                OE-00
                                                                                         •   4Ei04
                                           6E*04
                                                                                            2E»04
                                                                                                   c
                                                                                                   3
                                                                                                   |
                                                                                                    .
                                                                                                   C
                                                                                                   3

                                           OEiOO
                                    10
20           30
    Time (h)
40
50
                  Figure  10 Example plot of simulated agent and C02 concentrations
                  for the one-zone building

-------
            5.0E-03
            4.0E-03 •
         |
            3.0E-03 •
         1
            2.0E-03
            l.OE-03
            O.OE • 00
                                   10
20
40
50
                                                       Time (hi
                     Figure  11 Simulated agent concentrations in shelter for multiple
                     shelter leakage values
One-zone building model
The simulation results for the one-zone building model
are shown in Tables 3 through 6. Table 3 shows protection
factors after 1 h, 2 h, and 3h for different combinations
of building and shelter leakage under the cold and windy
weather conditions, as well as the building and shelter air
change rates for each combination of leakage values. The
building air change rates depend only on the building leakage
and range from 0.09 h'1 for the tightest building to 1.71 h'1
for the leakiest. The shelter air change rates describe the air
change rate with the rest of the building and depend only on
the shelter airtightness, covering a range of about 100 to  1
from the tightest to the leakiest shelter values.
Two different protection factors are presented in the table:
PF0, which is defined as the outdoor exposure divided by the
exposure in the shelter after the designated time interval, and
PFB, which is the exposure in the rest of the building divided
by the shelter exposure. The protection factors defined with
reference to the building are always lower than the PF0
because the building concentration remains elevated after
the outdoor concentration returns to zero. However,  PFB
          may be viewed as a better measure of protection for situations
          where the building occupants remain in the building during
          the sheltering period rather than go outdoors. The protection
          factors decrease over time,  since the agent remains in the
          shelter after the outdoor episode is over. As noted earlier, the
          protection factor decreases  as the building and shelter leakage
          increase. For a given duration of sheltering, the protection
          factors vary by more than two orders of magnitude; therefore,
          a very tight building and a very tight shelter can decrease the
          exposure in the shelter to less than 1% of what it would be
          under the leakiest conditions.

          Table 4 presents the air change rates and protection factors
          for the one-zone model for the mild and calm weather. The
          more temperate weather conditions reduce the building air
          change rates to about 25% of their values under the cold and
          windy conditions but do not impact the shelter-to-building
          air change rates. As a result, the values of PFB do not change
          very much relative to the values in Table 3. However,
          PF0 increases by roughly a factor of four, corresponding
          approximately to the reduction in the building air change rate.

-------
Table 3 Protection factors for one-zone model (cold  and windy weather)
II
1.0
1.0
1.0
1.0
1.0
1.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
20.0
20.0
11
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
111
I
0.09
0.09
0.09
0.09
0.09
0.09
0.43
0.43
0.43
0.43
0.43
0.43
0.85
0.85
0.85
0.85
0.85
0.85
1.71
1.71
1.71
1.71
1.71
1.71

I
I 1 !
0.01 182
0.05 37
0.11 19
0.27 8
0.54 4
1.08 3
0.01 171
0.05 35
0.11 18
0.27 7
0.54 4
1.08 2
0.01 160
0.05 33
0.11 17
0.27 7
0.54 4
1.08 2
0.01 143
0.05 29
0.11 15
0.27 6
0.54 3
1.08 2

1 I
fSi -| 1 1
2041
415
212
90
49
29
460
93
48
20
11
7
264
54
27
12
6
4
169
34
18
7
4
3

m !i;;;;f
89
18
10
4
3
2
80
16
9
4
2
2
71
14
8
3
2
1
61
13
7
3
2
1
i» i
f±, i
I I I
524
108
56
25
15
10
132
27
14
6
4
3
84
17
9
4
2
2
62
13
7
o
5
2
1


I'siii-iidiji
59
12
7
o
5
2
1
50
11
6
3
2
1
44
9
5
2
2
1
37
8
4
2
1
1
^i i
lit 1
240
50
27
13
8
6
67
14
7
4
2
2
46
10
5
3
2
1
38
8
4
2
1
1
PFB and PF0 are the protection factors with reference to the building and outdoor concentration, respectively.

-------
Table 4 Protection factors for one-zone building model  (mild and calm weather)
III
1.0
1.0
1.0
1.0
1.0
1.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
20.0
20.0
111
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
"III
I- ;11 IB
0.02
0.02
0.02
0.02
0.02
0.02
0.10
0.10
0.10
0.10
0.10
0.10
0.20
0.20
0.20
0.20
0.20
0.20
0.40
0.40
0.40
0.40
0.40
0.40
ii fy
1 Bl
i fiili ii 1?
0.01 18
0.05 3"
0.11 15
0.27 8
0.54 4
1.08 3
0.01 18
0.05 3"
0.11 V
0.27 8
0.54 4
1.08 3
0.01 17
0.05 3d
0.11 IS
0.27 8
0.54 4
1.08 2
0.01 17
0.05 3f
0.11 IS
0.27 7
0.54 4
1.08 2
U InlpllSti 'V I'll

jijl j: Piiplj j|
4 8516
' 1731
) 883
374
205
122
1 1753
356
182
77
42
25
8 908
185
94
40
22
13
2 487
99
50
21
12
7
|P
III
92
19
10
4
3
2
89
18
9
4
2
2
86
18
9
4
2
2
81
17
9
4
2
2
1 I
i^iifSJ i
2135
441
229
103
61
41
452
93
49
22
13
9
242
50
26
12
7
5
138
29
15
7
4
3
||l


jiMpiUIJiji i |!
61
13
7
3
2
1
58
12
6
3
2
1
56
12
6
3
2
1
51
11
6
o
3
2
1
'Misllljji
957
201
106
50
32
23
208
44
23
11
7
5
115
24
13
6
4
o
5
69
15
8
4
2
2
PFB and PF0 are the protection factors with reference to the building and outdoor concentration, respectively.

