September 2007
Environmental Technology
Verification Report
DAKOTA TECHNOLOGIES, INC.
BALLAST WATER EXCHANGE ASSURANCE METER
(BEAM)100
Prepared by
Battelle
Batreiie
"Ihe Business o/ Innovation
Under a cooperative agreement with
^jf CrTr\ U.S. Environmental Protection Agency
ET1/ET1/ET1/
-------
September 2007
Environmental Technology Verification
Report
ETV Advanced Monitoring Systems Center
DAKOTA TECHNOLOGIES, INC.
BALLAST WATER EXCHANGE ASSURANCE METER
(BEAM) 100
by
Mary Schrock
William Ivancic
Carlton Hunt
Zachary Willenberg
Amy Dindal
Battelle
Columbus, Ohio 43201
-------
Notice
The U.S. Environmental Protection Agency (EPA), through its Office of Research and
Development, has financially supported and collaborated in the extramural program described
here. This document has been peer reviewed by the Agency. Mention of trade names or
commercial products does not constitute endorsement or recommendation by the EPA for use.
11
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
nation's air, water, and land resources. Under a mandate of national environmental laws, the
Agency strives to formulate and implement actions leading to a compatible balance between
human activities and the ability of natural systems to support and nurture life. To meet this
mandate, the EPA's Office of Research and Development provides data and science support that
can be used to solve environmental problems and to build the scientific knowledge base needed
to manage our ecological resources wisely, to understand how pollutants affect our health, and to
prevent or reduce environmental risks.
The Environmental Technology Verification (ETV) Program has been established by the EPA to
verify the performance characteristics of innovative environmental technology across all media
and to report this objective information to permitters, buyers, and users of the technology, thus
substantially accelerating the entrance of new environmental technologies into the marketplace.
Verification organizations oversee and report verification activities based on testing and quality
assurance protocols developed with input from major stakeholders and customer groups
associated with the technology area. ETV consists of six environmental technology centers.
Information about each of these centers can be found on the Internet at http://www.epa.gov/etv/.
Effective verifications of monitoring technologies are needed to assess environmental quality
and to supply cost and performance data to select the most appropriate technology for that
assessment. Under a cooperative agreement, Battelle has received EPA funding to plan,
coordinate, and conduct such verification tests for "Advanced Monitoring Systems for Air,
Water, and Soil" and report the results to the community at large. Information concerning this
specific environmental technology area can be found on the Internet at http://www.epa.gov/etv/
centers/center 1 .html.
in
-------
Acknowledgments
The authors wish to acknowledge the support of all those who helped plan and conduct the
verification test, analyze the data, and prepare this report. We would like to thank the U.S. Coast
Guard Research and Development Center for providing co-funding to perform the verification
testing. Many thanks to Wei Huang, Ph.D. candidate in the Department of Environmental, Earth,
and Ocean Sciences at University of Massachusetts, Boston; and Battelle staff members Yixian
Zhang, Amy Dindal, Brenda LaSorsa, Tom Gulbransen, Matt Fitzpatrick, and Skip Newton for
collecting environmental water samples for use in testing. We also would like to thank Ms. Gail
Roderick, U.S. Coast Guard Research and Development Center; Dr. Robert Chen, University of
Massachusetts, Boston; and Dr. Darryl Keith, U.S. EPA, Office of Research and Development,
Atlantic Ecology Division, Narragansett, Rhode Island, for their careful review of the test/quality
assurance plan and this verification report.
IV
-------
Contents
Notice ii
Foreword iii
Acknowledgments iv
List of Abbreviations viii
Chapter 1 Background 1
Chapter 2 Technology Description 2
Chapters Test Design and Procedures 3
3.1 Introduction 3
3.2 Test Facility 4
3.3 Test Procedures 4
3.3.1 Test Sample Collection and Preparation 5
3.3.2 Test Sample Analysis Procedure 13
Chapter 4 Quality Assurance/Quality Control 15
4.1 Quality Control Samples 15
4.1.1 Negative Controls 15
4.1.2 Positive Controls 15
4.1.3 Calibration Checks 15
4.2 Audits 16
4.2.1 Performance Evaluation Audits 16
4.2.2 Technical Systems Audit 16
4.2.3 Data Quality Audit 16
4.3 QA/QC Reporting 17
4.4 Data Review 17
Chapters Statistical Methods 19
5.1 Accuracy 19
5.2 Linearity 20
5.3 Precision 20
5.4 Method Detection Limit (MDL) 20
5.5 Inter-unit Reproducibility 21
5.6 Temperature Effects 21
5.7 Matrix Effects 21
5.8 Data Completeness 21
5.9 Operational Factors 22
Chapter 6 Test Results 23
6.1 Accuracy 23
6.2 Linearity 27
6.3 Precision 30
-------
6.4 Method Detection Limit 30
6.5 Inter-unit Reproducibility 30
6.6 Temperature Effects 32
6.7 Matrix Effects 35
6.8 Data Completeness 37
6.9 Operational Factors 37
Chapter 7 Performance Summary 39
Chapter 8 References 41
Figures
Figure 2-1. Dakota Technologies, Inc.' s Ballast Water Exchange Assurance Meter 100 2
Figure 3-1. 3-D Plot of Seawater Collected in Sequim Bay, Washington State. Intensity and
features are representative of the received seawater samples with a maximum
intensity (IM) = 14 11
Figure 3-2. Seawater Collected at a Beach in Palm Beach, Florida. Intensity and features are
cleaner than typical seawater samples received with IM = 10 11
Figure 3-3. Seawater Collected in an Intercoastal Waterway in West Palm Beach, Florida.
Note the high intensity for this coastal sample where IM = 550 12
Figure 3-4. NASS-5 Standard. Features are representative of the received seawater samples,
but IM of 20 at excitation 300 nm is high for open-ocean water 12
Figure 3-5. Seawater from Duxbury Bay, Massachusetts. Intensity and features are
representative of the received seawater samples. The IM at excitation 300 nm
is 17 13
Figure 6-1. Comparison of BEAM and Reference Method F/R Values for Quinine Sulfate 24
Figure 6-2. Comparison of BEAM and Reference Method F/R Values for SR Fulvic Acid 24
Figure 6-3. Comparison of BEAM and Reference Method F/R Values for
Environmental Samples 25
Figure 6-4. Polynomial Correlation Between BEAM F/R and Reference Method F/R
Measurements With Quinine Sulfate 25
Figure 6-5. F (Upper Plot) and R (Lower Plot) Signals for the BEAM and Reference Method
Plotted Against Quinine Sulfate Solution Concentration 28
Figure 6-6. F (Upper Plot) and R (Lower Plot) Signals for the BEAM and Reference Method
Plotted Against Fulvic Acid Solution Concentration 29
Figure 6-7. Comparison of BEAM F/R Values for Quinine Sulfate (Upper Plot) and Fulvic
Acid (Lower Plot) at 4°C and 34°C to Those at 24°C 34
Figure 6-8. Matrix Effects Based on Percent Difference of BEAM Results Compared with
BEAM Equivalent (Based on Quinine Sulfate Correlation) Reference Method
Results 36
Tables
Table 3-1. Verification Test Samples 6
Table 3-2. Environmental Samples 9
Table 4-1. Summary of Data Recording Process 18
vi
-------
Table 6-1. Percent Difference Between the BEAM F/R Values and the BEAM Equivalent
Reference Method F/R Values 26
Table 6-2. Relative Standard Deviation of Triplicate Measurements with BEAMs and
Reference Method 31
Table 6-3. Method Detection Limits 32
Table 6-4. BEAM Inter-unit Reproducibility 33
Table 6-5. Percent Difference of 4°C and 34°C BEAM Measurements from 24°C BEAM
Measurements 35
Table 7-1. BEAM 100 Summary Table 39
vn
-------
List of Abbreviations
AMS
ASTM
BEAM
BWE
°C
CDOM
COC
EEM
EPA
ETV
F
FWHM
HPLC
IM
LRB
MDL
mL
NERL
nm
PC
PD
PE
ppb
PT
QA
QC
QCS
QMP
R
R2
RPD
RSD
SR
ISA
UV
Advanced Monitoring Systems
American Society for Testing and Materials
Ballast Water Exchange Assurance Meter
ballast water exchange
degrees Celsius
colored dissolved organic matter
chain of custody
excitation-emission matrix
U.S. Environmental Protection Agency
Environmental Technology Verification
intensity at 460 nm
full width at half maximum
high-performance liquid chromatography
maximum intensity
laboratory record book
method detection limit
milliliter
National Exposure Research Laboratory
nanometer
personal computer
percent difference
performance evaluation
parts per billion
performance test
quality assurance
quality control
quality control sample
quality management plan
intensity at 430 nm
coefficient of determination
relative percent difference
relative standard deviation
Suwannee River
technical systems audit
ultraviolet
Vlll
-------
Chapter 1
Background
The U.S. Environmental Protection Agency (EPA) supports the Environmental Technology
Verification (ETV) Program to facilitate the deployment of innovative environmental technologies
through performance verification and dissemination of information. The goal of the ETV Program
is to further environmental protection by accelerating the acceptance and use of improved and
cost-effective technologies. ETV seeks to achieve this goal by providing high-quality, peer-
reviewed data on technology performance to those involved in the design, distribution, financing,
permitting, purchase, and use of environmental technologies.
ETV works in partnership with recognized testing organizations; with stakeholder groups
consisting of buyers, vendor organizations, and permitters; and with the full participation of
individual technology developers. The program evaluates the performance of innovative
technologies by developing test plans that are responsive to the needs of stakeholders, conducting
field or laboratory tests (as appropriate), collecting and analyzing data, and preparing peer-
reviewed reports. All evaluations are conducted in accordance with rigorous quality assurance
(QA) protocols to ensure that data of known and adequate quality are generated and that the results
are defensible.
The EPA's National Exposure Research Laboratory (NERL) and its verification organization
partner, Battelle, operate the Advanced Monitoring Systems (AMS) Center under ETV. The AMS
Center recently evaluated the performance of Dakota Technologies, Inc.'s Ballast Water Exchange
Assurance Meter (BEAM) 100 in measuring colored dissolved organic matter (CDOM)
fluorescence as a tool for evaluating ballast water exchange (BWE).
-------
Chapter 2
Technology Description
The objective of the ETV AMS Center is to verify the performance characteristics of
environmental monitoring technologies for air, water, and soil. This report provides results for the
verification testing of Dakota Technologies, Inc.'s BEAM 100. The following is a description of
the BEAM 100, based on information provided by the vendor. The information provided below
was not verified in this test.
The BEAM 100 (Figure 2-1) is a portable, handheld
fluorimeter designed to generate a response relative to
the amount of CDOM in ballast water. The CDOM
related response is determined by exciting the sample
with near ultraviolet (UV) light and measuring the
resulting fluorescence to Raman scatter ratio.
The unit consists of a cuvette well permanently
mounted in the BEAM. The BEAM is operated
through four user-interface buttons. Acquired data are
shown in a display screen and can be transferred to a
personal computer (PC) for long-term storage.
Internally, the BEAM consists of electronics; a light-
emitting diode used as an excitation source; and two
photodetectors, each with different wavelength filters.
