EPA/600/R-11/020 | April 2012 | www.epa.gov/ord
United States
Environmental Protection
Agency
Multimedia Sampling During the
Application of Biosolids on a
Land Test Site
Office of Research and Development .L
National Risk Management Research Laboratory
-------
EPA/600/R-11/020
April 2012
MULTIMEDIA SAMPLING DURING THE APPLICATION OF
BIOSOLIDS ON A LAND TEST SITE
by
Eric A. Foote; Carolyn M. Acheson; Edwin F. Barth; Ronald F. Herrmann;
Richard C. Brenner; D. Bruce Harris; Steven J. Naber;
Robert H. Forbes, Jr.; Laura L. McConnell; and Patricia D. Millner
Contract No. EP-C-05-056
Work Assignment 4-48
Contractor
Pegasus Technical Services, Inc.
Cincinnati, OH 45219
Principal Subcontractor
Battelle
Columbus, OH 43201
U.S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
Land Remediation and Pollution Control Division
Cincinnati, OH 45268
-------
NOTICE
The U.S. Environmental Protection Agency (EPA), through its Office of Research and
Development (ORD), funded and managed the research described herein under Contract No.
EP-C-05-056. It has been subjected to the Agency's peer and administrative review and has been
approved for publication as an EPA document.
-------
FOREWORD
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
Nation's land, air, and water 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, EPA's research program
is providing data and technical support for solving environmental problems today and building a science
knowledge base necessary to manage our ecological resources wisely, understand how pollutants affect
our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for
investigation of technological and management approaches for preventing and reducing risks from
pollution that threaten human health and the environment. The focus of the Laboratory's research
program is on methods and their cost-effectiveness for prevention and control of pollution to air, land,
water, and subsurface resources; protection of water quality in public water systems; remediation of
contaminated sites, sediments, and ground water; prevention and control of indoor air pollution; and
restoration of ecosystems. NRMRL collaborates with both public and private sector partners to foster
technologies that reduce the cost of compliance and to anticipate emerging problems. NRMRL's research
provides solutions to environmental problems by: developing and promoting technologies that protect and
improve the environment; advancing scientific and engineering information to support regulatory and
policy decisions; and providing the technical support and information transfer to ensure implementation
of environmental regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan.
It is published and made available by EPA's Office of Research and Development to assist the user
community and to link researchers with their clients.
Cynthia Sonich-Mullin, Director
National Risk Management Research Laboratory
m
-------
This page intentionally left blank.
IV
-------
CONTENTS
NOTICE ii
FOREWORD iii
APPENDICES vii
FIGURES viii
TABLES x
ABBREVIATIONS AND ACRONYMS xi
ACKNOWLEDGMENTS xiii
EXECUTIVE SUMMARY xvi
1.0 PROJECT DESCRIPTION AND OBJECTIVES 1
1.1 Background and Introduction 1
1.2 Project Goal 2
1.3 Project Objectives 3
1.3.1 Task 1. Bioaerosol and Particulate Matter Sampling 3
1.3.2 Task 2. Volatile Organic Compound (VOC) and Odor Sampling 3
1.3.3 Task 3. Land Sampling 3
2.0 EXPERIMENTAL APPROACH AND TEST SITE SETUP 4
2.1 Experimental Approach 4
2.2 Site Design 6
2.3 Task 1. Bioaerosol and Particulate Sampling Design 6
2.3.1 Sampling Station Design 6
2.3.1.1 Transect Stations 6
2.3.1.2 Mobile Unit 8
2.3.1.3 Center Station 8
2.4 Task 2. VOC, Ammonia, Hydrogen Sulfide, and Odor Sampling 8
2.4.1 Open Path Fourier Transform Infrared Spectrometer 9
2.4.2 Advective Flux Measurements 9
2.4.3 Ammonia and Hydrogen Sulfide Measurements 10
2.4.4 Odor Measurements 11
2.4.5 In-Laboratory Biosolids Measurements 11
2.4.5.1 SVOC Analysis of the Biosolids 11
2.4.5.2 Headspace Analysis 11
2.5 Task 3. Land Sampling 12
2.5.1 Land Sampling Field Plot Design 12
2.5.2 Land Sampling Procedures and Plan 12
2.6 Schedule of Events 13
3.0 BIOSOLIDS APPLICATION 21
3.1 Product Selection 21
3.2 Application 21
4.0 BIOSOLIDS PRODUCT RESULTS 24
4.1 Samples Collected from the Delivery Truck 24
4.1.1 Microbial Enumeration of Biosolids Grab Samples 24
4.1.2 Flux Chamber Measurement of Biosolids (Measured from the Truck) 24
4.1.3 SPME Measurements from Samples Generated from the Flux Chambers 24
4.2 Samples Collected from the Biosolids Stockpile Prior to Application 24
-------
4.2.1 SVOC Analysis on Biosolids 25
4.2.2 Headspace Analysis Using VOA-7 25
4.3 Microbial and Physical/Chemical Characterization of the Biosolids Samples Collected
from the Stockpile 25
5.0 TASK 1. BIOAEROSOL AND PARTICULATE SAMPLING RESULTS 28
5.1 Objectives 28
5.2 Bioaerosol and Participate Matter Sampling 28
5.3 Overview of Field Operations 30
5.3.1 Application Schedule 30
5.3.2 Operational Schedule 31
5.3.3 Aerosol Sample Collection 31
5.3.3.1 Biosampler 32
5.3.3.2 Six-Stage Impactor 32
5.3.3.3 Button Sampler 33
5.3.3.4 GRIMM Particle Analyzer 33
5.4 Bioaerosol and Particulate Matter Results 34
5.4.1 Bioaerosol Results and Analysis 34
5.4.2 Particulate Matter 38
5.5 Conclusions 39
5.6 Discussion and Lessons Learned 39
6.0 TASK 2. VOLATILE ORGANICS AND ODOR SAMPLING RESULTS 40
6.1 Objectives 40
6.2 Volatile Emissions Sampling and Measurements from Biosolids 40
6.2.1 Headspace Analysis of Biosolids 40
6.2.2 Advective Flux Measurements 41
6.2.3 Off-Site Odor Panel Analysis 42
6.2.4 Direct Gas Measurements (Ammonia and Hydrogen Sulfide) 43
6.2.5 Open-Path Fourier Transform Infrared Spectrometer 43
6.3 Results and Discussion 43
6.3.1 Head Space Analysis of Biosolids 43
6.3.2 Advective Flux 44
6.3.3 Direct Gas Measurements (Ammonia and Hydrogen Sulfide) 47
6.3.4 Off-Site Odor Panel and On-Site Nasal Ranger® Analyses 47
6.3.5 Open-Path Fourier Transform Infrared Spectrometer Measurements 49
6.3.5.1 Baseline Ammonia Measurements 49
6.3.5.2 Ammonia Measurements During Biosolids Application 50
6.3.5.3 Ammonia Measurements After Biosolids Application 50
6.4 Conclusions 53
6.5 Discussion and Lessons Learned 53
7.0 TASK3. LAND SAMPLING RESULTS 55
7.1 Objectives 55
7.2 Overview of Field Plots 56
7.3 Sample Collection and Analysis 56
7.3.1 Collection Methods 56
7.3.2 Analyte Specific Sample Collection and Analysis Information 57
7.3.3 Statistical Analysis 58
7.4 Data and Results 59
7.4.1 Soil Characterization 59
VI
-------
7.4.2 Weather Data 60
7.4.3 Biosolids Distribution 61
7.4.4 PLFA 63
7.4.5 Fecal Colifbrms 68
7.4.6 Alkylphenol Ethoxylates 69
7.4.7 Ecotoxicity Screening 73
7.4.8 Other Microbial Measurements 76
7.5 Conclusions 79
7.6 Discussion and Lessons Learned 80
8.0 SUMMARY AND CONCLUSIONS 82
8.1 Introduction 82
8.2 Study Description 82
8.3 Study Site 83
8.4 Applied Biosolids 83
8.5 Field Results 83
8.5.1 Biosolids Characterization 84
8.5.2 Bioaerosol and Particulate Sampling 84
8.5.3 Volatile Organic and Inorganic Emissions and Odorant Sampling 85
8.5.4 Land Sampling 86
8.6 Lessons Learned 87
8.7 Recommendations 88
8.7.1 Bioaerosol Sampling 88
8.7.2 Particulate Sampling 89
8.7.3 Volatile Organic Compound and Ammonia Sampling 89
8.7.4 Odor Sampling 89
8.7.5 Land Sampling 90
9.0 REFERENCES 91
APPENDICES
Appendix A: Determination of Total Bacterial Bioburden from Impinger Samples Collected
During the NC Biosolids Land Application Study - Dr. Mark Hernandez, University of
Colorado
Appendix B: Parallel Sampling Approaches and Analysis of Impinger Samples Collected During the
NC Biosolids Land Application Study - Dr. Ian Pepper, University of Arizona
Appendix C: Endotoxin Sampling During a Post-Spring Cutting Event at the NC Biosolids Land
Application Study Site - Dr. Edwin Earth, EPA/NRMRL
Appendix D: Soil Agronomic Results for Land Samples and Fecal Coliform Results for Land Samples
VII
-------
FIGURES
Figure 2-1. Aerial View of the Test Site and Application Area 5
Figure 2-2. Bioaerosol and Particulate Sampling Array 7
Figure 2-3. Cart-Mounted Transect Sampling Equipment 8
Figure 2-4. Mobile Unit 9
Figure 2-5. Map of the Site Layout Showing the Location of the Vertical Radial Plume
Mapping Configurations and the Single-Path Measurements 10
Figure 2-6. Flux Chamber Design 11
Figure 2-7. Land Sampling Plan 14
Figure 3-1. Method of Biosolids Application 22
Figure 5-1. MOB Conducting Bioaerosol Sampling Approximately 8-10 m Behind Biosolids
Applicator 31
Figure 5-2. Downwind B (DWB) Biosampler and Operator 32
Figure 5-3. Distribution of BioSampler Impinger Fluid for Bacterial and Viral Analyses 33
Figure 5-4. Predominant Wind Directions and Velocities During Biosolids Control Trial and
Application Test 35
Figure 5-5. Bioaerosol Concentrations of Microorganisms for Mobile, Upwind, and Downwind
Sampling Locations 36
Figure 5-6. Mean Mass of Airborne Parti culates (< 5.0 (im) Captured by the GRIMM Sampler
Immediately Before and During Control and Application Sampling Periods 38
Figure 6-1. Aerial View of Test Site and Sampling Stations 41
Figure 6-2. Glass Vessel Used for Biosolids Headspace Analysis 42
Figure 6-3. Collecting a Flux Sample in the Field 42
Figure 6-4. Estimated Emission Factors Over Time of Biosolids Application 43
Figure 6-5. Calculated VOC Flux Rates for Acetone, Trimethylamine, Dimethyl Sulfide, and
Dimethyl Disulfide for up to 20 hr After Biosolids Application in the Test Area 45
Figure 6-6. SPME Fibers Exposed to Emission Samples Collected in Tedlar® Bags from Flux
Chamber Off-Gas 46
Figure 6-7. Concentration of VOCs from SPMEs Exposed to Air Emissions of Flux Chambers 46
Figure 6-8. Reconstructed Ammonia Plume from the Upwind VRPM Survey During Biosolids
Application on Day 0 51
Figure 6-9. Reconstructed Ammonia Plume from the Downwind VRPM Survey During
Biosolids Application on Day 0 51
Figure 6-10. Reconstructed Ammonia Plume from the Downwind VRPM Survey
Approximately 2 hr After Biosolids Application 52
Figure 6-11. Reconstructed Ammonia Plume from the Downwind VRPM Survey
Approximately 3 hr After Biosolids Application 52
Figure 7-1. Rainfall and Soil Moisture Measurements During Land Sampling Period 60
Figure 7-2. Soil Temperature During Land Sampling Period 61
Figure 7-3. Photographs of Land Sampling Plots Before and After Biosolids Application: A -
Before Application; and B - After Application 62
Figure 7-4. Distribution of Biosolids for Each Land Sampling Plot Based on Ash Mass 62
Figure 7-5. Photographs of Land Application of Biosolids on Plot 2 63
Figure 7-6. Total Biomass for Surficial Samples, 0-5 cm, in Each Plot Before and After
Biosolids Application 65
Figure 7-7. Total Biomass as a Function of Time After Biosolids Application for Surficial
Samples, 0-5 cm 66
Figure 7-8. Community Structure Based on PLFA Profile: A. 0-5 cm Depth; B. 10-15 cm
Depth; and C. 20-25 cm Depth 68
Figure 7-9. Fecal Coliform Concentrations as a Function of Time and Plot 69
vm
-------
Figure 7-10. NP Concentrations as a function of Time and Plot in Surficial Samples, 0-5 cm,
after Biosolids Application 70
Figure 7-11. Statistical Power for NP Samples as a Function of Sample Number for Day 0 and
Day 63 72
Figure 7-12. Root Elongation as a Function of Test Soil Concentration and Sample Event: A -
Lettuce Data; and B - Oat Data 74
Figure 7-13. Average Root Length for all Soil Concentrations as a Function of Treatment Time
for Lettuce and Oats 76
IX
-------
TABLES
Table 2-1. General Schedule of Field Events 15
Table 2-2a. Bioaerosols and Participate Sampling: Analyte, Method, Sample Frequency, and
Responsible Personnel for Collection/Analyses 16
Table 2-2b. VOC and Odor Sampling: Analyte, Method, Sample Frequency, and Responsible
Personnel for Collection/Analyses 17
Table 2-2c. Biosolids Sampling: Analyte, Method, Sample Frequency, and Responsible
Personnel for Collection/Analyses 18
Table 2-2d. Land Sampling: Analyte, Method, Sample Frequency, and Responsible Personnel
for Collection/Analyses 20
Table 3-1. Application Timeline 23
Table 4-1. SVOC Results for Biosolids Stockpile Composite Sample and Samples Collected
in the Field After Application on Days 1 and 2 26
Table 4-2. Microbial Analyses of Biosolids Collected from the Stockpile Composite Sample 26
Table 4-3. Physical/Chemical Constituents Results for Biosolids Collected from the Stockpile
Composite Sample 27
Table 5-1. Summary of Sampling Location Comparisons with Statistically Significant
Differences for THE and Total Fungi (Six-Stage Impactor Data Only) 37
Table 6-1. Results of Dynamic Dilution Olfactometry Analysis 48
Table 7-1. Sample Analytes, Events, and Sample Numbers for Land Sampling 57
Table 7-2. Soil Characterization Data for Land Sampling Plots 59
Table 7-3. Total Biomass ANOVA Results for Statistically Significant Factors 64
Table 7-4. Three-Way ANOVA Results for Root Elongation 75
Table 7-5. Other Microbial Indicators from Land Sampling 77
Table 7-6. Other Microbial Indicators Measured by the USDA Laboratory 78
Table 7-7. Fecal Coliform Results from Environmental Associates, Inc. and USDA 79
-------
ABBREVIATIONS AND ACRONYMS
ags above ground surface
ANOVA analysis of variance
APEs alkylphenol and alkylphenol ethoxylates
ASTM American Society for Testing and Materials
ATV all-terrain vehicle
BPA bisphenol-A
CEC cation exchange capacity
CFR Code of Federal Regulations
CPU colony forming unit
DT detection threshold
d/T dilutions-to-threshold
DWA Downwind Location A
DWB Downwind Location B
DWC Downwind Location C
EDC endocrine disrupting compound
EPA U.S. Environmental Protection Agency
FAME fatty acid methyl ester
FTIR Fourier transform infrared
GPS global positioning system
HCA Fiierarchal Cluster Analysis
LRPCD Land Remediation and Pollution Control Division
MANOVA multivariate analysis of variance
MD-GC-MS multidimensional gas chromatography/mass spectrometry
MOB mobile sampler
MPN most probable number
NAS National Academy of Sciences
NCDA&CS North Carolina Department of Agriculture & Consumer Services
NP nonylphenol
NRC National Research Council
NRMRL National Risk Management Research Laboratory
OP octylphenol
OP-FTIR open-path Fourier transform infrared
ORD Office of Research and Development
ORS optical remote sensing
PBDE polybrominated diphenyl ether
PBS phosphate buffered solution
XI
-------
PCA principal component analysis
PCB polychlorinated biphenyl
PCR polymerase chain reaction
PFLA phospholipid fatty acid
PFU plaque forming unit
PO primary objective
ppb parts per billion
ppm parts per million
ppmv parts per million by volume
QAPP Quality Assurance Project Plan
RSD relative standard deviation
RT recognition threshold
SAS Statistical Analysis Software
SPME solid-phase microextraction
SVOC semivolatile organic compound
THE total heterotrophic bacteria
UHP ultra-high purity
UWA Upwind Location A
USDA U.S. Department of Agriculture
VHO viable Helminth ova
VOA volatile organic analysis
VOC volatile organic compound
VRPM vertical radial plume mapping
WWTP wastewater treatment plant
XII
-------
ACKNOWLEDGMENTS
This report was prepared by Battelle Memorial Institute (Battelle), Columbus, OH, a
subcontractor to Pegasus Technical Services, Inc. (PTSI), Cincinnati, OH, under Contract No.
EP-C-05-056 between the U.S. Environmental Protection Agency (EPA) and PTSI. Stephen L. Wright
served as the EPA Project Officer, and Richard C. Brenner was the EPA Work Assignment Manager.
The Battelle project lead was Eric A. Foote. This project was conducted under the direction of EPA's
Environmental Stressors Management Branch (Branch Chief, Laurel J. Staley) of the Land Remediation
and Pollution Control Division (LRPCD), National Risk Management Research Laboratory (NRMRL).
This project was carried out at the Piedmont Research Station of the North Carolina Department
of Agriculture and Consumer Services (NCDA&CS) in Salisbury, NC. NCDA&CS staff generously
provided invaluable assistance to the researchers by making space available in their buildings and
facilities, supplying logistics support, lending tools and equipment, preparing (mowing) the application
field for receiving biosolids, consulting with research staff on equipment deployment, and courteously
responding to many questions and requests from the research team. Without the cooperation and
assistance of Piedmont Research Station Managers Joe Hampton and Ray Horton and their capable staff,
this project would not have been possible. Their service is commended, and their contributions are
greatly appreciated.
This project integrated research from several disciplines to evaluate the effects of land application
of biosolids on air and volatile emissions and soil microbial characteristics. Measurements included
chemical, physical, and microbiological analytes. Contributors to the integrated project planning included
Eric A. Foote from Battelle; Carolyn M. Acheson, Edwin F. Barth, Richard C Brenner, Paul R. de Percin
(retired), D. Bruce Harris (retired), Ronald F. Herrmann, Mark C. Meckes, James E. Smith, Jr. (retired),
Laurel J. Staley, Stephen L. Wright, and Lawrence D. Zintek from EPA; and Laura L. McConnell and
Patricia D. Millner from the U.S Department of Agriculture (USDA).
The following authors, utilizing their experience and expertise, analyzed, organized and
interpreted the large quantity of data collected during this study and assessed the peer-reviewed literature
for its relevance to this project. The technical interpretations of this report are based on their efforts,
which are gratefully acknowledged.
Executive Summary - Eric A. Foote (Battelle), Carolyn M. Acheson (EPA), and Richard C.
Brenner (EPA) with assistance from Laurel J. Staley (EPA), Mathew Morrison
(previously with EPA), and James E. Smith, Jr. (EPA-retired)
Sections 1, 2, 3, and 4 - Eric A. Foote (Battelle) and Richard C. Brenner (EPA) with assistance
from James E. Smith, Jr. (EPA-retired)
Section 5 - Edwin F. Barth (EPA), Eric A. Foote (Battelle), and Patricia D. Millner (USDA) with
assistance from Richard C. Brenner (EPA)
Section 6 - Edwin F. Barth (EPA), D. Bruce Harris (EPA-retired), Eric A. Foote (Battelle),
Laura L. McConnell (USDA), and Robert H. Forbes, Jr. (CH2M Hill) with assistance
from Richard C. Brenner (EPA) and Eben D. Thoma (EPA)
Section 7 - Carolyn M. Acheson (EPA), Ronald F. Herrmann (EPA), Eric A. Foote (Battelle),
and Steven J. Naber (Battelle) with assistance from Richard C. Brenner, Tracy A.
Dahling, Sally J. Stoll, Troy J. Strock, Stephen L. Wright, and Lawrence D. Zintek
from EPA; Melody J. Graves-Allen from Battelle; and J. Lee Heckman, M.
Jacqueline Tompkins, and Susan S. VonderHaar from Shaw Environmental and
Infrastructure
Section 8 - Carolyn M. Acheson (EPA), Richard C. Brenner (EPA), and Eric A. Foote (Battelle)
xm
-------
The Quality Assurance Project Plan (QAPP) was prepared by the following Battelle staff: Eric A.
Foote (leader), Melody J. Graves-Allen, Gregory L. Headington, Shane S. Walton, Steven J. Naber, and
Michael W. Holdren. The QAPP was reviewed by several EPA and USDA staff under the guidance of
LRPCD Quality Assurance Manager Scott A. Jacobs, who was responsible for final approval, quality
assurance (QA) auditing, QA oversight in the field, and QA review of work products such as this report.
A variety of field preparation activities including, among others, site layout, sampling and
auxiliary equipment fabrication and assembly, preliminary characterization sampling and analysis, and
temporary on-site trailer and laboratory deployments were critical to the conduct of this project. These
activities involved many individuals from the staffs mentioned above. Special thanks are due to Gregory
Headington, Shane S. Walton, and Melody Graves-Allen from Battelle; Tracy A. Dahling, Paul T.
McCauley, Kim A. McClellan, and Sally J. Stoll from EPA; Patricia D. Millner from USDA; and Susan
Boutros from Environmental Associates, Ltd.
Battelle staff organized and directed the sequence of events necessary to carry out this
intensive week-long field effort as well as the pre- and post-post application soil sampling studies.
Approximately 50 individuals from several organizations were involved on site in some capacity in this
complex, multi-faceted project. The contributions of the following Battelle personnel are acknowledged
and appreciated in orchestrating the orderly conduct of the many tasks required to bring this field study to
a successful conclusion: Eric A Foote (project leader), Gregory L. Headington (field staff leader), Melody
J. Graves-Allen, Shane S. Walton, Gary B. Carlin (Safety Officer), Hali L. Jackson, Lincoln G. Remmert,
Jan R. Satola, and R. Michael Woolfe.
Vital program support and advice have been provided by EPA senior leadership from the outset
of this project in dealing with policy issues that evolved during the planning and implementation stages of
the multi-discipline effort. The cooperation and support of the following EPA administrative personnel
are sincerely appreciated and acknowledged: Sally C. Gutierrez (Director, NRMRL), Hugh McKinnon
(retired Acting Director, NRMRL), Lawrence W. Reiter (former Acting Director, NRMRL), Benjamin L.
Blaney (retired Assistant Laboratory Director for Water, NRMRL), Robert A. Olexsey (retired Director,
LRPCD, NRMRL), Annette M. Gatchett (former Acting Director, LRPCD, NRMRL), Patricia M.
Erickson (former Acting Director, LRPCD, NRMRL), Robert K. Bastian (Office of Wastewater
Management, Office of Water), Richard R. Stevens (Office of Science and Technology, Office of Water),
and Patricia L. Schultz (retired Chief, Cincinnati Technical Communications and Outreach Staff).
The draft report for this project was subjected to external peer review by three acknowledged
experts in land application of biosolids and bioaerosol and soils sampling methods and approaches.
These three scientists were Dr. Charles Gerba, University of Arizona; Mr. Greg Kester, California
Association of Sanitation Agencies; and Dr. Jordan Peccia, Yale University. Solicitation of reviews and
collection and submittal of review comments were conducted under Contract with Versar, Inc.,
Springfield, VA via Contract No. EP-C-07-025, Task Order 72. Coordination of activities with Versar
and oversight of staff review of, response to, and reconciliation of peer review comments was carried out
under the leadership of EPA Branch Chief Laurel J. Staley.
In order to ensure community input to the development and conduct of this project, multiple
agency personnel acted as liaisons with the public through the Information Sharing Group (ISG). The
ISO's primary areas of interest have revolved around the potential impact of biosolids land application on
adjacent neighbors and landowners and methods chosen by the research team to evaluate the capture,
detection, identification, and quantification of chemical and biological constituents of biosolids released
to the environment. The ISG has been invited to review and comment on important documentation
produced during this study including the QAPP, data summaries, draft copies of the Executive Summary
from this report, and finally this report in its entirety. The comments provided by the ISG have been
xiv
-------
utilized, where appropriate, to improve the overall objectivity of and scientific approach to carrying out
this project.
The ISG was co-founded by John Walker (retired Biosolids Team Leader, EPA, Office of
Wastewater Management, Office of Water), Richard Giani (former Biosolids Coordinator, Pennsylvania
Department of Environmental Protection [PADEP]), and Patricia D. Millner (USDA). John Borland
(PADEP) provided technical and administrative assistance during the formative stages of the ISG. Laurel
J. Staley served as the EPA technical liaison with the ISG throughout the lengthy process of project and
work plan development, field study implementation, and report preparation and review. The collaborative
efforts of the following ISG members are acknowledged with appreciation: Ned Beecher (ISG
Coordinator, North East Biosolids and Residuals Association), Tom Albert (veterinarian and private
citizen, Laurel, MD), Ellen Harrison (retired, Cornell Waste Management Institute), Frank Hearl (Center
for Disease Control, NIOSH), Ronald Liebert (California Farm Bureau Federation), Albert Medvitz
(rancher and private citizen, Rio Vista, CA), Christopher Piot (District of Columbia Water and Sewer
Authority), Jay Snider (Wastewater Treatment Plant Operator, Borough of Ephrata, PA), and Henry
Staudinger (lawyer and Citizen's Representative of Virginia Biosolids Use Advisory Committee).
xv
-------
EXECUTIVE SUMMARY
The goal of this research study was to evaluate air and soil sampling methods and analytical
techniques for commercial land application of biosolids. Biosolids were surface applied at agronomic
rates to an agricultural field. During the period of August 2004 to January 2005, 35 groups of analytes
were measured using 13 sampling techniques. Several analytes were measured in more than one matrix.
For example, fecal coliforms were measured in biosolids, air, and soil. In total, 49 analyte-matrix
combinations were measured. The multimedia approach and numerous analyte-matrix combinations are
unique for a field study on the land application of biosolids. For 27 combinations, data met quality
criteria, and interpretation used conventional methods. Quality assurance (QA) criteria were not met, or
QA data were not reported for 12 combinations. The interpretation of these data sets was affected by QA
limitations, and conclusions from these data are more uncertain. No detections were observed for 10
microbial analytes. It is not clear if organisms were present but not detected or were absent.
In this study, odors were detected in the air, and chemicals and microbes were measured in the
soil after land application of biosolids. Odors had dissipated after 4 days. In shallow soils, most
microbial and chemical analytes remained elevated for the remainder of the study, 98 days. The
conclusions of this study may have been affected by the biosolids applied, weather conditions during and
after application, and other site-specific variables. Additional studies would be useful to determine if
these observations are consistent with other biosolids applications. Based on the results of this study, the
27 analyte-matrix combinations yielding usable data have been demonstrated at field scale and could be
used in future research within the QA context of this study. For the other analyte-matrix combinations,
additional QA samples, screening of analytical labs for compliance with QA, and continued methods
development are needed. This research, in combination with the work of others, may result in the
development of an integrated, multimedia protocol for use in field sampling of biosolids land application.
Background
In the United States, about 60% of biosolids, the solid residues produced by wastewater
treatment, are applied to land as an agricultural amendment. Many believe that biosolids application is a
beneficial use of this material. In 1993 under the Clean Water Act, the U.S. Environmental Protection
Agency (EPA) issued regulations governing land application of biosolids, commonly referred to as the
Part 503 Rule. Biosolids are defined as sewage sludge that has been treated to meet federal and state
regulations for land application. In the years since the regulations were issued, wastewater treatment
technologies and practices have changed and public concerns about the land application of biosolids have
grown. In 2002, the National Research Council of the National Academy of Science issued a report
entitled: "Biosolids Applied to Land: Advancing Standards and Practices " (NRC, 2002). The report
noted that no scientific evidence documented that the Part 503 Rule had failed to protect public health and
recommended additional scientific work to reduce uncertainties about the potential for adverse human
health effects from exposure to biosolids.
Motivated by this report and other research questions, a collaborative research team under the
leadership of the EPA's Office of Research and Development was assembled. A field-scale land
application study was undertaken to evaluate sampling methods and analytical techniques. The major
objective of this study was to screen many of the available methods for applicability and included: four
environmental matrices (air emissions, airborne particles, soil, and biosolids); 35 analyte groups; and 13
sampling methods. Air samples were measured before, during, and after application for volatile
compounds, odorants, microorganisms, and endotoxins as well as their short-range transport. Airborne
particulate levels were monitored. Microbial and chemical concentrations were determined for soil
samples before and after biosolids application.
xvi
-------
Study Design
This study was conducted at the North Carolina Department of Agriculture and Consumer
Services Piedmont Research Station in Salisbury, NC. Class B biosolids were surface applied to a fescue
field in a 100-m diameter circle by a commercial side discharge manure spreader at a target rate of 10 wet
tons/acre. Biosolids had not been applied to this land previously. Monitoring began before application in
August 2004 and continued through January 2005.
The sewage sludge used in this study was anaerobically digested, dewatered by centrifugation,
and treated with lime. Polymer was added during sludge treatment. This type of sludge treatment is
commonly used in wastewater treatment plants and is likely to produce biosolids with detectable odors
and generate aerosolized particulates via surface drying, flaking, and wind erosion during land
application. However, the biosolids delivered to the site were sticky, cohesive, and tended to rapidly
settle onto the ground in agglomerated clumps. This characteristic visually impacted the distribution of
the biosolids on the soil surface and may have introduced unplanned variance into soil/biosolids sample
collection.
Key Research Results
Biosolids
•\9
The biosolids used in this study had a solids content of 28% and contained 2.3 x 10 colony
forming units (CPUs) of fecal coliforms/g dry weight (gdw) total solids and 6.33 most probable number
(MPN) Salmonella spp ./gdw total solids. Several microbial characterizations were measured including
total heterotrophic bacteria (THE) at 1.6 x 1011 CFUs/gdw total solids, Escherichia coli at 4.35 x 1Q7
MPN/gdw total solids, total coliforms at 1.4 x 109 CFUs/gdw total solids, and Enterococcus spp. at 8.2 x
105 CFUs/gdw total solids. Samples were analyzed for Staphylococcus aureus, but none were detected.
Airborne Particles
Airborne particles were collected using impingers, impactors, and GRIMM samplers. These
samples were analyzed for microbial analytes and particulate mass. Two types of microbes, THE and
fungi, were detected during both the control trial and the biosolids application test, especially at sampling
points near the spreader. However, no specific bacterial pathogens (i.e., E. coli, Salmonella spp., S.
aureus, Clostridiumperfringens, and Enterococcus spp.), indicator microorganisms (i.e., fecal coliforms
and coliphage), or enteric viruses were detected. Organisms may have been present but not detected.
Standardized and more robust bioaerosol samplers and field QA samples may clarify this question.
Bacterial endotoxin samples were collected; however, due to operational and QA problems, data were
unusable. The mass of particulates <5.0 urn was statistically similar in samples collected immediately
before and during biosolids application. Based on these results, the mass of bioaerosol particles was not
changed during biosolids land application.
Air
Odors were monitored in the field for up to 4 days following application using Nasal Rangers®,
flux chambers, and off-site odor analysis. Nasal Rangers® are useful for detecting odors compared to
background levels; they do not identify specific chemicals. In the near-field application area, Nasal
Rangers® detected odor at approximately twice background levels for up to 2 days after application. By
Day 4, odors were not detected in the near field. Odor analysis was also conducted by sampling the
exhaust vents of flux chambers using Nasal Rangers® and off-site analysis by an odor panel. The flux
chamber temperatures were higher than ambient temperatures due to radiant heating. This situation was
xvii
-------
not anticipated and may have compromised observations. Exhaust gas resulted in Nasal Ranger® odor
detections at about twice background on Day 1, which dissipated to background levels on Day 2. Off-site
odor panel analysis resulted in an approximate five-fold increase from control levels to levels
immediately following application. Comparing the results for the control trial to 22 hr after application, a
total increase of about 50-fold was observed by the off-site odor panel. Concurrent with increased
temperatures, elevated levels of volatile organic sulfur compounds were observed in flux chamber
samples and may explain the increased odor observations.
Chemical measurements were made in conjunction with odor monitoring and included Draeger
tubes (for ammonia), a Jerome analyzer (for hydrogen sulfide), and vertical radial plume mapping
analysis using open-path (OP) Fourier transform infrared (FTIR) spectrophotometry. For all air samples,
ammonia was measured at less than 1 ppm ammonia by Draeger tubes. Near-field OP-FTIR plume
mapping following application showed an exponentially decreasing emission flux rate, initially detected
at 0.063 g/s and completely dissipating by 22 hr after application. Hydrogen sulfide concentrations were
similar during the control trial and biosolids application days. Samples of flux chamber exhaust from the
day of biosolids application contained 15 ppmv and 0.16 ppmv of ammonia and hydrogen sulfide,
respectively. Flux chamber samples for other days were below detection for ammonia and hydrogen
sulfide.
Soil
Soil samples were collected prior to biosolids application and for 4 months following application.
Soil measurements included: fecal coliforms as an indicator of pathogenic organisms; phospholipid fatty
acids to characterize the total biomass present as well as the microbial community structure; alkylphenol
ethoxylates (APEs) and metabolites such as octylphenol (OP) and nonylphenol (NP) that are potential
endocrine disrupting chemicals; and soil toxicity to plants and earthworms. Soil was sampled as deep as
25 cm for several analytes, but at this site, the 0-5 cm depth proved to be the most important in
understanding the potential effects of biosolids. Supporting information such as soil agronomic,
temperature, biosolids distribution, and weather data were also gathered.
Supporting information is useful in placing results from this study in the context of similar
research. For this study, rainfall was plentiful. As a result, soil was near moisture holding capacity
throughout the study, and desiccation was unlikely to affect soil microbes. Soil temperatures varied over
the sampling period from as high as 28 °C in August to 3°C in January. It is possible that cooler
temperatures, especially during the last month of the study, affected observations. Reduced microbial
die-off and lower aerobic degradation rates are often observed at lower temperatures and high moisture
levels. Measuring the quantity of biosolids after application was an easy method to document the
application rate, 7.3 to 9.5 wet tons/acre or 1.7 to 2.2 dry tons/acre, which was close to the planned rate.
At this site, land application of biosolids altered microbial and chemical soil concentrations at
shallow depths. Total biomass, fecal coliforms, and NP and OP displayed increases following application
that generally persisted for the 98-day post-application soil sampling period. Total biomass increased by
a factor of two. Fecal coliform measurements exhibited an increase of more than 100-fold between pre-
and post-application samples. However, because the laboratory reported semi-quantitative results for
48% of the samples, this finding is uncertain. Although APEs were not detected in the soil at any time,
after biosolids application, the metabolites OP and NP were detected at median concentrations of 5,400
and 215 ug/kg dry solids, respectively.
Some measurements showed transient changes or no change after biosolids application. For
example, the microbial community changed initially after application but returned to its pre-application
structure within 28 days. Enteric viruses, Salmonella spp., and viable Helminth ova were observed in the
xvin
-------
biosolids and in 20% of the soil samples throughout the study. Following biosolids application, soil
toxicity exhibited no changes in earthworm mortality and seed germination while root length data were
different for the two species tested. Since a limited set of ecotoxicity assays was used, it is difficult to
draw broad conclusions from this dataset.
Recommendations
Additional studies would be useful to determine if the discussed observations are consistent with
other types of biosolids, especially considering their nature and composition; site specific factors
including soil type, types of plants, and ambient bioaerosol levels; and differing weather conditions such
as wind directions and speeds. OP-FTIR techniques were useful for tracking the extent and longevity of
the plume. For future air sampling, data interpretation would benefit from an experimental plan that made
greater use of particulate matter sampling, meteorological monitoring, and air dispersion modeling. In
addition, the collection and operational methods used to sample bioaerosols in this study would benefit
from advancements such as the use of indicator organisms for positive controls and improved collection
of sensitive organisms. Due to radiant heating, unanticipated elevated temperatures were observed in the
flux chambers and compromised the volatile chemical and odor data. Improved flux chamber methods
are needed to produce data representative of field conditions. Future soil sampling efforts should evaluate
longer sampling periods and expand the chemical analyte list. Improved sampling procedures, such as
normalizing concentration with a biosolids-specific chemical, may reduce sample-to-sample variability
and thus increase confidence in conclusions. Pre-screening of labs analyzing microbial samples to
demonstrate QA compliance would be useful for future studies.
xix
-------
1.0 PROJECT DESCRIPTION AND OBJECTIVES
1.1 Background and Introduction
The historical approach taken by the U.S. Environmental Protection Agency (EPA) in managing
municipal wastewater treatment plant residuals or sewage sludge is largely based on the 1979 Regulation:
Criteria for Classification of Solid Waste Disposal Facilities and Practices (EPA, 1979) as modified by
the 40 CFR Parts 257, 403, and 503 Rule(s): Standards for the Use or Disposal of Sewage Sludge (EPA,
1993). The Part 503 Rule specifies standards for the treatment of municipal sewage sludge to be applied
on land. When sewage sludge has been treated to meet federal and state regulations for land application,
the resulting material is commonly called biosolids. The Part 503 Rule sets limits for land application of
biosolids based on metals concentrations and/or loadings, disinfection for reduction of pathogens, and
vector attraction reduction (for example, volatile solids destruction for anaerobically digested sludge).
In the years following issuance of the Part 503 Rule, the land application of biosolids has become
a subject of controversy. While some view this practice as a beneficial use of biosolids, others are
concerned by the practice. Anecdotal reports of illness from residents near some biosolids land
application sites have been made. At the request of EPA, the National Research Council (NRC) of the
National Academy of Science evaluated regulatory requirements and non-regulatory measures with
respect to land application of biosolids and provided an independent review of the technical basis of the
chemical and microbial contaminant regulations for biosolids land application as it pertains to human
health. In July 2002, NRC completed an 18-month study and issued a report entitled Biosolids Applied to
Land: Advancing Standards and Practices (NRC, 2002). NRC did not investigate individual reported
health incidents as part of its study. They did search the published scientific literature for evidence of
human health effects related to biosolids and found that no scientific studies of the alleged health
incidents had been published at the time of the review. Hence, in the "Overarching Findings" section of
their report, the NRC team stated: "There is no documented scientific evidence that the Part 503 Rule has
failed to protect public health. However, additional scientific work is needed to reduce the persistent
uncertainty about the potential for adverse human health effects from exposure to biosolids. There have
been anecdotal allegations of disease, and many scientific advances have occurred since the Part 503 Rule
was promulgated. To assure the public and protect public health, there is a critical need to update the
scientific basis of the Rule to: 1) ensure that the chemical and pathogen standards are supported by
current scientific data and risk assessment methods, 2) demonstrate effective enforcement of the Part
503 Rule, and 3) validate the effectiveness of biosolids-management practices" (NRC, 2002).
After careful study of the NRC report and current regulations, and with input from all relevant
stakeholders, EPA responded with a 14 Point Action Plan that had a goal of strengthening the beneficial
use and disposal program for municipal wastewater treatment plant residuals (Smith and Stevens, 2010).
The Action Plan (http://federalregister.gOv/a/03-32217) included such activities as: developing methods
for microbial pollutants such as Ascaris ova, viruses, fecal colifroms, and Salmonella; developing
analytical methods for pharmaceuticals and personal care products; conducting a targeted national sludge
survey for pollutants in biosolids; conducting field studies applying biosolids to land; participating in
meetings on incident tracking, exposure measurement, and sustainable land application; reviewing the
criteria for molybdenum in land-applied biosolids; and assessing available tools and methodologies for
conducting microbial risk assessments on pathogens. EPA has made significant progress on many Action
Plan activities and continues to address these and other activities.
This report details a field research project designed to: 1) evaluate multimedia sampling methods
and techniques prior to, during, and following the application of biosolids to agricultural grassland; and
2) address one of the activities (Field Studies of Application of Treated Sewage Sludge) from EPA's
-------
Action Plan response to the NRC report. The multimedia sampling techniques included methods for
measuring various components contained in the applied biosolids, entrained in aerosol emissions,
discharged into the air as volatile and semi-volatile gases and odorants, and collected from the soil surface
and subsurface of the applied grassland. This report documents the results of this study. This research
was not designed to investigate health-related incidents and, therefore, does not constitute a health effects
research study.
This report documents the approach, methodologies, results, and interpretation of a collaborative
research study conducted by EPA's Office of Research and Development (ORD), National Risk
Management Research Laboratory (NRMRL), Land Remediation and Pollution Control Division
(LRPCD); the U.S. Department of Agriculture (USDA); the North Carolina Department of Agriculture
and Consumer Services (NCDA&CS); Battelle; and other supporting groups and organizations. The
target audience for this report includes EPA's Office of Water; Regional and State Biosolids
Coordinators; and engineers, scientists, and consultants active in the wastewater treatment field, as well as
non-traditional stakeholders such as citizens' groups.
The study began in autumn, a typical application time for biosolids in the Southeastern portion of
the United States. Routine agronomic practices were utilized. In addition to evaluating sample collection
and analysis methodologies, related environmental and other conditions associated with the test
application were measured and/or monitored.
1.2 Project Goal
The goal of this research study was to investigate air and soil sampling methods and approaches.
Ultimately, this research along with the research of others may lead to the development of a protocol that
could be used in future studies to obtain data on the release of airborne and soil-bound contaminants
during the application of biosolids on land. The air was sampled for selected constituents of particulates,
microbes, and volatile compounds. Air particulate samples were analyzed for endotoxin. Air samples
were tested for indicator organisms and several pathogens and volatile compounds including malodorants.
The biosolids applied and the soil to which they were applied were also analyzed.
The study measured air emissions, their short-range transport, and soil microbial concentrations at
and around the test site during biosolids application, with a focus on qualitative and quantitative
characterization of the items described above. Soil microorganisms were evaluated based on the general
community and specific classes including fecal coliforms, Salmonella spp., viable Helminth ova (VHO),
and enteric viruses. Microorganisms enumerated from biosolids and air samples included fecal coliforms,
total heterotrophic bacteria (THE), E. coll, Salmonella spp., Enterococci, Staphylococcus aureus,
Clostridiumperfringens, male-specific coliphage, and enteric viruses.