-------
Figures 12 and 13 are graphical presentations of the data
in Tables 3 and 4, with Figure 12 presenting the protection
factor using the building exposure as a reference and Figure
                                                13 based on the outdoor exposure. Such figures are used to
                                                present the results for the other building models, rather than
                                                using the format in Tables 3 and 4.
          10000
           1000
Mild/Calm
Cold/Windy
• 1 hour
• 1 hour
• 2 hour
2 hour
•3 hour
3 hour
        I
        c
        •s
 Shelter  "-1
Building  1.0
                                                                                           20
                                              Leakage va
               Figure  12  Protection  factors,  referenced to building, for one-zone model
          10000
           1000
Mild/Calm
ColdWind.v
• 1 hour
• 1 hour
•2 hour
2 hour
•3 hoar
•3 hour
        c
        I
         Building
              1
          Shelter
                                                                                           2U
                                              Leakage value (cmz/m:)
               Figure 13 Protection factors, referenced to outdoors, for one-zone  model

-------
                  Table 5 Carbon dioxide concentrations for one-zone  model
                  (cold and windy weather)
1 III III III 1 III
IlilUHl
1 ll Hi

1.0
1.0
1.0
1.0
1.0
1.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
20.0
20.0
|| ll i|
ifiil
fid f """ 1

0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
1 1 II
llltillillii


13364
13113
12811
11973
10776
8965
13362
13104
12793
11929
10695
8824
13360
13095
12775
11888
10618
8690
13358
13084
12752
11833
10518
8517
i !' Hi 111!
iliiliiiiiillftilBiii


24787
23808
22668
19736
16107
11751
24781
23777
22607
19593
15849
11324
24776
23754
22562
19486
15655
11005
24771
23731
22518
19382
15467
10696
11


itpsypi k
till If if
36085
33933
31514
25721
19430
13184
36073
33874
31397
25443
18929
12377
36066
33837
31325
25271
18618
11878
36060
33809
31268
25132
18363
11475
I I
lllililliiiiiS
1 1
'Hit
7597
7470
7318
6895
6292
5379
7595
7461
7300
6852
6213
5244
7593
7452
7283
6811
6138
5116
7591
7441
7260
6758
6040
4953
HHntSl
iilliiltilllliiiiii
ll 1 11
-. 1 Bflfil?!
13308
12814
12239
10761
8932
6736
13301
12783
12180
10625
8696
6367
13297
12760
12136
10525
8522
6098
13292
12738
12093
10428
8358
5853
'111 I



18955
17871
16652
13732
10562
7415
18943
17813
16540
13482
10141
6788
18936
17778
16472
13332
9891
6425
18930
17751
16420
13217
9702
6157
                  PFB and PF0 are the protection factors with reference to the building and outdoor
                  concentration, respectively.
Table 5 shows the CO2 concentrations for the one-zone
model at 1 h, 2 h, and 3 h for two values of occupant density,
1 m2 per person and 2 m2 per person, under cold and windy
conditions. Each value in the table is the CO2 concentration
in the shelter at the specified leakage values and time
after sheltering commences. The concentrations are most
sensitive to the duration of sheltering and fairly insensitive
to the building leakage values. The three-to-one variation
in sheltering duration leads to an increase in concentration
by a factor of roughly two to three, which is similar to the
variation seen for the 100-to-l range in shelter leakage.
The shelter leakage has more of an effect than the building
leakage, but the sheltering duration is still the dominant
factor. None of the cases are as high as the ACGIH short-
term (1-min) exposure limit of 54 000 mg/m3, but several
exceed the threshold limit value of 9000 mg/m3 [42]. For
the sake of discussion, a reference value of 10,000 mg/m3 is
useful, but note that this value is not a health-based criterion
or a "safe" concentration limit. The higher value of occupant
density, 1 m2 per occupant, results in roughly a doubling of
the CO2 concentration for the same leakage and duration
values, with almost all of the values being above  10,000
mg/ m3. For the lower occupant density (2 m2 per occupant),
the concentration after 1 h of sheltering never attains this
reference concentration, even for the lowest leakage values.
The 2-h CO2 concentrations exceed that reference value for
all but the leakiest shelters, although having a leaky shelter
is counter to the goal of sheltering.