All measurements are recorded to the BEAM's
internal memory. The BEAM'S durable plastic
carrying case includes space for cuvette cleaning and
sample filtering accessories. The BEAM unit is
10.5 by 4.5 by 3.0 inches and weighs 2.5 pounds (with
batteries). The carrying case is 16 by 12 by 7 inches
and weighs approximately 10 pounds with the BEAM
unit and kit supplies in place. The BEAM 100 costs
approximately $6,000 per unit.
Figure 2-1. Dakota Technologies,
Inc.'s Ballast Water Exchange
Assurance Meter (BEAM) 100
-------
Chapter 3
Test Design and Procedures
3.1 Introduction
Mid-ocean ballast water exchange (BWE) is mandatory for all vessels entering U.S. waters from
outside the 200-mile exclusive economic zone. To support such regulation, accurate and portable
verification tools are needed for determining that BWE has taken place. One parameter proposed
as a means of distinguishing between coastal and open ocean water content in ballast water is
fluorescence due to colored dissolved organic matter (CDOM).(1'2'3) CDOM refers to the fraction
of dissolved organic matter that absorbs light and fluoresces in the UV and visible regions of the
spectrum.
The objective of this verification test was to evaluate the performance of the BEAM 100 in
measuring CDOM relative to a standard CDOM measurement approach using a laboratory bench-
scale excitation-emission spectrometer (Varian Gary Eclipse spectrometer) under controlled
laboratory conditions. This verification test was conducted from March to April 2007 according to
procedures specified in the Test/QA Plan for Verification of Ballast Water Exchange Screening
Tools including Amendments 1 and 2.(4) This evaluation assessed the capabilities of the BEAM
100 in both laboratory-prepared, performance test (PT) samples and real-world open-ocean and
coastal environmental samples. This test did not verify that the BEAM 100 successfully quantified
CDOM concentrations or detected BWE, but rather evaluated how well it measured fluorescence
from CDOM compared with a standard technique for measuring fluorescence. This test also did
not represent all types of waters that may be encountered in ballast water screening, but a range of
water (and subsequently the range of fluorescence measurements generated from various types of
water) that may be expected in practical application.
The BEAM 100 was verified by evaluating the following parameters:
• Accuracy—Comparison of the BEAM 100 CDOM measurement to CDOM measurements
generated by a Varian Gary Eclipse spectrometer with both instruments at ambient laboratory
temperature (approximately 24°C).
• Linearity—CDOM measurements from varying concentrations of standard analytes known to
fluoresce plotted against the analyte concentration. Linearity was evaluated based on linear
regression statistics (i.e., slope and correlation coefficients).
• Precision—The relative standard deviation (RSD) of replicate measurements of the same
sample.
-------
Method detection limit (MDL)—Analysis of seven replicates of known fluorescing analytes at
a concentration five times Dakota Technologies, Inc.'s expected detection limit for the analyte.
Inter-unit reproducibility—Relative percent difference (RPD) between the average of triplicate
CDOM measurements of the same sample taken at the same temperature using two different
BEAM 100 units.
Temperature effects—Comparison of the BEAM 100 CDOM measurements at approximately
4 degrees Celsius (°C) and 34°C with CDOM measurements at ambient laboratory temperature
(approximately 24°C).
Matrix effects—Evaluated by comparing the percent difference (PD) of the BEAM 100
measurements with the Varian Gary Eclipse spectrometer measurements for the various types
of samples analyzed during verification testing.
Data completeness—The number of valid measurements out of the total number of
measurements taken.
Operational factors—Observations and records related to maintenance needs, calibration
frequency, data output, consumables used, ease of use, repair requirements, waste production,
and sample throughput.
3.2 Test Facility
Laboratory analyses of the BEAM 100 were conducted in Battelle laboratories in Columbus, Ohio.
No field portability testing was conducted during this technology verification, although
temperature and matrices evaluated were varied to simulate field conditions.
3.3 Test Procedures
Test samples used in the verification test included performance test (PT) samples, environmental
samples, and quality control (QC) samples as summarized in Table 3-1. These various types of
samples are discussed in Section 3.3.1, with the exception of the QC samples, which are described
in Section 4.1. The PT and environmental samples were analyzed in triplicate and compared with
triplicate measurements taken with the reference method. These samples were evaluated for
accuracy compared with expected measurements based on the reference method CDOM analyses,
instrument linearity across the range of concentrations tested, and precision among the replicate
measurements obtained. Two BEAM 100 units were used to measure the test samples.
Measurements of aliquots of the same sample were taken sequentially with the two units and with
the reference method within minutes of each other. Inter-unit reproducibility was evaluated based
on the measurements taken with the two BEAM units. All measurements made for direct
comparison with the reference method were conducted at ambient room temperature. The
reference method spectrometer, a Varian Gary Eclipse spectrometer, was configured to be as
similar to the BEAM units as possible. Bandwidths were set the same as the BEAM at 10 nm full
width at half maximum (FWHM) for all comparison tests, cell geometry was positioned with 90
degrees between the excitation source and the emission detector, and a 1 cm path length cuvette
was used. During testing, temperatures were monitored by recording an air reading, a water
4
-------
reading, and the BEAM unit internal temperature reading. Ambient room temperatures observed
during testing were as follows: 19.3 to 24.4°C for air, 19.6 to 21.4°C for water, and 22 to 28°C for
BEAM internal measurements. Although the temperature varied slightly during the course of
testing, the ambient room temperature tests will be referred to as 24°C tests in this report. Also
note that while the BEAM internal temperature was consistently slightly higher than either the air
or water temperature readings, the instruments and test solutions were sufficiently equilibrated that
adding sample to the BEAM cell did not change the BEAM internal temperature and the BEAM
internal temperature remained constant during all measurements.
Because these technologies will be used in a wide range of temperatures in practical application
and because temperature can affect CDOM fluorescence, a subset of test samples was analyzed
using only the BEAM 100 units at two additional temperatures (approximately 4°C and 34°C) to
evaluate the BEAM 100's variability due to temperature effects. Testing at 4°C took place inside a
walk-in refrigerator and testing at 34°C took place inside a chamber where the elevated
temperature was created using space heaters and heat lamps. Actual temperatures measured during
the temperature extreme testing at 4°C were 4.8 to 7.9°C for air, 6.1 to 8.3°C for water, and 9 to
13°C for BEAM internal measurements. For testing at 34°C, the actual temperatures measured
were 33.4 to 35.2°C for air, 34.3 to 35.3°C for water, and 36 to 39°C for BEAM internal
measurements. Although the temperature varied during the temperature extreme tests, as also
during the ambient room temperature testing, these tests will be referred to as 4°C and 34°C tests
in this report. Note that while temperature is one of several variables that might affect practical
application (other possibilities include humidity, ambient light, and exposure to the elements), this
verification test evaluated only the effect of varying temperature (one temperature above and one
temperature below ambient laboratory temperature) on the BEAM 100's performance.
The procedures for preparing, storing, and analyzing test samples are provided below.
3.3.1 Test Sample Collection and Preparation
3.3.1.1 Performance Test (PT) Samples
PT samples were created by adding compounds known to cause fluorescence (i.e., quinine sulfate
and SR fulvic acid) at multiple concentration levels to Burdick and Jackson HPLC grade water.
Burdick and Jackson FIPLC grade water was selected because (1) it is certified as having low
levels of organic compounds and (2) it was checked for interferences in the wavelengths of interest
for verification testing [430 and 460 nanometers (nm)] and found to be clean, with a ratio of 460
nm/430 nm of approximately 0.02. The quinine sulfate samples were prepared in 0.1M sulfuric
acid solution, which was made with EMD Chemicals Inc. GR ACS grade sulfuric acid and the
Burdick and Jackson FIPLC grade water following ASTM E579-04,(5) including evaluating the
0.1M sulfuric acid solution for fluorescence prior to use and using the 0.1M solution as the
unspiked blank sample for the quinine sulfate samples. Fulvic acid samples were prepared in
Burdick and Jackson FIPLC grade water with no acidification. The fulvic acid solutions were
swirled and allowed to dissolve until no visible particles remained. The stock solution and any
subsequent dilutions of the stock were visibly checked to be free of precipitation before their use.
Because the fulvic acid solutions were to be prepared in water only, with no acidification as a
-------
Table 3-1. Verification Test Samples
Performance
Factor
PT Samples
Quinine sulfate prepared
in Burdick and Jackson
HPLCa grade water per
ASTMb E579-04(5)
Fulvic acid prepared in
Burdick and Jackson
HPLC grade water
Accuracy, linearity,
precision,
temperature effects
MDL
Accuracy, linearity,
precision,
temperature effects
MDL
Sample Description
Unspiked
1 ppb° quinine sulfate
5 ppb quinine sulfate
10 ppb quinine sulfate
50 ppb quinine sulfate
100 ppb quinine sulfate
Quinine sulfate at 1 ppb (5 x
Dakota-provided detection
limit of 0.2 ppb)
Unspiked
100 ppb SRd fulvic acid
500 ppb SR fulvic acid
1,000 ppb SR fulvic acid
5,000 ppb SR fulvic acid
10,000 ppb SR fulvic acid
SR fulvic acid at 100 ppb
(5 x Dakota-provided
detection limit of 20 ppb)
Replicates for Each
BEAM 100 unit
4°C
3
3
3
3
3
3
-
3
3
3
3
3
3
-
24°C
3
3
3
3
3
3
7
3
3
3
3
3
3
7
34°C
3
3
3
3
3
3
-
3
3
3
3
3
3
-
Environmental Samples
Location 1 -open ocean
Location 2-open ocean
Location 3 -coastal
Location 4-coastal
Location 5 -coastal
Location 6-coastal
Location 7-coastal
Location 8-coastal
Location 9-coastal
Location 10-coastal
Location 11 -coastal
Location 12-coastal
Matrix effects
Unspiked
-
-
-
-
-
-
-
-
-
-
-
-
3
3
3
3
3
3
3
3
3
3
3
3
-
-
-
-
-
-
-
-
-
-
-
-
QC Samples
Negative control
Positive control
Calibration check
N/Ae
N/A
N/A
Burdick and Jackson HPLC
grade water
5,000 ppb SR fulvic acid
10 ppb quinine sulfate
TOTAL
19
11
Single measurement minimally
every nine verification sample
measurements. A total of 24
calibration check measurements
were made.
212
Shading indicates samples that were also analyzed using the reference method.
a: HPLC = high-performance liquid chromatography
b: ASTM = American Society for Testing and Materials
c: ppb = parts per billion
d: SR = Suwannee River
e: N/A = not applicable, as QC samples were used to monitor BEAM and reference method performance during
verification testing.
-------
preservative, recording pH of the fulvic acid solutions was included in the test/QA plan(4) to ensure
that the fulvic acid solutions did not change over time. This pH measurement step was
inadvertently omitted during testing and is a deviation from the test/QA plan protocol. However,
during testing, the fulvic acid solutions were prepared fresh from a stock solution each testing day
and testing was completed within a 12-hour period. Because the fulvic acid solutions were made
new from the stock solution every day of testing, were not subject to long periods of time before
use, and were analyzed sequentially by the BEAM 100 and the reference method within minutes of
each other, there was no need to monitor the pH on a continual basis; thus, the absence of solution
pH measurements did not have any negative impact on the test results. The fulvic acid solutions
were prepared at a concentration 100 times that of the quinine sulfate solutions because, on an
equivalent weight basis, fulvic acid produces a much lower fluorescence yield compared to quinine
sulfate.
