December 2011
Environmental Technology
Verification Report
VERIFICATION OF BUILDING PRESSURE
CONTROL AS CONDUCTED BY GSI
ENVIRONMENTAL, INC. FOR THE
ASSESSMENT OF VAPOR INTRUSION
Prepared by
Battelle
Baltelle
Ins Business oj Innovation
Under a cooperative agreement with
U.S. Environmental Protection Agency
ET1/ET1/ET1/
-------
December 2011
Environmental Technology Verification
Report
ETV Advanced Monitoring Systems Center
VERIFICATION OF BUILDING PRESSURE
CONTROL AS CONDUCTED BY GSI
ENVIRONMENTAL, INC. FOR THE ASSESSMENT
OF VAPOR INTRUSION
by
Ian MacGregor, Mary Prier, Dale Rhoda, and Amy Dindal, Battelle
John McKernan, U.S. EPA
-------
Notice
The U.S. Environmental Protection Agency, through its Office of Research and Development,
funded and managed, or partially funded and collaborated in, the research described herein. It
has been subjected to the Agency's peer and administrative review. Any opinions expressed in
this report are those of the author(s) and do not necessarily reflect the views of the Agency,
therefore, no official endorsement should be inferred. Any mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
11
-------
Foreword
The EPA is charged by Congress with protecting the nation's air, water, and land resources.
Under a mandate of national environmental laws, the Agency strives to formulate and implement
actions leading to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, the EPA's Office of Research and
Development provides data and science support that can be used to solve environmental
problems and to build the scientific knowledge base needed to manage our ecological resources
wisely, to understand how pollutants affect our health, and to prevent or reduce environmental
risks.
The Environmental Technology Verification (ETV) Program has been established by the EPA to
verify the performance characteristics of innovative environmental technology across all media
and to report this objective information to permitters, buyers, and users of the technology, thus
substantially accelerating the entrance of new environmental technologies into the marketplace.
Verification organizations oversee and report verification activities based on testing and quality
assurance protocols developed with input from major stakeholders and customer groups
associated with the technology area. ETV consists of six environmental technology centers.
Information about each of these centers can be found on the Internet at http://www.epa.gov/etv/.
Effective verifications of monitoring technologies are needed to assess environmental quality
and to supply cost and performance data to select the most appropriate technology for that
assessment. Under a cooperative agreement, Battelle has received EPA funding to plan,
coordinate, and conduct such verification tests for "Advanced Monitoring Systems for Air,
Water, and Soil" and report the results to the community at large. Information concerning this
specific environmental technology area can be found on the Internet at
http://www.epa.gov/etv/centers/centerl.html.
in
-------
Acknowledgments
The authors wish to acknowledge the contribution of the many individuals, without whom this
verification testing would not have been possible. Quality assurance oversight was provided by
Jonathan Tucker, NAVFAC Atlantic, Michelle Henderson and Laurel Staley, U.S. EPA and
Rosanna Buhl, Battelle. We gratefully acknowledge Dr. Bart Chadwick and Dr. Ignacio Rivera-
Duarte at SPAWAR Systems Center Pacific for sponsoring this work through the Navy
Environmental Sustainability Development to Integration Program, as part of Project 424 on
"Improved Assessment Strategies for Vapor Intrusion." This technology was evaluated
concurrently in a project sponsored by the Environmental Security Technology Certification
Program (ESTCP), and the contribution of effort from ESTCP Project ER-0707 in the
implementation of this verification test is also gratefully acknowledged. We thank the various
participants on the panel of technical experts who provided input to and reviewed the Quality
Assurance Project Plan. Moreover, we thank Dr. Brian Schumacher (U.S. EPA), Ms. Donna
Cal dwell (NAVFAC Atlantic), and Dr. Ronald Mosley (US EPA, retired) for their review of this
verification report. Finally we gratefully acknowledge the work of the technology vendor, Dr.
Thomas McHugh and Ms. Lila Beckley, at GSI Environmental, Inc.
IV
-------
Contents
Page
Foreword iii
Acknowledgments iv
List of Abbreviations ix
Chapter 1 Background 1
Chapter 2 Technology Description 2
Chapters Test Design and Procedures 5
3.1 Test Overview 5
3.2 Test Site Descriptions 7
3.2.1 ASU VI Research House 7
3.2.2 Moffett Field Building 107 8
3.3 Experimental Design 9
3.3.1 Decision-making Support 17
3.3.1.1 Building Pressure Differential 18
3.3.1.2 Vapor Intrusion Enhancement and Reduction 18
3.3.1.3 Fractional Contribution of Vapor Intrusion to Indoor CoC Concentrations 20
3.3.2 Comparability 22
3.3.3 Operational Factors 22
3.3.4 Validation of Mosley Model Assumptions 23
Chapter 4 Quality Assurance/Quality Control 25
4.1 Quality Control Results 25
4.2 Data Quality Indicators 29
4.3 Audits 29
4.3.1 Technical Systems Audits 30
4.3.2 Audits of Data Quality 31
Chapters Statistical Methods 34
5.1 Decision-making Support 34
5.1.2 Vapor Intrusion Enhancement And Reduction 35
5.1.3 Fractional Contribution of Vapor Intrusion to indoor CoC concentrations 36
5.2 Comparability 37
5.3 Verification of Model Assumptions 37
Chapter 6 Test Results 40
6.1 Measurement Results From Both Buildings 40
6.1.1 Indoor/Outdoor and Cross-Foundation Pressure Differentials 40
6.1.2 Building Ventilation Rates 41
6.1.3 Concentrations of Compounds in Ambient Air 42
6.1.4 Concentrations of Compounds in Indoor Air 43
6.1.5 Concentrations of Compounds in Sub-Slab Soil Gas 47
6.1.6 Mass Discharges 49
6.2 Decision-making Support 53
6.2.1 Building Pressure Differential 53
6.2.2 Vapor Intrusion Enhancement and Reduction 54
6.2.3 Fractional Contribution of Vapor Intrusion to Indoor CoC Concentrations 56
-------
6.3 Comparability 59
6.4 Operational Factors 59
6.5 Validation of Model Assumptions 60
Chapter 7 Performance Summary 65
Chapters References 69
Appendix A Supplemental Information 72
VI
-------
Tables
Table 1. Mosley Model Notation Used for Description of Several Verification Parameters 7
Table 2. Days/times for Pressure Control Testing at Each Building 10
Table 3. Types of and Locations for Air Samples Collected During Each of the Three Pressure
Perturbation Periods 14
Table 4. Summary of Sample Types and Timing for Sample Collection at Each Test Building. 15
Table 5. Summary of Results of Various QC Procedures and Samples 26
Table 6. Summary of Frequency of Measurements Lower Than Estimated MDLs 29
Table 7. Indoor/outdoor pressure differentials 53
Table 8. Comparison of Radon Mass Discharges from Subsurface Sources to Determine VI
Enhancement and Reduction 55
Table 9. Comparison of Indoor and Ambient Air Radon Concentrations under Positive
Pressure 56
Table 10. Fractional Contribution of Ambient Sources, Indoor Sources, and VI to Indoor CoC
Concentrations Under Baseline Conditions 57
Table 11. Comparability of Building Pressure Control Results 59
Table 12. Validation of Model Assumption 1 61
Table 13. Validation of Model Assumption 2 62
Table 14. Validation of Model Assumptions 3, 4, and 5 62
Table 15. Validation of Model Assumptions 6, 7, and 8 63
Table 16. Minimum Detectable Differences for Model Assumptions 1 and 2 at Moffett Field
Building 107 64
vn
-------
Figures
Figure 1. Basis of Building Pressure Control Technique for the Assessment of the Impact of VI
on Concentrations of CoCs in Indoor Air 3
Figure 2. Delivery of SFe to the Building Atmosphere; Collection of SS Air Sample with a PVF
Bag; and Collection of an IA Sample into a Stainless Steel Canister 4
Figure 3. Photographs of the ASU VI Research House 8
Figure 4. Photographs of Moffett Field Building 107 9
Figure 5. Pressure Differential Measurements at ASU House 11
Figure 6. SFe Tracer Gas Delivery System as Deployed at Moffett Field Building 107 12
Figure 7. Fan Installed for Building Pressure Control at the ASU House 13
Figure 8. IA, AA, and SS sampling at ASU House 16
Figure 9. Indoor/Outdoor and Cross-Foundation Differential Pressure Measurements under
Three Different Pressure Conditions 41
Figure 10. Building Ventilation Rates Measured under Three Different Pressure Conditions.... 42
Figure 11. Concentrations of Compounds Measured in Ambient Air 43
Figure 12. Concentrations of Compounds Measured in Indoor Air 44
Figure 13. Average Indoor Air Concentrations Normalized by Ambient Concentrations 46
Figure 14. Concentrations of Compounds Measured in Sub Slab Soil Gas 47
Figure 15. Average Sub-Slab Concentrations Normalized by Average Indoor Air
Concentrations 48
Figure 16. Normalized Mass Discharges at ASU House for Radon, TCE, 1,1-DCE, SFe, Benzene
and Toluene 51
Figure 17. Normalized Mass Discharges at Moffett Field Building 107 for Radon, TCE, PCE,
SFe, Benzene, and Toluene 52
Vlll
-------
List of Abbreviations
1,1-DCE 1,1-dichloroethylene
AP differential pressure
AFyi error in FVI
AA ambient air
AER air exchange rate
ADQ audit of data quality
AMS Advanced Monitoring Systems
ASU Arizona State University
BL baseline
COA certificates of analysis
CoC(s) contaminant(s) of concern
DQIs data quality indicators
DQOs data quality objectives
EPA U.S. Environmental Protection Agency
ESTCP Environmental Security Technology Certification Program
ETV Environmental Technology Verification
Fa fractional contribution of ambient air to the indoor concentration of a CoC
F;n fractional contribution of indoor sources to the indoor concentration of a CoC
FVI fractional contribution of vapor intrusion to the indoor concentration of a CoC
GC/ECD gas chromatography with electron capture detection
GC/MS gas chromatography/mass spectrometry
HO null hypothesis
HI alternative hypothesis
Hg mercury
HVAC heating, ventilating and air conditioning
IA indoor air
I/O indoor/outdoor
LRB laboratory record book
MDL method detection limit
MDT Mountain Daylight Time
NAVFAC Naval Facilities Engineering Command
NIOSH National Institute of Occupational Safety and Health
NP negative pressure
pCi L"1 picocuries per liter
Pa Pascal
PCE perchloroethylene (tetrachloroethylene)
PDT Pacific Daylight Time
PP positive pressure
PVF polyvinyl fluoride
QA quality assurance
QAO Quality Assurance Officer
QAPP quality assurance project plan
QC quality control
QMP Quality Management Plan
Rn Radon
RPD relative percent difference
ix
-------
sulfur hexafluoride
SIM single ion monitoring
SPAWAR Space and Naval Warfare
SS subslab
TCE trichloroethylene
TSA technical systems audit
ug m"3 microgram per cubic meter
VI vapor intrusion
VOC volatile organic compound
VTC Verification Test Coordinator
-------
Chapter 1
Background
The U.S. Environmental Protection Agency (EPA) supports the Environmental Technology
Verification (ETV) Program to facilitate the deployment of innovative environmental
technologies through performance verification and dissemination of information. The goal of the
ETV Program is to further environmental protection by accelerating the acceptance and use of
improved and cost-effective technologies. ETV seeks to achieve this goal by providing high-
quality, peer-reviewed data on technology performance to those involved in the design,
distribution, financing, permitting, purchase, and use of environmental technologies.
ETV works in partnership with recognized testing organizations; with stakeholder groups
consisting of buyers, vendor organizations, and permitters; and with the full participation of
individual technology developers. The program evaluates the performance of innovative
technologies by developing test plans that are responsive to the needs of stakeholders,
conducting field or laboratory tests (as appropriate), collecting and analyzing data, and preparing
peer-reviewed reports. All evaluations are conducted in accordance with rigorous quality
assurance (QA) protocols to ensure that data of known and adequate quality are generated and
that the results are defensible. The definition of ETV verification is to establish or prove the
truth of the performance of a technology under specific, pre-determined criteria or protocols and
a strong quality management system. The highest-quality data are assured through
implementation of the ETV Quality Management Plan. ETV does not endorse, certify, or
approve technologies.
The EPA's National Risk Management Research Laboratory (NRMRL) and its verification
organization partner, Battelle, operate the Advanced Monitoring Systems (AMS) Center under
ETV. The AMS Center recently evaluated the performance of the building pressure control
technique for the assessment of the impact of vapor intrusion (VI) on the concentrations of
contaminants of concern (CoCs) in indoor air (IA). The pressure control technique was
conducted by the technology vendor, GSI Environmental, Inc.
-------
Chapter 2
Technology Description
This report provides results for the verification testing of a building pressure control technique
for the assessment of the impact of VI on the concentrations of CoCs in IA. This section provides
information on why developing such a technique is important, as well as a description of the
pressure control technique itself. GSI Environmental, Inc. was the technology vendor
conducting this technique for this verification test.
VI is the migration of volatile chemicals from the subsurface (from soils and/or groundwater)
into the air of overlying buildings.1 Adverse health effects may result from inhalation exposure
to certain CoCs such as the volatile organic compounds (VOCs) trichloroethylene (TCE),
tetrachloroethylene (perchloroethylene, PCE), 1,1-dichloroethylene (1,1-DCE), and benzene.1
Reducing or controlling the risk to human health related to inhalation exposure of CoCs due to
VI is the stated goal of many regulatory and governmental agencies. That said, many building
owners and regulated entities (such as the U.S. Navy)2'3 have developed policies and guidance to
state that they are not responsible for the mitigation of CoCs in the IA of structures in cases
where the CoCs are present due to natural or anthropogenic background sources.11 Thus, the
ability to distinguish concentrations of CoCs in background IA - defined for CoCs as everything
unrelated to the vapors that migrate into the overlying structure (from sources such as household
activities, consumer products, and building materials)4 - from CoCs present due to VI is of key
importance so that regulated entities can appropriately manage their limited resources when
making remediation and mitigation decisions. However, at present little guidance is available to
determine the impact of VI compared to the impact of natural or anthropogenic background
sources on indoor concentrations of CoCs. One technique that has shown promise for
distinguishing background indoor sources of CoCs from those present due to VI is the
manipulation of building pressure.5'6'7 Other work8 in this area has shown that radon occurs
naturally in soil gas due to the radioactive decay of uranium, and as a result, in ambient air (AA)
at concentrations of 0.2 to 0.7 picocuries per liter (pCi L"1).9 Therefore, radon may be used to
evaluate the VI of CoCs. The performance of the method of measuring radon and CoCs under
different building pressures to assess the impact of VI on the concentrations of CoCs in IA (the
"building pressure control technique") is the subject of this verification test.
Intentionally inducing negative pressure (NP) or positive pressure (PP) in the building- by use of
a fan to drive IA out of the building, or AA into the building, respectively - should enhance or
1VOC and CoC are used interchangeably throughout this report.
11 Navy guidance states that chemicals from background sources should not be considered CoCs. However, in his
report the term CoC may refer to chemicals from either background or VI sources.
-------
reduce VI. This is the conceptual basis for the building pressure control technique and is shown
in Figure 1. Under conditions of induced NP (top panel), VI should be enhanced; under induced
PP, VI should be stopped or reduced, as shown in the bottom panel. Arrows in the figures
indicate the expected direction of air flows. During implementation of the building pressure
control method, various types of air samples are collected to demonstrate VI manipulation, as
shown by the various symbols in the figure.
Induced
NEGATIVE
Building
Pressure
Induced
POSITIVE
Building
Pressure
Figure 1. Basis of Building Pressure Control Technique for the Assessment of the Impact
of VI on Concentrations of CoCs in Indoor Air. (Figure courtesy of GSI.)
Implementation of the building pressure control technique for the assessment of the impact of VI
on the IA at a given building takes place over approximately 3!/2 days. Over the first half day,
the building is prepared for testing. This includes installation of three subslab (SS) sampling
points through the building's concrete foundation as well as setting up and verifying the
operation of the various air sampling equipment and instrumentation. Over the next 24 hours,
the building is maintained under baseline (BL) pressure where the building pressure is not
intentionally manipulated. Over the following 24 hours, a NP is induced in the building. Over
the final 24 hours, a PP is induced in the building. To accomplish building pressurization and
depressurization, windows, and other openings are closed111 and a fan is installed in a doorway or
window.
During each 24 hour period of BL, NP, and PP testing, a known concentration of the tracer gas,
sulfur hexafluoride or SFe, is released at a known flow rate from a centralized location in the
m Doors and windows are closed, but sealing egresses and vents is not attempted.
3
-------
building. To the extent possible, indoor doors remain open throughout testing to enhance mixing
of the IA. Using the known flow rate of SF6 and measurements of indoor SF6 concentrations, the
flow rate of AA into the building, that is, the building's air exchange rate (AER) may be
determined. Real-time measurement of the differential pressure (AP) across the building
envelope (the indoor/outdoor (I/O) AP) and the building foundation are performed throughout
BL, NP, and PP testing.
Finally, several different types of air samples from inside, outside, and below the building - for
IA, AA, and SS soil gas, respectively - are also collected and analyzed to characterize
concentrations of various CoCs, SF6, and radon in these three compartments." Gas samples for
analysis of CoCs and SFe are collected into stainless steel sampling canisters; whereas samples
for radon analysis are collected into polyvinyl fluoride (PVF) DuPont™ Tedlar® gas sampling
bags, or measured in near real-time using an instrument designed for this purpose. While the
building is under each of the three pressure conditions, IA and SS concentrations of CoCs, SF6,
and radon are measured at three different spatially distributed locations throughout the building
and at a single outdoor location. Shown schematically in Figure 2 is the SFe delivery system, SS
sampling for radon into PVF bags, and IA sampling for VOCs and SFe into a stainless steel
canister. Canisters and PVF bags are delivered to separate off-site contract analytical
laboratories for gas analysis.
Figure 2. Delivery of SFe to the Building Atmosphere; Collection of SS Air Sample with a
PVF Bag; and Collection of an IA Sample into a Stainless Steel Canister. (Figure courtesy
ofGSI.)
1V For this verification test, more samples were collected than are needed for routine implementation of this
technology. For example, IA, AA, and SS air samples were collected. Routine implementation may be performed
without collection of SS samples.
-------
Chapter 3
Test Design and Procedures
3.1 Test Overview
This verification test was conducted according to procedures specified in the Quality Assurance
Project Plan for Verification of Building Pressure Control for the Assessment of Vapor
Intrusion1® (QAPP) and adhered to the quality system defined in the ETV AMS Center Quality
Management Plan (QMP).11 As indicated in the QAPP, the testing conducted satisfied EPA QA
Category III requirements. A panel of technical experts was convened to provide input to the
QAPP development. The following experts provided input to the QAPP and provided a peer
review of the QAPP and/or this verification report.
Ms. Donna Caldwell, U. S. Navy, NAVFAC Atlantic
• Mr. Douglas Grosse, EPA, National Risk Management Research Laboratory
• Dr. Ronald Mosley, EPA (retired)
Dr. Brian Schumacher, EPA, National Exposure Research Laboratory
• Ms. Lynn Spence, Spence Environmental Consulting
In addition, the VI technology category was reviewed with the broader AMS Center Stakeholder
Committees during regular stakeholder teleconferences, including the November 5 and 12, 2009
meetings, and input from those committees was solicited.
Battelle conducted this verification test with funding support from the U.S. Navy SPAWAR
Systems Center Pacific through funding from the Navy Environmental Sustainability
Development to Integration Program, as part of Project 424 on "Improved Assessment Strategies
for Vapor Intrusion". The subject technology is concurrently being evaluated in project ER-0707
sponsored by ESTCP.
The purpose of this verification test was to generate performance data on the use of the building
pressure control technique as a method to understand the impact of VI on the concentrations of
CoCs in IA. In general, the data generated from this verification test are intended to provide
organizations and users with information on the ability of this methodology to assess VI impacts.
GSI Environmental staff, with oversight from Battelle, implemented the building pressure
control technique at two different buildings (described later in this Chapter); testing was
executed in the autumn of 2010, over the course of 3.5 days at each building. The pressure
-------
control technique was evaluated at the two buildings using the following types of performance
parameters:
• Decision-making support
• Comparability
• Operational factors
The overall goal of implementing the building pressure control method is to obtain a better
understanding of VI in a building. For instance, if the control of building pressure results in
clear changes in building pressures and CoC and radon concentrations, the pressure control
method may yield results that are useful for decision-making (i.e., is VI a concern for this
building?). The effectiveness of the building pressure control method to support decision-
making was evaluated through three different metrics. The first metric under decision-making
support is to understand if the building pressure can be decreased and controlled and
subsequently elevated and controlled at each of the two buildings under induced NP and PP
conditions, respectively. The next metric was to determine, by inspection of the mass discharge
of radon from subsurface sources whether VI was in fact enhanced under NP and reduced (or
stopped) under PP. Demonstration of control of radon VI by manipulation of building pressure
is important since it should allow for concomitant control of CoC VI. The last sub-parameter
under decision-making support is the calculation of the fractional contribution of VI (Fvi) for
each of several different concentrations of indoor CoCs. FVI was calculated for four different
CoCs at each of the two test buildings. Of the four CoCs, two were among those expected to
have subsurface sources [trichloroethylene (TCE), and either 1,1-dichloroethylene (1,1-DCE) or
tetrachloroethylene (PCE)], and two others were CoCs not expected to be present in IA as a
result of VI (benzene and toluene). FVI for each CoC was calculated at each of the two buildings
under both NP and PP conditions according to an indoor air quality model developed by Dr.
Ronald Mosley (i.e., the Mosley Model).12 The error in each FVI (AFVi) calculation was also
estimated based on a Monte Carlo error estimation technique. Given FVI ± AFvi, decision-
makers may evaluate the impact of VI on the indoor atmosphere by calculation of the indoor
concentration of each CoC attributable to VI and comparison of this result to appropriate
regulatory criteria. Additional support to decision-makers was also provided by qualitative
trends, with respect to changes in building pressure, in concentrations of compounds in IA as
well as trends in the changes of compound mass discharges.
FVI was calculated using the Mosley Model that is presented and described in its entirety in the
QAPP.10 The Mosley Model notation is used throughout this report since this facilitates the
presentation of various results and verification metrics. Other notation was also developed based
on Mosley's use of superscripts and subscripts to specify building pressure, as shown in Table 1.
-------
Table 1. Mosley Model Notation Used for Description of Several Verification Parameters.
Parameter, units
Subscripts
Superscripts
R = radon concentration, pCi L"1 or pCi m"3
Q = flow rate, m3 h"1
C = CoC concentration, ug m"3
T = tracer gas concentration, ug m"3
G = generation rate of a compound by
indoor sources, ug h"1 or pCi h"1
E = entry rate of a compound from a
subsurface source, ug h"1 or pCi h"1
F = fractional contribution of the
concentration of a CoC, unitless
i = indoor air
a = ambient air
s = soil gas
T = tracer
C = CoC
R = radon
VI = vapor
intrusion
+ = positive pressure
- = negative pressure
(no superscript) = baseline conditions
(no pressure perturbation)
Other symbols and values:
V = building volume, m3
X = radioactive decay constant for 222Rn, 0.1813 d"1 = 0.007555 h"1 (half life = 3.823 d; reference 13)
Qi/V = building air exchange rate (AER), h"1
Beyond the three metrics comprising decision-making support, the metric of comparability was
assessed for the pressure control technique as the similarity of the building envelop differential
pressures achieved under induced NP and PP conditions at each of two buildings. The final
performance metric was comprised of an assessment of operational factors such as ease of
implementation of the pressure control technology, the expertise required to carry out the field
work and interpret the results, and costs to perform the testing.
3.2 Test Site Descriptions
3.2.1 ASU VI Research House
Arizona State University (ASU) purchased this research house (referred to as the "ASU House")
near Hill Air Force Base in Layton, UT, for use on Strategic Environmental Research and
Development Program Project ER-1686. This building overlies a dissolved plume of TCE and
1,1-DCE and as part of the work on ER-1686, ASU has confirmed that VI of these compounds is
occurring at this building. Hill Air Force Base has deployed a near real-time gas chromatograph
mass spectrometer (GC/MS), the HAPSITE® Smart Chemical Identification System (Inficon,
East Syracuse, New York), that measures the IA concentrations of CoCs every hour.
Photographs of the house are shown in Figure 3. The floor plan of the home is shown in Figure
Al in Appendix A. The building is an unoccupied single-family dwelling with a partially below-
grade finished basement and a single story living space above the basement. The area and
-------
volume of the living space in the building, determined by measurement of the inner building
dimensions, are 114 m2 and 273.5 m3, respectively. The area and volume calculations exclude
the garage, since during testing the door between the living space and garage generally remained
closed. The volume also excludes any attic space. During testing field staff were present in the
home between the hours of- 06:30 and 18:00 Mountain Daylight Time (MDT). During all
testing interior doors remained open (other than the door between the living space and garage),
windows were closed, the fan to induce the pressure perturbation was kept running, and the
building's heating, ventilation and air conditioning (HVAC) system operated normally. The
external garage door (that would allow ingress/egress of vehicles) remained closed during
testing, but otherwise building egresses were not strictly controlled and testing staff moved about
freely.
Figure 3. Photographs of the ASU VI Research House. Panel A, Front, and Panel B, Rear
of the Building.
3.2.2 Moffett Field Building 107
A number of buildings at Naval Air Station Moffett Field, near Palo Alto, CA, are impacted by
subsurface sources of TCE and PCE.14 The building selected for this verification test was
Building 107, used by the U.S. Navy. It is a single story slab-on-grade structure and is shown in
the photographs in Figure 4. The floor plan of the building is shown in Figure A2 in Appendix
A. The area and volume of the usable space of the building, determined by measurement of its
inner dimensions, are approximately 154 m2 and 365 m3, respectively. The volume calculation
excluded the void space between the drop ceiling and roof. The building was occupied by Navy
personnel and verification testing staff between the hours of- 06:30 and 18:00 Pacific Daylight
Time (PDT) on test days. During all testing interior doors remained open, the fan to induce the
appropriate pressure perturbation was kept running, exterior windows were closed, and the
building's HVAC system operated normally, but building egresses were not controlled and
building occupants were allowed to come and go freely.
-------
•SB-
4. Photographs of Moffett Field Building 107. Panel A Shows the West Side of
Building, Panel B the Southeast Corner, and Panel C the Northeast Corner.
3.3 Experimental Design
The test schedule and experimental procedures are discussed in detail in the QAPP. Two back-
to-back pressure control tests were conducted at each building. Both tests at each building were
included as part of GSI Environmental's ESTCP project ER-0707. Only the second pressure
control test at each building was included in the present verification. The initial pressure control
test that occurred at each building was nominally identical with respect to duration, types of
sampling performed, pressure control sequence, etc., to the ETV test that followed/ The ETV
portions of the field work were conducted Monday, October 4 through Thursday, October 7,
2010 at the ASU House and Sunday, October 31 through Wednesday, November 3, 2010 at
Moffett Field Building 107. Beginning late in the afternoon on the first day of testing, and
lasting over the next three consecutive days, each building was maintained for 24 hours at each
of the three pressure perturbation conditions (BL, NP, and PP). During the first 12 hours at each
v Conducting back-to-back building pressure tests may result in anomalous BL building conditions during the
second set of tests. However, results generated during the ESTCP-only tests demonstrate that contaminant
concentrations and mass discharges under BL conditions were similar for both this initial test and subsequent ETV
test. This outcome is consistent with the conclusion that BL results for the ETV test are representative of normal
building conditions.
-------
pressure condition, the building atmosphere was allowed to come to equilibrium, after which the
next 8 to 12 hours was taken to characterize the concentrations of various species in the building
atmosphere/1 Table 2 shows the timing for each of the pressure control tests at each of the two
buildings.