The multimedia approach that was used for the collection and analysis of air emissions at this test
site was unique in comparison with other projects in this area of study, i.e., others focus on one or more
components in individual classes of emissions (microorganisms, volatile organic compounds [VOCs], or
odors) independently of each other. Data gained from this project constitute a landmark set of
simultaneous multimedia information (qualitative and quantitative) associated with the application of
biosolids on land and will be used to further development of method protocols for sampling at other land
sites where biosolids are applied.
-------
1.3 Project Objectives
To achieve the goal stated in Section 1.2, the research team implemented three discrete tasks in
the field as described below, each with project-specific objectives. For each of these tasks, a primary
objective (PO) was identified.
1.3.1 Task 1. Bioaerosol and Particulate Matter Sampling. Selected bacteria, viruses, bacterial
endotoxins, and particulates were analyzed in samples taken of aerosol emissions from the biosolids pile
and from the field prior to, during, and after biosolids application. Section 5.0 describes in detail the
microorganisms monitored and the particulate matter analyses conducted. The PO for bioaerosol
sampling was to characterize the type and concentration of the suite of viable bioaerosol components
(seven bacteria, culturable enteric viruses, endotoxin, and male-specific coliphage) emitted and
transported to several downwind sampling stations.
1.3.2 Task 2. Volatile Organic Compound (VOC) and Odor Sampling. A select group of VOCs
and odorous compounds was monitored in emissions emanating from the application area prior to, during,
and following the application of biosolids. These compounds were measured using a variety of
instrumentation. Section 6.0 describes in detail the measurements of VOCs and odorous compounds.
The PO for Task 2 was to determine the presence and concentration of selected VOCs and odorants in
samples collected during field application, within the application area, and downwind of the application
area.
1.3.3 Task 3. Land Sampling. The PO of the land sampling effort was to measure the concentration
of microorganisms in biosolids applied to land in the test site and soil directly below the biosolids over
time and to screen the toxicity of the biosolids. In addition, the soil concentrations of nonylphenol
ethoxylates, nonylphenol, octylphenol, and Bisphenol A (BPA) were also measured. In Section 7.0, a
detailed description of the land sampling methods and analytes used to accomplish the PO is available.
-------
2.0 EXPERIMENTAL APPROACH AND TEST SITE SETUP
The overall test site design and approach used in setting up the field experiments are described in
this section along with the details of the site layout and the sampling techniques employed in the field.
2.1 Experimental Approach
This research was conducted on the property of the NCDA&CS Piedmont Research Station in
Salisbury, NC. The NCDA&CS provided land, personnel, and facilities in assisting with this research
effort.
The test site was designed to support the eventual development of a field sampling protocol by:
1) gathering data on the emission concentrations of chemical and biological parameters prior to, during,
and after biosolids land application at various upwind and downwind locations on the site, and
2) monitoring the persistence of selected microorganisms and chemicals on land. Atmospheric conditions
were recorded during the application process as they were expected to have an impact on emission
dynamics. Airborne microbes and particulates that may be associated with the application of biosolids
were collected in bioaerosol samples.
For the purposes of this study, a 100-m diameter circle (~2-acre area) served as the focal point for
this research and the location where the biosolids were applied. Various sampling and monitoring
activities were conducted prior to, during, and after the application. These activities included:
• Bioaerosol monitoring
• VOC sampling including the measurement of emissions using flux chambers; open-path
Fourier transform infrared (OP-FTIR) spectrometer measurements; optical remote sensing
(ORS); and the determination of ammonia, hydrogen sulfide, and odorant concentrations
• Land sampling activities including physical/chemical soil properties analysis, specific
chemical analysis, microbiological analysis, and ecotoxicity testing.
Hydrogen sulfide and ammonia measurements were conducted with hand-held monitors, and
odorants were monitored using Nasal Rangers® and in-lab analyses of vapors captured during flux
chamber sampling. These measurements, along with the bioaerosol monitoring program, were conducted
within and around the 2-acre area. Land sampling activities were only conducted within the 2-acre area
where biosolids had been applied. An aerial view of the test site and application area is shown in
Figure 2-1. This figure also denotes the global positioning system (GPS) locations of the various
sampling deployments and sample collection areas within and immediately adjacent to the application
area (identified by the circle).
Semivolatile organic compounds (SVOCs) were monitored because of the speculation that many
chemicals are released into a wastewater treatment facility's collection system that could potentially
accumulate in the biosolids. For example, organic compounds such as brominated flame-retardants have
been found to leach from biosolids into the environment (Anderson and MacRae, 2006). Inorganic
compounds, such as ammonia and hydrogen sulfide, are also commonly found in biosolids and were,
therefore, monitored during this study.
-------
Upwind Bioaerosol Stations
Land Sampling Test Plots
Nasal Ranger Detection Points
Flux Chambers
Downwind Bioaerosol Stations
Application Area (Center)
Barn Used for Staging
OP-FTIR Vertical Lifts
Biosolids Stock Pile
Meteorological Monitoring Station
Figure 2-1. Aerial View of the Test Site and Application Area
Total participates were monitored because of the possibility that small particulates from biosolids
material may be suspended in the air during land application. It has been speculated that adjacent
landowners can come in contact with these particles as they are transported off site (downwind).
In addition, it has also been speculated that pathogens, viruses and endotoxins can be adsorbed
onto suspended particulates and perhaps come into contact with the nearby human population; therefore,
-------
emissions were sampled using an array of bioaerosol sample collection equipment upwind and downwind
of the experimental test site prior to and during the application of the biosolids.
Testing for odors and their intensities occurred prior to, during, and up to 2 days after the
application of biosolids using hand-held olfactometers. In addition to these on-site analyses, in-lab
analyses of vapor emissions captured during flux chamber sampling were analyzed to determine odor
threshold.
2.2 Site Design
The test site was designed around a 100-m diameter circle so that an array of upwind and
downwind sampling units could be moved (see Figure 2-2) around the circle during sampling to
accommodate a major shift in wind direction should one occur. Sampling locations were positioned on
four discrete lines that ran parallel to each other. The upwind line, designated UWA (Upwind Location
A), was positioned 16 m above or northeast of the upwind perimeter of the circle. The first downwind
line, DWA (Downwind Location A), was positioned such that it transected the center of the circle. The
remaining two lines were positioned outside the downwind perimeter of the circle; the first, which was
designated DWB (Downwind Location B), was 16m from the circle perimeter; the second, designated
DWC (Downwind Location C), ran parallel to DWB and was an additional 34 m downwind.
2.3 Task 1. Bioaerosol and Particulate Sampling Design
The sampling array was set up such that three sampling stations were positioned with equal
spacing on UWA, DWB, and DWC. An additional station was placed in the center of the circle (DWA),
and a mobile sampler (MOB) followed immediately behind the applicator. The MOB was considered to
be a capture point for particulates and aerosols in a worse-case scenario. This sampling array design
resulted in a total of 11 sampling locations, one of which was constantly mobile.
Figure 2-2 depicts the site layout in respect to bioaerosol and particulate sampling collection
activities. At each of the nine stationary locations on sampling lines UWA, DWB, and DWC; two SKC
BioSamplers® (hereafter referred to as biosamplers); one Andersen six-stage impactor sampler (hereafter
referred to as a six-stage impactor); and two SKC Button Samplers (hereafter referred to as button
samplers) were used to collect air samples over the duration of the application. The mobile unit was
equipped with the same five samplers plus an additional six-stage impactor. The mobile unit was
deployed such that it was beyond the wake of the applicator yet within the boundary of airflow of the
suspended particles during application. The direction of the applicator relative to the position of the
mobile unit was carefully planned and is discussed in the following section. The center station consisted
of a GRIMM particle sampler for collecting particulate matter.
2.3.1 Sampling Station Design
2.3.1.1 Transect Stations. Although the 10 transect stations were intended to be stationary, nine of the
10 stations (all except the center station) were constructed such that they could be readily moved should a
major wind shift occur during sampling, and also so that they could be quickly and easily deployed into,
and removed from, the test area using minimal personnel. In order to meet these requirements, portable
sampling systems were designed and built on heavy duty garden carts. Each cart was equipped with a
Honda 2000EU generator to supply electrical power for the air samplers, a six-stage impactor, a Quick
Take 30 pump, two biosamplers with Vac-U-Go vacuum pumps, and two button samplers with pumps.
-------
Potential Downwind Receptors
Explanation
Upwind
(UW)
Downwind
(DW)
34.8 m
NOT TO SCALE
A 2 Impinger Biosamplers + 2 Button Filter Samplers (Endotoxin) + 1 Impactor Sampler
*
< A> Mobile": 2 Impinger Biosamplers + 2 Button Filter Samplers (Endotoxin) + 2 Impactor Samplers
D Grimm Sampler
Note
* Samplers were mounted on the front of an all-terrain vehicle (AW). The ATV was driven such that it
collected samples at a constant distance from the release point of the applicator.
Figure 2-2. Bioaerosol and Particulate Sampling Array
SAMPLEGRID06_02.CDR
-------
Figure 2-3. Cart-Mounted Transect Sampling Equipment
Each sampling system was mounted on a tripod with a platform on which the samplers were
positioned. The biosamplers were set up so that the biosampler orifice was located 1.5m above the land
surface. The six-stage impactor and duplicate button samplers were positioned at the same height on the
mast. The completed unit is shown in Figure 2-3.
2.3.1.2 Mobile Unit. The mobile sampling station (Figure 2-4) consisted of a four-wheeled, all-terrain
vehicle (ATV) equipped with two biosamplers, two six-stage impactors, and two button samplers with
ancillary pumps and equipment. The mobile sampler was set up in the same mode as the transect
samplers with the exception that it was outfitted with a second six-stage impactor.
2.3.1.3 Center Station. The center station consisted of a GRIMM sampler for collecting particulate
matter. This type of sampler allows fractionation of particle size ranges.
2.4 Task 2. VOC, Ammonia, Hydrogen Sulfide, and Odor Sampling
The experimental design also included selected in-field and in-lab measurements of VOCs,
ammonia, hydrogen sulfide, and various odorants. These measurements were collected in parallel with
other monitoring activities being conducted for Task 1.
-------
Figure 2-4. Mobile Unit
2.4.1 Open Path Fourier Transform Infrared Spectrometer. An OP-FTIR spectrometer was used
to measure "real-time" concentrations of VOCs and ammonia in air. The system consisted of an FTIR
spectrometer and a retro-reflector array. The OP-FTIR system was linked to a particle counter that
summed the total amount of energy that a target compound absorbed between the FTIR and retro-reflector
array. Concentrations of specific compounds were quantified using the measurement of energy absorbed
within selected regions of the spectrum. Figure 2-5 shows the FTIR and retro-reflector positioning used
for this research. A combination of single-path and vertical radial plume mapping (VRPM)
measurements was conducted. The equipment was maintained in these positions for the duration of the
sampling event described in the following sections.
2.4.2 Advective Flux Measurements. Flux chambers were used to estimate the rate and extent of
volatile emissions relative to a known surface area on the ground. Figure 2-6 illustrates the flux chamber
design that was used for this research. Each flux chamber was square, 120 cm on a side. One chamber
was deployed in each quadrant of the circular test area with one duplicate chamber placed in one of the
quadrants, for a total of five chambers. Each chamber was plumbed with a manifold assembly that was
designed to capture a total of up to five discrete samples for VOC and odorant analysis.
Each chamber had an open bottom that was driven approximately 6 cm into the ground surface.
The unit exhausted through an open stack that was covered loosely with aluminum foil to prevent any
downdraft air from entering the inside of the chamber during sampling. Air samples were pulled into
Tedlar®bags via a 12-volt miniature diaphragm pump (Gast Model # 10 D 1152-101-1052) powered by a
26-amp hour sealed battery. The pump pulled the sample from a 6.3-mm OD Teflon" tube that was
inserted into the top of the flux chamber. The sample was drawn through the pump and into an Aalborg
Instruments multi-tube flowmeter. The flowmeter contained a single inlet and a total of five
independently controlled outlets. Each of the flowmeter outlets was plumbed with a 4.8-mm OD Teflon"
-------
Explanation
Retrroreflector
I Vertical Tower
Figure 2-5. Map of the Site Layout Showing the Location of the Vertical Radial Plume
Mapping Configurations and the Single-Path Measurements
tube that connected to a total of up to four Tedlar®bags with a stainless steel tube connector. The Tedlar®
bags were sized larger than the target collection volumes. For each unit, three 5-L bags were used to
collect samples for solid-phase microextraction (SPME) analysis and one 10-L bag was used to collect
samples for odor threshold analysis (SKC Tedlar®bags, 5 L #231-05 and 12 L #231-10).
In addition, a 5-L Summa canister (under vacuum) was attached to the unit to collect samples for
EPA TO15 analysis. The sample rate was controlled by attaching a fixed orifice to the inlet of the
canister that allowed for the collection of a time-integrated sample.
Make-up air consisted of a pressurized cylinder of ultra-high purity (UHP) air that was metered
into the base of the chamber by a rotometer. The influent and effluent flow rates were set equally to
reduce the possibility of a pressure drop that would cause an influx of outside air through the stack of the
chamber. The UHP air was distributed evenly inside the chamber using a circular distribution manifold
that consisted of a perforated piece of Teflon® tubing. Each flux chamber was equipped with a
thermocouple and HOBO datalogger so that temperature could be recorded continuously within the
chamber.
2.4.3 Ammonia and Hydrogen Sulfide Measurements. In-field monitoring was conducted for
ammonia and hydrogen sulfide concentrations at various locations in and around the experimental site.
Ammonia and hydrogen sulfide measurements were acquired using a hand-held single sensor gas
detector. These data were used to develop a qualitative "footprint" for ammonia and hydrogen sulfide
about the site prior to, during, and after the application process.
10
-------
Flux Chamber
Exhaust
A
Thermocouple
NOT TO SCALE
"UHP = Ultra High Purity
• • .'.i ^
Figure 2-6. Flux Chamber Design
2.4.4 Odor Measurements. Odor, which is a primary mechanism by which the public typically
becomes aware of air emissions from biosolids processing and application, was measured using two
methods: in-field olfactometry (Nasal Ranger®) and in-laboratory dynamic dilution olfactometry analysis.
The latter was conducted using the 12-L Tedlar®bags that collected air emissions produced by the flux
chambers described above. These analyses are discussed in more detail in following sections of this
report.
2.4.5 In-Laboratory Biosolids Measurements
2.4.5.1 SVOC Analysis of the Biosolids. The unapplied (stockpiled) and applied biosolids were
collected and used to estimate the concentration of SVOCs in each sample. This work was conducted by
Severn Trent Laboratories under contract to Battelle. Modified EPA Method SW-846 8270 was used to
identify and determine the concentration of selected pesticides, polybrominated diphenyl ether (PBDE),
and selected poly chlorinated biphenyl (PCB) congeners.
2.4.5.2 Headspace Analysis. The headspace method of measurement was used to estimate the
equilibrium concentration and type of VOCs emitted from contained grab samples that were collected
from the experimental test site. Emissions in the gas space (headspace) above a containerized composite
sample were measured using two analytical methods.
One method contained the sample in a specialized bottle that was equipped with a sampling side
port for withdrawing the sample into a syringe for direct injection into a gas chromatograph. This sample
bottle was shipped to Battelle's Atmospheric Analysis Laboratory for analysis using volatile organic
analysis (VOA)-7 (a modification of EPA Method TO-15; EPA, 1999).
11
-------
The second sample was containerized in a specialized TeflonR bottle and submitted to USDA
laboratories, where it was analyzed for odorous chemicals using SPME and multi-dimensional gas
chromatography/mass spectrometry (MD-GC-MS) procedures. These procedures and a list of analytes
for both the headspace analysis approaches are described in Quality Assurance Project Plan (QAPP)
#390-QC-1 (Battelle, 2004).
2.5 Task 3. Land Sampling
The land sampling component of this research (Task 3) was conducted in parallel with Task 1 and
Task 2; however, the sampling regime for Task 3 extended several months beyond the day of biosolids
application. The land sampling schedule and specific details regarding the sampling plan and analyses for
this task are presented in the following subsections.
2.5.1 Land Sampling Field Plot Design. Unlike Tasks 1 and 2, the land sampling task was conducted
in the area of the biosolids application only and not the peripheral area. The land sampling approach was
developed such that three discrete plots of land were used for sampling activities. Each of the three plots
was randomly sited within the 2-acre area of the biosolids test site. The plots were 3 m across by 6 m
long. In the first half of the plot (3 m by 3 m), the soil was sampled at three randomly selected locations
during each sampling event using a grid system. Sampling locations were not used more than one time
(see Figure 2-7).
The distribution of biosolids on the day of application was measured using the second half of the
plot (3 m by 3 m). This was accomplished by creating a grid for the second half of the plot and randomly
selecting 20 locations within this area for sampling. Prior to biosolids application, these 20 locations
were covered with a 30-cm x 30-cm square of landscaping fabric that was secured to the soil surface.
Immediately after biosolids application, the squares were lifted off the soil and any biosolids on the
surface of the square (along with the fabric) were placed in a sample bag. These samples were analyzed
for biomass dry mass and volatile solids.
2.5.2 Land Sampling Procedures and Plan. Plant material on the soil surface and underlying soils
was sampled in this study. The soil was sampled using a coring device that removed a sample measuring
approximately 5 cm in diameter by 30 cm in depth. The sample handling was dependent on the analyte of
interest; in some cases, samples were subdivided and, in others, samples were composited.
Information on sample handling for the land sampling component of this study is included in
Section 7.0 of this report.
Land sampling included the following analytes:
• Microbial community by phospholipid fatty acid/fatty acid methyl ester (PLFA/FAME)
• Viable Helminth ova (VHO), Salmonella, enteric viruses, and male-specific coliphage
• Total heterotrophic bacteria (THE), fecal coliforms, Staphylococcus aureus, Enterococcus
spp., Escherichia coli, and Clostridiumperfringens
• Alkylphenol ethoxylates (APEs), their degradation products, and bisphenol-A (BPA)
• Soil characterization analyses (pH, percent organic matter, cation exchange capacity [CEC],
percent base saturation [Ca, Mg, Na, K, and H], disturbed bulk density, USDA textural class,
water-holding capacity, percent total nitrogen, phosphorous, and soluble salts)
• Biosolids dry mass/volatile solids
12
-------
• Terrestrial ecotoxicity by earthworm survival, seed germination, and root elongation
bioassays.
Samples of various depth ranges were collected and analyzed temporally.
2.6 Schedule of Events
Table 2-1 shows a list of sequential activities that occurred in the preparation and implementation
of this study. In general, on-site activities commenced 35 days prior to the day of application of the
biosolids to land in an effort to gather baseline data for the land sampling effort (Task 3). Other activities
related to bioaerosol and particulate sampling and to VOC and odor sampling (Tasks 1 and 2,
respectively) were conducted the day prior to application, immediately prior to application, during the
application, and up to 2 days after the application. Additional post-application land sampling efforts were
conducted up to 98 days after the biosolids had been land applied.
The day of application has been denoted as "Day 0". All sampling conducted prior to and after
the day of application are referenced in a positive or negative fashion relative to Day 0. For example,
pre-application land sampling conducted 35 days before Day 0 is referred to as Day -35.
Tables 2-2a through 2-2d present a composite list of the sampling and analyses that were
conducted for this research effort, with specific detail given by day relative to the application. Each of
Tables 2-2a through 2-2d are segregated by sample type (i.e., bioaerosols, VOCs, biosolids, and soil).
These tables also provide relevant information about the project-specific personnel who directed the
collection and/or analyses of the samples.
13
-------
Site Detail
3 Randomly Selected Sampling
Locations Using Grid System
for Each Sampling Event
(Sampling Locations were not j
Duplicated Through all
Sampling Events)
20 Randomly Selected Sampling
Locations Using Grid System for
Day of Application Sampling Event
(Measured by Removal of All
Visible Biosoiids from 30 cm x 30 cm
Area on Surface)
Application Area
(100 m Dia.)
/
m pi ing /
Astern /
'ent /
re not 1
all 1
t
i
f
t
t
i
t
t
/
ampling
tem for
ig Event
of All
n x 30 cm
^x
3mx
3
8
19
3
m x
15
3 m Sampl
3tT
12
4
6
ePI
1
otF
0
arti
*
tior
•
7
A
2
i Sample Plot Partition B
E
"
13
10
m
20
18
11
14
Field
Plots
(3 m x 6 ni)
Field
NOT TO SCALE
PLOTCONFIGURATION03.CDR
Figure 2-7. Land Sampling Plan
14
-------
Table 2-1. General Schedule of Field Events
Event
Site setup, plot marking, pre
application soil sampling
Initial site setup and mobilization for
the application study
Conference call to discuss field
schedule and logistics
On-site arrival of initial EPA field
personnel with supplies for site
preparation
On-site arrival of USD A bioaerosol
team
On-site arrival of USD A VOC and
odor sampling team
Pre-application soil sampling
Storm Delay - Morning meeting held
at hotel to discuss project schedule and
weather delay/ Afternoon meeting held
on-site to review bioaerosol sampling
plan
Biosolids arrived on-site at 3:30 PM
and were stored in trucks underneath
cover/Samples collected upon
arrival/Tractor and hopper equipment
arrived
Baseline bioaerosol and paniculate
sampling
Biosolids application day/Flux
measurements made on solids while in
the truck/Biosolids removed from the
trucks and dumped to form stockpile
near site/Specific sampling regime is
delineated in Table 4-3
Post application monitoring and field
sampling (Odor monitoring and SVOC
sampling-see Table 4-3)
Post application soil sampling - See
Table 4-3 and 7-1
Post application soil sampling - See
Table 4-3 and 7-1
Post application soil sampling - See
Table 4-3 and 7-1
Post application soil sampling - See
Table 4-3 and 7-1
Site demobilization
Performed By
EPA/Battelle/ NCDA&CS
Battelle/NCDA&CS
EP A/USD A/Battelle
and Battelle subcontractors/
NCDA&CS
EPA
USDA
USDA
EPA/Battelle/ NCDA&CS
EPA/USD A/Battelle/
Univ. of Colorado/Univ. of
Arizona/CH2MHill (morning
meeting only)/NCDA&CS
NCDA&CS/USDA/EPA
USDA/EPA/NCDA&CS
Battelle/Univ. of
Colorado/Univ. of Arizona/
CH2MHill/Biosolids Provider
USDA/EPA/NCDA&CS
Battelle/Univ. of Colorado/
Univ. of Arizona/
CH2MHill/Biosolids Provider
Battelle/NCDA&CS/
CH2MHill
EPA/Battelle/ NCDA&CS
EPA/Battelle/ NCDA&CS
EPA/Battelle/ NCDA&CS
EPA/Battelle/ NCDA&CS
Battelle/EPA/NCDA&CS
Date
(Day -35) August 25, 2004
(Day -7) September 23, 2004
(Day -5) September 25, 2004
(Day -4) September 26, 2004
(Day -4) September 26, 2004
(Day -3) September 27, 2004
(Day -3) September 27, 2004
(Day -2) September 28, 2004
(Day -2) September 28, 2004
(Day -1) September 29, 2004
(Day 0) September 30, 2004
(Day 1-3) October 1 - 3, 2004
(Day 14) October 14, 2004
(Day 28) October 28, 2004
(Day 63) December 7, 2004
(Day 98) January 4, 2005
May 11,2005
15
-------
Table 2-2a. Bioaerosols and Particulate Sampling: Analyte, Method, Sample Frequency,
and Responsible Personnel for Collection/Analyses
Analyte
THE
Fecal coliforms
Escherichia coli
Salmonella
Staphylococcus
aureus
Enterococcus spp.
Male-specific
coliphage
Clostridium
perfringens
Enteric viruses(a)
Bacterial endotoxins
Particulates
Method
USDA SOP 07.0 1A2
USD A SOP 7.0 1B2
USDA SOP 7.01C2
USDA SOP 7.02
USDA SOP 7.06
USDA SOP 7.03
USDA SOP 7.00 A, 7.09
USDA 7.08
EPA, 2003 (plaque forming
units [PFUs]) and EPA, 2001
(most probable number [MPN])
USDA SOP
EPA GRIMM
Responsible Party
USDA
Battelle/Environmental
Associates
USDA/EPA-
Cincinnati
EPA-Cincinnati
Sample Type
Biosamplers
(split sample)
Button Samplers
GRIMM Particle
Sampler
(in-field data collection)
Frequency
• Day -1
• Day 0, After
application
• Day -1
• Day 0, After
application
(a) Enumeration will reflect the possible presence of one or more of the following j
individual viruses cannot be provided for PFUs or MPN-cytopathic effect units
Coxsackie B, serotypes 1-34 of ECHO, Enterovirus serotypes 68-71, serotypes
groups of viruses; however, counts of specific
24 serotypes of Coxsackie A and six serotypes of
1-3 of OrthoReo, and three serotypes of polio.
-------
Table 2-2b. VOC and Odor Sampling: Analyte, Method, Sample Frequency,
and Responsible Personnel for Collection/Analyses
Analyte
OP-FTIR,EPATO-15,
and SPME
Ammonia and hydrogen
sulfide
Odor intensity
Odor concentration
Method
OP-FTIR
EPA TO- 15
SPME
Hand-held monitors
ASTM E544-99
ASTME679-91
Responsible
Party
EPA-RTP
Battelle
USDA
Battelle/CH2MHill
Sample Type
Monitored in the field
Flux Chambers
Monitored in the field
using hand-held
instrumentation
Monitored in the field
with olfactometers
Determined in the lab
by an Odor Panel
using samples from
the flux chambers
Frequency
• Day -1 through Day
1
• Day -1
• Day 0, Arrival of
biosolids at site
before application
• Day 0, After
application (1, 3, 4,
and 20 hours)
• Day -1 through Day
2 (periodically
throughout site)
• Day -1 through Day
2 (periodically
throughout site)
• Day 0, After
application (1,3,4,
and 20 hours)
selected flux
chambers
-------
Table 2-2c. Biosolids Sampling: Analyte, Method, Sample Frequency,
and Responsible Personnel for Collection/Analyses
Analyte
VOCs
SVOCs-Organochlorine
pesticides and brominated
diphenylether congeners
SVOCs
PLF A/FAME
APEs/BPA
Fecal coliforms
VHO
Salmonella
Enteric virases(a)
Male-specific coliphage
Method
GC/MS (VOA-7)
GC/MS (EPA
Modified SW-846
8270)
EPA Standard
Operating
Procedure
Region 5 SOP
EPA 1680 (Oct.,
2002) (MPN
Method)
EPA 2003
EPA 1682
EPA, 2003 (PFUs
and cell lines) and
EPA, 2001 (MPN)
EPA 1602
Responsible Party
Battelle
Battelle/Severn Trent
EPA-Cincinnati
EPA-Region 5
Battelle/Environmental
Associates
Battelle/Environmental
Associates
Sample Type
Headspace Analysis of
Biosolids
(composite of seven
randomly collected
biosolid samples)
Biosolids
(composite of seven
randomly collected
biosolid samples)
Biosolids (composite
sample)
Frequency
• Day 0, Arrival of biosolids
at site before application
• Days 1 and 2
• Day 0, Arrival of biosolids
at site before application
• Day 0, Arrival of biosolids
at site before application
• Day 0, Arrival of biosolids
at site before application
• Day 0, Arrival of biosolids
at site before application
-------
Table 2-2c (continued). Biosolids Sampling: Analyte, Method, Sample Frequency, and Responsible Personnel for
Collection/Analyses
Analyte
THE
Fecal coliforms
Escherichia coli
Staphylococcus aureus
Enterococcus spp.
Clostridium perfringens
pH
Organic matter
Cation exchange capacity (CEC)
Disturbed bulk density
% sand, silt, and clay
USDA texrural class
Water holding capacity
% total nitrogen
% total phosphorous
% soluble salts
% base saturation
Dry mass /volatile solids
Method
USDA SOP
07.01A
USDA SOP 7.0 IB
(Spiral Plating
Method)
USDA SOP 7.01C
USDA SOP 7.06
USDA SOP 7.03
USDA 7.08
SW-9045
Walkley Black
Titrimetric
SW-9081
SSSA Parti
SSSA Parti
SSSA Parti
SSSA Parti
SSSA Parti
SW-3050/6010
SSSA Parti
SSSA Parti
Standard Method
(SM) 2540
Responsible Party
USDA
Battelle/Agvise
Laboratory
EPA-Cincinnati
Sample Type
929 cm2 Surface
Sample
Frequency
• Day 0, Arrival of biosolids
at site before application
Immediately following
application
(a) Enumeration will reflect the possible presence of one or more of the following groups of viruses; however, counts of specific individual
viruses cannot be provided for PFUs or MPN-cytopathic effect units: 24 serotypes of Coxsackie A and six serotypes of Coxsackie B,
serotypes 1-34 of ECHO, Enterovirus serotypes 68-71, serotypes 1-3 of OrthoReo, and three serotypes of polio.
EPA, 2003. Environmental Regulations and Technology: Control of Pathogens and Vector Attraction in Sewage Sludge, EPA/625/R-92/013.
EPA, 2001. Manual of Methods for Virology, Chapter 15.
SSSA = Methods of Soil Analysis, Part 1 - Physical and Mineralogical Methods. Soil Science Society of America, Inc., Madison, WI. 1986.
-------
Table 2-2d. Land Sampling: Analyte, Method, Sample Frequency, and Responsible Personnel for Collection/Analyses
Analyte
PLF A/FAME
APEs/BPA
Fecal coliforms
VHO
Salmonella
Enteric viruses(a)
Male-specific coliphage
THE
Fecal coliforms
Escherichia coli
Staphylococcus aureus
Enterococcus spp.
Clostridium perfringens
Toxicity testing (earthworm
mortality, seed germination, root
elongation)
pH
Organic Matter
Cation Exchange Capacity (CEC)
Disturbed bulk density
% sand, silt, and clay
USDA textural class
Water holding capacity
% total nitrogen
% total phosphorous
% soluble salts
Method
LRPCD SOP
EPA-Region 5 SOP
EPA 1680 (Oct, 2002) (MPN
Method)
EPA, 2003
EPA 1682
EPA, 2003 (PFUs) and EPA, 2001
(MPN)
EPA 1602
USD A SOP 07.01 A
USDA SOP 7.0 IB (Spiral Plating
Method)
USDA SOP 7.01C
USDA SOP 7.06
USDA SOP 7.03
USDA 7.08
SOP (QAPP 33-Q3-0)
Walkley Black Titrimetric
SW-9081
SSSA Parti
SSSA Parti
SSSA Parti
SSSA Parti
SSSA Parti
SW-3050/6010
SSSA Parti
Walkley Black Titrimetric
Responsible Party
EPA-Cincinnati
EPA-Region 5
Battelle/Environmental
Associates
USDA
EPA-Cincinnati
Battelle/Agvise
Laboratory
Sample Type
0-5, 10-15, and 20-25
cm core segments,
measured on 65 mm
sieved samples
0-5 cm segment
0-5 cm segment
(composite of three
samples)
0-5 cm segment
(composite of three
samples)
3,375 cm3 (composite
of four 15 cm x 15 cm
x 15 cm samples)
3 samples composited
each by depth (0-5, 10-
15, and 20-25 cm core
segments)
Frequency
• Day 35
• Day 3
• Day 0, After
application
• Days 14, 28, 63, and
98
• Day 35
• Day 3
• Day 0, After
application
• Days 28, 63, and 98
• Day 35
• Day 28
• Day 98
• Day 35
• Day 0, After
application
• Day 98
• Day 35
• Day 0, After
application
• Day 28
• Day 98
(a) Enumeration will reflect the possible presence of one or more of the following groups of viruses; however, counts of specific individual viruses cannot be
provided for PFUs or MPN-cytopathic effect units: 24 serotypes of Coxsackie A and six serotypes of Coxsackie B, serotypes 1-34 of ECHO, Enterovirus
serotypes 68-71, serotypes 1-3 of OrthoReo, and three serotypes of polio.
EPA, 2003. Environmental Regulations and Technology: Control of Pathogens and Vector Attraction in Sewage Sludge, EPA/625/R-92/013.
EPA, 2001. Manual of Methods for Virology, Chapter 15.
SSSA = Methods of Soil Analysis, Part 1 - Physical and Mineralogical Methods. Soil Science Society of America, Inc., Madison, WI. 1986.
-------
3.0 BIOSOLIDS APPLICATION
3.1 Product Selection
Anaerobic, dewatered (centrifuged) biosolids generated at a municipal wastewater treatment plant
(WWTP) were specifically selected for this study. Polymer was added to the biosolids during dewatering
as a normal WWTP standard practice. Consideration was given to this type of biosolids because it was
desirable for the product to elicit odor and to generate particulates via surface drying, flaking, and wind
erosion. At the request of the researchers, this material was pretreated with only enough lime to adjust the
pH and suppress microbial growth to meet facility compliance for release (material met Class B
compliance at time of release from facility). As such, this material was atypical of biosolids that would
normally be released from this facility and are not produced in this manner on a regular basis. Larger
doses of lime are more consistent with the normal operation of the facility.
The material once released from the WWTP was stored for approximately 1.5 days inside the
truck and under cover at the NCDA&CS Piedmont Research Station prior to application. Samples were
collected from the biosolids in the truck immediately after arrival and from the biosolids stockpile
immediately before application, and the analytes listed in Table 2-2c were measured including fecal
coliforms. Permission to hold the biosolids under cover before application was granted in the form of a
permit by the State of North Carolina for the purpose of this research.
The decision to hold the biosolids prior to application was primarily driven by inclement weather
(Hurricane Jeanne) that occurred during the week of scheduled application and sampling.
3.2 Application
Biosolids were applied at a rate of 10 wet tons/acre. Other than modifications to facilitate taking
air and soil samples, application practices and equipment were typical of those used during normal
agronomic biosolids application.
The biosolids were land applied in the test area using a Knight 8030 hopper and tractor slinger
that distributed biosolids on the land surface from a discharge point on the forward left side of the hopper.
Biosolids were applied at an angle with the current wind flow, so that the unit itself was not obstructing
airflow to the downwind sampling array (Figure 3-1).
The applicator made one pass across the 2-acre area as depicted in Figure 3-1. When it reached
the circumference of the circle, it ceased to distribute biosolids and followed along outside the circle and
re-entered parallel to its first discharge point. It took 12 passes across the circle to complete the
application, expending a total of 44 minutes from start to finish. Table 3-1 shows the approximate start
and stop times of each pass, elapsed time per pass, the times when the hopper needed to reload, and the
total expended time. The Day -1 sampling event was conducted using an empty hopper so that the
particulate and fuel emissions that could potentially be produced from the movement of the equipment
across the field could be captured and quantified. Careful consideration was given to ensuring that the
applicator had been thoroughly cleaned in order to limit its contribution of bioaerosol emissions during
the baseline sampling.
The hopper was reloaded with biosolids four times during the application. Reloading times
ranged from 11 to 14 minutes measured from the point at which the hopper stopped applying biosolids to
the time that it started the application again. The hopper was pulled to the stockpile location that was a
21
-------
Upwind
(UW)
Downwind
(DW)
RE6EPT0R
NOT TO SCALE
A 2 Impinger Biosamplers + 2 Button Filter Samplers (Endotoxin) + 1 Impactor Sampler
*
Mobile*: 2 Impinger Biosamplers + 2 Button Filter Samplers (Endotoxin) + 2 Impactor Samplers
Grimm Sampler
Note
* Samplers were mounted on the front of an all-terrain vehicle (ATV). The ATV was driven such that
it collected samples at a constant distance from the release point of the applicator.
Figure 3-1. Method of Biosolids Application
SAMPLEGRID08_02.CDR
22
-------
Table 3-1. Application Timeline
September 29, 2004
Day -1 (Pre-application)**
Start
Time
14:07
14:20
14:30
14:39
14:45
14:52
14:59
15:07
15:13
15:20
15:26
15:32
Total
Elapsed
Time
Stop
Time
14:11
14:25
14:35
14:41
14:47
14:55
15:01
15:10
15:15
15:23
15:29
15:35
Time
Elapsed
(min)
4
5
5
2
2
3
2
3
2
3
3
3
37
Load
Yes/No/
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
September 30, 2004
Day 0 (Application)
Start
Time
9:37
9:44
10:00
10:19
10:25
10:30
10:47
10:54
10:59
11:15
11:20
11:24
Total
Elapsed
Time
Stop
Time
9:42
9:49
10:05
10:24
10:29
10:34
10:50
10:57
11:02
11:18
11:22
11:26
Time
Elapsed
(min)
5
5
5
5
4
4
3
3
3
3
2
2
44
Load
Yes/No/
NA
No
No
Yes
Yes
No
No
Yes
No
No
Yes
No
No
NA = not applicable
** = Tractor and empty hopper were deployed to simulate application during baseline sampling.
short distance away and on the other side of the tree-line from the area of main application (see Figure
2-1). The stockpile was located at some distance away from downwind airflow to minimize the potential
impacts to the in-field odor survey that was being conducted during application.
The times shown in Table 3-1, which correspond to the start and stop times of the hopper
application, also correspond to the start and stop times associated with the bioaerosol sampling stations.
Sampling at the nine transect upwind and downwind sampling points (three lines of three sampling points
each) and on the mobile unit commenced and terminated upon the initiation and completion of each pass.
This approach was designed to provide the best opportunity for capturing a detectable number of
organisms and diminishing the dilution effect that would result from a continuous capture mode of
sampling and prolonged waiting time during hopper reloading.
There was, however, some variance to the capture methods used during Day -1 and Day 0.
During Day -1, the upwind and first line of downwind sampling units were simultaneously started and
stopped in association with the application described. The second line of downwind sampling units was
stopped on a 20 second delay to account for the longer travel distance of the particle plume. These
samplers were started again simultaneously with the other sampling units. In practice, it turned out that
this was cumbersome and difficult to accomplish simultaneously for the nine total stations. Therefore,
during Day 0, this method of "staggering" sampler times was eliminated and all samplers were run for the
same duration, with all the samplers running for a period deemed suitable for capturing the particle
plume.
23
-------
4.0 BIOSOLIDS PRODUCT RESULTS
The biosolids stockpile was analyzed for a variety of constituents (SVOCs, VOCs, indicator and
pathogenic microorganisms, and physical/chemical characterization data) once it arrived on site. The
reasons for these analyses varied; however, specific baseline data were needed in an effort to complete the
objectives that were set for Tasks 1, 2, and 3. Therefore, biosolids as the studied matrix cross cut the
three specific tasks, and the data presented here are a compilation of the data generated from these three
tasks.
Upon arrival at the site, the biosolids were stored under cover at a facility approximately 1 mile
away from the application area on the property of the NCDA&CS Piedmont Research Station. Once the
biosolids stock arrived, it was sampled at two discrete time points (the first when the biosolids arrived on
site and the second approximately 1.5 days later when the biosolids were dumped from the truck to create
the stockpile).
4.1 Samples Collected from the Delivery Truck.
When the truck arrived at the site carrying the biosolids for the research study, a team of staff
collected one composite grab sample constructed from seven individual grab samples collected randomly
from the truck bed as specified in the QAPP.
In addition to this one composite grab sample, flux chamber measurements and sampling were
conducted approximately 1.5 days after the truck arrived at the site and immediately prior to forming the
biosolids stockpile on the day of application. The chamber was placed directly on top of the biosolids in
the truck. During operation of the flux chamber, gas samples were collected in three 5-L Tedlar® bags
and one 5-L Summa canister for SPME and TO-15 analyses, respectively.
4.1.1 Microbial Enumeration of Biosolids Grab Samples. The composite grab sample was collected
and then subsampled to determine the level of E. coli present. The E. coli MPN determined from this
analysis was 4.35 x 107/g-total biosolids.
4.1.2 Flux Chamber Measurement of Biosolids (Measured from the Truck). As described above,
in addition to the grab samples collected from the truck, a 0.5-hr flux chamber measurement was also
conducted immediately prior to removing the biosolids from the truck and creating the stockpile. This test
resulted in detectable levels of dimethylsulfide and dimethyldisulfide at flux values of 3.78 and
4.54 (ig/m2/hr, respectively.
4.1.3 SPME Measurements from Samples Generated from the Flux Chambers.
The results of SPME samples collected from flux chambers positioned on biosolids in the truck are
discussed in Section 6.3.2 along with the results of flux chamber SPME samples collected during
biosolids application.
4.2 Samples Collected from the Biosolids Stockpile Prior to Application
On the day of application, the biosolids were dumped from the truck onto the ground to create a
stockpile. Once the stockpile was formed, three composite samples (each constructed from seven
individual grab samples) were collected randomly within the pile using the method mentioned previously
and outlined in the QAPP. Sub-samples were removed from the composite sample for SVOC and
headspace analyses using the VOA-7 method.
24
-------
After these sub-samples were collected, the composite sample was contained in a sealed 5-gal
bucket at ambient temperature for approximately 1 day after collection prior to being sub-sampled for the
remaining suite of microbial and chemical analyses. The constituents that were analyzed from the day-
old stockpile composite were:
• PLFA/FAME (Task 3)
• Fecal Conforms (MPN Methodl680 [5-Tube]) (Task 3)
• Microbial Indicators (Task 3)
o VHO
o Salmonella
o Enteric virus
o Male-specific coliphage
• Microbes (Task 3)
o THE
o Fecal coliforms
o S. aureus
o Enterococcus spp.
o E. coli
o C. perfringens
• APEs/BPA (Task 3)
• Physical/Chemical Characterization (Task 3)
4.2.1 SVOC Analysis on Biosolids. The biosolids that were collected from the composite stockpile
sample were submitted for SVOC analysis using modified EPA Method SW-846 8270 for PBDE and
selected pesticides and PCB congeners. Samples were also collected in the field after the application on
Days 1 and 2. Three individual samples were collected from random locations within the application
area. The SVOC results for these samples are presented in Table 4-1.
4.2.2 Headspace Analysis Using VOA-7. An approximate 100-g aliquot of biosolids sample was
collected in triplicate and transferred into special bottles used to conduct headspace analysis. The
headspace analysis was conducted according to the procedures endorsed in the QAPP by removing a
headspace sample after allowing the sample to equilibrate for a period of 2 hr. The results of the analyses
conducted on the biosolids from the stockpile, as well as the samples that were collected after application
in the field on Day 1 and Day 2, are discussed in Section 6.3.1.
4.3 Microbial and Physical/Chemical Characterization of the Biosolids Samples
Collected from the Stockpile
Tables 4-2 and 4-3 Summarize Task 3 microbial indicator and physical/chemical analyses,
respectively. PLFA/FAME results for biosolids samples have been prepared such that they are relative to
the land sampling effort throughout the length of the study and, therefore, are presented in Section 7.0.
Microbial indicator analyses were performed by Environmental Associates, Ltd. under contract to
Battelle.
25
-------
Table 4-1. SVOC Results for Biosolids Stockpile Composite Sample and Samples
Collected in the Field After Application on Days 1 and 2
Analyte
% Moisture
% Solids
9/30/2004 (Stockpile)
Average
78
22
Std. Dev.