3.3.1.2 Environmental Samples
Many sources can contribute to CDOM in a sample.(1) These sources can vary from location to
location and at various times within the same location can contain large differences in fluorescing
materials. Environmental samples were included in verification testing to simulate real samples
that may be found in practical application of BWE screening and that would have more complex
fluorescence patterns than a simple standard such as quinine sulfate. A total of 15 environmental
samples were obtained, consisting of 13 coastal water samples that were collected from areas
around the United States between October 2006 and February 2007 and two open ocean samples
that were purchased standard reference materials (NASS-5 and MOOS-1) available from National
Research Council Canada (Ottawa, Ontario, Canada).
Prior to verification testing, the environmental samples from all 15 locations were screened for
their CDOM response using the reference method instrumentation to select a subset of these
samples for inclusion in testing. Table 3-2 lists locations where environmental samples were
collected and the CDOM screening value obtained for each location. Full excitation-emission
matrix (EEM) measurements using the reference method instrumentation were also obtained on all
environmental samples to provide additional spectroscopic information that may not be revealed
by a single emission scan to aid in selecting a subset to include in testing. EEMs are valuable in
analyzing seawater samples because of the many variables that can be included in the analysis.
These observations were made to remove any possible outliers such as those that might be
produced by observing samples contaminated from an oil slick or some other event in the
collection process. The aim was to choose a set of environmental samples that would span the
typical fluorescence patterns in "real-world" seawater samples. Because these EEMs were only
used to screen samples for use in testing, the excitation intervals were set wider (25-nm intervals)
than might be typical if detailed EEMs were needed (5-nm).
Figures 3-1 to 3-5 present three-dimensional (3-D) and contour EEM plots of excitation vs.
emission vs. intensity for five of the environmental samples collected for this test and are typical of
the EEMs observed overall. Labels of X, R, and F found in each image are the three wavelengths
used by the reference method and the BEAM 100 units in verification testing (excitation X = 375
nm, emission R = 430 nm, and emission F = 460 nm). The EEM data were obtained using a Varian
Gary Eclipse spectrometer with an automated collection routine in which the excitations were set
between 300 and 450 nm at 25-nm intervals and emission observations were measured between
350 and 550 nm at 5-nm intervals. The excitation wavelength was set first, and then a series of
-------
emission wavelengths were observed. Each series was stored electronically as an emission scan.
The light intensity at each data collection point in the scan results from fluorescence, Raman
emission and Rayleigh scatter. The fluorescence is emitted from chromophores in the sample
(e.g., CDOM). The Raman peak, which is observed at a wavelength shifted 3600 cm"1
(wavenumbers) from the excitation energy, is attributed to the water matrix. The Rayleigh scatter
occurs because all molecules have a cross section to the excitation energy that scatters 90 degrees
into the emission spectrometer.
The series of emission scans were combined by Grams 3-D software, which allows orientation for
optimum views. The Rayleigh scatter is seen as the series of pyramids toward the back of the
images in the 3-D plots in Figures 3-1 to 3-5. The Raman bands are also seen as pyramids (where
fluorescence does not overwhelm their intensity), but these pyramids are less intense and are
shifted a bit more towards longer wavelengths (e.g. towards red light) compared to the Rayleigh
scatter pyramids. The light intensities on the contour plots shown in Figures 3-1 to 3-5 are divided
into 19 equal bands ranging from 0 to the full scale intensity as plotted. IM is the maximum
fluorescence observed outside of the Raman or excitation signal regions (i.e., away from the
regions where the Rayleigh scatter and Raman pyramids are observed) and varied with each
sample. The maximum fluorescence intensity, IM, for each sample is listed in the figure captions. In
principle, the excitation and Raman signals could be subtracted to obtain a plot showing a
maximum resulting only from CDOM fluorescence. However, for sample screening purposes this
was not necessary.
The EEM data for each environmental sample were reviewed. All samples appeared to be free of
extraneous contamination, with the exception of the NASS-5 open-ocean seawater standard shown
in Figure 3-4. The NASS-5 open-ocean seawater standard had more fluorescence at excitations of
300 nm and 325 nm with emission in the 450-nm spectral region than would be expected, based on
experience and other clean samples. Three of the other four samples shown in the figures above
have IM values less than the IM of the NASS-5 sample even though they are coastal waters. It was
noted that the NASS-5 standard was packed in a plastic container. Plastic containers are known to
leach compounds that fluoresce in the low 400-nm region when excited in the UV region of the
spectrum, and this is a likely cause of the observed fluorescence. However, at 375-nm excitation,
which is the excitation wavelength used in the verification test, the emission intensities were
unchanged from the other clean waters and, as a result, the NASS-5 open-ocean seawater standard
was used in testing since the fluorescence, possibly due to the plastic container, would not affect
the comparison of the reference method measurements with BEAM 100 measurements.
Based on the EEMs and the screening CDOM responses, 12 of the 15 environmental samples were
selected for inclusion in the verification test. The environmental samples included in the
verification test are noted in Table 3-2.
-------
Table 3-2. Environmental Samples
Location
Duxbury Bay, MA
Boston Harbor, MA
Massachusetts Bay
NF7, MA
Massachusetts Bay
NF10, MA
Sequim Bay, WA
Puget Sound, WA
East Coast, FL-1
East Coast, FL-2
Description
Off the dock at Battelle
Duxbury Operations in
Duxbury, MA
Inside of Neponset Estuary in
Boston Harbor, MA
Nine miles east of Deer Island,
MA
Nine miles east of Deer Island,
MA
Off the dock at Battelle Marine
Sciences Lab in Sequim, WA
Outside of Ediz Hook in Port
Angeles, WA
Inter-coastal water way in West
Palm Beach, FL
Atlantic Ocean beach off Palm
Beach, FL
Sample Type
Coastal seawater
A mixture of freshwater
and coastal seawater
with expected high
CDOM fluorescence
Coastal seawater
Coastal seawater
Coastal seawater
Coastal seawater
A mixture of freshwater
and coastal seawater
with expected high
CDOM fluorescence
Coastal seawater
Collection
Date
1/24/2007
11/10/2006
11/18/2006
11/18/2006
1/30/2007
1/30/2007
1/28/2007
1/28/2007
Storage Conditions
Unfiltered, stored cool (~4 °C)
after collection.
Filtered with a 0.7-um glass
fiber filter upon return to the
laboratory. Frozen until shipped
to Battelle. Stored cool (~4 °C)
after receipt at Battelle.
Filtered in the field with a 0.7
um glass fiber filter, frozen
after collection until shipped to
Battelle. Stored cool (~4 °C)
after receipt at Battelle.
Filtered in the field with a 0.7
um glass fiber filter, frozen
after collection until shipped to
Battelle. Stored cool (~4 °C)
after receipt at Battelle.
Unfiltered, stored cool (~4 °C)
after collection.
Unfiltered, stored cool (~4 °C)
after collection.
Unfiltered, stored cool (~4 °C)
after collection.
Unfiltered, stored cool (~4 °C)
after collection.
Screening
CDOM
Ratio
0.47
1.40
0.43
0.44
0.33
0.28
1.28
0.24
Used in
Testing
No
Yes
Yes
No
No
Yes
Yes
Yes
-------
Table 3-2. Environmental Samples (continued)
Location
Open Ocean - 1
Open Ocean - 2
Long Island Sound,
NY
New York Harbor,
NY
New York Bight,
NY
San Diego Harbor,
CA
Narragansett Bay,
RI
Description
NASS-5 Open Ocean Seawater
Reference Material for Trace
Metals collected 35 kilometers
southeast of Halifax, NS,
Canada
MOOS-1 Seawater Certified
Reference Material for
Nutrients collected off the
northern tip of Cape Breton
Island, NS, Canada
Dock in Port Jefferson, NY
East River, NY
Atlantic Ocean sample from a
beach in South Hampton, NY
San Diego Harbor, CA
Off 2-14 Great Island Rd,
Narragansett, RI
Sample Type
Open ocean seawater
with expected low
CDOM fluorescence
Open ocean seawater
with expected low
CDOM fluorescence
Coastal seawater
A mixture of freshwater
and coastal seawater
with expected high
CDOM fluorescence
Coastal seawater
Coastal seawater
Coastal seawater
Collection
Date
Prior to
June 1998
6/24/1996
2/4/2007
2/4/2007
2/4/2007
1/22/2007
10/21/2006
Storage Conditions
Filtered through a 0.45 micron
filter and then acidified to pH
1.6 with ultrapure nitric acid.
Stored cool (~4 °C) after
collection.
Filtered through a 0.05 micron
cartridge filter after collection,
irradiated after bottling. Stored
cool (~4 °C) after collection.
Unfiltered, stored cool (~4 °C)
after collection.
Unfiltered, stored cool (~4 °C)
after collection.
Unfiltered, stored cool (~4 °C)
after collection.
Unfiltered, stored cool (~4 °C)
after collection.
Unfiltered, stored cool (~4 °C)
after collection.
Screening
CDOM
Ratio
0.31
0.20
0.63
0.85
0.46
0.37
0.53
Used in
Testing
Yes
Yes
Yes
Yes
Yes
Yes
Yes
-------
so ,/r -^
x
\ ^-^-""""'^ 4°°
^f~i 3&°
'^ 300 _.,xv\at\°0
Figure 3-1. 3-D Plot of Seawater Collected in Sequim Bay, Washington State. Intensity and
features are representative of the received seawater samples with a maximum intensity (!M)
= 14.
Figure 3-2. Seawater Collected at a Beach in Palm Beach, Florida. Intensity and features
are cleaner than typical seawater samples received with IM = 10.
11
-------
Figure 3-3. Seawater Collected in an Intercoastal Waterway in West Palm Beach, Florida.
Note the high intensity for this coastal sample where IM = 550.
Figure 3-4. NASS-5 Standard. Features are representative of the received seawater
samples, but IM of 20 at excitation 300 nm is high for open-ocean water.
12
-------
Figure 3-5. Seawater from Duxbury Bay, Massachusetts. Intensity and features are
representative of the received seawater samples. The IM at excitation 300 nm is 17.
3.3.1.3 Quality Control Samples
QC samples are discussed in Section 4.1.