Table 2. Days/times for Pressure Control Testing at Each Building
ASU VI Research House (times MDT)
Pressure condition Begin End
BL 10/04/2010 16:40 10/05/2010 17:50
NP 10/05/2010 17:50 10/06/2010 18:00
PP 10/06/2010 18:00 10/07/2010 18:05
Moffett Field Building 107 (times PDT)
Pressure condition Begin End
BL 10/31/201016:16 11/01/201016:21
NP 11/01/201016:36 11/02/201017:13
PP 11/02/201017:13 11/03/201016:00
As shown in the building floor plans in Appendix A, air sampling was conducted at various
locations interspersed throughout each of the two buildings. Before testing could begin, SS
sampling points needed to be installed. SS sampling points were already installed and available
on the lower level of the ASU House; at Moffett, SS sampling points were installed following
specifications provided in the QAPP. SS sampling points were spatially separated throughout
the building and located in unobtrusive areas. Installation at Moffett occurred on October 28,
2010, prior to commencement of the ESTCP portion of the field work.
At one of the SS sampling points, shown as "Foundation Pressure" in the floor plans, the cross-
foundation SS AP was measured over the entire test interval (approximately 72 hours). At the
location given as "Building Pressure Measurement" in the floor plans, the I/O AP was
determined over approximately the same time interval. Each differential pressure measurement
was performed using a separate calibrated Omniguard 4® (Engineering Solutions Inc., Tukwila,
WA) real-time differential pressure instrument. For the SS measurement, the reference port was
open to the building atmosphere and the other port was connected with 1A inch semi-rigid walled
tubing to the SS sampling port. For the I/O measurement, the reference pressure port on the
Omniguard 4® was open to the indoor atmosphere and the other port was connected to 1A inch
semi-rigid-walled tubing placed outside of the building envelope through a slightly opened
window. The open end of the tubing extended approximately 2 inches from the building.
Following installation through the open window, the window opening was sealed with tape. The
m Twelve hours is the minimum time for equilibration following a change in building pressure: at a minimum air
exchange rate of 0.25 h"1, 3 air changes would occur over 12 hours, after which indoor air concentrations would be
(1 - e~3)*100% = 95% of their expected final equilibrium concentrations. Moreover, given that integrated and other
air sampling must occur over the next twelve hours following establishment of the new indoor equilibrium
concentrations, twenty-four hours may be interpreted as the minimum required time for testing at each pressure
condition.
10
-------
same connections to the instruments were maintained throughout testing at both buildings so that
consistency of the observed sign of AP was maintained.™ Figure 5 shows the SS and I/O
pressure differential measurements at ASU House. Before and after each pressure condition, the
zero reading of each pressure transducer was verified. The minimum and maximum measured
pressure differential was recorded to an internal instrument datalogger every five minutes for the
duration of testing. Note that at the ASU House, I/O pressure differentials were not measured
under BL conditions. This issue was documented in QAPP Deviation 1; the lack of BL AP at
ASU House had only a minimal impact on test outcomes, given that these measurements were
not included specifically in any of the verification parameters.
Figure 5. Pressure Differential Measurements at ASU House. Panel A, SS AP Monitoring,
Showing Connection to SS Sampling Point; Panel B, I/O monitor; Panel C, Tube Extended
Outside Window for I/O AP Monitoring.
In order to measure building ventilation rates (air exchange rates, AERs), SFe tracer gas was
released at each building (at the locations shown as "SFe release" in the building floor plans in
Appendix A) over the entire duration of pressure testing at each building. Pure (> 99.8%) SFe
was delivered continuously; delivery was initiated at the start of the ESTCP testing that preceded
the ETV tests at each building, and delivery of the tracer continued uninterrupted until the
conclusion of the final PP test on October 7 and November 3, 2010, at ASU House and Moffett
Field Building 107, respectively. Tracer gas delivery was controlled using a rotameter, and
based on previous work and guidance in the QAPP, the target release rate of pure SF6 was
approximately 0.5 mL min"1. Figure 6 shows the tracer gas delivery system as deployed during
testing at Moffett Field Building 107.
Maintaining a steady tracer gas release rate is critical in order to obtain accurate estimates of the
building ventilation rates. Thus, the SFe release rate, as indicated by the rotameter, was checked
approximately every 16 hours and adjusted if found to have drifted by more than 10%.
™ With the reference port open to the interior of the building, AP was positive under NP conditions, and negative
under PP conditions.
11
-------
Furthermore, the SFe flow rate was independently verified, before and after each of the three
pressure conditions at each building, using a DryCal® DC-2 (Bios International Corporation,
Butler, NJ). Whereas the rotameter indicated that the delivery of SFe flow rate remained
relatively constant over the duration of testing at both buildings (as evidenced by fairly invariant
rotameter readings), the rotameter-determined flow rate differed from the DryCal®-determined
flow rate by more than the 10% acceptance criteria established in the QAPP. As documented in
QAPP Deviation 2, and described in more detail in Section 4.2, the flow rate of tracer gas as
measured by the DryCal was substituted for that indicated by the rotameter. Such a deviation
from the QAPP positively impacted the test given that more accurate building ventilation rates
were obtained by using the DryCal-determined SF6 flow rates. This was important, given that
the accuracy of the tracer gas flow rate measurement is one of the Data Quality Indicators (DQIs)
discussed in Section 4.2.
Figure 6. SF6 Tracer Gas Delivery System as Deployed at Moffett Field Building 107.
Under BL pressure conditions, each building's pressure was not intentionally manipulated.
However, in the late afternoon on Days 2 and 3 of testing at each building, the building pressure
was decreased (to induce a NP) or increased (to induce a PP). This was accomplished using a
Lasco® Model 3733 20" box fan installed in a window, either pushing air out of, or into, the
building, respectively. At ASU House, the fan was set to speed 2 (medium); at Moffett Field
Building 107, the fan was operated at its highest speed, speed 3.™ Fan locations at each building
are indicated on the floor plans in Appendix A by "Fan for pressure control." Figure 7 shows the
fans as installed at the ASU House. NP and PP pressure conditions were maintained for at least
12 hours before collection of AA, IA, or SS gas began the next morning, so as to best ensure that
the building atmosphere became well-mixed and to allow concentrations of the various gas-phase
™ At Moffett Field Building 107 the BL I/O AP was slightly negative due to the action of the building's HVAC
system. The fan was operated at its highest setting to better ensure that it could overcome the inherent negative I/O
AP and induce a positive pressure in the building.
12
-------
species to come to equilibrium. At the ASU House, the attainment of new equilibrium
concentrations of TCE and 1,1-DCE was investigated by inspection of measurements performed
by the on-site portable near-real time HAPSITE* GC/MS. Data generated from the HAPSITE*
GC/MS were only for diagnostic purposes and were not included in any verification parameters.
These data are discussed in subsequent Chapters.
Figure 7. Fan Installed for Building Pressure Control at the ASU House.
Beginning on the morning of the second, third and fourth days of testing; corresponding to
October 5, 6, and 7 at ASU House and November 1, 2, and 3 at Moffett Field Building 107; and
corresponding to BL, NP, and PP conditions, respectively; IA, AA and SS gas was collected to
measure various CoCs (VOCs), SF6, and radon. Given in Table 3 is the number of discrete
samples collected for each matrix. Also given are the locations where each of the three discrete
samples was collected; these locations correspond to those shown in the building floor plans in
Appendix A. Specific indoor sampling points were selected as a compromise between attaining
spatial representativeness while minimizing disturbance to building occupants and activities.
Ambient sampling locations were selected nominally upwind of the test building, away from
obvious VOC sources. Sampling procedures and types of samples collected are described in
additional detail below. Times when each type of sample was collected, and total numbers of
samples collected, are given in Table 4.
13
-------
Table 3. Types of and Locations for Air Samples Collected During Each of the Three
Pressure Perturbation Periods. VOCs, SFe, and Radon Were Measured in Each Matrix
and Location.
Matrix
Indoor air
Ambient air
Subslab soil gas
Number of
Locations
3
1
3
Location
Open area on lowest building level plus two additional
samples based on building layout; IA-1, IA-2, IA-3
Upwind location; AA-1
Three locations distributed across the
foundation; SS-1, SS-2, SS-3
building
In order to characterize the concentrations of VOCs, SFe, and radon in IA and AA, two different
types of air samples were collected, one each at IA-1, IA-2, IA-3, and AA-1. One 8-hour time-
integrated air sample for analysis of trace level VOCs/SFe (for IA and AA) was collected into an
evacuated 6-L stainless steel canister at each of the four sampling locations. IA and AA air
canister sampling is shown in Figure 8. Sampling commenced in the early morning and ended in
the early afternoon on each day. Three to four times throughout the day during sampling,
canister pressures were checked to ensure that each was filling at an approximately constant rate.
In one instance, under PP conditions at ASU House, the AA canister was found to be filling too
quickly. The rate of vacuum decrease indicated a leak in the canister valve or flow control
device. Thus another AA sample was collected; analysis results from this recollected sample
have been used for data interpretation and analysis.
At each IA and AA sampling location, a grab sample for radon analysis was collected into a 500-
mL PVF bag using a 50-mL polyethylene syringe connected to a polymer three-way valve. Each
PVF bag was filled with approximately 300 mL (6 syringes full) of air in approximately one
minute. A new syringe was used on each sampling day; AA samples were collected first,
followed by IA samples. Before collection of matrix into the bag, the syringe was purged three
times with the matrix.
14
-------
Table 4. Summary of Sample Types and Timing for Sample Collection at Each Test Building.
ASU House (times MDT)
Sample name
IA, AA (VOCs/SF6)
SS (VOCs/SF6)
IA, AA (Rn)
SS (Rn)d
Moffett Field Building
Sample name
IA, AA (VOCs/SF6)
SS (VOCs/SF6)
IA, AA (Rn)
SS (Rn)
Sample
type
Integrated
Grabb
Grab
Grab
Pressure
Condition,
Date
BL,
10/5/2010
Time
Start
06:50
14:54
13:52
15:00
Time
Stop
14:47
15:18
14:08
16:59
Pressure
Condition,
Date
NP,
10/6/2010
Time
Start
06:55
15:15
13:15
15:20
Time
Stop
15:06
15:40
13:23
17:10
Pressure
Condition,
Date
PP,
10/7/2010
Time
Start
06:46
15:29
14:12
15:33
Time
Stop
16:18
15:50
14:19
17:41
Total
samples3
) 24C
J
) 24ef
J
107 (times PDT)
Sample
type
Integrated
Grab
Grab
Grab
Pressure
Condition,
Date
BL,
11/1/2010
Time
Start
07:00
15:06
12:45
15:10
Time
Stop
15:00
15:23
12:50
16:33
Pressure
Condition,
Date
NP,
11/2/2010
Time
Start
06:58
15:01
13:15
15:05
Time
Stop
15:00
15:27
13:23
17:02
Pressure
Condition,
Date
PP,
11/3/2010
Time
Start
06:25
14:30
13:59
14:31
Time
Stop
14:25
14:44
14:03
15:52
Total
samples3
23g
] 24ef
aTotal number of samples includes those collected for quality control purposes
bCollection of individual grab samples was generally complete in under one minute; start/stop times indicate when each type of grab sampling was
initiated and completed, respectively
°Total canisters collected include: 9 IA, 3 AA, 9 SS, 1IA/AA duplicate, 1 SS duplicate, and 1 additional AA under PP conditions
dSS Rn measured both by sampling into PVF bags and onsite analysis using the RAD7
"Total number of PVF bags collected include: 9 IA, 3 AA, 9 SS, 1 field blank, 1 IA/AA duplicate, 1 SS duplicate
fTotal number of Rn samples does not include onsite sampling and analysis by RAD7
8Total canisters collected include: 9 IA, 3 AA, 9 SS, 1 IA/AA duplicate, and 1 SS duplicate
-------
The measurement of the concentrations of VOCs, SFe, and radon in SS gas required collection of
several different types of samples. For the determination of VOC and SF6 concentrations, one
grab sample each was collected at SS-1, SS-2, and SS-3 into individual evacuated 1-L stainless
steel canisters. Each canister grab sample at each location was filled in less than one minute.
Following the procedures used for grab sampling in IA and AA, grab samples were also
collected in PVF bags at each SS location for radon measurement. Prior to initiating SS
sampling at a given sample point, approximately 50 mL of gas was withdrawn from the sample
point using a polyethylene syringe. This SS purge gas was injected into a separate PVF bag for
discharge outdoors at a later time so as to avoid artificially elevating IA radon concentrations.
Also, before collection of each canister grab sample, the integrity of the plumbing connecting the
canister and three-way valve to the SS sampling line was confirmed by verifying, by inspection
of the canister pressure gauge, that a vacuum could be pulled using the polyethylene syringe.
SS radon was also determined using a near real-time instrument, the Durridge RAD7® radon
detector (Bedford, MA). A total of five RAD7® readings were performed at each sampling
point, each lasting 5 minutes. The average of the final three readings was taken as the radon
concentration at that sampling point. Whereas SS radon was measured using both sampling into
PVF bags and near-real time monitoring with the RAD7®, per the QAPP, the RAD7® will be
used for data interpretation in this verification test. SS sampling using the RAD7® is shown in
Figure 8.
Figure 8. IA, AA, and SS sampling at ASU House (Panels A, B, and C, respectively).
Before canister sampling commenced, canisters were checked for leaks by inspection of initial
canister vacuum. Following collection, final canister pressures were also recorded so that
canister integrity could be tracked until analysis. No canisters were rejected based on out-of-
bounds initial pressure; however, three canisters leaked between shipment and analysis (this is
subject of Deviation 3, that is discussed in more detail in Chapter 4). At each test building, two
duplicate canisters, one for IA/AA and one for SS gas, were collected. Details of the various QC
measures to ensure the validity of the canister sampling are given in Chapter 4.
All PVF bags were leak-checked prior to sampling by pulling a vacuum on the empty bag using
the polyethylene syringe. Each bag was checked again following sample collection, this time by
gently squeezing each bag to verify absence of leaks. Two PVF bags failed initial leak checks
and were discarded. At each test building, one field blank was generated by filling a bag with
16
-------
aged AA, and two duplicate PVF bags, one for IA/AA and one for SS gas, were collected. More
details of the various QC measures employed to ensure validity of the PVF bag sampling are
presented in Chapter 4.
At the completion of each day of testing, the canister samples were shipped by common carrier
to Columbia Analytical Services (Simi Valley, California) for analysis of VOCs and SF6, and the
PVF bags were similarly shipped to the University of Southern California, Department of Earth
Sciences, for radon analysis. Analyses were performed as specified in the QAPP. Briefly,
analysis of canister samples for VOCs was performed using cryogenic preconcentration GC/MS
according to the procedures outlined in EPA Compendium Method TO-1515, with TO-15 scan
for VOCs in SS gas and TO-15 with selected ion monitoring (SIM) for VOCs in IA/AA.
Canister samples for SFe were analyzed using GC/electron capture detection (ECD) according to
procedures in National Institute of Occupational Safety and Health (NIOSH) Method 6602.16
Radon concentrations were measured by way of alpha scintillation counting following
established EPA protocols.17 Additional details of the radon analysis method are given in
McHugh et al.8
The various verification parameters are described in the next several sections.
3.3.1 Decision-making Support
The goal of implementing the building pressure control method is to obtain a better
understanding of VI in a building. If the control of building pressure results in clear changes in
building conditions, such as I/O differential pressures and concentrations of radon and CoCs,
then the pressure control method may yield results that are useful for decision-making (i.e., is VI
a concern for this building?). The effectiveness of the building pressure control method to
support decision-making was evaluated through three metrics:
1. Building Pressure Differential: Did the pressure control method control building
pressure?
2. Vapor Intrusion Enhancement and Reduction: Did the pressure control method increase
the mass discharge of radon from subsurface sources through the building foundation
under induced NP conditions and/or decrease the mass discharge of radon from
subsurface sources through the building foundation under induced PP conditions?
3. Fractional Contribution of Vapor Intrusion to Indoor CoC Concentrations: Did the
pressure control method provide an improved understanding of the contribution of VI to
the concentration of individual CoCs detected in IA?
Each of these three quantitative verification metrics comprising decision-making support is
described in more detail in the following sections. In addition, qualitative metrics related to
providing support to decision-makers, metrics based on the inspection of trends in concentrations
of compounds in IA as well as mass discharges with respect to changing building pressure, are
described along with the presentation of the test results in Section 6.1.
17
-------
3.3.1.1 Building Pressure Differential
The first metric for the verification of the performance of the building pressure control
methodology is whether the building pressure could be decreased and subsequently elevated at
each of the two buildings under induced NP and PP conditions, respectively. The Omni guard 4®
pressure differential instrument measured the minimum and maximum I/O AP every five
minutes. The average AP for each five minute time interval was calculated as the arithmetic
mean of the minimum and maximum AP. Observed mean APs of less than -1 Pa (under NP
conditions) and greater than 1 Pa (under PP conditions) would verify that some degree of
building pressure control had been attained. More details of the data manipulation and statistical
tests applied to these data are described in Chapter 5.
3.3.1.2 Vapor Intrusion Enhancement and Reduction
The second verification metric under decision-making support is the effect of the pressure
control method on enhancement and reduction of radon VI. This metric was evaluated by
comparison of the mass discharge of radon from subsurface sources through the building
foundation under different building pressure conditions.18'1" For instance, under induced NP, the
mass discharge of chemicals with subsurface sources, including radon and CoCs, through the
building foundation and into IA may be enhanced. Similarly, under induced PP conditions, mass
discharge of radon and CoCs from subsurface sources into IA may either be reduced or
eliminated, where the latter condition indicates that VI has effectively been 'turned off by the
induction of PP.
Direct measurement of SS-to-IA flow rates is quite difficult; consequently, it is difficult to
directly measure the mass discharge of chemicals from subsurface sources. Nonetheless, the
mass discharge of radon (and by extension, CoCs) from subsurface sources may be estimated as
follows. The total mass discharge from the building of radon from all possible sources -indoor,
ambient, and subsurface - is calculated as in Equation 1 from the radon concentration in IA (R;,
pCi m"3) and the building ventilation rate, i.e., the flow of IA out of the building (Qj, m3 h"1).
Q! ' R! (1)
Similarly, the approximate mass discharge of radon from ambient sources into IA is found using
Equation 2.
Qa ' Ra « Q; ' Ra (2)
Where Qa is the flow rate of AA into the building and Ra is the ambient radon concentration." If
the mass discharge of radon into IA from indoor, but not ambient, sources is negligible,"1 then by
K Mass discharge is defined in Reference 18 as the strength of a source at a given time and location; it is actually a
rate and is defined in units of mass/time. In this report it is convenient to use mass discharge more genetically for
both radon and CoCs given the utility of the comparisons in the observed trends of radon and CoC mass discharges,
even though a mass discharge of radon has units of activity/time rather than mass/time. Furthermore, mass
discharge may refer to the generation rate of radon/CoCs from indoor source(s), the entry rate of radon/CoCs from
subsurface and/or ambient sources into the building, or the total discharge of radon/CoCs from the building.
x As described in the Mosley Model, this calculation assumes that Qa ~ CL i.e., that the building ventilation rate is
18
-------
mass balance the mass discharge from subsurface sources through the building foundation into
IA may be estimated by subtraction of Equation 2 from 1:
Qi • (R! - Ra) (3)
Note that the notation for Equations 1, 2, and 3 is appropriate for BL conditions.
For each building, three IA radon concentrations were measured at each pressure condition; the
mean of these three measurements was the mass discharge calculations; under BL conditions,
this concentration is given as R;. A single AA radon measurement was also taken at each
building under each pressure condition; under BL conditions this is denoted by Ra. Also, at each
building under each pressure condition, the building air flow rate between indoors and ambient
(the building ventilation rate) was determined at all three pressure conditions as (using the
notation appropriate for BL conditions) Q; = CT • QT / T;, where Q; is the building ventilation
rate, CT is the concentration of SFe (ug m"3), QT is the flow rate of the tracer gas (m3 h"1), and T;
is the mean of three spatially distributed measurements of the indoor concentration (ug m"3) of
SFe. CT was known (99.8% and converted to ug m"3 assuming T = 25°C, P = 1 atm; all
concentrations were converted to this same T/P scale) and for each pressure control test at each
building, the SFe flow rate was determined as the mean of two (one pre- and one post-test)
DryCal measurements.
The degree that VI was enhanced under induced NP conditions was determined by comparison
of Qi" • (R;" - Ra") to Q; • (R; - Ra). If Qi" • (R;" - Ra") > Qi • (Ri - Ra), that is, if the mass discharge
of radon from subsurface sources increased under induced NP compared to BL, then under
induced NP some degree of enhancement of VI has been verified. Similarly, the degree that VI
was reduced under induced PP conditions was determined by comparison of Q;+ • (R;+ - Ra+) to
Qi • (R; - Ra). If Q;+ • (R;+ - Ra+) < Qi • (Ri - Ra), that is, if the mass discharge of radon from
subsurface sources decreased under induced PP compared to BL, then under induced PP
conditions some degree of reduction of VI has been verified.
Under PP conditions, VI may also be reduced to the point that it has been 'turned off If under
PP conditions the radon concentration in IA (R;+) equals the radon concentration in AA (Ra+),
then there is some degree of confidence that VI has been stopped or 'turned off by the induction
of PP. As will be shown in Section 6.2, the mass discharge of radon from subsurface sources
into IA decreased at both buildings under PP compared to BL, thus R;+ = Ra+ was also checked
for both buildings. For each building, R;+ was calculated as the mean of the three IA radon
measurements and Ra+ was the single AA radon measurement.
much greater than the flow rate of soil gas through the foundation into the building, an assumption that is generally
valid.
H Under induced PP conditions, the mass discharge of radon from ambient sources into IA was greater than 75% of
the total mass discharge, indicating that AA was a non-negligible source of radon. As a result, VI enhancement and
reduction were evaluated as the mass discharge from subsurface sources through the building foundation using
Equation 3. This interpretation results from the assumption that radon emission from indoor sources is negligible
compared to subsurface sources (an assumption supported by the radon literature).19
19
-------
For all calculations, MDLs were substituted where concentrations were reported as either zero or
not detectable. Such substitutions were only performed for SF6 and CoC concentrations since in
all cases a measureable concentration of radon was determined in all IA and AA samples.
Further details of MDL substitutions may be found in Section 4. Additional details of the data
manipulation and statistical tests applied to these data are presented in Chapter 5.
3.3.1.3 Fractional Contribution of Vapor Intrusion to Indoor CoC Concentrations
The third verification metric that comprises decision-making support is the ability of the pressure
control technique to determine the fractional contribution of VI (Fyi) to the IA concentration of
several different CoCs. At each test building, two CoCs were selected that were expected to
have subsurface sources, and two CoCs were selected that were not expected to be present in the
subsurface but instead were expected to have only indoor, or potentially predominantly ambient
sources. At the ASU House, the four CoCs were TCE and 1,1-DCE (expected in the subsurface)
and benzene and toluene (not expected in the subsurface). At Moffett Field Building 107, the
four CoCs were TCE and PCE (expected in the subsurface) and benzene and toluene (not
expected in the subsurface).
The FVI calculation combines building ventilation rates and compound concentrations from either
(i) BL and NP (Fvf), or (ii) BL and PP (FVi+). In both cases, the calculation yields an estimate of
the fractional contribution of VI to a CoC's concentration under BL conditions. Both FVI" and
FVI+ and the error in each FVI (denoted as AFyi) were calculated for the four CoCs for each
building. Thus, a total of 16 different FVI ± AFyi combinations were calculated (2 buildings • 2
pressure conditions • 4 CoCs).
At each of the two buildings, Fvf for each of the four CoCs was calculated according to the
Mosley Model using BL and NP results by combining Equations 4 and 5. More detail of the
Mosley Model is provided in the QAPP.
E =
[Q- (R- - Ra}}-[Q,(R, - Ra}}
Qi and Qf were calculated as described in Section 3.3.1.2. For each building, mean IA
concentrations under BL and NP (R; and R;~), AA radon concentrations under BL and NP (Ra and
Ra~), and the corresponding building ventilation rates under BL and NP (Q; and Q;~) were
calculated as described in Section 3.3.1.2. C; and Q" were calculated for each of the four CoCs
at each building as the arithmetic mean of the three IA concentration measurements under BL
and NP conditions, respectively. Ca and Ca~ were the concentrations of each of the CoCs in the
single AA sample collected under BL and NP conditions, respectively.
20
-------
FVI+ may be determined by way of two different methods. One assumes that VI has been
reduced (but not stopped completely), and employs a calculation of complexity similar to that
under NP. The other is a more simplified calculation, where VI is assumed to have been halted
under PP. As is shown in Chapter 6, R;+ = Ra+ at both buildings, indicating that the more
simplified FVI+ calculation was the most appropriate for this verification test. Thus, at each of
the two buildings, the FVi+ values for each of the four CoCs were found according to the Mosley
Model using BL and PP results by combining Equations 5 and 6.
£c=a(c,-cj-a+(c+-c;) (6)
Qi+ was determined similarly to Q; and Qf. C;+ was calculated for each of the four CoCs at each
building as the arithmetic mean of the three IA concentration measurements under PP conditions.
Ca+ was the concentration of each of the CoCs in the single AA sample collected under PP
conditions.
Not only may CoCs be present in indoor air due to contributions as a result of VI (i.e., from
subsurface sources), they may also be present due to emissions from ambient and indoor sources.
Thus, in addition to calculating FVi, the fraction of each CoC's concentration in IA that was due
to ambient and indoor sources, Fa and F;n, respectively, were calculated for both NP and PP for
the four CoCs at each building. According to the Mosley Model, Fa is calculated using Equation
7.
Note that the expression simplifies to the ratio of the CoC's concentration in AA to its
concentration in IA, both under BL conditions. Thus only a single estimate of Fa is determined
using the pressure control technique.
Similar to FVI, F;n may be estimated two different ways. F;n" is found by combining Equations 4
and 8; F;n+ is calculated using Equations 6 and 8.
Q,(c,-ca)-Ec
F _
-
Note that for each of the two independent sets of fractional contribution calculations, Fa + F;n +
Fvi=l.
The error in each FVi, AFVi, was estimated using a Monte Carlo technique described in Chapter
5. Notionally, the FVIS for each CoC under NP and PP should be independent estimates of the
same quantity, and 0 < FVI < 1- Inspection of the FVIS and associated error intervals allows the
degree of confidence that the pressure control technique can ascribe a CoC's indoor
concentration to VI to be determined.
As with the calculations in Section 3.3.1.2, where concentrations were reported as either zero or
not detectable, estimated detection limits were substituted. Such substitutions were only
21
-------
performed for SFe and CoC concentrations since in all cases a measurable concentration of radon
was determined in all samples. Further details of MDL substitutions may be found in Section 4.
Additional details of the data manipulation and statistical tests applied to these data are presented
in Chapter 5.
3.3.2 Comparability
The verification metric of comparability is intended to evaluate the consistency of the pressure
control that was attained in different buildings.
The arithmetic mean of each time series of I/O pressure differentials was calculated according to
Section 3.3.1.1 to determine a total of four mean overall pressure differentials at the two different
buildings: (1) APf and AP2~, the mean differential pressure under induced NP at the ASU House
and Moffett Field Building 107, respectively; and (2) APi+ and AP2+, the mean differential
pressure under induced PP conditions at ASU House and Moffett Field Building 107,
respectively. The comparability of the building pressure control methodology was assessed by
comparison of the two NP differential pressures and the two PP differential pressures according
to the statistical calculations described in Chapter 5.
3.3.3 Operational Factors
Metrics related to operational factors are intended to evaluate primarily the cost and level of
effort associated with implementation of the pressure control method. Operational factors for
implementation of the entire building pressure control technology were evaluated based on
Battelle's observations and input from the technology vendor. General operational factors
include the knowledge, expertise, training, and costs required to carry out all aspects of the field
sampling campaign, including installation of the SS sampling points, measurement of pressure
differentials, and collection of all of the various air samples. The vendor provided cost
information, including information on purchase prices for the Omniguard 4® and RAD7® real-
time monitors, charges for off-site analysis of VOCs and SFe and radon, and costs for the
vendor's time in the field to plan and carry out the field work. Other factors included the
maintenance needs, calibration requirements and frequencies for the real-time pressure
differential and radon instruments, data output and analysis, and sustainability factors, such as
consumables required and used (if any), ease of use, and repair requirements (if any) of the real-
time monitors. Examples of other pertinent information include the number of canisters received
from the analytical laboratory, and number of PVF bags that were deemed unacceptable for
sample collection; the effort and/or cost associated with maintenance or repair of real-time
instruments; vendor effort (e.g., time on site) for repair or maintenance; the duration and causes
of any downtime for real-time instruments; Battelle's observations about ease of use, clarity of
the vendor's instruction manual; and overall convenience of the technologies and
accessories/consumables. During testing at the ASU House Battelle testing staff documented
observations in a laboratory record book (LRB). These observations were summarized to aid in
describing the technology performance.
