June 2000
Environmental Technology
Verification Report
SlGRIST
WTM500
On-Line Turbidimeter
Prepared by
Baffelle
Putting Technology To Work
Battel le
Under a cooperative agreement with
SEPA U.S. Environmental Protection Agency
ETtf ETtf E
-------
June 2000
Environmental Technology Verification
Report
ETV Advanced Monitoring Systems Pilot
Sigrist
WTM500 On-Line Turbidimeter
By
Kenneth Cowen
Thomas Kelly
Brian Canterbury
Karen Riggs
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 and recommended for public release.
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 (ORD) 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. At present, there are twelve environmental technology areas
covered by ETV. Information about each of the environmental technology areas covered by ETV
can be found on the Internet at http://www.epa.gov/etv.htm.
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 assess-
ment. In 1997, through a competitive cooperative agreement, Battelle was awarded EPA funding
and support 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/07/07_main.htm.
111
-------
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. In particular we would like to thank the
staff at the Dublin Road Water Plant, including Tom Camden and Terry Nichols. We also
acknowledge the participation of T.J. Medland of Peak Process Controls, Inc., and Dr. Arnd
Rogner of Sigrist in this verification test. We would like to thank the Hach Company for
supplying the two reference turbidimeters used in this test.
IV
-------
Contents
Notice ii
Foreword iii
Acknowledgments iv
List of Abbreviations x
1. Background 1
2. Technology Description 2
3. Test Design and Procedures 3
3.1 Introduction 3
3.2 Test Design Considerations 3
3.3 Experimental Apparatus 4
3.4 Reference Instruments 6
3.5 Off-Line Testing 7
3.5.1 Linearity 7
3.5.2 Accuracy and Precision 9
3.5.3 Water Temperature 9
3.5.4 Flow Rate 9
3.5.5 Color 10
3.6 On-Line Testing 10
3.6.1 Accuracy 11
3.6.2 Drift 11
4. Quality Assurance/Quality Control 13
4.1 Data Review and Validation 13
4.2 Deviations from the Test/QA Plan 13
4.3 Calibration 14
4.3.1 Reference Turbidimeters 14
4.3.2 Temperature Sensors 15
v
-------
4.3.3 Flow Meters 15
4.3.4 pHMeter 15
4.4 Data Collection 15
4.5 Assessments and Audits 16
4.5.1 Technical Systems Audit 16
4.5.2 Performance Evaluation Audit 16
4.5.3 Verification Test Data Audit 18
4.6 Audit Reporting 19
5. Statistical Methods 20
5.1 Off-Line Testing 20
5.1.1 Linearity 20
5.1.2 Accuracy 20
5.1.3 Precision 21
5.1.4 Water Temperature Effects 21
5.1.5 Flow Rate Sensitivity 21
5.1.6 Color Effects 21
5.2 On-Line Testing 22
5.2.1 Accuracy 22
5.2.2 Drift 22
6. Test Results 23
6.1 Off-Line Testing 23
6.1.1 Linearity 23
6.1.2 Accuracy 25
6.1.3 Precision 26
6.1.4 Water Temperature Effects 28
6.1.5 Flow Rate 30
6.1.6 Color Effects 31
6.2 On-Line Testing 33
6.2.1 Accuracy 33
6.2.2 Drift 35
vi
-------
6.3 Other Performance Parameters 40
6.3.1 Cost 40
6.3.2 Maintenance/Operational Factors 40
7. Performance Summary 41
8. References 43
Appendix A: Example Data Recording Sheet A-l
Appendix B: Technical Systems Audit Report B-l
Figures
Figure 2-1. Sigrist WTM500 On-Line Turbidimeter 2
Figure 3-1. Schematic Representation of Recirculation System 5
Figure 4-la. Control Chart for Performance Evaluation Calibration
Checks of ISO 7027 Reference Turbidimeter 17
Figure 4-lb. Control Chart for Performance Evaluation Calibration
Checks of Method 180.1 Reference Turbidimeter 17
Figure 6-la. Linearity Plot for Sigrist WTM500 Turbidimeter vs.
ISO 7027 Reference Turbidimeter 24
Figure 6-lb. Linearity Plot for Sigrist WTM500 Turbidimeter vs.
Method 180.1 Reference Turbidimeter 24
Figure 6-2a. Effect of Temperature on Sigrist WTM500 Turbidity Readings
vs. ISO 7027 at Both 0.3 and 5 NTU 29
Figure 6-2b. Effect of Temperature on Sigrist WTM500 Turbidity Readings
vs. Method 180.1 at Both 0.3 and 5 NTU 29
Figure 6-3. Effect of Sample Flow Rate on Sigrist WTM500 Turbidimeter Response 30
Figure 6-4a. Effect of Color on Relative Turbidity at 0.1 NTU with the
Sigrist WTM500 vs. ISO 7027 at Both 0.1 and 5 NTU 32
Figure 6-4b. Effect of Color on Relative Turbidity at 5 NTU with the
Sigrist WTM500 vs. Method 180.1 at Both 0.1 and 5 NTU 32
vii
-------
Figure 6-5. Summary of Stream Turbidity Data from
On-Line Testing of Sigrist WTM500 34
Figure 6-6. Twice-Weekly Calibration Checks During On-Line Testing of the
Sigrist WTM500 36
Figure 6-7a. Final Linearity Plot for Sigrist WTM500 vs. ISO 7027
Reference Turbidimeter 38
Figure 6-7b. Final Linearity Plot for Sigrist WTM500 vs. Method 180.1
Reference Turbidimeter 38
Tables
Table 3-1. Performance Characteristics Evaluated and Schedule of Verification Test 3
Table 3-2. Summary of Measurements for Off-Line Testing 8
Table 3-3. Summary of Measurements for On-Line Testing 11
Table 4-1. Results of Linearity Check of Reference Turbidimeters 14
Table 4-2. Summary of Flow Meter Calibration Check 15
Table 4-3. Results of Calibration Checks of Thermocouple Used
in the Verification Test 18
Table 6-1. Statistical Results of Initial Linearity Test on the
Sigrist WTM500 Turbidimeter 25
Table 6-2. Bias of Sigrist WTM500 Turbidimeter Relative to
Reference Measurements on Prepared Test Solutions 26
Table 6-3. Adjusted Turbidity Readings for Precision Calculations
on the Sigrist WTM500 27
Table 6-4. Precision of Sigrist WTM500 Turbidimeter and of the
Reference Turbidimeters 27
Table 6-5. Statistical Results of Temperature Test on the
Sigrist WTM500 Turbidimeter 28
Vlll
-------
Table 6-6. Statistical Results of Flow Rate Test for the
Sigrist WTM500 Turbidimeter 31
Table 6-7. Statistical Results of the Color Test with the
Sigrist WTM500 Turbidimeter 33
Table 6-8. On-Line Daily Accuracy Check Results from Water Stream Samples 35
Table 6-9. Results of Calibration Checks Performed During On-Line Testing 36
Table 6-10. Statistical Results of Final Linearity Test 39
Table 6-11. Comparison of Results from Linearity Tests at Beginning and
End of Verification 39
IX
-------
List of Abbreviations
AMS
Advanced Monitoring Systems
CU
color unit
DC
direct current
DRWP
Dublin Road Water Plant
EPA
U.S. Environmental Protection Agency
ETV
Environmental Technology Verification
FNU
formazin nephelometric unit
gpm
gallons per minute
LOD
limit of detection
NIST
National Institute of Standards and Technology
NPT
normal pipe thread
NTU
nephelometric turbidity unit
OD
outer diameter
PE
performance evaluation
QA
quality assurance
QC
quality control
QMP
Quality Management Plan
RSD
relative standard deviation
X
-------
Chapter 1
Background
The U.S. Environmental Protection Agency (EPA) has created the Environmental Technology
Verification (ETV) Program to facilitate the deployment of innovative environmental tech-
nologies through performance verification and dissemination of information. The goal of the
ETV Program is to further environmental protection by substantially accelerating the acceptance
and use of improved and cost-effective technologies. ETV seeks to achieve this goal by provid-
ing high quality, peer-reviewed data on technology performance to those involved in the design,
distribution, permitting, purchase, and use of environmental technologies.
ETV works in partnership with recognized testing organizations; with stakeholder groups
consisting of regulators, buyers and vendor organizations; and with the full participation of
individual technology developers. The program evaluates the performance of innovative tech-
nologies 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
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 and its verification organization partner,
Battelle, operate the Advanced Monitoring Systems (AMS) pilot under ETV. The AMS pilot has
recently evaluated the performance of on-line turbidimeters for use in water treatment facilities.
This verification report presents the procedures and results of the verification test for the Sigrist
WTM500 on-line turbidimeter.
1
-------
Chapter 2
Technology Description
The following description of the Sigrist WTM500 turbidimeter is based on information provided
by the vendor.
The WTM500 is an on-line turbidimeter, designed to meet ISO 7027(1), manufactured by Sigrist-
Photometer AG, that provides non-contact measurement of the 90° scattered light in a free-
falling water stream. Automatic adjustment using a fixed internal reference standard enhances
measurement reliability and minimizes the need for cleaning and calibration. The WTM500 has a
nominal range of 0 to 500 formazin nephelometric units (FNUs) in eight scale ranges, with a
maximum resolution of 0.001 FNU.
The control unit has a two-line
liquid-crystal display.
Because the flow cell and windows
have been eliminated, the turbidi-
meter optics do not need to be
cleaned regularly. The turbidimeter
is adjusted using a solid internal
reference standard. The formazin
calibration is checked at regular
intervals against a built-in solid
reference, and any deviations are
corrected automatically.