-------
Table 6 shows the CO2 concentrations for the one-zone model
under the mild and calm weather conditions. These values
exhibit the same strong dependence on duration of sheltering,
with less of an effect of shelter leakage and very little
dependence on building leakage, as was seen for the more
severe weather conditions. The concentrations themselves are
only slightly higher than those seen in Table 5 for the same
leakage and duration values.
                Table 6 Carbon dioxide concentrations for one-zone model
                (mild and calm weather)


liHIIiil

1.0
1.0
1.0
1.0
1.0
1.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
20.0
20.0
1 III 1 I
I I
I 11^' !

0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0

HI; lip!


13364
13115
12815
11982
10794
8996
13364
13113
12810
11970
10772
8958
13363
13110
12804
11957
10747
8914
13362
13105
12794
11932
10701
8834
llKprtiliililfiil
iHlWIIMIIMtfe

^f^lil^llllAlI
24788
23815
22683
19772
16172
11860
24786
23806
22665
19728
16093
11729
24784
23796
22644
19680
16006
11585
24781
23779
22611
19602
15865
11351




36088
33950
31547
25799
19572
13414
36084
33929
31507
25704
19401
13137
36079
33908
31465
25606
19223
12851
36073
33877
31403
25459
18958
12424

lllililiiliiiit


7597
7472
7322
6905
6309
5409
7597
7470
7317
6893
6288
5372
7596
7467
7311
6880
6263
5329
7595
7462
7301
6855
6218
5253
lililliliiil
Ililliiliilii
i« |
ill ill
13309
12822
12254
10796
8992
6832
13307
12812
12236
10754
8919
6717
13305
12802
12216
10708
8839
6591
13302
12785
12184
10634
8711
6390
1 ill



18959
17887
16684
13804
10684
7599
18954
17867
16645
13717
10537
7378
18950
17847
16605
13628
10386
7152
18944
17816
16546
13496
10165
6823

-------
           50000
           40000
        1
        £ 30000
           20000
           10000
               0
           Shelter

         Building
1 ill 'person
2 mVperson
• 1 hour
• 1 hour
•2 hour
2 hour
• 3 hour
• 3 hour

0.1
1.0
   0.5
    1.0
       I.U
          2.5
          1.0
              1.0
                  10
                 1.0
                     0.1
                     5.U
                        0.5
                        5.0
                           5.0
                               2.5
                               5.0
                                  5.U
                                      10
                                      5.0
                                         O.I
                                         1U
                                             10
                                                 Leakage value u-in-nv' j
             Figure 14 Shelter CO  concentration for one-zone  model - cold/windy weather
Figure 14 is a plot of the concentration data in Table 5
for the cold and windy weather conditions. This graphical
format will be used for the other building model results.
Since the concentrations are very similar for the mild and
calm conditions, a second plot is not included.
Two-story office building model
Table 7 contains the air change rates for the two-story office
building model as a function of building and shelter leakage
and for the two sets of weather conditions. The building air
change rates are  similar in magnitude to those seen for the
simple one-zone model. The  shelter rates are consistently
lower than for the one-zone model, as low as 20% of the
one-zone rates, but still of the same order of magnitude.
The simulated protection factors are shown in Figures 15
and 16. (Note that these figures have a different scale from
the one-zone results in Figures 12 and 13.) Figure 15 shows
the protection factor after 1 h, 2 h, and 3 h for different
                                          combinations of building and shelter leakage, using the
                                          building exposure as the reference. These building-based
                                          protection factors are almost an order of magnitude larger
                                          than those seen in the one-zone model for the mild weather
                                          conditions. This difference is due to the lower shelter air
                                          change rates. Under the cold-windy weather, the office
                                          building protection factors are somewhat higher than
                                          the one-zone model, but the difference is smaller, given that
                                          the air change rates are closer to those seen for the  one-zone
                                          case. Figure 16 shows the protection factors referenced to
                                          the outdoor exposure. These protection factors are again well
                                          above those seen for the one-zone model due to the lower
                                          building and shelter air change rates. Overall, these results
                                          display trends similar to those seen for the one-story building.
                                          Higher protection factors correspond to tighter buildings and
                                          tighter shelters, and shelter leakage has a more significant
                                          impact than building leakage.

-------
Table 7 Air change rates for two-story office building model
III
1.0
1.0
1.0
1.0
1.0
1.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
20.0
20.0
ill
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0


MMM^MAiMiRM^IIiilbMMi
fill If II
0.12
0.12
0.12
0.12
0.12
0.12
0.35
0.35
0.35
0.36
0.36
0.36
0.64
0.64
0.64
0.64
0.64
0.64
1.18
1.18
1.18
1.18
1.18
1.18
II!
0.04
0.07
0.09
0.11
0.18
0.31
0.05
0.10
0.13
0.16
0.20
0.34
0.06
0.11
0.14
0.17
0.21
0.35
0.06
0.12
0.14
0.17
0.23
0.37



II
0.02
0.02
0.02
0.02
0.02
0.07
0.07
0.07
0.07
0.07
0.07
0.12
0.12
0.12
0.12
0.12
0.12
0.12
0.21
0.21
0.21
0.21
0.21
0.21
4111111 i
1 ' if
0.01
0.02
0.04
0.08
0.15
0.28
0.01
0.03
0.04
0.08
0.15
0.28
0.01
0.03
0.04
0.08
0.15
0.28
0.01
0.03
0.04
0.08
0.15
0.28