22
-------
3. 3. 4 Validation ofMosley Model Assumptions
The Mosley Model was used for quantitative evaluation of the third decision-making metric (i.e.,
the fractional contribution of VI to CoC concentrations in IA, FVI). A number of different
assumptions are stated in the Mosley Model, several were explicitly tested using the data
collected in this verification of the pressure control methodology. Verifying the validity of the
assumptions helped to explain the outcomes of the FVi calculations.™ As described below, three
different groups of assumptions, with eight assumptions in total, were explicitly tested at each
building.
Group 1: Building pressure control has no significant effect on CoC and radon concentrations in
SS soil gas below the building foundation.
2. Rs = Rs
Cs and Cs~ for each CoC and Rs and Rs~ were calculated as the mean of the three SS concentration
measurements under BL and induced NP conditions, respectively.™
Group 2: Radon concentrations inAA are much lower than those in SS soil gas below the
building foundation.
3. Ra«Rs
4. Ra"«Rs"
5. Ra+«Rs+
The values of Ra, Ra~, and Ra+ were based on single grab samples of AA. Rs+ was found as the
mean of the three SS concentration measurements under induced PP conditions.
Group 3: In IA, the loss of radon through building ventilation is much greater than the loss due
to radioactive decay.
6. Q;»XV
7. Qi'
8. Q;+
Building volumes were estimated based on interior dimensions and are given in Section 3.1. The
values of Qi, Qf, and Q;+ were calculated as in section 3.3.1.2.
*" It should be noted, however, that some of the model assumptions cannot be verified using the data collected
during this verification test. For example, the Mosley Model assumes that the change in the magnitude of mass
transport though the building foundation under induced NP will be the same for radon and the CoCs.
™ Note that it is unnecessary to validate assumptions 1 and 2 above under PP conditions when it is determined that
VI has been 'turned off, i.e. when R;+ = Ra+, since the calculation of FVi no longer depends on the simplifying
assumption that Cs = Cs+ and RS = Rs+.
23
-------
Estimated detection limits were substituted where concentrations were reported as either zero or
not detectable. Errors in the various parameters were estimated as described in Chapter 5 along
with additional details of the statistical comparisons.
24
-------
Chapter 4
Quality Assurance/Quality Control
QA/quality control (QC) procedures were performed in accordance with the QMP11 for the AMS
Center and the QAPP for this verification test. There were a total of five deviations from the
QAPP. A deviation is an action or QC outcome that differs from QAPP procedures and
specifications. As detailed in the discussion of each deviation, there was little to no negative
impact on this verification test from any of the five deviations. Two deviations (1 and 2) were
described in Chapter 3 and related directly to the field testing, and the remaining three (3, 4, and
6); are described in this Chapter, along with the rationale for vacating Deviation 5. Also covered
here are the general QA/QC procedures employed for this verification test.
4.1 Quality Control Results
A variety of QC measures were implemented to ensure that data of the highest quality were
generated during this verification test. QC procedures were carried both in the field and at the
analytical laboratory, ranging from basic checks of instrument functionality to analytical
instrument calibrations; a number of different field QC samples were also generated for
subsequent laboratory analysis, including field blanks and duplicate samples; and various lab QC
samples were analyzed, such as replicates and method blanks. The specific QC procedures and
samples generated during the performance of this verification test, as well as applicable
acceptance criteria, are described in detail in the QAPP. The results of the implementation of
this verification test's QA/QC program are summarized in Table 5. In general nearly all of the
various QC measures were found to be within acceptable limits. Those that were not found to be
acceptable were the subject of findings or observations in one of two Technical Systems Audits
(TSAs) or Audits of Data Quality (ADQs). Findings were written up as QAPP deviations. The
impact of QAPP deviations 3, 4, and 6, as well as the observations from the first Audit of Data
Quality (ADQ), as referenced in Table 5, are described in more detail in Section 4.3.
25
-------
Table 5. Summary of Results of Various QC Procedures and Samples.
QC Sample Type
Instrument Calibration (VOC) IA/AA
Instrument Calibration (VOC) SS
Instrument Calibration (SF6)
Continuing Calibration Verification (VOC) IA/AA
Continuing Calibration Verification (VOC) SS
Continuing Calibration Verification (SF6)
Method Blank (VOC) IA/AA
Method Blank (VOC) SS
Method Blank (SF6)
Method Blank (radon)
Laboratory Replicate (VOC) IA/AA
Laboratory Replicate (VOC) SS
Laboratory Replicate (SF6)
Laboratory Replicate (radon)
Laboratory Replicate (IA/AA radon)
Laboratory Replicate (SS radon)
Calibration of radon counting cells
Differential Pressure Zero Check
Canister Pressure Pre-use Check
Canister Pressure Receipt Check
Canister Cleanliness Certification
Canister (Analytical) Hold Time (VOC)
Canister (Analytical) Hold Time (SF6)
PVF Bag (Analytical) Hold Time (radon)
PVF Bag (Analytical) Hold Time (IA/AA radon)
PVF Bag (Analytical) Hold Time (SS radon)
Lab/
field
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
F
F
L
F
L
L
L
L
L
Quantity Quantity
reviewed acceptable
4
5
5
4
5
17
4
5
5
various
4
5
5
10
2
11
various
15
47
47
47
47
47
48
28
20
4
5
5
4
5
17
4
5
5
all
4
5
5
10
1
11
all
15
47
11
47
47
47
48
28
20
Quantity
unacceptable Notes
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
o
36
0
0
0
0
o
0
closing CCVs not performed; QAPP deviation 4, no
impact to test, see text
closing CCVs not performed; QAPP deviation 4, no
impact to test, see text
QAPP deviation 6: no impact to test, see text
ADQ 1 observation 2; potential increased radon variability
at low concentrations, no corrective action necessary, see
text
SS radon in PVF bags not used for ETV test
QAPP deviation 6; no impact to test, see text
QAPP deviation 1 ; minimal impact to test, see text
QAPP deviation 3 ; 44 pass revised
acceptance criterion, minimal impact to test, see text
SS radon in PVF bags not used for ETV test
-------
Table 5. Summary of Results of Various QC Procedures and Samples (Continued)
QC Sample Type
Canister (Field) Duplicates (VOC) IA/AA
Canister (Field) Duplicates (VOC) SS
Canister (Field) Duplicates (SF6)
PVF Bag (Field) Duplicates (radon) IA/AA
PVF Bag (Field) Duplicates (radon) SS
PVF Bag (Field) Blank (radon)
Canister Matrix Spike (VOC) IA/AA
Canister Matrix Spike (VOC) SS
Canister Matrix Spike (SF6)
Radon matrix spike
Lab/
field
F
F
F
F
F
F
L
L
L
L
Quantity
reviewed
2
2
2
2
2
2
4
5
5
various
Quantity
acceptable
1
1
2
1
2
2
4
5
5
all
Quantity
unacceptable
1
1
0
1
o
0
0
o
0
0
Notes
ADQ 1 observation 4; potential increased toluene
variability at low concentrations; no corrective action
required, see text
ADQ 1 observation 4; potential increased PCE variability
at low concentrations; no corrective action required, see
text
ADQ 1 observation 5; potential increased variability at
low concentrations; no corrective action required, see text
SS radon in PVF bags not used for ETV test
ADQ 1 observation 3; minimal data quality impact, see
text
Data Quality Indicator, see text
Data Quality Indicator, see text
QAPP deviation 6; no impact on test, see text
-------
Another aspect of the quality system is the selection, use, and number of observations below
selected method detection limits (MDLs). Estimates of VOC and SF6MDLs used in this report
were those provided by the analytical laboratory. For radon, the selection of an appropriate and
applicable MDL is described below.
The analytical laboratory estimated that the lower limit of detection (LLD) for radon was
approximately 0.14 pCi L"1, that was estimated using procedures promulgated by the EPA.19
This guidance stated that the LLD is an "a priori estimate of the quantity of activity that will be
detected with a given confidence." However, the LLD "is [only] a prediction of measurement
capability;" to evaluate whether a radon measurement is greater than background, another metric,
the minimum significant measured activity (MSMA), defined as "the smallest measurement
interpreted to demonstrate the presence of activity in the sample," should be employed. In
general, both LLD and MSMA are calculated at a 95% confidence level where a 5% false
positive rate is deemed acceptable. MSMA varies on a per sample basis and depends on, for
example, the cell that the radon activity is measured (and specifically the cell volume,
background count rate, and efficiency factor), the count time, and the sample hold time.
Including counting cell-specific information to estimate the MSMA is important given that
certain counting cells have higher background count rates than others, and counting cells are
segregated on this basis. To effectively measure IA/AA radon, only those cells with the lowest
background counts (and lowest MDLs) may be used. MSMA is defined as 2.77-Sb, where Sb =
standard deviation of the background activity. The analytical laboratory provided Sb for each
radon measurement, and the radon MDL was set equal to the MSMA.
Where measured values of SFe and CoCs were reported as zero (below the applicable MDL), the
value of the MDL was substituted for the zero measurement in all calculations (of averages,
standard deviations, etc.). Radon concentrations were never reported as zero; thus, the actual
reported radon concentrations were used in all calculations. Although a reported result that is
below the MDL has a higher uncertainty than a reported result at a higher concentration, the
reported result is a more accurate characterization of the actual radon concentration in the sample
as compared to the estimated MDL. Such treatment of radon data is consistent with EPA
guidance.20' ^
In general, MDL substitutions negatively impacted the statistical calculations, for instance, by
precluding post-hoc power analysis. The impact of these substitutions is described in more detail
in Chapter 6. Specific instances where the reported concentrations were less than applicable
MDLs are shown in yellow highlight in the raw data that are presented in Tables A2 through A7
in Appendix A.
HV Reference 20, Appendix A states: "The result obtained in a measurement, which is a sample of the infinite
population of possible results, is the best estimate of the mean value of the population. These actual results, whether
greater than or less than the LLD, and whether positive, negative, or zero, should be used in averaging. Elimination
of results less than the LLD, or of results less than zero, introduces a bias into the overall average value."
28
-------
Table 6. Summary of Frequency of Measurements Lower Than Estimated MDLs.
Measurement Total #a #
-------
second ADQ of the results presented in this verification report was also conducted. Audit
procedures are described further below.
4.3.1 Technical Systems Audits
To ensure that the verification test was performed in accordance with the AMS Center QMP11
and with the QAPPthe NAVFAC Atlantic Quality Assurance Officer (QAO) for this verification
test performed two different TSAs, one at each of the buildings where the pressure control
technique was implemented. The QAO was onsite at each of the buildings for the entire duration
of each verification test. While onsite the QAO compared actual test procedures to those
specified and referenced in the QAPP, and reviewed all pertinent project documentation and data
acquisition and handling procedures. Moreover, the QAO observed all aspects of performing the
field work, including collecting air samples, operating (and in one instance, troubleshooting) the
real-time differential pressure and radon monitors, pre- and post-sampling canister/PVF bag
integrity checks, and all field QC measures listed in Table 5.
The first TSA at the ASU House resulted in three findings and four observations. The first
finding regarded maintenance of project records. The QAPP describes recordkeeping practices,
and states that all documents and records will be maintained by the VTC during the test and
transferred to secure storage at the conclusion of the test. However, technology vendor staff
were conducting the field work and were required ready access to the project records during
testing. Thus it was decided to allow the vendor to maintain the field data record sheets and logs
throughout the duration of the field work. The VTC, QAO, and technology vendor staff
discussed this procedural change and GSI Environmental agreed to provide the VTC photocopies
of all records - specifically, the project's data collection forms - at the end of each test day. This
solution was similar to the document maintenance and control procedures described in the QAPP
for future testing at Moffett Field (where the VTC knew in advance that he would be absent);
that is, GSI Environmental agreed to send electronic copies of all applicable project records to
the VTC at the end of each test day.
The second finding regarded the observed discrepancy in tracer gas flow rates as measured by
the DryCal® as compared to the rotameter, i.e. that they did not agree within ±10%. This finding
was ultimately addressed by substitution of the DryCal® flow rates for those indicated by the
rotameter, as discussed in QAPP Deviation 2 and in Sections 3.3 and 4.2 of this report.
The third finding was that the I/O pressure differential was not measured under BL conditions at
the ASU House because only one differential pressure monitor was available at the start of
testing. This finding was addressed in QAPP Deviation 1 and is discussed in Section 3.3.
Four observations also resulted from the first TSA. Observations were related to project records
and recordkeeping practices. In response to these observations, the following actions were taken.
• The certificate of analysis for the SF6 tracer gas was obtained for onsite review by the
QAO and added to the project records.
• The calibration record for the DryCal was similarly obtained for onsite review by the
QAO and added to the project records.
30
-------
• Data recording practices of field staff were improved so as to avoid future issues with
prompt recording of field-generated data onto data collection forms.
• Pre-printed sample labels were not used during testing at Moffett Field.
The NAVFAC Atlantic QAO also performed a ISA during testing at Moffett Field. This ISA
resulted in one finding and one observation. The finding further documented the discrepancy in
the tracer gas flow rates as indicated by the DryCal and the rotameter. This finding was
ultimately addressed as described in the summary of the first TSA above. The observation
regarded maintenance of the integrity of project records given that the VTC was absent from the
test site. This observation was addressed in advance of the test by requiring that the onsite test
team forward copies of project records to the VTC on a daily basis.
All of the findings and observations for both TSAs were determined to have either no or only
minimal impact on test outcomes. TSA reports were prepared and copies were distributed to the
EPA.
4.3.2 Audits of Data Quality
A Battelle technical staff member involved in this verification test reviewed all test records
before such were used to calculate, evaluate, or report verification results. The person
performing the review added his/her initials and the date to a hard copy of the record being
reviewed. The VTC reviewed 100% of the verification test data for quality. The data were
traced from the initial acquisition, through reduction and statistical analysis, to final reporting to
ensure the integrity of the reported results. Statistical manipulations were performed using
commercially available software (Stata and R) executing custom-written code; where applicable,
the VTC cross-checked statistical outputs against outputs derived from independent calculations
of results shown in the Appendix A.
In addition, the NAVFAC Atlantic QAO performed an ADQ where at least 10% of the data
acquired during the verification test and 100% of the calibration and QC data were audited and
compared against QAPP specifications. This first ADQ included a comprehensive audit of all
data generated by the laboratories that analyzed the canisters (for CoCs and SFe) and PVF bags
(for radon).
This ADQ resulted in 4 findings and six observations, and each of the four findings ultimately
resulted in a deviation from QAPP specifications. The first finding concerned the change in
canister pressure during the time that elapsed between sample shipment and sample receipt.
More details are provided in QAPP Deviation 3. Briefly, a total of 36 out of the 47 canisters
collected at both field sites failed the QAPP-specified pressure difference criterion of < 1 inch
Hg pressure change between sample shipment and receipt at the laboratory. Canister pressures
did change, but they all decreased (i.e., the measurements indicated greater vacuum upon receipt
at the laboratory). This is a physically impossible spontaneous phenomenon, and bias between
the pressure gauges used for the measurements was suspected. However, it is clear that
pressures in three of the SS canisters from the ASU Research House - BL-SS-VOC-1, NP-SS-
VOC-1, and PP-SS-VOC-3 - increased (i.e., their vacuum decreased) over the time interval
between laboratory receipt and analysis. These canisters fail the alternative acceptance criteria
31
-------
pressure, i.e. that canister pressure cannot increase (i.e., that vacuum cannot decrease) during the
time between completion of sample collection and analysis. For these three canisters, the results
that are affected are shown in red text in Table A4. For the three canisters, the change in
pressure ranged from 2.4 to 5.2 inch Hg; as a result, if the change in pressure was due to canister
leakage, this resulted in a 10% to 20% dilution of the sample. However, in no instance did any of
the final canister pressures reach 0 inch Hg gauge. Thus the impact on test outcomes is expected
to be minimal since results from these three SS canister samples are not included in the
calculation of any quantitative verification metrics, only in the verification of assumptions for the
Mosley Model.
The second finding regarded analysis of continuing calibration verification (CCV) standards
after completion of the analysis of all samples in a given batch. As detailed in Deviation 4, this
QAPP requirement was in error. Final CCV analysis is not required by U.S. EPA Compendium
Method TO-15,15 nor is such required by the laboratory's standard operating procedure. As
such, no impact on test outcomes is expected as a result of this deviation.
The third finding, written up as Deviation 5, was that not every batch of canisters analyzed for
CoCs in IA, AA and SS gas included a replicate analysis. Subsequent to the completion of the
first ADQ, additional replicate data were delivered to the QAO who determined that in fact a
replicate had been analyzed with every batch and that all replicates met the appropriate
acceptance criteria. Thus Deviation 5 no longer applied and was vacated.
The fourth finding covered the analysis of radon in PVF bags specifically that a matrix spike and
method blank were not analyzed with every sample batch. This finding resulted in deviation 6.
This deviation did not impact test outcomes given that the laboratory employed a wide variety of
appropriate and applicable alternative QC techniques, generally in accordance with guidance
provided by the EPA.17 This guidance document specifies that the calibration of the radon
measurement system be verified every 12 months, and states that the measurement system
background be checked, but does not explicitly specify a frequency for such background checks.
For convenience, the details of the actual radon analytical laboratory's QC measures are
summarized below.
Radon analysis for the ETV test samples was performed on 10/6/2010-10/13/2010 for the ASU
House batch and 11/3/2010-11/5/2010 for the Moffett Field samples. Leading up to the analysis
of these samples, three different quality control checks were performed at different times. The
first check was a channel confirmation using a cell containing 241Am. Since December 2009,
941
three Am cell checks have been performed; in December 2009, in May 2010, and on October
14th, 2010. Variances in the counts per minute for all channels were less than 1%. The second
QC check performed was the measurement of NIST-traceable 226Ra standards to calibrate the
efficiency of the cell/channel combinations. The efficiency test was performed September 23rd
and 24* ,2010, immediately before analysis of samples from the ASU House. The results
showed that the channels and cells were still within their calibrated efficiency range by
comparing concentrations to the calculated mean for all channels and cells. The calculated
results varied only ~ ± 5 % from the mean. The third check was a check of cell backgrounds;
cells were checked a variety of different times from June through October 2010. Cells are
segregated on the basis of background, with high background cells used to measure radon in SS
32
-------
samples, and low background cells for IA and AA. Background values are subtracted from
measured values to generate the reported disintegrations per minute and subsequently to radon
concentrations in pCi L"1. Background values are also used in the determination of MDLs.
The ADQ also revealed six observations, one related to data completeness and the remainder
regarding QC exceedances for replicate (1 observation) and duplicate (3 observations) analyses,
and the radon field blank. In response to these observations, the following actions were taken.
• The NAVFAC Atlantic QAO requested and received additional information from the
analytical laboratory, thereby completing the data package in question.
• The impact of excessive variability in the various replicate and duplicate analyses was
assessed. QC results are not specifically included in the quantitative verification metrics
for this test, thus direct impacts on test outcomes is minimal. However, these QC
exceedances demonstrate that, for the samples affected, there exists the chance for high
variability in all of the measurements performed during this verification test. Affected
samples are highlighted in orange text in Tables A2, A3, and A4.
• The concentration of radon in the field blank at the ASU House, 0.26 pCi L"1, while
exceeding the QAPP specification of 0.2 pCi L"1, was found to be less than the
corresponding MDL (0.36 pCi L"1) for the subject analysis.
The NAVFAC QAO also performed a final ADQ that assessed overall data quality, including
accuracy and completeness of this technical report. To ensure the integrity of the reported
results, the QAO traced data from initial acquisition, through reduction and statistical analysis, to
final reporting. The QAO confirmed that all audit findings and observations had been addressed,
verified the integrity of all hand entered and manually calculated results, and confirmed that all
formulae were accurate and consistent. The second ADQ revealed no findings or observations.
All of the findings and observations for the first ADQ were determined to have either no or only
minimal impact on test outcomes. Audit reports covering both ADQs were prepared and
distributed to the EPA.
33
-------
Chapter 5
Statistical Methods
The statistical methods used to evaluate the quantitative performance factors listed in Section 3.3
are presented in this chapter. Qualitative observations were also used to evaluate verification test
data.
5.1 Decision-making Support
5.1.1 Building Pressure Differential
The Omniguard 4® pressure differential instrument measured the minimum and maximum I/O
AP every five minutes. This generated a time series of approximately 288 observations over 24
hours for each pressure condition at each building. Pressure differentials were corrected to
account for the reference ports on each of the AP instruments being open to the indoor
atmosphere so that a positive AP indicates the potential for downward flow of air from the
building through the foundation and a negative AP indicates the potential for upward flow of soil
gas through the foundation into the building. The arithmetic mean of the minimum and
maximum AP for each observation in the time series was calculated, as was the overall mean of
the entire time series of observations, its standard deviation, and the standard deviation of the
mean. One-sided t-tests were performed to determine if the AP" at each building was statistically
significantly less than -1 Pa and if each AP+ was statistically significantly greater than 1 Pa. The
null (Ho) and alternative (Hi) hypotheses were formulated as follows.
Under NP:
H0: AF = -1 Pa
HI: AF<-lPa
Under PP:
H0: AP+ = 1 Pa
Hi: AP+> IPa
In addition, the percentage of the individual observations either less than -1 Pa (under NP) or
greater than 1 Pa (under PP) was calculated.
34
-------
5.1.2 Vapor Intrusion Enhancement And Reduction
The degree that VI was enhanced under induced NP conditions was determined by comparison
of Qi" • (Rf - Ra") to Q; • (R; - Ra). If Qi" • (Rf - Ra") > Qi • (R; - Ra), then under induced NP
conditions some degree of enhancement of VI has been verified. The degree that VI was
reduced under induced PP conditions was similarly determined comparison of Q;+ • (R;+ - Ra+) to
Qi • (R; - Ra). If Q;+ • (R;+ - Ra+) < Qi • (R; - Ra), then under induced PP conditions some degree
of reduction of VI has been verified. Q; • (R; - Ra), Qi" • (Rf - Ra "), and Q;+ • (R;+ - Ra+) for each
building were calculated as described in Section 3.3.1.2. The errors in the R;, Rf, and R; were
taken to be the standard deviation of the three spatially distributed measurements; relative errors
in Ra, Ra", and Ra+ were assumed to be equal to relative errors in corresponding triplicate R;, Rf,
and R; measurements, respectively. The error in Q;, Qf , and Q;+, the quantities (R; - Ra), (Rf -
Ra"), and (R;+ - Ra+), and the quantities Q; • (R; - Ra), Qi" • (Rf - Ra"), and Q;+ • (R;+ - Rf") were
estimated by propagation of errors.™ Results of these propagation of error calculations are given
in Table A2. Two-sample one-sided paired t-tests were conducted to determine if the above
inequalities could be verified statistically. Under NP, the following hypotheses were tested.
H0: Q, • (R: - Ra) = Qf • (Rf - Ra")
Hi:Q1-(R1-Ra)
-------
In instances where statistically significant differences were not detected, the feasibility of
performing a post-hoc calculation was investigated. Such a calculation would estimate the
minimum detectable difference, with 80% power and a 5% false positive rate, given the observed
sample size and variability. However, to perform such retrospective calculations, a number of
prerequisites had to have been met, including that a sufficient number of observations were
present (at least 3) for both samples in the comparison, the p-value was not significant, all
measurements were greater than corresponding MDLs, and in the case of paired t-tests, the
correlation between paired observations was positive. One or more of these prerequisites were
not met for the comparisons in this section; as such, no post-hoc power calculations were
performed.
5.1.3 Fractional Contribution of Vapor Intrusion to indoor CoC concentrations
The 16 FVI were calculated as described in Section 3.3.1.3. The error in each FVI, AFyi, was
estimated using a Monte Carlo technique instead of propagation of errors. The propagation of
errors error estimation technique given in the QAPP ignores more than one covariance and these
correlations cannot be assumed to be conservative. Furthermore, this experiment did not furnish
sufficient data to estimate the correlations.
A number of variables were transformed onto the natural log scale ahead of the Monte Carlo
analysis; these include Q, Cf, C;+, R;, Rf, Ra, and Ra".™ Those variables that were not
transformed included Ca, Ca~, Ca+, Q;, Q;~, and Q;+. On the appropriate scale, errors in Q, C;~, C;+,
R;, and R;~ were taken to be the standard deviation of the three spatially distributed
measurements; relative errors in Ra, and Ra~ were assumed to be equal to relative errors in
corresponding triplicate R; and Rf measurements, respectively. Errors in Ca, Ca~, Ca+ were
assumed to be equal to the accuracy limit for the TO-15 volatiles analysis, ±30%. Errors in Qj,
Qi", and Q;+ were estimated by propagation of error technique as described in Section 5. 1 .2.
Each Monte Carlo simulation generated random draws from the distributions of the quantities in
equations 4 and 5 (for Fyf) and 6 and 5 (for FVI+) (see Section 3.3.1.3). Calculating both Fyf and
Fyi+ required univariate and bivariate normal sampling of random variables. For those random
variables assumed to have correlations equal to zero, joint distributions were calculated as the
product of the corresponding marginal distributions (i.e., a univariate normal distribution that
was characterized by a mean and standard deviation). For those random variables known to have
nonzero correlations, the joint distribution was a bivariate normal characterized by two means,
two standard deviations, and the applicable correlation coefficient pXY (where X and Y are the
two variables in question). The formula for calculating FVI" required sampling from four
univariate normal distributions (Q;Ra, Q;Ca, Qi"Ra", and Qi"Ca") and four bivariate normal
distributions (QiQ, Qi'Cf, Q;R;, and QfRf); the Monte Carlo analysis for FVI+ required sampling
from two univariate normal (Q;Ca and Q;+Ca+) and two bivariate normal distributions (Q;C; and
™ All log-transformed values were exponentiated before final results were reported.
x™ Inspection of applicable physical phenomena revealed that building flow rates and ambient levels of CoC and
radon should be uncorrelated given that, to the first approximation, indoor air concentrations do not contribute
substantially to outdoor levels; thus pQjRa = pQA = pQrRa" = pQrCa" = pQ^C/ = 0. Furthermore, all of the
remaining applicable nonzero covariances (QA, QfQ", Qi+Q+, Q^, and QrRO should be negative under the
36
-------
For each building/CoC combination, eleven sets of correlation coefficients were constructed that
obeyed the appropriate ordering assumptions.™11 Confidence intervals were estimated for each
set of coefficients to determine whether the results would be sensitive to the choice of those
coefficients. For each combination of building, CoC, and correlation coefficients, N = 100,000
samples were generated from each sampling distribution required to calculate FVi+ and Fvf.
Results were combined using the appropriate formulae, and the 2.5 and 97.5 percentiles were
calculated to obtain a 95% confidence interval for the various FVI+ and Fvf estimates.
As explained in Chapter 6, post hoc power calculations were not performed for the FVi values.
5.2 Comparability
Comparability of the observed I/O differential pressures is assessed by calculation of the relative
percent difference (RPD) of the mean differential pressure under NP and PP conditions (PvPD,
AP" and RPD, AP+, respectively) using equations 4 and 5.
Apf-Ap,~
RPDW- = —
•100 (4)
•100
5.3 Verification of Model Assumptions
A total of eight assumptions were tested at each building and are organized into three groups.
The first group includes two assumptions to determine if inducing a NP in the building had a
significant effect on CoC and radon concentrations in SS soil gas below the building foundation.
2. RS = RS-
Cs and Cs~ for each CoC and Rs and Rs~ were calculated as the mean of the three SS concentration
measurements under BL and NP conditions, respectively. Errors in these quantities were
estimated as the standard deviations. All data were first transformed onto the natural log scale
before means and standard deviations were calculated. Two-sample 2-sided paired t-tests were
executed to investigate the following null and alternative hypotheses.
assumption that ambient levels of CoCs and radon are typically less than indoor levels.
x™ For instance, given that Ca < Q, increasing Q; will lead to a decrease in Q; thus pQA < 0. Reasoning along the
same lines, for chemicals expected to have VI sources, i.e., TCE, 1,1-DCE, and PCE, p(Qr, CO < p(Q;, CO < p(Qi+,
Q+) [i.e., p(Qr, Q") is more strongly correlated than p(Qj+, Q4)] and similarly, p(Qr, RO < p(Qi, RI); for chemicals
not expected to have VI sources, benzene and toluene, p(Qr, Q") ~ p(Q1; Q) ~ p(Qj+, Q+).
