svEPA
United States
Environmental Protection
Agency
Office o< Acid Deposition,
Environmental Monitoring and
Quality Assurance
Washington DC 20460
EPA/600/4-86/044
December 1986
Research and Development
National Stream
Survey - Phase I
Quality Assurance Plan
-------
SUBREGIONS OF THE NATIONAL STREAM SURVEY-PHASE I
Northern
Appalachians (2Cn)
Valley and Ridge (2Bn)
Southern Blue Ridge (2As)
(Pilot Study)
Poconos/Catskills (ID)
NY\
Mid-Atlantic
Coastal Plain (3B)
Ozarks/Ouacnitas 2D
Southern Appalachians (2X)
-------
EPA/600/4-86/044
December 1986
National Stream Survey
Phase I
Quality Assurance Plan
A Contribution to the
National Acid Precipitation Assessment Program
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
Environmental Monitoring Systems Laboratory - Las Vegas, NV 89119
Environmental Research Laboratory - Corvallls, OR 97333
-------
NOTICE
The information In this document has been funded wholly or in part by the U.S.
Environmental Protection Agency under contract number 68-03-3249 to Lockheed
Engineering and Management Services Company, Inc. It has been subject to the
Agency's peer and administrative review, and it has been approved for publication as an
Agency document.
Mention of trade names or commercial products does not constitute endorsement
or recommendation for use.
This document is one volume of a set which fully describes the National Stream
Survey - Phase L The complete document set includes the major data report, quality
assurance plan, analytical methods manual, field operations report, processing laboratory
operations report and quality assurance report. Similar sets are being produced for each
Aquatic Effects Research Program component project. Colored covers, artwork, and the
use of the project name in the document title serve to identify each companion
document set
Trie corract citation of this document is:
Drous6, S. K., D. C. HiMman, J. L Engete, L W. Creelman, and S. J. Siman. 1986.
National Surface Water Survey, National Stream Survey {Phase I - Pilot, Mid-Atlantic
Phase I, Southeast Screening, and Episodes Pilot) Quality Assurance Plan. EPA 600/4-
8BJ044. U.S. Environmental Protection Agency, Las Vegas, Nevada. 198 pp.
-------
ABSTRACT
The National Surface Water Survey of the National Acid Precipitation Assessment
Program is a three-phase project to evaluate the current water chemistry of lakes and
streams, determine the status of fisheries and other biotic resources, and select
regionally representative surface waters for a long-term monitoring program to study
future changes in aquatic resources. This manual describes the quality assurance plan
for the first four field components of the National Stream Survey: the Phase I - Pilot
Survey, the Phase I - Mid-Atlantic Survey, the Southeast Screening Survey, and the
Episodes Pilot Survey.
To ensure that procedures are performed consistently and that the quality of the
data generated can be determined, the Quality Assurance Project Plan for these four
elements of the National Stream Survey specifies the following measures:
• Provide detailed, written sampling methodology.
• Simultaneously train all personnel participating in sampling and processing
activities.
• Conduct site visits to each field operations base throughout the sampling
period to ensure that all methods and quality assurance procedures are being
performed properly.
• Perform extensive evaluation of analytical laboratories before their selection
and throughout their participation.
• Assess variability introduced at each level of activity in mobile processing and
analytical laboratories by utilizing audit samples (synthetic and natural lake
samples), duplicates, and blanks along with routine samples.
• Provide detailed, written analytical methodology.
• Use internal quality control procedures at the analytical laboratory to detect
potential contamination and to verify established detection limits.
• Enforce sample holding time requirements.
• Use quality control protocols in the field, at the mobile processing laboratory,
and at the analytical laboratory to confirm that reported data are correct.
• Enter data into the data base twice, and scan for outlying values to detect
and eliminate transcription errors.
• Verify data by means of range checks, internal consistency checks, and
quality assurance evaluations.
• Validate verified data by analysis of the reasonableness of data, based on the
values expected for the particular region or subregion involved.
iii
-------
ACKNOWLEDGMENT
Contributions provided by the following individuals were essential to the completion
of this quality assurance document and are gratefully acknowledged: C. Ariss, J. Messer
(U.S. Environmental Protection Agency); D. Chaloud, M. Faber. J. Fountain, Jr., C.
Hagley, B. Hess, D. Hoff, M. Knapp, C. Mayer, D. Peck, J. Potter, L. Stanley, M.
Stapanian (Lockheed Engineering and Management Services Company, Inc.); J. Coe, M.
Sale (Martin Marietta Energy Systems, Inc.); J. Eilers, K. Eshleman, J. Sprenger
(Northrop Services, Inc.); K. Schreiber (U.S. Department of the Interior).
Recognition belongs to R. D. Schonbrod (U.S. Environmental Protection Agency, Las
Vegas, Nevada) who served as project officer.