-------
  100000
   10000
c
fl
    1000
Mild/Calm
Cold/Wiitdy
• 1 hour
• 1 hour
•2 hour
2 hour
• 3 hour
•3 hour
       1
  Shelter I0'1
 Building 11-°
                                      Leakage value (cm2/m2)
    Figure 15 Protection factors, referenced to building, for office building model
  100000
   10000
    1000
     100
      10
       1
  Shelter |«-i
 Building  1.0
Mild/Calm
Cold/Windy
• 1 hour
• 1 hour
• 2 hour
2 hour
• 3 hour
• 3 hour
0.5
1.0
   1.0
      2.5
       1.1)
          1.0
             10
             1.0
                0.1
                5.0
                   0.5
                   5.0
                      5.0
                          2.5
                          S.O
                             5.0
                                10
                                5.0
                                   0.1
                                   10
                                      0.5
                                       10
                                          10
                                             2.5
                                             10
                                                10
                                                   10
                                                   10
                                                      0.1
                                                      20
                                                         0.5
                                                          20
                                                             20
                                                                2.5
                                                                20
                                                                   20
                                                                       10
                                                                      20
                                      Leakage value (cmVm1)
    Figure 16 Protection factors, referenced to outdoors, for office building model

-------
MJUUU
4UUUU
I
~, 30000
•
•|
e
u
8
10000 -


1


0
Shelter "-I o.s 1
Building i.o i.o I.O


IN

2.5 5 10 0.1 0.5 1
1.0 1.0 I.O 5.0 5.0 5.1

1 m1 person
2 mVperson
• 1 hour • 2 hour • 3 hour
• 1 hour 2 hour • 3 hour




- 4

2.5 5 10 1
5.0 5.0 5.0




).l
ID


1

U.5
10

|
1
10

1 1

!.S 5 10 0.1 O.f
10 10 10 20 20




1 2.5 5 10
20 20 20 20
                                               Leakage value (cmVm3)

         Figure 17 Shelter CO  concentration for office building model - cold/windy weather
Figure 17 shows the CO2 concentrations in the shelter at 1 h,
2 h, and 3 h for the two-story office building. Concentrations
are presented for the two different occupant densities and
the various building and shelter leakage values. As in the
case of the one-story model, the duration of sheltering is
more significant than building or shelter tightness. The lower
occupant density, 2 m2 per person, reduces the concentrations
significantly. The shelter concentrations are all above the
reference value of 10,000 mg/m3 except for the lower
occupant density after 1 h of sheltering.

-------
                      Table 8 Air change rates for ten-story building model
11
1.0
1.0
1.0
1.0
1.0
1.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
20.0
20.0
lf(v"*f|j j
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0
0.1
0.5
1.0
2.5
5.0
10.0


MMUSIMI^^MKMMiiiMMI
SI fill
0.08
0.08
0.08
0.08
0.08
0.12
0.38
0.38
0.38
0.38
0.38
0.38
0.76
0.76
0.76
0.76
0.76
0.76
1.52
1.52
1.52
1.52
1.52
1.52
HI
0.01
0.05
0.11
0.27
0.54
1.08
0.01
0.05
0.11
0.27
0.54
1.08
0.01
0.05
0.11
0.27
0.54
1.08
0.01
0.05
0.11
0.27
0.54
1.08
iPtfisiiJ
IfWimUB !
1; li JSi II j is i
I !
0.02
0.02
0.02
0.02
0.02
0.07
0.09
0.09
0.09
0.09
0.09
0.09
0.17
0.17
0.17
0.17
0.17
0.17
0.35
0.35
0.35
0.35
0.35
0.35

jj sill]
ill i
0.01
0.05
0.11
0.27
0.54
1.08
0.01
0.05
0.11
0.27
0.54
1.08
0.01
0.05
0.11
0.27
0.54
1.08
0.01
0.05
0.11
0.27
0.54
1.08
Ten-story building model
Table 8 contains the air change rates for the ten-story
building model as a function of building and shelter leakage
and for the two sets of weather conditions. The values are
very similar to those seen for the one-zone model. The
simulated protection factors are shown in Figures 18 and 19
for different combinations of building and shelter leakage,
which are also very similar to those seen in the one-zone
model. Given the configuration of the ten-story building,
these results are not very surprising as the primary difference
between this model and the one-zone is that the taller
building has stronger stack (or temperature difference driven)
pressures. However, the larger volume of the ten-story
building leads to somewhat lower air change rates.

-------
  100000
   10000
5   1000
D
Mild/Calm
ColdAVindy
• 1 hour
• 1 bour
• 2 hour
2 bour
•3 hour
• 3 hour
  Shelter P-1
Building
2(1
                                    Leakage value (cm2/m2)
  Figure  18 Protection  factors, referenced to building, for ten-story building model
  100000
   10000
o   1000
Mild/Calm
Cold/Windy
• 1 hour
• 1 hour
• 2 hour
2 hour
•3 hour
•3 hour
  Shelter  P'1
 Building  | '•»
2(1
                                    Leakage value (cin:/m-)
  Figure  19 Protection  factors, referenced to outdoors, for ten-story building model

-------
          50000
          40000
        E, 30000
        g
          20000
        u
         M
        c
        Q
1 in* person
2 mVperson
• 1 hour
• 1 hour
•2 hour
2 hour
• 3 hour
•3 hour
           10000 •*	
          Shelter
         Building
                                    2U
                                               Leakage value (cmVm')
            Figure 20 Shelter CO  concentration for ten-story model - cold/windy weather
Figure 20 shows the CO2 concentrations in the shelter at 1 h,
1 h, and 3 h for the ten-story office building. Concentrations
are presented for the two different occupant densities and the
various building and shelter leakage values. As with the other
models, the duration of sheltering is more significant than
building or shelter tightness and the lower occupant
density reduces the concentrations significantly. The
shelter concentrations are all above the reference value
of 10,000 mg/m3 except for the lower occupant density
after only 1 h of sheltering.