3.3.2 Test Sample Analysis Procedure
According to the cleaning and rinsing instructions in the BEAM 100 user manual, each unit was
flushed with distilled water, cleaned using the cleaning solution and swabs provided with the unit,
and then rinsed with copious amounts of distilled water. Once cleaned and rinsed, the units were
blank calibrated using Burdick and Jackson HPLC grade water, which was used to prepare all
calibration standards and PT samples used in verification testing. Blank calibration followed the
process listed in the instruction manual and was performed at the beginning of testing, prior to
quinine sulfate calibrations or any test sample analysis. The units were then ready for use in
verification testing. At the start of testing at each of the three temperatures used in verification
testing ( 24°C, 4°C, and 34°C), the BEAM 100 units were calibrated with quinine sulfate by
allowing the units and standard solution (10 ppb quinine sulfate) to come to the testing procedure
operating temperature with at least a 30-minute temperature equilibration time. Each unit was
calibrated with 10 ppb quinine sulfate following the "standard-point calibration process" listed in
the user manual, which consisted of filtering the calibration solution into the BEAM cell using the
syringes and 0.45-micron filters supplied with each unit, tightly capping the cell, and then simul-
taneously pressing the "CAL" and "RUN" buttons on the BEAM surface. The blank calibration and
quinine sulfate standard calibration was repeated only if it was necessary to address a testing
malfunction. After calibrating at the appropriate temperature, the units were ready for sample
measurement.
After the sample was allowed to equilibrate at the testing temperature for at least 30 minutes, a
sample measurement was acquired by first rinsing the unit with distilled water as outlined in the
user manual rinsing procedure. The sample was then filtered into the BEAM cell using the syringes
13
-------
and 0.45 micron filters which are supplied by the vendor and come with the BEAM kit. The cell
was rinsed with the sample, the rinse was discarded, and then the cell was filled with sample for
testing. Once filled with sample for testing, the cell cap was tightly closed and sample
measurements were taken by pressing the "RUN" button on the BEAM surface. After a few
seconds, the measured values were displayed on the BEAM unit and were manually recorded on
data sheets to provide a backup to the electronic data storage. After a day of analysis was completed
and prior to the next day of testing, the acquired data were downloaded to a PC using the software
provided by Dakota Technologies, Inc. and following the instructions listed in the user manual.
To compare measurements of the two BEAM units or the two BEAM units and the reference
method, aliquots of the same test sample were filtered into the cell of each instrument (BEAM and
reference spectrometer). Once all cells were full and caps tightened as appropriate, each unit's run
procedure was initiated.
When QC sample failures occurred (e.g., the quinine sulfate continuing calibration was outside the
0.41 to 0.45 acceptance criterion or the negative control had a reading >2,000 counts in the 460-nm
light channel) or a BEAM unit error occurred, the following corrective action process was followed.
First, the unit's cell well was re-rinsed and the sample was re-measured. Following a second
measurement error, the unit was re-cleaned, re-rinsed, and the sample re-measured. Following a
third measurement error, the unit was re-rinsed, then calibrated, and the sample re-measured. After
a fourth measurement error, the unit was re-rinsed, re-blank calibrated, re-calibrated, and the sample
re-measured. If none of these corrective actions helped resolve the problem, Dakota Technologies,
Inc. was contacted for technical support.
Two BEAM units were received for testing. The units were identified as 100-R2-08 and 100-R2-04
and are referred to as BEAM 08 and BEAM 04 in this report. BEAM 08 and BEAM 04 were used
for all 24°C room temperature testing. During testing, some technical difficulties (described in
Section 6.9) were encountered with BEAM 08; subsequently, this unit was replaced by Dakota
Technologies, Inc. with a unit identified as 100-R2-03 and is referred to as BEAM 03 in this report.
BEAM 03 and BEAM 04 were used for the testing at temperature extremes (4°C and 34°C).
14
-------
Chapter 4
Quality Assurance/Quality Control
QA/QC procedures were performed in accordance with the quality management plan (QMP) for the
AMS Center(6) and the test/QA plan for this verification test,(4) except for the deviation discussed in
Section 3.3.1.1. QA/QC procedures and results are described below.
4.1 Quality Control Samples
Steps were taken to maintain the quality of data collected during this verification test. This included
analyzing specific quality control samples (QCSs) at a regular frequency. QCSs included negative
controls, positive controls, and calibration checks.
4.1.1 Negative Controls
Burdick and Jackson HPLC grade water was analyzed as a negative control. Negative control
samples were used to help ensure that no sources of contamination were introduced in the sample
handling and analysis procedures. Dakota Technologies, Inc. indicated that the negative control
should provide a reading of <2,000 counts in the 460-nm light channel. If, at any time, the negative
control had more than 2,000 counts in the 460-nm light channel, the cell was cleaned and/or the
negative control solution and filter were replaced until a reading <2,000 counts was obtained.
4.1.2 Positive Controls
Throughout verification testing, positive control samples consisting of a 5,000-ppb SR fulvic acid
solution were analyzed to indicate to the operator that the BEAM 100 units were properly detecting
a positive response. CDOM ratio values between 0.86 and 0.89 were obtained for readings taken at
24°C. Slightly lower values (0.70 to 0.74) were obtained for the 5,000-ppb SR fulvic acid solution
at 34°C, and slightly higher values (0.93 to 1.16) were obtained at 4°C. The variations with
changing temperatures were not unexpected because of the influence temperature can have on
fluorescence.
4.1.3 Calibration Checks
Calibration checks of 10 ppb quinine sulfate were analyzed, at a minimum, after every nine
measurements of PT or environmental samples with the BEAM 100 units. The initial BEAM
quinine sulfate calibration set the CDOM ratio of a 10-ppb quinine sulfate solution at 0.43.
Subsequent calibration checks required that the CDOM ratio for a 10-ppb quinine sulfate solution
be between 0.41 and 0.45. If, at any time, the calibration check did not fall within these limits, the
cell was cleaned and the calibration check repeated. If the calibration check continued to remain
15
-------
outside the 0.41 to 0.45 limits, the affected BEAM unit was recalibrated. Analysis did not proceed
until a successful calibration check was obtained.
4.2 Audits
Three types of audits were performed during the verification test: a performance evaluation (PE)
audit of the reference method measurements made in this verification test, a technical systems audit
(ISA) of the verification test performance, and a data quality audit. Audit procedures are described
further below.
4.2.1 Performance Evaluation Audits
A PE audit was conducted to assess the quality of the reference method measurements made in this
verification test. The reference method PE audit was performed by supplying a second quinine
sulfate standard solution prepared from a different source of quinine sulfate than that used in
verification testing. The PE audit samples were analyzed in the same manner as all other samples,
and the analytical results for the PE audit samples were compared with the nominal concentration.
The target criterion for this PE audit was agreement of the analytical result within 3% of the
nominal concentration. This audit was performed once prior to the start of the test. The second
source PE standard was within 1.32% of the nominal value.
4.2.2 Technical Systems A udit
The Battelle Quality Manager performed one TSA during this verification test to ensure that the
verification test was being performed in accordance with the AMS Center QMP,(6) the test/Q A
plan,(4) and standard operating procedures. In the TSA, the Battelle Quality Manager reviewed the
reference methods used, compared actual test procedures with those specified or referenced in the
test/QA plan,(4) and reviewed data acquisition and handling procedures. Also in the TSA, the
Battelle Quality Manager observed testing, inspected sample chain-of-custody (COC)
documentation, and reviewed technology-specific record books. He also checked standard
certifications and technology data acquisition procedures and conferred with the technical staff. A
TSA report was prepared, including a statement of findings and the actions taken to address those
findings. The TSA findings were communicated to technical staff at the time of the audit. The
records concerning the TSA are permanently stored with the Battelle Quality Manager.
4.2.3 Data Quality Audit
At least 10% of the data acquired during the verification test were audited. Battelle's Quality
Manager traced the data from the initial acquisition, through reduction and statistical analysis, to
final reporting to ensure the integrity of the reported results. All calculations performed on the data
undergoing the audit were checked.
16
-------
4.3 QA/QC Reporting
Each audit was documented in accordance with Section 3.3.4 and 3.3.5 of the QMP for the ETV
AMS Center.(6) Once the audit reports were prepared, the Battelle Verification Test Coordinator
ensured that a response was provided for each adverse finding or potential problem and imple-
mented any necessary follow-up corrective action. The Battelle Quality Manager ensured that
follow-up corrective action was taken. The results of the TSA were submitted to the EPA.
4.4 Data Review
Records generated in the verification test received a one-over-one review before these records were
used to calculate, evaluate, or report verification results. Table 4-1 summarizes the types of data
recorded. A Battelle technical staff member involved in the verification test reviewed the data. The
person performing the review added his/her initials and the date to a hard copy of the record being
reviewed.
17
-------
Table 4-1. Summary of Data Recording Process
Data to Be
Recorded
Dates, times, and
details of test events,
BEAM 100
maintenance,
downtime, etc.
BEAM 100
calibration
information
BEAM 100 readings
Sample collection
and reference
method analysis
procedures,
calibrations, QA, etc.
Reference method
results
Responsible Party
Battelle
Battelle
Battelle
Battelle and others
assisting in sample
collection
Battelle
Where Recorded
ETV laboratory
record books (LRBs)
or data recording
forms
ETV LRBs, data
recording forms, or
electronically
Either recorded
electronically by the
BEAM 100 and
downloaded to an
independent
computer or hard
copy data printed by
the BEAM 100 and
taped into the ETV
LRB. Also hand
entered into ETV
LRBs or data
recording forms.
LRBs, COC, or other
data recording forms
Electronically or
manually into ETV
LRBs or data
recording forms.
Where possible at
least the same
number or a
maximum of one
number more
significant figures as
the BEAM 100 result
was reported for the
reference method.
How Often
Recorded
Start/end of test
procedure, and at
each change of a test
parameter or change
of BEAM 100 status
At BEAM 100
calibration or
recalibration
Recorded
continuously for
electronic data and
printed after each
measurement for
hard copy print-outs
and recorded
manually with each
reading
Throughout sampling
and analysis
processes
Every sample or QC
analysis
Disposition of
Data
Used to organize and
check test results;
manually
incorporated in data
spreadsheets as
necessary
Incorporated in
verification report as
necessary
Converted to or
manually entered
into spreadsheet for
statistical analysis
and comparisons
Retained as
documentation of
sample collection or
reference method
performance
Transferred to
spreadsheets for
calculation of results,
and statistical
analysis and
comparisons
18
-------
Chapter 5
Statistical Methods
The statistical methods used to evaluate the quantitative performance factors listed in Section 3.1
are presented in this chapter. Qualitative observations were also used to evaluate verification test
data.
5.1 Accuracy
Accuracy was determined by calculating the percent difference (PD) between the average of
triplicate CDOM measurements of a sample solution with the BEAM 100 (Mi) and the average of
triplicate BEAM equivalent CDOM measurement generated by a Varian Gary Eclipse spectrometer
(M2). As noted in Section B4 of the test/QA plan,(4) the CDOM ratios of the BEAM and reference
methods were not expected to be identical due to differences in grating efficiencies (the BEAM uses
filters to separate the light into the 430 and 460 nm wavelengths, whereas the Varian Eclipse
spectrometer uses gratings) and other conditions that can vary from instrument to instrument.
However, there should be a correlation between values obtained by the BEAM and the reference
method. This correlation was obtained by comparing the BEAM and Varian Gary Eclipse
spectrometer reference CDOM values for common concentration of quinine sulfate solution
measurements at 24°C. The regression statistics between the BEAM and Varian Gary Eclipse
spectrometer based on analyzing quinine sulfate solutions on both instruments were then used to
convert the Varian Gary Eclipse spectrometer CDOM values into BEAM equivalent values for
purposes of evaluating accuracy. The measurements were generated at a single temperature (i.e.,
data from 24°C measurements were used) for both PT and environmental samples using Equation 1.