0.6
0.6
10/1/2004 (Day 1)
Average
75
25
Std. Dev.
0.0
0.3
10/2/2004 (Day 2)
Average
74
26
Std. Dev.
1.5
1.1
SVOCs (SW-846 8270C), w/kg-dry wt.
bis(2-Ethylhexyl)phthalate
Di-n-octyl phthalate
3-Methylphenol & 4-
Methylphenol
Phenol
Benzo(b)fluoranthene
Pyrene
16,333
127
35,000
9,933
77
210
3,214.6
63.5
7,937.3
1,833.9
2.9
115.3
18,000
165
17,000
9,467
87
107
5,000.0
151.6
5,567.8
1,457.2
37.5
46.2
15,100
158
5,567
6,567
83
153
5,850.6
78.2
3,092.5
2,458.3
31.8
127.0
Pesticides (SW-846 8081A), fig/kg-dry wt.
Dieldrin
Heptachlor
Methoxychlor
13
9
21
6.9
0.3
0.6
11
8
22
8.7
0.9
31.4
13
8
58
9.4
1.2
36.2
Table 4-2. Microbial Analyses of Biosolids Collected from the Stockpile Composite Sample
Date Sampled
DayO
Virus
(MPN/4g)
<0.68
2.94
0.70
(PFU/4g)
<1
<1
<1
Salmonella
(MPN/4g)
>325
>325
>325
HOva
(No. viable/4g)
1.8
2.5
<0.8
Coliphage
Male-specific (PFU)
199
133
75
Somatic (PFU)
557
616
448
MPN - most probable number
PFU - plaque forming units
Additional pre-application analysis of the biosolids was performed by the USDA. The results
presented as averages of triplicate biosolids samples collected from the pile immediately prior to
application are as follows:
• THE: 1.6 x 1011 CFU/gdw total solids
Total Coliforms: 1.4 x 109 CFU/gdw total solids (consisted of only fecal coliforms)
• Fecal Coliforms: 2.3 x 109 CFU/gdw total solids
• Enterococcus spp.: 8.2 x 105 CFU/gdw total solids
• Staphylococcus aureus: none detected
• Salmonella: 6.33 MPN/gdw total solids (confirmed Salmonella enteritidis in all three
replicates)
Physical/chemical analyses were conducted by Agvise Laboratories under contract to Battelle.
Table 4-3 summarizes these data for the biosolids that were collected from the stockpile composite
sample.
26
-------
Table 4-3. Physical/Chemical Constituents Results for Biosolids Collected from the
Stockpile Composite Sample
Parameter
Sand (%)
Silt (%)
Clay (%)
USDA Textural class
Bulk Density (gm/cc)
Cation Exchange Capacity (meq/lOOg)
Moisture @l/3 bar (%)
Moisture @15 bar (%)
pH (H2O)
Total P (rag/kg)
Olsen Phosphorus (rag/kg)
Soluble Salts (mmhos/cm)
Value
5
78.6
16.4
Silt Loam
0.89
22
158.4 (supersaturated)
75.6
7.4
24,453
176
4.09
Base Saturation Data (mg/kg)
Ca
Mg
Na
K
H
823
2,231
627
221
26
27
-------
5.0 TASK 1. BIOAEROSOL AND PARTICULATE SAMPLING RESULTS
5.1 Objectives
In Task 1, aerosol emissions from biosolids were sampled to evaluate methods that measure the
concentrations of selected bacteria, fungi, viruses, bacterial endotoxins, and particulates. These aerosol
emissions were sampled prior to, during, and after biosolids application. Samplers were located upwind,
within, and downwind of the application area. The primary objectives for bioaerosol and particulate
sampling were to:
1. Characterize the type and concentration of the suite of viable bioaerosol components selected
for analysis (seven bacteria, culturable enteric viruses, and male-specific coliphage) as well as
particulates (< 5.0 |om)
2. Determine if the bioaerosol components were emitted and transported to several points
downwind of the biosolids application area under the circumstances investigated
3. Investigate the collection performance of the six-stage impactors, biosamplers, and GRIMM
sampler as applied in this field study setting.
Table 2.2a lists the analytes, methods, sample types, and sample frequency for the bioaerosol and
particulate sampling efforts. Analytical methods were evaluated based on whether data acceptance
criteria specified in the QAPP were achieved.
5.2 Bioaerosol and Particulate Matter Sampling
A limited number of quantitative field studies that involved bioaerosols associated with biosolids
management activities have been performed (Tanner, et al., 2008; Low et al., 2007; Paez-Rubio, et a\.,
2007; Baertsch, et al, 2007; Brooks et al, 2005; Brooks etal., 2004; Tanner, et al, 2005; Rusin, et al,
2003; NIOSH, 1998; and Pillai, et al, 1996). Some of these studies have shown downwind
concentrations of heterotrophic organisms at biosolids application sites to be greater than upwind
concentrations, but these levels were not considered to pose a public health concern because indicator
organisms (such as coliforms) were not detected at any significant distance from the biosolids source.
Bioaerosols were defined for this study as aerosolized particles of biological origin or activity that
may affect living things through infectivity, allergenicity, toxicity, pharmacological impacts, or other
processes. Particle sizes may range from aerodynamic diameters of ca. 0.5 to 100 um (Hirst, 1995). In
effect, bioaerosols have physical, biological, and chemical attributes and may contain fragments or parts
of the original intact organisms. For this research, a particle size analyzer (GRIMM) located in the center
of the application area (Figure 2-2) measured particles in the size range of 0.23 to 20 um. Specific
particle sizes captured by the bioaerosol sampling equipment (impactors and biosamplers) are unknown
but assumed to be within this size range.
Enteric viruses were assayed from air samples. The procedures used to conduct the assay
required that the virus be active and capable of infecting the Buffalo Green Monkey Kidney cell line.
Many enteric viruses that potentially could be present in the biosolids are not known to infect this cell
line, and others will not infect any cell line at all or only sporadically. Thus, the active virus populations
that were assayed reflected the possible presence of one or more of the following groups of viruses
(plaque-forming units [PFUs] or most probable number [MPN]-cytopathic effect units; 24 serotypes of
Coxsackie A and six serotypes of Coxsackie B; serotypes 1-34 of ECHO; Enterovirus serotypes 68-71;
serotypes 1,2,3 OrthoReo; and three serotypes of polio). Counts of specific individual viruses could not
28
-------
be provided because of limitations of the methods, and polymerase chain reaction (PCR) analyses were
not conducted for this work.
Coliphages are viruses that infect certain coliform bacteria, but not humans, animals, or plants.
Those investigated in this study infect E. coll. Methods of detection and enumeration can be conducted in
most microbiology laboratories using plaque assays. In contrast, culture-based methods of detection and
enumeration of enteric viruses require specialized facilities and training in tissue culture, a feature not
available in all microbiology labs. Coliphage analysis is significantly less expensive than analysis for
enteric viruses. Because of these features, coliphages have served as surrogates for enteric viruses and
they were included for assaying in this study.
Fecal coliforms were included as a reference standard method. The pre-eminent fecal coliform,
E.coli, historically has served as an indicator of fecal pollution. It has the good characteristics of a fecal
indicator, such as not normally being pathogenic to humans, and is present at concentrations much higher
than the pathogens it predicts (Scott et al., 2002). Methods for their detection in air are not standardized;
standard methods for their detection in water have been adapted for analyzing airborne organisms
collected using liquid impingement air sampling in these studies.
Salmonella spp. are fecal pathogens and are thus a public health concern. Methods for their
detection in biosolids are standardized (EPA Method 1682). Methods for their detection in air are not
standardized; standard methods for their detection in water were adapted for the analysis of airborne
organisms collected using liquid impingement air sampling in these studies.
Enterococcus spp. is a renamed subgroup of fecal streptococci; there are at least five species,
Enterococcus faecalis, E. faecium, E. durans, E. gallinarum, and E. avium, that are differentiated from
other streptococci by their ability to grow in the presence of stressors, specifically 6.5% NaCl, pH (9.6)
and45°C. Two species, E. faecalis and E. faecium, are most frequently found in humans. Enterococcus
spp. have been used successfully as indicators of fecal pollution and are especially reliable as indicators of
health risk in marine and recreational waters. However, environmental reservoirs of Enterococcus spp.
exist, and they may re-grow when released into the environment (Scott et al., 2002).
In contrast to the coliforms E. coll and Salmonella spp., which are gram-negative bacteria,
Enterococcus spp., Clostridiumperfringens, and Staphylococcus aureus are gram-positive bacteria.
C. perfringens is the only bacterium in the specific assay suite that is a strict anaerobe and endospore
former. They were selected for inclusion in this study because: 1) Enterococcus spp. are also indicators
of fecal contamination, 2) they are used in water quality evaluations, 3) S. aureus has been implicated as
the causal agent in human infections speculated to have resulted from direct contact with land applied
biosolids (EPA, 1992), and 4) C. perfringens is regarded by some researchers as a useful, very
conservative indicator of fecal bacterial contamination. Methods for their detection in air are not
standardized; standard methods for their detection in water were adapted for the analysis of airborne
organisms collected using liquid impingement air sampling in these studies.
THE were included as a positive control of sampler operation with regard to viable bacteria in
ambient air. Heterotrophic bacteria include the saprophytic aerobes and facultative anaerobes that are
naturally present in soil, on plant surfaces, in air, and in water. They include many beneficial,
non-pathogenic bacteria that degrade organic matter. Their concentrations in soil, water, and air depend
on vegetation type and amount, local climate and soil conditions, animal and human activities, and
circumstances in the general vicinity in which the soil, water, or air sample is obtained. In this study,
their presence in air samples upwind and downwind demonstrated that the air samplers were operating
sufficiently well to collect viable microbes and permitted characterization of concentration variations
across the sample grid. Methods for their detection in air are not standardized; standard methods for their
29
-------
detection in water were adapted for the analysis of airborne organisms collected using liquid impingement
air sampling in these studies.
Sampling for endotoxin (lipopolysaccharides containing Lipid A from all gram-negative bacteria)
was performed on air samples. When inhaled and respired, endotoxin can elicit a variety of well-
characterized responses in mucous membranes (eyes, nose, sinus, throat, bronchi, and lungs) and systemic
complaints (fever, malaise, and pulmonary function distress), some of which are among those commonly
described by persons reporting illness they associate with biosolids application near their residences.
Sampling and testing for airborne fungi were conducted to assess "mold" in air. Several
protocols for sampling are available. The sampling and culturing methods used in this study were
designed to provide an estimate of total viable fungal particles.
5.3 Overview of Field Operations
A total of nine bioaerosol sampling stations and one mobile bioaerosol sampling unit were
deployed for testing and were located within and outside a 100-m diameter (circular) study area described
in Section 2.3 and shown in Figure 2-2. The GRIMM optical scanning counter was placed in the center
of the application field.
One upwind (UWA) and two downwind (DWB and DWC) parallel sampling transects were
located outside the 100-m circular test area (see Figure 2-2). Each transect contained a total of three
sampling stations that were spaced laterally to increase the collection area of the sampling field, as
opposed to clustering the samplers within narrow zones or regions. The sampling station layout for
transects UWA, DWB, and DWC included three upwind stations (UWA- the center station located 16m
upwind from the top boundary of the application area), three stations in the first downwind transect
(DWB- the center station located 16 m downwind of the lower boundary of the application area), and
three stations in the second downwind transect (DWC- the center station located 50m downwind of the
lower boundary of the application area). The stations for transects UWA and DWB were 75 m apart
within each transect. The stations for transect DWC were approximately 90 m apart.
The field design allowed for the physical movement of each station because the samplers and
their ancillary equipment (pumps and power sources) were secured on pull carts so they could be
re-positioned readily if necessitated by varying wind patterns. Each of the sampling stations contained
two biosamplers, one six-stage impactor, and two button samplers. The cart-mounted bioaerosol
sampling equipment was shown previously in Figure 2-3.
A MOB was also deployed in the application area. The MOB consisted of an ATV with two
biosamplers, two six-stage impactors, and two button samplers affixed to a modified plate on the forward
side of the vehicle. The ATV followed a path approximately 8-10 m to the rear and downwind of the
Knight 8030 side-discharging hopper that applied the biosolids. When the hopper moved off site to
reload, the MOB waited inside the application area and commenced sampling again once biosolids
application was re-initiated. The MOB is shown sampling behind the side-discharging applicator in
Figure 5-1.
5.3.1 Application Schedule. Due to the logistics of testing and the amount of labor necessary for
on-site sample processing and analysis, the control trial and the biosolids application test were conducted
on separate consecutive days. The control trial was conducted first and involved all of the same activities
that occurred during the biosolids application, including the movement of the biosolids application
machinery, except that during the control trial the discharge hopper was not loaded with biosolids and
was cleaned by pressure spraying prior to use. Therefore, biological or particulate matter contribution in
30
-------
- •*• „.,•. -.- ,»•*. ;
Figure 5-1. MOB Conducting Bioaerosol Sampling Approximately
8-10 m Behind Biosolids Applicator
aerosols due to dust generation during movement of the application equipment could be determined
separately from that which was generated from biosolids.
5.3.2 Operational Schedule. For both the control trial and the actual biosolids application, the aerosol
samplers were operated intermittently, with sampling focused during the actual time that the applicator
moved through the application area and with a reasonable allowance time for particulate transport
(2 minutes) to downwind stations. When the applicator moved out of the application area, sampling was
disengaged in an effort to reduce overload on the impactor agar plates and evaporative loss of biosampler
fluid associated with longer collection periods (Barth, 2007).
The biosolids applicator moved across the field as shown in Figure 3-1 and moved off site to the
stockpile intermittently when it needed to reload. A total of 44 minutes were required to complete the
application. In a similar manner for the MOB, the biosamplers, six-stage impactor, and button samplers
were operated during the period when the applicator was traveling across the application area. The
impactor agar plates were replaced once during the application to reduce overloading on the agar plates.
5.3.3 Aerosol Sample Collection. The biosamplers were operated at an airflow rate of 12.5 L/min.
The collection fluid consisted of 20 mL of a sterile 0.02-M phosphate buffered (pH 7.4) solution (PBS)
that was subsequently transferred to R2A agar plates following APFŁA Method 9215 (APFŁA, 1992), with
subsequent analysis for bacteria (only). The biosamplers were foil covered to block ultraviolet light, and
periodic addition of the PBS was implemented to replace evaporated fluid.
The six-stage impactors were operated at an airflow rate of 28.7 L/min and contained two-section
agar plates that were separately analyzed for viable bacteria and fungi (with subsequent adjustment for the
positive control corrections). The fungi section of the agar plate contained Oxgall media (Difco)
supplemented with 50 mg/L of vancomycin and streptomycin to inhibit bacterial growth. The bacteria
section of the agar plate contained R2A agar (Difco), supplemented with 0.5% pyruvate to assist the
resuscitation of stressed bacteria.
31
-------
Participate size and mass were monitored immediately before and during both the control trial
and the biosolids application. The GRIMM Model 1108 optical scanning counter was operated at an
airflow rate of 1.2 L/min.
5.3.3.1 Biosampler. The biosamplers were secured on their tripod masts at approximately 1.5m above
the ground surface and inside the carts such that the orifices were oriented into the wind flow. When
sampling commenced, the Parafilm™ cover was removed from the orifices of the biosamplers, the
biosampler outlets were connected to a designated pump, and the pumps were turned on. Figure 5-2
shows the biosamplers being operated in the field during application testing.
Figure 5-2. Downwind B (DWB) Biosampler and Operator
When the sampling event was completed, each biosampler was disconnected from its pump and
immediately taken to the on-site mobile lab for sample processing and analysis. The liquid sample was
aseptically transferred into a sterile sample container for analysis. The sample volume was fractioned as
indicated in Figure 5-3 for the individual analyses already discussed in Section 5.2.
On-site processing was conducted in duplicate for each bioaerosol liquid sample. A 5.0-mL
aliquot was aseptically transferred into a separate screw-cap centrifuge tube and frozen for off-site
analysis of enteric viruses. Another 3.0 mL was used to conduct the Salmonella spp. and E. coll pathogen
assays. The remaining sample was distributed and analyzed for C. perfringens (1.0 mL), male-specific
coliphage (1.0 mL), Salmonella (MPN) for any samples that tested positive for the pathogen assay
(4 mL), THBs (1.0 mL), fecal coliforms (1.0 mL), E. coll (1.0 mL), Enterococcus (1.0 mL), S. aureus
(1.0 mL), and bacterial bioburden (1.0 mL). A 4.5-mL aliquot was transferred and stored in reserve at
4°C.
5.3.3.2 Six-Stage Impactor. When sampling was completed, each six-stage impactor was disconnected
from its pump and taken to the on-site mobile lab. Agar plates from the six-stage impactor were
aseptically removed and recovered with sterile lids, labeled, and incubated at 37°C. At 24 and 48 hr, the
32
-------
24-mL
Total from Biosampler
USDA Labs
Beltsville, MD
• 2mL
C. perfringens
Coliphage MS2
• 4 mL
Reserve Sample
University
of Colorado
• 1 mL
Total Bioburden
Assay
(bacterial)
Environmental
Associates
• 5mL
Enteric Viruses
USDA Labs
On-Site
• 3mL
Pathogen Assay
(Salmonella, and
E. coli)
•4mL
Salmonella (MPN)
•5mL
(1 mL) Total THB
(1 mL) Fecal coliform
(1 mL) E. coli
(1 mL) S. aureus
(1 mL) Enterococcus
Figure 5-3. Distribution of Biosampler Fluid for Bacterial and Viral Analyses
plates were examined for colony growth and the number of colonies corresponding to the stage sieve hole
pattern were counted. The positive hole correction table (Macher, 1989) was used to convert the count to
a maximum likelihood number of positives to account for the possibility of multiple impactions.
5,3.3.3 Button Sampler. The button samplers used to obtain bacterial endotoxin samples were operated
in parallel with the biosamplers and six-stage impactors. After sampling was completed, the sample
filters were harvested and stored at 4°C in the mobile laboratory until they were packaged for overnight
shipment to the laboratory for extraction and endotoxin analysis. Unfortunately, problems were
encountered in the field with the filters in many of the filter units, resulting in wrinkling and tearing of the
filters and ultimately negatively impacting their performance. Additional complications were
encountered in the laboratory during further processing and extraction. Due to these complications, the
validity of these data was questioned and, therefore, are not reported. There is no further discussion of
endotoxin in this report.
5.3.3.4 GRIMM Particle Analyzer. A GRIMM particle analyzer/dust monitor Model 1.108 was used
for the continuous measurement of particles in the air. The GRIMM monitored single-particle
33
-------
counts using a light-scattering technology and recorded the particle data on a data storage card. After
completion of the test, the card was removed and the data were downloaded.
5.4 Bioaerosol and Particulate Matter Results
The prevailing wind direction for this site is nominally from the north or northeast at this time of
the year according to historical records. Based on this information, the sample collection array was
designed with upwind samplers to the northeast and downwind samplers to the southwest of the
application area (see Figure 2-2). However, during the control trial (Day -1), the ground-level wind
direction was unexpectedly from the southwest (Figure 5-4) with an average magnitude of 1.1 m/s and
minimal gusting. As a result, the design as depicted in Figure 2-2 acted in reverse on this day with the
single upwind transect serving as the downwind sampling stations, and the double downwind transects
serving as the upwind sampling stations. Although this direction was not as expected, the control trial
results are still considered valid since no statistical differences in the upwind (acting as the downwind)
and two downwind (acting as the upwind) stationary zones were observed, as would be anticipated with
no biosolids being applied.
During biosolids application (Day 0), the wind direction was predominantly from the north and
the design depicted in Figure 2-2 functioned as expected. On this day, wind was very light with an
average magnitude of 0.6 m/s.
5.4.1 Bioaerosol Results and Analysis. Fecal coliforms, E.coli, Salmonella spp., S. aureus,
C. perfringens, Enterococcus spp., and coliphage were not detected in any of the bioaerosol samples
collected anywhere on the site either in the control trial or during biosolids application. This also was the
case for the MOB that sampled within 8-10 m of and directly behind the discharged biosolids.
Enteric virus analyses were conducted using the PFU and MPN procedures. Virus analyses were
only performed on the mid-line sample stations (see Figure 2-2; UWA-2, MOB, DWB-2, and DWC-2)
for the control trial and application test, resulting in a total of 16 samples for virus analysis. Four
additional blank samples were included for quality control purposes. The QAPP specified that PCR
would be performed on these 16 samples only if they were positive for PFU and MPN. The remaining 32
samples (from the remaining stations) were frozen and were to be analyzed for PFU/MPN and possibly
PCR only after it had been determined that the first 16 samples yielded positive results. There were no
positive results for enteric virus in the PFU and MPN analyses that were conducted for the initial 16
samples; therefore, no further analyses were conducted on these samples or on the remaining 32 samples
that were kept frozen.
THE were assayed and detected in all bioaerosol samples collected with the biosamplers for both
the control trial and biosolids application test. THE were also detected on agar plates in the six-stage
impactors. Their presence in both upwind and downwind air samples demonstrated that the air samplers
were operating sufficiently well to collect viable microbes. Total fungi were also assayed and detected in
the six-stage impactors.
The bioaerosol data for each of the three stations within a given transect (UWA, DWB, and
DWC) were averaged (e.g., the THE results for UWA stations 1, 2, and 3 were averaged, representing the
entire upwind zone). The data were analyzed to: 1) determine if there were significant differences in
THE and fungi counts between the control trial and the biosolids application test, and 2) determine if
there were significant differences in the THE and total fungi counts between upwind and downwind
locations.
34
-------
NORTH'
lA/EST '.
(A) Day -1 (Control Trial - September 29, 2004 from 2:00 - 3:30 PM)
' NORTH'• -.
INVEST
15% '
10%'
EAST:
1 .2 - 1 .4
1 - 1 .2
SOUTH
CH 0.6 -0.3
n 0.4 -0.6
• 0.2-0.4
• 0-0.2
(B) Day 0 (Application Test - September 30, 2004 from 9:30 - 11:30 AM)
Figure 5-4. Predominant Wind Directions and Velocities (m/s) During Biosolids
Control Trial and Application Test
-------
Descriptive and inferential statistical comparisons were performed. The specific statistical
methods used depended on whether the data distribution was normal or log-normal after log
transformation. The geometric mean and geometric standard deviation were used in statistical
comparisons when the data were transformed. All statistical analyses were performed using Statistical
Analysis Software (SAS) MEANS, TTEST, and MIXED procedures (SAS, 2003). Scheffe adjustments
were made for multiple comparisons, and /-tests were used for single comparisons. Probabilities < 0.05
were considered statistically significant.
The airborne concentrations (CFU/m3) of bacteria and fungi collected with the six-stage
impactors and the concentrations of bacteria (only) collected with the biosamplers for both the control
trial and the application test are presented in Figure 5-5. Biosampler data for bacteria for each sampling
transect tended to be more variable than were the six-stage impactor data, particularly for transect UWA
during the control trial and for MOB and transect DWC during the application test, even though more
biosamplers were used for each transect. Since the six-stage impactor is the traditional sampling device
used in bioaerosol sampling, no inferential statistics were attempted using the biosampler data. The
samples on which the enteric virus analyses were conducted were collected by the biosamplers.
5
4
3
2
1
0
5
4
Log10 CFU/m3 3
2
1
0
5
4
3
2
1
0
Total Heterotrophic Bacteria-Biosampler
i
EH Control Trial
I Application Trial
Total Heterotrophic Bacteria—Six-Stage Impactor
rt
Total Fungi—Six-Stage Impactor
UWA
MOB
DWB
DWC
Figure 5-5. Bioaerosol Concentrations of Microorganisms for Mobile, Upwind, and
Downwind Sampling Locations
Inferential statistics were applied to the six-stage impactor data to determine which sampling
locations exhibited statistically significantly different concentrations of THE and total fungi. Table 5-1
summarizes those comparisons of sampling locations during the control trial and the application test that
were found to have statistically significant differences (defined as p < 0.05). Note that the sampling
locations in the right-hand column are more downwind than the sampling locations in the left-hand
column to which they are compared.
36
-------
Table 5-1. Summary of Sampling Location Comparisons with Statistically
Significant Differences for THE and Total Fungi (Six-Stage Impactor Data Only)
Microorganism
THE
Total Fungi
Sampling
Period
Control Trial
Application
Test
Control
Trial
Application
Test
Zone Comparison
UWA vs. MOB
MOB vs. DWB
UWA vs. MOB
UWA vs. DWC
MOB vs. DWB
MOB vs. DWC
DWB vs. DWC
UWA vs. MOB
MOB vs. DWB
MOB vs. DWC
UWA vs. MOB
UWA vs. DWC
MOB vs. DWB
Probability
<0.02
<0.04
<0.01
<0.01
0.01
0.02
0.04
O.01
0.02
0.02
0.01
0.02
O.04
The mean airborne concentrations of THE collected with the six-stage impactors were greater
during the control trial than during the application test for the MOB and all stationary sampling locations.
As shown in Figure 5-5, similar concentration profiles were observed for the various sampling locations
during both the control trial and application test. During both sampling periods, statistically significant
differences were observed between UWA vs. MOB and between MOB vs. DWB. The fact that the MOB
sampler exhibited differences from two of the three upwind and downwind sampling transects for the six-
stage impactors during the control trial suggests that the application machinery (without biosolids)
aerosolized dust particles containing microorganisms; therefore, any differences among these locations
during the application trial may or may not be attributable to the biosolids source. During the application
test, statistically significant differences were also noted for UWA vs. DWC and MOB vs. DWC. A
statistically significant difference was also noted for DWB vs. DWC, where the mean THE
concentrations for DWC were greater than those for DWB. This was an unexpected observation as it
would be anticipated that the concentrations would decrease with distance from the application area due
to factors such as dispersion.
The mean airborne concentrations of fungi were slightly greater during the control trial than
during the application test for UWA and MOB. In contrast to the THE data, the mean concentrations
during the control trial were less for DWB and DWC than during the application test. As shown in Figure
5-5, consistent with the THE data, similar concentration profiles were observed over the various sampling
locations for the control trial and application test. Also consistent with the THE data, statistically
significant differences were observed during both sampling periods between UWA vs. MOB and between
MOB vs. DWB. Again, any differences noted among these locations during the application test may or
may not be attributable to the biosolids source. The concentration differences for MOB vs. DWC were
statistically significant for the control trial but not for the application test. During the application test,
statistically significant differences were also observed for UWA vs. DWC. Contrary to the THE data,
DWB was not statistically different from DWC during the application test as the standard error of the
mean zonal difference (log-scale) of the bacteria was smaller than for the fungi.
The differences in bioaerosol concentrations for both types of microorganisms (THE and total
fungi) during the two sampling periods may have been influenced by documented differences in
37
-------
environmental conditions such as the time of day of sampling (2:00-3:30 PM for the control trial vs.
9:30 - 11:30 AM the next day for the application test), ambient air temperature (25°C vs. 19.8°C,
respectively), relative humidity (50.2% vs. 86.0 %, respectively), and solar index (644 watts/m2 vs. 404
watts/m2, respectively). The effects of these variables or combinations thereof were not evaluated. The
differences in these environmental conditions could be partially mitigated by sampling at the same time
on consecutive days.
The accuracy and precision of the methods used for collecting microorganisms in this field setting
were not determined. Even with expected high collection efficiencies of airborne bacteria, laboratory
assays for individual microorganisms may indicate low recoveries. Microbial ecology studies have
shown the culturability of microorganisms is low compared to actual counts in many environmental
settings (Fabbian et al., 2004). For example, the recovery efficiency for S. aureus seeded in a biosolids
sample was 8.7% (Rusin etal., 2003). In a bioaerosol study, less than 10% of the aerosolized bacteria
were capable of forming visible colonies with culture techniques (Heidelberg etal., 1997).
5.4.2 Particulate Matter. The mean values for the total mass concentration (ug/m3) of all particulates
<5.0 um (in the general inhalable size range for many bacteria and fungi) detected per unit time by the
GRIMM sampler immediately before the control trial was initiated, during the control trial, immediately
before the application test was initiated, and during the application test are presented in Figure 5-6. The
control sampling period produced greater mass capture than did the pre-control sampling period after log-
transformation and Satterwaite adjustment of the t-test due to unequal variances. The increase in
suspended particulate mass captured (approximately 90 (ig/m3/min) was likely due to equipment activity
on the application field. No statistically significant differences in particulate mass captured were noted
between the pre-application trial period and the application sampling period. One reason a difference in
a:
(U
O
500
400
300
200
100
Pre-Control
Control Trial
Pre-Application
Application Test
Figure 5-6. Mean Mass of Airborne Particulates (< 5.0 um) Captured by the
GRIMM Sampler Immediately Before and During Control and Application
Sampling Periods
38
-------
mass captured was not observed between these two periods may have been due to the biosolids used in
this study, which appeared to have reduced friability. The adsorption of microorganisms to particulates in
biosolids is one of the factors thought historically to influence the amount of bioaerosolized
microorganisms collected during field studies (Tanner, 2005).
5.5 Conclusions
In this specific outdoor environment, differences in bacterial and fungal counts were observed for
the six-stage impactor data for the various sampling locations during both the control trial and application
test. There were expected concentration differences at the MOB sampling location, but unexpected
increases were noted in the mean concentrations for THE at DWC compared to DWB during the
application test. Fine particulates of microorganism size did not appear to be aerosolized during biosolids
application. This lack of observed aerosolization may have been influenced by environmental conditions
and the biosolids additives, primarily polymer, and processing operations. The biosolids tended to have a
sticky or more gel-like consistency that may have diminished their friability and ability to produce fine
aerosolized particles.
5.6 Discussion and Lessons Learned
The nature of the biosolids used in this research may have had an impact on the bioaerosol data
and other data sets in this study. The sludge processing at the municipal WWTP produced a biosolids
material that was sticky and gelatinous, substantially reducing its friability and perhaps limiting its ability
to be dispersed as fine particles into the air. These properties may, in turn, have negatively impacted the
capture and detection of aerosolized microorganisms. For future studies in which the primary objective is
to maximize dispersion of biosolids and associated chemicals and microorganisms, it is recommended
that a more friable biosolids matrix be selected that maximizes uniform distribution of fine, flaky particles
to the soil and into the air. Application of liquid biosolids may also yield a more uniform distribution of
droplets to the soil and into the air. Conversely, application of a more agglomerating biosolids product,
from a practical sense, may help restrict the applied material to the immediate area and limit the spread of
airborne particles to downwind receptors.
The bioaerosol conclusions from this study may have been affected by the physical site location
and weather conditions. It is plausible that land application of biosolids during a humid, cloudy day
might allow air-sensitive and UV-sensitive bacteria to survive longer. In future studies involving
bioaerosols, a flat site that minimizes elevation differences in collection devices is recommended.
Furthermore, sites with consistent wind velocities and directions will simplify data interpretation.
Finally, it would be helpful if the selected sites were not surrounded by heavy vegetation to minimize
external microorganism influences form grasses, plants, and trees.
Since it was of interest (particularly for Task 3) to conduct this work at a site where biosolids had
never been land applied, there was no opportunity within the study design to replicate the application
tests. It is recommended that for future work a study design with focus on bioaerosol monitoring be
considered so that tests can be replicated to increase the power of the study and reduce uncertainties.
In addition, it would be appropriate to consider the use of air dispersion models to estimate
airflow regimes and guide the placement of sampling stations in developing a sampling array design.
Consideration should also be given to the use and placement of additional particle analyzers for the
real-time detection of airborne particles. Strategically placed particle analyzers would be especially
helpful to the downwind sampling operator to assist in defining the cross-sectional extent of the plume so
that the location of the bioaerosol sampling equipment could be optimized and moved if needed during
testing.
39
-------
6.0 TASK 2. VOLATILE ORGANICS AND ODOR SAMPLING RESULTS
6.1 Objectives
The study evaluated methods that measure concentrations of a selected group of VOCs and
odorants in Task 2. Some methods involved sample collection, while others were real-time
measurements. Emissions samples were collected upwind, within, and downwind of the application area
of the biosolids land application test site. Emissions sampling began prior to biosolids application and
continued for 2 days after application. In addition, real-time measurements were conducted on these
emissions using VPRM and single path OP-FTIR spectroscopy. OP-FTIR measurements were performed
on the day before and the day of biosolids application.
The specific objectives for VOC and odor sampling were to evaluate methods that:
1. Characterize the specific compounds and concentrations of volatile organics and odorants
(such as ammonia and hydrogen sulfide) emitted
2. Determine if the VOCs and odorants that were emitted were transported downwind of the
biosolids application area.
Several site-specific factors such as wind velocity, atmospheric stability, temperature, humidity,
the amount of material being applied, and the particular application method may affect the degree of odor
generated at a biosolids application site. Biosolids processing variables also influence biosolids odor
(Gabriel et al. 2006). This section discusses the various gas monitoring techniques that were evaluated to
estimate post-application off-gassing of selected organic and inorganic compounds and the extent to
which these selected compounds were observed. Table 2.2b lists the analytes, methods, sample types, and
sampling frequencies for the VOC and odor sampling efforts. Analytical methods were evaluated based
on whether data acceptance criteria were achieved.
6.2 Volatile Emissions Sampling and Measurements from Biosolids
A selected group of organic, inorganic, and odorous compounds was monitored within various
areas associated with this biosolids land application study as described in Section 2.4. Analyses of air
emissions were performed on bulk biosolids samples that were collected from the delivery truck as they
arrived on site from the treatment plant, from a temporary biosolids stockpile staged near the application
site, periodically from the exhaust air of five independent advective flux chambers that each covered a
1.44-m2 footprint on the surface of the biosolids land application site, and from biosolids samples that
were collected from the ground surface approximately 1 hr, 24 hr, and 48 hr after biosolids application.
In addition, prior to, during, and following biosolids application, trained odorant professionals surveyed
the land application site and surrounding vicinity for the presence of ammonia and sulfide odors using an
in-field Nasal Ranger® protocol (ASTM E544-99). The field survey and advective flux chamber
sampling locations are shown in Figure 6-1.
6.2.1 Headspace Analysis of Biosolids. The headspace emissions of biosolid samples that were
progressively collected from within the on-site stockpile at the time of land application (0 hr) and at 24 hr
and 48 hr after application to the field were determined with the use of specially fabricated glass
containers shown in Figure 6-2. Each of these three progressive samples was collected in triplicate, with
each triplicate sample representing a composite of seven sampling locations. The containers were air
tight and equipped with a sealing cap and a septum-sealed sampling port. The samples were refrigerated
at 4°C until analysis. Each sample was allowed to equilibrate to laboratory temperature (20°C) before
40
-------
ft .',
BiosoSsKtockpile
M' A -Ł$, \
One Additional Sa
175-Meterstothe
Soulhwesl of this Station
Explanation
5 Randomly Placed Flux Chambers
Ammonia and Sulfide Sampling
Figure 6-1. Aerial View of Test Site and Sampling Stations
headspace samples were obtained for analysis. A total of 2.0 cm3 of air was withdrawn from each 100-g
(normalized) biosolids sample placed into the sealed container. These headspace samples were analyzed
for specific compounds following EPA Method TO-15. An estimated volatile emission factor for each of
the three samples was determined using Equation 1:
Emission Factor (ng/g) = concentration (ng/L) x volume (L) x mass of sample"1 (g) (#1)
6.2.2 Advective Flux Measurements. Five flux chambers (A, B, C, D, and E) were randomly placed
within the biosolids application area as shown in Figure 6-1. The flux chambers are discussed in detail in
Section 2.4.2, and a conceptual setup is shown in Figure 2-6. Each chamber was constructed of stainless
steel, 120 cm on a side and 60 cm high, tunneling upward to an opening of 7.6 cm for collection of
exhaust that was covered with aluminum foil to prevent downdrafts. Ultra-high purity air was evenly
introduced into the bottom of the chamber for make-up air (sweep air) to generate sufficient sample
volume. Air samples were pulled into a 5.0-L Summa canister by vacuum and submitted for EPA
Method TO-15 analysis by GC/MS. Figure 6-3 shows a flux chamber operating during the post biosolids
application sampling event in the field.
Flux chamber emissions were also captured in Tedlar® bags and subsequently analyzed using a
SPME absorption technique. The coated fused silica fibers (75-um carboxen-polydimethylsiloxane fiber)
were exposed to the inside of the Tedlar® collection bags for 1 hr for sample equilibrium, then stored on
dry ice before being analyzed by GC/MS. The SPME fibers were injected directly into a heated GC/MS
port, and the contaminants were thermally released into the GC/MS column. Specific analytes included
trimethylamine, carbon disulfide, dimethyl sulfide, dimethyl disulfide, ethyl mercaptan, propyl
mercaptan, and butyl mercaptan. Methyl mercaptan was too volatile and reactive to be included as an
41
-------
Figure 6-2. Glass Vessel Used for Biosolids Headspace Analysis
Figure 6-3. Collecting a Flux Sample in the Field
analyte and also caused calibration concerns. The analyte response for SPME analysis was calibrated
using gas standards generated from certified permeation devices (VICI Metronics, Inc., Poulsbo, WA)
containing the pure compound.
The estimated flux rate for each of the flux chambers was determined using Equation 2:
Flux Rate (ug/m2/hr) = amount of contaminant collected in SUMMA canister (fig) x
ground surface area of chamber-1 (m2) x collection time"1 (hr) (#2)
6.2.3 Off-Site Odor Panel Analysis. Air samples were also collected from the flux chamber
emissions and captured in 12.0-L Tedlar® air sample bags via an air pump for odor threshold analysis. A
certified odor panel was used to conduct the odor analysis at an off-site forced chamber olfactometer
(ASTM, 1991). The sample's dilution level (with air) at which an odor is barely detected from three
sources (two are odor free) by a panelist is expressed in the standard units for odor threshold
measurement, dilutions-to-threshold (d/T).
42
-------
6.2.4 Direct Gas Measurements (Ammonia and Hydrogen Sulfide). Field measurements of
ammonia were performed using chemical sensory tubes (Draeger Tubes® [Model No. 6733231]) coupled
with a hand-operated vacuum pump with a detection limit of 0.100 ppmv. In addition, a direct reading
instrument was used for hydrogen sulfide. This gas was measured with a Jerome™ gold-film analyzer
(Arizona Instruments) with a detection limit of 0.001 ppmv. Locations were determined in part using
Nasal Rangers®, where values determined with the Nasal Rangers® were recorded.
6.2.5 Open-Path Fourier Transform Infrared Spectrometer. An OP-FTIR spectrometer was used
to measure "real-time" concentrations of VOC and ammonia emissions from the surface of the land
application test site as discussed in Section 2.4.1. The OP-FTIR system was linked to a particle counter
that summed the total amount of energy that a target compound absorbed between the FTIR and retro-
reflector array. Concentrations of specific compounds were quantified using the measurement of energy
absorbed within selected regions of the spectrum.
6.3
Results and Discussion
6.3.1 Head Space Analysis of Biosolids. Detectable levels of acetone, 2-butanone, methylene
chloride, toluene, dimethyl sulfide, and dimethyl disulfide were associated with the biosolids that were
removed from the stockpile and applied to the field. As shown in Figure 6-4, the estimated emission
factor was highest for dimethyl sulfide (range of 221.1 to 658.6 ng/wet g) among the VOCs for all three
sampling times (0 hr, 24 hr, and 48 hr). These data represent the arithmetric mean of three discrete field
samples, and the error bars indicate the standard deviations of the means. The concentrations for all of
the compounds detected decreased for each of the following 2 days, except for dimethyl disulfide, which
remained relatively constant or showed a slight increase over time. While methylene chloride was
suspected as a laboratory contaminant, the fact that the concentration decreased over time suggests that it
also could have been a volatile emission from biosolids. Insufficient data were generated to allow for this
differentiation. The other detected compounds are likely organic byproducts of the anaerobic digestion of
municipal biosolids typically found in very low concentrations emitting from biosolids as volatile gases.
•d
1
_g
•s
I
Ł
o
d Acetone
• 2-Butanone
DMethylene chloride
DayO Dayl Day 2
(Pre-Application) (Post Application) (Post Application)
Figure 6-4. Estimated Emission Factors Over Time of Biosolids Application
43
-------
Results from SPMEs exposed to the headspace of biosolids for 1 hr also resulted in detectable
concentrations of dimethyl sulfide (1.75 to 8.0 ppmv) and dimethyl disulfide (0.75 to 2.0 ppmv). Trace
levels (0.25 ppmv) of carbon disulfide were also detected in the SPMEs, but only in the control trial test.
No significant decreasing trend was observed in the SPME headspace results over the time period (i.e.,
1 hr, 24 hr, and 48 hr after biosolids application) that samples were collected.
6.3.2 Advective Flux. The estimated flux rates from the flux chamber air emissions (collected within
the Summa canisters and analyzed by GC/MS) resulted in detectable concentrations for acetone,
trimethylamine, dimethyl sulfide, and dimethyl disulfide. Estimated flux rates for these compounds were
greater than 1.0 |og/m2/hr for several of the post-application sampling times (t = 0 hr, t = 3 hr, t = 4 hr, and
t = 20 hr). Figure 6-5 illustrates the flux rates for these compounds at each sampling location. Other
contaminants such as isopropyl alcohol, 2-butanone, carbon disulfide, methyl isobutyl ketone, toluene,
2-hexanone, styrene, 1,2,4-trimethylbenzene, and 1,4 dichlorobenzene, were detected at trace levels
(2.4to3.8(ig/m3).
Calculated flux rates for trace compounds are not presented in Figure 6-5. In general, the rate of
emissions declined with time after biosolids placement; however, dimethyl sulfide and dimethyl disulfide
emissions persisted into the 20th hr after biosolids application when sampling was terminated. A longer
monitoring period was needed to determine if emissions continued into the afternoon of Day 2 or if they
subsided.
For chambers A through D, the flux rate increased between the 3rd hr and 4th hr after biosolids
application. This trend was most likely due to compound volatility caused by increased temperatures
during the early afternoon (approximately 2:30 PM). During the 4th hr, internal chamber temperatures
reached approximately 42°C. Since the 4th hr after application was the last time point measured on the
day of application, it is not known if the increased temperature in the afternoon continued to enhance
emissions.
The SPME apparatus is shown in Figure 6-6. The SPME fibers were exposed to flux chamber
emissions captured in Tedlar® bags. Tedlar® bags were collected at each sampling event for all flux
chambers and submitted for headspace analysis using the SPME fiber and GC/MS techniques for
comparison against the other field methods used for sampling and detecting odorants. However, time did
not permit on-site calibration of the GC/MS unit in the field. The results shown here are derived from
SPMEs that were exposed to Tedlar® bags in the field and analyzed via a calibrated GC/MS in the
laboratory after the field event was completed. Therefore, these data should be considered semi-
quantitative as there may have been potential compound losses during extended holding times, and trip
blanks were not available to confirm the extent of these potential losses.