-------
                                                                        P.I
                    Figure 21 CONTAM sketchpad of simulated pressurization tests
4.3 CONTAM Simulations of SIP Leakage
Measurement
In order to investigate the impacts of interzone airflows and
pressures on the ability to reliably estimate shelter leakage
using a pressurization test, a series of CONTAM [34]
simulations was performed for a ten-story building model.
As described below, this ten-story model is different from
the model used in the protection factor calculations described
previously. In these simulations, a pressurization test to
measure shelter airtightness was simulated with CONTAM
by imposing an airflow on the SIP zone and then recording
the airflow rates out of the zone and zone-to-hallway pressure
differences as determined by the simulations. Figure 21 shows
the CONTAM sketchpad for one floor of the building, with
the SIP zone highlighted in green in the center of the floor,
surrounded by five other zones. The colored zones at the
upper edge of the floor plan represent stairway and elevator
shafts.
The goal of these simulations was to determine whether
erroneous leakage values might result if the SIP zone being
tested shared walls with other building zones, since the
pressure difference measured during a test is typically from
the SIP zone to the hallway. The impacts of ventilation system
airflows in other zones were also examined, as it may not
always be feasible to deactivate building ventilation systems
during these tests. Note that system airflows to the SIP zone
itself would need to be eliminated by turning off fans or
sealing vents. Several such simulations were conducted for
SIP zones on the first, fifth, and ninth floor of the ten-story
building. Simulations were conducted with the ventilation
system off and with 10% more supply airflow than return
airflow to those  parts of the building not being tested. Note
that there are no system flows into the SIP zone itself during
these simulations. In addition, pressurization tests were
simulated with the SIP zone pressurized and depressurized
relative to the building hallway. Finally, these simulations
were conducted under four different weather conditions:  zero
wind speed and  0 °C indoor-outdoor temperature difference
(T); 5 m/s wind  and 0 °C T; 0 m/s wind speed and 20 °C  T;
and 5 m/s wind  speed and 20  °C T.

-------
Table 9 shows the simulation results in terms of the effective
leakage area at 4 Pa calculated from the simulated pressure
test. The actual leakage values input into the CONTAM model
of the building are listed in the last row of the table. The
simulated tests on the first floor with the system off are very
close to the model value for all weather conditions. When the
system is on, the simulated test results are about 10% high if
the SIP zone is being pressurized during the test and 10% low
if it is being depressurized. The simulated test results on the
fifth and ninth floors are more sensitive to weather conditions,
in particular the outdoor air temperature. With a 20 °C indoor-
outdoor temperature difference and the air handler off, the
simulated test results are 15% to 20% low when the SIP zone
is being pressurized by the test and a similar magnitude high
when the zone is being depressurized. When the building air
handler is on, the simulated value is about 10% high when
the SIP zone is pressurized and about 10% low when it is
depressurized. Therefore, interzone, system operation, and
weather effects are expected to lead to roughly 10% errors
in SIP pressurization test results, with worse case conditions
leading to perhaps 20% errors.
          Table 9 Results of simulated  SIP tests
1 1 \
.-||
Air handler off, pressurized



Air handler on, pressurized



Air handler off, depressurize



Air handler on, depressurizec



Actual value in model

0
0
5
5
0
0
5
5
d 0
0
5
5
I 0
0
5
5



0
20
0
20
0
20
0
20
0
20
0
20
0
20
0
20
-


130.0
130.2
130.0
130.2
139.0
140.5
139.2
140.6
130.1
129.9
130.1
129.9
118.7
117.2
118.6
117.1
130.1

176.4
151.4
176.3
150.6
193.7
174.8
193.7
176.9
176.5
194.6
176.4
195.1
155.2
173.8
154.9
171.7
176.5
i
176.4
146.7
174.6
143.5
192.6
166.5
189.1
163.5
176.5
196.7
177.6
197.9
156.5
178.1
160.6
179.6
176.5

-------
4.4 Comparison of Predictions to
Measurements
While it is beyond the scope of this study to conduct the
experiments needed to perform a comprehensive validation
of the CONTAM predictions presented in Section 4.2,
a limited validation exercise is possible based on four
of the tested shelter spaces being part of a tracer gas study
conducted by EPA [13]. In particular, that study involved
the measurement of tracer gas decay rates in the four shelters,
which can be compared with the rates predicted from the
CONTAM simulations. Table 10 shows the decay rates
measured in the four spaces along with the air change rates
         predicted from the CONTAM simulations. The predicted
         rates are based on the measured shelter tightness values for
         each space, extrapolating between the airtightness values
         used in the predictions. The one-zone model predictions are
         used for the NC building spaces, based on the layout of that
         building, while the 2-story model is used for the two spaces in
         the more complex RTF building. In the latter case, predictions
         are presented for both the cold/windy and calm/mild weather
         conditions, as the actual weather conditions are not reported
         with the measurements. While the comparison is very
         limited, the measured and predicted values are in reasonable
         agreement considering the many unknowns impacting the
         predictions and the uncertainties in the measured values.
                Table  10 Comparison  of measured and predicted decay rates
                     NC-Off/A
0.72
   0.66
                     NC-Off/B
0.47
   0.51
                     RTP/C 2.8
0.18
0.17,0.09*
                     RTP/D 3.7
0.09
0.19,0.11*
                : Values predicted for cold/windy and mild/calm conditions.