The relationship between the BEAM and Varian Gary Eclipse spectrometer quinine sulfate curves
and additional information on how accuracy was determined using the BEAM and the adjusted
Varian Gary Eclipse spectrometer reference method CDOM values are discussed further in Section
6.1.
Mi- M
2
PD(%) = — x 100 (1)
PD values less than 20% were targeted as an acceptable demonstration of comparability between
the two measurements.
19
-------
5.2 Linearity
Linearity was determined by plotting the CDOM measurements (fluorescence values generated at a
single wavelength) while analyzing varying concentrations of analytes known to fluoresce (y-axis)
against the analyte concentration (x-axis) and performing linear curve fitting to determine the slope
(m) and intercept (b) in Equation 2.
y = mx + b (2)
•2\
Correlation coefficients such as the Pearson's r values and coefficient of determination (R ) values
,2
were calculated. A perfect regression line would have R values equal to 1
5.3 Precision
The standard deviation (S) of the results for the replicate analyses of the same sample was
calculated as follows:
r B i172
e 1 V (*' IT"9
S= 7L,(Mk-l
_n~ 1 k=l
(3)
where n is the number of replicate samples, A-4 is the CDOM measurement for the kih sample, and
M is the average CDOM measurement of the replicate samples. The BEAM 100 precision for each
sample was reported in terms of the relative standard deviation (RSD), which was calculated as
follows.
RSD(%) =
x 100 (4)
M
RSD values less than 10% were targeted as an acceptable indication of precise measurements.
5.4 Method Detection Limit (MDL)
The MDL was determined according to procedures described in Appendix B in Chapter 40 of the
Code of Federal Regulations Part 136 (40 CFR 136)(7) and assessed from seven replicate analyses
of a fortified sample. Fortified samples were generated by adding known fluorescing compounds
(quinine sulfate and SR fulvic acid) to Burdick and Jackson FtPLC grade water. The target analyte
was added at a concentration approximately five times Dakota's stated detection limit. The MDL
was calculated using Equation 5:
MDL = txS (5)
20
-------
where t is the Student's value of 3.143 for a 99% confidence level when the degrees of freedom
(N-l, where N equals the total number of measurements in the set) equals six, and S is the standard
deviation of the replicate samples.
5.5 Inter-unit Reproducibility
Inter-unit reproducibility was determined by evaluating the relative percent difference (RPD)
between the average of triplicate measurements for each sample tested using two separate units of
the BEAM 100. The equation for RPD, reported in percent, is as follows:
Mi-M2
RPD(%) = — —
Mi + A/2
x 200 (6)
where MI is the average of triplicate measurements made by the first BEAM 100 andM2 is the
average of triplicate measurements made by the second BEAM 100. RPD values less than 20%
were targeted as an indication of good agreement between the two units.
5.6 Temperature Effects
Temperature effects were determined by measuring the PD (using Equation 1) between the average
of triplicate measurements for each sample at 4°C and 34°C using the BEAM instruments (Mi)
against the average measurements at 24°C using the BEAM instruments (M2).
5.7 Matrix Effects
Matrix effects were determined by comparing the PD measurements between the BEAM results and
the reference method results for each type of sample used in testing (the two PT sample types:
quinine sulfate and fulvic acid, and the environmental samples). The PD measurements are
determined as described in Section 5.1. Trends in PD from the reference method were assessed
based on sample type.
5.8 Data Completeness
Data completeness was calculated as the percentage of the total possible data. Completeness was
determined by dividing the number of valid data measurements generated by each BEAM 100
(Mvaitd) by the total number of data measurements included in verification testing (Mtotai).
IVlvalid frl\
Completeness^/*) = x 100 (')
Mtotai
The cause of any substantial loss of data was established from operator observations or BEAM 100
records and noted in the discussion of the data completeness results.
21
-------
5.9 Operational Factors
There were no statistical calculations applicable to operational factors. Operational factors were
determined based on documented observations of the testing staff and the Verification Test
Coordinator.
22
-------
Chapter 6
Test Results
The results of the verification tests of the BEAM 100 are presented below for each of the
performance parameters.
6.1 Accuracy
Accuracy was determined by comparing the BEAM 100 CDOM measurements (calculated as the
intensity at 460 nm [F] divided by the intensity at 430 nm [R] or F/R) and the reference method F/R
results generated by the Varian Gary Eclipse spectrometer for all PT and environmental sample
analyses performed at 24°C. As shown in Figures 6-1, 6-2, and 6-3, the BEAM F/R measurements
tracked the reference method F/R measurements, but were offset (i.e., the reference measurements
ranged from approximately 8% to 26% higher than the BEAM measurements for quinine sulfate
and from approximately 20% to 40% higher than the BEAM measurements for fulvic acid and
environmental samples). Additionally the F/R measurements of both the BEAM and reference
method plateau at higher concentrations, possibly due to internal quenching. The F/R ratios of the
BEAM instruments and reference method instrument were not expected to be exactly the same
because of differences in type and efficiency of gratings, detectors, the light source, and other
conditions that vary from instrument to instrument. However, the instrumental differences can be
partially compensated for by correlating the BEAM and reference method results based on the
relationship between standards analyzed on each instrument. For ETV testing, quinine sulfate
standards were used to generate a correlation between the BEAMs and the reference method. This
relationship is shown in Figure 6-4, where the F/R measurements of both BEAM units are plotted
against the reference method F/R measurements for quinine sulfate. Using the polynomial
correlation between the BEAM F/R values and the reference method F/R values shown in Figure 6-
4 [BEAM equivalent F/R = 0.1159* (reference method F/R)2 + 0.7031 * (reference method F/R)],
the reference method F/R values for all of the test samples were converted to BEAM equivalent F/R
values to assess the accuracy of the BEAM measurements in comparison to reference method
measurements. Table 6-1 shows the PD between the BEAM equivalent reference method F/R
values and the BEAM F/R values for each BEAM unit tested. The PD was <20% for all test
samples except the unspiked solutions that were processed with the quinine sulfate and fulvic acid
PT samples. The PD values of unspiked solutions are not representative because their very low F/R
measurements result in small differences in F/R creating large PD values. In general, the results in
Table 6-1 show that the lower the CDOM value, the greater the variability in the result. There was
good agreement, however, between results for the same sample generated using two different
BEAM units. Inter-unit reproducibility is discussed further in Section 6.5.
23
-------
Quinine Sulfate at 24 °C
1 /inn -,
I .4-UU
1 nnn
I .UUU
«%> n Qnn
IL n Rnn
0.400
0.000 J
C
Zl
I
*
•
r»
iii ^^
) 20 40 60 80 100
Quinine Sulfate (ppb)
* Reference • BEAM 04 BEAM 08
Figure 6-1. Comparison of BEAM and Reference Method F/R Values for Quinine Sulfate
(prior to correlating the BEAM and reference method results using quinine sulfate)
Suwanee River (SR) Fulvic Acid at 24 °C
1 Ann -n
I .*HJU
1 onn
I .ZUU
1 nnn
I .UUU
-. n snn
>v u.ouu
U_ n r?nn
U.DUU
0/nn
.^•UU
0.200
0.000 '
(
*
r1
0
) 2000 4000 6000 8000
SR Fulvic Acid (ppb)
* Reference • BEAM 04 BEAM 08
t
1
10000
Figure 6-2. Comparison of BEAM and Reference Method F/R Values for SR Fulvic Acid
(prior to correlating the BEAM and reference method results using quinine sulfate)
24
-------
Environmental Samples at 24 °C
1 ROO -,
1 400
1 200
1 000
9r 0 ROO
0 ROO
0/inn
.4UU
0 900 -H
0 000
,
•
•
• r<
*»»*••
**•***""
^ r f
• Reference
• BEAM 04
BEAM 08
0 4 8 12
Environmental Site Location
Figure 6-3. Comparison of BEAM and Reference Method F/R Values for
Environmental Samples (prior to correlating the BEAM and reference method results using
quinine sulfate)
BEAM vs Reference F/R Measurements for
Quinine Sulfate
1.400
1.200 -
2 «
m
1.000
c 0.800
(7)
0.600
0.400
0.200
0.000 -SK
= 0.1159x2 + 0.7031x
R2 = 1
0.000 0.200 0.400 0.600 0.800
Reference F/R Signal
1.000
1.200
1.400
Figure 6-4. Polynomial Correlation Between BEAM F/R and Reference Method F/R
Measurements With Quinine Sulfate
25
-------
Table 6-1. Percent Difference
Reference Method F/R Values
Quinine Sulfate (ppb)
0
1
5
10
50
100
SR Fulvic Acid (ppb)
0
100
500
1000
5000
10000
Environmental Samples
(listed in order of increasing F/R)
MOOS-1
East Coast, FL-2
Puget Sound, WA
NASS-5
San Diego Harbor, CA
Massachusetts Bay NF7
New York Bight, NY
Narragansett Bay, RI
Long Island Sound, NY
New York Harbor, NY
East Coast, FL-1
Boston Harbor, MA
Between the BEAM F/R Values
PD
BEAM 04 (%)
53.8
4.7
0.8
0.8
0.3
0.1
41.9
11.1
19.9
18.7
11.8
8.6
11.6
17.4
16.1
11.3
16.8
16.9
15.0
16.9
13.8
9.3
3.1
7.9
and the BEAM Equivalent
PD
BEAM 08 (%)
18.3
3.2
0.4
0.1
0.3
0.2
9.2
11.1
18.8
18.3
11.3
8.3
14.0
17.9
15.0
12.2
15.3
15.0
13.6
13.2
10.7
7.8
2.4
6.2
26
-------
While the BEAM and reference method F/R measurements track each other and are related to each
other within a PD of 20% after correlating the BEAM and reference method measurements using
quinine sulfate standards, this offset between BEAM and reference method F/R values may indicate
that any action or cut-off F/R limits for BWE screening with the BEAM should be based on
historical values for open-ocean or coastal water that have been generated on a BEAM and not on
any other instrument unless the other instrument's data have been correlated to BEAM F/R values.
This is discussed further in the matrix effects Section 6.7. Additionally, while the test/QA plan(4)
defined using quinine sulfate to correlate the BEAM and reference method data due to its use as a
calibration standard, it should be noted that quinine sulfate is not the only standard that could be
used to relate the two different instruments. For example, a correlation could also have been made
using the fulvic acid standards, or other standards could have been selected that might be more
similar in composition to environmental samples. Use of different standards for correlation would
result in different PD values for accuracy between the BEAM and reference methods for the various
types of samples tested. For example, use of fulvic acid as the comparison standard would have
likely improved PD for fulvic acid and environmental samples, while resulted in large PD values
for quinine sulfate. For the purposes of ETV testing, only one standard, quinine sulfate, was used to
correlate the BEAM to a reference method; however, it should be understood that quinine sulfate is
not the only standard that could be used to correlate BEAM data to a reference method, nor is it
necessarily the standard which most closely represents actual ballast water samples.