IV
-------
Section T of C
Revision 4
Date: 9/88
Page 1 of 4
CONTENTS
Page
1.0 INTRODUCTION 1 of 2
2.0 PROJECT DESCRIPTION 1 of 4
2.1 Components of the National Stream
Survey 1 of 4
2.1.1 Phase I - Pilot Survey 1 of 4
2.1.2 Mid-Atlantic Phase I Survey. 1 of 4
2.1.3 Southeast Screening Survey. 1 of 4
2.1.4 Episodes Pilot Survey. 3 of 4
2.2 Data Quality Objectives ' 3 of 4
2.3 Specifications of the Quality
Assurance Project Plan 4 of 4
3.0 PROJECT ORGANIZATION. . 1 of 3
4.0 QUALITY ASSURANCE OBJECTIVES FOR
PRECISION, ACCURACY, COMPLETENESS,
REPRESENTATIVENESS, AND COMPARABILITY . . 1 of 3
4.1 Precision and Accuracy 1 of 3
4.2 Completeness . 1 of 3
4.3 Representativeness 3 of 3
4.4 Comparability 3 of 3
5.0 SAMPLING STRATEGY 1 of 10
5.1 Selection of Subregions for Sampling 1 of 10
5.2 Selection of Reaches for Potential
Sampling (First Stage Sample) 1 of 10
5.3 Selection of Sampling Sites (Second
Stage Sample) 4 of 10
5.4 Selection of Types and Locations of
Measurements 4 of 10
6.0 FIELD OPERATIONS 1 of 19
6.1 Sampling Team Activities 1 of 19
6.1.1 Field Base Activities Before
Sampling Trip 1 of 19
6.1.2 Stream Site Activities 1 of 19
Revision
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
-------
CONTENTS (Continued)
Section T of C
Revision 4
Date: 9/88
Page 2 of 4
Page
6.1.3 Field Base Activities After
Sampling Trip 9 of 19
6.2 Field Base and Mobile Processing
Laboratory Activities 9 of 19
6.2.1 Reagent Preparation 9 of 19
6.2.2 Sample Processing 9 of 19
6.3 Training 19 of 19
7.0 FIELD MEASUREMENT QUALITY CONTROL CHECKS.. 1 of 5
7.1 Stream Site Measurements . 1 of 5
7.1.1 Site Measurements of Chemical
Parameters 1 of 5
7.2 Mobile Processing Laboratory
Measurements 2 of 5
7.2.1 Dissolved Inorganic Carbon .... 2 of 5
7.2.2 pH 3 of 5
7.2.3 Turbidity 4 of 5
7.2.4 True Color 4 of 5
7.2.5 Nonexchangeable and Total
PCV-Reactive Aluminum 4 of 5
7.2.6 Specific Conductance 4 of 5
8.0 ANALYTICAL PROCEDURES 1 of 1
9.0 ANALYTICAL INTERNAL QUALITY CONTROL 1 of 13
9.1 Sample Receipt 1 of 13
9.2 Sample Analysis 1 of 13
9.3 Analytical Laboratory Documentation
for Quality Control 1 of 13
9.4 Internal Quality Control Within Each
Method 2 of 13
9.5 Overall Internal Quality Control 9 of 13
9.5.1 Anion-Cation Balance 9 of 13
9.5.2 Conductance Balance 9 of 13
9.6 Instrumental Detection Limits^ 11 of 13
9.7 Data Reporting 11 of 13
9.8 Daily Evaluation of Quality Control
Data 11 of 13
Revision
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
-------
CONTENTS (Continued)
Section T of C
Revision 4
Date: 9/88
Page 3 of 4
ion Page Revision
10.0 PERFORMANCE AND SYSTEM AUDITS 1 of 4 4
10.1 Performance Audit Samples 1 of 4 4
10.1.1 Field Audit Samples 1 of 4 4
10.1.2 Laboratory Audit Samples 1 of 4 4
10.1.3 Application of Audit Sample
Data 3 of 4 4
10.2 Quality Assurance System Audits
(On-Site Evaluations) 3 of 4 4
10.2.1 Field and Mobile Processing
Laboratory Operations
On-Site Evaluation 3 of 4 4
10.2.2 Analytical Laboratory On-Site
Evaluation 4 of 4 4
11.0 ACCEPTANCE CRITERIA 1 of 2 4
11.1 Audit Sample Results 1 of 2 4
11.2 Duplicate Analysis Results 2 of 2 4
11.3 Blank Analysis Results. 2 of 2 4
11.4 Corrective Action 2 of 2 4
12.0 DATA BASE MANAGEMENT SYSTEM 1 of 9 4
12.1 Raw Data Set 1 of 9 4
12.2 Verified Data Set 1 of 9 4
12.3 Validated Data Set 9 of 9 4
12.4 Final Data Set 9 of 9 4
13.0 DATA EVALUATION AND VERIFICATION 1 of 7 4
13.1 Field Data Review 1 of 7 4
13.2 Analytical Data Review 1 of 7 4
13.2.1 Daily Quality Assurance
Communications 1 of 7 4
13,2.2 Preliminary Review of Sample
Data Package 3 of 7 4
13.2.3 Review of Quality Assurance
and Quality Control Data 3 of 7 4
13.2.4 Computer Evaluation of DIG, pH,
ANC, and BNC Data by Proto-
fyte Analysis 3 of 7 4
-------
CONTENTS (Continued)
Section T of C
Revision 4
Date: 9/88
Page 4 of 4
Section
Page
Revision
13.2.5 Follow-up with Analytical