37
-------
Assumption 1:
TT . /"I /-I -
rlQ. LxS — LxS
Assumption 2:
HO : Rs = Rs
HI : Rs ^ Rs
As described in more detail in Chapter 6, the assumptions above were tested both including and
excluding the results from SS-1 at ASU House, and post hoc power calculations of minimum
detectable differences were performed only for TCE, PCE, benzene and radon at Moffett Field.
The second group of assumptions whose validity were verified related to if radon concentrations
in AA were in fact much lower than those in SS soil gas below the building foundation.
3. Ra«Rs
5. Ra+«Rs+
Values of Rs and Rs~ were calculated as the mean of the three SS concentration measurements
under BL and NP conditions, respectively; Rs+ was calculated similarly as the mean under PP
conditions. The values of Ra, Ra, and Ra+ were taken as the results of a single grab sample of
AA; the estimated relative error in their concentrations will be assumed to be equal to the relative
error in the corresponding triplicate R;, Rf, and R;+ measurements, respectively.^ All data were
transformed to the natural log scale before conducting a 2-sample 1-sided (unpaired) t-test with
unequal variances. Hypotheses were formulated as follows.
Assumption 3:
H0: Ra - Rs = 0
HI: Ra-Rs<0
Assumption 4:
HQ: Ra - Rs = 0
^ The error in Ra~ at ASU House was assumed equal to the standard deviation of Rr; that is, relative errors were not
used. Assuming a relative error unreasonably inflated the error estimate for Ra~ given the high absolute values of Rj"
in comparison to Ra~.
38
-------
Assumption 5:
H0: Ra+ - Rs+ = 0
Hi:Ra+-Rs+<0
As described in more detail in Chapter 6, the assumptions above were tested both including and
excluding the results from SS-1 at ASU House, and post hoc power calculations of minimum
detectable differences were not performed.
The third and final group of assumptions tested included those to determine if, in IA, the loss of
radon through building ventilation is much greater than the loss due to radioactive decay.
6. Q;»XV
7. Qi'»XV
8. Q;+»XV
Each of the two building's volumes was calculated using interior dimensions of each and the
error in the building volume was conservatively estimated to be ± 3Q%™ The decay rate of
radon was found in the literature13 and assumed to be known quite accurately (estimated error of
± 1%). The values of Q;, Q;~, and Q;+ and estimates of their errors were calculated as described
in section 5.1.3. Data were not log transformed. The number of standard deviations that the
mean of XV was from the mean of Q; was calculated and a one-sided p-value was generated
assuming 2 degrees of freedom (Q; was regarded as having a sample size of 3 and XV a sample
size of 1). Minimum detectable differences were calculated if the null hypotheses were not
rejected. The null and alternative hypotheses were formulated as follows.
Assumption 6:
H0: Q, = XV
Hi:Qi>XV
Assumption 7:
H0: Qf = XV
Hi: Qf>XV
Assumption 8:
H0: Qf = XV
HI: Q;+>XV
** This conservative estimate of error in building volume is insignificant compared to estimated errors in building
ventilation rates. See Section 6.5.
39
-------
Chapter 6
Test Results
The results of the verification test of the building pressure control technique are presented in this
Chapter. Presented first in Section 6.1 are the results of the various field measurements,
including differential pressures, calculations of building ventilation rates, presentation of the
radon, SF6, and CoC concentrations, and mass discharges of the various compounds. Where
necessary for clarity, descriptions of the data manipulation methods are discussed. Presented in
Sections 6.2 to 6.4 is the evaluation of the three different quantitative performance metrics:
decision-making support, comparability, and operational factors. The IA concentration data in
Section 6.1.4 and the mass discharges in Section 6.1.6 illustrate qualitative trends that aid in
evaluating the performance of the pressure control technique in terms of decision-making
support.
The data generated during the verification test, both used in the calculations presented in this
Chapter and ancillary to the test, are presented in Appendix A.
6.1 Measurement Results From Both Buildings
6.1.1 Indoor/Outdoor and Cross-Foundation Pressure Differentials
During each verification test, both the I/O and cross foundation differential pressures were
measured. Treatment of the AP data was described in detail in Section 5.1.1. The averages of
the minimum and maximum APs for each five-minute observation are plotted in Figure 9; also
shown is the overall average AP for each of the three pressure conditions.
For the induced NP period, negative APs were measured at both buildings, both I/O and across
the foundation. Similarly, for the induced PP period, positive APs were measured both I/O and
across the foundation at both buildings. Such results indicate that the building pressure control
technique was successful at manipulating building pressure.
The control of I/O AP is one of the quantitative performance metrics for this verification test; it
and the cross-foundation differential pressures are discussed in more detail in Section 6.2.
40
-------
Baseline
Not recorded
Negative
Avg.-5.18
Positive
Avg. 3.87
„
i D 0
I 8
! 2 -2
i Q!
ii -4
: 0)
10/4/1012:00 10/5/1012:00 10/6/1012:00 10/7/1012:00
Baseline
Avg. 0.35
Negative
Avg.-2.12
Positive
Avg. 2.56
Timestamp
3 0 1
8
-------
Negative Positive
PO
_£
9
Baseline Negative Positive
Figure 10. Building Ventilation Rates Measured under Three Different Pressure
Conditions at ASU House (Panel A) and Moffett Field Building 107 (Panel B).
6.1.3 Concentrations of Compounds in Ambient Air
The ambient concentrations for the various compounds at each building are show in Figure 11.
Plots in panels A and B in this Figure, and in Figures in subsequent sections, are divided to show
three compounds that were expected to have predominately subsurface sources (radon, TCE, and
1,1-DCE for ASU House and radon, TCE, and PCE for Moffett Field Building 107) and the three
compounds that were expected to have predominately indoor or ambient sources (SFe, benzene,
and toluene for both buildings).
For instances where compound concentrations were reported as not detectable, the MDL is
shown and the data are flagged with "ND." In instances where reported radon concentrations
were less than the MDL, the reported concentration is shown and the result is flagged with
"
-------
c
o
c
0)
u
c
o
u
SUBSURFACE SOURCES
INDOOR
SOURCE
AMBIENT SOURCES
Baseline
Negative
Positive
Radon TCE 1,1-DCE SF6 Benzene Toluene
(pCi/L) (ug/m3) (ug/m3)x2 (ug/m3)/10 (ug/m3) (ug/m3)/10
o
'4-*
(D
0)
U
O
u
SUBSURFACE SOURCES
INDOOR
SOURCE
AMBIENT SOURCES
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
I Baseline
I Negative
Positive
Radon
(pCi/L)
TCE
(ug/m3)
PCE
(ug/m3)
SF6
(ug/m3)/10
Benzene
(ug/m3)
Toluene
(ug/m3)/10
Figure 11. Concentrations of Compounds Measured in Ambient Air at ASU House (Panel
A) and Moffett Field (Panel B).
6.1.4 Concentrations of Compounds in Indoor Air
Concentrations of the various compounds in IA were measured at three spatially separated
locations in each test building. Figure 12 shows the average compound concentrations across the
three measurement locations and error bars as ± 1 standard deviations. Where individual
concentrations were reported as not detectable, the MDL was substituted into the average and
standard deviation calculations. For these and future plots in this Section, asterisks (*) denote
instances where at least one such MDL substitution was performed. No such substitutions were
performed for radon, but plus signs (+) indicate instances where at least one reported radon value
was less than its corresponding MDL.
43
-------
c
o
c
s
c
o
u
SUBSURFACE SOURCES
INDOOR
SOURCE
AMBIENT SOURCES
Baseline
Negative
Positive
Radon TCE 1,1-DCE SF6 Benzene Toluene
(pCi/L) (ug/m3) (ug/m3)x2 (mg/m3) (ug/m3) x 10 (ug/m3)
c
o
c
0)
u
c
o
u
SUBSURFACE SOURCES
AMBIENT SOURCES
Baseline
Negative
Positive
Radon TCE PCE SF6
(pCi/L) (ug/m3) (ug/m3) (mg/m3)
Benzene Toluene
(ug/m3) (ug/m3) /10
Figure 12. Concentrations of Compounds Measured in Indoor Air at ASU House (Panel A)
and Moffett Field (Panel B).
For the ASU House, TCE was expected to have predominately a subsurface source. However,
after the verification study was completed, it was discovered that a liquid TCE laboratory
standard had been stored in a refrigerator in the garage throughout the duration of testing,
thereby creating the potential for an unexpected TCE source at this location. This may explain
the high IA TCE concentration under BL conditions at ASU House.™ In the absence of such an
indoor source, it was expected that its BL IA concentration would be lower than that under
induced NP, similar to the trends observed for radon and 1,1-DEC, other compounds with
™- The impact of the indoor TCE source may also be seen in the plot of the real-time TCE concentrations as
measured by the HAPSITE GC/MS (Figure A3) in which TCE is quite high under BL despite the fact that the cross-
foundation pressure differential was positive, indicating that VI should be suppressed under BL. In the absence of
such a source, TCE concentrations in IA likely would have been lower than observed.
44
-------
subsurface sources. This is in contrast to Moffett Field Building 107 where compounds with
subsurface sources had higher IA concentrations under BL compared to NP since (a) VI was
effectively turned on under BL conditions, as evidenced by the negative cross-foundation AP and
(b) the combined effect of enhancing VI under NP was countered by the concomitant increase in
ventilation rate. Nonetheless, in all cases for compounds with expected subsurface sources,
concentrations in IA were lower under the induced PP condition compared to the induced NP
condition.
For compounds with expected indoor or ambient sources, there was little change in
concentrations between NP and PP. IA SF6 concentrations did decrease under induced NP and
PP compared to BL - due to dilution from the increased building ventilation rate, behavior
consistent with having a dominant indoor source - but its concentration did not change between
the two different induced pressure conditions given relatively constant building ventilation rates.
Changes in the indoor concentrations of benzene and toluene between BL and NP/PP were
similar to the changes in their AA concentrations, consistent with the expectation that AA was
the major source of these compounds to IA.
Figure 13 shows the average compound concentrations measured in IA normalized by their
concentrations in AA. Note that the graphs of normalized concentrations use a log scale. These
figures show the relationship between the concentrations of the various compounds in IA
compared to AA and demonstrate the differences and similarities in the sources of the
compounds.
45
-------
SUBSURFACE SOURCES
1000
INDOOR
SOURCE
AMBIENT SOURCES
100
Baseline
Negative
Positive
CO
CCL
c
o
c
0)
u
c
O
U
Radon TCE 1,1-DCE SF6 Benzene Toluene
SUBSURFACE SOURCES
1000
INDOOR
SOURCE
AMBIENT SOURCES
100
Baseline
Negative
Positive
CO
CCL
c
o
c
0)
U
c
O
U
Radon TCE
PCE
SF6 Benzene Toluene
Figure 13. Average Indoor Air Concentrations Normalized by Ambient Concentrations at
ASU House (Panel A) and Moffett Field Building 107 (Panel B).
For both buildings, compounds with expected subsurface sources (i.e., radon, TCE, 1,1-DCE,
and PCE) had IA concentrations greater then AA (i.e., normalized concentrations > 1) under the
induced NP condition but had IA concentrations similar to AA (i.e., normalized concentration
close to 1) under the induced PP condition. For the compound with expected indoor source (i.e.,
SF6), the IA concentration was greater than ambient for all pressure conditions. For the
compounds with expected ambient sources (i.e., benzene and toluene), IA concentrations were
similar to ambient for all pressure conditions. These plots demonstrate the ability of the building
pressure technique to discern sources of various CoCs and judge the potential that certain CoCs
may be present in IA due to VI. That is, compounds with expected subsurface sources - TCE,
1,1-DCE, and PCE - have patterns in their IA concentrations similar to radon, that has a known
subsurface source. This indicates that these CoCs are likely present in IA under BL conditions
due to VI. On the other hand, benzene and toluene have different concentration patterns
46
-------
compared to radon, suggesting that VI is likely not a concern for these compounds. Decision-
makers could use the qualitative information derived from such plots to evaluate compounds that
are a VI concern at a specific building.
6.1.5 Concentrations of Compounds in Sub-Slab Soil Gas
SS soil gas samples were collected from three spatially distributed locations in each building.
All three locations in each building were used to calculate the average SS compound
concentrations (Figure 14). Error bars in the plots are the ± 1 standard deviation of the measured
concentrations. Figure 15 presents the same data normalized by the indoor concentrations.
c
o
c
0)
u
c
o
u
SUBSURFACE SOURCES
INDOOR
SOURCE
AMBIENT SOURCES
Baseline
Negative
Positive
Radon TCE 1,1-DCE SF6 Benzene Toluene
(pCi/L) (ug/m3) (ug/m3) (mg/m3)xlOO (ug/m3)xlOO (ug/m3)xlO
c
o
c
s
c
o
u
SUBSURFACE SOURCES
INDOOR
SOURCE
AMBIENT SOURCES
Baseline
Negative
Positive
Radon
(pCi/L)/100
TCE
(ug/m3)
PCE
(ug/m3)
SF6
(mg/m3)
Benzene
(ug/m3)
Toluene
(ug/m3)
Figure 14. Concentrations of Compounds Measured in Sub Slab Soil Gas at ASU House
(Panel A) and Moffett Field Building 107 (Panel B).
47
-------
As shown in the normalized concentration graphs, at ASU House, the three compounds with
nominally subsurface sources were present in SS gas at concentrations higher than measured in
IA (except for TCE under BL conditions), consistent with a subsurface source for these
compounds. For Moffett Field Building 107, radon, but not PCE or TCE, was present at a higher
concentration below the building foundation. For both houses, concentrations of the
indoor/ambient source compounds were similar in IA in SS soil gas. These plots further
demonstrate the ability of the building pressure technique to discern sources of various CoCs.
INDOOR
SOURCE
AMBIENT SOURCES
Baseline
Negative
Positive
0.1
Radon TCE 1,1-DCE SF6 Benzene Toluene
Baseline
Negative
Positive
0.1
Radon
TCE
PCE
SF6
Benzene
Toluene
Figure 15. Average Sub-Slab Concentrations Normalized by Average Indoor Air
Concentrations at ASU House (Panel A) and Moffett Field Building 107 (Panel B).
48
-------
6.1.6 Mass Discharges
Mass discharge is the mass of a given compound that moves through the building per unit time.
The mass discharge for each pressure condition is calculated by multiplying the building
ventilation rate (Q;) by the compound concentration. The total mass discharge is calculated as
the product of Q; and the compound's IA concentration (R;, C;, or T;) and the mass discharge
from ambient sources is calculated as the product of Q; and the compound's AA (Ra, Ca, or Ta)
concentration. The difference between the total mass discharge and the mass discharge from
ambient sources provides the mass discharge from subsurface and indoor sources. As examples
of these calculations, in Section 3.3.1.2 the total mass discharge, mass discharge from ambient
sources, and mass discharge from subsurface and indoor sources were determined for radon
using Equations 1, 2, and 3, respectively.™1
The change in mass discharge between pressure conditions accounts for changes in both building
ventilation rate and compound concentrations; thus, the change in mass discharge between
pressure conditions provides a more comprehensive evaluation of the effect of the pressure
condition on VI. Figures 16 and 17 provide the mass discharge for each compound under BL
conditions, induced NP, and induced PP, normalized by the total mass discharge under the BL
condition. In these Figures, normalized mass discharges are labeled in instances where the
values are small relative to the chart scale. "S/I" and "A" refer to mass discharge from
subsurface and indoor sources and from ambient sources, respectively. In a few cases, the mass
discharge calculations yielded negative values. Although actual mass discharge cannot be
negative, variability in measured compound concentrations can yield negative calculated values
of mass discharges from indoor and subsurface sources, reflecting the uncertainty associated
with small measured differences in compound concentrations in IA and AA.
In Figures 16 and 17, values greater than one indicate that the mass discharge was higher than
under BL and values less than one indicate that mass discharge was lower than under BL. For
example, for TCE in the ASU House, the normalized total mass discharge of 3.6 under induced
NP and 0.05 under induced PP indicate that the total mass discharge increased over baseline (by
3.6 times) under the induced NP condition and decreased (by 95%) under induced PP. At both
buildings, for compounds with expected subsurface sources, the total mass discharge was greater
under induced NP than under induced PP. For benzene, toluene, and SFe, (expected ambient and
indoor sources), the total mass discharge was similar for both pressure conditions.
Furthermore, since the normalized BL mass discharge from ambient sources is equivalent to Fa,
the fractional contribution of AA to the IA concentration of a given compound (see Section
3.3.1.3, Equation 7), much can be learned about the sources of the various compounds by
inspection of these values.™11 For instance, Fas are low and in many cases nearly 0 for radon,
xm Note that, as explained in Section 3.3.1.2, indoor sources of radon are assumed to be negligible compared to
subsurface sources, thus Equation 3 is the mass discharge of radon into IA from only subsurface sources.
xxm Normalized mass discharges are also the basis of other Mosley Model calculations. For instance, under PP in
which VI is 'turned off (R,+ = Ra+), FVi+ = [Qr(Q - Ca) - QrCC^ - Ca+)]/(Qr Q) = the normalized mass discharge
from subsurface and indoor sources under BL - normalized mass discharge from subsurface and indoor sources
under PP (Section 3.3.1.3, combination of Equations 5 and 7). In addition, Fm+ = Q+-CC + - Ca+)/(QrQ) =
normalized mass discharge from subsurface and indoor sources under PP (Equations 6 and 8). The calculation for
FVI" and Fm" are similar but somewhat more complicated since scaling factors are added based on radon.
49
-------
TCE, 1,1 -DCE, and PCE, indicating the predominance of subsurface/indoor sources (TCE and
PCE have weak ambient sources). Moreover, an Fa of zero for SF6 is consistent with it being the
indoor tracer, and Fas of nearly 1 for benzene and toluene indicate strong ambient sources.
While the Fa information serves as indicator of a compound's source, the observed similarity of
changes in mass discharges following building pressure perturbation is a more powerful, albeit
still qualitative, CoC source attribution technique. That is, similar to the information gleaned
from Figure 13 (Section 6.1.4), the mass discharges for compounds with expected subsurface
sources - TCE, 1,1-DCE, and PCE - vary under application of the building pressure control
technique similarly to radon (that has a known subsurface source) thereby indicating that these
CoCs are likely present in IA under BL conditions due to VI. On the other hand, benzene and
toluene have a different mass discharge pattern under pressure perturbation compared to radon,
suggesting that VI is likely not a concern for these compounds. Such trends in mass discharges
could aid decision-makers in evaluating compounds that are likely VI concerns at a specific
building.
50
-------
Radon
40
Baseline Negative Positive
TCE
Baseline Negative Positive
400
1,1-DCE
Baseline Negative Positive
Benzene
Toluene
-i
Baseline Negative Positive
Baseline Negative Positive
Baseline Negative Positive
Figure 16. Normalized Mass Discharges at ASU House for Radon, TCE, and 1,1-DCE (Compounds With Primarily
Subsurface Sources) and SFe (a Compound with a Dominant Indoor Source), and Benzene and Toluene (Compounds with
Dominant Ambient Sources).
-------
Radon
I From Subs. &
Indoor
Sources
Baseline Negative Positive
TCE
Baseline Negative Positive
PCE
-2
Baseline Negative Positive
to
Benzene
Toluene
Baseline Negative Positive
Baseline Negative Positive
Baseline Negative Positive
Figure 17. Normalized Mass Discharges at Moffett Field Building 107 for Radon, TCE, PCE (Compounds with Primarily
Subsurface Sources) and SFe (a Compound with a Dominant Indoor Source), and Benzene and Toluene (Compounds with
Dominant Ambient Sources).
-------
6.2 Decision-making Support
6.2.1 Building Pressure Differential
The first verification metric for decision-making is the ability of the pressure control method to
control building pressure. Building pressure control was verified by inspection of the mean I/O
AP that was attained for the 24-hour periods of induced NP and PP at each of the two buildings.
Statistical significance was tested to determine whether the observed mean APs were less than -1
Pa under induced NP and greater than 1 Pa under induced PP. In each instance rejection of the
null hypotheses - that the building pressure differentials were greater than or equal to -1 Pa
(under NP) and less than or equal to 1 Pa (under PP) - at the 95% confidence level (5% false
positive rate) indicates that building pressure control had been achieved. Results are shown in
Table 7 below and were shown graphically in Figure 9, Panels A and C (Section 6.1.1).
Table 7. Indoor/outdoor pressure differentials at ASU House and Moffett Field Building
107.
ASU House
Test
BLe
NP
PP
AP
APi
APf
AP:+
Mean
AP,Pa
N/A
-5.18
3.87
Std
Dev, Pa
N/A
1.01
0.66
Na
N/A
290
289
Std devA/Nb,
Pa
N/A
0.059
0.039
A?" < -1 Pa?c
N/A
yes; p< 0.0001
N/A
N below
-IPa
N/A
288 (99%)
N/A
AP+ > 1 Pa?d
N/A
N/A
yes; p< 0.0001
N above
IPa
N/A
N/A
289 (100%)
Moffett Field Building 107
Test
BL
NP
PP
AP
AP2
AP2-
AP2+
Mean
AP,Pa
-0.83
-2.47
1.03
Std
Dev, Pa
0.60
0.37
0.27
N
290
291
273
Std devA/N,
Pa
0.035
0.021
0.016
AP" < -1 Pa?
N/A
yes; p< 0.0001
N/A
N below
-IPa
N/A
291 (100%)
N/A
AP+ > 1 Pa?
N/A
N/A
yes; p = 0.044
N above
IPa
N/A
N/A
164 (60%)
aNumber of observations
bStandard deviation of the mean
°H0: mean = -1; Hj: mean < -1 Pa; H0 rejected when p < 0.05
dH0: mean = 1; Hj: mean > 1 Pa; H0 rejected when p < 0.05
"Data are not available for BL conditions at ASU House; see Section 3.3
During each of the pressure perturbations at the two buildings, the mean building pressures were
either below (under NP) or above (under PP) the target pressure of -1 Pa and 1 Pa, respectively,
indicating that some degree of pressure building control was achieved under both pressure
perturbation conditions at both buildings. In three of the four instances, mean building pressure
differentials were either substantially less than the target pressure (under NP) or greater than the
target pressure (under PP), as indicated by the low p values generated by the one-sided t-tests.
53
-------
Indeed, not only were the target building pressures achieved (as evidenced by the overall mean),
in these three instances the building pressures were maintained at the target AP over 99% of the
time. Higher percentages indicate that the building pressure was maintained at the target
pressure differential for a greater duration of time.
In one instance, under PP at Moffett Field Building 107, the observed AP was above 1 Pa (p =
0.044), but only 60% of the individual AP observations were greater than 1 Pa. While pressure
control was achieved at Moffett Field Building 107 under PP, the attained AP was not as large as
at ASU House and was only slightly above the target AP of 1 Pa. As can be seen in Figure 9
Panel C, Moffett Field Building 107 was under a slight negative pressure under BL conditions,
ostensibly due to the action of the building's HVAC system; it appears that application of
positive pressure to the building envelop using the window fan was just able to overcome this
inherent negative pressure. This slight negative pressure under BL conditions was also observed
in the cross-foundation AP measurements (Table Al).
6.2.2 Vapor Intrusion Enhancement and Reduction
The second verification metric for decision-making is the effect of the pressure control method
on the enhancement and reduction of radon VI. This metric was evaluated using the mass
discharge of radon from subsurface sources through the building foundation. Under induced NP,
the mass discharge of chemicals with subsurface sources - including radon and CoCs - into the
building may increase compared to BL. Similarly, under induced PP, the mass discharge from
subsurface sources may either decrease compared to BL or be reduced to zero, where the latter
condition indicates that VI has effectively been stopped.
As described in Sections 3.3.1.2 and 5.1.2, the radon mass discharge from subsurface sources
was calculated under BL, induced NP, and induced PP for both buildings to determine if VI was
enhanced under induced NP and reduced under induced PP. Results of these comparisons at
both buildings are shown in Table 8. Mass discharges are in pCi h"1.
54
-------
Table 8. Comparison of Radon Mass Discharges from Subsurface Sources to Determine VI
Enhancement and Reduction.
ASUHouse
error in
Pressure Qj-(Ri-Ra) Qr(RrRa) Qr-(R-Ra)> Qr(R-Ra)?a Q+-(R+-Ra+)< Q.-(R-Ra)?b
BL
NP
PP
14830
706231
8143
8704
999852 No; p = 0.1769
32347
No; p = 0.3794
Moffett Field Building 107
error in
Pressure QrCR-Ra) Qr(R-Ra) Qr-(R-Ra)>Qr(R-Ra)?a Q+- (R+-Ra+) < Q.- (R-Ra)?b
BL
NP
PP
82556
213492
-16153C
11579
137656 No; p = 0.1203
93610
No; p = 0.1326
aH0: Qi- (R-R.) = Qf- (R -Ra ); HI: Q, • (Rf-Ra ) > Q;- (R-Ra); H0 rejected when p < 0.05
bH0: Qi- (Ri-Re) = Qi+" (Ri+-R.+); HI: Q+- (R+-Ra+) < Qi*(RrR.); H0 rejected when p < 0.05
°While mass discharges cannot be negative, this calculated value was < 0 due to the measurement
variability at low radon concentrations. For the calculation of the p-value, Q1+-(R+-Ra+) was set equal to 0.
Qualitatively, VI was observed to have been enhanced under induced NP and reduced under
induced PP at both buildings; this is apparent by comparison of the magnitudes of the mass
discharges under NP vs. BL and PP vs. BL, respectively. Thus the building pressure
manipulation method was shown to control VI to some extent. However, in none of the four
cases was there sufficient statistical evidence to reject the null hypotheses and conclude that VI
was enhanced under induced NP or reduced under induced PP. The failure to find statistically
significant differences is due to the estimated errors in the various radon mass discharges. These
estimated errors are driven by variability in the spatially distributed IA SFe concentrations (this
led to large estimated errors in the calculated building ventilation rates - see the error bars on Q;s
in Figure 10, Section 6.1.2) as well as variability in the spatially distributed IA radon
concentrations (see error bars on R; data presented in Figure 12, Section 6. 1 .4). Another driver
of the observed variability is that several IA and AA radon measurements were at or below
estimated method detection limits where measurement uncertainty is magnified.
Given that VI was qualitatively determined to have been reduced under induced PP, whether VI
had been 'turned off was investigated by comparison of R;+ to Ra+ at each building. Results of
the comparisons are summarized in Table 9. At neither building was there evidence sufficient to
conclude that R;+ was different from Ra+; VI was thus concluded to have been stopped under PP
and the more simplified version of the Mosley Model was used to calculate the FVI+ values for
each of the CoCs at each building.
55
-------
Table 9. Comparison of Indoor and Ambient Air Radon Concentrations under Positive
Pressure.
Building3
ASU House
Moffett Field Building 107
R+
-2.74
-1.24
std dev R,+
1.17
0.37
Ra+ est error in Ra+
-2.67
-1.11
1.15
0.33
R+ =
Yes; p
Yes; p
Ra+?b
= 0.949
= 0.690
aData were log transformed
bH0: R+ = Ra+; HI: R+ ± Ra+; H0 rejected when p < 0.05
6.2.3 Fractional Contribution of Vapor Intrusion to Indoor CoC Concentrations
The third verification metric under decision-making support is the ability of the pressure control
method to provide an improved understanding of the contribution of VI to IA CoC
concentrations. This metric was assessed by calculation of two independent estimates of FVi
under BL using (i) CoC measurements from BL and induced NP (Fyf) and (ii) CoC
measurements from BL and induced PP (Fyi+) for each of four CoCs at both buildings. Error
estimates in each FVI were also calculated. Prior to the field program, two of the CoCs selected
were expected to have primarily subsurface sources, TCE and 1,1-DCE at ASU House and TCE
and PCE at Moffett Field Building 107; whereas the other two CoCs were expected only to have
indoor/ambient sources (and no appreciable subsurface sources), benzene and toluene at both
buildings.XX1V Fyf and FVI+ for each CoC at each building, as well as corresponding 95%
confidence intervals generated by the Monte Carlo error analysis, are provided in Table 10.™ Fa,
Fin" and F;n+ are also given in Table 10 to provide the comprehensive picture of CoC source
attribution that results from application of the building pressure technique at these two buildings.