SPME results from these field emission tests are shown in Figure 6-7. Dimethyl sulfide was
detected in all but two samples, and the levels remained approximately the same, ranging from 0.012 to
0.11 ppmv through t = 20 hr. Dimethyl disulfide was not detected at t = 0 hr and t = 4 hr, but was
detected att = 20 hr (range 0.07 to 0.15 ppmv). Trimethylamine concentrations were highest (range 0.01
to 0.04 ppmv) at t = 0 hr, then decreased. These levels are above the human detection threshold. Carbon
disulfide was detected at low levels in all flux chamber samples but was also present in the control trial
samples, indicating a potential source independent of the biosolids or possible interference within the
Tedlar® bags. Ethyl mercaptan, propyl mercaptan, and butyl mercaptan were not detected.
44
-------
4.75 -
4.50 -
4.25 -
4.00 -
3.75 -
f 3.50 -
™Ł 325-
Ł 3.00 -
™ 2.75
— 2.50 -
3^:
o: 2.00 -
X 1 .75 -
E 1'5°-
1.25
1.00 -
0.75 -
0.50 -
0.25 -
• acetone D trimethylam
|
Day-1 DayOt=0
ne
•
D dimethyl sulfide
H.
-1
-
• dimethyl disulfide
^
Day 0 1=3 hrs Day 0 1=4 hrs Day 0 1=20 hrs
Chamber A Time (Sampling Event)
4.75
4.50
4.25
4.00
— 3.75
Ł 3.50
™~~ 3.25
.i 3.00
|>2.75
— 2.50
S 2.25
tŁ 2.00
X 1.75
= 1.50
"- 1.25
1.00
0.75
0.50
0.25
• acetone n tr methylamine n dimethyl sulfide • dimethyl disulfide
n
H
•
rl
—
i
•
_ rl
Day-1 DayOt=0 DayOt=3hrs DayOt=4hrs DayOt=20hrs
ChamberC Time (Sampling Event)
• acetone n trimethylamine n dimethyl sulfide • dimethyl disulfide
r 00
4.75 -
4.50 -
4.25 -
4.00 -
_ 3.75 -
Ł 3.50 -
<Ł 3.25 -
.Ł 3.00 -
|> 2.75
— 2.50 -
Ł 2.25
Ł 2.00 -
X 1.75 -
= 1.50 -
"- 1.25
1.00 -
0.75 -
0.50 -
0.25 -
I — i
-i
i — H
jn
—
JM ,_, | | |
Day-1 DayOt=0 DayOt=3hrs DayOt=4hrs DayOt=20hrs
ChamberE Time (Sampling Event)
5.00
4.75
4.50
4.25
4.00
_ 3.75
^ 3.50
™~~ 3.25
= 3.00
™ 2.75
— 2.50
% 225
o: 2.00
X 1.75
2 1.50
"• 1.25
1.00
0.75
0.50
0.25
Chamber B
• acetone n trimethylamine n dimethyl sulfide • dirr
J
—
nil
ethyl disulfide
.1
Day-1 DayOt=0 DayOt=3hrs DayOt=4hrs Day 01=20 hrs
Time (Sampling Event)
4.75 -
4.50 -
4.25 -
4.00 -
3.75 -
^ 3.50 -
J- 3.25 -
E 3.00 -
™ 2.75
— 2.50 -
1 2.25
o: 2.00 -
X 1.75 -
"• 1.25
1.00 -
0.75 -
0.50 -
0.25 -
• acetone n trimethylam ne n dimethyl sulfide •
D
Chamber D
n
J
n
1
Lr
dimethyl disulfide
l.m
ay-1 DayOt=0 DayOt=3hrs DayOt=4hrs Day 01=20 hrs
Time (Sampling Event)
Figure 6-5. Calculated VOC Flux Rates for Acetone, Trimethylamine, Dimethyl Sulfide, and
Dimethyl Disulfide for up to 20 hr After Biosolids Application in the Test Area
45
-------
Figure 6-6. SPME Fibers Exposed to Emission Samples Collected in
Tedlar® Bags from Flux Chamber Off-Gas
0.16-
Qi
,S
6
0 0.1
L-
E
O
X
O.QZ-
IB 1
*^-
^.c
s
c
c
lelivery Truck
ample
MDS = 0.91
IMS = 2.9
1
r
D Dimethyl Disulfi
§ Carbon Disulfid
H Trimelhylamine
3e (DMOS)
s (CDS)
(DMS)
(TMA)
n
I .
,
||
"V^OO^Q ^-^0^^^^
kit ,
•
71
'G'O'^'^' ^' ^' o
*•' v-
r
—
I
—
i
—
I
« ' O ' O ' O ' <<, '
Day -1
Baseline
Day 0
Time = 0 h
Day 0
Time = 3 h
Day 0
Time = 4 h
Day 1
Time = 20 h
Figure 6-7. Concentration of VOCs from SPMEs Exposed to Air Emissions of
Flux Chambers
46
-------
A comparison of flux chamber emission results between the SPME samples collected in Tedlar®
bags and the Summa canister method analyzed by GC/MS revealed that dimethyl sulfide, dimethyl
disulfide, and trimethyl amine were found consistently with both methods. Overall, concentration results
from the canister method were an order of magnitude higher than with the SPMEs (data not shown),
indicating potential losses over time as previously discussed or other interferences resulting in inadequate
sorption onto the SPME fiber. On-site processing and analysis of exposed SPME fibers may have
produced higher volatile emission concentrations resulting in more comparable results between the two
sampling approaches.
6.3.3 Direct Gas Measurements (Ammonia and Hydrogen Sulfide). Ammonia was not detected in
above-ground air samples within the application zone area during the control trial. Immediately after the
application test, ammonia was detected within the range of 0.10 to 0.90 ppmv for near-ground samples
within the application area, and from a flux chamber exhaust sample at 15 ppmv. Hydrogen sulfide was
detected at levels near the recognition threshold at concentrations from 0.002 to 0.050 ppmv within the
application zone area during the control trial, and at levels of 0.007 to 0.021 ppmv directly behind the
moving biosolids application equipment. The exhaust from the biosolids applicator machinery may have
been responsible for some of the hydrogen sulfide detected. Immediately after the application trial,
hydrogen sulfide was detected at concentrations within the application area slightly lower than those
measured during the control trial, within the range of 0.001 to 0.007 ppmv for near-ground air samples,
and from a flux chamber exhaust sample at 0.160 ppmv. The highest measurements for each of the gases
never approached any health criterion or guidance level. As expected because of their high vapor
pressures, the concentration of both gases decreased during the 2nd day (the day after application) to the
detection limit, and was below detection limits within 4 days. The concentration of both gases was below
detection limits 400 m downwind of the application area during the application trial.
6.3.4 Off-Site Odor Panel and On-Site Nasal Ranger* Analyses. On-site odor measurements were
conducted on the biosolids application site using hand-held olfactometers (Nasal Rangers®, St. Croix
Sensory, Inc., Lake Elmo, MN). These included measurements (ASTM E544-99 [ASTM, 2004]) made
on ambient air at selected locations around the application area. In addition, emission samples for off-site
odor analysis (ASTM E679-91 [ASTM, 1991]) were collected from flux chambers B and D in 12.0-L
Tedlar® bags via an air pump.
The 12.0-L Tedlar® bag samples obtained by collecting emissions from flux chambers B and D
were forwarded to St. Croix Sensory, Inc. for olfactometry analysis by a certified odor panel of three
individuals to confirm the presence and level of odor using ASTM E679-91, as described above. In one
of three evaluation ports presented to the panel, odor-free air was diluted with increasing levels of
contaminated air from the Tedlar® bags. Odor-free air was provided to the other two ports at the same
airflow rate. Each panelist was asked to identify the port containing the diluted odor.
The sample dilution level at which odor is first smelled by the panel members is termed the
detection threshold (DT). The sample being analyzed by the panel members is further diluted with
contaminated air until the panel members recognize the source from a quadrant of odor categories, e.g.,
rotten cabbage, rotten meat, rotten eggs, strongly fecal, etc. This dilution level is called the recognition
threshold (RT).
The unit used in this report to express both DT or RT is dilutions-to-threshold or d/T, where d/T
is defined as the volume of uncontaminated odor-free air provided to each panelist at the beginning of the
test divided by the volume of contaminated air required to be bled into the uncontaminated air source to
reach the respective threshold (either DT or RT). For example, a d/T of 100 is equivalent to one volume
of odor-free air divided or diluted by 1/100 volume of contaminated air, while a d/T of 1 is equivalent to
one volume of odor-free air divided or diluted by one volume of contaminated air. Note that the smaller
47
-------
the d/T value becomes the greater becomes the dilution of odor-free air with contaminated air. Odor
concentrations for each sample were reported as the geometric mean of the individual panelist's
thresholds.
The results for the dynamic dilution and odor threshold analyses performed on Tedlar® bag
samples collected from the flux chamber emissions are shown in Table 6-1. DT and RT increased with
time indicating that the odors became stronger with time. This is thought to have been due to increasing
volatilization resulting from rising temperatures throughout the day, and also because of anaerobic
degradation of organic sulfur compounds in the biosolids. For the same sample, DT will always be larger
than RT because less contaminated air is needed to detect the odor than to recognize the type of odor.
In Table 6-1, laboratory DT values defined for the control trial (before application) were 70 to 90
d/T. This is a typical background DT range for most rural agricultural areas as measured by a highly-
trained, highly-odor-sensitive, off-site panel under controlled laboratory conditions. The panel measured
odor DT at 500 to 1,000 d/T from flux chamber sample bags during the first 4 hr. After 22 hr, the panel
measured 2,500 to 6,100 d/T from the two flux chamber sample bags. These data suggest that the volatile
odors associated with degradation products increased near the ground surface during biosolids application
and may increase for a limited time period after application due to ongoing biodegradation and
volatilization. Off-site odor panel analyses were not conducted on samples taken more than 22 hr after
biosolids application. These odor data are in agreement with observations of increasing concentrations
for organic sulfur compounds and other emissions over time shown in Figure 6-7 for samples collected
from the same flux chambers.
Table 6-1. Results of Dynamic Dilution Olfactometry Analysis
Flux
Chamber
B
D
B
D
B
D
B
D
B
D
Event
Control Trial
After
Biosolids
Application
Time
(hr)
NA
0
o
J
4
22
Laboratory Olfactometry Analysis
ASTM E679-91
Detection
Threshold (d/T)
70
90
500
330
620
540
620
1000
2500
6100
Recognition
Threshold (d/T)
55
70
310
240
330
290
360
540
1400
2500
NA = not applicable
With the exception of the data for flux chamber D at 22 hr, RT values fell within a fairly narrow
range of 54% - 79% of the corresponding DT values. This observation indicates the panelists were
recognizing odor categories at roughly the same levels of increased dilution with contaminated air after
the DTs had been established for nine of the 10 sampling conditions. The ratio of RT-to-DT of 41% for
flux chamber D at 22 hr lies outside this range, suggesting either a data outlier or a more difficult-to-
recognize odor.
48
-------
With regard to the odor measurements made on site with the Nasal Ranger® instruments, only
ambient trace odor levels were detected at 1.5 m ags within the biosolids application area prior to
biosolids application. Immediately after biosolids application, the odor DT levels at 1.5 m ags were 2 to 7
d/T at 25 m upwind of the application area and 15 to 30 d/T approximately 25 m downwind of the
application site. Odors were not detected at distances greater than 75 m from the application area.
Approximately 22 hr after biosolids application, odor DT levels at 1.5 m ags were roughly 15 d/T in the
application area and odor was undetectable above background at levels elsewhere in the project area.
After 48 hr, odors were barely detectable downwind of the site, and after 196 hr, no odors were detected
above background in any location on the site, consistent with other biosolids application studies (Krach et
a/., 2008; Hamel et a/., 2004). As expected due to the high rates of air dispersion associated with
measurements made in ambient air, the odor measurements decreased vertically above ground surface
(ags) and horizontally at increased distances from the application area. The downwind values may have
been greater if the biosolids application area was larger or if the biosolids loading was greater.
As opposed to the pristine laboratory conditions under which off-site odor measurements are
conducted, on-site odor measurements are carried out by roving observers or receptors using hand-held
olfactometers substantially less sensitive than the trained human nose. In addition, on-site observers are
subjected to background odors to which the off-site panel are not exposed. Understandably, then, the on-
site DT values determined with the Nasal Rangers® were substantially less than those measured for the
flux chambers by the off-site panel. Also, as expected, the measured off-site DT values decreased with
time after biosolids application as odors were dispersed, rather than increased with time as did the levels
measured in the flux chamber emissions due to enhanced volatility and biodegradation of trapped organic
compounds. While the Nasal Rangers® are not as sensitive as the off-site panel, the results obtained with
these instruments should be considered the more valid data set for accurately representing on-site odor
conditions to an on-site human receptor.
6.3.5 Open-Path Fourier Transform Infrared Spectrometer Measurements. VRPM was
performed by deploying 10 mirrors in various locations on a vertical plane in line with the scanning
OP-FTIR sensor. The vertical plane was configured as close to perpendicular to the prevailing wind
direction as possible. By combining measured wind data with the path-integrated concentration data, the
emission flux through the vertical plane was calculated. Two VRPM configurations were used, one
upwind and one downwind. Each VRPM configuration consisted of five mirrors, three placed along the
surface and two mounted on a vertical structure (scissorjack) positioned approximately 8 m and 12 m ags.
A Midac™ OP-FTIR was used in the upwind configuration, and an IMACC™ OP-FTIR was used in the
downwind configuration. See Figure 2-5 for the on-site configuration that was used to deploy the OP-
FTIRs.
An additional IMACC™ OP-FTIR was deployed along a single path (one-way path length of
approximately 141m) over the center of the application area. Data from this instrument were collected to
measure ammonia concentrations and determine whether VOCs were present in the application area. No
interpretive data were found relative to VOCs and, therefore, are not presented in this report.
The ammonia measurements were collected on Day -1, Day 0 during application, and Day 0 for
several hours after application. On Day -1, baseline measurements helped to determine the VOC
contribution from various generators and equipment and from the tractor during the baseline sampling
event. On Day 0, measurements were collected during the entire application. Measurements continued
for several hours after the application ended to evaluate the decrease of emissions overtime.
6.3.5.1 Baseline Ammonia Measurements. During Day -1, baseline measurements were collected along
each of the VRPM configurations. The upwind VRPM configuration failed to detect the presence of any
ammonia plumes during the entire duration of this measurement period. The downwind VRPM
49
-------
configuration measured negligible ammonia concentrations throughout the duration of the baseline
measurements.
6.3.5.2 Ammonia Measurements During Biosolids Application. During the period of biosolids
application, measurements were collected along each of the VRPM configurations. The VRPM
procedure calculates the concentration values for every square elementary unit in a vertical plane. Then,
the VRPM procedure integrates the values, incorporating wind speed data at each height level to compute
the flux. The concentration values are converted from parts per million by volume (ppmv) to grams per
cubic meter (g/m3) taking into consideration the molecular weight of the target gas. This enables the
direct calculation of the flux in grams per second (g/s) using wind speed data in meters per second (m/s).
The calculated ammonia flux from the upwind VRPM configuration during and immediately after
biosolids application was 0.006 g/s. Figure 6-8 shows the reconstructed ammonia plume for the upwind
VRPM configuration. The calculated ammonia flux measured from the downwind VRPM configuration
for the same time period was 0.063 g/s. The reconstructed ammonia plume from these measurements is
shown in Figure 6-9. The bold vertical lines in Figures 6-8 and 6-9 and in the figures to follow in Section
6.3.5.3 represent the physical height of the retro-reflectors on the ground and in the tower. The angled
lines projecting from the origin indicate the path length. The positioning of the retro-reflectors and the
path lengths is illustrated for clarity in Figure 6-8.
Upwind ammonia concentration contours are shown in Figure 6-8. Concentrations were greatest
near the ground and in the immediate vicinity of the biosolids application area at approximately 0.018
ppm and dissipate radially to 0.004 ppm at a height of approximately 14m and over a horizontal span of
approximately 95 m.
The ammonia concentration plume in the downwind configuration (Figure 6-9) is dispersed over
greater horizontal and vertical distances (approximately 21m ags and 180 m laterally). The ammonia
concentrations were an order of magnitude greater than those for the upwind configuration at 0.15 ppm
near the ground and 0.03 ppm for the most distant radial contour.
6.3.5.3 Ammonia Measurements After Biosolids Application. In order to investigate the rate of
emissions decay, measurements continued for several hours after biosolids application. The upwind
VRPM configuration measured negligible ammonia concentrations during the post-application period,
and, therefore, the results are not shown graphically. However, the downwind VRPM configuration
detected ammonia plumes for several hours after the application ended. Figures 6-10 and 6-11 depict
reconstructed ammonia plumes in the downwind location at 2 and 3 hr, respectively, after biosolids
application.
The calculated emission flux rates for ammonia 2 and 3 hr after biosolids application were
approximately 0.036 and 0.009 g/s, respectively. There is an observed decrease in ammonia emissions
over the 2 to 3 hr time interval.
50
-------
Concentrations are in ppm
Flux = 0 006 gts
Ground Retro-Reflectors
40 SO 80
Crosswind Distance [meters]
100
120
Figure 6-8. Reconstructed Ammonia Plume from the Upwind VRPM Survey
During Biosolids Application on Day 0
22
20
18
16
•5T 14
r 12
•5
I 10
Concentrations are in ppm
Flux = 0.063 g/s
80 100 120
Crosswind Distance [meters]
160 180
Figure 6-9. Reconstructed Ammonia Plume from the Downwind VRPM Survey
During Biosolids Application on Day 0
51
-------
22
20
IS
16
"uT 14
V
| 12
I 10
8
e
4
2
Concentrations are in ppm
Flux = 0.038 g/s
20
40
60
80 100 120 140
Crosswind Distance [meters]
160
Figure 6-10. Reconstructed Ammonia Plume from the Downwind VRPM Survey
Approximately 2 hr After Biosolids Application
Concentrations are in ppm
Flux = 0.009 g/s
I 10
8
6
4
2
80 100 120 140
Crosswind Distance [meters]
160
180
Figure 6-11. Reconstructed Ammonia Plume from the Downwind VRPM Survey
Approximately 3 hr After Biosolids Application
52
-------
The height of the downwind ammonia plume decreased to approximately 16m ags 2 hr after
application from an initial height (t = 0 hr) of approximately 21m and maintained this height at 3 hr. The
lateral plume dimension at 2 hr after application decreased to approximately 25 m in width. At 3 hr after
application, the lateral extent of the plume unexplainably expanded to a width of 100 m but decreased to a
concentration that was an order of magnitude less than that observed at 2 hr.
6.4 Conclusions
During and after biosolids application, various organic and inorganic odors were detected above
background levels approximately 1.5 m ags and in concentrated samples collected from flux chambers.
Analysis of air samples collected from flux chambers using GC/MS confirmed that odors were primarily
associated with compounds such as dimethyl sulfide, dimethyl disulfide, and trimethylamine. Acetone,
which was quantified by GC/MS, is not believed to be a common constituent of biosolids or their
biodegradation breakdown byproducts; however, it appeared at relatively high flux rates (1.25 (ig/m2/hr)
in flux chamber samples. SPME analysis of headspace samples that were equilibrated with biosolids
confirmed the presence of dimethyl sulfide, dimethyl disulfide, and trimethylamine at approximately an
order of magnitude less than those concentrations determined through GC/MS analysis.
Volatilization and further degradation of biosolids resulted in increasing detectable concentrations
of odors captured in flux chamber emission samples for 3 to 22 hr after application. Biosolids application
also increased the near-surface concentration of hydrogen sulfide and ammonia immediately after
application, but concentrations at 1.5 m ags were similar to background conditions, as observed in
Draeger Tube® sampling. Concentrations of ammonia and hydrogen sulfide at ground level began to
decrease within 3 to 4 hr after application, then to non-detectable levels within 20 hr.
OP-FTIR results confirmed the in-field detection of ammonia collected via Draeger Tube® and
Nasal Ranger® measurements. The VPRM protocol showed a decreasing ammonia emission flux rate,
initially measured at 0.063 g/s, decreasing to 0.036 g/s approximately 2 hr after the application ended, and
further reducing to 0.009 g/s 3 hr after the application ended.
6.5 Discussion and Lessons Learned
While odor was virtually undetectable using Nasal Rangers® 2 days after biosolids application at
this site, and observed still to be undetectable above background levels at 4 days after application, it is not
known to what extent changing weather patterns may have impacted this apparent trend and whether or
not odors may have "rebounded" after monitoring was terminated. Furthermore, differences were
observed in environmental conditions between the control trial and the biosolids application test that may
have influenced volatilization results. For the control trial (Day -1), which occurred between the hours of
2:07 PM to 3:35 PM on September 29, 2004, ambient air temperature was approximately 25°C, relative
humidity was 50%, and the solar index was 644 w/m2. These values are compared to the application test
(Day 0 ) where biosolids application was conducted between 9:37 AM and 11:26 AM, the ambient air
temperature was approximately 20°C, the relative humidity was 86% (with a visible light fog), and the
solar index was 404 w/m2.
Although the use of chambers seemed to be an effective approach for measuring flux emissions,
the elevated temperatures inside the chambers most likely increased the volatility of the organic
compounds measured and may have enhanced microbial activity on the ground. Also, in retrospect, a
more focused sampling schedule with better resolution between 4 and 22 hr after biosolids application,
combined with other weather effects data, would have provided a better understanding of the extent of
volatile emissions and the recalcitrance of particular compounds.
53
-------
Lastly, the post-application field sampling efforts in the application zone associated with this task
and other tasks for this study created logistical challenges. Activity in the field from vehicles used to
transport samples and equipment in some instances obstructed the OP-FTIR pathway to reflectors
positioned on the ground surface and may have created aerosols that were introduced into the beam path.
A modified design should be considered for future studies that reduces or eliminates these types of
interferences. Development of such a design is beyond the scope of this project, but most certainly would
begin with a reduction in the number of simultaneous sampling activities that can lead to interference.
54
-------
7.0 TASK 3. LAND SAMPLING RESULTS
7.1 Objectives
The soil sampling portion of this study focused on methods to measure the concentrations of
microbes and chemicals before and for 98 days after biosolids application. The specific objectives for
Task 3 were to evaluate methods that:
1. Characterize the quantity and distribution of biosolids applied
2. Characterize the microbial community quantity and structure
3. Measure the fecal coliform density
4. Measure the concentration of alkylphenol ethoxylates (APEs) and their degradation products
5. Screen samples for terrestrial ecotoxicity.
Several concepts were considered in selecting these objectives. Objectives 1 and 2 were
identified to better describe observations of this study. Biosolids application methods can vary
considerably. One objective was to evaluate a method to measure the quantity and distribution of
biosolids applied at this site. Biosolids distribution was determined by measuring the dry mass and ash
mass of applied material for each replicate plot. These data were evaluated for consistency between plots
and for spatial variation within a plot. PLFA measurements were used to characterize the size and
structure of the microbial community. Biosolids may alter the microbial community by adding nutrients,
organic matter, and microbes. PLFA data provided insight into the magnitude and diversity of these
changes. Objectives 3 and 4 were based on specific recommendations of the 2002 NRC report to study
the impacts of pathogens and chemicals, such as the surfactants used in cleaning products and detergents.
Fecal coliforms were measured as potential indicators of pathogenic bacteria. The NRC report (2002)
expressed concern about the persistence of organic compounds in environmental matrices and the
potential for transport within soils and to other environmental media. Surfactants, such as APEs, are
produced and used in large volumes. These compounds and their degradation products, such as
octylphenol (OP) and nonylphenol (NP), have been reported to disrupt endocrine activity. Objective 5
was included to evaluate whether biosolids introduce toxicity based on observed responses in some
ecologically relevant organisms. Soil toxicity was screened using a 14-day earthworm mortality bioassay
and a 5-day assay measuring seed germination and root elongation in lettuce and oats. One stated reason
to land apply biosolids is to serve as a soil amendment, increasing the soil organic matter and supplying
nitrogen and phosphorus to enhance plant growth. In this study, observed responses were compared
before and after application to evaluate whether biosolids application improved growth or survival in
these toxicity assays.
Data collection was not limited to information needed to meet study objectives. For example,
supporting information such as soil agronomic characterization, temperature, and weather data were
gathered. In addition, microbial indicators such as VHO, Salmonella, Enterococcus spp., THE, enteric
viruses, and male-specific coliphage were analyzed on selected samples.
This section discusses the data collected and, where possible, interpretation of results and
recommendations for future studies. Table 2-2d lists the analytes, methods, sample types, and sampling
frequencies for the land sampling effort. Analytical methods were evaluated based on whether data
acceptance criteria specified in the QAPP were satisfied. When data quality was adequate, further
analysis was conducted to interpret the data. Recommendations were made to improve methods and
55
-------
sampling techniques in the discussion of the data sets. Recommendations for future studies were also
included.
7.2 Overview of Field Plots
For this field-scale research project, a research material comprised of anaerobically digested
biosolids mixed with lime and dewatering additives was applied at agronomic levels to a fescue field (see
Section 3.0 for more information on biosolids application). Biosolids had not been applied to this field in
the past. This material was applied in a 100-m diameter circle by a side discharge manure spreader. The
soil sampling occurred in three replicate plots of 3 m across by 6 m long, randomly located within the
biosolids application area (see Figure 2-7). All samples were collected based on a predetermined
randomized sampling plan. In the first half of each plot (3 m by 3 m), the distribution of biosolids at Day
0 (day of application) was measured. Ecotoxicity samples were also collected in this section. In the
second half of each plot (3 m by 3 m), the soil was sampled at three selected locations (30 cm by 30 cm)
for each sample event using a grid system. In each location, multiple samples were removed, but each
location was sampled only once. Soil samples were collected from the three replicate plots prior to
biosolids application (two sample events) and for 98 days following application (five sample events). The
post-application sample period was selected based on site restrictions for fields where Class B biosolids
have been applied. Site restrictions are dependent on whether biosolids are allowed to remain on the
surface or are incorporated within a 4-month period. Thus, the sample period was selected to be slightly
less than 4 months. Sample collection methods, sample events, sample depths, and sample compositing
varied depending on the analyte.
7.3 Sample Collection and Analysis
7.3.1 Collection Methods. For most land sampling activities, a coring method was used to obtain the
sample. Biosolids distribution and ecotoxicity samples were collected using different methods. Coring
equipment consisted of a 60 cm stainless steel split spoon sampler (6.3-cm diameter) that was driven with
a hydraulic hammer to the appropriate depth based on the analyte of interest for that sample. Specific
sample locations within a grid were selected to include biosolids based on visual inspection after biosolids
application. The spoon was pulled out of the soil using a core pulling device, and the core sample was
retrieved by opening up the split spoon and processing the soil core to remove the specific depth intervals
using a decontaminated and sterile putty knife. Soil segments were transferred into clean glass or plastic
jars. Sterilized sample jars were used for samples that were to undergo microbiological analyses. Split
spoon samplers were sterilized by autoclave prior to use in the field. Samples were shipped for analysis.
Biosolids distribution was collected by securing 30-cm x 30-cm squares of geotextile to the soil
surface at preselected random locations prior to biosolids application. Within 1 hr after application, the
squares were lifted off the soil and any biosolids on the surface of the square (along with the geotextile)
was placed in a sample bag. Samples were shipped for analysis of dry mass, ash mass, and volatile solids.
A sample collected for measuring ecotoxicity was generated by compositing soil samples from
four preselected random subsample locations in a plot. Each subsample was a soil cube of 15 cm x 15 cm
x 15 cm (LxWxD). For each sample, the subsamples were placed in a single bucket and shipped for
analysis.
Weather data were collected on site using an HOBO Weather Station Model H21-001. Prior to
biosolids application, the weather station was placed adjacent to the field site. Immediately after
biosolids application, the weather station was placed on the application area next to Plot 3. The weather
station monitored parameters that could affect the study variables, including rainfall, soil water
content/moisture, soil temperature, air temperature, relative humidity, dew point, and solar radiation.
56
-------
7.3.2 Analyte Specific Sample Collection and Analysis Information. Table 7-1 lists the sample
events and the number of samples collected for each analyte. The pre-application soil sampling event
took place 3 days before air sampling in order to avoid confusion in logistics with the air sampling team.
Biosolids distribution was measured on the day of application as described in Section 7.3.1.
Sample collection for PLFA analysis used the coring technique. In each plot, PLFA samples
were collected at three locations and at three depths (0 to 5, 10 to 15, and 20 to 25 cm) for each location.
Samples were not composited. Thus, each sample event generated a total of 27 PLFA samples.
Table 7-1. Sample Analytes, Events, and Sample Numbers for Land Sampling
Analyte
Biosolids dry
mass/volatile solids
FAME
Fecal coliforms(b)
APEs, Bisphenol A
Soil characterization
Ecotox
Microbial indicators1^
Microbes(d)
Sampling Event
Soil
Day
-35
NS
27
9
27
9
o
J
3
o
5
Day
-3
NS
27
9
27
NS
NS
3
NS
Day
0
60
27
9
27
9
o
J
3
NS
Day
14
NS
27
9
27
NS
NS
3
NS
Day
28
NS
27
9
27
9
NS
3
o
6
Day
63
NS
27
9
27
NS
NS
3
NS
Day
98
NS
27
9
27
9
o
J
3
o
J
Biosolids3
Day
0
NS
3
3
3
1
NS
3
3
NS = no sample collected
(a) A composite sample formed by combining seven subsamples from random locations in the biosolids
pile prior to biosolids being applied to land.
(b) MPN method analyzed by Environmental Associates, Inc.
Sent to Environmental Associates, Inc. for VHO, Salmonella, enteric viruses, and male-specific
(c)
(d)
coliphage analysis.
Sent to USD A for total heterotrophs, fecal coliforms, S. aureus, Enterococcus spp., E.coli, and
C. perfringens analysis.
Samples collected for fecal coliform analysis used the coring technique and sterile equipment. In
each plot, these samples were collected at three locations at the 0 to 5 cm depth. Thus, each sample event
generated a total of nine fecal coliform samples.
Samples for measuring APE concentrations were collected using the coring technique. In each
plot, these samples were collected at three locations and at three depths (0 to 5, 10 to 15, and 20 to 25 cm)
per location. APE samples were not composited. Thus, each sample event generated a total of 27 APE
samples.
Similarly, the coring technique was used to collect soil characterization samples. In each plot,
these samples were collected at three locations and at three depths (0 to 5, 10 to 15, and 20 to 25 cm) for
each location. These samples were composited based on depth. For example, the three 0 to 5 cm samples
from a plot were mixed together. Thus, each sample event generated a total of nine soil characterization
samples.
57
-------
Samples for ecotoxicity measurements were collected as described in Section 7.3.1 and in
accordance with the sampling schedule (Table 7-1). At each sample event, a total of three composite
samples were collected.
Samples for microbial indicator and microbe analysis were collected using the coring technique
and sterile equipment. Three cores were collected from a plot, and the 0 to 5 cm portions were
composited to form a microbial indicator sample. This procedure was repeated in each plot. Thus, a total
of three microbial indicator samples was generated at each sample event listed in Table 7-1. Samples
collected to quantify specific groups of microbes (microbe samples) were collected in a similar fashion.
Additional information about sample collection, compositing, and analysis was specified in the
QAPP and is discussed in later sections.
7.3.3 Statistical Analysis. The same general statistical approach was used in the analyses of the data
for Objectives 1, 3, 4, and 5. General linear analysis of variance (ANOVA) models were fitted to the data
for each study component, and a backward selection method was used to reduce the terms in the model.
In backward selection, insignificant variables are removed from the model one at a time, starting with the
least significant, until only significant variables or variables of interest remain in the model. For the
objectives where more than one variable was measured (dry and ash mass, APEs, and soil
characterization data), multivariate ANOVA (MANOVA) models were used to account for correlation
between the measured variables. In addition, the distribution of the model residuals was explored after
each model was fitted to determine the extent to which the underlying assumptions of ANOVA were met.
In general, interactions between the factors were included in the model, where appropriate. The
predictive factors that were included in the models for each analysis were:
• Biosolids distribution - plot, row, and column locations (separate models fitted to each plot)
• Biomass by PLFA - time, plot, depth, and two-way interactions
• APE concentration - time, plot, depth, and time-by-plot interactions
• Fecal coliform density - time, plot, and time-by-plot interactions
• Terrestrial ecotoxicity - time, plot, sample soil concentration, and all two-way interactions.
For the root elongation component of the ecotoxicity data, separate models were fitted for lettuce
and oats.
For those analyses where predictive factors were found to be statistically significant, multiple
comparison procedures were applied to determine the specific ways in which the measured variable was
affected by the significant factor. When time was a significant factor, the multiple comparison analyses
compared post-application levels to pre-application levels (controls) as well as comparing among
post-application levels. For other factors, all factor levels were compared. Multiple comparisons were
performed using Scheffe's method to control for overall error rate among all possible contrasts. For those
analyses where there were statistically significant interactions, multiple comparisons were performed by
fixing one of the interacting factors at each of its levels (e.g., fixing depth at each of its three values) and
comparing the means among the levels of the second interacting factor.
For the microbial community structure part of Objective 2, the PLFA data were subjected to
multivariate data analysis in order to characterize the relationships between the samples by category
prediction methods. Hierarchal cluster analysis (HCA) and principal component analysis (PCA) were the
two exploratory methods used in examining this dataset. All PLFA data were transformed to a mass
percent basis prior to analysis. HCA organized samples by similarity (least distance). The methods of
HCA used in this study were incremental, centroid, group average, and median. The incremental method
58
-------
of HCA gave the deepest branching and tightest clusters. PCA was run with preprocessing by mean-
center, maximum factors of 10, no rotation, and cross validation. The scores and loadings plots were
examined for the presence or absence of groupings and for the possibility of outliers. If outliers were
suspected, then an outlier diagnostic was performed by plotting sample residuals vs. Mahalanobis
distance (Beebe, et.al., 1998). If the sample was more than two threshold values away, it was removed as
an outlier, a new PCA was performed, and the process was repeated.
7.4
Data and Results
7.4.1 Soil Characterization. Soil characterization data collected throughout the study, available in
Appendix D, were used to define the conditions at this site. Statistical analysis identified differences in
soil properties with depth, plot, and time as well as interaction effects between plot and depth.
Differences by depth were observed in most soil characterization measurements, followed by differences
between plots. Statistical differences based on sample time were observed in six measurements, but these
differences did not exhibit consistent trends (such as an increase or decrease after biosolids application).
Similarly, interactions between plot and depth were observed in a few soil characterization measurements,
but did not exhibit consistent trends.
Soil properties data are summarized in Table 7-2. Since the depth parameter showed the most
differences, data have been grouped based on depth. For example, statistical analysis identified that the
bulk density was different in each horizon; therefore, averages and standard deviations have been
calculated for each depth. For cation exchange capacity, the surficial depth was distinct from the deeper
samples, so data were averaged in two groupings (0 to 5 cm depth range, or the combined 10 to 15 and 20
to 25 cm depth ranges). For each measurement type, the final column lists other variables for which
statistical differences were observed.
Table 7-2. Soil Characterization Data for Land Sampling Plots
Soil Measurement
Soil Composition
Sand (%)
Silt (%)
Clay (%)
Bulk density (g/cm3)
Cation exchange capacity
(meq/100 g)
Moisture at 1/3 bar (%)
pH
Organic matter (%)
Total N (%)
Total P (mg/kg)
Olsen P (mg/kg)
Soluble salts (mmhos/cm)
Biosolids
5.0
78.6
16.4
0.89
22.0
158.4
7.4
NA(a)
NA
24453
176
4.09
Soil Depth
0-5 cm
10-15 cm
40.7 + 6.5
20-25 cm
30.7 + 7.0
24.8 + 5.7
34.5+6.5
0.84 + 0.05
12.4 + 0.6
(b)
33.7 + 3.2
6.0 + 0.2
7.2 + 1.1
1.04 + 0.03
44.5 + 8.4
0.99 + 0.06
10.2 + 0.8
26.1 + 1.7
6.7 + 0.7
1.9 + 0.2
(b)
33.25+4.6
6.8 + 0.2
1.2 + 0.2
0.412 + 0.500
1152 + 114
43+6
0.24 + 0.13
447 + 44
322 + 28
6 + 2
0.11+0.04
Other
Statistical
(c)
Differences
T
P,DP
P,T
T, P, DP
P, T
DP
P,DP
T,DP
P
T
DP
(a) NA not analyzed
(b) The two depth ranges 0 to 5 and 20 to 25 cm are statistically similar.
(c) Other statistical differences for soil measurements: P, plot; T, time; and DP, interaction of depth by plot.
59
-------
Soil characterization data were useful in documenting conditions at this site, and similar data
collection efforts are recommended for other sites. If budgets do not permit data collection at this level,
soil characterization efforts could be reduced. In this study, shallow samples displayed more dynamic
effects.
7.4.2 Weather Data. Rainfall and soil moisture data are presented in Figure 7-1. A substantial rainfall
occurred immediately prior to biosolids application. In addition, several smaller rain events occurred
throughout the study. Consequently, the soil moisture remained close to water holding capacity
throughout the study and, thus, desiccation was not an issue in this study.
Figure 7-2 displays the soil temperature data from Day -3 (September 27, 2004) to the end of the
sampling period. In this graph, sample events are noted by arrows. The horizontal line marks 8°C, the
upper temperature limit for holding microbial cultures. Temperatures declined throughout the study and
were low during the last sample events. The lower temperature may have affected the observations for
this study. For example, degradation of APEs may have been retarded by colder temperatures.
Weather data collection was also useful in documenting site conditions. Based on experience at
this site, similar data collection efforts are recommended for other studies. If budgets permit, it would be
useful to collect soil moisture and temperature data for each replicate plot since plot-to-plot variations
were seen for some measurements in this study.
Rainfall
(mm)
45
40 -
Significant Rainfall Event
Sept. 2004
Jan.
Soil
Moisture
(mVm3)
Figure 7-1. Rainfall and Soil Moisture Measurements During Land Sampling Period
60
-------
Temperature (°C)
(D
O O ^ ^ N
O O O O O
M
T- CD T-
O O T-
Figure 7-2. Soil Temperature During Land Sampling Period. (The probe was placed 5 cm
beneath the soil surface. Arrows indicate sampling events. The horizontal line marks 8°C, the
upper temperature limit for storing microbial samples.)
7.4.3 Biosolids Distribution. The goal of measuring the quantity and distribution of biosolids after
application was to characterize biosolids application for this study. Two photographs are presented in
Figure 7-3. In A, the squares of geotextile were pinned to the soil prior to biosolids application. In B, a
post-application photograph is shown. Twenty squares were randomly positioned in each plot. Each
sample was analyzed for the wet, dry, and ash mass of biosolids on each 900 cm2 square of geotextile.
Analysis met all data quality objectives from the QAPP.
One-factor ANOVA models were fitted to three measurements (dry mass, wet mass, and ash
mass) to determine if significant differences existed among the three plots. These analyses showed that
Plot 2 received significantly higher levels than Plots 1 and 3 for all three measurements. A boxplot of the
ash mass data illustrates this observation in Figure 7-4.
To determine whether spatial variability was observed within plots, a separate ANOVA model
was fitted to the three measurements for each plot. The model included predictive factors for both the
row and column locations (see Figure 2-7). It was not possible to include interaction between these
factors because there were too few observations (less than 20) to estimate differences between 20
different factor levels (10 rows and 10 columns). For Plots 1 and 3, the three measured variables were not
affected by either the row or column locations. For Plot 2, both the row and column variables had
significant effects on all three responses. This analysis showed that biosolids application was even in
Plots 1 and 3 and uneven in Plot 2.
The statistical observations regarding spatial variability in Plot 2 were consistent with field
observations. During application, the spreader applied biosolids aerially, but a trail of biosolids was left
from one side of the spreader. This trail left a banding pattern across the application area as seen in
Figure 7-5. The spreader drove diagonally across Plot 2, and the trail of biosolids, visible in Figure 7-5,
may have contributed to the uneven distribution observed.
61
-------
•"??•.-
*'•*-*
M-MIlL
ftŁ***&m-%*\/*q**FFz-:,Ł. -•';'T-~'" tf"
Ł*SW»*i
-ii«C'
ra^
!.2>>^'"V*?
Figure 7-3. Photographs of Land Sampling Plots Before and After Biosolids Application:
A - Before Application (black squares are geotextile used to measure biosolids distribution); and
B - After Application (These photographs were from different plots.)
30
25
Ash
Mass
(g)
15
10
T
Plotl
Plot 2
Plot3
Figure 7-4. Distribution of Biosolids for Each Land Sampling Plot Based on Ash Mass (Data are
displayed using boxplots. Plots 1 and 3 were statistically similar, 10.8 + 4.4 g ash mass of biosolids/
900 cm2 geotextile, and showed an even distribution of biosolids across the plot. Plot 2 had a
statistically higher level of biosolids ash mass, 20.0 + 5.6 g ash mass of biosolids/ 900 cm2 geotextile,
and an uneven distribution of biosolids across the plot.)
62
-------
Figure 7-5. Photographs of Land Application of Biosolids on Plot 2 (In A, the banding pattern on
the field from the spreader crosses the photograph. In B, Plot 2 is shown on Day 14; the band of
biosolids runs below the yellow line.)
Using all the data from the geotextile squares, the application rate was 7.3 to 9.5 wet tons/acre or
1.7 to 2.2 dry tons/acre (95% confidence interval). This rate closely approximated the planned rate of 10
wet tons/acre.
Based on a review of the peer-reviewed literature, biosolids land application studies have not
characterized biosolids distribution. In general, dry weight application rates were reported. In a few
cases, large tarps were used to measure application rates. In this study, most samples were collected in
discrete soil cores within specific plots. In addition, statistical analysis identified plot-specific
observations for several analytes. Therefore, global estimates of biosolids application are not as
informative as the data produced by the technique used in this study. The technique used to measure the
amount and distribution of applied biosolids worked well for this study. Quantities were documented,
and data and statistical analysis supported qualitative observations about distribution. Based on the
experience at this site, similar data collection efforts are recommended for other studies.
7.4.4 PLFA. Phospholipid fatty acid (PLFA) analysis is a useful measurement to characterize the total
microbial community inhabiting a soil sample. This measurement involves extracting a soil sample and
isolating the phospholipids that comprise the cell membranes of microbes. Microbes produce different
fatty acid molecules in their cell membranes based on the type of microbe and in response to
environmental conditions. PLFA does not rely on culturing the cells; a snapshot of the community
present is obtained rather than an estimate based on culturable organisms. Total biomass is measured
using the quantity of lipid phosphate extracted from a sample, while community structure is characterized
by the relative abundance of individual phospholipid fatty acids in a sample (Vestal and White, 1989).