The WTM500 turbidimeter's
measuring wave length is 880 nm,
and it has eight configurable scale
ranges. The turbidimeter is
designed to be operated at
temperatures between 0 and 40 °C
and at a flow rate between 0.84 and
1.06 gallons per minuts (gpm).
Figure 2-1. Sigrist WTM500 On-Line Turbidimeter
2
-------
Chapter 3
Test Design and Procedures
3.1 Introduction
The verification test was conducted according to procedures specified in the Test/QA Plan for
Verification of On-Line Turbidimeters,(2) Performance characteristics evaluated in the verification
test are listed in Table 3-1 along with the dates that data were collected for these evaluations. The
test was conducted at a full-scale municipal water treatment facility in Columbus, Ohio. The
verification test described in this report was conducted from September 9 through October 26,
1999, as indicated in Table 3-1.
Table 3-1. Performance Characteristics Evaluated and Schedule of Verification Test
Performance Characteristic
Date Data Collected
Off-Line Phase
Linearity
Accuracy
Precision
Water temperature effects
Flow rate sensitivity
Color effects
On-Line Phase
Accuracy
Calibration checks
September 9 to 10; October 20 to 21
September 9 to 10
September 9 to 10
September 10, 14 to 15
September 15 to 16
October 25 to 26
September 17 to October 18
September 23, 24, 27, 30 and October 6, 8, 12, 18
3.2 Test Design Considerations
Since turbidity is a measurement of light scattering, a number of factors can influence the
measurement of turbidity in a given sample solution. Instrument design, including light source
selection and geometric differences, may result in significant differences between the responses
of different turbidimeters. Further differences may result from the variable nature of both the size
and composition of particles typically found in water streams, relative to those in standard solu-
tions made with formazin or with polymer beads. These issues were addressed in this verification
3
-------
test in two ways: (1) by using different instrumental designs for the reference turbidimeters, and
(2) by evaluating a variety of samples.
To avoid potential bias associated with a single method of comparison, the verification test used
two reference methods for data comparisons: ISO 7027, "Water Quality—Determination of
Turbidity,"(1) and EPA Method 180.1, "Determination of Turbidity by Nephelometry."(3) Both of
these methods measure turbidity using a nephelometric turbidimeter, but they differ in the type of
light source and the wavelength used. ISO 7027 calls for an infrared light source, whereas
Method 180.1 calls for a visible light source. The Sigrist WTM500 is designed to conform to the
requirements of ISO 7027, and thus that method is the appropriate reference for verification of
the WTM500's performance. Verification results presented in this report, and summarized in the
Verification Statement, are based on comparisons with the ISO 7027 data. However, secondary
comparisons also are shown in this report, based on data from the WTM500 and Method 180.1.
These secondary comparisons are of interest because Method 180.1 is widely recognized in the
U.S. and is designated as the required method for drinking water compliance measurements.
These secondary comparisons are shown only to illustrate the performance capabilities of the
WTM500 and should not be taken as having equal weight as the comparisons with ISO 7027.
Additionally, to assess the response of the Sigrist WTM500 turbidimeter to both prepared
solutions and real-world water samples, verification involved both off-line and on-line phases.
The off-line phase challenged the turbidimeter with a series of prepared standards and other test
solutions to verify performance under controlled conditions. The on-line phase assessed long-
term performance under realistic operating conditions by monitoring a sample stream in a
municipal water treatment plant under normal operation. With the cooperation of the City of
Columbus' Water Division, both off-line and on-line phases were performed at the division's
Dublin Road Water Plant in Columbus, Ohio.
3.3 Experimental Apparatus
On-line turbidimeters measure turbidity continuously on flowing sample streams, as opposed to
the static grab samples analyzed by the bench-top reference turbidimeters. Consequently, great
care was taken to ensure that the samples collected for reference analysis were representative of
the sample flow measured by the Sigrist WTM500 turbidimeter. A cylindrical distribution
manifold provided identical sample streams to sample ports spaced equally around the
circumference of the manifold. Throughout the verification test, three ports were used for the
turbidimeters being verified and one port provided a stream for the grab samples. A single port
centered in the bottom of the distribution manifold introduced the sample stream to the manifold.
All the ports were tapped for V2" male NPT fittings, and hard plastic compression fittings were
used to connect the distribution manifold to the tubing (V2" OD polyethylene) used in the
recirculation system. Using a consistent tubing size and fitting style enabled rapid switching of
the turbidimeters on a scheduled basis among the ports on the distribution manifold. Providing
4
-------
On-line Sample
Introduction Port
Figure 3-1. Schematic Representation of Recirculation System
identical samples to each of the manifold ports minimized biases arising from water quality or
turbulence issues; rotation of the technologies to each of these ports was used to identify if biases
existed.
A schematic representation of the recirculation system is provided in Figure 3-1, where T1
through T3 represent the three on-line turbidimeters undergoing verification testing. T2 repre-
sents the Sigrist WTM500 turbidimeter. Prepared solutions were supplied to all three turbidi-
meters simultaneously in a closed-loop recirculation system that used a 40-L reservoir and a
centrifugal pump. Stream water from the plant was sampled from a pressurized source in a once-
through configuration (i.e., without the use of the pump or reservoir). In-line particle filters were
inserted into the water flow, using appropriate valving, when reduction of turbidity levels was
needed.
Before verification testing began, a series of five grab samples was collected from each port on
the cylindrical manifold while recirculating a formazin solution with a nominal turbidity of
0.5 NTU. These samples were analyzed with the reference turbidimeters and compared to ensure
5
-------
uniformity of the turbidity of the solution. Comparison of the sample analyses indicated agree-
ment in turbidity readings within ± 5% of the average turbidity among all of the ports.
Before verification testing began, the on-line turbidimeters verified in this test were installed in
the test apparatus at the Dublin Road Water Plant. Much of the recirculation system, including
the flow meters and the distribution manifold, was mounted to a '/V-thick aluminum panel
installed in the water plant specifically for this verification test. The WTM500 turbidimeter
(Serial Number 980952), installed by a representative of Sigrist, was mounted on an "L" bracket
bolted to the aluminum panel. The bracket was secured to the panel using two bolts (~ V2"
diameter), and the WTM500 turbidimeter was secured to the bracket with two similar bolts.
During installation, the turbidimeter was leveled to optimize the optical qualities of the falling
water stream in the turbidimeter and to ensure that the sample was properly collected and
discharged. The sample inlet line entered the WTM500 at the base of the turbidimeter housing.
The turbidimeter and the recirculation system were connected with soft plastic tubing that was
connected at one end to a nipple in the base of the turbidimeter housing and at the other end to a
barbed fitting on the tubing in the recirculation system. A 1 "-diameter flexible hose was
connected to the outlet of the turbidimeter. To prevent overflow of the sample stream inside the
turbidimeter housing, the outlet line was significantly larger than the inlet; no flow obstructions
were added to the outlet stream. Consequently, the flow meter was installed on the inlet line of
the WTM500, as illustrated in Figure 3-1.
The control unit for the Sigrist WTM500 was installed above the turbidimeter and mounted to
the aluminum panel using two of four available bolt holes in the back of the control unit housing.
The output was converted from a 4-20 mA signal to a DC voltage using a precision resistor, and
was recorded every 10 seconds throughout the test using LabTech Notebook software, which was
run on a personal computer at the test site. Since the intent of the verification test was to assess
performance in routine unattended operation, the WTM500 was operated in its 0 to 10 NTU
range throughout all test activities. In this range the smallest observable increment in turbidity
was approximately 0.006 NTU.
3.4 Reference Instruments
Owing to the nature of turbidity measurement and the inherent differences in response arising
from different instrumental designs, separate bench-top turbidimeters meeting the design criteria
detailed in ISO 7027(1) and EPA Method 180.1(3) were used as reference instruments in this test.
Both methods measure the nephelometric light scattering of a formazin solution, albeit with
different prescribed instrumental designs. The primary difference between these two methods is
in the choice of light source. Method 180.1 requires the use of a broadband visible incandescent
tungsten lamp, while ISO 7027 requires the use of a narrowband IR source. Since the WTM500
is designed to comply with ISO 7027 requirements, that reference method is the basis for this
verification. Comparisons of data with Method 180.1 are also shown because of the widespread
recognition and use of that method. However, Method 180.1 comparisons are secondary to the
ISO 7027 comparisons used for verification. The bench-top turbidimeters used as the reference
methods were the Hach 2100AN (Serial Number 980300001366) and the Hach 2100AN IS
6
-------
(Serial Number 950700000173), which, according to the manufacturer's literature, comply with
the design specifications described in EPA Method 180.1(3) and ISO 7027(1), respectively.
Throughout the test the reference turbidimeters were operated in the non-ratio mode.
3.5 Off-Line Testing
The off-line phase of the verification test involved off-line sample introduction aimed at
assessing the linearity, accuracy, and precision of the on-line turbidimeter relative to the
reference methods. Additionally, response to various upset conditions was quantified. As a
means of testing these parameters, the off-line test phase included the introduction of standard
formazin solutions or other samples and the intentional manipulation of flow and water quality
parameters.
Throughout the verification test, continuous turbidity measurements from the Sigrist WTM500
turbidimeter were recorded at preset intervals using LabTech Notebook software. Grab samples
were collected simultaneously with some of these recorded measurements and analyzed using the
bench-top reference turbidimeters to provide a basis of comparison for the performance evalua-
tions. The collection of grab samples was timed to coincide within 10 seconds with the recording
of real-time turbidity measurements from the Sigrist WTM500, and the grab samples were
analyzed within three minutes after collection to minimize possible temperature and settling
effects.