XX1V As noted in Section 6.1.4, after completion of the verification test, it was discovered that TCE was being stored
in the ASU house garage creating an additional potential indoor source of TCE.
"^ The 95% confidence interval estimates presented in Table 10 are the result of averaging the bounds over the
eleven combinations of correlation coefficients for the given building/CoC combination. The bounds of the
confidence intervals were quite consistent across different combinations of correlation coefficients. This implies
that under the assumptions used in this analysis, the choice of correlation between any two dependent random
variables in these two formulae does not substantively affect the 95% confidence interval of FVi+ or FVI".
56
-------
Table 10. Fractional Contribution of Ambient Sources, Indoor Sources, and VI to Indoor
CoC Concentrations Under Baseline Conditions.
ASU House
Using NP and BL results
Using PP and BL results, Rj+ = R/
Compound
TCE
1,1-DCE
Benzene
Toluene
Fa
0.01
0.30
0.85
1.04
Fin'
0.94
-6.87
0.14
-0.09
FVI"
0.05
7.57
0.01
0.05
LBa AF^
-9.2
-1077
-18
-23
UBb AFvi"
5.3
529
17
14
Fa
0.01
0.30
0.85
1.04
Fin+
0.02
0.02
1.81
1.37
FVI+
0.97
0.68
-1.66
-1.42
LB AF^
0.84
-6.0
-25
-18
UB AF^
1.1
8.2
22
14
Moffett Field Building 107
Using NP and BL results
Using PP and BL results, R;+ = R/
Compound
TCE
PCE
Benzene
Toluene
Fa
0.02
0.04
1.00
0.90
Fin'
-0.56
-0.68
-0.21
0.02
FVI'
1.55
1.64
0.21
0.09
LB AFvj-
-5.8
-6.2
-10
-21
UB AFvj-
2.5
2.7
10
19
Fa
0.02
0.04
1.00
0.90
Fin+
0.03
-1.00
0.73
-2.32
F^ LB AFv!+ UB AF^
0.95
1.96
-0.73
2.42
0.65
-1.6
-20
-33
1.3
6.3
21
42
aLower bound of 95% error interval
bUpper bound of 95% error interval
By definition, FVi is expected to be between 0 and 1 (i.e., between 0% and 100% of the CoC
concentration in IA is attributable to VI) and the sum of the fractional contributions from all
possible sources (ambient, indoor, and subsurface) is defined to be 1 (i.e., Fa + F;n + FVI = 1). FVI
values of less than zero or greater than one are indicative of either variability in the dataset used
for the calculations or incorrect model assumptions. Of the 16 FVI values, five were greater than
one and three were less than zero. In addition, the Monte Carlo uncertainty analysis indicated
that uncertainties in the FVI values are likely larger than calculated values. The best bounded
estimates of FVI were for TCE calculated using the BL and PP data at both the ASU House and
Moffett Field, where FVI+ was 0.97 and 0.95 with 95% confidence intervals of 0.84 to 1.1 and
0.65 to 1.3, respectively.
Although the results of the FVI calculations provide only modest quantitative information
regarding the fraction of each CoC in IA attributable to VI, the qualitative pattern was generally
as predicted and is similar to the observed qualitative trends in mass discharges discussed in
Section 6.1.6. For instance, the FVI values for the two CoCs with expected subsurface sources
were close to or greater than one in seven of eight cases. The exception (Fyi = 0.05) was for
TCE at the ASU House. As noted previously, an indoor source of TCE was found in the
building after completion of the study; in this instance, the model may have predicted the
presence of the indoor source. The FVI values for the two CoCs without expected subsurface
sources were close to or less than zero in seven of eight cases. The exception (toluene at Moffett
Field Building 107) is discussed further below. Moreover, as described in Section 6.1.6, the Fa
values are low and in many cases nearly 0 for radon, TCE, 1,1-DCE, and PCE, indicating the
57
-------
predominance of subsurface/indoor sources; whereas Fas are nearly one for benzene and toluene,
pointing to predominant ambient sources.
It is helpful to consider limitations in the Mosley Model to better understand the quantitative FVI
calculations. One limitation is the assumption that, under NP, radon may be used a tracer gas for
the movement of subsurface vapors into the overlying structure. In the model radon is applied as
a scaling factor [in the form of the term Qj • (R; - Ra)/{Qi" • (Rf - Ra") - Qi • (Ri - Ra)}] to the
observed difference of NP and BL mass discharges for each CoC [the term Q;~ • (C;~ - Ca~) - Q; •
(C; - Ca)]. That several of the mean FVI" values were outside the [0, 1] interval may be due to
radon serving as an imperfect indicator for the movement of other subsurface gases.
Specifically, FVI" values of greater than one suggest that the induction of NP had a greater effect
on the mass discharge of CoCs through the building foundation compared to radon.
Another limitation of the Mosley Model is its sensitivity to short-term variability in
concentrations of CoCs and radon.12 This sensitivity is illustrated by the FVI+ results for toluene
for Moffett Field Building 107. Here the AA and IA concentrations of toluene were similar
under all three pressure conditions, suggesting that AA is the predominant source of toluene in
IA (reflected in the calculated value of Fa of 0.9). Under BL conditions, C; > Ca, but under
induced PP, C;+ < Ca+ (see Figures 11 and 12, Sections 6.1.3 and 6.1.4). Although these
observed differences are likely attributable to measurement variability, the Mosley Model yields
nonsensical FVI+ and F;n+ values of 2.42 and -2.32, respectively, because the model attributes the
measured "decrease" in the IA concentration of toluene relative to AA as a decrease in the
contribution from VI.
Highlighted in red in Table A3 are other specific instances where modeled FVI values were
impacted by measurement uncertainty or failure of specific model assumptions."™ In general,
FVI estimates may benefit from improvements in the sensitivity of critical measurements such as
radon and CoCs in ambient and IA™* and from more homogeneous distribution of the SFe tracer
gas such that building ventilation rates are more accurately known. It may also be beneficial to
perform sampling over longer time intervals to reduce short-term variability in CoC and radon
concentrations and to have more than a single observation of AA concentrations of CoCs and
radon at each pressure condition to better characterize the true distributions of these
concentrations.
x™ FVI" estimates were impacted by the following issues. For example, at ASU House, Ra and Ra~ were both
nondetects; measurements at such low concentrations are subject to elevated uncertainty which likely decreased the
fidelity of the FVi" estimates. Also, as explained in more detail in Section 6.4, the assumption that Cs = Cs~ was not
valid for benzene; and if SS-1 was excluded, the assumption was also not valid for TCE and 1,1-DCE. The failure
of this assumption demonstrates that the FVi" values for these three CoCs may be subject to greater uncertainty. The
FVI" for toluene may also be of limited value given that Q < Ca which is contradictory given that one of the model
assumptions is that a CoC's concentration in AA is less than in IA. Other examples of problems observed for FVi+
estimates include, for instance: 1,1-DCE at ASU House, where a number of ND results impacted the calculation;
PCE at Moffett Field Building 107, where Ca+ > Q+; benzene at both buildings, where Q, • (Q - Ca) < Q/ • (Q+ -
Ca+); toluene at ASU House in which Ca > Q. Substitution of nondetects degrades the fidelity of model outputs; in
all of the other FVi+ calculations, a model assumption was violated which most likely led to FVI+ results outside of
the interval of 0 and 1.
xx™ Monte Carlo error estimates may have been artificially elevated given that several IA/AA radon and CoC
measurements were at or below estimated method detection limits where measurement uncertainty is magnified.
58
-------
Post hoc power analysis was not performed to the FVI calculations. The 95% confidence
intervals for FVi that were calculated via Monte Carlo simulation extended well below zero and
well above one, meaning that these data do not meaningfully narrow down the portion of the
interval [0,1] likely to contain the true value of FVI. A retrospective calculation to estimate
minimum detectable differences is not useful in this context because the magnitude of AFyi is so
large that FVi would have to be nonsensically large to outweigh AFVi in a statistically significant
manner.
6.3 Comparability
The comparability of the building pressure control technique as implemented at two different
buildings is shown in Table 11. In general lower RPDs indicate better comparability. Given that
each building's characteristics were different (ASU House is a single family home compared to
Moffett Field Building 107 - a commercial building) and especially that the HVAC system at
Moffett Field Building 107 caused a slight negative pressure differential under BL, it is not
surprising that high RPDs were observed.
Table 11. Comparability of Building Pressure Control Results.
AP Value, Pa RPD
APf
AP2
APi+
AP2+
-5.18
-2.47
3.87
1.03
j 71
J
.
I 116
J
Moreover, the ability to evaluate comparability was limited because testing was conducted at
only two buildings.
6.4 Operational Factors
The technology vendor executed all aspects of the building pressure control test at both
buildings. Battelle oversaw testing at both sites; the VTC was onsite at the ASU House and
conducted daily briefings with the team during testing at Moffett Field Building 107. A
minimum of two people were required to execute the field work, and one staff person had to
have the experience and specialized knowledge in indoor and outdoor air sampling and use of
analytical instrumentation required for implementation of a typical VI sampling program. Also
required for the field team was the ability to install SS sampling points. GSI Environmental is
currently preparing an instruction manual with detailed guidance on how to execute the building
pressure control technique. However, at the time of the verification test, no detailed instruction
manual or written guidance was available, beyond the project QAPP, the vendor's ESTCP
report5, and instruction manuals for the Omniguard 4® and RAD7®. Examples of test procedures
that should be included in such an instruction manual include: selection of appropriate locations
for IA, AA, and SS sampling and tracer gas release; installation of SS sampling points; delivery
59
-------
of tracer gas; receipt, pre- and post-sampling sample media integrity checks; and guidance on
such issues as sample collection, shipment of canisters, flow control devices, and PVF bags.
It was determined that the Omniguard 4®, set up to measure I/O AP at ASU House, was not
logging data during BL conditions. (Note that the lack of BL ASU House I/O AP data is the
subject of QAPP Deviation 1.) Troubleshooting was performed by the vendor, the instrument
was reconfigured in about 10 minutes, and datalogging was re-enabled coincident with the
beginning of the NP condition. The instrument had not malfunctioned; rather, its default settings
were inappropriate for the intended use. Inspection of the instrument's instruction manual
helped to resolve the problem. No issues were encountered with the RAD7 . No canisters (out
of 47) were rejected during pre-sampling integrity checks (based on pressures as received). The
pressures in three canisters (out of 47; 6%) did increase after sampling and before analysis
(subject of QAPP Deviation 3), indicating that these samples had been slightly diluted with gas
of unknown composition. Two PVF bags (out of a total 48; 4%) failed pre-sampling checks and
were not used; these failures did not impact study outcomes. One bag (out of 48; 2%) arrived at
the analytical laboratory at a lower volume than the others, potentially indicating a leak.
The vendor required approximately one day to set up the equipment to conduct the field work,
followed by three days for project execution. Labor, travel, and expenses for testing at both sites
totaled approximately $23,000. Each Omniguard 4 was $1,500 and requires annual
recalibration. Similarly, the RAD7 was $6,000 and requires annual recalibration. The total cost
for the rental of canisters and flow control devices, purchase of PVF bags, and for the various
analyses described in Section 3.3 was approximately $21,000. Moreover, miscellaneous gas
sampling equipment and accessories were required. Thus, the total cost for implementing this
technology for this verification test at two sites over 3.5 days at each site, excluding any data
reduction, interpretation, and reporting, was approximately $50,000.
For the routine implementation of the technology at a given site, the field work is expected to
require approximately 80 person-hours (2 staff- 4 days • 9 hours/day). Additional costs would
include travel and expenses, as well as time for data evaluation and reporting after the field work
is completed. One differential pressure instrument ($1,500) is required to perform the I/O
monitoring. The cost for laboratory analysis of the basic set of canisters (VOCs and SFe) and
PVF bags (radon) samples is approximately $6,000, including media and shipping. This cost
covers analysis of 9 IA samples, 3 AA samples, and two field duplicates. Routine
implementation would not require SS sampling, either for gas-phase species or cross-foundation
differential pressures.
6.5 Validation of Model Assumptions
The results of the validation of the various model assumptions, specifically those that can be
explicitly tested, are presented in this section. Inspection of the verification of these assumptions
helps to explain the outcomes of the FVI calculations.
A total of eight assumptions were tested at each building and are organized into three groups.
The first group includes two assumptions to determine if inducing a NP in the building had a
significant effect on CoC and radon concentrations in SS soil gas below the building foundation.
60
-------
Shown in Tables 12 and 13 are the results of testing if Cs = Cs~ (model assumption 1) for the
various CoCs and for Rs = Rs~ (radon; model assumption 2), respectively, at both buildings. For
FVI" to be most meaningful, the SS concentration of CoCs and radon should not change when NP
is applied to the building; that is, source strengths and distributions of the CoCs and radon should
not change under induced NP, given that such are required assumptions of the Mosley Model.
As can be seen by inspection of the individual SS data points for the various CoCs and radon
(see Tables A4 and A5 for the non-log transformed data), concentrations of CoCs and radon at
SS-1 at the ASU House were quite different than those at the other two SS sampling points.
Also, SS-1 was located in the garage, and the door connecting the garage to the remainder of the
usable space (see Figure Al) was mainly kept closed during pressure testing. Thus additional
statistical tests were conducted by excluding SS-1 to test the assumptions that Cs = Cs~ and Rs =
As can be seen in Table 12, Cs = Cs~ did not hold for benzene at ASU House and was also not
valid for TCE and 1,1-DCE (if SS-1 was excluded). The failure of this assumption demonstrates
that Fyf for these three CoCs may be subject to greater uncertainty since the strength of the
subsurface source changed under induced NP. Applied building pressures were effectively
smaller at Moffett Field Building 107, likely explaining the lower observed perturbation in soil
gas CoC concentrations. SS radon concentrations at both buildings remained relatively constant
under induced NP compared to BL (Table 13; see also Table A5 for individual data points and
non-log transformed data).
Table 12. Validation of Model Assumption 1.
ASUHouse
Compound3
TCE
1,1-DCE
Benzene
Toluene
ASUHouse
Compound
TCE
1,1-DCE
Benzene
Toluene
cs
2.29
0.52
-0.94
0.95
std dev Cs
0
1
0
0
.35
.83
.07
.35
cs
4.
4.
69
74
-0.35
1.
29
std dev Cs
2.2
1
0
1
.82
.15
.03
C = C
V-'S V-'S
Yes;
Yes;
No;
Yes;
P
P
P
P
= 0
= 0
= 0.
= 0
?b
.2390
.1804
0075
.4952
(excluding SS-1)
cs
2.12
-0.51
-0.98
0.76
Moffett Field Building
Compound
TCE
PCE
Benzene
Toluene
cs
0.57
0.55
0.04
1.29
std dev Cs
0
0
0
0
107
.25
.65
.02
.17
std dev Cs
0
0
0
0
.67
.73
.93
.69
cs
5.
5
-0
0.
97
.8
.44
69
cs
0.
0.
-0
1.
52
52
.13
23
std dev Cs~ Cs =
0
0
0
.09
.01
0
No;
No;
No;
Yes;
P
P
P
P
std dev Cs Cs =
0
1
0
0
.47
.22
.76
.51
Yes;
Yes;
Yes;
Yes;
P
P
P
P
= cs
= 0.
= 0.
= 0.
= 0
= cs
= 0
= 0
-?
0287
0401
0247
.6746
-?
.8384
.9384
= 0.2943
= 0
.6635
aData were log transformed
bH0: Cs = Cs~; H^ Cs ^ Cs~; H0 rejected when p < 0.05
61
-------
Table 13. Validation of Model Assumption 2.
Building3
ASU House
ASU House (excluding SS-1)
Moffett Field
Rs
4.42
3.59
5.28
std dev RS
1.65
1.13
0.97
Rs
5.15
4.70
5.75
std dev RS"
0.80
0.05
0.63
Rs = Rs ?'
Yes; p = 0.3502
Yes; p = 0.4109
Yes; p = 0.2884
aa were og ransorme
bH0: RS = RS ; Hj: RS + RS ; H0 rejected when p < 0.05
The second group of assumptions whose validity were verified related to if radon concentrations
in AA were in fact much lower than those in SS soil gas below the building foundation. The
results of the validation of model assumptions 3, 4, and 5, that is, that Ra « Rs, Ra" « Rs~, and
Ra+ « Rs+, are shown in Table 14. (In addition, see Table A5 for individual data points and non-
log transformed data.) In all cases, including those where SS-1 was excluded at ASU House, all
assumptions were statistically validated. That these assumptions were validated supports the use
of the Mosley Model and the veracity of the FVI calculations.
Table 14. Validation of Model Assumptions 3, 4, and 5.
Baseline
Building3
ASU House
ASU House (excluding SS-1)
Moffett Field Building 107
Ra
-2.26
-2.26
-1.73
std dev Rb
0.78
0.78
1.96
Rs
4.42
3.59
5.28
std dev RS
1.65
1.13
0.97
Ra < RS ?C
Yes;p
Yes;p
Yes; p
= 0.0046
= 0.0187
= 0.0062
Negative Pressure
Building
ASU House6
ASU House (excluding SS-l)e
Moffett Field Building 107
Ra
-3.65
-3.65
-1.36
std dev Rf
1.20
1.20
0.66
Rs
5.15
4.70
5.75
std dev Rs"
0.80
0.05
0.63
Ra < Rs ?"
Yes; p
Yes;p
Yes; p
= 0.0005
= 0.0034
< 0.0001
Positive Pressure
Building
ASU House
ASU House (excluding SS-1)
Moffett Field Building 107
Ra+
-2.67
-2.67
-1.11
std dev R,+
1.15
1.15
0.33
Rs+
3.88
2.81
4.64
std dev RS+
2.11
1.43
1.76
Ra+<
Yes;p
Yes;p
Yes;p
** 13 i~
~~~ AVt- :
= 0.0084
= 0.0256
= 0.0131
aData were log transformed
bStandard deviation of AA radon based standard deviation of IA radon; see text Chapter 5
CH0: Ra - RS = 0; Hj: Ra - RS < 0; H0 rejected when p < 0.05
dH0: Ra~ - RS" = 0; Hj: Ra" - RS" < 0; H0 rejected when p < 0.05
Standard deviation of AA radon assumed equal to standard deviation of IA radon; see text
Chapter 5
dH0: Ra+ - RS+ = 0; Hj: Ra+ - RS+ < 0; H0 rejected when p < 0.05
With respect to the validation of model assumptions 1 through 5, it is important to recognize that
a number of the Cs and Rs observations were at or below estimated detection limits. In such
62
-------
cases measurement uncertainty is magnified, and statistical difference testing is likely adversely
impacted.
The third and final group of assumptions tested included those to determine if, in IA, the loss of
radon through building ventilation is much greater than the loss due to radioactive decay. Shown
in Table 15 are the results of validating model assumptions 6, 7, and 8, that is, Q; » XV, Q;~ »
XV, and Q;+ » XV. These results are also shown graphically in Figures A4 through A9. Under
all three pressure conditions at both buildings, building ventilation rates were indeed larger than
XV (by at least a factor of 26 in all cases), but only under BL conditions was there sufficient
statistical evidence to conclude that Q; > XV. Failure to reject the null hypotheses under induced
NP and induced PP was due to the large variability in Q;, that is due in large part to the observed
spatial heterogeneity in the IA SFe concentrations that determine Q; (see error bars in Figure 10,
Section 6.1.2). That these assumptions were either statistically validated, or shown to be verified
at least qualitatively, supports the simplification performed in the Mosley Model, that is, that XV
is negligible compared to Q;, Qf, and Q;+.
Table 15. Validation of Model Assumptions 6, 7, and 8.
ASUHouse
Pressure
BL
NP
PP
Q,
54.5
384
364
est error
Q,
14.2
351
342
XV
2.07
2.07
2.07
est error
XV
0.62
0.62
0.62
Q[
Yes;
No;
No;
>
P
P
P
XV?a'b'c
= 0.0332
= 0.1955
= 0.2004
#SD that
XV xv?
= 0.0084
= 0.0740
= 0.1823
#SD that
XV XV; H0 rejected when p < 0.05
bNP: H0: Qf = XV; HI: Qf > XV; H0 rejected when p < 0.05
bPP: H0: Q,+ = XV; HI: Q,+ > XV; H0 rejected when p < 0.05
As shown in Table 16, post-hoc power analyses (assuming 80 % power and 5% false positive
rate) were performed to determine the minimum detectable differences between Q; and XV under
induced NP and induced PP at both buildings. Such post-hoc analyses were performed only in
those instances where the prerequisites were met, including that a sufficient number of
observations were present (at least 3) for both samples in the comparison, the p-value was not
significant, and all concentrations were greater than MDLs.
Minimum detectable differences between Q;~ and XV and Q;+ and XV indicate that, assuming the
measured variability in building ventilation rates remain constant, ventilation rates would need to
63
-------
be on the order of 1200 to 2500 m3 h"1, or roughly 3.3 to 6.8 air changes h"1 to detect a
statistically significant difference between the ventilation rates and the radon radioactive decay
constant. One important conclusion from these calculations of minimum detectable differences
is that the variability in the building flow rates are quite high under induced NP and induced PP,
given that ventilation rates must exceed XV by a factor of 1000 before it can be concluded with
statistical confidence that the former is greater than the latter. An alternative, and potentially
superior, method to better ensure a statistically significant difference in future testing would be
to improve the spatial homogeneity of the concentration of the indoor atmosphere, as the error in
the building ventilation rates is driven in large part by the observed spatial heterogeneity of the
SF6 IA concentrations.
In a similar way, minimum detectable differences were calculated for Cs and Cs~ for TCE, PCE,
and benzene (assumption 1) and for radon (Rs and Rs~) at Moffett Field Building 107 (Table 16).
These differences may be interpreted as how much larger the concentrations measured under NP
conditions would have to be in order for the difference to be observed as statistically significant.
Essentially differences in SS TCE and PCE would have to be fairly large (relative to observed
mean concentrations of ~ 2 ug m"3) in order to detect differences. Rs~ would have to be even
larger than Rs (by nearly 15,000 pCi L"1). The calculated minimum detectable difference for
benzene reveals a limitation in the post-hoc power calculation itself, since only a nonsensically
large difference between BL and NP SS benzene concentrations is predicted to be statistically
observable.
Table 16. Minimum Detectable Differences for Model Assumptions 1 and 2 at Moffett Field
Building 107.
Compound Minimum Detectable Difference3
TCE 27
PCE 1,620
Benzene 92,107,564
Radon 14,933
aConcentrations expressed in ug m"3 for CoCs, and pCi L"1 for radon
The benzene result notwithstanding, the outcomes of the various minimum detectable difference
estimates suggest that the observed variability in these SS concentrations are quite large;
consequently, the ability to assess the accuracy of assumptions 1 and 2 - and thereby confirm the
utility of the Mosley Model - is negatively impacted. Essentially, the large observed variability
in concentrations combined with the relatively few spatially distributed measurements increase
the likelihood of false negatives (incorrectly retaining the null hypotheses formulated for
assumptions 1 and 2 , that Cs = Cs~ and Rs = Rs~), thereby leading to the conclusion that the Fyf
estimates are of a higher fidelity than they otherwise may be.
64
-------
Chapter 7
Performance Summary
The objective of this verification test was to generate performance data on the use of the building
pressure control technique, as conducted by GSI Environmental, as a method to understand the
impact of VI on the concentrations of CoCs in IA. The data generated from this verification test
are intended to provide organizations and users with information on the utility of such a
methodology.
The pressure control technique was evaluated at two different buildings where VI was a known
concern using the following types of performance parameters.
• Decision-making support
• Comparability
• Operational factors
In general, the goal of implementing the building pressure control method is to obtain a better
understanding of VI in a building. If the control of building pressure results in clear changes in
building conditions, such as I/O differential pressures and concentrations of radon and CoCs,
then the pressure control method may yield results that are useful for decision-making (i.e., is VI
a concern for this building?). The effectiveness of the building pressure control method to
support decision-making was evaluated via three metrics.
1. Building pressure differential: Did the pressure control method control building pressure?
2. Vapor intrusion enhancement and reduction: Did the pressure control method increase
the mass discharge of radon from subsurface sources through the building foundation
under induced NP conditions and/or decrease the mass discharge of radon from
subsurface sources through the building foundation under induced PP conditions?
3. Fractional contribution of vapor intrusion to indoor CoC concentrations: Did the pressure
control method provide an improved understanding of the contribution of VI to the
concentration of individual CoCs detected in IA?
Additional support to decision-makers was also provided by qualitative trends, with respect to
changes in building pressure, in concentrations of compounds in IA, as well as trends in the
changes of compound mass discharges.
65
-------
Beyond the three metrics comprising decision-making support, the performance metric of
comparability was assessed for the pressure control technique as the similarity of the I/O
differential pressures achieved under induced NP and PP conditions at each of two buildings.
The final performance metric was comprised of an assessment of operational factors such as ease
of implementation of the pressure control technology, the expertise required to carry out the field
work and interpret the results were also determined, and costs to perform the testing. The results
of the verification are summarized below.
Building Pressure Differential
For both buildings, the building pressure control method achieved a measureable negative
pressure gradient both across the building envelope (the I/O differential pressure) and the
building foundation under induced NP, as well as a measureable positive pressure gradient across
the building envelope and building foundation under induced PP. Furthermore, during each of
the pressure perturbations at the two buildings, the mean I/O differential pressures were either
below (under NP) or above (under PP) the target pressure of -1 Pa and 1 Pa, respectively. These
results indicate that some degree of building pressure control was achieved under both pressure
perturbation conditions at both buildings.
Vapor Intrusion Enhancement and Reduction
At both buildings, the building pressure control method had the expected qualitative effect on the
mass discharge of radon from subsurface sources through the building foundation. That is,
under induced NP, the mass discharge of radon from subsurface sources through the building
foundation increased compared to BL, indicating that radon vapor intrusion had been enhanced;
and under induced PP, the mass discharge of radon from subsurface sources through the building
foundation decreased compared to BL, indicating that radon vapor intrusion had been reduced.
However, in none of these four cases (NP and PP comparisons to BL at two buildings) was the
difference in mass discharges found to be statistically significant - due to the large estimated
errors in the measured mass discharges. Radon concentrations in IA and AA under induced PP
were also compared to ascertain if radon vapor intrusion had been stopped under induced PP.
For both buildings IA and AA radon concentrations were not found to be statistically different,
thus indicating an absence of radon vapor intrusion under induced PP.
Fractional Contribution of Vapor Intrusion to Indoor CoC Concentrations
The pressure control method had the expected qualitative effect on CoC concentrations in IA.
For both radon (that has a known subsurface source) and the CoCs with expected subsurface
sources (TCE, 1,1-DCE, and PCE), concentrations in IA were greater than in AA under induced
NP, but similar to concentrations in AA under induced PP. For the CoCs without expected
subsurface sources (benzene and toluene), concentrations in IA were similar to concentrations in
AA for all pressure conditions. Similar trends were seen in mass discharges. Mass discharges of
the CoCs with expected subsurface sources varied under application of the building pressure
control technique similarly to radon, but compounds without expected subsurface sources had a
pattern different than radon under pressure perturbation.
66
-------
The building pressure control technique generated less definitive quantitative results. FVI is the
fraction of the measured IA concentration of a given CoC (under BL conditions) that is due to
VI. By definition, FVI is expected to be between 0 and 1 (i.e., between 0% and 100% of the CoC
concentration in IA is attributable to VI). Under each induced pressure condition, and at both
buildings, a total of 16 FVIS were calculated. For the two CoCs expected to have subsurface
sources - TCE and 1,1-DCE at ASU House and TCE and PCE at Moffett Field Building 107 -
and for two CoCs expected only to have indoor/ambient sources - benzene and toluene at both
buildings (2 buildings • 2 pressure conditions • 4 CoCs). FVI values of less than zero or greater
than one are indicative of either variability in the dataset used for the calculations or incorrect
model assumptions. Of the 16 FVi values, five were greater than one and three were less than
zero. In addition, the uncertainty analysis indicated that uncertainties in the FVI values are likely
larger than calculated values. Nonetheless, the FVI values for the two CoCs with expected
subsurface sources were close to or greater than one in seven of eight cases, and the FVI values
for the two CoCs without expected subsurface sources were close to or less than zero in seven of
eight cases. In general, the variability in measured concentrations limited the quantitative
interpretation of the FVI values.