This technique is powerful for community level insight, but since pathogens comprise a relatively small
fraction of the microbial biomass, PLFA does not provide insight into pathogen levels.
For this study, the objective of PLFA biomass and community structure data collection was to
characterize microbial conditions during this study. PLFA biomass data were collected at three locations
and three depths from each plot during each sample event. ANOVA was used to evaluate the effect of
63
-------
plot, sample time, and sample depth as well as two-way interactions. The p-values of statistically
significant factors are shown in Table 7-3. Further analysis revealed that plot differences were significant
in surface samples, and, therefore, discussions of plot and time differences are focused in this horizon.
Table 7-3. Total Biomass ANOVA Results for Statistically
Significant Factors
Factor
Plot
Time
Depth
Time by depth
P-value
0.0062
O.001
O.001
0.0012
The total biomass in surface samples for each plot is shown in Figure 7-6 before and after
biosolids application. Initially, total biomass varied somewhat among the plots, but this variation was not
significant compared to the variation within each plot. Biosolids application increased total biomass by
60 nmol/g dry weight (gdw) on average. This change may have been due to added organisms as well as
growth of indigenous microbes. Unfortunately, due to variation within plots, the difference between pre-
and post-application biomass could not be statistically separated for the plot pairs.
The surface (0 to 5 cm) total biomass data are graphed as a function of time in Figure 7-7. The
mean total biomass for the samples at 10 to 15 and 20 to 25 cm depths are also shown. Statistical analysis
of these data revealed that total biomass data decreased with depth. The decrease in biomass as a function
of soil depth has been shown with many different soils (Vestal and White, 1989). The subsurface samples
were consistent among the three plots and did not change with time, and, therefore, data from all time
events and plots were used to compute the 10 to 15 cm mean and error bars depicted in Figure 7-7. The
20 to 25 cm graphical depiction was computed in a similar fashion. The surface samples demonstrated
time-dependent behavior. Following biosolids application, Day 14 values were 100% higher than the
pre-application levels. Unfortunately, sample-to-sample variation was substantial in these samples, often
around 30%. Thus, it is not possible to state that the total biomass changed in the post-application period.
Community structure data were evaluated using PCA and HCA. PCA representations are shown
in Figure 7-8A for the surficial samples (0 to 5 cm) and in Figures 7-8B and 7-8C for the 10 to 15 cm and
20 to 25 cm depths, respectively. The axes in each graph were determined by the statistical algorithms of
PCA and each dataset. In general, axes are a weighted linear combination of several measurements for
each sample. The distance between points indicates differences. The numbers next to each point in the
three graphs correspond to the sample event. Based on PCA, the surficial microbial community broke
into three groups: biosolids, samples from immediately after application (Days 0 and 14), and
pre-application samples clustered with Day 28 and later samples. Within the pre-application and more
than 28 days post-application cluster, Day -3 samples were shifted to one side. This shift was the result of
a rain event earlier that day. The rain caused a change in fatty acids related to cell dormancy (18: Iw7c,
16: Iw7c, cy!7:0, and cy!9:0). This break in dormancy, increased proportions of 18: Iw7c and 16: Iw7c
with decreased proportions of cy!7:0 and cy!9:0, could have been a function of decreased osmotic stress
and/or end of starvation/cellular growth (Kaur et al., 2005). Overall, the plot of surface PLFA data
suggested that the microbial community present on the soil surface was changed by biosolids application,
but the community returned to the pre-application state by the Day 28 sample event.
The PCA of the microbial community structure for the 10 to 15 cm depth showed a clustering of
the Day -3 samples (Figure 7-8B). The microbial fatty acids associated with dormancy had shifted as in
64
-------
200
150
Total
Biomass 100
(nmol/g dry solids)
50
AB
AB
AB
Plot 1
Plot 2
Plots
Plot 1
Plot 2
Plots
Pre-Application
Post-Application
Figure 7-6. Total Biomass for Surficial Samples, 0-5 cm, in Each Plot Before and After Biosolids
Application (The means are plotted, and the error bars are twice the standard error. The letter
below each box indicates the statistical grouping based on ANOVA. Symbols that share a letter are
statistically similar. For example, all pre-application samples share an A designation and are
similar. Pre-application Plot 2 and post-application Plot 1 are designated A and B, respectively,
and have statistically distinct levels of total biomass.)
the surface samples. As noted in the field observation book, on Day -3 (September 27, 2004), samples
were collected during a light rain/drizzle and the soil was evenly moist down to 25 cm. The rest of the
samples did not display any discernable pattern, and, therefore, the biosolid application did not appear to
affect the soil microbial population at this depth for the length of this study.
The PCA of the microbial community structure from the 20 to 25 cm depth showed no
discernable pattern for all soil samples taken (Figure 7-8C). At this depth, the microbial community
remained stable for the length of the study.
A review of the peer-reviewed literature identified one relevant study (Peacock et al., 2001). A field
was amended with manure and ammonium nitrate. The organic (manure) amendment only influenced the
microbial community structure in the 0 to 5 cm depth, similar to the observations in this report. However,
the literature study differed in that field treatments had been applied six times over 5 years, and the
conclusion was based on a single sampling event after treatment. The long-term manure applications
changed the soil structure and, therefore, changed the microbial community structure. The single
application evaluated in the reported study did not result in a long-term change in the microbial
community structure.
65
-------
200
Total
Bio mass
(nmol/g dry solids)
100
Al
c*
D *
-50
ABl
, All
10-15 cm
20-25 cm
Bl
1
1
HAB
J-
*.
ABll
ABll
•L
0 50 1C
Time (days)
Figure 7-7. Total Biomass as a Function of Time After Biosolids Application for
Surficial Samples, 0-5 cm (The means are plotted, and error bars show twice the standard error.
Samples at deeper depths, 10-15 and 20-25 cm, are also shown. Since statistical analysis showed no
significant temporal variation, graphed values were based on all data at that depth over the study.
The letter near each symbol indicates the statistical grouping based on ANOVA. Symbols that
share a letter are statistically similar. For example, the surficial data at Days -35 and -3 share an A
designation and are statistically similar. The Day 14 level is denoted B and is statistically different
from Days -35 and -3 data.)
Based on the experience of this study, PLFA measurements documented short-term changes in
the community structure during the sampling period. PLFA-based community structure assessments
displayed changes correlating to changes in soil conditions, such as the rainfall immediately prior to
biosolids application. In addition, changes were observed following biosolids application. The microbial
community returned to pre-application structure after 28 days (see Figure 7-8A).
PLFA-based biomass measurements document differences with soil depth (see Figure 7-7). Data
indicate there may be differences based on sample time and plot, but these differences were not
statistically significant. In future studies, more replicates of shallow samples are recommended. In
addition, it would be useful to collect PLFA-based biomass measurements for a longer period after
application. If budgets limit sample collection and analysis, sample collection should be focused on the
shallow soil horizon as this depth was the most dynamic.
66
-------
7-8A
-3 -3
Soil Depth
0-5 cm
Factors
Pre-application
and Day 28 and
later
actGiJ
Days 0 and 14
Biosolids
7-8B
Soil Depth
10-15 cm
"-35
.
Biosolids
67
-------
7-8C
Soil Depth
20-25 cm
°96 *
Biosolids
Figure 7-8. Community Structure Based on PLFA Profile: A. 0-5 cm Depth; B. 10-15 cm Depth;
and C. 20-25 cm Depth (Numbers next to each point correspond to sample event.)
7.4.5 Fecal Coliforms. Fecal coliform data were collected during this study as an indication of
pathogen bacterial behavior following land application of biosolids. The fecal coliform samples were
taken from the three different plots at seven different time points (two pre-application, one on the day of
application, and four post application). The samples were sent to Environmental Associates, Inc. for
analysis. Sample values were reported, but data quality information was not. Results are tabulated in
Appendix D (Table D-2). The laboratory reported semi-quantitative values for 48% of the samples
(values of less than or greater than a specific number). The QAPP did not anticipate this
semi-quantitative data in the statistical analysis plan. To facilitate statistical analysis of the data, censored
values were assigned the value of the boundary number for that sample. For example, at Day -36, one
sample density was reported as <0.22 MPN/gdw. This sample was assigned a value of 0.22 MPN/gdw
for statistical purposes. In another example, one of the Day 63 samples was observed to have a density of
>2.19 x 104 MPN/gdw. For statistical analysis, this sample was assigned a value of 2.19 x 104 MPN/gdw.
Figure 7-9 displays the data using this convention for all three plots. Censored values are noted as open
symbols, while quantitative data are shown as closed symbols.
Statistical analysis of these data involved several steps. The distribution of the fecal coliform
densities was extremely right-skewed, and, thus, natural logarithm transformation of the density was used
so that the ANOVA assumptions could be met. ANOVA indicates that time and time-by-plot interaction
were both statistically significant in the model, with p-values of less than 0.0001 and 0.0203,
respectively. Therefore, the temporal trend in densities was different for each plot. Because of the
significant interaction effect, multiple comparisons were performed within each plot to examine
68
-------
Fecal
Coliform
Concentration
(MPN/gdw)
10b
103
104
0
1
•'
1
•
•
0
a
+
•
1
1 .
i
>
g
•
•
*
n
•
•
+
±
*
*
t
• Plot 1
0 Plot 1 censored
• Plot 2
o Plot 2 censored
* Plots
O Plot 3 censored
101
10U
1(T
-40 -20 0 20 40 60 80 100
Time (days)
Figure 7-9. Fecal Coliform Concentrations as a Function of Time and Plot
differences in fecal coliform densities overtime. Scheffe's method was used to provide a set of
confidence intervals with joint 95% confidence. No differences in fecal coliform densities were present
between the two pre-application times for any plot. For all three plots, statistically significant differences
were identified between pre-application and post-application sample events, with higher estimated
densities for the post-application events. In the 28 days after application, differences were observed in the
plots. For Plots 1 and 2, the fecal coliform concentrations increased following application and remained
higher throughout the post-application period. For Plot 3, the fecal coliform concentration increased after
application, reached a maximum on Day 14, and remained elevated through Day 96. For all plots, the
densities are stable from Day 28 to the end of the study, but higher than the pre-application levels. The
soil temperature began dropping after Day 28, which may have contributed to this plateau in fecal
coliform densities.
Fecal coliform measurements exhibited a statistically significant increase following land
application of biosolids of more than 100-fold. Post-application levels were stable from Day 28 through
the end of the sample period at Day 98. Differences were observed between plots. Since data quality
information was not available and analysis included censored data, uncertainty in these conclusions was
substantial. Follow-on studies, including the use of replicate plots for fecal coliform sampling are
recommended. In addition, more sample replicates within a plot, more replicate plots, and more precise
data from the analytical laboratory may enable discrimination of temporal changes and those changes due
to differences in biosolids application. In addition, extending the sampling period following biosolids
application may be useful.
7.4.6 Alkylphenol Ethoxylates. The APE soil concentration was measured as a function of time to
better understand the persistence of these chemicals following biosolids land application. APE samples
were collected at three depths in each plot at each sample event. They were not observed in any sample.
69
-------
The concentration of all analytes was below the detection limit in all pre-application samples. Since the
field had no prior exposure to biosolids, pre-application concentrations of NP and OP were set at zero for
statistical purposes. The APE degradation products OP and NP were observed after biosolids application.
NPs were observed in shallow samples but not consistently at lower depths. OPs exhibited similar
behavior. Thus, the statistical analysis was limited to surficial data. After biosolids application,
concentrations reported as below detection limit were assigned the reporting limit for statistical purposes.
The NP data are graphed in Figure 7-10 as a function of time and plot.
10"
10J
NP
Concentration
(|_ig/kg dry solids)
10
1 1 1 1
i i i i
i i i i
i i i i
i i i i
i i i i
i i i i
: : : :
i i i i
• Plot 1
k. Plot 2
A Plot 3
_„. .j .j. j f _
i i i i
i i i i
m
I \
i i i i
1 1^1
m
^
1 •
L
i i i • i
: : : :
1 A 1 i i
^ 1 M
„ j. .^, |. j. i
* \ * \ 1 1
^
i i i i
i i i i
i i i i
i i i i
1 1 1 1
A
k^
•_
^
^
k
•
20
40
Time (days)
60
80
100
Figure 7-10. NP Concentrations as a function of Time and Plot in Surficial Samples, 0-5 cm, after
Biosolids Application (Concentrations of each replicate are graphed. Prior to biosolids application,
samples did not contain detectable levels of NP. For the purposes of statistical analysis, these pre-
application samples were set at zero.)
Data quality measures for APEs were within acceptance criteria with one exception, surrogate
recoveries. Surrogate recoveries were lower than acceptance criteria in 26% of the samples. However,
specific acceptance criteria were not established for this soil. Lower surrogate recoveries are consistent
70
-------
with higher clay contents, and this soil had significant clay content. Thus, reported concentrations may be
biased low but were useful for understanding NP/OP behavior in this setting.
A MANOVA model was used to examine the effects of the two factors (plot and date) on NP and
OP concentrations. This analysis revealed that the date-by-plot interaction was not statistically
significant, so it was dropped from the model, leaving only the main effects of date and plot. Wilks'
Lambda test indicated that date was a significant predictor of the joint NP/OP responses (p-value =
0.0032), but plot effect was not found to be statistically significant, with a p-value of 0.5248. However,
even though the multivariate Wilks' Lambda test indicated that the date was statistically significant in
jointly predicting NP/OP concentrations, multiple comparison could not discriminate differences among
the seven dates for NP and OP concentrations. In other words, the statistical analysis did not identify any
changes in concentration.
Other researchers have also considered the fate of NP in biosolids-soil systems. When anaerobic
biosolids are combined with soils to form homogeneous mixtures, removal can be quick: 80 to 2.4 mg/kg
dry soil in 30 days (97% removal, Hseu, 2006), and 0.246 to 0.064 mg/kg dry soil in 30 days (74 %
removal, Mortensen and Kure, 2003). However, two studies report slower removal: NP mineralization of
58 % in 38 days (Hesselsoe etal, 2001), and 10 % in 150 days (Dettenmaier and Doucette, 2007). Two
field studies have shown very disparate results. In one, anaerobically digested sludge was applied as a
liquid to a grass field at 6 dry tons/acre and monitored for 320 days (Marcomini et al., 1989). Initial
removal rates were high; concentrations decreased from 4.7 mg/kg dry soil to about 1 mg/kg in 3 weeks.
At that point, the rate slowed, and a plateau concentration persisted at 0.5 mg/kg dry soil (89% removal)
through the remainder of the study. When anaerobically digested biosolids were applied as a wet cake at
8 dry tons/acre, NP was not detected 31 days after application, but after 156 days, NP was measured at
3.6 + 0.8 mg/kg dry soil (mean + standard deviation, Kinney et al., 2008). The results for the second field
study are similar to the results for this study, persistent and variable observations of NP following land
application of biosolids.
In a series of studies, Hesselsoe et al. (2001) observed that NP degradation was fastest in
homogeneous aerobic mixtures of soil and biosolids and degradation was retarded in non-homogeneous
mixtures containing biosolids aggregates. NP mineralization in 4-cm aggregates was 51% of the
mineralization observed in 2-cm aggregates after 119 days. Oxygen penetration into these aggregates is
slow, and the aerobic volume of an aggregate correlated to NP mineralization. These observations were
consistent with available field data. Residual concentrations were lower for liquid application where
aggregates should be small (Marcomini etal., 1989) than for cake application (Kinney etal., 2008).
Biosolids particle size was not measured in this study, but as shown in Figure 7-3B, the biosolids did not
form a fine granular material when applied. Rather, they formed clumps of varying size and thickness.
This physical distribution and the high moisture content would limit oxygen penetration into this material
and, thus, NP removal.
Nine replicates at each sample point were not sufficient to statistically distinguish concentration
changes with time after application; however, this dataset can be used to evaluate the effect of additional
samples on statistical power (1 - P). Van Belle and Martin (1993) developed expressions relating
statistical power to sample replication, variability, and difference between samples. Using the equations
for log normally distributed data, ZI_P was calculated for sample sizes ranging from 3 to 54 replicates
using the summary data from the five sample events. In this calculation, relative standard deviation (a/jo,,
RSD) was assumed to be equal for both samples and the ratio of means (m/^2 or f) was set at 1.25. An
Excel spreadsheet was used to compute 1 - (3 at a = 0.05. Results from these calculations for the lowest
and highest RSD observed in this study are presented in Figure 7-11. Power increases with sample
replication. For the case with the lowest RSD, power does not exceed 60% with more than 50 samples
71
-------
Statistical
Power
1.0
0.8
0.6
0.4
0.2
0.0
m
I
I
0
63
0
10 20 30
40
50
60
Sample Number
Figure 7-11. Statistical Power for NP Samples as a Function of Sample Number
for Day 0 and Day 63 (These sample events had the highest and lowest RSD,
respectively. The ratio of means, f, was set at 1.25, and the relative
standard deviation was equal. Lognormal calculations were used.)
while the higher RSD barely exceeds 20%. This level of replication would be too expensive for many
research projects, and these power levels are too low. This analysis suggests that a different sampling
approach is needed. Sampling options to consider include larger samples with homogenization, sample or
extract composites, or normalization with a biosolids marker. Field studies are needed to select an
acceptable method.
APEs were not observed during the sampling period. Their metabolites, OP and NP, were
observed consistently in post-application shallow samples but not in deeper samples. Since no plot
effects were observed, conclusions were based on nine replicates. Concentrations increased following
application, but variability was too high to identify temporal changes. Based on the experience in this
study, a different approach to sampling may be useful in characterizing the persistence of compounds
such as APEs. A longer sample period is also recommended. In addition, future data collection to
evaluate the persistence of endocrine disrupting compounds (EDCs) should include a broader spectrum of
chemicals.
72
-------
7.4.7 Ecotoxicity Screening. Soil toxicity was screened to evaluate whether biosolids application
enhanced or degraded the soil environment. Several assay characteristics were considered in the process
of selecting bioassays including test organism, exposure period, and assay endpoint. In general, it is
preferable for assays to cover a range of species, exposure periods, and assay endpoints. For example, the
assays selected for this study included ecologically relevant plants and animals as test organisms. Also,
assays may address a variety of endpoints such as mortality, mutagenicity, growth, or endocrine system
disruption. For the assays selected, the endpoints included mortality and growth.
For this study, 14-day earthworm mortality and 5-day seed germination and root elongation
assays were used to screen for soil toxicity or enhancement. These assays were generally selected based
on their relevance to soil toxicity (Chang et al., 1997; Dorn et a/., 1998; Environment Canada, 1994;
Hund and Traunsurger, 1994; Meier et al, 1997; Salintro et al, 1997; Simini et al, 1995; EPA, 1988,
1989). The list was limited and would be strengthened by the addition of chronic sublethal animal assays
such as earthworm growth and reproduction, earthworm avoidance, or longer exposure plant assays.
However, the current assay list provides some toxicological information; testing costs and development
time prevented using a larger collection of assays.
Soil samples for ecotoxicity measurements were collected before (Day -35) and after land
application (Days 0 and 98) and were transported to the laboratory for testing. In the laboratory, sample
soil was mixed with artificial soil to produce five different concentrations of test soil. Then, organisms
were placed in the soil mixtures and incubated for 5 days (plants) or 14 days (earthworms). The response
to the concentrations of the test soils was measured.
The toxicity tests for these samples followed the QAPP with one exception, oat seed age. The
seeds used in the oat seed germination and root elongation assays were taken from a shipment of oat seeds
received in 2002. This seed holding time exceeds QAPP specifications. However, since subsequent
shipments have not met QAPP accuracy requirements for root elongation in controls and the 2002 seeds
continue to meet these requirements, the 2002 seeds were used.
Prior to performing ecotoxicity tests, various soil parameters, including soil pH, were evaluated
so that the tests could be conducted in a consistent fashion. For all but one soil sample from this study,
pH was low, ranging from 4.7 to 5.7. The QAPP specified that if soil pH was below 6.0, calcium
carbonate (CaCO3) would be added to raise the pH to at least 6.5. Preliminary testing showed that 1-mass
percent CaCO3 raised the pH to an acceptable level. Thus, all samples were amended with 1-mass
percent CaCO3to avoid any question about inconsistent dilution. One sample was tested to determine if
the CaCO3 amendment affected the results; no difference was observed.
Ecotoxicity assays met data quality objectives with one exception. The reference toxicant controls
for the Day 98 lettuce test did not display the required root reduction. This test was repeated, and the
reference toxicant controls met data quality objectives. The retest data were used for statistical analysis.
No earthworm mortality was observed in any sample or for any soil concentration. No consistent
changes in seed germination effects were observed. Thus, these assays were not sensitive to any changes
associated with land application of biosolids at this site and under these conditions.
Root elongation data for lettuce and oats demonstrated changes. The data consisted of three
replicates from each of three plots at three time points and five different test soil concentrations. Data are
graphed in Figure 7-12. In the presence of a toxic soil, roots are usually longer in the 0% sample soil and
decrease as sample soil concentration increases. In this study, the root length data for both lettuce and
oats did not display the typical pattern as a function of soil concentration. The 0 and 100% soil
concentrations showed longer roots than the intermediate concentrations. The cause of this unusual
73
-------
A
40
35
30 ]
Mean j
Lettuce 20
Root
Length 15
(mm)
10
5
n
1
>
•
I
id
•
•
*
1
*
•
I
• Day -36
• DayO
* Day 96
i
|
(
i
»
»
0 20 40 60 80 100
Soil Concentration (%, sample soil to total soil)
B 100 r
250
Mean 20°
Oat
Root 150
Length
(mm) 100
50
0
f
A
1
•
•
»
i
•
•
..a
•
•
• Day -36
• DayO
>
'
1
;
1
0 20 40 60 80 100
Soil Concentration (%, sample soil to total soil)
Figure 7-12. Root Elongation as a Function of Test Soil Concentration and Sample Event:
A - Lettuce Data; and B - Oat Data
74
-------
behavior is not known. The characteristics of the sample soil and artificial soil were quite different. It
was possible that varying soil texture and nutrient levels may have played a role in these observations.
This unusual pattern was observed in all samples, before and after biosolids application.
Data were evaluated using a three-way ANOVA for the two species. Results are shown in
Table 7-4. Root lengths did not differ between plots. Differences were observed in response to sample
time and sample concentration for both lettuce and oat roots. For lettuce, the time-by-concentration
interaction was also statistically significant, which indicates that the effect of sample time on root lengths
differed for the five sample concentrations. Statistical analysis of the concentration dependent data was
used to maximize knowledge gained from the study. However, the focus of this data collection effort was
to understand the effect of biosolids application on soil organisms, and, thus, the time dependent changes
are of greater interest.
Table 7-4. Three-Way ANOVA Results for Root Elongation
Effect
Plot
Time
Concentration
Time -by-Concentration
Lettuce p-value
0.8825
<0001
<0001
0.0121
Oats p-value
0.7237
0.0004
<0001
For lettuce, root lengths increased over the course of the study. In Figure 7-12A, the Day 98
samples appear consistently longer than for the other days. Statistical analysis indicates that each time
point was distinct, as illustrated in Figure 7-13, and root lengths for lettuce increased with time, both from
pre-application day to the application day and from the application day to the final sample. This
observation supports the use of biosolids as a beneficial soil amendment.
The observations for oat roots were different in that the roots decreased in length throughout the
course of the study. The root lengths for the final sample event appear to be generally lower than for the
other events in Figure 7-12B. The statistical analysis confirmed this observation. Oat root lengths
decreased with collection time. The decrease was statistically significant between the pre-application and
application days, but there was no statistically significant change between the application day and the
final sample. This dataset does not support the use of biosolids as a beneficial soil amendment.
Ecotoxicity following use of biosolids as a soil amendment was evaluated by Banks et al. (2006).
They evaluated anaerobically digested biosolids from 19 locations where biosolids are typically land
applied. Samples of biosolids and soils to which biosolids had been applied were tested. Earthworm tests
included a 14-day biomass assay and a 7-week assay measuring biomass, numbers of juvenile hatchlings,
and cocoon numbers and hatchability. Plant assays included a 5-day seed germination assay using
lettuce, radish, and millet and a 10-day root elongation test using lettuce. The paper describes slightly
different 5-day seed germination assay methods than were used in this study. In Phase I, biosolids were
added to soil at one concentration. Lettuce germination was rarely affected, but earthworm biomass
accumulation was reduced in six of the 19 biosolids samples. Five samples from Phase I were evaluated
further in Phase II. Lettuce germination was significantly reduced in one sample, possibly due to elevated
salinity and low pH in that sample. Radish and millet exhibited similar differences. Lettuce root lengths
were shorter in biosolids amended soils, but the difference was not statistically significant. Earthworm
biomass gain was lower than controls in two samples and similar to controls in three samples. Both
75
-------
Root
Length
(mm)
<1JU
200
150
100
50
0
-4
*A
• c
1
1
1 B
> D
• Lettuce
• Oat
B*
E •
0 -20 0 20 40 60 80 100
Time (day)
Figure 7-13. Average Root Length for all Soil Concentrations as a Function
of Treatment Time for Lettuce and Oats (The 95% confidence limits are
indicated by the error bars. The letter near each symbol indicates the statistical
grouping based on ANOVA. Symbols that share a letter are statistically similar.)
samples with lower biomass gain displayed higher ammonia levels; one sample also had elevated saline
and low pH. In summary, Banks et al. (2006) used a more varied selection of bioassays to evaluate the
ecotoxicity of biosolids application to land. Based on six bioassays measuring 10 endpoints, 19 of 110
toxicity tests showed a negative effect, but no consistent trends were observed in biosolids applied in a
manner consistent with 503 Regulations.
Therefore, based on the literature, screening ecotoxicity results did not exhibit a consistent
pattern. In this study, no change was observed in earthworm mortality and seed germination data
following biosolids application. Since both assays displayed a maximal response prior to application, no
added benefit of biosolids could be demonstrated. Neither assay showed a negative response to biosolids.
Lettuce root length displayed enhanced growth following biosolids application in this study.
However, oat root length showed reduced growth. Based on this information, continued toxicity
screening is recommended for future studies. In addition to the current assays, chronic earthworm tests
with non-lethal endpoints, such as biomass accumulation, and longer-term plant studies would be useful
in gaining a fuller understanding of the effect of biosolids. If a particular organism is of interest, assays
directed toward that organism are recommended. A longer sampling period may be useful.
7.4.8 Other Microbial Measurements. The microbial community was evaluated using several
techniques for this study. PLF A/FAME measurements characterized the size and diversity of the
76
-------
community. Fecal coliforms were measured as a potential indicator of pathogenic bacteria. In addition,
other microbial indicators were measured including VHO, Salmonella, enteric viruses, coliphage, total
heterotrophs, fecal coliforms, and Enterococcus spp. Due to the number of tests and costs of analysis,
this sampling was conducted with less replication than the replication used for the fecal coliform analysis
discussed earlier. One composite sample was collected from each plot during each sample event (see
Table 7-1).
Most samples were sent to Environmental Associates, Inc. for analysis. Sample results were
reported, but data quality information was not. Results from this data analysis are shown in Table 7-5.
With the exception of biosolids samples, rare detections were observed for the enteric viruses,
Salmonella, and VHO following biosolids application. Somatic coliphage detections were more common.
Table 7-5. Other Microbial Indicators from Land Sampling
Analyses Performed by Environmental Associates, Inc.
Sample
Event
Day -35
Day -3
DayO
Day 28
Day 63
Day 98
Day 1
Date
Sampled
8/25/2004
9/27/2004
9/30/2004
10/28/2004
12/7/2004
1/4/2005
10/1/2004
Plot
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Biosolids
Virus(a)
(MPN/4g)
0.82
O.84
0.80
O.80
0.67
O.67
0.67
1.37
0.68
O.67
0.67
O.68
0.68
O.68
0.68
O.68
0.68
O.68
0.68
2.94
0.70
Salmonella
(MPN/4g)
0.33
O.32
0.32
O.32
0.32
O.32
0.36
O.35
O.37
6.06
1.59
O.36
0.694
20.85
1.41
O.38
1.22
O.43
>325
>325
>325
VHO
(No.
viable/4g)
0.56
O.47
0.48
O.57
0.53
O.60
0.47
O.56
0.68
O.44
0.48
O.48
0.32
0.54
0.52
1.14
1.13
O.67
1.78
2.46
0.77
Coli
Male-
specific
(PFU)
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
199
133
75
phage
Somatic
(PFU)
<1
<1
<1
3
16
7
1
2
4
1
<1
<1
1
<1
1
1
1
<1
557
616
448
(a) All samples <1 PFU/4g
In this study, fecal coliforms were measured as indicators of pathogenic bacteria such as
Salmonella. Salmonella were detected in a total of six soil samples during the study period between Days
28 and 98. Increased Salmonella levels in soil samples during this period may have been due to additions
to the soil community from the biosolids or growth of indigenous Salmonella on biosolids substrates (Unc
et al., 2006; Winfield and Groisman, 2003). Fecal coliforms levels, which were higher than pre-
application levels and stable during this period of the study, were about three orders of magnitude higher
than the Salmonella levels observed. Both microbial analytes were temporally steady during this period.
77
-------
Thus, fecal coliforms were more easily measured than Salmonella, and showed a pattern of behavior
consistent with the available Salmonella data. In this study, fecal coliforms were indicators of Salmonella
behavior during static periods; no data were available to evaluate whether fecal coliforms were indicators
of Salmonella behavior during dynamic conditions.
The Sustainable Agricultural Systems Laboratory, USDA, analyzed samples at three time events:
Day -35, Day 28, and Day 98. Data were not reported with information concerning holding times,
positive or negative controls, or other measures of data quality. Therefore, these data are of unknown
quality. The results from the USDA data analysis are shown in Table 7-6.
Table 7-6. Other Microbial Indicators Measured by the USDA Laboratory
Sample
Total Heterotrophs
Fecal Coliforms
Enterococci
Biosolids Samples
Replicate 1
Replicate 2
Replicate 3
6.61 x 1010
6.43 x 1010
1.51 x 1011
1.49 x 1011
4.95 x 1010
4.32 x 1010
1.04 x 109
1.08 x 109
4.87 x 109
5.57 x 109
5.49 x 108
6.30 x 108
2.75 x 105
8.26 x 105
9.54 x 105
2.07 x 106
1.80 x 105
6.30 x 105
Day 28 Samples
Plot 1, Replicate 1
Plot 1, Replicate 2
Plot 2, Replicate 1
Plot 2, Replicate 2
Plot 3, Replicate 1
Plot 3, Replicate 2
1.66 x 108
1.58 x 108
1.22 x 108
1.35 x 108
3.63 x 106
1.80 x 107
1.79 x 106
1.48 x 106
8.92 x 105
1.00 x 106
8.85 x 104
1.06 x 105
2.60 x 104
2.85 x 104
2.88 x 104
3.30 x 104
1.14x 104
1.28 x 104
Day 98 Samples
Plot 1, Replicate 1
Plot 1, Replicate 2
Plot 2, Replicate 1
Plot 2, Replicate 2
Plot 3, Replicate 1
Plot 3, Replicate 2
7.65 x 107
7.42 x 107
6.20 x 107
8.97 x 107
2.57 x 107
1.95 x 107
5.43 x 105
6.87 x 105
2.48 x 105
2.17 x 105
2.27 x 105
1.56 x 105
1.80 x 104
1.29 x 104
5.45 x 103
6.00 x 103
1.76 x 104
1.90 x 104
Units on all measurements
thus, were not included. S.
detection limit.
are CFU/g dry solids. Data for Day -35 samples were not
aureus and Salmonella were measured by spread plating,
reported on a dry basis and,
but all values were below the
While both labs measured fecal coliforms and Salmonella, it is difficult to compare results.
Salmonella detections were very infrequent; there are little data to compare. Fecal coliform samples are
processed in different ways in the two laboratories, and results (Figure 7-9 and Table 7-6) are reported in
different units. Fecal coliforms results from Day 28 and 98 for both labs were shown in Table 7-7. In
general, the MPN results were two orders of magnitude lower than the CPU results. However, the change
between days within a measurement technique was minimal. This observation was consistent with
statistical analysis of the Environmental Associates, Inc. data, i.e., no significant changes were observed
between Day 28 and the final sample event at Day 98.
78
-------
Table 7-7. Fecal Coliform Results from Environmental Associates, Inc. and USDA (Data were
presented as a function of plot and time because statistical analysis of Environmental Associates,
Inc. data identified statistically significant time-by-plot interactions. Data quality for these analyses
were unknown as neither laboratory supplied this information.)
Plot
Results from EAI
(MPN/g dry wt)
Replicate 1
Replicate 2
Replicate 3
Results from SASL, USDA
(CFU/g dry solid)
Replicate 1
Replicate 2
Day 28
I
2
3
7.00 x 103
>2.13 x 104
2.28 x 103
6.02 x 103
> 1.96 x 104
1.25 x 103
1.23 x 104
>2.05 x 104
1.21 x 104
1.79 x 106
8.92 x 105
8.85 x 104
1.48 x 106
l.OOx 106
1.06 x 105
Day 98
I
2
3
> 1.95 x 104
2.38 x 104
2.26 x 104
2.25 x 103
>2.09x 104
1.95 x 103
1.88 x 103
>2.18x 104
2.21 x 104
5.43 x 105
2.48 x 105
2.27 x 105
6.87 x 105
2.17 x 105
1.56 x 105
The biosolids used in this study contained many microbial indicators, but detections in soil
samples were rare. Since other measurements displayed high variability, future studies should evaluate
whether the microbial indicator sample strategy yields representative samples. Otherwise, this sample
approach is appropriate and indicates that other microbial analytes were rarely observed following land
application of biosolids at this site. A similar approach may be useful in future studies.
7.5 Conclusions
A field-scale research project was conducted in 2004-2005 to evaluate land application of
anaerobically digested biosolids at agronomic levels. For this study, biosolids were applied to a fescue
field in a 100-m diameter circle by a side discharge manure spreader. Biosolids had not been applied to
this land previously. Soil samples were collected from three replicate plots within the application area
prior to biosolids application and for 98 days following application. Study conditions were characterized
by measuring biosolids mass and distribution after application, PLFA, and agronomic and weather data.
Fecal coliform, APEs, total biomass, community structure, and ecotoxicity data were used to evaluate the
effects of biosolids land application. Limited data for specific pathogens were collected. Salmonella and
VHO were observed in biosolids samples but rarely in soil samples.
The measurements used to characterize the study were generally informative. Biosolids
distribution for each replicate plot was determined by measuring the dry mass and ash mass of applied
material in each plot. Statistical analysis revealed plot-to-plot variations in the amount and distribution of
biosolids. This observation supports a recommendation for using replicate plots within a study. In
addition, large-scale measures of biosolids application, such as loading rate, may not reflect the biosolids
content of discrete samples. Biosolids distribution data documented the wet and dry mass applied per
acre. Sampling techniques that enable measurement of biosolids loading and analytes of interest in a
single sample may be useful in reducing uncertainties in these types of studies. Soil agronomic data
varied primarily with depth and did not exhibit consistent changes following biosolids application.
Weather data displayed fairly moist conditions and falling temperatures following application. The soil
and weather data collected may be useful when placing this study in context with other similar studies in
order to draw general conclusions about land application of biosolids.
79
-------
Measurements used to evaluate the effect of biosolids application often exhibited changes.
However, data interpretation was complicated by semi-quantitative data, sample-to-sample variation, and
inconsistent results. PLFA measurements were used to characterize the size and diversity of the
microbial community. Total biomass based on PLFA varied from plot to plot, with depth, and with time
for surficial samples. Total biomass in shallow samples increased following application, but changes
with time after application were not statistically significant. Based on PLFA distribution, the microbial
community was shifted following biosolids application, but returned to pre-application structure within 28
days. Fecal coliforms were measured as indicators of pathogenic bacteria. The laboratory reported semi-
quantitative fecal coliform data. Statistical analysis identified plot-to-plot variations. An increase in fecal
coliform density of more than 100-fold was observed after application that generally persisted for the
duration of the study. Therefore, results from total biomass amount and fecal coliform density were
consistent, i.e., increased levels following biosolids application and remaining throughout the sampling
period. In addition, both exhibited plot-to-plot variations. PLFA community structure results were
different, showing a transient change after biosolids application. The transient period in community
structure results corresponded to the time period when fecal coliforms and total biomass levels increased.
When fecal coliform and total biomass data were stable, before application and from 28 to 98 days after
application, community structures were similar. Two characteristics of these measurements should be
considered in comparing this information. Both total biomass and fecal coliform data reflect absolute
amounts of analyte, while community structure is determined by the relative amounts of PLFAs. In
addition, community structure analyses incorporated much more data and, as a result, may be more
sensitive to changes in the microbial population than fecal coliform and total biomass measurements.
APEs, including degradation products such as OP and NP, were of interest as potential EDCs and
persistent chemicals. Samples were collected at three soil depths. The degradation products were
observed in shallow samples following application. Statistical analysis could not discriminate temporal
changes in concentration after application. Since these compounds are aerobically degradable in soil
biosolids mixtures, the persistence of the chemicals may reflect exposure to anaerobic environments
within biosolids particles or microbial preferences for other substrates in biosolids.
Soil toxicity was screened using the 14-day earthworm mortality and 5-day seed germination and
root elongation in lettuce and oat bioassays. Soil samples were screened for toxicity prior to, immediately
after, and 98 days after biosolids application. These results did not demonstrate a consistent pattern.
Earthworm mortality and seed germination were neutral, with no enhancement or reduction following
application. Lettuce and oat root elongation showed enhanced and reduced growth, respectively.
Based on all of the measures of biosolids effects, it appeared that soil samples changed following
biosolids application. In many cases, these changes persisted throughout the study. In the future, longer
sampling periods and improved sampling procedures are recommended to refine observations. In light of
the biosolids distribution information, sampling procedures that express data based on the biosolids
present in a discrete sample may facilitate data interpretation.
7.6 Discussion and Lessons Learned
Several aspects of the experimental design were useful in describing the conditions throughout
the study and evaluating the effects of biosolids application to land. For example, the general
experimental design using replicate plots with subsampling of each plot with time was practical.
Replicate plots differed based on several analytes. The use of replicate plots facilitated separating plot
effects from time or other variables in statistical analysis. Replicate plots are recommended in future
studies.
80
-------
The measurements included in this study were selected to characterize the experiment and
evaluate questions regarding biosolids application to land. Within the characterization measurements,
biosolids distribution and weather measurements were novel and useful. In particular, measurement of
biosolids distribution documented mass applied per acre. PLFA measurements provided characterization
data as well as documenting changes with time. Measurements to evaluate biosolids land application
included indicators of pathogens, the microbial community, chemicals of emerging interest, and assays to
evaluate ecological effects. This range of variables is appropriate to evaluate use of biosolids as a soil
amendment. Future studies should include this range of variables and, if possible, expand the
measurements within each category. For example, chemicals of emerging interest in future studies could
include steroid hormones, fluorotelomer residuals, pharmaceuticals, personal care products, pesticides,
brominated diphenyl ethers, atrazine, and vinclozolin. Ecotoxicity testing could also be expanded to
include earthworm biomass, longer-term plant assays, and a broader range of test species. Other
measures of microbial indicators, such as Salmonella or coliphage, could be included to better evaluate
residual concentrations following biosolid application. However, as with this study, the availability and
cost of analytical methods for these measurements, particularly in a biosolids/soil mixture, may limit
inclusion in future studies.
The sampling plan for this study evaluated several variables with time and depth. Changes with
time were a question of primary interest. In the cases of total biomass, fecal coliforms, and OP and NP,
the concentrations before and after biosolids application were different. Unfortunately, the variability in
measurements was high relative to the concentration changes after application. Two recommendations
for future studies may improve the ability to draw statistically-based conclusions: 1) evaluate biosolids
application over a longer period, and 2) use a different sampling strategy to reduce variability. This study
also evaluated whether analytes were transported through the soil. At this site, very little downward
migration was observed; however, the site soil contained a high fraction of fine particles (clay) that tend
to slow downward migration. Based on these observations, future studies may focus sampling on shallow
samples and include fewer samples with depth. This study did not consider whether surface runoff would
distribute analytes across the study area or beyond. Future studies may want to incorporate sampling to
evaluate the runoff.
Data quality for most measurements was acceptable. Study plans included corrective action for
most instances when data quality samples were outside of acceptance levels. Two situations were not
considered: 1) semi-quantitative fecal coliform data, and 2) low surrogate recoveries for APEs. The semi-
quantitative fecal coliform data limited the ability to draw conclusions on residual concentrations. This
variable was of critical interest in this study, and, thus, the data quality gap is a concern. It is
recommended that future studies carefully evaluate analytical laboratories to assess whether they can
meet study requirements. The APE data had a more minor problem. Surrogate recoveries were low in
26% of samples, possibly due to the high clay content of the soil. Analysis of preliminary samples may
have identified the need for site-specific data acceptance criteria for these analytes. Analysis of
preliminary samples is recommended for future studies, particularly if the analyte list for chemicals of
emerging interest is expanded.
81
-------
8.0 SUMMARY AND CONCLUSIONS
8.1 Introduction
This report documents the approach, results, and interpretation of a collaborative research study
conducted by EPA's National Risk Management Research Laboratory, USDA, NCDA&CS, Battelle, and
other organizations to evaluate the land application of Class B biosolids The overall goal of this research
was to investigate air, volatile emissions, and soil sampling methods and analytical techniques. To
accomplish this goal, samples were collected using a variety of sampling methods and equipment and
analyzed for a broad matrix of chemical, physical, and microbial species. It is anticipated this study along
with other research will eventually lead to the development of a standardized protocol that can be used in
future studies on the application of biosolids to land.
This research was conducted under EPA Quality Assurance Project Plan (QAPP) No.l63-Q10-2,
and it represents the first known comprehensive study evaluating this variety of sampling methods and
analytical techniques simultaneously in the field before, during, and after the land application of
biosolids. The study commenced in August 2004 at the NCDA&CS Piedmont Research Station in
Salisbury, NC. Biosolids application was conducted in September 2004, and field monitoring continued
until January 2, 2005.
8.2 Study Description
Land application occurred on a fescue field with no previous exposure to biosolids. Other than
modifications to facilitate sampling, application practices and equipment were typical of those used
during normal agronomic biosolids application. Biosolids were land applied at a target rate of 10 wet
tons/acre.