-------
                                                                                                           5.0
                                                                                          Discussion
The purpose of this project was to develop and demonstrate
methods to assess shelter airtightness in relation to
the protection shelters provide in the event of outdoor
contaminant releases. In actual application, a building owner
or manager would select spaces for use as shelters based on
a number of qualitative considerations identified previously
and perhaps make modifications to increase the degree of
protection offered by the shelter. In the case of unventilated
shelters intended for short-term sheltering, which are the
subject of this study, a key modification is to increase the
airtightness of the shelter through sealing of the boundaries
to adjacent spaces. This project has focused on the
relationship of shelter and building airtightness to the
protection provided, with space pressurization testing
examined as an evaluation method.
Note that this effort focused on unventilated shelters,
in which building occupants are expected to reside for only
an hour or two. When shelters are ventilated and employ
air-cleaning and space pressurization strategies, occupants
can stay in the shelter for longer periods of time because
the level of CO2 will not increase as much or as quickly, nor
will the level of oxygen decrease. Space airtightness is still
important as it impacts the airflow required to pressurize the
space, and pressurization testing is still the method of choice
for quantifying space airtightness. Note also that there is a
wealth of guidance on these more sophisticated, longer-term
SIP approaches [2], and this study does not address ventilation
requirements, air cleaning equipment, and other issues  related
to ventilated shelters. Other important considerations such
as structural design, communications, signage,  medical care,
decontamination of people or the building, and community
planning are addressed in many of the documents cited in
this report.
In reviewing existing SIP guidance, the lack of quantitative
guidance on the shelter airtightness is notably lacking.
There are many recommendations to tighten shelters, but
no information on how tight or how to assess that tightness.
As shown in this study, while tighter shelters (and tighter
buildings) result in better protection against outdoor releases,
they also limit the duration of occupancy in the shelter due to
CO2 buildup. Therefore, one must balance these benefits and
concerns when considering shelter tightness and the use of
unventilated shelters.
Room pressurization testing was seen to be relatively
straightforward as applied in this study. While it is based on a
standard test method for whole building airtightness testing,
its application to individual rooms has yet to be standardized.
Nevertheless, the simulations in this report showed that
interzone pressure effects and system operation in non-test
zones impacted the measurement results by about 10%. The
measured values of airtightness were surprisingly consistent
among shelters tested, as well as the percentage increase in
airtightness through sealing. Under limited sealing the ELA
values were in a relatively narrow range from somewhat
above 1 cmVm2 to about 5 cm2/m2. The sealed values were
lower, as expected, and ranged from 0.25 cmVm2 to just
under 1 cnvYm2. The percent reduction due to sealing was
surprisingly consistent for the four spaces, ranging from
about 60% to 90%.
The simulation of occupant exposure showed that tighter
buildings and tighter shelters reduce exposure of SIP
occupants, with shelter tightness having a greater impact
than building tightness. Longer duration sheltering reduced
the protection factor, as expected, which highlights the
importance of obtaining and communicating reliable
information on when the outdoor hazard has ended and it is
time to end the sheltering period. Based on these simulation
results, CO2 buildup over time may be more critical than
the reduction in exposure. For the building and shelter
airtightness values considered, the duration of sheltering
was seen to be more important to the shelter CO2 level than
airtightness. None of the predicted CO2 concentrations were
as high as the ACGIH short-term (1-min) exposure limit of
54,000 mg/m3, but several exceeded the threshold limit
value (TLV), based on an 8-h exposure over a 40-h work
week, of 9000 mg/m3 [42]. The TLV is not really relevant
to a 1-h or 2-h exposure, but the high values seen in the
simulations are of potential concern. Occupant density,
or floor area per shelter occupant, is obviously an important
determinant of these CO2 concentrations. Many guides
recommend 1 m2 per occupant, but this value produced high
CO2 concentrations for many of the simulations in this study.
A lower value of occupant density, e.g., 2 m2 per occupant,
might provide more habitable conditions for longer periods
of time and merits consideration in future recommendations.
While the simulation results provided some useful insights,
the results are highly dependent on the building models
employed, including layout, airtightness, outdoor weather,
and indoor temperatures. However, the relative results
between the simulated cases are less sensitive to these
factors than the  absolute results.
This study does suggest some additional work to advance
these results in the future. This additional work would include
more field testing of different SIP spaces and the development
of a standard test protocol. Ultimately, these efforts could
support definitive airtightness criteria for unventilated
shelters. The results of this study support a target airtightness
value of 1 cm2/m2 and limiting unventilated sheltering to 2 h
at the most, but these suggestions should not be considered
as universally applicable recommendations.
The results of this study could be further supported by tracer
gas studies. These studies could simulate an outdoor release