Comparability
The comparability of the building pressure differential achieved at two buildings was assessed as
the RPD between the mean pressure differentials measured under both induced NP and induced
PP. RPDs were 71% and 116% under NP and PP, respectively. In general lower RPDs indicate
better comparability. Thus, while pressure control was achieved at both buildings, the magnitude
of the induced pressure gradients varied, most likely due to differences in building characteristics
such as HVAC systems. Moreover, implementation of the pressure control method in only two
buildings provided a limited dataset for evaluation of comparability.
Operational Factors
A minimum of two people were required to execute the field work, and at least one of these
personnel must have the experience and specialized knowledge in indoor and outdoor air
sampling, the use of analytical instrumentation required for a typical VI field investigation, and
the ability to install SS sampling points. No detailed instruction manual or written guidance was
available that provided guidance on how to execute various test procedures; however, such
guidance is expected to be available in the future. Settings on one pressure differential
measurement instrument had to be reconfigured during the test; no issues were encountered with
the real-time radon instruments. No canisters (out of 47) were rejected during pre-sampling
integrity checks (based on pressures as received). The pressures did increase (i.e., the vacuum
decreased) in three canisters (out of 47; 6%) after sampling and before analysis, indicating that
these samples had been slightly diluted with gas of unknown composition. Two PVF bags (out
of a total 48; 4%) failed pre-sampling checks and were not used; these failures did not impact
study outcomes. One bag (out of 48; 2%) arrived at the analytical laboratory at a lower volume
than the others, potentially indicating a leak.
For the routine implementation of the technology at a given site, the field work is expected to
require approximately 80 person-hours (2 staff • 4 days • 9 hours/day). Additional costs would
67
-------
include travel and expenses, as well as time for data evaluation and reporting after the field work
is completed. One differential pressure instrument ($1,500) is required to perform the I/O
monitoring. The cost for laboratory analysis of the basic set of canisters (VOCs and SFe) and
PVF bags (radon) samples is approximately $6,000, including media and shipping. This cost
covers analysis of 9 IA samples, 3 AA samples, and two field duplicates. Note that SS sampling
is not required for routine implementation of this technology.
68
-------
Chapter 8
References
1. U. S. EPA (2002). OSWER Draft Guidance for Evaluating the Vapor Intrusion to Indoor Air
Pathway from Groundwater and Soils (Subsurface Vapor Intrusion Guidance). November
2002 EPA 530-D-02-004.
2. U.S. Navy (2004). Memorandum entitled "Navy Policy on the Use of Background Chemical
Levels." January 30, 2004. Available at
http://web.ead.anl.gov/ecorisk/policy/pdf/Final_Navy_Background_Policy.pdf, accessed
December 11,2011.
3. U.S. Navy (2008). Memorandum entitled "Navy/Marine Corps Policy on Vapor Intrusion."
April 29, 2008. Available at
https://portal.navfac.navv.mil/portal/page/portal/navfac/navfac ww_pp/navfac nfesc_pp/env
ironmental/erb/resourceerb/don%20vi%20policy-final.pdf, accessed December 11, 2011.
4. Interstate Technology and Regulatory Council (ITRC) 2007. "Vapor Intrusion Pathway: A
Practical Guideline." Washington, DC, January 2007. Available at
http://www.itrcweb.org/Documents/VI-1.pdf accessed December 11, 2011.
5. GSI Environmental, Inc. (2009). "Results and Lessons Learned Interim Report; Proposed
Tier 2 Screening Criteria and Tier 3 Field Procedures for Evaluation of Vapor Intrusion."
ESTCP Project ER-0707, October 30, 2009.
6. McAlary T., R. Ettinger, P. Johnson, B. Eklund, H. Hayes, D.B. Chadwick, I. Rivera-Duarte
(2009). Review of Best Practices, Knowledge and Data Gaps, and Research Opportunities of
the U.S. Department of Navy Vapor Intrusion Focus Areas. Technical Report 1982,
SPAWAR Systems Center Pacific, May 2009. Available at
http://www.spawar.navy.mil/sti/publications/pubs/tr/1982/trl982cond.pdf, accessed
December 11,2011.
7. McAlary, T.A., R. Ettinger and P. Johnson, 2005, "Reference Handbook for Site-Specific
Assessment of Subsurface Vapor Intrusion to Indoor Air," EPRI, Palo Alto, CA, 2005, EPRI
Document #1008492.
69
-------
8. McHugh, I.E., D.E. Hammond, T. Nickels, and B. Hartman (2008). "Use of radon
measurements for evaluation of volatile organic compound (VOC) vapor intrusion."
Environmental Forensics 9: 107-114.
9. U.S. EPA (1993). A Physician's Guide to Radon. U.S. EPA Office of Air and Radiation,
EPA-402-K-93-008, Washington, DC, September 1993. Available at
http://www.epa.gov/radon/pubs/physic.html, accessed December 11, 2011.
10. Battelle, Quality Assurance Project Plan for Verification of Building Pressure Control for the
Assessment of Vapor Intrusion. Prepared by Battelle, Columbus, Ohio, October 1, 2010.
Available at http://www.epa.gov/nrmrl/pubs/600rl0157/600rl0157.pdf, accessed December
11,2011.
11. Battelle, Quality Management Plan for the ETV Advanced Monitoring Systems Center,
Version 7.0, U.S. EPA Environmental Technology Verification Program. Prepared by
Battelle, Columbus, Ohio, November 2008.
12. Mosley, R.B., D. Greenwell, and C. Lutes (2010). "Use of Integrated Indoor Concentrations
of Tracer Gases and Volatile Organic Compounds to Distinguish Soil Sources from Above-
Ground Sources." Poster #968, Presented at Seventh International Conference, Remediation
of Chlorinated and Recalcitrant Compounds, Monterey, CA, May 24-27. Record ID: 220117.
13. CRC Handbook of Chemistry and Physics, 85th ed., David R. Lide, editor. Boca Raton: CRC
Press, 2004.
14. Brenner, D. (2010). "Results of a Long-Term Study of Vapor Intrusion at Four Large
Buildings at the NASA Ames Research Center." Journal of the Air & Waste Management
Association 60: 747-758.
15. U.S. EPA Compendium Method TO-15, "Determination of Volatile Organic Compounds
(VOCs) In Air Collected In Specially-Prepared Canisters and Analyzed By Gas
Chromatography/Mass Spectrometry (GC/MS)." Second edition. Available at
http://www.epa.gov/ttnamtil/files/ambient/airtox/to-15r.pdf, accessed December 11, 2011.
16. NIOSH Method 6602, "Sulfur hexafluoride by portable GC" Available at
http://www.cdc.gov/niosh/docs/2003-154/pdfs/6602.pdf, accessed December 11, 2011.
17. U.S. EPA (1992). Indoor Radon and Radon Decay Product Measurement Device Protocols.
Washington, DC: U.S. EPA Office of Air and Radiation, 402-R-92-004, July 1992.
Available at http://www.epa.gov/radon/pubs/devprotl.html, accessed December 11, 2011.
18. ITRC (Interstate Technology & Regulatory Council) (2010). Use and Measurement of Mass
Flux and Mass Discharge. MASSFLUX-1. Washington, D.C.: Interstate Technology &
Regulatory Council, Integrated DNAPL Site Strategy Team.
http://www.itrcweb.org/Documents/MASSFLUX1.pdf, accessed December 11, 2011.
70
-------
19. Nazaroff, W.W., H. Feustel, A.V. Nero, K.L. Revzan, D.T. Grimsrud, M.A. Essling, R.E.
Toohey (1985). "Radon transport into a detached one-story house with a basement."
Atmospheric Environment 19: 31-46.
20. U.S. EPA (1997). National Radon Proficiency Program Guidance on Quality Assurance.
Montgomery, AL: National Air and Radiation Environmental Laboratory, 401-R-95-012,
October 1997. http://www.nrsb.org/images/file/OualityAssurranceProgram.pdf, accessed
December 11,2011.
71
-------
Appendix A
Supplemental Information
72
-------
18'
18'
GARAGE
DOWNSTAIRS
Figure Al. Floorplan for ASU VI Research House. Shown are the locations for IA, AA,
and SS sampling, cross-foundation and building (I/O) differential pressure measurements,
the release point for the SFe tracer gas, and placement of the fan for pressure control.
Dimension are of the building interior and are in feet (') and inches ("). Ceiling height
upstairs is 8', downstairs is 7'6". Figures courtesy of GSI.
73
-------
MOFFETT BUILDING 107
54' 6"
Figure A2. Floorplan for Moffett Field Building 107. Shown are the locations for IA, AA
and SS sampling, cross-foundation and building (indoor/ourdoor) differential pressure
measurements, the release point for the SFe tracer gas, and placement of the fan for
pressure control. Dimension are of the building interior and are in feet (') and inches (").
Ceiling height 7'9". Figure courtesy of GSI.
74
-------
20
18
16
14
_
Q.
Q.
12
10
OJ
u
• TCE
*1,1-DCE
Baseline
Negative Pressure
Positive Pressure
'
1.8
1.6
Q.
Q.
1.2 o
1 OJ
u
E
O
U
0.8
0.6
0.4
0.2
Figure A3. Real-time indoor air data from the HAPSITE at ASU House.
75
-------
ASU-Q_ivs. lambda'V
(Assumption 6)
20
40
60
Q i
100
120
Figure A4. Graphical presentation of validation of Mosley Model assumptions:
compared to XV at ASU House.
Moffett - Q_i vs. lambda*V
(Assumption 6)
w ^
3 sd below mean
lambda*V est
3 sd lambda'V
50
100
Q i
150
200
Figure A5. Graphical presentation of validation of Mosley Model assumptions:
compared to XV at Moffett Field.
76
-------
ASU -Q_i_neg vs. lambda*V
(Assumption 7)
i 9
3 sd below mean
lambda*V est
(under green)
3 sd lambda'V
-1000
-500
500
1000
1500
Q i
Figure A6. Graphical presentation of validation of Mosley Model assumptions: Qf
compared to XV at ASU House.
in
8
Moffett - Q_i_neg vs. lambda*V
(Assumption 7)
3 sd below mean
lambda'V est
(under green)
3 sd lambda'V
-500
500
1000
1500
Q i
Figure A7. Graphical presentation of validation of Mosley Model assumptions:
compared to XV at Moffett Field.
77
-------
ASU - Q_i_pos vs. lambda*V
(Assumption 8)
3 sd below mean
lambda'V est
(under green)
3 sd lambda'V
-1000
-500
500
1000
1500
Q i
Figure A8. Graphical presentation of validation of Mosley Model assumptions:
compared to XV at ASU House.
Moffett - Q_i_pos vs. lambda*V
(Assumption 8)
» 9
3 sd below mean
lambda'V est
(under green)
3 sd lambda'V
-1000
1000
2000
Q i
Figure A9. Graphical presentation of validation of Mosley Model assumptions:
compared to XV at Moffett Field.
78
-------
Table Al. Mean subslab pressure differentials (in Pa) at ASU House and Moffett Field.
Also presented are standard deviations, number of observations (N), and standard
deviation of the mean. Pressures have been corrected to account for the reference port
being open to the indoor atmosphere.
ASU House
Test Mean Std Dev N Stddev/VN
BL
NP
PP
0.35
-2.12
2.56
0.16 290
0.25 290
0.27 289
0.009
0.015
0.016
Moffett Field
Test
BL
NP
PP
Mean
-0.24
-1.21
0.47
Std Dev
0.45
0.24
0.22
N
278
275
256
Std dev/VN
0.027
0.015
0.014
Tables A2 through A8, presented on the following pages, provide the raw IA, AA, and SS data
generated during the testing at both buildings, as well as quantities calculated using and derived
from these raw data. Data are organized and presented in tabular format according to which
verification parameter or assumption required the use of which raw data. Names of specific field
samples are in the nomenclature described in the Data Collection Forms in Appendix D of the
QAPP, such as 1-PP-IA-VOC-l to indicate the positive pressure IA sample for VOCs and SFe
collected at IA-1, either at ASU House or Moffett Field. The system of nomenclature in the
table headers generally follows that presented in Table 1 of Chapter 3, with the exception that
subscripts have been replaced by underscores followed by normal text, and superscripts have
simply become normal text: for instance T;+ is now T_i+. Note that quantities presented using
nomenclature from Table 1 are generally derived from several other data; for instance, T_i+ is
the arithmetic mean of three SF6 measurements: 1-PP-IA-VOC-l, l-PP-IA-VOC-2, and 1-PP-
IA-VOC-3. Moreover, tags such as std_dev_, delta_, _MDL, _AVG, _SF_6, _Rn, _TCE, _DCE,
_PCE, _Ben, _Tol. These tag names stand for, respectively, standard deviation, estimated error
(according to principles in Appendix C of the QAPP), estimated method detection limit,
arithmetic mean, SFe, radon, trichloroethylene, 1,1-dichloroethylene, tetrachloroethylene,
benzene, and toluene, have been appended to names, as appropriate, to clarify the identity of the
quantify presented. Additional information is also provided under the tables, including a
roadmap of variable names, units for all results, references to concentration data that were less
than MDLs, and to data impacted by various quality control issues. Finally, for calculation of
the statistics and results of hypothesis testing presented in Chapter 6 of the present report, much
of the raw data were first transformed onto the log scale; as such, derived quantities such as
standard deviations and error estimates will not match those presented in Tables A2 through A8.
79
-------
Table A2. Data used in the calculation of VI enhancement and reduction
Pressure
BL
NP
PP
Pressure
BL
NP
PP
Variable names
Baseline
Negative Pressure
Positive Pressure
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
Q T
4.755E-05
5.022E-05
5.301E-05
Q T
5.83E-05
5.60E-05
5.64E-05
Q_T
Q_T-
Q_T+
delta Q T
10%
10%
10%
delta Q T
10%
10%
10%
delta_Q_T
delta_Q_T-
delta_Q_T+
IA-1 SF 6
4,000
360
82
IA-1 SF 6
3,000
310
790
1-BL-IA-VOC-l
1-NP-IA-VOC-l
1-PP-IA-VOC-l
IA-1 SF 6 MDL
120
21
9.9
IA-1 SF 6 MDL
91
8.8
11
1-BL-IA-VOC-1_MDL
1-NP-IA-VOC-1_MDL
1-PP-IA-VOC-1_MDL
IA-2 SF 6
5,100
380
820
IA-2 SF 6
3,400
600
790
l-BL-IA-VOC-2
l-NP-IA-VOC-2
l-PP-IA-VOC-2
IA-2 SF 6 MDL
96
19
11
IA-2 SF 6 MDL
88
9.8
9.6
1 -BL-IA-VOC-2_MDL
1 -NP-IA-VOC-2_MDL
1-PP-IA-VOC-2_MDL
oo
o
Units
pCi/h none
reported value < MDL
italics: data used in calcs
orange text: ADQ observations
bold text: deviation 5
m3/h
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A2. Data used in the calculation of VI enhancement and reduction (Continued)
Pressure
BL
NP
PP
Pressure
BL
NP
PP
Variable names
Baseline
Negative Pressure
Positive Pressure
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
IA-3 SF 6
6,500
1,600
1,700
IA-3 SF 6
3,500
780
0
l-BL-IA-VOC-3
l-NP-IA-VOC-3
l-PP-IA-VOC-3
IA-3 SF 6 MDL
910
22
93
IA-3 SF 6 MDL
90
8.8
10
1 -BL-IA-VOC-3_MDL
1 -NP-IA-VOC-3_MDL
1-PP-IA-VOC-3_MDL
T i
5,200
780
867
T i
3,300
563
530
T_i
T_i-
T_i+
std dev T i
1,253
710
810
std dev T i
265
237
450
std_dev_T_i
std_dev_T_i-
std_dev_T_i+
C T
5.96E+09
5.96E+09
5.96E+09
C T
5.96E+09
5.96E+09
5.96E+09
C_T
C_T-
C_T+
delta C T
0.2%
0.2%
0.2%
delta C T
0.2%
0.2%
0.2%
delta_C_T
delta_C_T-
delta_C_T+
V
273.5
273.5
273.5
V
364.8
364.8
364.8
V
V
V
Units
pCi/h none
reported value < MDL
italics: data used in calcs
orange text: ADQ observations
bold text: deviation 5
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A2. Data used in the calculation of VI enhancement and reduction (Continued)
Pressure
BL
NP
PP
Pressure
BL
NP
PP
Variable names
Baseline
Negative Pressure
Positive Pressure
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
delta V
30%
30%
30%
delta V
30%
30%
30%
delta_V
delta_V
delta_V
Q i
54
384
364
Q i
105
592
634
Q i
Q i-
Q_i+
delta Q i
14
351
342
delta Q i
13
256
542
delta_Q_i
delta_Q_i-
delta_Q_i+
IA-1 Rn
0.277
1.103
0.017
IA-1 Rn
1.008
0.520
0.346
1-BL-IA-Rn-l
1-NP-IA-Rn-l
1-PP-IA-Rn-l
IA-1 Rn MDL
0.244
0.478
0.095
IA-1 Rn MDL
0.284
0.211
0.247
l-BL-IA-Rn-l_MDL
l-NP-IA-Rn-l_MDL
l-PP-IA-Rn-l_MDL
IA-2 Rn
0.326
4.124
0.153
IA-2 Rn
0.921
0.530
0.372
l-BL-IA-Rn-2
l-NP-IA-Rn-2
l-PP-IA-Rn-2
IA-2 Rn MDL
0.438
0.863
0.057
IA-2 Rn MDL
0.327
0.334
0.206
l-BL-IA-Rn-2_MDL
l-NP-IA-Rn-2_MDL
l-PP-IA-Rn-2_MDL
oo
to
Units
pCi/h none
reported value < MDL
italics: data used in calcs
orange text: ADQ observations
bold text: deviation 5
m3/h
m3/h
pCi/L
pCi/L
pCi/L
pCi/L
-------
Table A2. Data used in the calculation of VI enhancement and reduction (Continued)
Pressure
BL
NP
PP
Pressure
BL
NP
PP
Variable names
Baseline
Negative Pressure
Positive Pressure
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
IA-3 Rn
0.533
0.376
0.104
IA-3 Rn
0.955
0.800
0.189
l-BL-IA-Rn-3
l-NP-IA-Rn-3
l-PP-IA-Rn-3
IA-3 Rn MDL
0.295
0.189
0.072
IA-3 Rn MDL
0.228
0.336
0.196
l-BL-IA-Rn-3_MDL
l-NP-IA-Rn-3_MDL
l-PP-IA-Rn-3_MDL
R i
377
1868
91
R i
961
617
302
R_i
R_i-
R_i+
std dev R i
138
1987
69
std dev R i
44
159
99
std_dev_R_i
std_dev_R_i-
std_dev_R_i+
R a
104
26
69
R a
777
256
328
R_a
R_a-
R_a+
R a MDL
289
220
55
R a MDL
175
237
165
R_a_MDL
R_a-_MDL
R_a+_MDL
delta R a
38
28
52
delta R a
8
66
107
delta_R_a
delta_R_a-
delta_R_a+
oo
OJ
Units
pCi/h none
reported value < MDL
italics: data used in calcs
orange text: ADQ observations
bold text: deviation 5
pCi/L
pCi/L
pCi/m3
pCi/m3
pCi/m3
pCi/m3
pCi/m3
-------
Table A2. Data used in the calculation of VI enhancement and reduction (Continued)
Pressure
BL
NP
PP
Pressure
BL
NP
PP
Variable names
Baseline
Negative Pressure
Positive Pressure
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
R i-R a
272
1841
22
R i-R a
784
360
-25
R_i-R_a
R_i»R_a-
R_i+-R_a+
delta (R i-R a)
143
1987
86
delta (R i-R a)
45
172
146
delta_(R_i*R_i)
delta_(R_i-*R_i-)
delta_(R_i+*R_i+)
Q*(R i-R a)
14830
706231
8143
Q*(R i-R a)
82556
213492
-16153
Q_i*R_i
Q_i-*R_i-
Q_i+*R_i+
delta (Q i*(R i-R a)]
8704
999852
32347
delta [Q i*(R i-R a)]
11579
137656
93610
delta_[Q_i*(R_i-R_a)]
delta_[Q_i-*(R_i--R_a-)]
delta_[Q_i+*(R_i+-R_a+)]
Units
pCi/h
reported value < MDL
italics: data used in calcs
orange text: ADQ observations
bold text: deviation 5
none
pCi/m3
pCi/m3
pCi/h
pCi/h
-------
Table A3. Data used for FVI calculations
F i
F i
Calculation
F vi-
F vi+
vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Q i
54
384
364
Q i
105
592
634
delta Q i
14
351
342
delta Q i
13
256
542
R i
377
1868
91
R i
961
617
302
std dev R i
138
1987
69
std dev R i
44
159
99
R a
104
26
69
R a
777
256
328
R a MDL
289
220
55
R a MDL
175
237
165
delta R a
38
28
52
delta R a
8
66
107
R i/R a
3.6
70.5
1.3
R i/R a
5.4
2.4
0.9
Roadmap of variable names
F_vi-
F_vi+
Building
Building
Q_i
Q_i-
delta_Q_i
delta_Q_i-
R_i
R_i-
std_dev_R_i
std_dev_R_i-
R_a
R_a-
R_a_MDL
R_a-_MDL
delta_R_a
delta_R_a-
R_i/R_a
R_i-/R_a-
F_vi+_(R_i+=R_a+)
Building
delta_Q_i+
R
std dev R i+
R a+
R a+ MDL
delta R a+ R i+/R a+
Units
none none
reported value < MDL
italicized text: data used in coles
ilem with model calculat
bold text: deviation 5
m3/h
m3/h
pCi/m3
pCi/m3
pCi/m3
pCi/m3
pCi/m3
unitless
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F_vi-
F_vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in coles
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
none
TCE=trichloroethene
IA-1 TCE
18
4
0.11
IA-1 TCE
4.7
2.3
0.12
1-BL-IA-VOC-l
1-NP-IA-VOC-l
1-PP-IA-VOC-l
ug/m3
IA-1 TCE MDL
0.05
0.044
0.041
IA-1 TCE MDL
0.038
0.037
0.045
1-BL-IA-VOC-1_MDL
1-NP-IA-VOC-1_MDL
1-PP-IA-VOC-1_MDL
ug/m3
IA-2 TCE
22
16
0.15
IA-2 TCE
5
3
0.1
l-BL-IA-VOC-2
l-NP-IA-VOC-2
l-PP-IA-VOC-2
ug/m3
IA-2 TCE MDL
0.04
0.04
0.047
IA-2 TCE MDL
0.037
0.041
0.040
1 -BL-I A- VOC-2_MDL
1-NP-IA-VOC-2_MDL
1-PP-IA-VOC-2_MDL
ug/m3
IA-3 TCE
16
8.4
0.18
IA-3 TCE
5.7
4.1
0.13
l-BL-IA-VOC-3
l-NP-IA-VOC-3
l-PP-IA-VOC-3
ug/m3
ilem with model calculation
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F_vi-
F_vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
IA-3 TCE MDL
0.038
0.047
0.039
IA-3 TCE MDL
0.038
0.037
0.043
1 -BL-IA-VOC-3_MDL
1-NP-IA-VOC-3_MDL
C i TCE
18.7
9.5
0.15
C i TCE
4.9
3.1
0.12
C_i
C_i-
std dev C i TCE
3.1
6.1
0.04
std dev C i TCE
0.2
0.9
0.02
std_dev_C_i
std_dev_C_i-
C a TCE
0.17
0.15
0.084
C a TCE
0.084
0.12
0.089
C_a
C_a-
C a TCE MDL
0.037
0.039
0.042
C a TCE MDL
0.037
0.033
0.038
C_a_MDL
C_a-_MDL
delta C a TCE
30%
30%
30%
delta C a TCE
30%
30%
30%
delta_C_a
delta_C_a-
F_vi+_(R_i+=R_a+)
Building
l-PP-IA-VOC-3 MDL
std dev C i+
C a+
C a+ MDL
delta C a+
Units
none none
reported value < MDL
italicized text: data used in coles
ilem with model calculati
bold text: deviation 5
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A3. Data used for FVI calculations (Continued)
oo
oo
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F_vi-
F_vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
C i/C a TCE
110
63
1.7
C i/C a TCE
59
26
1.3
C_i/C_a
C_i-/C_a-
Q i-*(C i-C a-) TCE
3574
none
none
Q i-*(C i-C a-) TCE
1785
none
none
none
Q_i-*(C_i--C_a-)
Q i*(C i-C a) TCE
1008
none
none
Q i*(C i-C a) TCE
511
none
none
Q_i*(C_i-C_a)
none
Q i+*(C i+-C a+) TCE
23
none
none
Q i+*(C i+-C a+) TCE
18
none
none
Q_i+*(C_i+-C_a+)
none
Q i-*(C i-C a-)-
Q i*(C i-C a) TCE
2566
none
none
Q i-*(C i-C a-)-
Q i*(C i-C a) TCE
1274
none
none
Q i-*(C i-C a-)-
Q_i*(C_i-C_a)
none
F_vi+_(R_i+=R_a+)
Building
C i+/C a+
none
none
none
none
Units
none none
reported value < MDL
italicized text: data used in coles
ilem with model calculation
bold text: deviation 5
Unitless
ug/h
ug/h
ug/h
ug/h
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F_vi-
F_vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Q i*(C i-C a)-
Q i+*(C i+-C a+) TCE
985
None
None
Q i*(C i-C a)-
Q i+*(C i+-C a+) TCE
493
None
None
Q_i*(C_i-C_a)-
None
Q i*(R i-R a)
14830
none
none
Q i*(R i-R a)
82556
none
none
Q_i*(R_i-R_a)
none
Q i-*(R i-R a-)
706231
none
none
Q i-*(R i~R a-)
213492
none
none
(Li-*(RJ-R_a-)
none
Q i+*(R i+-R a+)
8143
none
none
Q i+*(R i+-R a+)
^fm^
none
none
Q_1+*(R_1+-R_a+)
none
Q i*(R i-R a)/[Q i-*(R i-
R a-)-Q i*(R i-R a)]
0.021
none
none
Q i*(R i-R a)/[Q i-*(R i-
R a-)-Q i*(R i-R a)]
| 0.631
none
none
Q i*(R i-R a)/[Q i-*(R i-
R_a-)-Q_i*(R_i-R_a)]
none
F_vi+_(R_i+=R_a+)
Building
None
none
none
none
none
Units
none none
reported value < MDL
italicized text: data used in coles
ilem with model calculation
bold text: deviation 5
ug/h
pCi/h
pCi/h
pCi/h
none
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
Building
Q i*(R i-R a)/[Q i*(R i-
R a)-Q i+*(R i+-R a
Q i*C i TCE
E C TCE
F a TCE
F in TCE
F vi TCE
F_vi-
F vi+
F_vi+_(R_i+=R_a
ASU
ASU
ASU
2.22
None
None
1017
none
none
55
2184
985
0.01
0.01
0.01
Calculation
F_vi-
F_vi+
F vi+ (R i+=R
Building
Moffett
Moffett
Moffett
Q_i*(R_i-R_a)/[Q_i*(R_i-
R_a)-Q_i+*(Ri+-F
1B^
None
None
i*C i TCE
519
none
none
E C TCE
804
493
F a TCE
0.02
0.02
0.02
F in TCE
0.19
0.03
F vi TCE
Roadmap of variable names
F_vi- Building
F_vi+ Building
F_vi+_(R_i+=R_a+) Building
Units
none none
reported value < MDL
italicized text: data used in coles
Problem with model calculation
bold text: deviation 5
None
None
None
Q_i*C_i E_C- F_a
none E_C+ none
none E_C_(R_i+=R_a+) none
ug/h
ug/h
none
F_in-
F in+
F_vi-
F vi+
F_in_(R_i+=R_a+) F_vi+_(R_i+=R_a+)
none
none
-------
Table A3. Data used for FVI calculations (Continued)
DCE= 1 , 1 -dichloroethene
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
IA-1 DCE
a 12
2.3
0
IA-1 DCE
N/A
N/A
N/A
IA-1 DCE MDL
0.05
0.044
0.041
IA-1 DCE MDL
N/A
N/A
N/A
IA-2 DCE
0.13
11
0
IA-2 DCE
N/A
N/A
N/A
IA-2 DCE MDL
0.04
0.04
0.047
IA-2 DCE MDL
N/A
N/A
N/A
IA-3 DCE
0.12
5.4
0
IA-3 DCE
N/A
N/A
N/A
Roadmap of variable names
F vi-
F_vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in
Building
Building
Building
none
coles
1-BL-IA-VOC-l
1-NP-IA-VOC-l
1-PP-IA-VOC-l
ug/m3
1-BL-IA-VOC-1_MDL
1-NP-IA-VOC-1_MDL
1-PP-IA-VOC-1_MDL
ug/m3
l-BL-IA-VOC-2
l-NP-IA-VOC-2
l-PP-IA-VOC-2
ug/m3
1 -BL-I A- VOC-2_MDL
1-NP-IA-VOC-2_MDL
1-PP-IA-VOC-2_MDL
ug/m3
l-BL-IA-VOC-3
l-NP-IA-VOC-3
l-PP-IA-VOC-3
ug/m3
ilem with model calculation
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
to
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F vi-
F_vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
IA-3 DCE MDL
0.038
0.047
0.039
IA-3 DCE MDL
N/A
N/A
N/A
1 -BL-IA-VOC-3_MDL
1-NP-IA-VOC-3_MDL
C i DCE
0.12
6.2
0.0423
C i DCE
N/A
N/A
N/A
C_i
C_i-
std dev C i DCE
0.006
4.4
0.004
std dev C i DCE
N/A
N/A
N/A
std_dev_C_i
std_dev_C_i-
C a DCE
0
0
0
C a DCE
N/A
N/A
N/A
C_a
C_a-
C a DCE MDL
0.037
0.03P
0.042
C a DCE MDL
N/A
N/A
N/A
C_a_MDL
C_a-_MDL
delta C a DCE
30%
30%
30%
delta C a DCE
N/A
N/A
N/A
delta_C_a
delta_C_a-
F_vi+_(R_i+=R_a+)
Building
1-PP-IA-VOC-3_MDL
std_dev_C_i+
C_a+
C_a+_MDL
delta_C_a+
Units
none none
reported value < MDL
italicized text: data used in coles
ilem with model calc
bold text: deviation 5
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F vi-
F_vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
C i/C a DCE
3.3
160
1.0
C i/C a DCE
N/A
N/A
N/A
C_i/C_a
C_i-/C_a-
Q i-*(C i-C a-) DCE
2376
none
none
Q i-*(C i-C a-) DCE
N/A
N/A
N/A
none
Q_i-*(C_i»C_a-)
Q i*(C i-C a) DCE
4.7
none
none
Q i*(C i-C a) DCE
N/A
N/A
N/A
Q_i*(C_i-C_a)
none
Q i+*(C i+-
C a+) DCE
0.1
none
none
Q i+*(C i+-
C a+) DCE
N/A
N/A
N/A
Q_1+*(C_1+-C_a+)
none
Q i-*(C i-C a-)-
Q i*(C i-C a) DCE
2371
none
none
Q i-*(C i-C a-)-
Q i*(C i-C a) DCE
N/A
N/A
N/A
Q i-*(C i-C a-)-
Q_i*(C_i-C_a)
none
F_vi+_(R_i+=R_a+)
Building
C i+/C a+
none
none
none
none
Units
none none
reported value < MDL
italicized text: data used in coles
ilem with model calc
bold text: deviation 5
Unitless
ug/h
ug/h
ug/h
jig/h
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F_vi-
F_vi+
F_vi+_(R_i+=R_a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
Q i*(C i-C a)-
Q i+*(C i+-
C a+) DCE
4.6
None
None
Q i*(C i-C a)-
Q i+*(C i+-
C a+) DCE
N/A
N/A
N/A
Q_i*(C_i-C_a)-
Q i+*(C i+-C a+)
None
None
Q i*(R i-R a)
14830
none
none
Q i*(R i-R a)
N/A
N/A
N/A
Q_i*(R_i-R_a)
none
none
Q i-*(R i-R a-)
706231
none
none
Q i-*(R i-R a-)
N/A
N/A
N/A
(U-*(R_i-R_a-)
none
none
Q i+*(R i+-R a+)
8143
none
none
Q i+*(R i+-R a+)
N/A
N/A
N/A
Q_1+*(R_1+-R_a+)
none
none
Q i*(R i-R a)/[Q i-*(R i-
R a-)-Q i*(R i-R a)]
0.021
none
none
Q i*(R i-R a)/[Q i-*(R i-
R a-)-Q i*(R i-R a)]
N/A
N/A
N/A
Q i*(R i-R a)/[Q i-*(R i-
R_a-)-Q_i*(R_i-R_a)]
none
none
Units
none none
reported value < MDL
italicized text: data used in coles
ug/h
pCi/h
pCi/h
pCi/h
none
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F_vi+_(R_i+=R_a+)
Roadmap of variable names
F vi-
F_vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
Q i+*(R i+-R a+)]
2.22
none
none
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
Q i+*(R i+-R a+)]
N/A
N/A
N/A
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
Q_i+*(R_i+-R_a+)]
none
Q i*C i DCE
6.7
none
none
Q i*C i DCE
N/A
N/A
N/A
Q_i*C_i
none
E C DCE
51
10
5
E C DCE
N/A
N/A
N/A
E_C-
E_C+
F a DCE
0.30
0.30
0.30
F a DCE
N/A
N/A
N/A
F_a
none
F in DCE
-6.87
_A O1
0.02
F in DCE
N/A
N/A
N/A
F in-
F_in+
F_vi+_(R_i+=R_a+)
Building
none
none
E_C_(R_i+=R_a+)
none
F_in_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in calcs
Problem with model calculation
bold text: deviation 5
none
none
ug/h
ug/h
none
none
-------
Table A3. Data used for FVI calculations (Continued)
PCE=tetrachloroethene
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
F vi DCE
7.57
1.51
0.68
F vi DCE
N/A
N/A
N/A
IA-1 PCE
N/A
N/A
N/A
IA-1 PCE
2.7
1.5
0.44
IA-1 PCE MDL
N/A
N/A
N/A
IA-1 PCE MDL
0.038
0.037
0.045
IA-2 PCE
N/A
N/A
N/A
IA-2 PCE
2.9
1.9
0.41
IA-2 PCE MDL
N/A
N/A
N/A
IA-2 PCE MDL
0.037
0.041
0.040
Roadmap of variable names
F_vi-
F_vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in
Building
Building
Building
none
coles
.