6.2 Linearity
Because the F/R ratio values plotted against concentration are nonlinear as evidenced in
Figures 6-1, 6-2, and 6-3, linearity of the BEAM units as compared to the reference method was
determined by plotting the individual F and R measurements while analyzing varying
concentrations of analytes known to fluoresce (y-axis) against the analyte concentration (x-axis).
Figures 6-5 and 6-6 show how the linearity of each BEAM unit F or R value compares with the
linearity of the reference method F or R value. Because the signal output by the BEAM units and
the reference method are of different intensity, the reference method values were multiplied by a
factor of 1,000 to get the BEAM and reference F and R signals on the same scale. As demonstrated
in Figures 6-5 and 6-6, both the BEAM and reference method F and R signals were linear across the
concentration levels of quinine sulfate and fulvic acid analyzed, with R2 values greater than 0.99.
The slopes of the BEAM and reference method signals are different, however. A difference in slope
might be expected because of differences in the BEAM units and reference instrument (e.g.,
differences in type and efficiency of gratings, detectors, the light source, etc.).
Given the non-linearity of the F/R measurement with both the BEAM and reference method
instruments (shown in Figures 6-1, 6-2, and 6-3) and that the BEAM reporting unit is the F/R ratio,
it should be noted that it will be more difficult to distinguish between higher concentration CDOM
samples that have higher F/R values. Any ballast water screening action limits would need to avoid
the area in which the F/R ratios plateau.
27
-------
4^nnnn -,
4nnnnn
350000
c ^nnnnn
C/5
E
c onnnnn
S§ 1
-------
F (460 nm) vs Fulvic Acid Concentration
350000
300000
15 250000
O)
" 200000
£
1 150000
tO
a^
u- 100000
50000
0,
yBEAM04= 28.31 4x + 371 0.7
R2- 0.998 _<,
^f
,.*•-'
^..-c- "" yBEAMOS = 27.858X + 4125.6
^f* R2 = 0.9962
*""^'
******
^-""* ^_______-
y Reference= 1 0.584x + 1 1 83.2
J*-'' R2- 0.9991
^£-^
T* i i ii
0 2000 4000 6000 8000 10000
Fulvic Acid (ppb)
• BEAM 04 » BEAM 08 * Reference x 1000
R (430 nm) Signal
R (430 nm) vs Fulvic Acid Concentration
onnnnn
OUUUUU
ocnnnn
ZOUUUU
onnnnn
zuuuuu
1 cnnnn
IOUUUU
mnnnn
IUUUUU
cnnnn '
yBEAM04 = 21 .1 1 8x + 65902
R^= 0.9987
„ .- .-=-"• V"BEAM08= 20.705X + 64762
R^= 0.9959
•'^'•'"
^^-^"
,,--^"
„.-.'**'•
*-*'*"
*~* . ^ *
OUUUU yReference = 6.8048x + 1 4354
R2= 0.9922
r> T*"
0 2000 4000 6000 8000 10000
Fulvic Acid (ppb)
• BEAM 04 » BEAM 08 A Reference x 1 000
I inmr /'Poforonro v 1nnn^ I inmr fRF t\M nft^ I inmr /'RF/\ft/l n/1^
Figure 6-6. F (Upper Plot) and R (Lower Plot) Signals for the BEAM and Reference Method
Plotted Against Fulvic Acid Solution Concentration
29
-------
6.3 Precision
The precision among triplicate measurements evaluated as RSD was comparable between the
BEAM unit measurements and the reference method measurements at 24°C. Table 6-2 shows the
RSD for each of the triplicate measurements made during verification testing. At 24°C, the RSDs of
both the BEAM and reference measurements were less than 10%, with the exception of the
unspiked fulvic acid solution analyzed with the reference method. Because the raw values of
unspiked solutions were so low, small differences caused large RSDs. At the temperature extremes
(4°C and 34°C) where only BEAM measurements and not reference method measurements were
made, RSDs were less than 10% for all but the unspiked quinine sulfate solution (BEAM 03) and
the 1-ppb quinine sulfate solution (BEAM 04). Again, the lower concentration-solution RSDs were
affected by small differences in low-level measurements. The implication for BWE screening is that
BEAM measurements for lower concentration CDOM samples, such as open-ocean samples, will
be less precise than those for higher-concentration CDOM samples. However, out of 96 BEAM
measurements evaluated for precision, only three had BEAM RSD values greater than 10%.
6.4 Method Detection Limit
MDLs were evaluated using both quinine sulfate and fulvic acid solutions by measuring seven
replicates of each solution at concentrations five times the detection limit concentration specified by
Dakota Technologies, Inc. Quinine sulfate MDLs were evaluated using a 1-ppb solution and fulvic
acid MDLs were evaluated using a 100-ppb solution. MDL results for the BEAMs and reference
method are listed in Table 6-3. Note that the reference method results were not adjusted to BEAM
equivalent results using quinine sulfate for the MDL calculations. The calculated MDL F/R values
following this 40 CFR 136 Appendix B(7) approach for both the BEAMs and the reference method
are lower than the unspiked blanks analyzed with quinine sulfate and fulvic acid solutions, which
had F/R measurements of approximately 0.01. It is possible that the MDL calculated in this way
does not represent a practical detection limit, in part, because the water used to make up the
standard solutions, as purified as it is, fluoresces above the MDL of the instruments. For the
BEAMs, a practical MDL lies between the F/R values generated by the lowest concentration
standards analyzed (F/R = 0.07 for 1 ppb quinine sulfate and F/R = 0.06 for fulvic acid) and the F/R
of the unspiked blank samples (approximately 0.01) and is similar to the reference method (F/R =
0.09 for 1 ppb quinine sulfate, F/R = 0.1 for fulvic acid, and F/R = 0.01 for unspiked blank
solutions).
6.5 Inter-unit Reproducibility
The RPDs between measurements of aliquots of the same test solution using two different BEAM
units are shown in Table 6-4. Most measurements were within 10% RPD. However, the lower-
concentration solutions that resulted in low CDOM F/R measurements were affected by small
changes causing large RPDs. While RPDs were generally less than 10%, there was a noticeable
increase in RPD for measurements at 4°C and 34°C compared with those at 24°C. Excluding the
unspiked solutions, the RPDs between the two BEAM measurements of quinine sulfate and fulvic
acid averaged 6.9% at 4°C and 6.6% at 34°C, compared to 0.6% at 24°C.
30
-------
Table 6-2. Relative Standard Deviation of Triplicate Measurements with BEAMs and Reference Method
Sample Description
PT Samples
Quinine sulfate
prepared in Burdick
and Jackson HPLC
grade water per ASTM
E579-04(5)
Fulvic acid prepared in
Burdick and Jackson
HPLC grade water
Unspiked
1 ppb quinine sulfate
5 ppb quinine sulfate
10 ppb quinine sulfate
50 ppb quinine sulfate
100 ppb quinine sulfate
Unspiked
100 ppb SR fulvic acid
500 ppb SR fulvic acid
1,000 ppb SR fulvic acid
5,000 ppb SR fulvic acid
10,000 ppb SR fulvic acid
Environmental Samples
Location 1 -open ocean
Location 2-open ocean
Location 3 -coastal
Location 4-coastal
Location 5 -coastal
Location 6-coastal
Location 7-coastal
Location 8-coastal
Location 9-coastal
Location 10-coastal
Location 11 -coastal
Location 12-coastal
NASS-5
MOOS-1
Long Island Sound, NY
New York Harbor, NY
New York Bight, NY
East Coast, FL-1
East Coast, FL-2
San Diego Harbor, CA
Narragansett Bay, RI
Puget Sound, WA
Massachusetts Bay NF7
Boston Harbor, MA
RSD (%)
24°C
Reference
Method
3.4
0.7
1.3
0.3
0.4
0.2
26.0
2.6
1.2
1.6
0.2
0.1
BEAM
04
0.0
0.8
0.6
0.4
0.3
0.2
0.0
0.0
0.5
0.9
0.5
0.7
BEAM
08
5.6
0.8
0.6
0.7
0.4
0.3
5.1
2.0
1.2
0.6
0.1
0.3
4°C
BEAM
04
6.7
22.9
1.6
1.5
0.3
0.4
0.0
3.5
1.2
1.2
2.2
0.9
BEAM
03
10.2
6.1
9.1
2.3
0.3
0.5
6.7
6.0
3.1
1.6
2.9
1.3
34 °C
BEAM
04
6.2
5.1
2.8
1.1
0.9
1.1
6.9
1.3
2.3
0.8
1.2
2.5
BEAM
0.3
22.9
2.4
0.9
0.5
0.6
1.1
6.3
0.9
0.5
2.4
0.4
1.9
3.3
0.6
1.0
0.5
1.3
0.1
0.9
0.8
0.8
0.1
1.1
0.3
0.9
0.0
0.4
0.4
1.1
0.7
0.4
0.9
1.1
1.4
1.6
0.7
1.6
1.1
0.5
0.1
0.9
0.5
0.4
1.3
0.4
0.4
0.6
0.1
Yellow highlight indicates RSD > 10%.
-------
Table 6-3. Method Detection Limits
Sample Description
PT Samples
1 ppb Quinine sulfate
prepared in Burdick
and Jackson HPLC
grade water per
ASTM E579-04(5)
100 ppb Fulvic acid
prepared in Burdick
and Jackson HPLC
grade water
REPLICATE-1
REPLICATE-2
REPLICATE-3
REPLICATE-4
REPLICATE-5
REPLICATE-6
REPLICATE-7
SD
MDL
Unspiked blank
REPLICATE-1
REPLICATE-2
REPLICATE-3
REPLICATE-4
REPLICATE-5
REPLICATE-6
REPLICATE-7
SD
MDL
Unspiked blank
CDOM F/R Value
24°C
Reference Method
0.0895
0.0955
0.0936
0.0930
0.0952
0.0939
0.0962
0.0022
0.0070
0.012
0.1001
0.1002
0.1001
0.0999
0.1003
0.0981
0.0979
0.0011
0.0033
0.013
BEAM 04
0.071
0.072
0.073
0.072
0.072
0.071
0.072
0.00069
0.0022
0.013
0.062
0.062
0.062
0.062
0.061
0.062
0.061
0.00049
0.0015
0.013
BEAM 08
0.071
0.072
0.072
0.071
0.072
0.071
0.070
0.00076
0.0024
0.010
0.064
0.062
0.062
0.061
0.060
0.061
0.058
0.0019
0.0059
0.010
Because one BEAM unit (BEAM 08) was exchanged between the 24°C testing and the
temperature extreme testing (BEAM 03), it is not clear whether the differences in RPD values
result from the temperature extremes or other differences between BEAM 08 and BEAM 03.
However, the inter-unit reproducibility was generally within 10% overall.