-------
and then be used to calculate actual protection factors for
comparison to simulation results. In this way, the relationship
between shelter airtightness and building protection could be
more reliably demonstrated. In addition, it may be possible
to develop tracer gas methods to determine protection factors
directly without the intermediate step of calculating interzone
airflows. These methods would involve releasing a tracer gas
into a shelter space and then monitoring its decay over time,
which in turn could potentially be related to the protection
factor in the face of an outdoor release. While such a method
has not been developed and would require a much higher
level of expertise than pressurization testing, such an approach
may be more accurate than an estimate of protection based on
shelter airtightness.

-------
                                                                                       6.0
                                                            Acknowledgments
This effort was supported by the U.S. Environmental         The authors express their appreciation to John Chang, Jacky
Protection Agency under Interagency Agreement            Rosati, and Jim Jetter at EPA. William Bahnfleth reviewed
No. DW-13-92178301-0 but was not subjected to EPA peer     this report and made many helpful suggestions that are greatly
review. The conclusions in this paper are therefore those of     appreciated by the authors.
the authors and are not necessarily those of the U.S. EPA.

-------

-------
                                                                                                      7.0
                                                                                      References
[1] NIOSH. 2002. Guidance for Protecting Building Environments from Airborne Chemical, Biological, or
Radiological Attacks. National Institute of Occupational Safety and Health, DHHS (NIOSH) Publication
No. 2002-139.
[2] FEMA. 2006. Design Guidance for Shelters and Safe Rooms. Federal Emergency Management Agency
Report No. FEMA 453.
[3] Janney, C., Janus, M., Saubier, L.F., and Widder, J. 2000. System Effectiveness Test of Home/
Commercial Portable Room Air Cleaners. Battelle Report to U.S. Army Soldier, and Biological Chemical
Command (SBCCOM), Contract No. SPO900-94-D-0002.
[4] Cristy, G.A. and Chester, C.V 1981. Emergency Protection from Aerosols. Oak Ridge National
Laboratory, Report No. ORNL-5519.
[5] Blewett, W.K. and Area, V.J. 1999. Experiments in Sheltering in Place: How Filtering Affects Protection
Against Sarin and Mustard Vapor. U.S. Army Edgewood Chemical Biological Center, Report No. ECBC-
TR-034.
[6] Fradella, J. and Siegel, J. A. 2005. An Evaluation of Shelter-in-Place Strategies in Four Industrial
Buildings. Proceedings of 10th International Conference on Indoor Air Quality and Climate: 3360-3364.
[7] Engelmann, R.J. 1990. Effectiveness of Sheltering in Buildings and Vehicles for Plutonium. U.S.
Department of Energy, Report No. DOE/EH-0159T UC-160.
[8] Engelmann, R.J. 1992. Sheltering Effectiveness Against Plutonium Provided by Buildings, Atmospheric
Environment, 26A(11): 2037-2044.
[9] Sohn, M.D., Sextro, R.G., and Lorenzetti, D.M. 2005. Assessing Sheltering-in-Place Responses to
Outdoor Toxic Releases.  Indoor Air, 15(11):  1792-1796.
[10] Rogers, G.O., Watson, A.P, Sorensen, J.H., Sharp, R.D., and Carnes, S.A. 1990. Evaluating Protective
Actions for Chemical Agent Emergencies. Oak Ridge National Laboratory, Report No. ORNL-6615.
[11] Yuan, L.L. 2000. Sheltering Effects of Buildings from Biological Weapons. Science & Global Security,
8:287-313.
[12] Jetter, J.J. and Whitfield, C. 2005. Effectiveness of Expedient Sheltering in Place in a Residence.
Journal of Hazardous Materials, A119: 31-40.
[13] Jetter, J. and Proffitt, D. 2006. Effectiveness of Expedient Sheltering in Place in Commercial Buildings.
Journal of Homeland Security and Emergency  Management, 3 (2), Article 4.
[14] Blewett, W.K., Reeves, D.W., Area, V.J., Fatkin, D.P, and Cannon, B.D. 1996. Expedient Sheltering
in Place: An Evaluation for the Chemical Stockpile Emergency Preparedness Program. Chemical Research,
Development & Engineering Center, Report No. ERDEC-TR-336.
[15] Sorenson, J.H. and Vogt, B.M. 2001. Expedient Respiratory and Physical Protection: Does a Wet Towel
Work to Prevent Chemical Warfare Agent Vapor Infiltration? Oak Ridge National Laboratory, Report No.
ORNL/TM-2001/153.
[16] Sorenson, J.H. and Vogt, B.M. 2001. Will  Duct Tape and Plastic Really Work? Issues Related to
Expedient Shelter-in-Place. Oak Ridge National Laboratory, Report No. ORNL/TM-2001/154.
[17] CSEPP 2001. Report of the Shelter-in-Place Work Group. Chemical Stockpile Emergency
Preparedness Program.
[18] Shumpert, B. 2003.  Questions and Answers Regarding Actions to Take When Ending Shelter-in-Place.
Oak Ridge National Laboratory, Report No.  ORNL/TM- 2003/230.
[19] Sorenson, J., Shumpert, B., and Vogt, B. 2002. Planning Protective Action Decision-Making: Evacuate
or Shelter-in-Place? Oak Ridge National Laboratory, Report No. ORNL/TM- 2002/144.
[20] Yantosik, G.D., Lerner, K., Maloney, D., and Wasmer, F. 2001. When and How to End Shelter-in-Place
Protection From a Release of Airborne Hazardous Material: Report on a Decision-Making Concept and
Methodology. Argonne National Laboratory.