Additionally, off-line testing included monitoring the instrumental responses of the Sigrist
turbidimeter to variations in water temperature, flow rate, and color. Each of these parameters
was varied within a range consistent with conditions encountered under typical plant operation.
The following subsections describe the procedures used for the off-line phase of the verification
test.
Table 3-2 provides a summary of the parameters tested in the off-line phase, the test solutions
used, and the number of readings recorded for each parameter.
3.5.1 Linearity
Linearity was measured in the range from approximately 0.05 to 5 NTU as an initial check in the
off-line phase. The recirculation system was filled with distilled, deionized water, which was
then recirculated and filtered in the test apparatus using a 0.2-|j,m pleated polypropylene filter for
24 hours. After filtering, the in-line filter was bypassed and the turbidity of the water in the
recirculation system was measured by the reference turbidimeters to be approximately 0.05 NTU.
A series of five turbidity measurements was taken at that turbidity level, with intervals of at least
five minutes between successive measurements. A corresponding set of five measurements also
was recorded at approximately 0.3, 0.5, 2, and 5 NTU. To reach each turbidity level, a small
amount of 4000 NTU StablCal formazin stock solution was diluted in the recirculation system
and allowed to flow through the recirculation system unfiltered for at least 15 minutes before
7
-------
Table 3-2. Summary of Measurements for Off-Line Testing
Parameter Tested
Test Solution
Number of Readings
Linearity
Filtered Water
5
(<0.1 NTU)
Linearity (accuracy, precision)®
0.3 NTU Formazin
5
Linearity (accuracy, precision)
0.5 NTU Formazin
5
Linearity (accuracy, precision)
2 NTU Formazin
5
Linearity (accuracy, precision)
5 NTU Formazin
5
Water Temperature Effect
0.3 NTU Formazin
5 each at 16, 21, 27°C
Water Temperature Effect
5 NTU Formazin
5 each at 16, 21, 27°C
Flow Rate Effect
0.3 NTU Formazin
5
Flow Rate Effect
5 NTU Formazin
5
Color Effect
-0.1 NTU
5 each at 5, 15, 30 CU
Color Effect
5 NTU Formazin
5 each at 35, 45, 60 CU
a: () indicates additional parameters analyzed using collected data.
turbidity readings were recorded. At each turbidity level, a series of five turbidity readings was
recorded with at least a five-minute interval between successive readings.
These readings were compared to the reference measurements of grab samples collected simul-
taneously with each reading; that is, the turbidity of the solutions was determined by measure-
ment with the reference turbidimeters, rather than simply by calculations based on the dilution
process. After the prescribed measurements were recorded at each turbidity level, additional
formazin stock solution was added to the recirculation system to increase the turbidity of the
solution to the next value in the series.
Before measurements were recorded, the calibration of the reference turbidimeters was checked
using a 0.5 NTU StablCal formazin solution purchased from Hach Company, Loveland,
Colorado. Pursuant to the requirements of the test/QA plan,(2) agreement between the reference
measurement and the certified turbidity of the standard was required to be within 10% before
recording any series of measurements. After each series of measurements, the calibration of the
reference turbidimeters was again checked with the same standard, and the same acceptance
limits were applied. In addition to the 0.5 NTU calibration checks, before and after the measure-
ments on the filtered water level, a < 0.1 NTU blank standard also was measured to ensure proper
calibration of the reference instruments at low levels. The < 0.1 NTU standard also was
8
-------
purchased from Hach Company; agreement between the reference measurement and the turbidity
reported on the certificate of analysis was required to be within 0.02 NTU.
3.5.2 Accuracy and Precision
Data obtained from the linearity measurements were used to establish the accuracy and precision
of the Sigrist WTM500 turbidimeter in measuring formazin solutions. Accuracy was assessed by
comparing continuous turbidity measurements with those from the ISO 7027 reference turbidi-
meter. Precision was assessed from the five replicate results at each turbidity level.
3.5.3 Water Temperature
Variations in the temperature of the water stream were introduced to simulate a range of
conditions under which the on-line turbidimeters may typically operate. During off-line testing,
the temperature of the recirculating water equilibrated in the range from 27 to 29 °C, which was
approximately 3 to 6°C above the room temperature at the water plant during testing. To assess
the effect of temperature on the turbidimeter performance, the temperature of the recirculating
solution was lowered using an immersion type chiller, and replicate turbidity measurements were
recorded at approximately 21 °C and again at 16°C. In these tests, the solution temperature in the
reservoir was held within 2.5 °C of the nominal 16°C and 21 °C targets, while a series of five
measurements was recorded at each temperature. To ensure equilibration, the solution was
allowed to recirculate for one hour before the turbidity measurements were recorded. For the
temperature tests at 16°C and 21 °C, the temperature of the sample stream was recorded at the
grab sample port within 30 seconds of sample collection, and the temperature of the grab sample
was measured within 30 seconds of completion of the reference measurement. To assess
temperature effects at different turbidities, this test was conducted with both 0.3 and 5 NTU
solutions.
3.5.4 Flow Rate
The flow rate of the sample stream through the Sigrist WTM500 turbidimeter was manipulated to
assess the response of the turbidimeter to various realistic operational conditions. A manual ball
valve and needle valve were included upstream of the WTM500 turbidimeter and were adjusted
to vary the flow rate through the turbidimeter. Owing to the nature of the WTM500 design, the
flow requirements for the sample stream cover a narrow range of approximately 0.8 to 1.0 gpm.
Flow rates below the minimum specification may result in an unsteady or uneven column of
falling water, whereas flow rates above the maximum specification may result in overflow of the
capture reservoir and subsequent flooding of the turbidimeter housing. During normal testing, the
flow rate was held in the range from 0.85 to 0.95 gpm. The flow test was performed at a
minimum flow rate of 0.85 gpm and at a maximum flow rate of 0.95 gpm. To assess the effect of
flow rates on performance, measurements were made at both the minimum and maximum flow
rates at turbidity levels of both 0.3 NTU and 5 NTU.
9
-------
3.5.5 Color
Changes in water color were introduced by spiking the sample stream with colored solutions
prepared from commercial food coloring dye. Stock solution was added to the system reservoir to
give sample solutions approximately 5, 15, and 30 color units (CU) successively, and the instru-
mental response to these color changes was monitored. Five measurements were made for each
color level at both low turbidity (~ 0.1 NTU) and higher turbidity (~ 5 NTU).
The color of the recirculated solution was determined by analyzing the grab samples instru-
mentally using the Hach 2100AN reference turbidimeter with the supplied light filter. The
reference turbidimeter was calibrated for color measurements according to the instrument
manual. Solutions used in the color calibration of the reference turbidimeter were prepared by
dilution of a commercial cobalt-platinum color standard(4) (Hach Company, Loveland, Colorado).
At ~ 0.1 NTU, the color of the solution before addition of the dye was approximately 0 CU.
However, at the 5 NTU level, light scattering from the presence of formazin introduced an
apparent color to the solution of approximately 30 CU. Consequently, for the 5 NTU test, dye
solution was added to increase the color by 5, 15, and 30 CU; i.e., to bring the absolute color to
approximately 35, 45, and 60 CU, respectively.
3.6 On-Line Testing
The on-line test phase focused on assessing the long-term performance of the Sigrist turbidimeter
under realistic unattended operating conditions and assessing its accuracy in monitoring an actual
sample stream. Specifically, this phase of testing addressed the calibration and drift character-
istics of the turbidimeter over a five-week period of monitoring a sample stream from the water
plant. Routine reference measurements were used for comparison with the on-line readings to
assess accuracy, and a re-evaluation of the calibration at the end of the test period helped
establish drift characteristics. Natural meteorological and demand changes contributed to the
variability of water quality in the treatment facility and provided a natural range of turbidity for
characterizing performance.
Table 3-3 provides a summary of the parameters tested in the on-line phase, the test solutions
used, and the number of readings recorded for each parameter.
10
-------
Table 3-3. Summary of Measurements for On-Line Testing
Parameter Tested Test Solution Number of Readings
Accuracy
Drift
Drift
Drift
Drift
3.6.1 Accuracy
In the on-line testing, the accuracy of the Sigrist WTM500 turbidimeter relative to the ISO 7027
reference method was assessed on water samples from the plant stream. A sample stream was
drawn from a flocculation settling basin at the Dublin Road Water Plant facility, containing
unfiltered water that had been treated with lime, caustic, and alum. The sample stream was
directed to the Sigrist WTM500 turbidimeter through the distribution manifold. Two grab
samples of this stream were collected and analyzed by the reference turbidimeters each weekday
(Monday through Friday) for the four weeks of testing. The reference measurements of these
samples were compared with the simultaneous results from the Sigrist WTM500 turbidimeter.
The observed range of turbidity in the sample stream was approximately 0.1 to 1.0 NTU.
3.6.2 Drift
Drift was determined in two ways: (1) through off-line calibration checks conducted regularly
throughout the course of the verification test using formazin solutions, and (2) through a
comparison of multi-point linearity checks performed initially during the off-line phase
(described in Section 5.1) and subsequently after the completion of the on-line phase. The
turbidimeter was calibrated by the vendor prior to shipment and installation at the water plant.