Measurements were made of air emissions (volatile odorants and microorganisms) and their
short-range transport; airborne particulates; and soil microbial and chemical concentrations at and around
the test site before, during, and after biosolids application. To achieve the overarching goal of the project,
the research was implemented via three tasks, each with its own discrete goals and sets of hypotheses:
Task 1. Bioaerosol and Paniculate Matter Sampling. Select bacteria, fungi, viruses, bacterial
endotoxins, and particulates were sampled in the aerosol emissions from biosolids prior to, during, and
after biosolids application. The primary objectives for Task 1 were to: 1) characterize the types and
measure the concentrations of the suite of viable bioaerosol components (seven bacteria, enteroviruses,
and male-specific coliphage) and particulates; 2) determine if these bioaerosol components were emitted
and transported to several points downwind of the biosolids application area under the circumstances
investigated; and 3) evaluate the collection performance of the six-stage impactors, biosamplers, and
GRIMM sampler in this field application study.
Task 2. Volatile Organic Carbon and Odorant Monitoring and Analysis. The presence and
concentration of a selected group of inorganic and organic compounds and odorants were measured in
emissions generated upwind, within, and downwind of the application area of the biosolids land
application test site. The objectives of this task were to: 1) quantify the concentrations of specific
compounds identified in the emissions including VOCs and odorants; and 2) determine the transport (if
any) of these chemicals downwind of the biosolids application area at this site.
Task 3. Land Sampling. The soil sampling component of this research involved a longer
sampling period than did other tasks and focused on measuring the concentrations of microbes and
82
-------
chemicals before and for 4 months after biosolids application. The specific objectives for Task 3 were to:
1) characterize the quantity and distribution of biosolids applied, 2) characterize the microbial community
quantity and structure in the soil, 3) measure the fecal coliform concentrations in the soil, 4) measure the
concentration of alkylphenols and alkylphenol ethoxylates (APEs) in the soil, and 5) screen soil samples
for terrestrial ecotoxicity.
In addition to the work described here, two additional studies were carried out during the
biosolids land application and one study was completed in the spring of 2005 following biosolids
application at the site. These investigations were not part of the project-specific quality assurance plan
and, therefore, are not reported in the body of this report. The work, however, is related to this study, and
the results are provided in the report appendices as follows: Appendix A. Determination of Total
Bacterial Bioburden from Impinger Samples Collected During the NC Biosolids Land Application Study -
Dr. Mark Hernandez, University of Colorado; Appendix B. Parallel Sampling Approaches and Analysis
of Impinger Samples Collected During the NC Biosolids Land Application Study - Dr. Ian Pepper,
University of Arizona; and Appendix C. Endotoxin Sampling During a Post-Spring Cutting Event at the
NC Biosolids Land Application Study Site - Dr. Edwin Earth, EPA/NRMRL.
8.3 Study Site
The test site, shown previously in Figure 2.2, consisted of a 100-m diameter circle application
area (approximately 2 acres) located within the selected fescue field. It was designed to accommodate an
array of upwind and downwind bioiaerosol sampling units that could be moved around a circular center
point if needed due to changes in wind direction during application. A total of nine stationary sampling
locations were positioned on each of three parallel sampling lines (three per line), one upwind and two
downwind. In addition to the stationary sampling locations, one mobile sampler (MOB) consisting of an
ATV equipped with a similar sampling equipment array was deployed to follow within 15m of the
application plume behind the hopper.
8.4 Applied Biosolids
The Class B biosolids applied during the study consisted of anaerobically digested and
mechanically dewatered (centrifugation combined with polymer addition) municipal wastewater
treatment plant sludge. At the request of the researchers, this material was pretreated with only enough
lime to adjust the pH and suppress microbial growth to meet facility compliance for release (material met
Class B compliance at time of release from facility). Heavier doses of lime are more consistent with the
normal operation of the facility. Therefore, this material was atypical of biosolids that would normally be
released from this facility. As a result of this modified sludge treatment regimen, researchers hoped that
viable microbial populations would be available in the aerosol phase for collection and analysis once the
study was implemented. This specially prepared/treated sewage sludge was also expected to elicit odors
and generate particulates via surface drying, flaking, and wind erosion during land application. The
parameters measured in the biosolids were consistent with those used throughout the investigation for
comparison purposes.
8.5 Field Results
In addition to characterizing the applied biosolids, researchers simultaneously conducted
bioaerosol sampling, volatile organic and inorganic compounds collection and analysis, odorant
monitoring and sampling, and land sampling activities within and near the application test area prior to,
during, and following land application. A comprehensive description of all sample collection procedures
and analytical methodologies can be found in Section 2.0.
83
-------
The measurements conducted and the analytical data generated on this project are discussed in
detail in Section 4.0 for biosolids characterization, Section 5.0 for bioaerosol and particulate sampling,
Section 6.0 for volatile organic and inorganic emissions and odorant sampling, and Section 7.0 for land
sampling. Key results and conclusions are summarized below in Sections 8.5.1, 8.5.2, 8.5.3, and 8.5.4,
respectively.
8.5.1 Biosolids Characterization. The biosolids used in this study had a solids content of 28% and
contained 2.3 x 109 CPU fecal coliforms/g dry weight (gdw) total solids and 6.33 most probable number
(MPN) Salmonella spp./gdw total solids. Microbial measurements included total heterotrophic bacteria
(THE) at 1.6 x 1011 CFU/gdw total solids, Escherichia coli at 4.35 x 1Q7 MPN/gdw total solids, total
coliforms at 1.4 x 109 CFU/gdw total solids, and Enterococcus spp. at 8.2 x 105 CFU/gdw total solids.
Samples were assayed for Staphylococcus aureus, but none were detected. In addition, a number of
organic SVOCs and VOCs were measured in the stockpile prior to application.
8.5.2 Bioaerosol and Particulate Sampling. In the aerosol phase of this study, THE, which include
the saprophytic aerobes and facultative anaerobes that are naturally present in soil, on plant surfaces, in
air, and in water, were detected in all biosamplers and on agar plates in all six-stage impactors for both
the control trial and biosolids application test. Their presence in air samples upwind and downwind
demonstrated that the air samplers were operating sufficiently well to collect viable microbes. Fungi were
also assayed and detected in all six-stage impactors.
THE and the fungi data were evaluated to determine if there were significant differences in the
THE and total fungi counts between: 1) upwind and downwind locations and 2) the MOB and other
locations. In addition, apparent differences in THE and fungi data between the control trial and the
biosolids application test were compared. For Comparison 1, statistically significant differences in THE
concentrations were observed between the upwind sampling transect and the farthest downwind sampling
transect during the application of biosolids to the land. Similar behavior was observed for fungi data.
However, no statistically significant differences were observed during the control trial. For Comparison
2, multiple statistically significant differences for THE were observed between the MOB and the
stationary sampling locations. Once again, fungi results were similar to the THE observations. These
observations were noted in both the control trial and application test. No apparent differences in counts
were observed between the control trial and the biosolids application test for THE and fungi data.
Detection of pathogenic and indicator microorganisms was also anticipated due to the bacterial
counts observed in the bulk biosolids. However, fecal coliforms, E.coli, Salmonella spp., S. aureus, C.
perfringens, Enterococcus spp., and coliphage were not detected in any of the bioaerosol samples
collected at the stationary sampling locations on the site either in the control trial or during biosolids
application. These organisms also were not detected in any of the samples collected by the MOB.
Enteric virus analyses were conducted using the plaque forming unit (PFU) and MPN procedures
initially on samples collected from mid-line stationary samplers and MOB for the control trial and
application test. No positive results were found for enteric virus in the PFU and MPN analyses conducted
on these initial samples. Therefore, no further viral analyses were conducted on any additional samples.
Total mass concentrations (ug/m3) of particulates <5.0 um were measured using the GRIMM
particle analyzer. Particulate mass increased by approximately 90 (ig/m3 during the onset of field activity
during the control trial. No statistical differences in mass were noted either between samples collected
during the control trial and the biosolids application sampling period or between samples taken
immediately before and during biosolids application.
84
-------
8.5.3 Volatile Organic and Inorganic Emissions and Odorant Sampling. Headspace analyses
indicated that detectable levels of acetone, 2-butanone, methylene chloride, toluene, dimethyl sulfide, and
dimethyl disulfide were associated with the biosolids removed from the stockpile and applied to the field.
Among the volatile compounds, the estimated emission factor was highest for dimethyl sulfide (range of
230 to 660 ng/g wet solids) for all three sampling periods (0 hr, 24 hr, and 48 hr). The emissions factor
for all of the compounds detected decreased for each of the 2 days following application, except for
dimethyl disulfide, which remained relatively constant or showed a slight increase over time. While
methylene chloride was suspected as a laboratory contaminant, the other detected compounds are likely
organic byproducts of the anaerobic digestion of municipal biosolids typically found in very low
concentrations emitting from biosolids as volatile gases.
Exposure of solid-phase microextraction (SPME) fibers to the headspace of biosolids for 1 hr also
resulted in detectable concentrations of dimethyl sulfide (1.8 to 8.0 ppmv) and dimethyl disulfide (0.75 to
2.0 ppmv). Trace levels (0.25 ppmv) of carbon disulfide also appeared in the SPMEs only in the control
trial. No significant trend was observed in the SPME headspace results for the time period samples (i.e.,
1 hr, 24 hr, and 48 hr after biosolids application).
Flux chamber air emissions (collected within Summa canisters and analyzed by gas
chromatography/mass spectrophotometry) produced detectable concentrations for acetone,
trimethylamine, dimethyl sulfide, and dimethyl disulfide. Estimated flux rates for these compounds were
greater than 1.0 (ig/m2/hr for several of the post-application sampling times (0, 3, 4 hr, and 20 hr). The
rate of flux chamber emissions declined with time after biosolids placement. Dimethyl sulfide and
dimethyl disulfide emissions persisted into the 20th hour after biosolids application when sampling was
terminated. Flux chamber information should be considered carefully as temperatures within the flux
chambers were often higher than ambient temperatures. Since higher temperatures may enhance volatility
and chemical reactions, this emission information may not be representative of biosolids land application.
SPME results obtained from flux chamber samples have been treated as semi-quantitative since
significant holding time delays occurred. Comparison of results between the SPME samples collected in
Tedlar® bags and the Summa canister method analyzed by GC/MS indicated that dimethyl sulfide,
dimethyl disulfide, and trimethylamine were found consistently with both methods. However,
concentrations determined with the canister method were an order of magnitude higher than those
produced with the SPMEs, indicating potential losses over time or other interferences resulting in
inadequate sorption onto the SPME fiber.
Odor was monitored in the field using Nasal Rangers®. The dilution at which the odor was barely
detected, i.e., the detection threshold (DT), was determined using ASTM E544-99 (ASTM, 2004). Odor
units were expressed as the number of dilutions-to-threshold (d/T) before odor was barely detected.
Odorous materials have smaller d/T values than do less odorous materials. For the on-site monitoring
conducted using Nasal Rangers®, DTs during biosolids application were 15 to 30 d/T approximately 25 m
downwind of the application site and 2 to 7 d/T approximately 25 m upwind of the application area. At
distances greater than 75 m, odors were undetected at both upwind and downwind locations.
Approximately 22 hr after biosolids application, odor levels in the application area were comparable to
the levels at 25 m upwind and downwind during application, and odor was undetectable above
background levels elsewhere in the project area. After 48 hr, odors were barely detectable downwind of
the site, and after 196 hr, no odors were detected above background in any location on the site.
Additional samples were collected from flux chamber exhaust in 12-L Tedlar® bags for shipment
to a laboratory and off-site olfactometry analysis to confirm odor presence and determine odor
concentrations using ASTM E679-91 (ASTM, 1991). Substantially increasing odors were detected by the
off-site panel on samples collected during biosolids application up to 22 hr after application. It is
85
-------
believed the increasing odors noted into the afternoon and throughout the day of application were an
artifact of the flux chamber units. Solar heating elevated air temperatures inside the chambers, thereby
increasing volatility and anaerobic degradation of organic sulfur compounds contained in the applied
biosolids. Therefore, while flux chambers are useful for estimating odor generation during and for
perhaps up to an hour after biosolids application, they are not considered practical, effective, and accurate
for determining odor levels over longer sampling periods stretching into multiple hours and days.
In-field measurements of ammonia and hydrogen sulfide were conducted using Draeger tubes and
the Jerome® analyzer, respectively. Sampling locations were identified during the Nasal Ranger®
monitoring activities. Ammonia was not detected in above-ground air samples during the control trial
within the application area. Immediately after the application test, ammonia was detected at
concentrations of 0.10 to 0.90 ppmv for near-ground samples within the application area and from a flux
chamber exhaust sample at 15 ppmv. Within the application area, hydrogen sulfide was detected at levels
near the recognition threshold, 0.002 to 0.050 ppmv during the control trial and 0.007 to 0.021 ppmv
directly behind the moving biosolids application equipment during biosolids application. The exhaust
from the biosolids applicator machinery may have been responsible for some of the hydrogen sulfide gas
measured. Immediately after the biosolids application test, hydrogen sulfide was detected within the
application area at slightly lower concentrations than those detected during the control trial (within the
range of 0.001 to 0.007 ppmv for near-ground air samples) and from a flux chamber exhaust sample at
0.160 ppmv. The highest measurements for each of the gases never approached any health criterion or
guidance level, and both gases were below detectable limits (0.001 ppmv) 400 m downwind of the
application area during the application trial. Due to their high vapor pressures and the existing field
conditions, the concentration of ammonia and hydrogen sulfide decreased by the second day after
application and were below detectable limits within 4 days.
The OP-FTIR also confirmed the presence of ammonia, and the data were used to calculate a flux
emission rate across the site after the addition of biosolids to the test area. Immediately after biosolids
addition and upwind of the application area, the calculated ammonia flux rate was 0.006 g/s while
downwind it was 0.063 g/s. In order to investigate the rate of emissions decay, measurements continued
for several hours after biosolids application. The upwind VRPM configuration measured negligible
ammonia concentrations during the post-application period. However, the downwind VRPM
configuration detected ammonia plumes for several hours after the application ended. The calculated
emission flux from the reconstructed ammonia plume was 0.036 g/s at 2 hr after application, decreasing
further to 0.009 g/s approximately 3 hr after application.
8.5.4 Land Sampling. The soil sampling portion of this study focused on the concentrations of
microbes and chemicals found in the soil matrix prior to and for 4 months following the applications of
biosolids. The quantity and distribution of applied biosolids were characterized by measuring dry mass
and ash mass in three replicate plots. Statistical analysis determined that biosolids were evenly
distributed in two plots at 10.8 + 4.4 g ash/900 cm2. In the other plot, biosolids distribution was uneven at
20.0 + 5.6 g ash/900 cm2. These data and further statistical analysis are consistent with qualitative
observations made in the field (see Section 7.0 for photographs showing the distribution of biosolids at
the test site). Measured wet and dry masses were used to calculate the amount of biosolids applied, 7.3 to
9.5 wet tons/acre, which is equivalent to 1.7 to 2.2 dry tons/acre. This measured application rate was
slightly lower but in the range of the planned agronomic application rate of 10 wet tons/acre.
The quantity and diversity of the microbial community were characterized at three depths. Total
biomass was measured based on the quantity of lipid phosphate in a sample, while community structure, a
measure of microbial diversity, was characterized by the relative abundance of individual PLFAs.
Microbes produce different PLFAs depending on the types of organisms present and environmental
conditions. Biosolids application may alter the microbial community of the soil by adding nutrients,
86
-------
organic matter, chemicals, and microbes. PLFA-based community structure assessments displayed small
changes correlating to changes in soil conditions, such as the rainfall immediately prior to biosolids
application. More substantial changes were observed following biosolids application. However, by Day
28, community structure was similar to its pre-application state.
The PLFA-based biomass measurements documented differences as a function of soil depth with
the majority of the microbial mass occurring within the top 5 cm of the soil matrix. Biosolids application
increased the total biomass by 60 nmol/g dry weight on average. Sample variability, often around 30%,
limited the statistical conclusions that could be drawn. Post-application levels were stable from Day 28
through the end of the sampling period at Day 98. Statistically significant differences were observed
between plots.
Assessments of pathogenic organisms relied on measurement of fecal coliforms as an indicator
organism. A limited number of samples were also analyzed for enteric viruses, Salmonella spp., and
viable Helminth ova (VHO). Semi-quantitative fecal coliform data were reported by the analytical
laboratory, complicating statistical evaluations. Fecal coliform density displayed a statistically significant
increase of more than 100-fold between pre-application and post-application samples. The post-
application density generally persisted through Day 98 following application. About 20% of the samples
contained detectable concentrations of enteric viruses, Salmonella spp., and VHO prior to, during, and
after biosolids application.
APEs and their metabolites OP and NP were of interest as potential endocrine disrupting
chemicals (EDCs). APEs were not detected in the soil at any time during the sampling period. OP and
NP were observed frequently following application in shallow samples but not in deeper samples.
Downward migration of these compounds may have been affected by the physical/chemical structure of
the biosolids, the affinity of these compounds for the biosolids matrix, and the high clay content of the
soil in this study. Variability in OP and NP concentrations was too high to identify temporal changes.
Ecotoxicity was screened using earthworm mortality and 5-day seed germination and root
elongation assays with lettuce and oats. Results did not exhibit a consistent pattern. Earthworm mortality
and seed germination exhibited no changes following biosolids application. Since both assays indicated
maximal response prior to application, no added benefit of biosolids could be demonstrated. Neither of
the above assays produced a negative response to biosolids. Lettuce root length, however, demonstrated
enhanced growth following biosolids application, while oat root length showed reduced growth.
8.6 Lessons Learned
Implementing this large-scale and fairly complex strategy for sampling during an active biosolids
land application event had inherent challenges, and difficulties were encountered during the field trial that
may have influenced the final results. The study relied on meteorological data collected in the northeast
quadrant of the test area. The wind direction varied substantially between the control trial and application
test, while the velocity varied only slightly. Subtle changes may have been observed by operators across
the application area. In retrospect, real-time wind direction and wind speed recordings within individual
transects or even stations would have been informative and may have helped in the interpretation of the
variability observed between individual stations on the same and/or differing transect lines.
Even though two-way radios and headsets were used by operators at each station, communication
among personnel assigned to each sampling station was still relatively ineffective, primarily due to noise
and obstruction of view from the applicator. This situation may have resulted in slight discrepancies in
sampling times on sample lines and did not permit the samplers to be repositioned into the prevailing
wind direction during wind shifts as originally planned.
87
-------
Participate matter was only recorded in the center of the application area and was presumed to
travel some distance downwind. Future studies should consider the use of real-time particulate analyzers
in the upwind and downwind sampling locations. The dataset produced from this bioaerosol sampling
effort was highly variable.
The study was designed to accommodate a number of different objectives for other specific tasks
in addition to the aerosol sampling component. For example, it was of interest, particularly for Task 3, to
conduct this work at a site where biosolids had never been land applied. However, based on the
variability observed in this trial, the bioaerosol sampling component would have benefited from replicated
application tests and subsequent focused sampling events to increase the number of observations and
reduce uncertainties.
In retrospect, the bioaerosol sampling design may have benefited from the use of air dispersion
modeling once the application site was identified and predominant weather conditions could be estimated.
When sampling for studies that span multiple days, care should be taken to collect the samples during the
same approximate time intervals each day. This practice may help decrease the impacts that may be
caused by variable environmental conditions, such as temperature, wind direction, humidity, etc.
The nature of the biosolids used may have had a significant impact on the data generated in this
study. The biosolids were sticky and cohesive, possibly due to polymer addition during sludge
dewatering or other sludge processing operations. It is believed the viscid nature of the biosolids
substantially reduced their friability and perhaps limited their ability to be dispersed as fine particles into
the air. Consequently, the biosolids (when slung from the application vehicle and applied to the test site)
tended to settle onto the ground in small agglomerated clumps rather than as discrete particles. This
characteristic visually impacted the distribution of the biosolids on the soil surface and was not
anticipated in the soil/biosolids sample collection plan. Further, the cohesive nature of the biosolids may
have decreased the number of particles aerosolized and, consequently, the capture and detection of
aerosolized microorganisms. For future studies in which the primary objective is to maximize dispersion
of biosolids and associated chemicals and microorganisms, it is recommended that preliminary screening
evaluate biosolids friability and likely distribution into the air. Application of liquid biosolids may also
yield a more uniform distribution of droplets to the soil and into the air. In this study, application of a
more agglomerating biosolids product appears to have restricted the applied material to the immediate
area and limited the spread of airborne particles to downwind receptors.
8.7 Recommendations
Additional work will be needed in order to develop a detailed protocol for future biosolids land
application studies. At the completion of this study, the following recommendations, coupled with
suggestions for improvement where appropriate, can be offered for similar activities:
8.7.1 Bioaerosol Sampling. Careful consideration should be given to the analyte list when designing a
bioaerosol sampling protocol. The customarily-used sampling devices have significant limitations in
enabling the conduct of a comprehensive bioaerosol sampling program, particularly for capturing stress-
sensitive pathogenic bacteria. Until such time as more robust equipment is designed and developed
specifically for outdoor applications, the types and species of microorganisms and/or indicators selected
for assay should be those that are relevant to project objectives but also relatively easy to culture.
Meteorological data such as wind velocity and patterns should be acquired for a substantial period
prior to the study and used to develop predictive models, which in turn could be used to develop and
optimize bioaerosol sampling array designs. Ultimately, development of a sampling array design that is
independent of wind direction (i.e., a design that does not need to consider a shift in sampling locations to
-------
acquire representative samples) is a desirable goal. Also, incorporation of elevated samplers in the
sampling array design to accommodate capture of vertically dispersed bioaerosols would enhance the
overall comprehensiveness of the design.
In addition, when sampling aerosols at time intervals for studies that span multiple days, care
should be taken to collect the samples at the same approximate time period each day. This practice may
help decrease the impacts that may be caused by variable environmental conditions, such as temperature,
wind direction, humidity, etc.
Endotoxin sampling may be useful to indicate the bioaerosol emissions associated with biosolids
and should be evaluated further. Even though analytical, logistical, and quality control problems made it
impossible to interpret data from this study, endotoxins are generally easy to sample and may be present
in sufficiently high numbers to permit the conduct of a statistical analysis.
8.7.2 Particulate Sampling. Particulate sampling should be conducted both upwind and downwind of
biosolids application areas and in close association with bioaerosol samplers so that bioaerosol and
particulate data can be correlated. The design implemented in this study, which confined particulate
sampling to the center of the application area, did not allow for this correlation of data. The GRIMM
particle analyzer was utilized in this study to develop mean mass information on airborne particulates for
different phases of the project. For future studies, this particulate monitoring device could also be used to
take advantage of its real-time data acquisition capabilities to facilitate in-field decisions such as
placement or movement of aerosol sampling equipment.
8.7.3 Volatile Organic Compound and Ammonia Sampling. Collecting samples using VRPM via
an OP-FTIR spectrophotometer is expensive and complex, but this technique results in a three-
dimensional map of plumes emanating from biosolids land application. From these measurements,
chemical-specific flux estimates can be determined. In this study, ammonia plumes were detected during
and following biosolids application, but not during baseline sampling. No interpretive data were
developed regarding VOCs during the study. Given these results, while use of this technique cannot be
conclusively recommended for detecting and quantifying VOCs, it's use can be recommended for
producing real-time estimates of ammonia plume generation and dissipation.
Flux chambers used in combination with Summa canisters can also be used to estimate VOC
emissions; however, internal chamber temperatures can become elevated influencing chemical
volatilization and microbial activity, and possibly biasing emission results. It is recommended that future
studies employ the use of Summa canisters without flux chambers and that they be positioned in upwind
and downwind locations, as well as in the application area. Predictive models can also be used to identify
appropriate Summa canister stations. SPME-based analytical methods may also be useful, but care
should be exercised to stay within sample holding times.
Ammonia sampling using Draeger tubes and hydrogen sulfide measurements using the Jerome
analyzer provided inexpensive and reliable measurements of these two odiferous compounds at specific
locations downwind of the biosolids application site. These measurements provided useful information
regarding the geographic distribution of odiferous chemicals in the area surrounding the biosolids land
application site.
8.7.4 Odor Sampling. Field olfactometry using on-site Nasal Rangers® is an acceptable method to
establish real-time geographic distribution of odor and worked well for this study. The protocol would
likely benefit from sampling for a longer duration to account for temporal changes due to local area
impacts such as changing temperature, humidity, or wind direction. As discussed earlier, flux chamber
samples may have been influenced by higher internal temperatures. Nasal Rangers® are more flexible,
89
-------
effective, and accurate for determining on-site near- and far-field odor levels over extended sampling
periods.
8.7.5 Land Sampling. A successfully implemented field protocol for determining the effects of
biosolids on land is dependent on a carefully designed statistical approach due to the inherent
heterogeneity of the microbial population in the soil. The use of replicate sample plots and replicate
sampling within plots is strongly recommended to facilitate separation of plot effects from time or other
variables for statistical analysis.
Since biosolids may not be applied to the land uniformly, it is necessary to determine the spatial
distribution of the biosolids material as applied. The method of pinning geotextile fabric to the ground to
collect biosolids mass and distribution samples for this study worked quite well and is recommended for
additional studies.
PLFA measurements reliably track changes in the soil microbial community that result from
biosolids land application. Because PLFA results vary significantly by depth, it is important that
sufficient numbers and types of samples be collected in order to facilitate the statistical analyses needed to
identify differences that are important to the study. Furthermore, the results from this study would
suggest using an approach that focuses sample replication in the shallow soil horizon as minimal
downward analyte migration was observed. However, this observation may have been impacted by the
high level of fine-grain particles (clay) present in the soil at this site.
The reliance on fecal coliforms as an indicator organism was useful as detection frequency was
high enough to facilitate interpretative analysis of the data. However, due to data quality problems,
uncertainties in the conclusions that could be drawn from the fecal coliform data were substantial.
Follow-on studies are recommended. More sample replicates within a plot, more replicate plots, and
more precise data from the analytical laboratory may reduce these uncertainties. In addition, extending
the sampling period following biosolids application may be useful.
The results of the APE testing indicated that the analytical methods used were appropriate for
biosolids and biosolids/soil mixtures. However, since APEs persisted in the soil matrix for the duration
of the study, a longer sampling period may be needed for future projects. Due to the variability in data
observed at this site, future APE sampling designs may benefit from a different approach to sampling that
considers such options as larger samples with homogenization, sample or extract composites, or
normalization with a biosolids marker. Finally, expanding the list of EDC analytes to include a broader
spectrum of chemicals is recommended.
Toxicity screening produced information characterizing the response of selected organisms to
biosolids application and is recommended for future studies. In addition to the assays used in this study,
chronic earthworm tests with non-lethal endpoints, such as biomass accumulation, and longer-term plant
studies would be useful in gaining a fuller understanding of the effects of biosolids. A longer sampling
period may also be useful.
Soil characterization was relatively inexpensive compared to many of the other analyses
conducted and provided useful supporting data for this study. This characterization, particularly for sites
such as this one with high soil clay content, should be focused on the shallower depth horizon due to
microbial population dynamics. Similarly, weather data collection was relatively inexpensive and
provided useful information.
90
-------
9.0 REFERENCES
Acheson, C.M., Q. Zhou, Y. Shan, G.D. Sayles, and M.J. Kupferle. 2004. "Comparing the Solid Phase
and Saline Extract Microtox(R) Assays for Two Polycyclic Aromatic Hydrocarbon-Contaminated
Soils," Environmental Toxicology and Chemistry. 23, 245-251.
Anderson, T.D. and J.D. MacRae. 2006. "Polybrominated Diphenyl Ethers in Fish and Wastewater
Samples from an Area of the Penobscot River in Central Maine," Chemosphere. 62, Issue 7,
1153-1160.
ASTM Standard E679-91. 1991. "Standard Practice for Determination of Odor and Taste Thresholds by a
Forced-Choice Ascending Concentration Series Method of Limits," ASTM International, West
Conshohocken, PA, www.astm.org.
ASTM Standard E544-99. 2004. "Standard Practices for Referencing Suprathreshold Odor Intensity,"
ASTM International, West Conshohocken, PA, www.astm.org.
APHA. 1992. Method 9215 C. "Spread Plate Method". Standard Methods for the Examination of Water
and Wastewater, 17th Edition, Washington, DC.
Baertsch, C., T. Paez-Rubio, E. Viau, and J. Peccia. 2007. "Source Tracking of Aerosols released from
Land-Applied Class B Biosolids During High-Wind Events," Applied and Environmental
Microbiology. 739 (14), 4522-4531.
Banks, M.K., A.P. Schwab, N. Cofield, J.E. Alleman, M. Switzenbaum, J. Shalabi, and P. Williams.
2006. "Biosolids Amended Soils: Part I. Effect of Biosolids Application on Soil Quality and
Ecotoxicity," Water Environ. Res. 78, 2217-2230.
Beebe, K.R., R.J. Pell, and M. Seashultz. 1998. Chemometrics: A Practical Guide, John Wiley and
Sons, New York, NY.
Brooks, J, B. Tanner, K. Josephon, C. Gerba, C. Haas, and I. Pepper. 2005. "A National Study on the
Residential Impact of Biological Aerosols from the Land Application of Biosolids," Journal of
AppliedMicrobiolg. 98(2), 397-405.
Brooks, J., B. Tanner, K. Josephson, C. Gerba, and I. Pepper. 2004. "Bioaerosols from the Land
Application of Biosolids in the Desert Southwest USA," Water Science and Technology. 50(1), 7-
12.
Barth, E. 2007. "Lessons Learned Regarding the Use of an Impinger for Collecting Airborne Bacteria
during a Biosolids Application Field Study." Proceedings of the Biosolids Exposure
Measurement Workshop, Office of Research and Development, EPA, Cincinnati, OH.
Battelle. 2004. Final Quality Assurance Project Plan (QAPP) for Multi-Media Sampling at Biosolids
Land Application Sites #390-QC-1. September.
Chang, L.W., J.R. Meier, and M.K. Smith. 1997. "Application of Plant and Earthworm Bioassays to
Evaluate Remediation of a Lead-contaminated Soil," Arch. Environ. Contamin. Tox. 32, 166-171.
91
-------
Dettenmaier, E. and W.J. Doucette. 2007. "Mineralization and Plant Uptake of 14C-labelled Nonylphenol,
Nonylphenol Tetraethoxylate and Nonylphenol Nonylethoxylate in Biosolids/Soil Systems
Planted with Crested Wheatgrass," Environmental Toxicology and Chemistry. 26, 193-200.
Dorn, P.B., T.E. Vipond, J.P. Salanitro, and H.L. Wisniewski. 1998. "Assessment of the Acute Toxicity
of Crude Oils in Soils Using Earthworms, Microtox®, and Plants," Chemosphere. 37, 845-860.
Environment Canada. 1994. A Review of Whole Organism Bioassays for Assessing the Quality of Soil,
Freshwater Sediment, and Freshwater in Canada (Scientific Series No. 198), Ottawa, ON.
Fabian, M., S. Miller, T. Reponen, and M. Hernandez. 2005. "Ambient Bioaerosol Indices for Indoor Air
Quality Assessments of Flood Reclamation," Aerosol Science, 36, pp. 763-783.
Gabriel, S., S. Vilalai, S., C. Peot, and M. Ramirez. 2006. "Statistical Modeling to Forecast Odor Levels
of Biosolids Applied to Reuse Sites" Journal of Environmental Engineering, 132(5), 479-488.
Hamel, K., L. Walters, C. Sulerud, and M. McGinley. 2004. "Land Application Odor Control Case
Study," Water Environmental Federation Residuals and Biosolids Management Conference, Salt
Lake City, Utah.
Heidelberg, J., M. Shahamat, M. Levin, I. Rahman, G. Stelma, C. Grim, and R. Colwell. 1997. "Effect of
Aerosolization on Culturability and Viability of Gram-Negative Bacteria," Applied and
Environmental Microbiology, 63(9), pp. 3585-3588.
Hesselsoe, M., D. Jensen, K. Skals, T. Olesen, P. Moldrup, P. Roslev, G.K. Mortensen, and K. Henriksen.
2001. "Degradation of 4-Nonylphenol in Homogeneous and Nonhomogeneous Mixtures of Soil
and Sewage Sludge," Environ. Sci. Technol. 35, 3695-3700.
Hirst, J.M. 1995. "Bioaerosols: Introduction, Retrospect and Prospect," C.S. Cox and C.M. Wathes, eds.,
Bioaerosols Handbook. CRC Press, Inc., Lewis Publishers, Boca Raton, FL, pp. 5-25.
Hseu, Z.Y. 2006. "Response of Microbial Activities in Two Contrasting Soils to 4-nonylphenol Treated
with Biosolids," Chemosphere. 64, 1769-1776.
Hund, K. and W. Traunsurger. 1994. "Ecotox-evaluation Strategy for Soil Bioremediation Exemplified
for a PAH-contaminated Site," Chemosphere. 29, 371-390.
Kaur, A., A. Chaudhary, A. Kaur, R. Choudhary, and R. Kaushik. 2005. "Phospholipid Fatty Acid - A
Bioindicator of Environment Monitoring and Assessment in Soil Ecosystem," Current Science.
89, 1103-1112.
Kinney, C.A., E.T. Furlong, D.W. Kolpin, M.R. Burkhardt, S.D. Zaugg, S.L. Werner, J.P. Bossio, and
M.J. Benotti. 2008. "Bioaccumulation of Pharmaceuticals and Other Anthropogenic Waste
Indicators in Earthworms from Agricultural Soil Amended with Biosolid or Swine Manure,"
Environ. Sci. Technol. 42(6), 1863-1870.
Krach, K., B. Burns, B. Li, A. Shuler, C. Colee, and Y. Xie. 2008. "Odor Control for Land Application
of Lime Stabilized Biosolids," Water Air & Soil Pollution: Focus. 8(3-4), 368-378.
92
-------
Low, S., T. Paez-Rubio, C. Baertsch, M. Kucharski, and J. Peccia. 2007. "Off-site Exposure to
Respirable Aerosols Produced During the Disk-Incorporation of Class B Biosolids," Journal of
Environmental Engineering. 133(10), 987-994.
Macher, J.M. 1989. Positive-hole Correction of Multiple-jet Impactors for Collecting Viable
Microorganisms. Amer. Inc. Hyg. Assoc. J. 50, 561-568.
Marcomini, A., P.O. Capel, Th. Lichtenseiger, P.H. Brunner, and W. Giger. 1989. "Behaviour of
Aromatic Surfactants and PCBs in Sludge-treated Soil and Landfills," Journal of Environmental
Quality. 18, 523-528.
Meier, J.R., L.W. Chang, S. Jacobs, J. Torisella, M.C. Meckes, and M.K. Smith. 1997. "Use of Plant and
Earthworm Bioassays to Evaluate Remediation of Soil from a Site Contaminated with
Polychlorinated Biphenyls," Environ. Tox. Chem. 16, 928-938.
Mortensen, O.K. and L.K. Kure. 2003. "Degradation of Nonylphenol in Spiked Soils and in Soils Treated
with Organic Waste Products," Environ. Tox. Chem. 22, 718-721.
Moce-Llivina, L., M. Muniesa, M. Pimenta-Vale, F. Lucena, and J. Jofre. 2003. "Survival of Bacterial
Indicator Species and Bacteriophages after Thermal Treatment of Sludge and Sewage." Appl.
Environ. Microbiol. 69, 1452-1456.
National Research Council (NRC). 2002. Biosolids Applied to Land — Advancing Standards and
Practices, The National Academies Press, Washington, DC.
NIOSH, 1998. Biosolids Land Application Process - LeSourdsville, OH. HETA 98-0118-2748,
Cincinnati, OH
Paez-Rubio, T., A. Ramarui, J. Sommer, H. Xin, and J. Peccia. 2007. "Emission Rates and
Characterization of Aerosols Produced During the Spreading of Class B Biosolids,"
Environmental Science and Technology. 41, 3537-3544.
Peacock, A.D., M.D. Mullen, D.B. Ringelberg, D.D. Tyler, D.B. Hedrick, P.M. Gale, and D.C. White.
2001. "Soil Microbial Community Responses to Dairy Manure or Ammonium Nitrate
Applications," Soil Biology and Biochemistry. 33, 1011-1019.
Pillai, S., K. Widmer, S. Dowd, and S. Ricke. 1996. "Occurrence of Airborne Bacteria and Pathogen
Indicators during Land Application of Sewage Sludge" Applied Environmental Microbiology.
62(1), 296-299.
Rusin, P., S. Maxwell, J. Brooks, C. Gerba, and I. Pepper. 2003. "Evidence for the Absence of
Staphylococcus aureus in Land Applied Biosolids," Environmental Science and Technology. 37,
4027-4030.
Salintro, J.P., P.B. Dorn, M.H. Huesemann, K.O. Moore, I.A. Rhodes, L.M. Rice Jackson, T.E. Vipond,
M.M. Western, and H.L. Wisniewski. 1997. "Crude Oil Hydrocarbon Bioremediation and Soil
Ecotoxicity Assessment," Environ. Sci. Technol. 31, 1796.
Scott, T.M., J.B. Rose, T.M. Jenkins, S.R. Farrah, and J. Lukasik. 2002. "Microbial Source Tracking:
Current Methodology and Future Directions."^/)/. Environ. Microbiol. 68: 5796-5803.
93
-------
Statistical Analysis Software. 2003. SAS Version 9.1, Gary, NC.
Simini, M., R.S. Wentsel, R.T. Checkai, C.T. Phillips, N.A. Chester, M.A. Major, and J.C. Amos. 1995.
"Evaluation of Soil Toxicity at Joliet Army Ammunition Plant," Environ. Tox. Chem. 14,
623-630.
Tanner, B., J. Brooks, C. Gerba, C. Haas, K. Josephson, and I. Pepper. 2008. "Estimated Occupational
Risk from Bioaerosols Generated During land Application of Class B Biosolids," Journal of
Environmental Quality. 37, 2311-2321.
Tanner, B., J. Brooks, C. Haas, C. Gerba, and I. Pepper. 2005. "Bioaerosol Emission Rate and Plume
Characteristics during Land Application of Liquid Class B Biosolids." Environ. Sci. Techno!.
396, 1584-1590.
Tanner, B., J. Brooks, C. Haas, C. Gerba, and I. Pepper. 2005b. "Bioaerosol Emission Rate and Plume
Characteristics during Land Application of Liquid Class B Biosolids," Environmental Science
and Technology. 396,1584-1590.
U.S. Environmental Protection Agency (EPA). 1988. Protocols for Short-Term Toxicity Screening of
Hazardous Waste Sites, EPA/600/3-88/029, Corvallis Environmental Research Laboratory:
Corvallis, OR.
U.S. Environmental Protection Agency (EPA). 1989. Ecological Assessment of Hazardous Waste Sites,
EPA/600/3-89/013, Corvallis Environmental Research Laboratory, Corvallis, OR.
U.S. Environmental Protection Agency (EPA). 1992. Technical Support Document for Part 503
Pathogen and Vector Attraction Reduction Requirements in Sewage Sludge, National Technical
Information Service, NTIS No. PB89-136618, Springfield, VA.
U.S. Environmental Protection Agency (EPA). 1999. Measurement of Volatile Organic Compounds in
Biosolid Samples by Headspace Analysis Using Capillary Column Gas Chromatography/Mass
Spectrometry (GC/MS).
U.S. Environmental Protection Agency (EPA). 2001. Manual of Methods for Virology, Chapter 15.
U.S. Environmental Protection Agency (EPA). 2003. Environmental Regulations and Technology:
Control of Pathogens and Vector Attraction in Sewage Sludge, EPA/625/R-92/013.
Unc, A., J. Gardner, and S. Springthorpe. 2006. "Recovery of Soil E. coll after Addition of Sterile
Organic Wastes," Appl. Environ. Microbiol. 72, 2287-2289.
Van Belle, G. and B.C. Martin. 1993. "Sample Size as a Function of Coefficient of Variation and Ratio of
Means," The American Statistician. 47, 165-167.
Vestal, J.R and D.C. White. 1989. "Lipid Analysis in Microbial Ecology," Bioscience. 39, 535-541.
Winfield, M.D. and E. Groisman. 2003. "Role of Non-host Environments in the Lifestyles of Salmonella
and Escherichia coli," Appl. Environ. Microbiol. 69, 3687-3694.
94
-------
APPENDIX A
Determination of Total Bacterial Bioburden from Impinger Samples Collected During the NC
Biosolids Land Application Study -Dr. Mark Hernandez, University of Colorado
-------
PROJECT REPORT
Molecular-based Identification of Bacterial Constituents
from Aerosols Collected During the Land Application of Biosolids
Department of Civil, Environmental and Architectural Engineering, and Department of Molecular, Cellular, and
Developmental Biology, University of Colorado at Boulder, University of Colorado, Boulder, Colorado, 80309
ABSTRACT
The ability of standard laboratory methods to detect microorganisms that could be potentially liberated to
the environment during the land application of biosolids has not been well studied. Bacteria which may
be aerosolized during the land application process, or during the subsequent weathering of biosolids,
have only recently been investigated with modern genetic methods. The purpose of this study was to use
the polymerase chain reaction (PCR) to amplify 16S rRNA genes for the detection of microbial
indicators and/or pathogens commonly associated with biosolids. Air samples were obtained from liquid
impingers immediately circumventing a well defined biosolids application area before, during and after
spreading on a grass field. Samples were collected as close to the application vehicle as safety allowed.
Liquid impinger samples were sent to the USDA microbiology laboratory for standard enrichments, and
aliquots of those samples were processed to isolate and purify bacterial DNA, which was subsequently
analyzed using broad-range rRNA PCR. Of 16 aerosol samples analyzed, 5 contained DNA that could
be amplified by PCR. Genetic amplification detected an average of 35 different organisms in those 5
samples (range 28-51). In total, 439 rRNA clones were screened, and 157 phylotypes (DNA sequence
relatedness groups) were identified. Using the most recent genetic library available from GenBank, the
most abundant lineages/species were previously uncultured groups of bacteria that could not be classified
by current systematic taxonomomy (18%), followed by a Beta Proteobacterium (6%), an uncultured LI 1
bacterium (4%), Corynebacterium segmentosum (3%), a Ralstonia spp. (3%), Lactobacillus lactus SL3
(3%), and a Sphingomonas sp. SKJH-30 (2%). Markedly less abundant species also were detected in
some of the samples that may be considered as indicators or pathogens associated with sewage sludge,
including two Clostridium species, several Lactobacillus species, and a CDC Group DF-3 isolate.
Overall, the data demonstrated a low level of concordance with classical indicator organisms or
pathogens associated with sewage sludge biosolids.