6.6 Temperature Effects
The BEAM units and the test solutions were equilibrated at 4°C and 34°C for a minimum of
30 minutes prior to taking any measurements. Once equilibrated at the appropriate temperature,
the BEAM units were calibrated with the 10-ppb quinine sulfate calibration solution. The
calibration procedure set the F/R value for the 10-ppb quinine sulfate calibration solution at a
value of 0.43. The upper plot in Figure 6-7 shows that, at 4°C, the BEAM quinine sulfate
measurements agreed quite well with the 24°C measurements. At 34°C, however, the BEAM
quinine sulfate measurements begin to diverge from the 24°C measurements, especially at F/R
values greater than 0.5. This is also shown in Table 6-5, which lists the PD values between the
temperature extreme measurements and the 24°C measurements. When measurements at 4°C are
compared with those at 24°C for quinine sulfate, all PD values are less than 10%, with the
exception of the unspiked solution. However, when measurements at 34°C are compared with
those at 24°C, the PD values are less than 10% only for the 5-ppb and 10-ppb quinine sulfate
solutions, which had F/R values near the 0.43 F/R calibration level. As the F/R values move
away from the calibrated 0.43 level, the PD between the 34°C and 24°C measurements
32
-------
Table 6-4. BEAM Inter-unit Reproducibility
Sample Description
PT Samples
Quinine sulfate
prepared in Burdick
and Jackson HPLC
grade water per
ASTM E579-04(5)
Fulvic acid prepared in
Burdick and Jackson
HPLC grade water.
unspiked
1 ppb quinine sulfate
5 ppb quinine sulfate
10 ppb quinine sulfate
50 ppb quinine sulfate
100 ppb quinine sulfate
unspiked
100 ppb SR fulvic acid
500 ppb SR fulvic acid
1000 ppb SR fulvic acid
5000 ppb SR fulvic acid
10000 ppb SR fulvic acid
Environmental Samples
Location 1 -open ocean
Location 2-open ocean
Location 3 -coastal
Location 4-coastal
Location 5-coastal
Location 6-coastal
Location 7-coastal
Location 8-coastal
Location 9-coastal
Location 10-coastal
Location 11 -coastal
Location 12-coastal
NASS-5
MOOS-1
Long Island Sound, NY
NY Harbor, NY
NY Bight, NY
East Coast FL-1
East Coast FL-2
San Diego Harbor, CA
Narragansett Bay, RI
Puget Sound, WA
Massachusetts Bay NF7
Boston Harbor, MA
RPD (%)
24°C
BEAM 04
vs. BEAM
08
26.1
1.4
0.4
0.7
0.6
0.3
26.1
0.0
1.3
0.5
0.6
0.3
4°C
BEAM 04
vs. BEAM
03
14.0
12.4
7.5
10.4
5.5
7.6
18.2
3.4
6.2
4.3
7.8
4.0
34°C
BEAM 04
vs. BEAM
03
92.3
17.2
2.2
5.8
0.8
0.8
57.8
20.7
9.8
3.2
0.0
5.9
1.0
2.8
3.5
1.6
1.6
0.8
0.7
1.7
4.3
1.2
2.2
1.8
Yellow highlight indicates RPD> 10%.
increases. This divergence away from the 0.43 calibration level illustrates that a sample's
fluorescence differs with temperature. This is a typical fluorescence phenomenon, with lower
temperatures increasing fluorescence and higher temperatures decreasing fluorescence, and
should not be interpreted as a function of the BEAM units.
Similar comparisons for fulvic acid are shown in the lower plot in Figure 6-7 and the PD data
listed in Table 6-5. As for quinine sulfate, fulvic acid measurements at 34°C agree well near the
calibrated 0.43 F/R level, but the PD increases as the F/R values move away from the calibrated
0.43 level.
At 4°C, considerable PD (26% and 98%) exists between the fulvic acid F/R values and the 24°C
measurements across all solutions tested.
33
-------
BEAM Quinine Sulfate Measurements at 4 °C and 34 °C vs 24 °C
1.400
O 1 -200 f
o
CO
o 1.000
o
^
4-1
9; 0.800
0.600
0.400
0.200
0.000
• 34C BEAM03
A 34C BEAM04
+ 4C BEAM03
X 4C BEAM04
Poly. (4C BEAM04)
— - Poly. (34C BEAM04)
Poly. (34C BEAM03)
— - Poly. (4C BEAM03)
0.000
0.200
0.400 0.600 0.800
BEAM Quinine Sulfate FIR at 24 °C
1.000
1.200
BEAM Fulvic Acid Measurements at 4 °C and 34 °C vs 24 °C
1.600
1.400
O
*r
CO
,_ 1.200
O
o
•s
a:
LL
1.000
0.800
0.600
0.400
00
0.200
0.000
0.000
A 34C BEAM03
A 34C BEAM04
O 4C BEAM03
D 4C BEAM04
Poly. (4C BEAM04)
— Poly. (4C BEAM03)
- - Poly. (34C BEAM03)
Poly. (34C BEAM04)
0.200
0.400 0.600 0.800
BEAM Fulvic Acid F/R at 24°C
1.000
1.200
Figure 6-7. Comparison of BEAM F/R Values for Quinine Sulfate (Upper Plot) and Fulvic
Acid (Lower Plot) at 4°C and 34°C to Those at 24°C
34
-------
Table 6-5. Percent Difference of 4°C and 34°C BEAM Measurements from 24°C BEAM
Measurements
Sample Description
PT Samples
Quinine sulfate
prepared in
Burdick and
Jackson HPLC
grade water per
ASTM E579-04(5)
Fulvic acid
prepared in
Burdick and
Jackson HPLC
grade water
Unspiked
1 ppb quinine sulfate
5 ppb quinine sulfate
10 ppb quinine sulfate
50 ppb quinine sulfate
100 ppb quinine sulfate
unspiked
100 ppb SR fulvic acid
500 ppb SR fulvic acid
1,000 ppb SR fulvic acid
5,000 ppb SR fulvic acid
10,000 ppb SR fulvic acid
Percent Difference (%)
Avg. BEAM 03/04 at 4°C
vs. Avg. BEAM 08/04 at
24°C
87.0
2.8
0.4
0.7
5.9
6.7
43.5
98.3
60.0
42.5
25.9
30.0
Avg. BEAM 03/04 at 34°C
vs. Avg. BEAM 08/04 at
24°C
13.0
31.9
8.8
0.9
16.2
19.5
95.7
22.9
5.8
2.1
14.2
18.8
These results illustrate the importance of calibrating the BEAM units at the testing temperature
and ensuring that the calibration solution has an F/R close to any action level developed for
BWE screening to maximize the accuracy of measurements near action-level concentrations. The
results also underscore the fact that temperature can cause deviations in F/R values in the PT
sample solutions, particularly when the solution F/R values differ from the 0.43 calibration level.
Additionally, the differences between the quinine sulfate and fulvic acid F/R responses at the
various temperatures suggest that temperature effects may vary depending on the sample
composition.
6.7 Matrix Effects
The PD data in Table 6-1 (excluding the data for the unspiked blanks), which is also displayed
graphically in Figure 6-8, show that fluorescence observed by the BEAM and reference method
instruments is affected by the matrix. As noted in Section 6.1 above, the reference method
results were converted to BEAM equivalent results based on the correlation of the BEAM and
reference method when analyzing quinine sulfate. Using the reference method results adjusted
based on the quinine sulfate correlation, the environmental samples and the fulvic acid samples
yield a different relationship to the reference method measurements (PDs ranging from
approximately 2 to 20%) than the quinine sulfate samples (all PDs less than 5%). Differing
results based on matrix are not unexpected based on instrumental differences between the BEAM
and the reference method (i.e., use of gratings in the reference method instrument versus using
filters in the BEAM) and how different compounds fluoresce. The amount of fluorescence
depends upon the amount of absorption occurring with all of the species at the excitation
wavelength. The more spectroscopically complex the matrix, the greater the variability is likely
to be. For example, compounds present in the fulvic acid and environmental samples may
obscure the excitation energy, causing the fluorescing compounds to fluoresce less than with a
simple standard such as quinine sulfate. The composition of different environmental samples can
also result in different amounts of fluorescence quenching. Changes in quenching will change
35
-------
both shape and intensity of fluorescence signals. Seawater is a very complex mixture of
compounds, and so matrix effects could be expected.
%PD Fulvic Acid , Environmental and Quinine Sulfate Sample F/R
BEAM vs BEAM Equivalent Reference Method
25.00%
20.00%
^ 15.00%
V
o
10.00%
Q
•£
u
5.00%
0.00%
x x
X
•*-& . ^
V
• FA Beam 04
& FA Beam 08
x EV Beam 04
x EV Beam 08
• QS Beam 04
• QS Beam 08
Poly. (QS Beam 04)
Poly. (EV Beam 08)
- - - Poly. (FA Beam 04)
Poly. (FA Beam 08)
— - Poly. (EV Beam 04)
Poly. (QS Beam 08)
x
X
X
0.000
0.200
0.400
10.600
0.8001
1.000
1.200
1.400
-5.00%
1.600
F/R Values
Figure 6-8. Matrix Effects Based on Percent Difference of BEAM Results Compared with
BEAM Equivalent (Based on Quinine Sulfate Correlation) Reference Method Results
When using quinine sulfate to correlate the BEAM to the reference method, the differences in the
way the two instruments respond to different matrices (and not compounding with temperature
effects, etc.) result in the environmental sample BEAM F/R values varying approximately 15%
from the Varian Gary BEAM equivalent (quinine sulfate adjusted) reference method F/R values
for samples with F/R values <0.6. As discussed previously, the F/R ratios of the BEAM and
reference method instruments were not expected to be the same because of differences in type
and efficiency of gratings, detectors, the light source, and other conditions that vary from
instrument to instrument. This would be the case between any hand-held fluorimeter compared to
any laboratory bench-top spectrometer and is not unique to the BEAM. The implication of this
for BWE screening is that, among instruments, target values likely will be different for each
instrument design. This implies a need to set BWE action limits on an instrument specific basis.
Likewise, any comparison of screening instrument results to those generated using a laboratory
based reference method will be more accurate by correlating the screening instrument to the
reference method based on the relationship between the same standards analyzed on each
instrument, as was conducted for this verification test. While quinine sulfate was used for this
verification test, it is beyond the scope of this test to determine what standard serves as the best
for correlating ballast water screening results between different instruments for any kind of
regulatory purpose.
36
-------
6.8 Data Completeness
Data completeness for this verification test was 100%. All data measurements expected to be
taken were completed and were usable.
6.9 Operational Factors
The BEAM units were easy to operate. The data display was easy to read, and the data were easy
to download to a PC. Testing staff received a 4-hour training session from Dakota Technologies,
Inc., which was more than sufficient to familiarize the staff with BEAM operation and data
downloading procedures. The BEAM units contained an instruction manual with clearly written
information and illustrations. While the instructions for each specific procedure such as blank
calibrations, quinine sulfate calibrations, and calibration checks were easy to follow, more
information on required frequency of each procedure and QC pass/fail criteria for the procedure
would be useful to ensure that the operator knows whether the BEAM units are functioning
properly. Such limits were agreed upon with Dakota Technologies, Inc. for use during
verification testing (e.g., acceptable CDOM measurement ranges for the calibration checks,
maximum readings for negative control samples, etc.) and were useful for ensuring that
instrumentation was clean and operating properly. Contents of the BEAM kit were listed in the
instruction manual; however, it would be useful for the actual vial containing cleaning solution
to be labeled. From a safety perspective, it would also be useful to identify for the user, either on
the vial or in the instruction manual, any hazard associated with the cleaning solution or special
precautions necessary in case of spillage or user exposure. Instructions include information for
storing the BEAM cell with clean water to prevent spotting, but no information for storing the
cleaning solution. Given the potential for the BEAM units to be used at temperature extremes
during practical application, guidance as to any precautions for storage under such conditions
may be useful. Storage conditions for the quinine sulfate calibration solution, which is not
included in the BEAM kit but needed for operation, are discussed in the ASTM method
referenced in the instruction manual, but are not discussed in the instruction manual itself.