-------
[21] Yantosik, G.D., Lerner, K., and Maloney, D.M. 2001. Temporary Shelter-in-Place as Protection Against
a Release of Airborne Hazardous Material: Report of a Literature Search. Argonne National Laboratory.
[22] Yantosik, G.D., Maloney, D., and Wasmer, F. 2003. Comparison of Two Concepts and Methods to
Decide When to End Temporary Shelter-in-Place Protection. Argonne National Laboratory.
[23] Vogt, B.M., Hardee, H.K., Sorensen, J.H., and Shumpert, B.L. 1999. Assessment of Housing Stock Age
in the Vicinity of Chemical Stockpile Sites. Oak Ridge National Laboratory, Report No. ORNL/TM- 13742.
[24] FEMA. 1999. National Performance Criteria for Tornado Shelters. Federal Emergency Management
Agency, Report No. 361.
[25] Chester, C.V. 1988. Technical Options for Protecting Civilians from Toxic Vapors and Gases. Oak
Ridge National Laboratory, Report No. ORNL/TM-10423.
[26] Mannan, M.S. and Kilpatrick, D.L. 2000. The Pros and Cons of Shelter-in-Place, Process Safety
Progress. 19(4)210-218.
[27] NICS.  1999. Shelter in Place at Your Office. National Institute of Chemical Studies.
[28] NICS.  2001. Sheltering in Place as a Public Protective Action. National Institute of Chemical Studies.
[29] AFCES. 2001. Protective Actions for a Hazardous Material Release. Air Force Civil Engineer Support
Agency, Tyndall Air Force Base.
[30] Chan, W.R. and Gadgil, AJ. 2004.  Sheltering in Buildings from Large-Scale Outdoor Releases. Air
Infiltration and Ventilation Centre, Report VIP No. 10.
[31] U.S. Army Corps of Engineers. 1998. Design of Chemical Agent Collective Protection Shelters for
New and Existing Facilities. Report No. ETL 1110-3-490.
[32] U.S. Army Corps of Engineers. 1999. Design of Chemical Agent Collective Protection Shelters to
Resist Chemical, Biological, and Radiological (CBR) Agents. Report No. ETL 1110-3-498.
[33] Price, P.N., Sohn, M.D., Gadgil, A.J., Delp, W.W., Lorenzetti, D.M., Finlayson, E.U., Thatcher, T.L.,
Sextro, R.G., Derby, E.A., and Jarvis, S.A. 2003. Protecting Buildings From a Biological or Chemical
Attack: Actions to take before or during a release. Lawrence Berkeley National Laboratory, Report No.
LBNL-51959.
[34] Walton, G.N. and Dols, WS. 2005. CONTAMW 2.4 User Guide and Program Documentation. National
Institute of Standards and Technology, Report No. NISTIR 7251.
[35] Persily, A.K. and Ivy, E.M. 2001. Input Data for Multizone Airflow and IAQ Analysis. National
Institute of Standards and Technology, Report No. NISTIR 6585.
[36] ASTM. 2003. Standard E779-03, Standard Test Method for Determining Air Leakage Rate by Fan
Pressurization. American Society for Testing and Materials.
[37] ASTM. 2007. Standard £1827-96(2007), Standard Test Method for Determining Airtightness of
Buildings Using an Orifice Blower Door. American Society for Testing and Materials.
[38] ASTM. 2008. Standard £1258-88(2008), Standard Test Method for Airflow Calibration of Fan
Pressurization Devices. American Society for Testing and Materials.
[39] ASHRAE. 2005. Fundamentals Handbook. American Society of Heating, Refrigerating and Air-
Conditioning Engineers.
[40] Emmerich, SJ. and Persily, A.K. 2005. Airtightness of Commercial Buildings in the U.S. Proceedings
of 26th AIVC Conference: 65-70.
[41] ASHRAE. 2007. ANSI/ASHRAE Standard 62.1-2007, Ventilation for Acceptable Indoor Air Quality.
American Society of Heating, Refrigerating and Air-Conditioning Engineers.
[42] ACGIH. 2001. 2001 Threshold Limit Values (TLVs) for Chemical Substances and Physical Agents and
Biological Exposure Indices (BEIs). American Conference of Governmental Industrial Hygienists.

-------

-------
&EPA
     United States
     Environmental Protection
     Agency
PRESORTED STANDARD
 POSTAGES FEES PAID
         EPA
   PERMIT NO. G-35
     Office of Research and Development
     National Homeland Security Research Center
     Cincinnati, OH 45268


     Official Business
     Penalty for Private Use
     $300
             Recycled/Recyclable
             Printed with vegetable-based ink on
             paper that contains a minimum of
             50% post-consumer fiber content
             processed chlorine free

-------