After that calibration, no further manual calibration or adjustment was performed for the
duration of the verification test period. However, during the course of the on-line testing,
automatic calibration adjustments were performed daily by the WTM500 using its internal
reference standard. Also, the housing of the Sigrist WTM500 turbidimeter was opened twice
during the on-line testing, at which times the reservoir was cleaned to remove accumulated
deposits.
The Sigrist WTM500 turbidimeter was taken offline briefly twice each week for routine
calibration checks against a 0.5 NTU formazin solution. These intermediate calibration checks
were performed twice weekly for four consecutive weeks. Freshly diluted StablCal solutions
were used as the standards for these calibration checks.
Plant Water
0.3 NTU Standard
0.5 NTU Standard
2 NTU Standard
5 NTU Standard
2 per weekday for 4 weeks
(40 total)
5 for final linearity check
5 each for eight calibration checks (40 total)
and 5 for final linearity check
5 for final linearity check
5 for final linearity check
11
-------
Upon completion of the four-week period, calibration and linearity were checked again, through
a comparison with the reference measurements using standard solutions of 0.3, 0.5, 2, and
5 NTU. A linear fit of these data was compared with the initial linearity check performed in the
off-line phase to assess the degree of calibration drift.
12
-------
Chapter 4
Quality Assurance/Quality Control
Quality control (QC) procedures were performed in accordance with the quality management
plan (QMP) for the AMS pilot(5) and the test/QA plan(2) for this verification test.
4.1 Data Review and Validation
Test data were reviewed and approved according to the AMS pilot QMP,(5) the test/QA plan,(2)
and Battelle's one-over-one policy. The Verification Test Coordinator, or the Verification Test
Leader, reviewed the raw data and the data sheets that were generated each day and approved
them by adding their signature and the date. Laboratory record notebook entries were also
reviewed, signed, and dated.
4.2 Deviations from the Test/QA Plan
During the preparation and performance of the verification test, deviations from the test/QA plan
were implemented to better accommodate differences in vendor equipment and other changes or
improvements. Any deviation required the approval signature of Battelle's Verification Testing
Leader. A planned deviation report form was used for documentation and approval of the
following changes:
1. Commercial food coloring dye was used for the color test instead of diluted color
standard owing to the strongly acidic nature of the cobalt-platinum standard solution.
2. Calibration of the pH meter was performed only once during the test, and the meter was
not readjusted to account for variations in ambient temperature. Recalibration was to be
performed under the conditions of the test. However, the pH measurements were used
only to assess changes and not for absolute measurements.
3. Only one in-line filter was used in the recirculation system.
4. The schedule of tests was lengthened and the order of testing was changed to better
group series of parameter evaluations and to respond to unexpected experimental results.
These deviations had no significant impact on the test results used to verify the performance of
the on-line turbidimeters.
13
-------
4.3 Calibration
4.3.1 Reference Turbidimeters
The reference turbidimeters were calibrated according to the procedures described in their
respective instrument manuals. The calibrations were performed on August 23, 1999. Calibration
was performed using a blank, and 20, 200, 2000, and 7500 NTU StablCal calibration standards
(Hach Company, Loveland, Colorado). After calibration and before proceeding with the
verification test, the calibration of each reference turbidimeter also was checked through a five-
point linearity test using solutions with the following turbidities: < 0.1, 0.3, 0.5, 2, and 5 NTU.
The < 0.1, 0.3, and 0.5 NTU solutions were purchased and used as is, whereas the 2 and 5 NTU
solutions were prepared by diluting a purchased 20 NTU StablCal formazin standard solution.
The results of the linearity check are summarized in Table 4-1, indicating that the two reference
turbidimeters gave essentially identical results. For each reference turbidimeter, the slope of this
linear fit was within the 0.90 and 1.10 limits prescribed in the test/QA plan,(2) and each fit had an
r2 > 0.98 as called for in the test/QA plan.(2)
Table 4-1. Results of Linearity Check of Reference Turbidimeters
Parameter
Hach 2100AN IS (ISO 7027)
Hach 2100AN (180.1)
Slope
1.086
1.086
Intercept
0.0038
0.0101
r2
0.9991
0.9996
The calibration of each reference turbidimeter also was checked both before and after each series
of test measurements, using a nominal 0.5 NTU StablCal standard solution. The reference
turbidimeters were to be recalibrated if agreement between the turbidity reading and the certified
0.521 NTU turbidity value of this standard solution was not within ± 10% (i.e., 0.469 - 0.573
NTU). If this calibration check criterion was met before but not after a series of test measure-
ments, those measurements were to be repeated after recalibration of the reference turbidimeters.
Throughout the course of the verification test, neither reference turbidimeter was ever found to
be out of calibration, and consequently no recalibration of the reference turbidimeters was
performed.
Before the background readings were measured for the detection limit determination, an addi-
tional calibration check with <0.1 NTU standard also was performed on the reference turbidi-
meters to ensure proper calibration at low levels. These calibration checks were performed on
September 9, 1999, for the initial linearity test, and October 20, 1999, for the final linearity test.
The results showed agreement within 0.02 NTU between the turbidity reading of the <0.1 NTU
standard and the value as reported on the certificate of analysis.
14
-------
4.3.2 Temperature Sensors
A Fluke 52 thermocouple (Battelle Asset Number 570080) was used throughout the verification
test to determine water temperature, and the ambient room temperature. This thermocouple was
calibrated on June 30, 1999, against a calibrated temperature standard (Fluke 5500A, Battelle
Asset Number SN-714755).
4.3.3 Flow Meters
The flow meter used in the verification test to measure the water flow through the Sigrist
WTM500 turbidimeter was a panel-mounted, direct-reading meter purchased from Cole-Parmer
(Catalog Number P-03248-56), capable of measuring up to 1 gpm. The flow meter was factory
calibrated and was checked once during the verification test by measuring the time required to
fill a container of known volume through the meter at a setting of 0.8 gpm. Table 4-2
summarizes the results of the flow rate checks.
Table 4-2. Summary of Flow Meter Calibration Check
Flow Meter Setting
Volume
Time
Calculated Rate
(gpm)
(gallon)
(seconds)
(gpm)
0.8
2
143
0.84
0.8
4
299
0.80
The calibration check was performed on August 26, 1999, and indicated agreement within the
10% criterion established in the test/QA plan(2) at this flow rate.
4.3.4 pHMeter
Calibration of the pH meter was performed once during the verification test with no further
adjustment of the meter. Calibration included standardization at a pH of 7 and a pH of 10 using
buffer solutions. Calibration checks performed during the color test indicated a bias of 0.1 to 0.3
pH units. Biases above 0.2 pH units fall outside of the acceptance criterion for the verification
test and introduce an uncertainty to the absolute magnitude of the pH readings. However, the pH
readings were recorded as a means of assessing if changes in the acidity of the solution occurred
as a result of adding the color solution, rather than as an absolute measure of the pH itself. The
pH readings recorded during the test indicated no evidence of pH change in the test solution as
the result of adding dye to the test solution.
4.4 Data Collection
Electronic data were collected and stored by a PC-based data acquisition system using LabTech
Notebook software (Version 8.0.1). Data were collected from the Sigrist WTM500 turbidimeter
every 10 seconds over much of the course of verification testing. These data were saved in ASCII
15
-------
files along with the time of collection. Data files were stored electronically both on the hard drive
of the data collection system and on floppy discs for backup purposes. Data collected manually
included turbidity readings of the reference turbidimeters, flow rates, and water and ambient air
temperature measurements. An example of the data recording sheet used to record these data is
shown in Appendix A.
4.5 Assessments and Audits
4.5.1 Technical Systems Audit
Battelle's Quality Manager performed a technical systems audit once during the verification test.
The purpose of this audit was to ensure that the verification test was performed in accordance
with the test/QA plan(2) and that all QA/QC procedures were implemented. In this audit, the
Quality Manager reviewed the calibration standards and reference methods used, compared
actual test procedures with those specified in the test/QA plan, and reviewed data acquisition and
handling procedures. A report on this audit is provided in Appendix B.
4.5.2 Performance Evaluation Audit
Performance evaluation audits were conducted to assess the quality of the measurements made in
the verification test. These audits addressed only those measurements made by Battelle staff in
conducting the verification test, i.e., the reference turbidimeter readings and temperature
measurements. The audits were conducted by analyzing the standards or comparing them with
references that were independent of those used in the verification test. Each audit was made at
least once during the verification test.