All items included in the BEAM kit were easy to open. The Luer-lock syringe provided with the
BEAM kit required considerable hand strength for processing large quantities of samples. A user
may want to investigate other types of syringes if planning to process more than two or three
samples in a short time period. Reagents were easy to prepare; however, since the 10-ppb
quinine sulfate calibration solution was made daily following the ASTM guidance, preparation
of calibration solution took a fair amount of time each analysis day during verification testing.
With the exception of the 10-ppb quinine sulfate calibration solution and distilled water for
rinsing, all reagents were supplied with the kit.
The BEAM kit provided a container for rinse water, but the BEAM carrying case did not provide
for the calibration solution. Assuming that the calibration solution must be carried along with the
BEAM for use in the field, a container or holding spot for this solution in the carrying case
would be useful. The BEAM kit included all necessary equipment, with the exception of
pipettes, flasks, balances, and containers used to prepare and store the quinine sulfate calibration
solution and to store waste. The BEAM unit's exterior was easily wiped clean. Other than
keeping the cell clean, no routine maintenance was required. Approximately 12 milliliters (mL)
of sample waste and 20 mL of water rinse waste were generated with every sample
measurement.
37
-------
The number of samples that can be processed continuously with a BEAM, assuming unlimited
access to distilled water for rinsing the BEAM cells, is limited by the life of the batteries in the
BEAM and by the BEAM'S internal memory size. During verification testing, the six AA
batteries were replaced once after processing approximately 100 samples. A spare set of batteries
in the BEAM kit would be useful. Spare batteries were supplied in the BEAM kits provided for
verification testing, but are not listed in the instruction manual as standard items with the kit. The
BEAM's internal memory will hold 256 measurements before it overwrites previous readings.
This is not an issue if data are also recorded manually or downloaded to a PC. If the operator
must rely only on the distilled water in the provided rinse bottle, approximately 15 samples could
be processed before more water would be needed. This estimate assumes that the cell and cap are
rinsed with water twice between each analysis. Overall sample throughput was between 20 and
25 samples per hour (approximately three samples per minute) when manually recording the
CDOM F/R value and properly rinsing the cell between each sample reading.
During verification testing, some technical difficulties were encountered with the data displays
and system interlocks. For example, after tightening down the cell cap, an interlock countdown
occurs before a measurement reading can be taken. This interlock prevents light-leaks that could
impact measurements. However, in some instances, this interlock countdown would repeat
multiple times before a measurement reading would be taken, even though the cap was tightly
secured. Display errors observed during verification testing included the 460-nm column header
sometimes not displaying properly and calibration values sometimes appearing in the display
instead of sample results. The frequency of technical difficulties increased when operating at the
temperature extremes. At 34°C, one of the BEAM units (BEAM 08) had unusual calibration
results with the 10-ppb quinine sulfate solution and subsequently generated system errors every
time a sample analysis was attempted. When cooled to 24°C, the calibration and sample readings
were successful, but failed again when exposed to 34°C a second time. Similar problems with
this unit occurred with 4°C testing. This particular unit was replaced before the 34°C and 4°C
testing. Dakota Technologies, Inc. provided phone support for troubleshooting and replacement
equipment when difficulties were encountered during verification testing. The replacement
BEAM unit (BEAM 03) did not have problems at the temperature extremes; however, this unit
did have difficulties with the interlock issues described above. The number of BEAM units
tested was too small to determine whether the problems with BEAM 08, potentially induced by
temperature extremes, is a weakness in the instrumentation or a random, chance instrument
failure. It also should be noted that the BEAMs power down after 7 minutes of idle time and then
require a 90-count warm-up period before restarting, which can add to analysis time.
38
-------
Chapter 7
Performance Summary
Table 7-1. BEAM 100 Summary Table
Performance
Factor
Sample Information
Result
Accuracy
Five concentrations of quinine sulfate prepared in
Burdick and Jackson HPLC grade water per
ASTM E579-04(5) plus one unspiked blank; five
concentrations of SR fulvic acid plus one unspiked
blank; and 12 environmental (natural water)
samples.
All testing was performed at approximately 24°C.
PD from reference method measurements (using a
quinine sulfateA correlation between the BEAM and
reference method results) was less than 20% for both
quinine sulfate and fulvic acid samples, except for the
unspiked, blank samples. PD was also less than 20%
for environmental samples. PD values increased with
lower CDOM measurements.
Linearity
Five concentrations of quinine sulfate prepared in
Burdick and Jackson HPLC grade water per
ASTM E579-04(5) plus one unspiked blank; five
concentrations of SR fulvic acid plus one unspiked
blank.
All testing was performed at approximately 24°C.
Individual signals at 460 nm and 430 nm were linear
across the concentrations tested and had R2 values
>0.99 for both quinine sulfate and fulvic acid test
solutions.
Precision
Five concentrations of quinine sulfate prepared in
Burdick and Jackson HPLC grade water per
ASTM E579-04(5) plus one unspiked blank; five
concentrations of SR fulvic acid plus one unspiked
blank; and 12 environmental (natural water)
samples. Testing was performed at approximately
24°C, 4°C, and 34°C.
RSD of triplicate measurements of each test sample
was <10% except for low CDOM concentration
samples such as the unspiked blank samples for which
the highest RSD was 22.9%.
MDL
Seven replicates of 1 ppb quinine sulfate and
seven replicates of 100 ppb SR fulvic acid
analyzed following 40 CFR 136 Appendix B(7)
procedures. Concentrations were set at five times
the vendor-specified detection limit for each
compound.
All testing was performed at approximately 24°C.
Calculated MDLs were lower than the CDOM values
of the unspiked blank samples (<0.01) and, therefore,
may not represent practical detection limits. The
BEAMs proved capable of detecting CDOM values
<0.06 to 0.07, which were the CDOM values of the
lowest concentration quinine sulfate and fulvic acid
standards analyzed.
39
-------
7-1. BEAM 100 Summary Table (continued)
Inter-unit
Repro-
ducibility
All test samples. Testing was performed at
approximately 24°C, 4°C, and 34°C.
RPD values between the average of triplicate
measurements were mostly <10% at all testing
temperatures. RPD increased as CDOM concentration
decreased.
Temperature
Effects
Five concentrations of quinine sulfate prepared in
Burdick and Jackson HPLC grade water per
ASTM E579-04(5) plus one unspiked blank; five
concentrations of SR fulvic acid plus one unspiked
blank.
Testing was performed at temperature extremes of
approximately 4°C and 34°C and compared with
results obtained at approximately 24°C (ambient
conditions).
For the spiked samples, PD values ranged as follows:
QS solutions: 0.4 to 6.7% for 4°C vs 24°C
0.9 to 31.9% for34°Cvs24°C
SRFA solutions: 25.9 to 98.3% for 4°C vs 24°C
2.1 to 22.9% for 34°C vs 24°C
For the unspiked blanks, the PD values ranged from
13.0 to 95.7%.
The results indicate that temperature changes can cause
deviations in performance and illustrate the importance
of calibrating the BEAM units at the testing
temperature.
Matrix Effects
Five concentrations of quinine sulfate prepared in
Burdick and Jackson HPLC grade water per
ASTM E579-04(5) plus one unspiked blank; five
concentrations of SR fulvic acid plus one unspiked
blank; and 12 environmental (natural water)
samples.
All testing was performed at approximately 24°C.
The accuracy PD measurements comparing
BEAM CDOM to reference method measurements
(using a quinine sulfate correlation between the
BEAM and reference method results) of the same
solution were evaluated for differences between
matrix type.
Distinct differences in correlation to reference method
values were observed based on matrix type.
Environmental samples and fulvic acid samples were
between 2 and 20% PD from BEAM equivalent
reference method measurements (using a quinine
sulfateA correlation between the BEAM and reference
method results), whereas quinine sulfate samples were
all less than 5% PD.
Data
Completeness
All test samples.
Data completeness was 100%. All intended analyses
and measurements were performed and all
measurements were valid and usable.
Operational
Factors
The BEAM 100 units were portable, convenient, easy to use, and came with a clearly written instruction
manual. Sample throughput was 20 to 25 samples per hour. Waste generated while processing each sample
included approximately 12 mL of sample waste and 20 mL of water rinse waste. Factors limiting
continuous operation of the BEAM include battery life (six AA batteries had to be replaced after
100 measurements), BEAM internal memory size (data are overwritten after 256 measurements), access
to distilled water (rinse bottle provided with BEAM holds enough distilled water for ~15 samples), and
operator hand strength (each sample must be filtered through a 0.45-micron filter using a Luer-lock
syringe). Technical difficulties with displays and system interlocks resulted in one BEAM unit being
replaced by the vendor during testing. Technical difficulties increased when testing at approximately 4°C,
and 34°C. Not enough BEAM units were evaluated to know whether these technical difficulties indicate
more than a random instrument failure.
A Quinine sulfate was selected to correlate the BEAM and reference instruments because of its use as a spectroscopic
standard. Use of other standards with properties closer to the environmental samples may have improved PD values for
the environmental samples; however, this was not verified as part of this test.
40
-------
Chapter 8
References
1. Hunt, Carlton D.; Tanis, Deborah; Bruce, Elizabeth; and Taylor, Michael. In Press. "Optical
Signatures of Seawater and their Potential Use in Verification of At Sea Ballast Water
Exchange," Marine Ecology and Progress Series.
2. Murphy, K.; Ruiz, G.; Dunsmuir, W.; Waite, T. "Optimized Parameters for Fluorescence-
Based Verification of Ballast Water Exchange by Ships," Environ. Sci. Technol., 40, 2357-
2362, 2006.
3. Murphy, K.; Boehme, 1; Coble, P.; Cullen, 1; Field, P.; Moore, W.; Perry, E.; Sherrell, R.;
and Ruiz, G. "Verification of mid-ocean ballast water exchange using naturally occurring
coastal tracers," Marine Pollution Bulletin, 48, 711-730, 2004.
4. Test/QA Plan for Verification of Ballast Water Exchange Screening Tools, Battelle,
Columbus, Ohio, January 2007.
http://www.epa.gov/etv/pdfs/testplan/600etv07004/600etv07004.pdf
5. ASTM E579-04, "Standard Test Method for Limit of Detection of Fluorescence of Quinine
Sulfate in Solution," ASTM International, 2007.
6. Quality Management Plan (QMP)for the ETV Advanced Monitoring Systems Center.,
Version 6.0, U.S. EPA Environmental Technology Verification Program, Battelle,
Columbus, Ohio, November 2005. http://www.epa.gov/etv/pdfs/qmp/ams qmp ver6.pdf
7. Code of Federal Regulations, Title 40, Part 136, Appendix B, Definition and Procedure for
the Determination of the Method Detection Limit-Revision 1.11, 2006.
41
------- |