F_vi-
F_vi+
F_vi+_(R_i+=R_a+)
none
1-BL-IA-VOC-l
1-NP-IA-VOC-l
1-PP-IA-VOC-l
ug/m3
1-BL-IA-VOC-1_MDL
1-NP-IA-VOC-1_MDL
1-PP-IA-VOC-1_MDL
ug/m3
l-BL-IA-VOC-2
l-NP-IA-VOC-2
l-PP-IA-VOC-2
ug/m3
1 -BL-I A- VOC-2_MDL
1-NP-IA-VOC-2_MDL
1-PP-IA-VOC-2_MDL
ug/m3
ilem with model calculati
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F_vi-
F_vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
IA-3 PCE
N/A
N/A
N/A
IA-3 PCE
2.9
2.6
0.41
l-BL-IA-VOC-3
l-NP-IA-VOC-3
IA-3 PCE MDL
N/A
N/A
N/A
IA-3 PCE MDL
0.038
0.037
0.043
1 -BL-IA-VOC-3_MDL
1-NP-IA-VOC-3_MDL
C i PCE
N/A
N/A
N/A
C i PCE
2.8
2.0
0.42
C_i
C_i-
std dev C i PCE
N/A
N/A
N/A
std dev C i PCE
0.12
0.56
0.02
std_dev_C_i
std_dev_C_i-
C a PCE
N/A
N/A
N/A
C a PCE
0.12
0.21
0.89
C_a
C_a-
C a PCE MDL
N/A
N/A
N/A
C a PCE MDL
0.037
0.033
0.038
C_a_MDL
C_a-_MDL
F_vi+_(R_i+=R_a+)
Building
l-PP-IA-VOC-3 l-PP-IA-VOC-3 MDL
C i
std dev C i+
C a+
C a+ MDL
Units
none none
reported value < MDL
italicized text: data used in coles
ilem with model calculation
bold text: deviation 5
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
delta C a PCE
N/A
N/A
N/A
delta C a PCE
30%
30%
30%
C i/C a PCE
N/A
N/A
N/A
C i/C a PCE
24
10
0.47
Q i-*(C i-C a-) PCE
N/A
N/A
N/A
Q i-*(C i-C a-) PCE
1060
none
none
Q i*(C i-C a) PCE
N/A
N/A
N/A
Q i*(C i-C a) PCE
286
none
none
Q i+*(C i+-C a+) PCE
N/A
N/A
N/A
Q i+*(C i+-C a+) PCE
•••^H
none
none
Roadmap of variable names
F_vi-
F vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in
Building
Building
Building
none
coles
delta_C_a
delta_C_a-
delta_C_a+
%
C_i/C_a
C_i-/C_a-
C_i+/C_a+
unitless
none
Q_i-*(C_i~C_a-)
none
ug/h
Q_i*(C_i-C_a)
none
none
ug/h
Q_1+*(C_1+-C_a+)
none
none
jig/h
ilem with model calculation
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F vi-
F vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Q i-*(C i-C a-)-
Q i*(C i-C a) PCE
N/A
N/A
N/A
Q i-*(C i-C a-)-
Q i*(C i-C a) PCE
775
none
none
Q i-*(C i-C a-)-
Q_i*(C_i-C_a)
none
Q i*(C i-C a)-
Q i+*(C i+-C a+) PCE
N/A
N/A
N/A
Q i*(C i-C a)-
Q i+*(C i+-C a+) PCE
^•••^H
none
none
Q_i*(C_i-C_a)-
none
Q i*(R i-R a)
N/A
N/A
N/A
Q i*(R i-R a)
| 82556
none
none
Q_i*(R_i-R_a)
none
Q i-*(R i~R a-)
N/A
N/A
N/A
Q i-*(R i~R a-)
213492
none
none
(Li-*(RJ-R_a-)
none
Q i+*(R i+-R a+)
N/A
N/A
N/A
Q i+*(R i+-R a+)
^•-16153^^1
none
none
Q_1+*(R_1+-R_a+)
none
F_vi+_(R_i+=R_a+)
Building
none
none
none
none
none
Units
none
reported value < MDL
italicized text: data used in coles
none
ilem with model calculation
bold text: deviation 5
ug/h
ug/h
pCi/h
pCi/h
pCi/h
-------
Table A3. Data used for FVI calculations (Continued)
o
o
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F_vi-
F_vi+
F vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Q i*(R i-R a)/[Q i-*(R i-R a-)-
Q i*(R i-R a)]
N/A
N/A
N/A
Q i*(R i-R a)/[Q i-*(R i~R a-)-
Q i*(R i-R a)]
0.631 |
none
none
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
Q i+*(R i+-R a+)]
N/A
N/A
N/A
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
Q i+*(R i+-R a+)]
^H 084 ^H
none
none
Q i*C i PCE
N/A
N/A
N/A
Q i*C i PCE
| 298
none
none
E C PCE
N/A
N/A
N/A
E C PCE
488
r488
584
Roadmap of variable names
F_vi-
F_vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in
Building
Building
Building
none
coles
Q_i*(R_i-R_a)]
None
none
none
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
none
none
none
Q_i*C_i
none
none
ug/h
E_C-
E_C+
E_C_(R_i+=R_a+)
Ug/h
ilem with model calculation
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F_vi-
F_vi+
F vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
F a PCE
N/A
N/A
N/A
F a PCE
0.04
0.04
0.04
Roadmap of variable names
F vi-
F_vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in
Building
Building
Building
none
coles
F_a
none
none
none
F in PCE
N/A
N/A
N/A
F in PCE
-0.68
-0.68
-1.00
F in-
F_in+
F_in_(R_i+=R_a+)
none
F vi PCE
N/A
N/A
N/A
F vi PCE
1.64
1.64
1.96
F_vi-
F_vi+
F_vi+_(R_i+=R_a+)
none
Ben=benzene
IA-1 Ben
0.47
0.44
0.56
IA-1 Ben
1.1
1.5
1.5
1-BL-IA-VOC-l
1-NP-IA-VOC-l
1-PP-IA-VOC-l
ug/m3
IA-1 Ben MDL
0.15
0.13
0.12
IA-1 Ben MDL
0.11
0.11
0.14
1-BL-IA-VOC-1_MDL
1-NP-IA-VOC-1_MDL
1-PP-IA-VOC-1_MDL
ug/m3
IA-2 Ben
0.46
0.47
0.57
IA-2 Ben
1.1
1.7
1.6
l-BL-IA-VOC-2
l-NP-IA-VOC-2
l-PP-IA-VOC-2
ug/m3
ilem with model calculation
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
o
to
F i
F i
Calculation
F vi-
F vi+
ri+ (R i+=R a+)
Calculation
F vi-
F_vi+
ri+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
IA-2 Ben MDL
0.12
0.12
0.14
IA-2 Ben MDL
0.11
0.12
0.12
IA-3 Ben
0.44
0.46
0.59
IA-3 Ben
1.1
1.5
1.5
IA-3 Ben MDL
0.11
0.14
0.12
IA-3 Ben MDL
0.11
0.11
0.13
C i Ben
0.46
0.46
0.57
C i Ben
1.1
1.6
1.5
std dev C i
0.015
0.015
0.015
std dev C i
0.00
0.12
0.06
Ben
Ben
Roadmap of variable names
F vi-
F_vi+
Building
Building
1 -BL-I A- VOC-2_MDL
1-NP-IA-VOC-2_MDL
l-BL-IA-VOC-3
l-NP-IA-VOC-3
1 -BL-IA-VOC-3_MDL
1-NP-IA-VOC-3_MDL
C_i
C_i-
std_dev_C
std_dev_C
i
i-
F_vi+_(R_i+=R_a+)
Building
l-PP-IA-VOC-2 MDL
l-PP-IA-VOC-3
l-PP-IA-VOC-3 MDL
C i
std dev C i+
Units
none
reported value < MDL
italicized text: data used in coles
none
ilem with model calculation
bold text: deviation 5
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A3. Data used for FVI calculations (Continued)
o
OJ
F i
F i
Calculation
F vi-
F vi+
vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
C a Ben
0.39
0.42
0.45
C a Ben
1.1
1.5
1.4
C a Ben MDL
0.11
0.12
0.13
C a Ben MDL
0.11
0.1
0.11
delta C a Ben
30%
30%
30%
delta C a Ben
30%
30%
30%
C i/C a Ben
1.2
1.1
1.3
C i/C a Ben
1.0
1.0
1.1
Q i-*(C i-C a-)
14
none
none
Q i-*(C i-C a-)
39
none
none
Ben
Ben
Roadmap of variable names
F vi-
F_vi+
Building
Building
C_a
C_a-
C_a_MDL
C_a-_MDL
delta_C_a
delta_C_a-
C_i/C_a
C_i-/C_a-
none
Q_i-*(C_i-C_£
i-)
F_vi+_(R_i+=R_a+)
Building
C a+
C a+ MDL
delta C a+
C i+/C a+
none
Units
none
reported value < MDL
italicized text: data used in coles
none
ilem with model calculation
bold text: deviation 5
ug/m3
ug/m3
unitless
ug/h
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F_vi-
F vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Q i*(C i-C a) Ben
4
None
None
Q i*(C i-C a) Ben
0.0
None
None
Q_i*(C_i-C_a)
None
Q i+*(C i+-
C a+) Ben
45
none
none
C~a+) Ben
85
none
none
Q_1+*(C_1+-C_a+)
none
Q i-*(C i-C a-)-
Q i*(C i-C a) Ben
10
none
none
Q i-*(C i-C a-)-
Q i*(C i-C a) Ben
39
none
none
Q i-*(C i-C a-)-
Q_i*(C_i-C_a)
none
Q i*(C i-C a)-Q i+*(C i+-
C_a+)_Ben
none
none
Q i*(C i-C a)-Q i+*(C i+-
C a+) Ben
-85
none
none
Q i*(C i-C a)-Q i+*(C i+-
C_a+)
none
Q i*(R i-
R a)
14830
none
none
Q i*(R i-
R a)
82556
none
none
Q i*(R i-
R_a)
none
F_vi+_(R_i+=R_a+)
Building
None
none
none
none
none
Units
none none
reported value < MDL
italicized text: data used in coles
ilem with model calculat
bold text: deviation 5
ug/h
ug/h
ug/h
ug/h
pCi/h
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Q i-*(R i-R a-) Q i+*(R i+-R a+)
706231 8143
None none
None none
Q i-*(R i-R a-) Q i+*(R i+-R a+)
213492 ^^|-16153 ^|
None none
None none
Q i*(R i-R a)/[Q i-*(R i-R a-)-
Q i*(R i-R a)]
0.021
none
none
Q i*(R i-R a)/[Q i-*(R i-R a-)-
Q i*(R i-R a)]
| 0.631 |
none
none
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
Q i+*(R i+-R a+)]
2.22
none
none
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
Q i+*(R i+-R a+)]
~^HHi^H~
none
none
Roadmap of variable names
F vi-
F_vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in
Building
Building
Building
none
coles
.
(U-.OU-R.a-) Q_1+*(R_1+-R_a+)
None none
None none
pCi/h pCi/h
Q i*(R i-R a)/[Q i-*(R i-R a-)-
Q_i*(R_i-R_a)]
none
none
none
Q_i*(R_i-R_a)/[Q_i*(R_i-R_a)-
none
none
none
ilem with model calculati
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
Tol=toluene
Calculation
Building
F_vi- ASU
F_vi+ ASU
F vi+ (R i+=R a+) ASU
Q_i*C_i_Ben
E C Ben
F a Ben
F in Ben
F vi Ben
IA-1 Tol
6.8
2.1
1.8
Calculation
Building
F_vi- Moffett
F_vi+ Moffett
F_vi+_(R_i+=R_a+) MofFett
Roadmap of variable names
Q_i*C_i_Ben
E C Ben
F a Ben
F in Ben
F vi Ben
IA-1 Tol
3.7
5.4
8.6
F_vi- Building
F_vi+ Building
F_vi+_(R_i+=R_a+) Building
Units
none none
reported value < MDL
italicized text: data used in calcs
Problem with model calculation
bold text: deviation 5
Q_i*C_i E_C- F_a
None E_C+ none
None E_C_(R_i+=R_a+) none
ug/h ug/h none
F_in-
F in+
F_vi-
F vi+
F_in_(R_i+=R_a+) F_vi+_(R_i+=R_a+)
none
none
1-BL-IA-VOC-l
1-NP-IA-VOC-l
1-PP-IA-VOC-l
ug/m3
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
IA-1 Tol MDL
0.2
0.18
0.17
IA-1 Tol MDL
0.15
0.15
0.18
IA-2 Tol
2.2
4.4
2.3
IA-2 Tol
4
8.3
8.4
IA-2 Tol MDL
0.16
0.16
0.19
IA-2 Tol MDL
0.150
0.160
0.160
IA-3 Tol
2 2
2.7
2.7
IA-3 Tol
4
15
8.5
IA-3 Tol MDL
0.15
0.19
0.16
IA-3 Tol MDL
0.15
0.15
0.17
Roadmap of variable names
F vi-
F_vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in
Building
Building
Building
none
coles
1-BL-IA-VOC-1_MDL
1-NP-IA-VOC-1_MDL
1-PP-IA-VOC-1_MDL
ug/m3
l-BL-IA-VOC-2
l-NP-IA-VOC-2
l-PP-IA-VOC-2
ug/m3
1 -BL-IA-VOC-2_MDL
1-NP-IA-VOC-2_MDL
1-PP-IA-VOC-2_MDL
ug/m3
l-BL-IA-VOC-3
l-NP-IA-VOC-3
l-PP-IA-VOC-3
ug/m3
1 -BL-IA-VOC-3_MDL
1-NP-IA-VOC-3_MDL
1-PP-IA-VOC-3_MDL
ug/m3
ilem with model calculation
bold text: deviation 5
-------
Table A3. Data used for FVI calculations (Continued)
o
oo
F i
F i
Calculation
F vi-
F vi+
vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
C i Tol
3.7
3.1
2.3
C i Tol
3.9
9.6
8.5
std dev C i Tol
2.7
1.2
0.45
std dev C i Tol
0.17
4.9
0.10
C a Tol
3.9
1.9
1.5
C a Tol
3.5
9.4
10
C a Tol MDL
0.15
0.16
0.17
C a Tol MDL
0.15
0.13
0.15
delta C a Tol
30%
30%
30%
delta C a Tol
30%
30%
30%
C i/C a Tol
1.0
1.6
1.5
C i/C a Tol
1.1
1.0
0.9
Roadmap of variable names
F vi-
F_vi+
Building
Building
C_i
C_i-
std_dev_C_i
std_dev_C_i-
C_a
C_a-
C_a_MDL
C_a-_MDL
delta_C_a
delta_C_a-
C_i/C_a
C_i-/C_a-
F_vi+_(R_i+=R_a+)
Building
std dev C i+
C a+
C a+ MDL
delta C a+
C i+/C a+
Units
none
reported value < MDL
italicized text: data used in coles
none
ilem with model calculation
bold text: deviation 5
ug/m3
ug/m3
ug/m3
ug/m3
unitless
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F_vi-
F_vi+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Q i-*(C i-C a-) Tol
448
none
none
Q i-*(C i-C a-) Tol
99
none
none
none
Q_i-*(C_i--C_a-)
Q_i*(C_i-C_a)_Tol
none
none
Q i*(C i-C a) Tol
42.1
none
none
Q_i*(C_i-C_a)
none
Q i+*(C i+-C a+) Tol
279
none
none
Q i+*(C i+-C a+) Tol
gci
none
none
Q_i+*(C_i+-C_a+)
none
Q i-*(C i-C a-)-
Q i*(C i-C a) Tol
457
none
none
Q i-*(C i-C a-)-
Q i*(C i-C a) Tol
57
none
none
Q i-*(C i-C a-)-
Q_i*(C_i-C_a)
none
F_vi+_(R_i+=R_a+)
Building
none
none
none
none
Units
none
reported value < MDL
italicized text: data used in coles
ilem with model calculation
bold text: deviation 5
none
ug/h
ug/h
ug/h
ug/h
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Roadmap of variable names
F vi-
F_vi+
F_vi+_(R_i+=R_a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
Q i*(C i-C a)-
Q_i+*(C_i+-C_a+)_Tol
None
None
Q_i*(C_i-C_a)-
^^^^ll^^^H
None
None
Q_i*(C_i-C_a)-
None
None
Q i*(R i-
R a)
14830
none
none
Q i*(R i-
R a)
82556
none
none
Q i*(R i-
R_a)
none
none
Q i-*(R i-R a-)
706231
none
none
Q i-*(R i-R a-)
213492 |
none
none
(U-*(R_i-R_a-)
none
none
Q i+*(R i+-R a+)
8143
none
none
Q i+*(R i+-R a+)
^••i^
none
none
Q_1+*(R_1+-R_a+)
none
none
Q i*(R i-R a)/[Q i-*(R i-
R a-)-Q i*(R i-R a)]
0.021
none
none
Q i*(R i-R a)/[Q i-*(R i-
R a-)-Q i*(R i-R a)]
| 0.631
none
none
Q i*(R i-R a)/[Q i-*(R i-
R_a-)-Q_i*(R_i-R_a)]
none
none
Units
none none
reported value < MDL
italicized text: data used in coles
Problem with model calculation
bold text: deviation 5
ug/h
pCi/h
pCi/h
pCi/h
none
-------
Table A3. Data used for FVI calculations (Continued)
Calculation
F vi-
F vi+
F vi+ (R i+=R a+)
Calculation
F vi-
F_vi+
F vi+ (R i+=R a+)
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Q_i*(R_i-R_a)/[Q_i*(R_i-
R a)-Q i+*(R i+-R a+)]
2.22
none
none
Q_i*(R_i-R_a)/[Q_i*(R_i-
R a)-Q i+*(R i+-R a+)]
^•••^H
none
none
Q i*C i Tol
203
none
none
Q i*C i Tol
| 411
none
none
E C Tol
10
-639
-288
E C Tol
36
831
993
F a Tol
1.04
1.04
1.04
F a Tol
0.90
0.90
0.90
F in Tol
-0.09
1.37
F in Tol
0.02
-1.92
-2.32
Roadmap of variable names
F_vi-
F vi+
F_vi+_(R_i+=R_a+)
Units
none
reported value < MDL
italicized text: data used in
Building
Building
Building
none
coles
^-Q^CR^Ca^)1]"
none
none
none
Q_i*C_i
none
none
ug/h
E_C-
E_C+
E_C_(R_i+=R_a+)
ug/h
F_a
none
none
none
F in-
F_in+
F_in_(R_i+=R_a+)
none
F vi Tol
0.05
^^^H
-1.42
F vi Tol
0.09
2.02
2.42
F vi-
F vi+
F_vi+_(R_i+=R_a+)
none
lem with model calculation
bold text: deviation 5
-------
Table A4. Data used to verify model assumption Cs = Cs
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
TCE=trichloroethene
SS-1 TCE
14
8.6
13
SS-1 TCE
3
1.9
0.73
1-BL-SS-VOC-l
1-NP-SS-VOC-l
1-PP-SS-VOC-l
SS-1 TCE MDL
0.35
0.35
0.39
SS-1 TCE MDL
0.73
0.4
0.42
1-BL-SS-VOC-1_MDL
1-NP-SS-VOC-1_MDL
1-PP-SS-VOC-1_MDL
SS-2 TCE
7
390
8.6
SS-2 TCE
0.83
1
0
l-BL-SS-VOC-2
l-NP-SS-VOC-2
l-PP-SS-VOC-2
SS-2 TCE MDL
0.38
0.4
0.39
SS-2 TCE MDL
0.37
0.41
0.41
1-BL-SS-VOC-2_MDL
1-NP-SS-VOC-2_MDL
1-PP-SS-VOC-2_MDL
SS-3 TCE
9.9
390
7.2
SS-3 TCE
2.2
2.5
2.3
l-BL-SS-VOC-3
l-NP-SS-VOC-3
l-PP-SS-VOC-3
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
excluding SS-1
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
SS-3 TCE MDL
0.37
2
0.35
SS-3 TCE MDL
0.39
0.42
0.52
1-BL-SS-VOC-3_MDL
1-NP-SS-VOC-3_MDL
1-PP-SS-VOC-3_MDL
C s TCE
10.3
263
9.6
C s TCE
2.0
1.8
1.1
C s
C_s-
C_s+
std dev C s TCE
3.5
220
3.0
std dev C s TCE
1.1
0.75
1.0
std_dev_C_s
std_dev_C_s-
std_dev_C_s+
C s/C i TCE
0.6
28
65
C s/C i TCE
0.4
0.6
9.8
C s/C i
C_s-/C_i-
C_s+/C_i+
C s TCE
8.5
390
7.9
C s TCE
1.5
1.8
1.4
C s
C_s-
C_s+
std dev C
2.1
0
1.0
std dev C
1.0
1.1
1.3
std_dev
std_dev_
std_dev_
s TCE
s TCE
C s
C_s-
C_s+
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
unitless
ug/m3
ug/m3
omit SS-1
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
DCE= 1 , 1 -dichloroethene
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
SS-1 DCE
13
14
20
SS-1 DCE
N/A
N/A
N/A
1-BL-SS-VOC-l
1-NP-SS-VOC-l
1-PP-SS-VOC-l
SS-1 DCE MDL
0.35
0.35
0.39
SS-1 DCE MDL
N/A
N/A
N/A
1-BL-SS-VOC-1_MDL
1-NP-SS-VOC-1_MDL
1-PP-SS-VOC-1_MDL
SS-2 DCE
0
310
0.43
SS-2 DCE
N/A
N/A
N/A
l-BL-SS-VOC-2
l-NP-SS-VOC-2
l-PP-SS-VOC-2
SS-2 DCE MDL
0.38
0.4
0.39
SS-2 DCE MDL
N/A
N/A
N/A
1-BL-SS-VOC-2_MDL
1-NP-SS-VOC-2_MDL
1-PP-SS-VOC-2_MDL
SS-3 DCE
0.95
350
2.2
SS-3 DCE
N/A
N/A
N/A
l-BL-SS-VOC-3
l-NP-SS-VOC-3
l-PP-SS-VOC-3
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
excluding SS-1
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
SS-3 DCE MDL
0.37
0.39
0.35
SS-3 DCE MDL
N/A
N/A
N/A
1-BL-SS-VOC-3_MDL
1-NP-SS-VOC-3_MDL
1-PP-SS-VOC-3_MDL
C s DCE
4.8
225
7.5
C s DCE
N/A
N/A
N/A
C s
C_s-
C_s+
std dev C s DCE
7.1
184
10.8
std dev C s DCE
N/A
N/A
N/A
std_dev_C_s
std_dev_C_s-
std_dev_C_s+
C s/C i DCE
39
36
178
C s/C i DCE
N/A
N/A
N/A
C_s/C_i
C_s-/C_i-
C_s+/C_i+
C s DCE
0.7
330
1.3
C s DCE
N/A
N/A
N/A
C s
C_s-
C_s+
std dev C s
0.4
28
1.3
std dev C s
N/A
N/A
N/A
std_dev_C
std_dev_C
std_dev_C_
DCE
DCE
_s
_s-
_s+
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
unitless
ug/m3
ug/m3
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
PCE=tetrachloroethene
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
SS-1 PCE
N/A
N/A
N/A
SS-1 PCE
2.2
1.4
0.69
1-BL-SS-VOC-l
1-NP-SS-VOC-l
1-PP-SS-VOC-l
SS-1 PCE MDL
N/A
N/A
N/A
SS-1 PCE MDL
0.73
0.4
0.42
1-BL-SS-VOC-1_MDL
1-NP-SS-VOC-1_MDL
1-PP-SS-VOC-1_MDL
SS-2 PCE
N/A
N/A
N/A
SS-2 PCE
0.77
0.55
1
l-BL-SS-VOC-2
l-NP-SS-VOC-2
l-PP-SS-VOC-2
SS-2 PCE MDL
N/A
N/A
N/A
SS-2 PCE MDL
0.37
0.41
0.41
1-BL-SS-VOC-2_MDL
1-NP-SS-VOC-2_MDL
1-PP-SS-VOC-2_MDL
SS-3 PCE
N/A
N/A
N/A
SS-3 PCE
3.1
6.2
3.2
l-BL-SS-VOC-3
l-NP-SS-VOC-3
l-PP-SS-VOC-3
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
excluding SS-1
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
SS-3 PCE MDL
N/A
N/A
N/A
SS-3 PCE MDL
0.39
0.42
0.52
1-BL-SS-VOC-3_MDL
1-NP-SS-VOC-3_MDL
l-PP-SS-VOC-3 MDL
C s PCE
N/A
N/A
N/A
C s PCE
2.0
2.7