*Corresponding Investigator: Mark Hernandez,
Electronic mail: Mark.Hernandez@Colorado.Edu: Phone: 303 492 5991
A-l
-------
INTRODUCTION
The aerosolization of bacteria from biosolids applications has long been a point of controversy and needs
further investigation. Biosolids are repositories of both aerobic and anaerobic bacterial species, some of which are
pathogenic to humans and other mammals. Whether standard culturing techniques can detect the full microbial
array in biosolids as compared to newer molecular based techniques has not been fully investigated. Conventional
laboratory techniques involve the use of bacterial plates with defined media that may also select for specific
organisms. Less well understood is the symbiotic nature of mixed bacterial communities in biosolids with potential
pathogens, and how land application and weathering may affect their aerosolization potential.
Knowledge of the complexities of the microbial communities found within biosolids is minimal.
Traditionally, the identification and enumeration of microbial species from biosolids or their associated aerosols has
depended entirely upon pure-culture techniques. However, the difficulty with which some types of microorganisms
are cultured, particularly those with fastidious and/or anaerobic physiologies, means that the more easily grown
species in a mixed microbial community likely are overrepresented by cultivation and plate counts. In many
complex natural environments, for example, less than one percent of viable microbes present can be cultured under
standard conditions (Pace, 1997).
Recently, culture-independent molecular methods of microbial identification and characterization have
been developed and applied in the context of microbial ecology. Several of these techniques involve the use of
ribosomal RNA (rRNA) gene sequences as tools for species identification by means of phylogenetic sequence
analysis. rRNA genes can be amplified by polymerase chain reaction (PCR) directly from mixed-community DNA
preparations, cloned, and individual clones sequenced. The occurrence of a particular rRNA gene sequence in a
clone library indicates that the organism that encodes this sequence is present in the sampled community. On the
basis of rRNA sequence comparisons, species-specific DNA- or RNA-hybridization probes subsequently can be
designed and used to enumerate the various types of organisms present in an aerosol or biosolids sample. Only
recently have some environmental laboratories used rRNA-based molecular techniques in order to identify and
characterize human pathogens and commensals aerosolized from the purposeful land application of sewage (Paez-
Rubio, 2005). These studies have identified a plethora of microbes associated with human waste and sewage, many
of which represent novel genera that previously were not described at the molecular level.
Less than twenty rRNA-based studies of bacterial or fungal bioaerosols have been published to date
(included in reference list), with limited numbers of samples and rRNA sequences analyzed. To further investigate
the biodiversity of aerosols associated with biosolids and identify potential pathogens liberated during the land
application of biosolids, we used broad-range 16S rRNA PCR, cloning, and subsequent sequencing to characterize
air samples obtained from liquid impingers immediately circumventing a well defined biosolids application area
before, during and after their spreading. This information will be compared to bacteria recovered and identified
A-2
-------
using standard clinical laboratory techniques.
MATERIALS AND METHODS
Sample Collection. Bioaerosol samples for genetic analyses were collected using swirling liquid impingers
according to accepted methods and manufacturer's specifications (BioSampler, SKC Inc., Eighty Four, PA). The
efficiency of the BioSampler filled with 20 ml of water is 79% for 0.3 jam particles, 89% for 0.5 jam particles, 96%
for 1 |j,m particles and 100% for 2 |am particles. Particle-free, autoclaved 0.01 M phosphate-buffer saline (PBS)
containing 0.01% Tween 80 (SIGMA, St. Louis, MO) was used as the collection medium in all impingers. 1 ml
aliquots were aseptically transferred in a commercial PCR prep hood to DNA free microcentrifuge tubes and were
immediately shipped on dry ice to the University of Colorado at Boulder for molecular analysis, following custody
protocols approved by the USEPA.
DNA Sample Preparation. A rigorous solvent and grinding DNA extraction protocol was used to process the
aerosol samples. Samples were placed in 2 ml microcentrifuge tubes to which 700 |al of Buffer (200 mM NaCl, 200
mM Tris-Cl pH 8.0, 20 mM EDTA, 5% SDS), 500 (ol phenol/chloroform and 0.5 g zirconium beads (Biospec
Products Inc, Bartlesville, OK) were added. The samples were agitated in a Mini Beadbeater-8 (Biospec Products
Inc, Bartlesville, OK) on the highest setting for four minutes and then subjected to centrifugation (13000 rpm) for 5
minutes. The aqueous phase was extracted with phenol/chloroform. DNA was precipitated by addition of NaCl
(280 mM final cone.) and 2.5 volumes of ethanol followed by centrifugation (13000 rpm) for 30 minutes. DNA
pellets were washed once with 70% ethanol, allowed to dry in a laminar flow hood, and resuspended in 50 |al sterile
TE (10 mM Tris-Cl pH 8.0, 1 mM EDTA). Extracted samples were either placed on ice or stored at -20C before
PCR analysis. All DNA extraction and PCR amplification was conducted by Dr. Daniel Frank and Mari Rodriguez
in the Phylogentics laboratory of Norman Pace. Although samples were not processed simultaneously, frozen
aliquoted reagents were used for DNA extractions and PCR amplifications in order to minimize sample-to-sample
variation.
PCR, Cloning, and Sequence Analysis. Small subunit rRNA (SSU rRNA) genes were amplified from DNA
samples by PCR with oligonucleotide primers with broad-range specificity for all bacterial SSU rRNA genes: 8F
(5'AGAGTTTGATCCTGGCTCAG) and 805R (5'GACTACCAGGGTATCTAAT). Each 30 (ol PCR reaction
includedS (ol lOx PCR Buffer, 2.5 jaldNTP mix (2.5 mM each dNTP), 1.5(4,1 50 mMMgC!2, 75 ng of each primer, 1
unit Taq polymerase, and 1 (ol genomic DNA lysate (following manufacturers' protocols). PCR reagents from
Bioline (Biolase polymerase) and Eppendorf (MasterTaq polymerase) produced indistinguishable results. Thirty
cycles of amplification (92C 30 sec., 52C 60 sec., 72C 90 sec.) usually were sufficient to obtain a product of the
appropriate length that was visible in ethidium bromide stained agarose gels (Kodak Inc.). Two negative control
PCR reactions, one with lysate from control extractions, and the other with sterile H2O serving as template, were
A-3
-------
performed for each set of samples processed in order to assess whether contamination of reagents had occurred.
Positive control PCR reactions, which used an environmental genomic DNA sample as template, were also
performed for each set of samples processed. For PCR inhibition controls, equal quantities of positive control DNA
templates were added to each of two PCR reactions set up in parallel, one of which was supplemented with known
quantities of previously sequenced DNA.
DNA fragments were excised from agarose gels (1% agarose gel in tris-borate EDTA) and purified by the
QIAquick® gel extraction kit (Qiagen Inc., Valencia, CA). A portion of each PCR product was cloned into the
pCR4®-TOPO® vector of the Invitrogen (TOPO® TA Cloning kit following the manufacturer's instructions
(Invitrogen Corp., Carlsbad, CA). For each clone library that was constructed, 96 transformants were grown
overnight at 37C in a 96-well culture plate filled with 1.5 mis of 2xYT medium per well. In order to sequence the
inserts of positive transformants, 20 |al of each overnight culture was added to 20 |al of 10 mM Tris-Cl (pH 8.0),
heated 10 minutes at 95C, and centrifuged 10 minutes at 4000 rpm in a 96-well plate centrifuge (Eppendorf Inc.,
Westbury, NY). One |al of culture supernatant was used as template in a 30 |al PCR reaction (38 cycles of the
program listed above) with vector specific primers (T7 and T3 sites). Ten |al of each PCR product were first treated
with the ExoSap-IT kit (USB Corp, Cleveland, OH) and then subjected to cycle-sequencing with the Big-Dye
Terminator kit (Applied Biosystems, Inc., Foster City, CA) following the manufacturers' protocols. Sequencing was
performed on a MegaBACE 1000 automated DNA sequencer (Amersham Biosciences, Piscataway, NJ).
Sequence base calling and assembly were performed with proprietary software. Initial microbial species
identifications were made by a batch BlastN search (National Center for Biological Information; NCBI) using the
client application BlastC13 (NCBI). SSU rRNA gene sequences were aligned to an existing database of rRNA gene
sequences using the computer application ARE. Phylogenetic analyses, including phylogenetic tree estimations,
utilized ARE and PAUP*. Statistical analyses were performed using the R software package (v.2.0.1; www.r-
project.org). The hypothesis that mean values of species/phylotypes were identical between sample sets was tested
by the paired t-test and the paired Wilcoxon rank sum test, at a significance level of a = 0.01.
RESULTS
As summarized in Table A-l, broad-range rRNA PCR was successful for 5 of 16 aerosol samples. The identities of
microorganisms in the samples were preliminarily determined by querying GeneBank with the wound rRNA
sequences using BLAST (basic local alignment search tool). For each sequence analyzed, we defined a "best
BLAST hit", which was the GenBank entry with the highest BLAST bit score. In order to cull sequences of poor
quality, sequences with lengths less than 500 nucleotides or BLAST bit scores less than 500 were dropped from
further analysis. A total of 439 rRNA clones, comprising 5 clone libraries, constituted the final data set.
The distribution of BLAST percent identity scores between the bioaerosol sample sequences and their best BLAST
hits provides an estimate of the extent of novel sequence diversity in the clone libraries. Table A-l shows only the
A-4
-------
sample designations for Clone libraries that were constructed and sequenced from each of the mixed-community
PCR reactions. The mean percent identity for all sequences was 97% (range 90 - 100%). More than 15% of the
clones were identical to previously characterized rRNA sequences. Following alignment of the bioaerosol rRNA
data set, sequences were clustered into phylotypes, or relatedness groups defined by > 97% intra-group sequence
identity. Although there is no objective criterion for differentiating microbial species on the basis of rRNA
sequence similarity, we used a cutoff of 97% identity to define phylotypes because this value provides a
conservative estimate of species diversity. By this criterion, the 439 bioaerosol rRNA clones represent 157
phylotypes. Table A-l also lists the species most closely related (but not necessarily identical) to each of the
phylotypes and the number of clones belonging to each phylotype. The prevalences of each phylotype within the
specimen set are summarized in Table A-l as well. Using the most recent genetic library available from GenBank,
the most abundant lineages/species were previously uncultured groups of bacteria that could not be classified by
current systematic taxonomy (18%), followed by a Beta Proteobacterium (6%), an uncultured LI 1 bacterium (4%),
Corynebacterium segmentosum (3%), aRalstonia spp. (3%), Lactobacillus lactus SL3 (3%), and a Sphingomonas
sp. SKJH-30 (2%). Markedly less abundant species also were detected in some of the samples that may be
considered as indicators or pathogens associated with sewage sludge, including two Clostridium species, several
Lactobacillus species and a CDC Group DF-3 isolate.
Bacteria identified by DNA sequence analysis listed in Table A-l were grouped according to their major lineage
classifications (family, groups, genus etc). Pie graphs are provided (Figures A-l to A-6) to show the higher order
lineage groupings from each sample where DNA could be extracted for successful PCR. The number of DNA
sequences that were amplified in order to make a statistically valid observation from each sample is provided along
with the relative abundance of each of the major lineage groupings.
DISCUSSION
In examining the distribution and abundance of the clone libraries compiled from these bioaerosol samples, no
trends emerged which suggested that significant amounts of bacteria recovered from aerosols were generated during
or after the period when the biosolids were land-applied. This conclusion is based on the following observations: (i)
only 5 of the samples could recover DNA in sufficient quantity, and free of inhibition, for successful PCR
amplification; (ii) where PCR amplification was successful, biosamplers recovered some types of microorganisms
associated with soils regardless of their positions; (iii) while some of the DNA recovered from bioaerosol samples
are closely related to potential pathogens or known enteric microorganisms, (e.g. Clostridium spp. and Lactobacillus
spp.), the relative abundance of these sequences was markedly low with respect to other DNA sequences present,
and not distributed among the samplers closest to the biosolids application unit; and (iv) no sequences appear in
abundance that have been associated with biosolids by conventional culturing techniques, or other recent molecular
surveys (WERF, Peccia et al, 2004).
A-5
-------
In conclusion, broad-range rRNA PCR provided a new perspective to our understanding of aerosol microbiology
near biosolids application sites. Given that wind speeds were negligible during this sampling campaign, these
results may or may not be indeterminate; indeed they may serve as indication of background in the absence of wind
and weathering processes. A molecular survey of microorganisms that are present in the biosolids themselves would
increase the yield of this type of study in the context of determining which microbial populations may serve as
reliable tracers for bioaerosol aerosolization potential — this level of microbal tracer work is currently being carried
out with support from the Water Environment Research Foundation (WERF, Peccia et al., 2004). Until recently,
characterizing bioaerosols associated with these environments was executed by classical culturing methods, which
can be severely limited for their potential to identify and detect a broad range of microorganism; certainly this
potential was demonstrated by juxtaposing the clone library and culturing recoveries here. Finally, although this
phylogentic study provided a static picture of the bioaerosols collected pre- and post-land spreading, it sets a
baseline for future, longitudinal studies that may address the dynamics of airborne microbial ecology associated with
biosolids applications.
REFERENCES
Alvarez, A. J., M. P. Buttner, et al. (1995). "PCR for bioaerosol monitoring: sensitivity and environmental
interference." Applied and Environmental Microbiology 61: 3639-3644.
Amann, R., W. Ludwig, et al. (1995). "Phylogenetic identification and in situ detection of individual microbial cells
without cultivation." Microbial Reviews 59: 143-169.
Angenent, L., S. Kelley, et al. (2005). "Molecular identification of potential pathogens in water and air of a hospital
therapy pool." Proceedings of the National Academay of Science 102: 4860 - 4865.
Boreson, J., A. M. Dillner, et al. (2005). "Correlation bioaerosol load with PM2.5 and PMlOcf concentrations: a
comparison between natural desert and urban fringe aerosols." Atmospheric Environment
Burge, H. A. (1990). "Bioaerosols: Prevalence and health effects in the indoor environment." Journal of Allergy and
Clinical Immunology 85: 687-701.
Burke, J. F., Ed. (1996). PCR: Essential Techniques. Chichester, England, John Wiley and Sons.
Buttner, M. P., K. Willeke, et al. (1997). Sampling and Analysis of Airborne Microorganisms. Manual of
Environmental Microbiology. C. J. Hurst, G. R. Knudsen, M. J. Mclnerney, L. D. Stetzenbach and M. V. Walter.
Washington D.C., ASM Press: 629-640.
Calderon, C., E. Ward, et al. (2002). "Detection of airborne fungal spores sampled by rotating-arm and Hirst-type
traps using polymerase chain reaction assays." Aerosol Science 33: 283-296.
Calderon, C., E. Ward, et al. (2002). "Detection of airborne fungal spores sampled by rotating-arm and Hirst-type
spore traps using polymerase chain reaction assays." Journal of Aerosol Science 33: 283-296.
Eduard, W. and D. Heederik (1998). "methods for quantitaitve assessment of airborne levels of noninfectious
microorganisms in highly contaminated work environments." American Industrial Hygiene Association Journal 59:
113-127.
Maher, N., H. K. Dillon, etal. (2001). "Magnetic bead capture eliminates PCR inhibitors in samples collected form
A-6
-------
the airborne environment permitting detection of Pneumoncystis carinii DNA." Applied and Environmental
Microbiology 67: 449-452.
Mastorides, S. M, R. L. Oehler, et al. (1997). "Detection of airborne Mycobacteria tuberculosis by Air Filtration
and polymerase chain reaction." Clinical and Infectios Disease 25: 756.
Mastorides, S. M., R. L. Oehler, et al. (1999). "The detection of airborne Mycobacterium tuberculosis using
micropore membrane air sampling and polymerase chain reaction." Chest 115: 1-25.
McCluskey, R., R. Sandin, et al. (1996). "Detection of airborne cytomegalovirus in hospital rooms of
immuncompromised patients." Journal of Virological Methods 56(115-118).
Myatt, T., S. L. Johnston, et al. (2004). "Detection of airborne Rhinovirus and its relation to outdoor air supply in
office environments." American Journal of Respiratory and Critical Care Medicine 169: 1187-1190.
NRC (2002). Biosolids Applied to Land- Advancing standards and practices. Washigton, D.C., The National
Academies Press.
Nugent, P. G., J. Cornett, et al. (1997). "Personal monitoring of exposure to genetically modified microorganisms in
bioaerosols: rapid and sensitive detection using PCR." Journal of Aerosol Science 28: 525-538.
Paez-Rubio, T., E. Viau, et al. (2005). "Source bioaerosol concentration and rDNA-based identification of
microorganisms aerosolized at a flood irrigation wastewater reuse site." Applied and Environmental Microbiology
71: In Press.
Palmer, C. J., G. F. Bonilla, et al. (1995). "Detection of Legionella species in reclaimed water and air with the
EnvironAmp Legionella PCR kit and direct fluorescent antibody staining." Applied and Environmental
Microbiology 61: 407-412.
Pascual, L., S. Perez-Luz, et al. (2001). "Detection of Legionella pneumophila in boaerosols by polymerase chain
reation." Canadian Journal of Microbiology 47: 341-347.
Peccia, J. (2004) Characterization of Aerosols during the Land Application of Biosolids. Progress Report to the
Water Environment Research Foundation, Alexandria, VA.
Pillai, S., K. Widmer, et al. (1996). "Ocurrence of Airborne Bacteria and Pathogen Indicators during Land
Application of Sewage Sludge." Applied and Environmental Microbiology 62(1): 296-299.
Sawyer, M., Y. Wu, et al. (1994). "Detection of Varicella-Zoster virus DNA in air samples from hospital rooms."
The Journal of Infectious Diseases 169: 91-94.
Sorber, C. and K. Outer (1975). "Health and Hygience aspects fo spray irrigation." American Jounal of Public
Health 65: 47-52.
Stark, K. D. C., J. Nicolet, et al. (1998). "Detection ofMycoplasma hyopneumoniae by air sampling with a nested
PCR assay." Applied and Environmental Microbiology 64: 543-548.
Stetzenbach, L. D., M. P. Buttner, et al. (2004). "Detection and enumeration of airborne biocontaminants." Current
Opinion in Biotechnology 15: 170-174.
Wakefield, A. E. (1996). "DNA sequences identical to Pneumocytis carinii f. sp. carinii w&Pneumocytis carinii f.
sp. hominis in samples of air spora." Journal of Clinical Microbiology 34: 1754-1759.
Wan, G., S. Lu, et al. (2004). "Polymerase chain reaction used for the detection of airborne Mycobacterium
tuberculosis in health care settings." AJIC 32: 17-22.
A-7
-------
Wan, G.-H., S.-C. Lu, et al. (2004). "Polymerase chain reaction used for the detection of airbonre Mycobacterium
tuberculosis in health care settings." American Journal of Infection Control 32: 17-21.
Williams, R. H., E. Ward, et al. (2001). "Methods for integrated air sampling and DNA analysis for detection of
airborne fungal spores." Applied and Environmental Microbiology 67: 2453-2459.
A-8
-------
Table A-l. Identity of Clones Based on DNA Similarity of 16s Ribosomal DNA Sequeces Catalogued with GENEBANK
Blast Association
Uncultured organism clone M8907A05 small subunit ribosomal RNA
Uncultured beta proteobacterium clone SM1G08 16S ribosomal RNA
Uncultured bacterium clone Lll 16S ribosomal RNA gene, partial
Corynebacterium segmentosum partial 16S rRNA gene, strain NCTC 934
Uncultured bacterium 16S rRNA gene, clone cD0266
Ralstonia sp. 1F2 16S ribosomal RNA gene, partial sequence
Uncultured proteobacterium clone TAF-B73 16S ribosomal RNA gene,
Lactococcus lactis strain SL3 16S ribosomal RNA gene, complete
Sphingomonas sp. SKJH-30 16S ribosomal RNA gene, partial sequence
Uncultured bacterium clone TM06 16S ribosomal RNA gene, partial
Acinetobacter seohaensis 16S ribosomal RNA gene, partial sequence
Uncultured organism clone MC061215 small subunit ribosomal RNA gene,
Staphylococcus epidermidis partial 16S rRNA gene, isolate SLF1
Bacteroidetes bacterium LC9 16S ribosomal RNA gene, partial
Nocardioides sp. 43/50 16S ribosomal RNA gene, partial sequence
Streptococcus mitis bv2 16S ribosomal RNA gene, partial sequence
Acidobacteriaceae bacterium TAA43 16S ribosomal RNA gene, partial
Bacterium SM2-6 16S ribosomal RNA gene, complete sequence
Lactobacillus plantarum gene for 16S rRNA, partial sequence
Lactococcus lactis subsp. cremoris gene for 16S rRNA, partial
Uncultured alpha proteobacterium SBR6alpha8 16S ribosomal RNA gene,
Uncultured bacterium 16S rRNA gene, clone cD0261 1
Uncultured bacterium clone BREC 40 16S ribosomal RNA gene, partial
Uncultured Corynebacterium sp. clone ACTINO9C 16S ribosomal RNA
Agricultural soil bacterium clone SC-I-64, 16S rRNA gene (partial)
Lachnospira pectinoschiza 16S ribosomal RNA gene, partial sequence
Leuconostoc pseudomesenteroides DNA for 16S ribosomal RNA, strain
Uncultured bacterium clone D8A 5 16S ribosomal RNA gene, partial
uncultured Bacteroidetes bacterium partial 16S rRNA gene, clone
Uncultured Corynebacterium sp. clone ACTINO9B 16S ribosomal RNA
A.calcoaceticus gene for 16S rRNA
Acinetobacter sp. HPC270 16S ribosomal RNA gene, partial sequence
Alpha proteobacterium F820 16S ribosomal RNA gene, partial sequence
Bacterium PSD-1-3 16S ribosomal RNA gene, partial sequence
Beta proteobacterium Ellinl52 16S ribosomal RNA gene, partial
% DNA Similarity
97-100 (99)
97 - 100 (99)
99-100 (99)
98-99 (98)
99-100 (99)
98 - 100 (99)
99
99 - 100 (99)
99-100 (99)
99-100 (99)
98-99 (98)
99
96 - 100 (98)
95 - 96 (95)
94-95 (94)
99
96-99 (97)
98-99 (98)
98-99 (98)
99-100 (99)
99-100 (99)
96-99 (97)
96-98 (97)
99
91-95 (93)
99
99
99
95 - 96 (95)
96-99 (98)
100
98-99 (98)
96 - 97 (96)
98
97
# of Clones in Sample Designation
10003 10007 10028 10052 10035
_
_
_
14
_
-2
-
9
-
-_
-
_
6
_
__
5
_
_
_
4
-
-
4
4
_
_
_
-1
_
3
_
-
-
-
-
26
_
13
12
-
-
1
-
_
_
_
_
2
_
_
-
-
4
-
_
_
_
_
_
_
_
-
-
1
-
26
_
5
_
_
_
9
1
4
-
-
2
_
_
5
_
_
_
4
-
-
-
-
_
_
3
3
3
_
_
_
-
-
-
-
28
_
3
_
_
_
1
-
4
-
1
1
_
5
_
_
2
_
_
-
-
-
-
_
1
_
_
_
3
_
_
2
2
1
2
25
_
6
_
_
_
1
-
-
7
6
4
_
_
_
_
_
4
_
-
4
-
-
_
1
_
_
_
_
_
2
-
-
-
-
Total
79
26
16
14
13
12
11
10
8
8
7
7
6
5
5
5
4
4
4
4
4
4
4
4
3
3
3
3
3
3
2
2
2
2
2
Bacteria; Lineage
unclassified; environmental samples.
Proteobacteria; Betaproteobacteria; environmental
Bacteria environmental samples.
Actinobacteria; Actinobacteridae; Actinomycetales;
Bacteria environmental samples.
Proteobacteria; Betaproteobacteria; Burkholderiales;
Proteobacteria; environmental samples.
Firmicutes; Lactobacillales; Streptococcaceae;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Bacteria environmental samples.
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
unclassified; environmental samples.
Firmicutes; Bacillales; Staphylococcus.
Bacteroidetes.
Actinobacteria; Actinobacteridae; Actinomycetales;
Firmicutes; Lactobacillales; Streptococcaceae;
Acidobacteria; Acidobacteriales; Acidobacteriaceae.
Bacteria.
Firmicutes; Lactobacillales; Lactobacillaceae;
Firmicutes; Lactobacillales; Streptococcaceae;
Proteobacteria; Alphaproteobacteria; environmental
Bacteria; environmental samples.
Bacteria; environmental samples.
Actinobacteria; Actinobacteridae; Actinomycetales;
Bacteria; environmental samples.
Firmicutes; Clostridia; Clostridiales; Lachnospiraceae;
Firmicutes; Lactobacillales; Leuconostoc.
Bacteria; environmental samples.
Bacteroidetes; environmental samples.
Actinobacteria; Actinobacteridae; Actinomycetales;
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Bacteria; Proteobacteria; Alphaproteobacteria.
Bacteria.
Bacteria; Proteobacteria; Betaproteobacteria.
A-9
-------
Table A-l (continued). Identity of Clones Based on DNA Similarity of 16s Ribosomal DNA Sequeces Catalogued with GENEBANK
Blast Association
Brachybacterium sp. SKJH-25 16S ribosomal RNA gene, partial
Clostridium lactatifermentans 16S ribosomal RNA gene, partial
Corynebacterium pseudogenitalium partial 16S rRNA gene, strain
Curtobacterium sp. 1594 16S ribosomal RNA gene, partial sequence
D.pigrum 16S rRNA gene (partial)
Ketogulonogenium vulgarum strain 266- 13B small subunit ribosomal
Klebsiella pneumoniae isolate 521 16S ribosomal RNA gene, partial
Klebsiella sp. PN2 gene for 16S rRNA
Lactobacillus brevis 16S ribosomal RNA gene, partial sequence
Lactobacillus brevis gene for 16S rRNA, strain:B4101
Pseudomonas aeruginosa strain PD100 16S ribosomal RNA gene, partial
Pseudomonas saccharophila 16S ribosomal RNA gene, partial sequence
Pseudomonas sp. (strain BKME-9) 16S rRNA gene, partial
Pseudomonas sp. pDLOl 16S ribosomal RNA gene, partial sequence
Rhizosphere soil bacterium clone RSC-II-81, 16S rRNA gene (partial)
Sphingomonas sp. Alphal-2 16S ribosomal RNA gene, partial sequence
Sphingomonas sp. gene for 16S ribosomal RNA
Sphingomonas yunnanensis strain YIM 003 16S ribosomal RNA gene,
Streptococcus pneumoniae strain Kor 145 16S ribosomal RNA gene,
Uncultured bacterium clone DSBR-B020 16S ribosomal RNA gene,
Uncultured bacterium partial 16S rRNA gene, clone SHD-12
Uncultured earthworm cast bacterium clone c256 16S ribosomal RNA
Uncultured forest soil bacterium clone DUNssu554 16S ribosomal RNA
Uncultured soil bacterium clone G9-1338-5 small subunit ribosomal
A.calcoaceticus 16S rRNA gene (DSM30009)
Aeromicrobium erythreum 16S ribosomal RNA gene, partial sequence
Agricultural soil bacterium clone SC-I-92, 16S rRNA gene (partial)
Agrobacterium sp. NCPPB1650 gene for 16S ribosomal RNA, complete
Alpha proteobacterium 34619 16S ribosomal RNA gene, partial
Alpha proteobacterium PI GH2. 1 .D5 small subunit ribosomal RNA gene,
Anaerococcus prevotii strain CCUG 41932 16S ribosomal RNA gene,
Bacterium RBA-1-13 16S ribosomal RNA gene, partial sequence
Bacterium RBA-1-31 16S ribosomal RNA gene, partial sequence
Bacterium RSD-1-9 16S ribosomal RNA gene, partial sequence
Beta proteobacterium 9c-3 16S ribosomal RNA gene, partial sequence
% DNA Similarity
98-99 (98)
92-93 (92)
99 - 100 (99)
99
98
96-99 (97)
99 - 100 (99)
99
99
99
99
99
99
99
90-95 (92)
96
95 - 96 (95)
99
99
99
98 - 99 (98)
96-98 (97)
99
98-99 (98)
99
96
96
99
99
96
98
97
96
99
100
# of Clones in Sample Designation
10003
_
_
2
-
2
-
-
-
_
_
_
_
_
_
_
_
-
1
2
-
_
1
_
2
1
_
_
-
-
1
-
-
_
1
1
10007
_
_
_
2
-
-
-
-
_
_
2
2
_
2
1
_
-
-
-
-
_
1
_
_
_
_
1
-
-
-
-
-
1
_
-
10028
2
_
_
-
-
2
-
-
_
_
_
_
_
_
_
_
-
-
-
-
_
_
_
_
_
1
_
-
1
-
-
-
_
_
-
10052
_
2
_
-
-
-
-
-
_
_
_
_
_
_
1
_
2
1
-
-
_
_
_
_
_
_
_
1
_
-
1
1
_
_
-
10035
_
_
_
-
-
-
2
2
2
2
_
_
2
_
_
2
-
-
-
2
2
_
2
_
_
_
_
-
_
-
-
-
_
_
-
Total
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
Bacteria; Lineage
Actinobacteria; Actinobacteridae; Actinomycetales;
Firmicutes; Clostridia; Clostridiales; Clostridiaceae;
Actinobacteria; Actinobacteridae; Actinomycetales;
Actinobacteria; Actinobacteridae; Actinomycetales;
Firmicutes; Lactobacillales; Carnobacteriaceae;
Proteobacteria; Alphaproteobacteria; Rhodobacterales;
Proteobacteria; Gammaproteobacteria; Enterobacteriales;
Proteobacteria; Gammaproteobacteria; Enterobacteriales;
Firmicutes; Lactobacillales; Lactobacillaceae;
Firmicutes; Lactobacillales; Lactobacillaceae;
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Proteobacteria; Betaproteobacteria; Burkholderiales;
Bacteria; Proteobacteria.
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Bacteria; environmental samples.
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Firmicutes; Lactobacillales; Streptococcaceae;
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Actinobacteria; Actinobacteridae; Actinomycetales;
Bacteria; environmental samples.
Proteobacteria; Alphaproteobacteria; Rhizobiales;
Bacteria; Proteobacteria; Alphaproteobacteria.
Bacteria; Proteobacteria; Alphaproteobacteria.
Bacteria; Firmicutes; Clostridia; Clostridiales;
Bacteria.
Bacteria.
Bacteria.
Bacteria; Proteobacteria; Betaproteobacteria.
A-10
-------
Table A-l (continued). Identity of Clones Based on DNA Similarity of 16s Ribosomal DNA Sequeces Catalogued with GENEBANK
Blast Association
Brevundimonas bacteroides DNA for 16S ribosomal RNA, strain CB7
Brevundimonas diminuta 16S ribosomal RNA gene, partial sequence
Butyrate-producing bacterium SRI/1 16S ribosomal RNA gene, partial
Caulobacter sp. DNA for 16S ribosomal RNA, strain FWC38
CDC Group DF-3 16S LMG 11519 ribosomal RNA gene, partial sequence
Clostridium sp. ArC6 16S ribosomal RNA gene, partial sequence
Comamonas testosteroni gene for 16S rRNA, partial sequence
Corynebacterium accolens partial 16S rRNA gene, strain CIP104783T
Cytophagales str. MBIC4147 gene for 16S rRNA, partial sequence
Diaphorobacter sp. PCA039 16S ribosomal RNA gene, partial sequence
Earthworm burrow bacterium B33D1 16S ribosomal RNA gene, partial
Flexibacter cf. sancti 16S ribosomal RNA gene, partial sequence
Glacial ice bacterium G200-C18 16S ribosomal RNA gene, partial
Klebsiella pneumoniae strain 542 16S ribosomal RNA gene, partial
Lactobacillus brevis gene for 16S rRNA, strain:NRIC 1684
Lactobacillus parabuchneri gene for 16S ribosomal RNA, partial
Lactobacillus sp. oral clone CX036 16S ribosomal RNA gene, partial
Leuconostoc citreum 16S ribosomal RNA gene, partial sequence
Leuconostoc pseudomesenteroides 16S ribosomal RNA gene, partial
Leuconostoc pseudomesenteroides gene for 16S rRNA, partial
Mesorhizobium sp. Ellinl89 16S ribosomal RNA gene, partial sequence
Mesorhizobium sp. M01 gene for 16S rRNA, partial sequence
Methylobacterium sp. RKT-5 16S ribosomal RNA gene, partial sequence
Mycobacterium sp. Ellinl 18 16S ribosomal RNA gene, partial sequence
Nocardioides OS4 16S rRNA
Nocardioides sp. MWH-CaK6 partial 16S rRNA gene, isolate MWH-CaK6
Pantoea ananatis partial 16S rRNA gene, strain 0201935
Paracraurococcus ruber partial 16S rRNA gene, isolate CP2C
Peptostreptococcaceae bacterium 19gly3 16S ribosomal RNA gene,
Potato plant root bacterium clone RC-III-8, 16S rRNA gene (partial)
Pseudomonas aeruginosa 16S ribosomal RNA gene, partial sequence
Pseudomonas sp. 3B 8 16S ribosomal RNA gene, partial sequence
Pseudomonas sp. MFY69 16S ribosomal RNA gene, partial sequence
Pseudomonas sp. TB3-6-I 16S ribosomal RNA gene, partial sequence
Pseudomonas veronii gene for 16S rRNA, strain: IN A06
% DNA Similarity
100
98
99
97
94
94
98
99
95
100
95
91
98
99
99
99
99
99
99
99
98
98
99
98
98
97
99
96
94
95
99
99
99
100
99
# of Clones in Sample Designation
10003
_
_
_
-
-
-
-
1
_
_
1
_
1
_
_
_
-
-
-
-
1
_
_
_
_
_
_
-
1
-
-
1
1
_
-
10007
_
1
_
-
-
-
-
-
_
_
_
1
_
_
_
_
-
-
-
-
_
_
_
_
1
1
1
-
-
-
1
-
_
_
-
10028
_
_
_
-
1
-
1
-
_
1
_
_
_
_
_
1
-
1
1
1
_
_
_
_
_
_
_
1
-
-
-
-
_
1
1
10052
1
_
1
1
-
1
-
-
1
_
_
_
_
_
_
-
-
-
-
-
_
1
_
1
_
_
_
-
-
1
-
-
_
_
-
10035
_
_
_
-
-
-
-
-
_
_
_
_
_
1
1
_
1
-
-
-
_
_
1
_
_
_
_
-
-
-
-
-
_
_
-
Total
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Bacteria; Lineage
Proteobacteria; Alphaproteobacteria; Caulobacterales;
Proteobacteria; Alphaproteobacteria; Caulobacterales;
Bacteria; Firmicutes; Clostridia; Clostridiales.
Proteobacteria; Alphaproteobacteria; Caulobacterales;
Bacteroidetes; Bacteroidetes (class); Bacteroidales;
Firmicutes; Clostridia; Clostridiales; Clostridiaceae;
Proteobacteria; Betaproteobacteria; Burkholderiales;
Actinobacteria; Actinobacteridae; Actinomycetales;
Bacteria; Bacteroidetes.
Proteobacteria; Betaproteobacteria; Burkholderiales;
Actinobacteria; Rubrobacteridae; Rubrobacterales;
Bacteroidetes; Sphingobacteria; Sphingobacteriales;
Bacteria.
Proteobacteria; Gammaproteobacteria; Enterobacteriales;
Firmicutes Lactobacillales; Lactobacillaceae;
Firmicutes Lactobacillales; Lactobacillaceae;
Firmicutes Lactobacillales; Lactobacillaceae;
Firmicutes Lactobacillales; Leuconostoc.
Firmicutes Lactobacillales; Leuconostoc.
Firmicutes Lactobacillales; Leuconostoc.
Proteobacteria; Alphaproteobacteria; Rhizobiales;
Proteobacteria; Alphaproteobacteria; Rhizobiales;
Proteobacteria; Alphaproteobacteria; Rhizobiales;
Actinobacteria; Actinobacteridae; Actinomycetales;
Actinobacteria; Actinobacteridae; Actinomycetales;
Actinobacteria; Actinobacteridae; Actinomycetales;
Proteobacteria; Gammaproteobacteria; Enterobacteriales;
Proteobacteria; Gammaproteobacteria; Enterobacteriales;
Firmicutes; Lactobacillales; Leuconostoc.
Bacteria; environmental samples.
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
Proteobacteria; Gammaproteobacteria; Pseudomonadales;
A-ll
-------
Table A-l (continued). Identity of Clones Based on DNA Similarity of 16s Ribosomal DNA Sequeces Catalogued with GENEBANK
Blast Association
Rhizosphere soil bacterium clone RSC-II-92, 16S rRNA gene (partial)
Roseburia faecalis strain M6/1 16S ribosomal RNA gene, partial
S.trueperi 16S rRNA gene
Sejongia jeonii strain AT1047 16S ribosomal RNA gene, partial
Sphingomonas oligophenolica gene for 16S rRNA, partial sequence
Sphingomonas sp. 44/40 16S ribosomal RNA gene, partial sequence
Sphingomonas sp. ATCC 53159 16S ribosomal RNA gene, partial
Sphingomonas sp. CJ-5 partial 16S rRNA gene, isolate CJ-5
Sphingomonas sp. KIN163 16S ribosomal RNA gene, partial sequence
Sphingomonas sp. KIN169 16S ribosomal RNA gene, partial sequence
Sphingomonas sp. SAFR-028 16S ribosomal RNA gene, partial sequence
Sphingomonas sp. SRS2 16S rRNA gene, strain SRS2
Sphingomonas sp. YT gene for 16S rRNA
Spirosoma sp. 2.8 partial 16S rRNA gene, strain 2.8
Staphylococcus epidermidis AB1 11112 16S ribosomal RNA gene, partial
Uncultured alpha proteobacterium 16S rRNA gene, clone BIriil3
Uncultured alpha proteobacterium clone ALPHA5C 16S ribosomal RNA
Uncultured alpha proteobacterium clone S1-10-CL6 16S ribosomal RNA
Uncultured alpha proteobacterium SBR8alpha5 16S ribosomal RNA gene,
Uncultured bacterium clone 1974b-28 16S ribosomal RNA gene, partial
Uncultured bacterium clone 33-FL34B99 16S ribosomal RNA gene,
Uncultured bacterium clone 76 16S ribosomal RNA gene, partial
Uncultured bacterium clone A4 16S ribosomal RNA gene, partial
Uncultured bacterium clone ABLCf6 16S ribosomal RNA gene, partial
Uncultured bacterium clone B8 16S small subunit ribosomal RNA gene,
Uncultured bacterium clone D3 16S ribosomal RNA gene, partial
Uncultured bacterium clone E9 16S small subunit ribosomal RNA gene,
Uncultured bacterium clone FB33-24 16S ribosomal RNA gene, partial
Uncultured bacterium clone FBP249 16S ribosomal RNA gene, partial
Uncultured bacterium clone LJ3 16S ribosomal RNA gene, partial
Uncultured bacterium clone LO13. 1 1 16S ribosomal RNA gene, partial
Uncultured bacterium clone mek62a01 16S ribosomal RNA gene, partial
Uncultured bacterium clone O-CF-31 16S ribosomal RNA gene, partial
Uncultured bacterium clone REC6M 59 16S ribosomal RNA gene, partial
Uncultured bacterium clone S1-1-CL4 16S ribosomal RNA gene, partial
% DNA Similarity
97
99
96
97
98
98
98
97
97
96
97
96
97
93
99
97
93
99
98
96
98
98
97
99
99
98
97
98
90
96
98
96
96
99
94
# of Clones in Sample Designation
10003
_
_
1
-
-
-
-
-
_
_
_
_
_
1
1
1
-
-
1
1
_
_
_
_
_
_
_
1
-
-
-
-
_
_
1
10007
_
_
_
-
-
-
-
-
_
_
_
1
_
_
_
-
-
-
-
-
_
_
_
_
_
1
1
-
-
-
1
-
_
_
-
10028
_
1
_
-
-
-
-
-
_
_
_
_
_
_
_
_
-
-
-
-
_
_
_
1
_
_
_
-
-
-
-
1
_
_
-
10052
1
_
_
-
1
-
_
1
_
_
_
1
1
-
-
1
_
1
_
1
_
_
-
1
1
-
-
1
1
-
10035
_
_
_
1
-
1
-
-
_
_
_
_
_
_
_
_
-
-
-
-
_
1
_
_
_
_
_
-
-
-
-
-
_
_
-
Total
Bacteria; Lineage
Bacteria; environmental samples.
Firmicutes; Clostridia; Clostridiales; Lachnospiraceae;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Bacteria; Bacteroidetes; Flavobacteria; Flavobacteriales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Proteobacteria; Alphaproteobacteria; Sphingomonadales;
Firmicutes; Bacillales; Staphylococcus.
Proteobacteria; Alphaproteobacteria; environmental
Proteobacteria; Alphaproteobacteria; environmental
Proteobacteria; Alphaproteobacteria; environmental
Proteobacteria; Alphaproteobacteria; environmental
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
A-12
-------
Table A-l (continued). Identity of Clones Based on DNA Similarity of 16s Ribosomal DNA Sequeces Catalogued with GENEBANK
Blast Association
Uncultured bacterium clone synarJM02 16S ribosomal RNA gene,
Uncultured bacterium clone ZEBRA 19 16S ribosomal RNA gene, partial
Uncultured bacterium partial 16S rRNA gene, clone MCS2/83
Uncultured bacterium partial 16S rRNA gene, clone StaO-39
Uncultured Bacteroides sp. clone TNHul-10 16S ribosomal RNA gene,
Uncultured beta proteobacterium clone FTL217 16S ribosomal RNA
Uncultured eubacterium 16S rRNA gene, clone LKB47
uncultured eubacterium WD208 partial 16S rRNA gene, clone WD208
uncultured eubacterium WD286 partial 16S rRNA gene, clone WD286
Uncultured forest soil bacterium clone DUNssu642 16S ribosomal RNA
Uncultured organism clone M8907G12 small subunit ribosomal RNA
Uncultured soil bacterium clone 359 small subunit ribosomal RNA
Uncultured soil bacterium clone 5-1 small subunit ribosomal RNA
Uncultured soil bacterium clone 749-2 16S ribosomal RNA gene,
Uncultured soil bacterium clone Tcl24-Cl 1 16S ribosomal RNA gene,
Uncultured yard-trimming-compost bacterium clone S-19 16S ribosomal
Unidentified bacterium clone W4-B50 16S ribosomal RNA gene, partial
Zoogloea sp. AI-20 16S ribosomal RNA gene, partial sequence
% DNA Similarity
96
94
94
99
92
99
99
95
98
99
95
95
94
98
95
94
95
98
# of Clones in Sample Designation
10003
_
_
_
1
-
1
-
-
_
_
_
_
_
1
_
_
-
-
81
10007
_
_
1
-
-
-
-
1
1
_
_
1
1
_
1
_
-
-
90
10028
_
1
_
-
-
-
-
-
_
1
_
_
_
_
_
_
-
-
86
10052
1
_
_
-
-
-
1
-
_
_
1
_
_
_
_
1
-
1
95
10035
_
_
_
-
1
-
-
-
_
_
_
_
_
_
_
_
1
-
87
Total
1
439
Bacteria; Lineage
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteroidetes; Bacteroidetes (class); Bacteroidales;
Proteobacteria; Betaproteobacteria; environmental
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
unclassified; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Bacteria; environmental samples.