The audit of the reference turbidimeters was performed by analyzing a reference solution that
was independent of the formazin standards used for calibration of the reference turbidimeters
during the verification test. The independent reference solution was an AMCO-AEPA-1
0.500 NTU standard solution obtained from APS Analytical Standards, Redwood City,
California. This audit was conducted once daily throughout the verification test and served as an
independent verification of the calibration of the reference turbidimeters. Agreement between the
National Institute of Standards and Technology (NIST) traceable turbidity value of the AMCO-
AEPA-1 solution and the turbidity readings from each reference turbidimeter was recorded and
tracked graphically using a control chart. Furthermore, similar calibration assessments were
performed daily using a purchased 0.521 NTU StablCal formazin standard (Hach Company,
Loveland, Colorado), as described in Section 4.3.1. The results of these StablCal daily calibration
assessments always showed agreement between the turbidity reading from each reference turbidi-
meter and the certified turbidity within ± 10%, as required in the test/QA plan.(2) The results of
the daily calibration assessments are shown in Figures 4-la and 4-lb for both the AMCO-AEPA-
1 standard and the formazin standard, on the 2100AN IS (ISO 7027) and 2100AN (180.1)
reference turbidimeters, respectively. (The dashed lines in the upper parts of Figures 4-la and
4-lb are at intervals of 0.05 NTU, but are for visual reference only and are not exactly the ± 10%
control limits of the calibration checks. The bottom portion of each figure shows the ± 10%
16
-------
0.575
0.525
ec
o
Pi
0.475
0.425
- AMCO AEPA-1 (0.500 NTU)
- StableCal Formazin (0.521 NTU)
09/09/1999
09/23/1999
10/07/1999
10/21/1999
Q
a
o
dn
10
5
0
-5
-10
09/09/1999
A a » ¦ =*rl
1 1 1
09/23/1999
10/07/1999
10/21/1999
Figure 4-la. Control Chart for Performance Evaluation Calibration
Checks of ISO 7027 Reference Turbidimeter
0.575
0.525
oo
c
¦3
-------
control limits.) Throughout the course of the verification test, readings of the AMCO-AEPA-1
standard, as measured by the ISO 7027 reference turbidimeter, ranged from - 1.2% to + 6.2%
relative to the certified turbidity value for that standard, and were on average ~ 2.7% higher. For
the Method 180.1 turbidimeter, the range was - 1.2% to + 5.4% with an average reading that was
1.7% higher than the certified turbidity value. The daily fluctuations in these measurements
resulted in standard deviations of - 1.7% for each reference turbidimeter. Similarly, readings of
the formazin standard ranged from - 5.2% to + 8.4% for the ISO 7027 turbidimeter and from
- 6.9% to + 8.3%) for the Method 180.1. The average readings were higher than the certified
turbidity value by ~ 2.2% when measured by the ISO 7027 turbidimeter, and by ~ 1.6% for the
Method 180.1 turbidimeter, with standard deviations of 3.5% and 3.7%, respectively. Although
the average deviations from the true turbidity values for these standards were approximately the
same, the scatter in the readings was greater in the formazin readings.
The audit of the thermocouple used during the verification test consisted of a comparison of the
temperature readings from the thermocouple with those of an independent temperature sensor.
The thermocouple was checked for accuracy by comparison with an American Society for
Testing and Materials mercury-in-glass thermometer in the Battelle Instrument Laboratory on
October 13, 1999, and again on November 1, 1999. Those comparisons were done at ambient
temperature, and the results are shown in Table 4-3.
Table 4-3. Results of Calibration Checks of Thermocouple Used in the Verification Test
October 13,1999
November 1,1999
Fluke 52 Thermocouple
27.2°C
29.5°C
ASTM Mercury-in-Glass Thermometer
27.2°C
29.7°C
Agreement between the thermocouple used in the verification test and the mercury-in-glass
thermometer was well within the two-degree specification established in the test/QA plan.(2)
4.5.3 Verification Test Data Audit
Battelle's Quality Manager audited at least 10% of the verification data acquired during the
verification test. The Quality Manager traced the data from initial acquisition, through reduction
and statistical comparisons, and to final reporting. All calculations performed on the data
undergoing the audit were checked.
18
-------
4.6 Audit Reporting
Each assessment and audit was documented in accordance with Section 2.9.7 of the Quality
Management Plan for the AMS pilot.(5) The assessment report included the following:
¦ Identification of any adverse findings or potential problems
¦ Response to adverse findings or potential problems.
A copy of the Technical Systems Audit Report is included as Appendix B of this report.
19
-------
Chapter 5
Statistical Methods
5.1 Off-Line Testing
The turbidimeter performance characteristics were quantified on the basis of statistical
comparisons of the test data. This process began by converting the files that resulted from the
data acquisition process into spreadsheet data files suitable for data analysis. The following
statistical procedures were used to make the comparisons.
5.1.1 Linearity
Linearity was assessed by linear regression, with the reference turbidity reading (R) as an
independent variable and the turbidimeter response (7) as a dependent variable. The regression
model was
T — Hi x R+ j8
where |_i, and (3 are the slope and intercept of the response curve, respectively. The turbidimeter
performance was assessed in terms of the slope, intercept, and the square of the correlation
coefficient of the regression analysis.
5.1.2 Accuracy
The accuracy of the turbidimeter with respect to the reference method was assessed in terms of
the average relative bias (B), as follows:
(
B
(R- Tf
R
x 100
where R is the turbidity reading of the reference turbidimeter, and T is the corresponding
turbidity reading of the Sigrist WTM500 turbidimeter.
Accuracy relative to the reference turbidimeter was assessed both for the prepared solutions and
the samples from the plant water stream. The accuracy of the Sigrist WTM500 turbidimeter was
assessed relative to the ISO 7027 reference method for verification purposes and relative to the
180.1 reference method as an additional illustration of performance.
20
-------
5.1.3 Precision
Precision was reported in terms of the percent relative standard deviation (RSD) of a group of
similar measurements. For a set of turbidity measurements given by Tx, T2, Tn, the standard
deviation (S) of these measurements is
S =
1 11 — n
—~ X (\ ~ t/
n -1 k=i
1/2
where T is the average of the turbidity readings. The RSD is calculated as follows:
RSD = 1x100
T
and is a measure of the dispersion of the measurements relative to the average value of the
measurements. This approach was applied to the groups of replicate measurements on each test
solution. In some cases, the turbidity of the prepared solution changed approximately linearly
with time, due to loss of particles in the recirculation system. In those cases, a linear regression of
the data was performed to assess the slope of the turbidity change as a function of time. This
slope was used to adjust the individual turbidity readings to approximately the initial concen-
tration. The precision was then calculated on the adjusted values as described above.
5.1.4 Water Temperature Effects
The effect of water temperature on the response of the Sigrist WTM500 at 0.3 NTU and 5 NTU
was assessed by trend analysis. The turbidity readings relative to the ISO 7027 reference turbidi-
meter were analyzed as a function of water temperature to identify trends in the relative turbidity
at each of the two levels of turbidity. The calculations were performed using separate linear
regression analyses for the data at each turbidity level. A similar calculation was done for
illustrative purposes using the 180.1 reference data.
5.1.5 Flow Rate Sensitivity
Analysis of flow rate influence on turbidity readings was similar to that for water temperature
effects. The turbidimeter response relative to the ISO 7027 reference turbidimeter was analyzed
as a function of flow rate to assess trends in the response of the turbidimeter with changes in
sample flow rate. The analyses were performed separately for the 0.3 NTU and 5 NTU data. A
similar calculation was done for illustrative purposes using the 180.1 reference data.
5.1.6 Color Effects
The influence of color on turbidity was assessed through a linear regression analysis of the
turbidity measured for each color relative to the ISO 7027 reference turbidimeter. Separate
21
-------
analyses were performed for the measurements recorded at 0.1 NTU and those recorded at
5 NTU. A similar calculation was done for illustrative purposes using the 180.1 reference data.
5.2 On-Line Testing
5.2.1 Accuracy
As described in Section 5.1.2, accuracy in the on-line measurements was determined as a bias
relative to the ISO 7027 reference turbidimeter. Daily reference measurements of the sample
stream from the water plant were used to assess accuracy. A similar calculation was done for
illustrative purposes using the 180.1 reference data.
5.2.2 Drift
Drift was assessed in two ways. The drift in the calibration of the Sigrist WTM500 turbidimeter
was assessed by a comparison of the regression analyses of the multi-point linearity tests per-
formed at the beginning and end of the verification test. This comparison was used to establish
any long-term drift in instrumental calibration during the verification test. Also, the reference and
on-line turbidity results in monitoring the plant water stream were used to assess drift associated
with the operation of the instrument (e.g., fouling of the optics, etc.). Trends in the intermediate
calibration data toward a positive bias were used to identify when the turbidimeter needed
cleaning.
22
-------
Chapter 6
Test Results
The results of the verification test are presented in this section, based upon the statistical methods
of comparison shown in Chapter 5. For all performance characteristics verified, two sets of
results are shown. The primary verification results are based on comparisons with the ISO 7027
reference method; a secondary illustration of performance is based on comparisons with the
180.1 reference method.
6.1 Off-Line Testing
Off-line testing was performed to assess the performance of the Sigrist WTM500 turbidimeter
when measuring known solutions under controlled conditions. The first of the off-line tests was
performed to establish the linearity of the turbidimeter response in the range from < 0.1 to
5 NTU. Data from the linearity test also were used to assess the accuracy and precision of the
WTM500 in this turbidity range. After the linearity test, the effects of sample temperature,
sample flow rate, and sample color were evaluated. The results of these tests are described in this
section.
6.1.1 Linearity
The verification data from the initial linearity test are shown in Figure 6-la, relative to the
ISO 7027 reference turbidimeter. A series of at least five data points was recorded at each of the
five nominal turbidity levels (approximately 0.05, 0.3, 0.5, 2, and 5 NTU.) At the two highest
NTU levels, a decrease in turbidity was observed in the readings of both the Sigrist WTM500
turbidimeter and the reference turbidimeter. This decrease can be seen graphically as a spread in
the data along the slope of the linearity plots. Between the first and fifth readings at 2 NTU, the
decrease in turbidity represented 4 to 5% of the initial turbidity as measured by the reference
turbidimeter. This decrease in turbidity was likely the result of formazin being lost from the
solution in the recirculation system. In an attempt to prevent the formazin loss, the solution was
stirred magnetically. A second series of five measurements was recorded at the 2 NTU level after
magnetic stirring of the formazin solution was introduced. After magnetic stirring was intro-
duced, the decrease in turbidity was still observed, however, to a slightly lesser extent
(approximately 2 to 4%).