1.6
C s
C_s-
C s+
std dev C s PCE
N/A
N/A
N/A
std dev C s PCE
1.2
3.0
1.4
std_dev_C_s
std_dev_C_s-
std dev C s+
C s/C i PCE
N/A
N/A
N/A
C s/C i PCE
0.7
1.4
3.9
C s/C i
C_s-/C_i-
C s+/C i+
C s PCE
N/A
N/A
N/A
C s PCE
1.9
3.4
2.1
C s
C_s-
C s+
std dev C s
N/A
N/A
N/A
std dev C s
1.6
4.0
1.6
std_dev_C
std_dev_C
std dev C
PCE
PCE
_s
_s-
s+
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
unitless
ug/m3
ug/m3
omit SS-1
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
Ben=benzene
SS-1 Ben
0.42
0.84
0
SS-1 Ben
0.98
0.75
0.75
1-BL-SS-VOC-l
1-NP-SS-VOC-l
1-PP-SS-VOC-l
SS-1 Ben MDL
0.35
0.35
0.39
SS-1 Ben MDL
0.73
0.4
0.42
1-BL-SS-VOC-1_MDL
1-NP-SS-VOC-1_MDL
1-PP-SS-VOC-1_MDL
SS-2 Ben
0
0.64
0
SS-2 Ben
0.42
0.45
0.83
l-BL-SS-VOC-2
l-NP-SS-VOC-2
l-PP-SS-VOC-2
SS-2 Ben MDL
0.38
0.4
0.39
SS-2 Ben MDL
0.37
0.41
0.41
1-BL-SS-VOC-2_MDL
1-NP-SS-VOC-2_MDL
1-PP-SS-VOC-2_MDL
SS-3 Ben
0.37
0.65
0.58
SS-3 Ben
2.7
2
1.5
l-BL-SS-VOC-3
l-NP-SS-VOC-3
l-PP-SS-VOC-3
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
excluding SS-1
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
SS-3 Ben MDL
0.37
0.39
0.35
SS-3 Ben MDL
0.39
0.42
0.52
1-BL-SS-VOC-3_MDL
1-NP-SS-VOC-3_MDL
1-PP-SS-VOC-3_MDL
C s Ben
0.39
0.71
0.45
C s Ben
1.37
1.07
1.03
C s
C_s-
C_s+
delta C s Ben
0.03
0.11
0.11
delta C s Ben
1.19
0.82
0.41
std_dev_C_s
std_dev_C_s-
std_dev_C_s+
C s/C i Ben
0.85
1.55
0.79
C s/C i Ben
1.24
0.68
0.67
C_s/C_i
C_s-/C_i-
C_s+/C_i+
C s Ben
0.38
0.65
0.49
C s Ben
1.61
1.10
0.47
C s
C_s-
C_s+
delta C s Ben
0.01
0.01
0.13
delta C s Ben
1.61
1.10
0.47
std_dev_C_s
std_dev_C_s-
std_dev_C_s+
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
unitless
ug/m3
ug/m3
omit SS-1
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
Calculation
C s
C s-
C s+
Calculation
C_s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
Tol=toluene
SS-1 Tol
3.8
12
0
SS-1 Tol
o
3.3
3.5
1-BL-SS-VOC-l
1-NP-SS-VOC-l
1-PP-SS-VOC-l
SS-1 Tol MDL
1.8
1.8
1.9
SS-1 Tol MDL
3.7
2
2.1
1-BL-SS-VOC-1_MDL
1-NP-SS-VOC-1_MDL
1-PP-SS-VOC-1_MDL
SS-2 Tol
0
0
0
SS-2 Tol
0
0
4
l-BL-SS-VOC-2
l-NP-SS-VOC-2
l-PP-SS-VOC-2
SS-2 Tol MDL
1.9
2
2
SS-2 Tol MDL
1.8
2.1
2.1
1-BL-SS-VOC-2_MDL
1-NP-SS-VOC-2_MDL
1-PP-SS-VOC-2_MDL
SS-3 Tol
2.4
0
3.7
SS-3 Tol
7.1
5.8
4.8
l-BL-SS-VOC-3
l-NP-SS-VOC-3
l-PP-SS-VOC-3
to
o
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
-------
Table A4. Data used to verify model assumption Cs = Cs~ (Continued)
excluding SS-1
Calculation
C s
C s-
C s+
Calculation
C s
C_s-
C s+
Roadmap of variable
C s
C_s-
C_s+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
SS-3 Tol MDL
1.9
2
1.8
SS-3 Tol MDL
2
2.1
2.6
1-BL-SS-VOC-3_MDL
1-NP-SS-VOC-3_MDL
1-PP-SS-VOC-3_MDL
C s Tol
2.7
5.3
2.5
C s Tol
4.2
3.7
4.1
C s
C_s-
C_s+
delta C s Tol
1.0
5.8
1.0
delta C s Tol
2.7
1.9
0.7
std_dev_C_s
std_dev_C_s-
std_dev_C_s+
C s/C i Tol
0.7
1.7
1.1
C s/C i Tol
1.1
0.4
0.5
C_s/C_i
C_s-/C_i-
C_s+/C_i+
C s Tol
2.2
2.0
2.9
C s Tol
4.5
4.0
4.4
C s
C_s-
C_s+
delta C s Tol
0.4
0.0
1.2
delta C s Tol
3.7
2.6
0.6
std_dev_C_s
std_dev_C_s-
std_dev_C_s+
Units
ug/m3 none
reported value < MDL
italics: data used in calcs
Red text: can leak; deviation 3
bold text: deviation 5
orange text: ADQ observations
ug/m3
ug/m3
ug/m3
unitless
ug/m3
ug/m3
omit SS-1
-------
Table A5. Data used to verify Mosley Model assumptions Rs = Rs~, Ra « RS9 Ra « RS'•, and Ra+ « Rs+
to
to
Calculation
R s,R a
R_s-,R_a-
R s+,R a+
Calculation
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
SS-1 1 Rn
404
429
376
SS-1 1 Rn
147
175
81
SS-1 1 Rn SD
42
43
40
SS-1 1 Rn SD
26
28
20
SS-1 1 Rn MDL
115
119
111
SS-1 1 Rn MDL
71
77
54
SS-1 2 Rn
488
425
447
SS-1 2 Rn
166
185
107
SS-1 2 Rn SD
46
43
44
SS-1 2 Rn SD
27
29
22
SS-1 2 Rn MDL
127
118
122
SS-1 2 Rn MDL
76
79
62
Variable names
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Units
pCi/L
reported value
italicized text:
Building
Building
Building
none
-------
Table A5. Data used to verify Mosley Model assumptions Rs = Rs~, Ra « RS9 Ra « RS'•, and Ra+ « Rs+ (Continued)
to
OJ
Calculation
R s,R a
R_s-,R_a-
R s+,R a+
Calculation
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
SS-1 3 Rn
419
446
410
SS-1 3 Rn
180
209
108
SS-1 3 Rn SD
42
44
42
SS-1 3 Rn SD
28
30
22
SS-1 3 Rn MDL
117
121
116
SS-1 3 Rn MDL
79
84
62
SS-2 1 Rn
76
118
58
SS-2 1 Rn
72
246
18
SS-2 1 Rn SD
19
23
17
SS-2 1 Rn SD
19
33
10
SS-2 1 Rn MDL
53
65
48
SS-2 1 Rn MDL
51
91
29
Variable names
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Units
pCi/L
reported value
italicized text:
Building
Building
Building
none
-------
Table A5. Data used to verify Mosley Model assumptions Rs = Rs~, Ra « RS9 Ra « RS'•, and Ra+ « Rs+ (Continued)
Calculation
R s,R a
R_s-,R_a-
R s+,R a+
Calculation
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
SS-2 2 Rn
88
112
41.1
SS-2 2 Rn
81
249
19
SS-2 2 Rn SD
21
23
15
SS-2 2 Rn SD
20
33
11
SS-2 2 Rn MDL
57
64
41.2
SS-2 2 Rn MDL
54
92
29
SS-2 3 Rn
77
86
37
SS-2 3 Rn
93
278
18
SS-2 3 Rn SD
20
21
14
SS-2 3 Rn SD
21
35
10
SS-2 3 Rn MDL
54
57
40
SS-2 3 Rn MDL
58
97
29
Variable names
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Units
pCi/L
reported value
italicized text:
Building
Building
Building
none
-------
Table A5. Data used to verify Mosley Model assumptions Rs = Rs~, Ra « RS9 Ra « RS'•, and Ra+ « Rs+ (Continued)
Calculation
R s,R a
R_s-,R_a-
R s+,R a+
Calculation
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
SS-3 1 Rn
15
101
4.8
SS-3 1 Rn
528
592
613
SS-3 1 Rn SD
11
22
7.3
SS-3 1 Rn SD
47
50
51
SS-3 1 Rn MDL
30
61
20
SS-3 1 Rn MDL
132
139
141
SS-3 2 Rn
18
105
4.8
SS-3 2 Rn
587
648
633
SS-3 2 Rn SD
11
22
8.5
SS-3 2 Rn SD
50
53
52
SS-3 2 Rn MDL
31
62
24
SS-3 2 Rn MDL
139
146
144
Variable names
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Units
pCi/L
reported value
italicized text:
Building
Building
Building
none
-------
Table A5. Data used to verify Mosley Model assumptions Rs = Rs~, Ra « RS9 Ra « RS'•, and Ra+ « Rs+ (Continued)
Calculation
R s,R a
R_s-,R_a-
R s+,R a+
Calculation
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Variable names
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
SS-3 3 Rn
15
135
8.6
SS-3 3 Rn
549
684
611
BL-SS-3_3_Rn
NP-SS-3_3_Rn
PP-SS-3_3_Rn
SS-3 3 Rn SD
11
25
7.9
SS-3 3 Rn SD
48
54
51
BL-SS-3_3_Rn_SD
NP-SS-3_3_Rn_SD
PP-SS-3_3_Rn_SD
SS-3 3 Rn MDL
31
70
22
SS-3 3 Rn MDL
134
150
141
BL-SS-3_3_Rn_MDL
NP-SS-3_3_Rn_MDL
PP-SS-3_3_Rn_MDL
SS-1 Rn AVG
437
433
411
SS-1 Rn AVG
164
190
99
BL-SS-l_Rn_AVG
NP-SS-l_Rn_AVG
PP-SS-l_Rn_AVG
std dev SS-1 Rn
45
11
36
std dev SS-1 Rn
16
18
15
std_dev_BL-SS-l_Rn
std_dev_NP-SS-l_Rn
std_dev_PP-SS-l_Rn
Units
pCi/L none
reported value < MDL
italicized text: data used in coles
pCi/L
pCi/L
pCi/L
pCi/L
pCi/L
-------
Table A5. Data used to verify Mosley Model assumptions Rs = Rs~, Ra « RS9 Ra « RS'•, and Ra+ « Rs+ (Continued)
Calculation
R s,R a
R_s-,R_a-
R s+,R a+
Calculation
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
SS-2 Rn AVG
80
106
46
SS-2 Rn AVG
82
258
18
std dev SS-2 Rn
7.0
17
11
std dev SS-2 Rn
11
18
0.5
SS-3 Rn AVG
16
113
6.0
SS-3 Rn AVG
555
641
619
std dev SS-3 Rn
1.7
18
2.2
std dev SS-3 Rn
30
47
12
R s
178
218
154
R s
267
363
245
std dev R s
227
187
223
std dev R s
253
243
326
Variable names
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Units
pCi/L
reported value
italicized text:
Building
Building
Building
none
-------
Table A5. Data used to verify Mosley Model assumptions Rs = Rs~, Ra « RS9 Ra « RS'•, and Ra+ « Rs+ (Continued)
to
oo
excluding SS-1
Calculation
R s,R a
R_s-,R_a-
R s+,R a+
Calculation
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Variable names
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
R s
48
110
26
R s
N/A
N/A
N/A
R_s
R_s-
R_s+
std dev R s
45
5
28
std dev R s
N/A
N/A
N/A
std_dev_R_s
std_dev_R_s-
std_dev_R_s+
R a
0.104
0.026
0.069
R a
0.177
0.256
0.328
R_a
R_a-
R_a+
R a MDL
0.289
0.220
0.055
R a MDL
0.175
0.237
0.165
R_a_MDL
R_a-_MDL
R_a+_MDL
delta R a
0.038
0.028
0.052
delta R a
0.008
0.066
0.107
delta_R_a
delta_R_a-
delta_R_a+
R s/R i
472
116
1686
R s/R i
278
589
811
R_s/R_i
R_s-/R_i-
R_s+/R_i+
R s/R a
1703
8215
2232
R s/R a
1507
1416
748
R_s/R_a
R_s-/R_a-
R_s+/R_a+
delta (R s/R a)
2258
11239
3642
delta (R s/R a)
1427
1017
1024
delta_(R_s/R_a)
delta_(R_s-/R_a-)
delta_(R_s+/R_a+)
Units
pCi/L none
reported value < MDL
italicized text: data used in coles
pCi/L
pCi/L
pCi/L
pCi/L
pCi/L
unitless
unitless
unitless
-------
Table A5. Data used to verify Mosley Model assumptions Rs = Rs~, Ra « RS9 Ra « RS'•, and Ra+ « Rs+ (Continued)
Calculation
R_s,R_a
R_s-,R_a-
R s+,R a+
Calculation
R_s,R_a
R_s-,R_a-
R s+,R a+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
R s/R
463
4137
373
R s/R
N/A
N/A
N/A
excluding SS-1
a delta (R s/R a)
466
4407
492
a delta (R s/R a)
N/A
N/A
N/A
Variable names
R_s,R_a
R_s-,R_a-
R_s+,R_a+
Units
pCi/L
Building
Building
Building
none
R_s/R_
R_s-/R_
R_s+/R_
a delta_(R_s/R_a)
a- delta_(R_s-/R_a-)
a+ delta_(R_s+/R_a+)
unitless Unitless
reported value < MDL
italicized text: data used in coles
-------
Table A6. Data used to verify Qi» XV, Qi » XV, Qi+ » XV
OJ
o
Calculation
Q_i,lamda*V
Q_i-,lamda*V
Q i+,lamda*V
Calculation
Q_i,lamda*V
Q_i-,lamda*V
Q i+,lamda*V
Roadmap of variable
Q_i,lamda*V
Q_i-,lamda*V
Q_i+,lamda*V
Units
m3/h
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
none
Q i
54
384
364
Q i
105
592
634
Q_i
Q_i-
Q_i+
m3/h
delta Q i
14
351
342
delta Q i
13
256
542
delta_Q_i
delta_Q_i-
delta_Q_i+
m3/h
lambda
0.0075546
0.0075546
0.0075546
lambda
0.0075546
0.0075546
0.0075546
lambda
lambda
lambda
/h
delta lambda
1%
1%
1%
delta lambda
1%
1%
1%
delta_lambda
delta_lambda
delta_lambda
%
V
273.5
273.5
273.5
V
364.8
364.8
364.8
V
V
V
m3
delta V
30%
30%
30%
delta V
30%
30%
30%
delta_V
delta_V
delta_V
%
lambda*V
2.07
2.07
2.07
lambda*V
2.76
2.76
2.76
lambda*V
lambda*V
lambda*V
m3/h
delta (lambda*V)
0.62
0.62
0.62
delta (lambda*V)
0.83
0.83
0.83
delta_(lambda*V)
delta_(lambda*V)
delta_(lambda*V)
m3/h
-------
Table A6. Data used to verify Qi» XV, Qi » XV, Qi+ » XV (Continued)
Calculation
Q_i,lamda*V
Q_i-,lamda*V
Q i+,lamda*V
Calculation
Q_i,lamda*V
Q_i-,lamda*V
Q i+,lamda*V
Roadmap of variable names
Q_i,lamda*V
Q_i-,lamda*V
Q_i+,lamda*V
Units
m3/h
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
None
Q i/(lambda*V)
26
186
176
Q i/(lambda*V)
38
215
230
Q_i/(lambda*V)
Q_i-/(lambda*V)
Q_i+/(lambda*V)
unitless
delta [Q i/(lambda*V)]
10
179
174
delta [Q i/(lambda*V)]
12
113
209
delta_[Q_i/(lambda*V)]
delta_[Q_i+/(lambda*V)]
delta_[Q_i+/(lambda*V)]
unitless
-------
Table A7. Indoor air, ambient air, and sub-slab SFe data
OJ
to
Calculation
T_s,T_i/T_s
T_s-,T_i-/T_s-
T s+,T i+/T s+
Calculation
T_s,T_i/T_s
T_s-,T_i-/T_s-
T_s+,T_i+/T_s+
Roadmap of variable names
Q_i*R_i
Q_i-*R_i-
Q_i+*R_i+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Building
Building
Building
AA SF 6
0
12
0
AA SF 6
18
16
18
1-BL-AA-VOC-l
1-NP-AA-VOC-l
1-PP-AA-VOC-l
AA SF 6 MDL
8.8
9.4
10
AA SF 6 MDL
8.7
7.9
9.1
1-BL-AA-VOC-1_MDL
1-NP-AA-VOC-1_MDL
1-PP-AA-VOC-1_MDL
SS-1 SF 6
290
120
370
SS-1 SF 6
1,800
200
570
1-BL-SS-VOC-l
1-NP-SS-VOC-l
1-PP-SS-VOC-l
SS-1 SF 6 MDL
17
8
9
SS-1 SF 6 MDL
110
10
10
1-BL-SS-VOC-1_MDL
1-NP-SS-VOC-1_MDL
1-PP-SS-VOC-1_MDL
SS-2 SF 6
260
14
1,000
SS-2 SF 6
7,700
86
750
l-BL-SS-VOC-2
l-NP-SS-VOC-2
l-PP-SS-VOC-2
Units
pCi/h none
reported value < MDL
italicized text: data used in calcs
ug/m3
ug/m3
(ig/m3
ug/m3
Hg/m3
-------
Table A7. Indoor air, ambient air, and sub-slab SFe data (Continued)
OJ
OJ
excluding SS-1
Calculation
T_s,T_i/T_s
T s-,T i-/T s-
T s+,T i+/T s+
Calculation
T_s,T_i/T_s
T_s-,T_i-/T_s-
T_s+,T_i+/T_s+
Roadmap of variable
Q_i*R_i
Q_i-*R_i-
Q i+*R i+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
SS-2 SF 6MDL
9
10
19
SS-2 SF 6MDL
88
10
10
1-BL-SS-VOC-2_MDL
1-NP-SS-VOC-2_MDL
l-PP-SS-VOC-2 MDL
SS-3 SF 6
7,700
24
1,600
SS-3 SF 6
700
72
90
l-BL-SS-VOC-3
l-NP-SS-VOC-3
l-PP-SS-VOC-3
SS-3 SF 6 MDL
88
9.4
84
SS-3 SF 6 MDL
9
10
10
1-BL-SS-VOC-3_MDL
1-NP-SS-VOC-3_MDL
l-PP-SS-VOC-3 MDL
T s
750
53
990
T s
1,200
119
470
T_s
T_s-
T s+
std dev T s
823
59
615
delta T s
954
70
341
std_dev_T_s
std_dev_T_s-
std dev T s+
T s
980
19
1300
T s
N/A
N/A
N/A
T_s
T_s-
T s+
std dev T
1018
7
424
std dev T
N/A
N/A
N/A
std_dev_T
std_dev_T
std dev T
s
s
_s
_s-
s+
Units
pCi/h none
reported value < MDL
italicized text: data used in calcs
ug/m3
ug/m3
ug/m3
ug/m3
ug/m3
jig/m3
(ig/m3
-------
Table A7. Indoor air, ambient air, and sub-slab SFe data (Continued)
excluding SS-1
Calculation
T s,T i/T s
T_s-,T_i-/T_s-
T s+,T i+/T s+
Calculation
T s,T i/T s
T_s-,T_i-/T_s-
T_s+,T_i+/T_s+
Roadmap of variable
Q_i*R_i
Q_i-*R_i-
Q_i+*R_i+
Building
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
names
Building
Building
Building
T i
5200
780
867
T i
3300
563
530
T_i
T_i-
T i+
std dev T i
1253
710
810
std dev T i
265
237
450
std_dev_T_i
std_dev_T_i-
std dev T i+
T i/T a
591
65
87
T i/T a
183
35
29
T_i/T_a
T_i-/T_a-
T i+/T a+
T s/T i
0.14
0.07
1.1
T s/T i
0.36
0.21
0.89
T_s/T_i
T_s-/T_i-
T s+/T i+
delta (T s/T i)
0.16
0.10
1.3
delta (T s/T i)
0.29
0.15
1.0
delta_(T_s/T_i)
delta_(T_s-/T_i-)
delta_(T_s+/T_i+)
T s/T i
0.19
0.02
1.5
T s
N/A
N/A
N/A
T_s/T_i
T_s-/T_i-
T s+/T i+
delta (T s/T i)
0.20
0.02
1.5
std dev T s
N/A
N/A
N/A
delta_(T_s/T_i)
delta_(T_s-/T_i-)
delta_(T_s+/T_i+)
Units
pCi/h none
reported value < MDL
italicized text: data used in calcs
ug/m3
ug/m3
unitless
unitless
unitless
unitless
unitless
-------
Table A8. Calculation of Mass Discharges
Pressure
BL
NP
PP
BL/BL
NP/BL
PP/BL
Pressure
BL
NP
PP
BL/BL
NP/BL
PP/BL
Building
ASU
ASU
ASU
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Moffett
Moffett
Moffett
Q i
54
384
364
N/A
N/A
N/A
Q i
105
592
634
N/A
N/A
N/A
R i
377
1868
91
N/A
N/A
N/A
R i
961
617
302
N/A
N/A
N/A
R a
104
26
69
N/A
N/A
N/A
R a
177
256
328
N/A
N/A
N/A
Q i-R i
20520
716388
33310
1.0
35
1.6
Q i-R i
101219
365347
191721
1.0
3.6
1.9
Radon
Q i-R a
5690
10157
25167
0.28
0.49
1.2
Q i-R a
18663
151855
207874
0.18
1.5
2.1
Q i-(R i-
R a) Q
14830
706231
8143
0.72
34.4
0.40
Q i-(R i-
Ra) Q
82556
213492
-16153
2.1
-0.16
i-C i TCE
1017
3631
53
1.0
3.6
0.05
i-C i TCE
519
1856
74
1.0
3.6
0.14
0
Q
TCE
i-C a TCE
9.3
58
31
0.01
0.06
0.03
i-C a TCE
8.8
71
56
0.02
0.14
0.11
Q i-(C i TCE-
C a TCE)
1008
3574
23
0.99
3.5
0.02
Q i-(C i TCE-
C a TCE)
511
1785
18
0.98
3.4
0.03
Values highlighted in green are plotted in Figures 16 and 17.
-------
Table A8. Calculation of Mass Discharges (Continued)
Pressure
Building
Q_i-C_i_DCE
1,1-DCE PCE
Q i-C a DCE Q i-(C i DCE-C a DCE) Q i-C i PCE Q i-C a PCE
_i- (C_i_PCE-C_a_PCE)
BL
NP
PP
BL/BL
NP/BL
PP/BL
ASU
ASU
ASU
ASU
ASU
ASU
6.7
2391
15.4
1.0
356
2.3
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Pressure
Building
Q i-C i DCE Q i-C a DCE Q i-(C i DCE-C a DCE) Q i-C i PCE Q i-C a PCE Q i-(C i PCE-C a PCE)
BL
NP
PP
BL/BL
NP/BL
PP/BL
Moffett
Moffett
Moffett
Moffett
Moffett
Moffett
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
298
1185
266
1.0
4.0
0.89
-------
Table A8. Calculation of Mass Discharges (Continued)
Pressure
BL
NP
PP
BL/BL
NP/BL
PP/BL
Pressure
BL
NP
PP
BL/BL
NP/BL
PP/BL
Building
ASU
ASU
ASU
ASU
ASU
ASU
Building
Moffett
Moffett
Moffett
Moffett
Moffett
Moffett
Q i-T i
283293
299201
315823
1.0
1.06
1.11
Q i-T i
347459
333696
336020
1.0
0.96
0.97
SF6
Q i-T a
479
4603
3641
0.00
0.02
0.01
Q i-T a
1895
9478
11412
0.01
0.03
0.03
Q i'(T i-
T a) Q
282814
294598
312182
1.00
1.04
1.10
Q i'(T i-
T a) Q
345564
324219
324608
0.99
0.93
0.93
i-C i Ben Q
24.9
175
209
1.0
7.0
8.4
i-C i Ben Q
115.8
928
972
1.0
8.0
8.4
Benzene
i-C a Ben
21.2
161
164
0.85
6.6
i-C a Ben
116
889
888
1.00
7.7
7.7
Q i-(C i Ben-
C a Ben)
3.6
14
45
0.15
0.6
Q_i-(C_i_Ben-
C a Ben)
0.0
39
85
0.00
0.7
Q i-C i Tol
203
1176
825
1.0
5.8
4.1
Q i-C i Tol
411
5667
5389
1.0
13.8
13.1
Toluene
Q i-C a Tol
212
729
546
1.04
3.6
2 7
Q i-C a Tol
369
5568
6340
0.90
13.6
15 4
Q i-(C i Tol-
C a Tol)
-9
448
279
-0.04
M
Q i-(C i Tol-
C a Tol)
42
99
-951
0.10
0.24
------- |