Proteobacteria; Betaproteobacteria; Rhodocyclales;
A-13
-------
Total
• Bacteria
DActinobacteria
QBacteroidetes
nBacteria; environmental
samples
DFirmicutes
n = 439
Figure A-l.
IOO52
• Bacteria
DActinobacteria
DBacteroidetes
DBacteria; environmental
samples
DFirmicutes
n =95
Figure A-2.
A-14
-------
IOOO3
Figure A-3.
10007
n=81
• Bacteria
DActinobacteria
Q Bacteroidetes
a Bacteria; environmental
samples
• Bacteria
DActinobacteria
a Bacteroidetes
n -90
Figure A-4.
A-15
-------
10028
28
/\
// 12\
DActinobacteria
DBacteroidetes
D Bacteria; environmental
samples
DFirmicutes
oProteobacteria;
Alphaproteobacteria
11 = 36
Figure A-5.
IOO035
• Bacteria
QBacteroidetes
nBacteria; environmental
samples
DFirmicutes
n = 87
Figure A-6.
A-16
-------
APPENDIX B
Parallel Sampling Approaches and Analysis oflmpinger Samples Collected During the NC Biosolids
Land Application Study -Dr. Ian Pepper, University of Arizona
-------
North Carolina Field Aerosol Sampling
Dates: 9-27-04 - 9-30-04
Objectives:
1. To collect aerosol samples downwind of a land application of biosolids site located in Salisbury, NC and
analyze for microbial content.
Results:
Aerosol samples were collected in multiples of 6 from approximately 2 m downwind from the site perimeter,
constituting a sampling array. This was conducted 3 times during biosolids land application. In addition a set of 6
samples were collected 2 m from the site perimeter prior to biosolids application, termed background samples (BG).
Samples were analyzed for the presence of heterotrophic plate count bacteria (HPC), total conforms, Escherichia
coli, Clostridium perfringens, and coliphage. In addition, a 5 ml aliquot from each sample was set aside for
pathogenic virus analysis. To increase sensitivity, the 5 ml from the 6 samples collected during each separate array
were combined and concentrated using Centriprep 50 concentrators to generate 1 sample from the array of 6. The
final volume (< 1.0 ml) was used in enteric virus viability assays (cell culture - BGMK cells, incubation 14 d, two
passages) and used for reverse transcriptase polymerase chain reaction (RTPCR) for the detection of enterovirus,
hepatitis A virus, and norovirus ribonucleic acid (RNA). Finally in addition to aerosol samples a sub-sample of
Class B biosolids was analyzed for the presence of the previously mentioned microbial indicators and pathogens.
Overall, aerosol samples were negative for the presence of microbial indicators and microbial pathogens.
Aerosolized heterotrophic plate count bacteria densities were approximately 7.27 x 103 during biosolids land
application, while during background sample collection, HPC concentrations were approximately 3.18 x 103 (Table
B-l). HPC concentrations during background and downwind sample collections were similar, possibly due to the
soil moisture and ambient relative humidity levels, which led to overall low levels of aerosolized HPC bacteria from
both biosolids and soil. In addition, the location of the sample placement was such that the biosolids applicator
began application approximately 10 m upwind of the sampler locations, even though the sampler location was 2 m
from the edge of the field perimeter. In addition biosolids application proceeded from right to left in a circular
fashion relative to the sampler location. Upon proceeding to the opposite end of the circle (180 °) the samplers were
no longer in line of sight of the biosolids application, due to the presence of an uneven field in which the center of
the field was approximately 2 - 4 m above the edges of the field, forming a convex field. As such exposure to
biosolids only took place for the brief instant that the applicator was in range of the samplers, thus limiting aerosol
concentrations.
B-l
-------
Table B-l. Aerosol microbial concentrations detected during September 29 and September 30, 2004.
* NA - Refers to no data collected
RTPCR Cell Culture
Sample
1
2
3
4
5
6
Avg
7
8
9
10
11
12
Avg
13
14
15
16
17
18
Avg
19
20
21
22
23
24
Avg
Sample
Placement
BG
BG
BG
BG
BG
BG
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
DW
Ambient Climate Conditions
Temp RH WS HPC
C % ms1
NA NA NA 5.60E+03
1.87E+03
3.47E+03
1.33E+03
4.27E+03
2.40E+03
3.16E+03
20.2 88.0 1.0 6.13E+03
5.60E+03
8.00E+03
9.33E+03
9.33E+03
8.27E+03
7.78E+03
22.0 52.0 0.8 1.22E+04
5.65E+03
8.78E+03
4.71 E+03
7.22E+03
5.02E+03
7.27E+03
29.4 39.0 0.0 7.82E+03
4.27E+03
7.47E+03
4.98E+03
3.56E+03
6.40E+03
5.75E+03
Total Coliform E. coli
C. peifiingens
Virus Presence/Absence
Coliphage Enterovirus HAV Norovirus Enterovirus HAV
Norovirus
CPU, MPN, PFU m-3
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 Neg Neg Neg Neg Neg
0
0
0
0
0
0 Neg Neg Neg Neg Neg
0
0
0
0
0
0 Neg Neg Neg Neg Neg
0
0
0
0
0
0 Neg Neg Neg Neg Neg
Neg
Neg
Neg
Neg
Neg - Refers to negative results, none detected
CPU - Colony forming unit
MPN - Most probable number
PFU - Plaque forming unit
Temp - Temperature
RH - Relative Humidity
WS - Windspeed
BG - Background aerosol sample
DW - Downwind aerosol sample
B-2
-------
APPENDIX C
Endotoxin Sampling During a Post-Spring Cutting Event at the NC Biosolids Land Application Study
Site-Dr. Edwin Barth, EPA/NRMRL
-------
RESEARCH
on
E. EARTH*, R. HERRMANN, T. DAHLING, R. BRENNER, S. WRIGHT and P. CLARK
National Risk Management Research Laboratory, Office of Research and Development, USEPA, Cincinnati, OH
ABSTRACT: Public health concerns have been expressed regarding inhalation expo-
sure associated with the application of biosolids on cropland, which is due to the poten-
tial aerosolization of microorganisms, cell wall products, volatile chemicals, and nui-
sance odors. Endotoxin is a component of the cell walls of Gram-negative bacteria and is
likely present in many biosolids. The application of biosolids to cropland may result in an
immediate exposure or a delayed exposure to these microbial agents, such as when the
crops are harvested. Upwind and downwind airborne concentrations of endotoxin were
compared among and within two adjacent established hayfields, one with and one with-
out previously applied biosolids, during grass raking and bailing activities. The mean
downwind concentration of airborne endotoxin was significantly higher than the mean
upwind concentration at the site where biosolids had been previously applied. The mean
downwind concentration of endotoxin was not significantly different than the mean up-
wind concentration at the control field where biosolids had not previously been applied.
When comparing the mean concentrations of airborne endotoxin among the sites, sig-
nificant main effects were noticed for wind direction and field type, and an interactive ef-
fect was noticed for direction and field type. It is not known if the increased mean con-
centration of endotoxin associated with the downwind air samples at the applied
biosolids field were due to the residual biosolids that were previously applied or due to
endotoxin associated with plant material. The results apply to this investigation only
since there was no treatment replication of each type of field. The downwind endotoxin
concentrations observed during the raking and bailing activities were lower than various
health effects criteria that have been recommended for airborne endotoxin.
INTRODUCTION
ENDOTOXIN is a term associated with the toxic
characteristics of the outer membrane of Gram-
negative bacteria [1], specifically the fragments of
the Gram-negative cell wall that contain
lipopolysaccharides [2]. Lipopolysaccharides are es-
sential forthe physical organization and function of the
outer membrane, and thus for bacterial growth and mul-
tiplication [3]. Endotoxin consists of a family of mole-
cules called lipopolysaccharide (LPS). The LPS con-
tains a lipid region (lipid A), and a long covalcntly
linked heteropolysaccharide. The polysaccharide por-
tion is divided into a core portion and the O-specific
chain [2,4]. Endotoxin is present in the environment as
whole cells, large membrane fragments, or
macromolecular aggregates of about one million
Daltons [5].
* Author to whom correspondence should be addressed.
E-mail: bartli.edfiJlepa.gov
The multiple biological activities associated with
endotoxin reside in the lipid A component [6.7]. The bi-
ological activity of endotoxin is not dependent on bac-
terial viability [8J. Human inhalation studies or worker
exposure cases involving endotoxin have shown ad-
verse physiological and symptomatic respiratory re-
sponses [9,10]. Inhalation of the components of
bioaerosols may result in several allergenic-type reac-
tions or lung diseases, such as bronchitis, reactive air-
way disease, organic dust toxic syndrome (ODTS). and
hypersensitivity pneumonitis (HP) [11]. There is de-
bate whether early childhood exposure to endotoxin is
positively or negatively associated with the onset or se-
verity of asthma [12,13].
Endotoxin is released into the environment after bac-
terial cell lysis or during active cell growth [14]. Since
bacteria, fungi, and endotoxin may be associated with
biosolids, there is an inhalation concern with these
bioaerosol components both during and after biosolids
land application. Bacteria in biosolids may survive for
long periods of time, depending upon the method of
Journal of Residuals Science & Technology, Vol. 6, No, 2—April 2009
1544-8053/09/02061-05
© 2009 DEStech Publications, Inc.
61
-------
62
E. EARTH, R. HERRMANN, T. DAHLING, R. BRENNER, S. WRIGHT and R CLARK
management and environmental conditions [15,16,17].
Elevated levels (above background) of endotoxin were
associated with sites receiving biosolids application
with a mechanical slinger [18]. There is no published
study regarding airborne endotoxin concentrations dur-
ing subsequent crop management activities.
The primary objective of this study was to determine
if a statistically significant difference existed between
the mean upwind and mean downwind airborne con-
centrations of endotoxin, during grass raking and bail-
ing activities among and within two proximal hayfields
(grass), with one of the sites having been previously
treated with biosolids as a soil amendment.
MATERIALS AND METHODS
The sampling approach for this study involved two
separate sampling events for aerosolized endotoxin. One
sampling event occurred at an established hayfield that
did not receive any biosolids application (control field).
The other sampling event occurred at an established
hayfield that had received surface applied biosolids (ap-
plication field). Anaerobically digested biosolids were
similar to Class B biosolids, but were pre-treated with a
limited amount of lime to ensure a viable microbial pop-
ulation was present for monitoring purposes. The
biosolids contained approximately 109 colony forming
units per gram-dry weight (CPU gdw"1) total coliforms
with a solids content of 22%. The biosolids were applied
to the application field within a 100 m diameter area, ap-
proximately nine months earlier, and were applied to the
surface via a hopper truck with a mechanical slinger at a
rate of 10 dry tons per acre.
Each of the two sampling events occurred during sep-
arate grass raking and bailing activities (3 dry-days af-
ter grass cutting) for approximately 60 minutes. For
each sampling event, five upwind and five downwind
stations (containing two endotoxin sampling devices
each) were placed along parallel lines, perpendicular to
the prevailing wind direction, as shown in Figure 1. The
exact orientation of the zones was determined based
upon the weather station wind direction data collected
by a Davis Instruments Weather Monitor II weather sta-
tion (Hayward, California). The samplers were oriented
around a 40 m X 40 m monitoring area (within a 40 m X
80 m area that had been cut). For the application site, the
monitoring area was within the original 100m diameter
biosolids application area. Five upwind and five down-
wind sampling stations, each containing two endotoxin
samplers, were located 10 m apart from each other
t
LU
O
N
WIND DIRECTION
- 20m >
Biosolids CuJttng/BaflinoAfea
<^-
- 4Qjrf-
^ 20m
LU
Z
O
N
1,2 3,4 5,6 7,8 9,10
KEY: • Individual sampling station (containing two endotoxin samplers)
• Weather Station
Figure 1. Sampling station orientation for both control and applica-
tion fields.
within the respective zone. For both the upwind and
downwind zones, the distance from the samplers to the
corresponding external edge of the biosolids raking and
bailing area was 10m. The bailing machine was oper-
ated parallel to the sampler lines. After each pass, the
raking and bailing equipment (two distinct farm ma-
chine vehicles in series) temporarily left the sampling
zone, turned around, and performed another pass in the
opposite direction (endotoxin samplers continued to
operate during the turn-around). The samples were col-
lected near the personal breathing zone (PBZ) height at
1.5 m above the ground surface by mounting the
endotoxin samplers on portable tripods. The weather
station was placed 20m upwind and on the mid-line of
the upwind sampling line.
The control field (of the same size) contained the
same grass cover as the biosolids field. It was located
approximately 400-500 m from the biosolids field. To
reduce field variation, one initial fertilizer application
was applied to the control field within three months of
the demonstration, since the application site received
nitrogen loading from the applied biosolids.
-------
Evaluation of Airborne Endotoxin Concentrations Associated with Management of a Crop Grown on Applied Biosolids 63
Prolonged wind direction changes of more than 45
degrees, for longer than two minutes, or any strong
wind gust (greater than 15 MPH for at least two min-
utes), or any precipitation event would immediately
halt sample collection activities until they subsided.
The bailing equipment was instructed to shut-down at
this time as well. If the sampling was shut down for
more than 30 continuous minutes, the sampling event
would have been considered to be invalid.
Various sampling methods for collecting airborne
endotoxin have been used in occupational settings [9,
19]. There are several variables which will possibly in-
fluence the endotoxin concentration collected in air
samples, such as filter type, extractant fluid, and sample
preservation time [20]. The sampling method used in-
volved the use of commercially available 37-mm cas-
sette filter (0.45 pm polycarbonate filters) assemblies
(Acrotech Laboratories, Phoenix, AZ), which were
manufactured to be endotoxin-free. Two cassettes were
mounted to each tripod for each of the sampling sta-
tions. The cassettes were separately attached to a vac-
uum pump (GAST Manufacturing, Benton Harbor,
MI). The desired collection flow rate for the cassette as-
semblies was 4.0 L mkr1. Each air collection pump was
calibrated pre- and post-sampling in the field immedi-
ately before and after each sampling event with a pri-
mary standard calibrator (Gillian Model 2 primary stan-
dard pump calibrator). Any pre- and post-flow rates
that differed by more than 10% were not used in subse-
quent data calculations.
After each sampling event, the cassettes were
capped, placed in plastic bags, and then placed into an
iced cooler for transport back to the analytical labora-
tory within 8 hours. After arrival at the laboratory, cas-
settes were opened, cassette filters were ascptically re-
moved, and then the filters were placed into a pyrogen
free 50 ml centrifuge tube containing 6 ml of pyrogen
free water. The 50 ml centrifuge tube was capped and
shaken on a mechanical shaker for one hour to complete
the extraction procedure for endotoxin.
After the endotoxin was extracted, it was assaved us-
ing the Kinetic-QCL method [21]. The field samples
were mixed with a substrate, placed in the kinetic
reader, and monitored (time) for the appearance of a
yellow color. A standard dilution curve ranging from
0.005-50 Endotoxin Units (EU) ml"1, using a control
standard endotoxin (CSE), was prepared during the as-
say. A positive product control spike (PPC) for each di-
lution was incorporated into the assay to determine re-
covery (-50%-200%). The solutions were delivered
onto a 96-wcll microplatc, which was then inserted into
the BioWhittaker Kinetic QCL reader.
Descriptive and inferential statistical methods were
used on the collected data. Two approaches were consid-
ered for the analysis. In the first approach, the sites were
considered independent of each other. Inferential statis-
tics included the parametric student t-test assuming nor-
mality of the data distribution. The null hypothesis was
that there is no statistically significant difference in air-
borne endotoxin concentration between the mean up-
wind zone concentration (10 samples) and mean down-
wind zone concentration (10 samples) of endotoxin, for
each sampling event. The second approach analyzed the
data as a completely randomized design with a two-way-
treatment structure (2 X 2), wind direction (upwind,
downwind) and field type (control, biosolids) using
ANOVA (PROC GLM Procedure). The /-test and
ANOVA analyses were performed using SAS [22].
RESULTS AND DISCUSSION
The airborne concentrations of endotoxin at each
identified station, and the mean concentrations for each
trial, are provided in Table 1. The values in Table 1
were adjusted for concentrations of endotoxin detected
in the field and blank samples. Interactive box plots
(field type by wind direction) of the endotoxin data is
presented in Figure 2, showing a potential outlier sam-
ple value for each trial. None of the environmental con-
ditions that would have invalidated the results were en-
countered, so the sample collection effort was
considered valid.
Table 1, Concentration of Airborne Endotoxin at Individual Sampling Stations (EU m~3).
Field
Control
Control
Application
Application
Zone
Upwind
Downwind
Upwind
Downwind
1
11.3
11.6
7,7
24.3
2
6.1
38.1*
2,9
27.1
3
4.8
14.9**
2.0
38,2**
4
3.1
6.3
2.5
28.9
5
8.6
9.7
0.3
43.5
6
9.5
15.6
0.4
44.8
7
8.7
16.1
0.0
42.7
8
8.7**
7.4
2,5**
34.2
9
7.9
14.2
1.5
39.9
10
10,6
11.9
2.0
38.3
Mean
7.8
14.5
2.1
36.0
*Represent potential outlier data.
"Stations greater than 10% difference in pre/post flow calibration.
-------
64
E. EARTH, R. HERRMANN, T. DAHLING, R. BRENNER, S. WRIGHT and P. CLARK
3-
LU
Bioslids Down Biosolids Up Control Down Control Up
Trial Location
Figure 2. Boxplots of upwind and downwind endotoxin concentra-
tions for control and biosolids application trials.
For relative comparisons, the mean airborne
endotoxin concentrations observed for each trial were
greater than some of the other published background
range levels detected in outside environments that
vary from 0.0005-0.74 EU m~3 in outdoor environ-
ments in Germany to 2.0-3.8 EU m~3 for outdoor sites
in the United States [23,24,18]. The levels observed
were lower than the mean concentration of 114 EUnr3
observed within 10m downwind of a limited number
of biosolids application sites in the southwestern
United States, but in the range of the mean concentra-
tion of 6 EU m~3 observed further downwind on these
sites.
The mean airborne endotoxin concentrations ob-
served for each trial were less than published occupa-
tion exposure levels, and less than the large range of
other published human exposure criteria for endotoxin.
Inhalation of endotoxin in concentrations as low as
4-15 ng m~3 (40-150 EU m~3) has been associated with
acute and chronic airway inflammation and lung func-
tion decrements [23]. The International Committee on
Occupational Health (ICOH) Committee on Organic
Dust observed toxic pneumonitis at endotoxin levels of
200 ng m~3 (2000 EU m~3), systemic reactions at 100 ng
m~3 (1000 EU m~3), and airway inflammation at 10 ng
m~3 (100 EU m~3) [25]. Experimental studies of human
exposure to cotton dust and field studies suggest an
endotoxin threshold for acute airflow obstruction in the
range of 45 to 330 EU nr3 [26]. The ACGIHhas recom-
mended an indoor endotoxin concentration less than 30
to 100 times the ambient (outdoor) concentration [26].
The Dutch Expert Committee on Occupational Stan-
dards of the National Health Council has proposed a
health-based recommended limit value of 4.5 ng m~3
(0.45 EU m~3) over an eight-hour exposure period [27].
The NIOSH Recommended Exposure Limit (REL) for
airborne metalworking particulates, that may contain
endotoxin from recirculated fluids, is limited to 0.4 mg
m~3 for thoracic particle mass (0.5 mg m~3 total particu-
late mass) [28].
The mean concentration of downwind airborne
endotoxin samples was significantly higher than up-
wind concentration mean during grass raking and bail-
ing operations within the application trial field where
biosolids had previously been applied. The mean up-
wind and downwind concentrations were not statisti-
cally significantly different within the control trial field
where biosolids had not previously been applied. It was
not determined if the increased concentrations of
endotoxin in the downwind air samples at the biosolids
application field were due to the biosolids residual or
due to plant material grown on the field. Even though
the control field did receive fertilizer, the density of
plant material appeared to be visually higher on the
biosolids application site, though any type of measure-
ment for this property was not performed.
The mean concentration of the downwind air samples
at the application site was statistically different than the
other three means (upwind control, downwind control,
upwind application). However, the downwind control
trial mean was higher than the upwind means for both
trials (before multiple comparison adjustment) and
higher than the upwind application trial site mean even
after adjustment (via Scheffe's approach). ANOVA
analysis between the two sites (four groups) indicated
that there were statistically significant main effects
among the sites in wind direction and field type; there
was also a statistically significant interaction effect
with wind direction and field type. The residuals from
the ANOVA are not normal, but also are not skewed, so
a transformation (such as the logarithmic) was not use-
ful to normalize the distributions. However, the box
plots did identify possible influential outliers. Sus-
pected outliers made no difference in interpretation,
and after removing these outliers, the relationship be-
tween the endotoxin concentration and the type of field
and sampler location is strengthened. After removal of
the outliers, the distribution of the residuals becomes
normal, indicating that the two outliers contributed to
the non-normality of the original data distribution but
were not influential. The results apply to this investiga-
tion only since there was no treatment replication of
each type of field.
-------
Evaluation of Airborne Endotoxin Concentrations Associated with Management of a Crop Grown on Applied Biosolids 65
CONCLUSIONS
The mean downwind concentration of airborne
endotoxin associated with raking and bailing of grass
was significantly higher than the mean upwind concen-
tration at a specific hayfield site where biosolids had
been applied approximately nine months prior to the
sampling event. It was not determined if the increased
mean concentration of endotoxin in the downwind air
samples at the biosolids application field were due to
biosolids residuals or due to plant material grown on the
field. In contrast, the mean downwind concentration of
airborne endotoxin for the same activities at a close
proximity site (control site) that did not receive
biosolids application was not significantly higher than
the mean upwind concentration.
ACKNOWLEDGEMENTS
The cooperation of several individuals from the Ag-
ricultural Research Station in Salisbury. NC is greatly
appreciated. Larry Wctzel of the USEPA participated
in the sample collection effort.
REFERENCES
1. Rietschel, E. and H. Brade. "Bacterial Endotoxin", Sci. Amer. Vol.
267, No. 2, 1992, pp. 54-61.
2. Jacobs, R. D. Heederik, J. Douwes, and U. Zahringer. "Endotoxin
Structure", Int. J. Occ. Health. Vol. 3(1 Suppl), 1997, pp. S6-S7.
3. Moran, A. "Structure-Bioactivity Relationship of Bacterial Endotox-
ins", J. Toxic. Toxin Rev., Vol." 14, No.l, 1995, pp. 47 83.
4. Schletter. J., H. Heine. A. Ulmer. E. Rietschel. "Molecular Mecha-
nisms of Endotoxin Activity". Arch. ofMicrob. Vol. 164, 1995, pp.
383-389.
5. Milton. D. 1995. Endotoxin. Bioaerosols, H. Surge, editor, CRC
Press, Boca Raton, FL.
6. Uliner, A. "Biochemistry and Cell Biology of Endotoxins. Human
Challenge Studies with Endotoxins" Int. J of Occ. Health, Vol. 3, No.
1(1), 1997, pp. S8-S15.
7. Shiba. T. and S. Kusumoto. 1984. Chemical Synthesis and Biological
Activity of Lipid A Analogs. Handbook of Endotoxin, Vol. 1, Chemis-
try of Endotoxin, E. Rietschel, editor, Elsevier Science Publishers.
8. Jacobs, R. "Airborne Endotoxins: An Association with Occupational
Lung Disease". App. Ind. Hyg., Vol. 4, No. 2, 1989, pp. 50-56.
9. Olenchock, S. 1997. Airborne Endotoxin. Manual of Environmental
Microbiology, C. Hurst, editor. ASM press. Washington, DC.
10. Rylander, R., B. Rake, J. Fischer, and I. Helander. "Pulmonary Func-
tion and Symptoms After Inhalation of Endotoxin", Amer. Rev. Resp.
Dis., Vol. 140, No. 4, 1989, pp. 981-986.
11. Harber, P., M. Schenker. and J. Balmes. 1996. Occupational and Envi-
ronmental Respiratory Diseases, Mosby Publishing, 1st edition, St.
Louis, ML
12. Salem, M., Y. Li, B. Langholz and F. Gillialand. "Early-Life Environ-
mental Risk Factors for Asthma" Env. Health Persp.. Vol. 112, No.6,
2004, pp. 760-765.
13. Braun-Fahrlander, C.. U. Flerz, W. Eder, L. Grize, S. Maisch, D. Carr,
F. Gerlach, A. Buff, R. Schierl, H. Renz, D. Nowak, and E. von Mutius.
"Environmental Exposure to Endotoxin and its Relation to Asthma
School Aged Children". New Eng. J. Med. Vol. 347, No.l 2, 2002, pp.
869-877.
14. Olenchock, S. "Health Effects of Biological Agents: The Role of
Endotoxins". App. Occ. Env.Hyg., Vol. 9, No. 1, 1994, pp. 62-64.
15. Gibbs, R., C. Flu, G. Ho, and I. Unkovich. "Regrowth of Faecal
Coliforms and Salmonella in Stored Biosolids and Soil Amended with
Biosolids", Wat. Set. Tech., Vol. 35, No.l 1/12, 1997, pp. 269-275)
16. Westphal, P. and G. Christensen. "Lime Stabilization: Effectiveness of
Two Process Modifications", J. Wat.Poll. Control Fed., Vol. 55, No.
11, 1983, pp. 1381-1386.
17. Yeager, J. and R. Ward. "Effects of Moisture cvontent on Long-Term
Survival and Regrowth of Bacteria and Wastewater Sludge".^/?/?, and
Env.Microb. Vol. 41, No.5, 1981, 1117 1122.
18. Brooks, J., B. Tanner, C. Gerba, and 1. Pepper. "The Measurement
of Aerosolized Endotoxin From Land Application of Class B
Biosolids in Southeast Arizona", Can. J. Microbiol. Vol. 52, 2006,
pp. 150-156.
19. Castellan, R., S. Olenchock, K. Kinsley, and J. Hankinson. "Inhaled
Endotoxin and Decreased Spirometric Values". New Eng. J.Med. Vol.
317, No. 10. 1987, pp. 605 610.
20. Laitinen. S. "Importance of Sampling Extraction, and Preservation for
the Quantification of Biologically Active Endotoxin", Env. Med., Vol.
6, 1999, pp: 33 38.
21. BioWhittaker. 1996. LimulasAmebocyteLysate Assay. Walkersville,
MD.
22. SAS. 2003. Statistical Analysis Software Version 9.1. Cary, NC.
23. Carty, C. . U. Gehring. J. Cyrus. W. Bischof, and J. Heinreich. "Sea-
sonal Variability of Endotoxin in Ambient Fine Particulate Matter". J.
Env.Momt. Vol. 5, No. 6, 2003, pp. 953 958.
24. Milton. D. 1996.BacterialEndotoxins: AReview ofHealthEffects and
Potential Impact in the Indoor Environment, Ind. Air and Human
Health, 2cd edition, R. Gammage and B. Bervan, editors, CRC Press
Inc., Lewis Publishers, Boca Raton, FL.
25. Simpson.J.. R. Niven. C. Pickering. L. Oldham. A. Fletcher, and H.
Francis. "Comparative Personal Exposure to Organic Dusts and
Endotoxin", Ann.Occ.Hyg.. Vol. 43, No. 2, 1999, pp. 107-115.
26. ACGIH. 1999. Bioaerosols: Assessment and Control. Janet Macher.
editor. American Conference of Government Industrial Hygienist.
Cincinnati, OH.
27. Epstein, E., N. Wu, C. Youngberg, and G. Croteau. "Controlling Dust
and Bioaerosols at a Biosolids Compost Facility". BioCycle. Vol. 42,
No. 4, 2001, pp. 50-54.
28. NIOSH, 1998. Criteria for a Recommended Standard: Occupational
Exposure to .Metal Working Fluids. Publication, No. 98-102.
Cincinnati, OH.
-------
APPENDIX D
Soil Agronomic Results for Land Samples and Fecal Coliform Results for Land Samples
-------
Table D-l. Soil Agronomic Results for Land Samples
SOIL CHARACTERIZATION
Sample
Date
8/25/04
8/25/04
8/25/04
8/25/04
8/25/04
8/25/04
8/25/04
8/25/04
8/25/04
10/1/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
9/30/04
10/28/04
10/28/04
Sample ID
PLOT 1 0-5 CM
PLOT 1 10-15 CM
PLOT 1 20-25 CM
PLOT 2 0-5 CM
PLOT 2 10-15 CM
PLOT 2 20-25 CM
PLOT 3 0-5 CM
PLOT 3 10- 15 CM
PLOT 3 20-25 CM
Biosolid stockpile
P1-A6-0-5 CM
P1-A6-10-15 CM
P1-A6-20-25 CM
Pl-DO-0-5 CM
PI -DO- 10- 15 CM
Pl-DO-20-25 CM
P1-G8-0-5 CM
P1-G8-10-15 CM
P1-G8-20-25 CM
P2-B2-0-5 CM
P2-B2-10-15 CM
P2-B2-20-25 CM
P2-J1-0-5 CM
P2-J1-10-15 CM
P2-J1-20-25 CM
P2-IO-0-5 CM
P2-IO-10-15 CM
P2-IO-20-25 CM
P3-B8-0-5 CM
P3-B8-10-15 CM
P3-B8-20-25 CM
P3-G2-0-5 CM
P3-G2-10-15 CM
P3-G2-20-25 CM
P3-J1-0-5 CM
P3-J1-10-15 CM
P3-J1-20-25 CM
PI -0-5 CM
Pl-10-15 CM
% Sand
43.0
55.0
29.0
43.0
35.0
29.0
47.0
33.0
35.0
5.0
38.8
39.0
17.7
30.1
27.4
20.1
35.1
32.9
19.2
36.4
39.0
24.9
37.1
15.9
23.2
35.0
37.3
23.5
36.2
37.2
19.2
40.7
39.3
20.1
40.6
37.5
30.8
42.0
36.0
%
Silt
28.0
6.0
30.0
32.0
30.0
28.0
26.0
24.0
20.0
78.6
20.7
18.6
23.6
20.2
27.6
27.8
25.1
30.2
31.9
22.8
23.4
29.6
24.2
56.5
33.7
25.8
22.0
25.1
17.2
18.5
18.7
14.6
16.8
11.0
14.4
16.9
14.8
32.0
30.0
%
Clay
29.0
39.0
41.0
25.0
35.0
43.0
27.0
43.0
45.0
16.4
41.0
42.4
58.7
49.7
45.0
52.1
39.8
36.9
48.9
40.8
37.6
45.5
38.7
27.6
43.1
39.2
40.7
51.4
46.6
44.3
62.1
44.7
43.9
68.9
45.0
45.6
54.4
26.0
34.0
USDA Textural
Class
CLAY LOAM
SANDY CLAY
CLAY
LOAM
CLAY LOAM
CLAY
SANDY CLAY
LOAM
CLAY
CLAY
SILT LOAM
CLAY
CLAY
CLAY
CLAY
CLAY
CLAY
CLAY LOAM
CLAY LOAM
CLAY
CLAY
CLAY LOAM
CLAY
CLAY LOAM
SILTY CLAY
LOAM
CLAY
CLAY LOAM
CLAY
CLAY
CLAY
CLAY
CLAY
CLAY
CLAY
CLAY
CLAY
CLAY
CLAY
LOAM
CLAY LOAM
Bulk
Density
(gm/cc)
0.79
1.01
0.99
0.87
1.05
1.01
0.75
0.97
0.90
0.89
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.83
1.07
Cation
Exchange
Capacity
(meg/lOOg)
12.0
10.8
10.2
12.9
9.9
9.6
12.3
9.8
10.0
22.0
14.1
10.3
10.5
12.1
10.0
10.6
11.7
9.4
9.9
11.2
9.0
9.6
11.3
9.4
9.7
11.3
8.9
9.7
13.2
9.6
9.8
12.5
9.1
10.4
13.2
9.8
10.2
11.6
10.2
%
Moisture
@l/3 bar
33.0
27.7
32.7
31.3
24.9
29.8
NA
29.7
38.4
158.4
32.3
23.2
31.9
31.3
26.3
32.1
34.4
23.5
28.6
32.6
22.7
28.1
22.4
30.5
30.4
31.3
24.6
30.7
34.7
24.3
43.1
35.7
26.1
41.0
36.7
27.3
29.5
34.4
24.4
%
Moisture
@15 bar
20.8
17.7
21.3
18.6
14.7
19.5
NA
18.5
25.0
75.6
17.3
13.6
22.2
17.6
16.4
20.3
17.2
14.3
19.5
17.5
13.4
18.0
14.0
15.2
18.4
15.4
13.7
19.1
18.5
15.9
26.8
18.4
15.4
25.5
18.5
16.7
18.2
22.3
15.6
%
Organic
Matter
8.0
1.9
1.1
5.2
1.8
1.1
8.3
2.3
1.6
NA
6.1
2.0
0.8
6.0
1.6
1.0
6.6
1.9
0.8
7.0
1.9
1.0
6.2
1.5
0.9
6.4
1.7
1.1
9.4
2.0
0.9
8.2
1.9
1.2
7.4
1.9
1.5
7.0
2.1
pH
(H20)
5.9
6.2
6.8
6.5
6.4
6.9
5.9
6.7
6.8
7.4
6.5
6.4
6.7
5.8
6.2
6.4
6.4
6.4
6.7
5.9
6.2
6.6
6.0
6.4
6.8
5.5
6.6
6.8
6.2
6.9
7.1
6.1
6.8
6.9
6.2
7.1
7.2
5.7
6.2
%
Total
N
1.075
2.926
0.850
0.574
0.299
0.252
0.729
0.356
0.363
NA
0.595
0.346
0.181
0.554
0.272
0.141
0.457
0.326
0.128
0.548
0.111
0.084
0.386
0.155
0.158
0.363
0.185
0.235
0.353
0.450
0.171
0.568
0.222
0.104
0.719
2.250
0.262
0.386
0.225
Total P
(ppm)
1159
495
330
1079
446
301
1209
424
339
24453
1183
493
275
1313
392
424
1133
608
396
1373
609
350
1223
428
231
993
404
301
1183
387
303
1063
471
311
1203
418
394
1339
526
Olsen
Phosphorus
(ppm)
45
8
o
J
37
4
4
33
6
4
176
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Soluble Salts
(mmhos/cm)
0.14
0.11
0.10
0.26
0.07
0.12
0.14
0.20
0.13
4.09
0.35
0.14
0.08
0.22
0.07
0.07
0.29
0.07
0.06
0.20
0.11
0.10
0.15
0.09
0.07
0.14
0.07
0.07
0.19
0.13
0.10
0.29
0.11
0.12
0.34
0.14
0.11
0.26
0.08
Base Saturation Data (ppm)
K
150
33
25
178
57
50
211
65
54
823
293
48
60
173
33
36
139
23
21
196
57
66
93
21
20
132
27
22
152
38
27
199
28
32
185
32
23
149
29
Ca
925
883
898
1124
802
819
1007
797
773
2231
1351
898
917
926
821
922
1008
758
834
818
697
780
922
759
830
821
741
845
1196
802
757
1053
758
849
1169
857
901
915
832
Mg
273
206
190
252
203
197
252
227
233
627
334
193
225
265
205
225
281
195
226
256
211
227
258
203
234
236
195
219
344
268
272
308
246
294
408
305
300
256
187
Na
26
23
23
20
22
18
24
21
20
221
18
12
10
12
27
14
18
14
13
9
9
12
14
13
13
10
11
10
12
21
11
21
15
12
14
14
11
13
10
H
46
45
39
46
39
37
46
37
40
26
38
40
38
48
40
40
38
38
38
44
36
36
42
38
35
48
34
35
39
31
36
41
31
36
34
29
31
45
43
D-l
-------
Table D-l (continued). Soil Agronomic Results for Land Samples
SOIL CHARACTERIZATION CONT.
Sample
Date
10/28/04
10/28/04
10/28/04
10/28/04
10/28/04
10/28/04
10/28/04
1/4/2005
1/4/05
1/4/05
1/4/05
1/4/05
1/4/05
1/4/05
1/4/05
1/4/05
2/8/05
2/8/05
2/8/05
2/8/05
2/8/05
Sample ID
PI -20-25 CM
P2-0-5 CM
P2-10-15 CM
P2-20-25 CM
P3-0-5 CM
P3-10-15 CM
P3 -20-25 CM
PI -0-5 CM
Pl-10-15 CM
PI -20-25 CM
P2-0-5 CM
P2-10-15 CM
P2-20-25 CM
P3-0-5 CM
P3-10-15 CM
P3 -20-25 CM
OTest Soil/
lOODiluent Soil
25Test Soil/
75Diluent Soil
50Test Soil/
SODiluent Soil
75Test Soil/
25Diluent Soil
lOOTest Soil/
ODiluent Soil
% Sand
30.0
42.0
36.0
32.0
36.0
36.0
32.0
51.0
41.0
45.0
49.0
41.0
39.0
51.0
47.0
31.0
73.0
63.0
51.0
49.0
37.0
%
Silt
28.0
26.0
28.0
28.0
30.0
24.0
20.0
24.0
26.0
24.0
26.0
28.0
28.0
18.0
20.0
20.0
8.0
12.0
16.0
16.0
26.0
%
Clay
42.0
32.0
36.0
40.0
34.0
40.0
48.0
25.0
33.0
31.0
25.0
31.0
33.0
31.0
33.0
49.0
19.0
25.0
33.0
35.0
37.0
USDA Textural
class
CLAY
CLAY LOAM
CLAY LOAM
CLAY LOAM
CLAY LOAM
CLAY LOAM
CLAY
SANDY CLAY
LOAM
CLAY LOAM
SANDY CLAY
LOAM
SANDY CLAY
LOAM
CLAY LOAM
CLAY LOAM
SANDY CLAY
LOAM
SANDY CLAY
LOAM
CLAY
SANDY LOAM
SANDY CLAY
LOAM
SANDY CLAY
LOAM
SANDY CLAY
LOAM
CLAY LOAM
Bulk
Density
(gm/cc)
1.04
0.86
1.05
1.04
0.92
1.05
0.91
0.80
1.04
0.99
0.87
1.07
1.08
0.83
1.02
0.92
0.91
0.92
0.95
0.95
0.96
Cation
Exchange
Capacity
(meg/lOOg)
10.1
12.8
9.4
10.0
11.8
10.3
11.9
12.9
12.2
11.0
12.7
9.8
10.0
12.9
10..2
11.7
8.1
13.2
16.4
20.9
23.7
%
Moisture
®l/3 bar
27.7
34.1
24.6
28.9
35.9
26.6
40.6
39.8
28.1
32.6
29.5
24.9
29.8
35.8
26.4
39.8
20.3
25.1
29.1
35.2
37.7
%
Moisture
@15 bar
18.1
21.1
14.9
18.1
20.9
17.6
29.4
22.5
17.8
21.6
18.8
14.6
18.6
21.2
17.9
26.7
11.1
14.4
16.3
19.0
21.1
%
Organic
Matter
1.3
7.0
1.7
1.2
6.4
1.8
1.4
8.9
1.8
1.1
6.4
1.7
1.1
8.6
1.8
1.3
3.8
4.4
5.4
5.4
5.8
pH
(H20)
6.4
5.8
6.3
6.7
6
6.8
7.1
5.7
6.7
6.5
5.8
6.4
6.7
5.9
7.2
7.2
7.6
7.6
7.5
7.4
7.3
%
Total
N
0.329
0.443
0.144
0.148
0.309
0.114
0.188
0.430
0.100
0.070
0.370
0.100
0.060
0.410
0.090
0.060
0.047
0.094
0.156
0.231
0.289
Total P
(ppm)
313
1149
425
334
970
443
347
1214
430
342
925
394
268
1224
383
298
51
292
531
870
1238
Olsen
Phosphorus
(ppm)
NA
NA
NA
NA
NA
NA
NA
49
8
8
48
6
6
44
7
5
12
24
36
41
48
Soluble Salts
(mmhos/cm)
0.05
0.61
0.11
0.08
0.19
0.10
0.14
0.22
0.20
0.11
0.19
0.11
0.13
0.15
0.12
0.11
0.23
0.23
0.30
0.33
0.47
Base Saturation Data (ppm)
K
25
328
47
47
167
41
41
172
34
27
206
60
48
152
33
41
22
44
61
88
115
Ca
876
985
740
852
952
899
994
1012
1123
948
1061
816
891
1102
961
1090
1331
2183
2627
3300
3615
Mg
189
275
195
205
255
259
305
300
259
226
290
214
217
302
284
308
49
100
142
195
264
Na
11
14
9
10
10
11
13
14
14
11
12
12
11
11
12
15
36
35
26
23
13
H
40
46
39
38
45
35
42
48
42
42
44
38
36
44
29
35
8
12
18
24
30
D-2
-------
Table D-2. Fecal Coliforms Results for Land Samples
Plot
1
2
3
Biosolids
Date
8/25/2004
9/27/2004
9/30/2004
10/14/2004
10/28/2004
12/7/2004
1/4/2005
8/25/2004
9/27/2004
9/30/2004
10/14/2004
10/28/2004
12/7/2004
1/4/2005
8/25/2004
9/27/2004
9/30/2004
10/14/2004
10/28/2004
12/7/2004
1/4/2005
10/1/2004
Fecal Coliforms by Grid Location (MPN/g dry wt)
Replicate 1
O.22
0.23
>2.21E03
>2.27E05
7.00E+03
974
>1.95E04
O.23
0.54
>2.24E03
1.72E+04
>2.13E04
>2.19E04
2.38E+04
O.23
0.23
>2.26E03
4.61E+04
2.28E+03
115
2.26E+04
>8.08E06
Replicate 2
O.22
2.15
67.8
>2.1E05
6.02E+03
1.36E+04
2.25E+03
O.22
0.55
614
1.43E+04
>1.96E04
>2.32E04
>2.09E04
O.23
448
>2.12E03
4.43E+04
1.25E+03
239
1.95E+03
>8.08E06
Replicate 3
O.22
0.22
2.20E+03
1.48E+03
1.23E+04
2.38E+03
1.88E+03
O.22
0.23
>2.15E03
2.27E+04
>2.05E04
1.22E+04
>2.18E04
O.22
0.22
>2.24E03
1.19E+05
1.21E+04
44
2.21E+04
>8.08E06
D-3
------- |