The data from the linearity test were fit using a linear regression as described in Section 5.1.1,
and the results of these fits are shown in Table 6-1. The secondary comparison with the Method
180.1 data is shown in Figure 6-lb, with the regression results shown in Table 6-1.
23
-------
6.0
£
H
£
tao
-3
C3
"3
oi
-3
4.5
3.0
1.5
0.0
0.0 1.5 3.0 4.5 6.0
Reference Turbidity (NTU)
Figure 6-la. Linearity Plot for Sigrist WTM500
Turbidimeter vs. ISO 7027 Reference Turbidimeter
Figure 6-lb. Linearity Plot for Sigrist WTM500
Turbidimeter vs. Method 180.1 Reference
Turbidimeter
0 1.5 3 4.5 6
Reference Turbidity (NTU)
-------
Table 6-1. Statistical Results of Initial Linearity Test on the Sigrist WTM500 Turbidimeter
Linear Regression Parameter
Verification Results3
Secondary Comparisonb
Slope (std. error)
1.013 (0.005)
1.026 (0.004)
Intercept (std. error)
0.001 (0.010)
0.014 (0.009)
r2 (std. error)
0.9994 (0.0404)
0.9996 (0.0359)
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
The verification results of the linear regression indicate that the Sigrist WTM500 turbidimeter
responded linearly to turbidity throughout the range of about 0.05 to 5 NTU. The slope of the
response curve was 1.3% higher than unity with respect to the ISO 7027 reference method. Based
on the uncertainty of the reference measurements, the 95% confidence interval of the slope
relative to the reference method includes unity. A near zero intercept was determined for the
linearity plot; the 95% confidence interval for the intercept includes zero.
The secondary comparison in Table 6-1 shows that the WTM500 also exhibited good linearity
relative to Method 180.1.
6.1.2 Accuracy
Data obtained from the initial linearity test were used to assess accuracy for the off-line tests. The
results of the accuracy verification are given in Table 6-2 and are presented as the average
difference between the Sigrist WTM500 turbidimeter and the reference turbidimeter, as well as
the relative bias of the Sigrist WTM500 turbidimeter with respect to the reference measurements.
Negative values indicate a negative bias in the Sigrist WTM500 turbidimeter readings when
compared with the reference turbidimeter, and positive numbers indicate a positive bias in the
Sigrist WTM500 readings.
The verification results in Table 6-2 show a bias of about 1 to 5% over all turbidity levels from
0.3 to 5 NTU, resulting from average measured differences of 0.005 to 0.065 NTU. No trend of
the average bias with NTU level is evident. The observed bias is comparable to the degree of
fluctuations in the daily calibration checks of the reference turbidimeter.
The secondary comparison in Table 6-2 shows similar performance relative to Method 180.1,
with average bias results of 1 to 7%.
25
-------
Table 6-2. Bias of Sigrist WTM500 Turbidimeter Relative to Reference Measurements on
Prepared Test Solutions
Verification Results3 Secondary Comparisonb
Nominal Turbidity
of Test Solution (NTU)
Average
Difference
(NTU)
Relative
Bias
(%)
Average
Difference
(NTU)
Relative
Bias
(%)
0.3
-0.0211
-5.2
0.0045
1.3
0.5
-0.0054
-0.9
0.0258
4.9
2
0.0650
5.1
0.0970
7.0
5
0.0390
0.8
0.1170
2.4
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
6.1.3 Precision
Data from the linearity test were used to calculate precision at 0.3, 0.5, 2, and 5 NTU. At both the
2 NTU and 5 NTU levels, a decrease in turbidity was observed as a function of time during the
test procedure. To account for this variability in turbidity, the readings at these two levels were
analyzed by linear regression against time and adjusted to approximately the initial turbidity
value using correction factors based on the regression results. The adjusted values (T/) were
calculated using the following equation:
Ti = Tt + c(tt - tj
where Tt is the ith turbidity reading, /, is the time at which the ith sample was collected, l0 is the
time of collection for the initial sample in the series, and c is the slope of the line determined
from the linear regression results of turbidity versus time at 2 NTU or at 5 NTU. The results of
the adjustment calculations are given in Table 6-3 for the Sigrist WTM500 turbidimeter. Similar
corrections were applied to the reference readings since the reference readings showed the same
trend of decreasing turbidity with time.
The precision was calculated from the raw data at the 0.3 and 0.5 NTU levels, and from the
corrected data at the 2 NTU and the 5 NTU levels. The results of these calculations are shown in
Table 6-4. For comparison, the calculated precision values for the two reference turbidimeters are
also included in that table. The values presented in this table are based on five readings at each
level, with the exception of the 2 NTU levels, which included ten readings.
26
-------
Table 6-3. Adjusted Turbidity Readings for Precision Calculations on the Sigrist WTM500
Nominal
Actual
Corrected
Nominal
Actual
Corrected
Value
Reading
Reading
Value
Reading
Reading
(NTU)
(NTU)
(NTU)
(NTU)
(NTU)
(NTU)
2
1.8091
1.8091
5
5.0012
5.0012
2
1.7786
1.7929
5
5.0195
5.0360
2
1.7419
1.7705
5
4.9280
4.9610
2
1.7297
1.7725
5
4.9158
4.9653
2
1.6992
1.7563
5
4.9707
5.0367
2
1.7114
1.7828
2
1.7053
1.7910
2
1.6748
1.7748
2
1.6809
1.7951
2
1.6687
1.7972
Table 6-4. Precision of Sigrist WTM500 Turbidimeter and of the Reference Turbidimeters
Nominal
Turbidity
Sigrist WTM500
SD RSD (%)
ISO 7027 (2100AN IS)
SD RSD (%)
Method 180.1
(2100 AN)
SD RSD (%)
0.3 NTU
0.0127
3.4
0.0161
4.1
0.0051
1.4
0.5 NTU
0.0093
1.7
0.0118
2.1
0.0095
1.8
2 NTU
0.0157
0.88
0.0123
0.71
0.0108
0.64
5 NTU
0.0366
0.73
0.0126
0.26
0.0088
0.18
27
-------
The results of these calculations indicate that the Sigrist WTM500 turbidimeter has approxi-
mately the same precision as the reference turbidimeter through the range of turbidity measured
in this verification test. From 0.3 to 5 NTU, the WTM500 exhibited precision of 3.4 to 0.7% as
RSD.
6.1.4 Water Temperature Effects
The verification data obtained for the temperature test are shown in Figure 6-2a. As a result of
loss of formazin in the recirculation system during the temperature test, additional formazin
solution was added between each set of temperature measurements to maintain turbidity levels at
approximately 0.3 NTU and 5 NTU. Consequently, the absolute turbidity readings alone cannot
be used as an indication of temperature effects. Therefore, the readings recorded for the Sigrist
WTM500 turbidimeter were normalized to the corresponding reference readings to get a relative
measure of turbidity. These relative values (i.e., ratios of WTM500 to ISO 7027 data) are shown
in Figure 6-2a and were analyzed by linear regression to assess the effect of water temperature on
turbidity reading. The results of the regression analysis are given in Table 6-5.
Table 6-5. Statistical Results of Temperature Test on the Sigrist WTM500 Turbidimeter
Linear Regression
Parameter
Verification Results3
0.3 NTU 5 NTU
Secondary Comparisonb
0.3 NTU 5 NTU
Slope
-0.0009
0.0011
-0.0094
0.0007
(std. error)
(0.0018)
(0.0014)
(0.0018)
(0.0017)
Intercept (std. error)
0.8676
1.025
0.9493
1.024
(0.0418)
(0.030)
(0.0417)
(0.036)
r2
0.0164
0.046
0.6614
0.012
(std. error)
(0.0418)
(0.024)
(0.0417)
(0.030)
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
These verification results indicate that, relative to the ISO 7027 reference turbidimeter, the
Sigrist WTM500 shows no statistically significant relation between the turbidity readings and the
water temperature at either 0.3 NTU or at 5 NTU, since the 95% confidence interval includes
zero slope in both cases, i.e., water temperature has no effect on WTM500 readings within the
tested temperature range.
The secondary results in Figure 6-2b and Table 6-5 suggest a slight negative dependence of
turbidity reading on temperature at 0.3 NTU. However, the difference in measurement method
28
-------
1.1
1.0
£
'¦e
£ 0.9
0.8
0.7
1~
~
I
I
~
~
~ 0.3 NTU
~ 5 NTU
15
20
25
30
Temperature (C)
Figure 6-2a. Effect of Temperature on Sigrist WTM500 Turbidity Readings
vs. ISO 7027 at Both 0.3 and 5 NTU
1.2
1.0
^ O.f
~
I
~
-k-
0.6
0.4
15
~ 0.3 NTU
A 5 NTU
20
25
30
Temperature (C)
Figure 6-2b. Effect of Temperature on Sigrist WTM500 Turbidity Readings
vs. Method 180.1 at Both 0.3 and 5 NTU
29
-------
between the WTM500 and Method 180.1 may account for this result, rather than any actual
temperature dependence of the WTM500.
6.1.5 Flow Rate
The results of the flow rate test are summarized in Figure 6-3. The data are again presented and
analyzed as relative turbidity readings, rather than absolute turbidity readings, to account for loss
of formazin during the testing. The results of the statistical analysis of the flow data are presented
in Table 6-6.
~ ISO 7027
A EPA 180.1
-3
£
-------
Table 6-6. Statistical Results of Flow Rate Test for the Sigrist WTM500 Turbidimeter
Parameter
Verification Results3
Secondary Comparisonb
Slope (std. error)
0.271 (0.189)
0.131 (0.153)
Intercept (std. error)
0.655 (0.167)
0.736 (0.138)
r2 (std. error)
0.211 (0.029)
0.084 (0.024)
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
The results show no statistically significant effect of sample flow rate on the response of the
Sigrist WTM500 turbidimeter in the range of 0.85 to 0.95 gpm, since at the 95% confidence level
the slope values are not significantly different from zero. It should be noted that the
specifications of the Sigrist WTM500 turbidimeter require a small range of operational flow
rates. It is reasonable that within the specified range no flow rate effects are present.
6.1.6 Color Effects
The verification data obtained from the color tests are shown in Figure 6-4a. In this figure, the
data at each color level are plotted as relative values with respect to the reference turbidimeter
readings, and the statistical analysis of these data involved a linear regression analysis of the
relative data as a function of solution color. At 5 NTU, the background color reading of
approximately 30 CU was subtracted, and only the effect of color added during the test is shown.
The results of the statistical calculations are summarized in Table 6-7.
The verification results in Table 6-7 show that at the 0.1 NTU level color has no significant
effect (95% confidence) on the response of the Sigrist WTM500 turbidimeter. At the 5 NTU
level, color has a small but statistically significant effect on the response of the Sigrist WTM500
turbidimeter. This effect amounts to a decrease in turbidity determined by the Sigrist WTM500
turbidimeter on the order of 0.1% per CU increase relative to the reference turbidimeter. Very
similar results also are shown in the secondary comparison in Table 6-7.
It is somewhat surprising that the magnitude of this effect is the same relative to both reference
turbidimeters. Since the two reference turbidimeters use light sources with different peak wave-
lengths, the effect of color relative to the two reference turbidimeters would be expected to be
different. However, it should be noted that both reference turbidimeters were used in the non-
ratio mode in this test, and thus their readings were not compensated for color. Since the
WTM500 readings are compensated, this difference may be the cause of the apparent color effect
on WTM500 readings.
31
-------
1.2
1.0
£
'¦e
^ 0.8
_>
Qi
0.6
0.4
A
I
~
~
i
I 0.1 NTU
~ 5 NTU
10
~
~
15 20
Color (CU)
25
30
35
Figure 6-4a. Effect of Color on Relative Turbidity with the Sigrist WTM500
vs. the ISO 7027 at Both 0.1 and 5 NTU
1.2
1.0
£
€
3
H 0.!
>
0.6
0.4
I-
A
~
«_
~ 0.1 NTU
~ 5 NTU
10
~
15 20
Color (CU)
25
4
A
30
35
Figure 6-4b. Effect of Color on Relative Turbidity with the Sigrist WTM500
vs. Method 180.1 at Both 0.1 and 5 NTU
32
-------
Table 6-7. Statistical Results of the Color Test with the Sigrist WTM500 Turbidimeter
Reference:
Verification Results3
Secondary Comparisonb
Parameter
0.1 NTU
5 NTU
0.1 NTU
5 NTU
Slope
-0.0003
-0.0014
0.0003
-0.0014
(std. error)
(0.0011)
(0.0003)
(0.0010)
(0.0004)
Intercept
0.7838
0.9925
0.8076
1.028
(std. error)
(0.0224)
(0.0058)
(0.0198)
(0.008)
r2
0.0054
0.634
0.0056
0.471
(std. error)
(0.0455)
(0.012)
(0.0403)
(0.016)
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
6.2 On-Line Testing
Figure 6-5 shows the results from the four weeks of on-line testing. In this figure, data from the
Sigrist WTM500 and the reference turbidimeters are shown, along with additional data supplied
by the Dublin Road Water Plant (DRWP). Data from the DRWP are from a turbidimeter in the
plant sampling the same water stream at a different location, for plant operational purposes.
These DRWP data are shown to illustrate the trends in turbidity of the water stream sampled for
this test. No quantitative comparisons with the DRWP data should be made, since these data
were not collected at the same location as samples for this verification test. For convenience,
only one data point per hour is shown for the Sigrist WTM500, although data were recorded at
intervals of 10 seconds throughout the on-line testing. Breaks in the data from the Sigrist
turbidimeter indicate periods during which the turbidimeter was taken off-line for calibration
checks, or for cleaning.
In general, Figure 6-5 illustrates correlation and sometimes close quantitative agreement between
the WTM500 and the reference measurements. Also, the varying turbidity levels shown by the
WTM500 indicate a temporal pattern similar to that of the DRWP data. Two episodes near the
end of the four-week period show large deviations in the Sigrist data from the benchtop
measurements. These episodes occurred after adjustments to the recirculation system and are
likely to be associated with these modifications. Consequently, data recorded between 10/8 to
10/10, and after 10/12, will not be included in the following discussions of accuracy.
6.2.1 Accuracy
The results from the four weeks of on-line accuracy testing are given in Table 6-8. The results
shown in the table are given as the average of the two readings taken each day on water stream
samples for the Sigrist WTM500 turbidimeter and the reference turbidimeter. In cases where
more than the prescribed two readings were recorded, all the values are included in the reported
33
-------
Figure 6-5. Summary of Stream Turbidity Data from On-Line Testing of Sigrist WTM500
average. Additionally, the bias in the Sigrist readings relative to the reference turbidimeter is
reported.
The verification results in Table 6-8 show that the WTM500 generally read about 0.05 to
0.2 NTU higher than the ISO 7027 reference turbidimeter within a reference turbidity range of
about 0.16 to 0.6 NTU. Positive biases of about 15 to 40% characterize most of this data range.
However, near the end of the on-line test the lowest turbidities and best accuracy were observed
(i.e., biases of about 0 to -20%). The WTM500 and reference data exhibited a linear regression of
the form WTM500 = 1.288 (ISO 7027) + 0.025 NTU, with r = 0.513.
The secondary comparison in Table 6-8 shows somewhat lower accuracy of the WTM500
relative to Method 180.1, as expected, but the same degree of correlation (r2 = 0.513).
A similar positive bias was observed for all of the on-line turbidimeters tested in this verification
test, suggesting a systematic bias in the reference data. It should be noted that, since visible
granular deposits accumulated in the test apparatus during the on-line testing, it is possible that a
systematic negative bias may have existed as a result of large particles settling in the grab sample
vial between sample collection and reference analysis.
34
-------
Table 6-8. On-Line Daily Accuracy Check Results from Water Stream Samples
Sigrist WTM500 Verification Results3 Secondary Comparisonb
NTU (Relative bias %) (Relative bias %)
9/17/99
0.7044
0.5320 (32.4)
0.4315 (63.2)
9/20/99
0.6067
0.5090 (19.2)
0.4165 (45.7)
9/21/99
0.6250
0.5115 (22.2)
0.4215 (48.3)
9/22/99
0.7944
0.5855 (35.7)
0.4338 (83.1)
9/23/99
0.4389
0.3725 (17.8)
0.2990 (46.8)
9/24/99
0.4755
0.3790 (25.4)
0.3220 (47.7)
9/27/99
0.4755
0.4040 (17.7)
0.3245 (46.5)
9/28/99
0.8814
0.4885 (80.4)
0.4230(108.4)
9/29/99
0.6006
0.4155 (44.5)
0.3560 (68.7)
9/30/99
0.7593
0.5310 (43.0)
0.4690 (61.9)
10/1/99
0.5151
0.3140 (64.0)
0.2710 (90.1)
10/4/99
0.8661
0.3040 (184.9)
0.2550 (239.6)
10/5/99
0.5803
0.4043 (43.5)
0.3607 (60.9)
10/6/99
0.4023
0.4240 (-5.1)
0.3605 (11.6)
10/7/99
0.3901
0.3065 (27.3)
0.2755 (41.6)
10/8/99
0.3320
0.4173 (-20.4)
0.3627 (-8.5)
10/11/99
0.1459
0.1660 (-12.1)
0.1320 (10.5)
10/12/99
0.1581
0.1570 (0.7)
0.1340 (17.9)
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
6.2.2 Drift
6.2.2.1 Calibration Checks
The results from the twice weekly calibration checks at 0.5 NTU with formazin standards are
shown in Figure 6-6 and summarized in Table 6-9.
35
-------
0.75
°.5 - • | '
1
.-s
-3
£
£
0.25 -
• ISO 7027
I EPA 180.1
i Sigrist
0 J 1 1 1
09/23/1999 09/30/1999 10/07/1999 10/14/1999
Date
Figure 6-6. Twice-Weekly Calibration Checks During On-Line Testing of the Sigrist
WTM500
Table 6-9. Results of Calibration Checks Performed During On-Line Testing
Sigrist WTM500 Verification Results3 Secondary Comparisonb
(NTU) (Relative Bias %) (Relative Bias %)
09/23/99
0.454
0.473 (-4.1)
0.461 (-1.6)
09/24/99
0.459
0.512 (-10.4)
0.492 (-6.8)
09/27/99
0.485
0.526 (-7.9)
0.518 (-6.4)
09/30/99
0.514
0.542 (-5.1)
0.524 (-1.9)
10/06/99
0.435
0.458 (-5.2)
0.432 (0.6)
10/08/99
0.518
0.530 (-2.3)
0.514 (0.6)
10/12/99
0.459
0.505 (-9.0)
0.498 (-7.7)
10/18/99
0.520
0.556 (-6.5)
0.544 (-4.4)
10/18/99
0.505
0.569 (-11.3)
0.544 (-7.2)
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
Turbidimeter Cleaned
10/8/99
Turbidimeter Cleaned
Between Calibration
Checks
10/18/99
36
-------
The verification results in Table 6-9 show that the WTM500 read about 2 to 11% lower than the
ISO 7027 reference turbidimeter on the twice-weekly calibration solutions, with an average
negative bias of 6.8% (and an average uncertainty of ± 4.6%).
During the initial linearity check (Section 6.1.2), data from the WTM500 showed an average bias
of -0.9% ± 3.0%) relative to the ISO 7027 reference turbidimeter. No significant drift can be
inferred from the results of the initial linearity check and the on-line calibration checks because
the uncertainties in these measurements overlap.
Drift associated with optics fouling was not a concern since the Sigrist WTM500 turbidimeter
measures turbidity on a falling stream of water, and at no point does the sample stream contact
the optics of the turbidimeter.
The secondary comparison data in Table 6-9 show slightly better agreement of the WTM500
with the EPA Method 180.1 results than with the ISO 7027 results (i.e., -3.9%> average bias
compared to -6.8%> relative to ISO 7027).
6.2.2.2 Final Linearity Check
Data from the final linearity check are shown in Figure 6-7a. These data were recorded after
completion of the four weeks of on-line testing and after the Sigrist WTM500 turbidimeter had
been cleaned. As with the data from the initial linearity test, these data were analyzed by linear
regression. The results are summarized in Table 6-10. In Table 6-11, the results of the final
linearity test are compared with those from the initial linearity check conducted at the start of the
verification as part of the off-line phase.
The verification results of the regression analysis show a high degree of linearity, with a slight
negative bias in the slope with respect to the reference turbidimeter and a small negative intercept
of less than 0.05 NTU.
37
-------
0 1 2 3 4 5
Reference Turbidity (NTU)
Figure 6-7a. Final Linearity Plot for Sigrist WTM500
vs. ISO 7027 Reference Turbidimeter
Reference Turbidity (NTU)
Figure 6-7b. Final Linearity Plot for Sigrist WTM500
vs. Method 180.1 Reference Turbidimeter
38
-------
Table 6-10. Statistical Results of Final Linearity Test
Reference Turbidimeter
Linear Regression
Verification Results3
Secondary Comparisonb
Slope (std error)
0.949 (0.004)
0.985 (0.003)
Intercept (std error)
-0.042 (0.008)
-0.045 (0.006)
r2 (std. error)
0.9993 (0.0436)
0.9997 (0.0302)
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
Table 6-11. Comparison of Results from Linearity Tests at Beginning and End of
Verification
Verification Results3
Secondary Comparisonb
Slope
Intercept
Slope
Intercept
Initial Linearity Test
1.013
0.001
1.026
0.014
Final Linearity Test
0.949
-0.042
0.985
-0.045
Difference
-0.064
-0.043
-0.041
-0.059
% Difference
6.3
-
4.0
-
a Comparison with ISO 7027 reference method (2100AN IS reference turbidimeter).
b Comparison with EPA Method 180.1 (2100AN reference turbidimeter).
The verification results in Table 6-11 show a change of 6.3% in the slope of the WTM500
response relative to the ISO 7027 reference method between the initial and final linearity tests.
Based on the results of the daily calibration checks, the average difference between the reference
turbidimeter readings and the stated turbidity value of the formazin standard used for the checks
was 3.2%, with a standard deviation of 2.4%. With these uncertainties, at the 95% confidence
level, the initial and final slopes are not significantly different from unity and no drift can be
inferred from the difference between the slopes. A very slight change in intercept (0.04 NTU)
also was observed. Again, these changes are within the total estimated uncertainty of the
reference method, and thus do not definitively indicate significant drift in the WTM500
calibration.
The secondary comparison shown in Table 6-11 leads to a similar conclusion, as a difference of
only 4% in slope was observed, relative to the Method 180.1.
39
-------
6.3 Other Performance Parameters
6.3.1 Cost
As tested, the cost of the Sigrist WTM500 was approximately $4,000.
6.3.2 Maintenance/Operational Factors
Time requirements for installation were not assessed in this test, as temporary installation in the
test apparatus did not simulate permanent installation in a water treatment facility. However, the
primary time-consuming activities in the installation procedure were securing the "L" bracket to
the panel and leveling of the turbidimeter housing. After installation, the Sigrist WTM500
turbidimeter required no operator input and provided data continuously throughout the
verification test.
The only maintenance of the Sigrist WTM500 turbidimeter involved occasional cleaning of the
sample reservoir. The reservoir was cleaned twice during the on-line testing to remove residues
and material deposits that had accumulated. Since the optics in the Sigrist WTM500 turbidimeter
never contact the water sample, the optics remained clean throughout the verification test and
required no maintenance.
The primary concern with the Sigrist WTM500 turbidimeter is ensuring that the flow through the
system remains within the specified narrow limits. On at least one occasion during the test, the
housing of the Sigrist WTM500 turbidimeter was flooded as a result of a flow rate above the
specified limit. However, flooding caused no damage to the turbidimeter or the housing. After
the flooding occurred, the turbidimeter housing was opened and the inside of the housing was
dried with paper towels. Proper control of the flow must be maintained to eliminate this potential
problem. However, design changes in the WTM500 are being considered to prevent overflow.
40
-------
Chapter 7
Performance Summary
The Sigrist WTM500 is an on-line turbidimeter designed to provide continuous, real-time
measurement of turbidity of aqueous solutions. The Sigrist WTM500 turbidimeter provided
linear response over the tested range of < 0.1 to 5 NTU. The slope of the response curve from
<0.1 to 5 NTU for the Sigrist WTM500 turbidimeter relative to the ISO 7027 reference
turbidimeter was 1.013 at the beginning of this test, with an intercept of 0.001 NTU and r2
> 0.999.
In measuring standard formazin solutions in the range of 0.3 to 5 NTU, the Sigrist WTM500 and
the ISO 7027 reference turbidimeter agreed within 5%, which was comparable to the fluctuations
in the daily calibration checks of the reference turbidimeter. The precision in the measurements
of the Sigrist WTM500 ranged from 3.4% to 0.7% RSD at turbidities of 0.3 to 5 NTU. These
results were approximately the same as for the reference turbidimeter throughout this range of
turbidity.
Water temperature had no effect on the response of the Sigrist WTM500 turbidimeter relative
to the ISO 7027 method at low turbidity (0.3 NTU) or at higher turbidity (5 NTU). In contrast,
there was an effect of color on readings at high turbidity (5 NTU), but none at low turbidity
(~ 0.1 NTU). The color effect at 5 NTU resulted in a decrease in the observed turbidity of - 0.1%
per each CU increase relative to the reference turbidimeters. In the narrow range of flow rates
tested for the Sigrist WTM500 turbidimeter (0.85 to 0.95 gpm), there was no statistically
significant effect on the turbidity readings as a function of sample flow rate.
In reading the turbidity of treated, unfiltered water from a municipal drinking water plant with a
turbidity range of ~ 0.1 to 0.6 NTU, the Sigrist WTM500 turbidimeter usually showed a positive
bias of typically ~ 0.05 to 0.2 NTU relative to the reference turbidimeter. Similar results were
seen with other on-line turbidimeters verified in this same test. On the other hand, calibration
checks of the WTM500 turbidimeter using a nominal 0.5 NTU formazin solution showed a
negative bias of 4 to 7% with respect to the reference turbidimeter, indicating a difference in
response between the formazin and plant water streams. A systematic bias in the reference
readings may have been present in the on-line test phase and may have contributed to the
observed differences between the Sigrist WTM500 and reference readings on the water stream
samples.
A change of approximately 6% in the slopes of the response curves between the beginning and
end of the verification test was observed; however, this change is within the combined experi-
mental uncertainty of the reference measurements over this time period, and does not definitively
41
-------
indicate a calibration drift. A change of approximately 0.04 NTU was observed in the values of
the intercepts calculated from the initial and final linearity checks. This degree of change is
within the experimental uncertainty of the reference measurements. No apparent drift was
observed in the Sigrist WTM500 calibration throughout the on-line testing on the plant water
stream.
The Sigrist WTM500 turbidimeter is easy to use and provides continuous on-line turbidity
readings. The turbidimeter was cleaned during the test to remove residues and material deposits
from inside the system reservoir. However, cleaning the optics was not required since the optics
are never in contact with the sample. The flow rate of the sample solution through the turbidi-
meter must be closely controlled to avoid producing an uneven column of falling water or
internal flooding of the turbidimeter housing.
42
-------
Chapter 8
References
1. "Water Quality—Determination of Turbidity," International Standard ISO 7027, Second
Edition, International Organization for Standardization, Geneva, 1990.
2. Test/QA Plan for Verification of On-Line Turbidimeters, Battelle, Columbus, Ohio,
June 3, 1999.
3. "Determination of Turbidity by Nephelometry," Methods for the Determination of
Inorganic Substances in Environmental Samples, Method 180.1, EPA/600/R-93/100,
U. S. Environmental Protection Agency, Cincinnati, Ohio, August 1993.
4. "Color in Water by Visual Comparison to Standards," Standard Methods for the
Examination of Water and Wastewater, 18th Edition, Method 2120-B, American Public
Health Association, 1992.
5. Quality Management Plan (QMP) for the ETV Advanced Monitoring Systems Pilot, U. S.
EPA Environmental Technology Verification Program, Battelle, Columbus, Ohio,
September 1998.
43
-------
Appendix A
Example Data Recording Sheet
A-l
-------
Appendix B
Technical Systems Audit Report
B-l
------- |