Laboratories 6 of 7
13.2.6 Evaluation of Outliers Gener-
ated by the Corvallis Staff ... 6 of 7
13.2.7 Preparation and Delivery of
Verification Tapes 6 of 7
14.0 DATA VALIDATION. 1 of 8
14.1 Overview 1 of 8
14.2 Detection of Outliers 2 of 8
14.2.1 Frequency Analyses 2 of 8
14.2.2 Univariate Analyses 3 of 8
14.2.3 Principal Components Analysis . . 3 of 8
14.2.4 Bivariate Analyses 5 of 8
14.2.5 Multivariate Analyses 5 of 8
14.3 Detection of Systematic Error 6 of 8
14.4 Treatment of Outliers and Systematic
Differences 7 of 8
15.0 REFERENCES 1 of 3
APPENDICES
APPENDIX A - Data Forms for Reporting
Analytical Results 1 of 16
APPENDIX B - Field Sampling and Mobile
Processing Laboratory On-Site
Evaluation Questionnaire 1 of 17
APPENDIX C - Analytical Laboratory On-Site
Evaluation Questionnaire. ..... 1 of 53
APPENDIX D - National Stream Survey Preaward
Audit Sample Scoring Sheet. .... 1 of 3
APPENDIX E - National Stream Survey
Verification Report . 1 of 16
4
4
4
4
4
4
4
4
4
4
-------
FIGURES
Figures Page
2.1 Organization of the National Surface Water
Survey 2 of 4
3.1 National Surface Water Survey internal
management structure 2 of 3
5.1 National Stream Survey subregions 2 of 10
5.2 Description of sampling procedure for
National Stream Survey study reaches 3 of 10
5.3 National Stream Survey Phase I - Pilot
Survey second stage sampling sites 5 of 10
5.4 National Stream Survey Mid-Atlantic Phase I
Survey second stage sampling sites 6 of 10
5.5 National Stream Survey Southeast Screening
Survey second stage sampling sites,
Southern Appalachian subregions 7 of 10
5.6 National Stream Survey Southeast Screening
Survey second stage sampling sites, Ozark
and Ouachita Mountains subregion 8 of 10
5.7 National Stream Survey Southeast Screening
Survey second stage sampling sites,
Florida subregion 9 of 10
6.1 Flowchart of sampling activities for the
National Stream Survey. 2 of 19
6.2 National Surface Water Survey Form 7 -
Watershed Characteristics 2 of 19
6.3 National Surface Water Survey Form 4 -
Stream Data 5 of 19
6.4 National Surface Water Survey Form 4A -
Hydrologic Data 5 of 19
Section Figures
Revision 4
Date: 9/88
Page 1 of 2
Revision
4
4
-------
FIGURES (Continued)
Figures
Page
6.5 National Surface Water Survey Form 6 -
Stream Episode Data 7 of 19
6.6 Field sample label 8 of 19
6.7 Flowchart of daily sampling and mobile
processing laboratory activities for the
National Stream Survey 10 of 19
6.8 Aliquot and Audit Sample Labels 13 of 19
6.9 National Surface Water Survey Form 5 -
Batch/QC Field Data (National Stream
Survey) 15 of 19
6.10 National Surface Water Survey Form 3 -
Shipping 17 of 19
6.11 Flow scheme for field data forms 18 of 19
12.1 Data base management for the National
Surface Water Survey 2 of 9
12.2 Aquatics Analysis System (AQUARIUS II) .... 8 of 9
13.1 Flowchart for data verification process 2 of 7
13.2 National Surface Water Survey Form 26 - Data
Confirmation/Reanalysis Request 7 of 7
14.1 Flowchart for data validation process 8 of 8
Section Figures
Revision 4
Date: 9/88
Page 2 of 2
Revision
4
4
4
4
4
4
4
4
4
4
4
-------
TABLES
Table
1.1 Sections in This Report and in the
Analytical Methods Manual that Address
Quality Assurance Subjects
4.1 Data Quality Objectives for Precision,
Accuracy, and Detectability
5.1 Parameters to be Measured During Phase I
of the National Stream Survey
6.1 Aliquots, Containers, Preservatives, and
Corresponding Parameters
6.2 Sample Codes
8.1 Parameters and Corresponding Measurement
Methods
9.1 Maximum Holding Times
9.2 Maximum Control Limits for Quality Control
Check Samples
9.3 Summary of Internal Quality Control Checks
for Analysis Methods
9.4 Data Forms Used by the Analytical
Laboratory.
9.5 Chemical Reanalysis Criteria
9.6 Conductance Factors of Ions
9.7 List of Decimal Place Reporting
Requirements
9.8 National Surface Water Survey Lab/Field
Data Qualifiers (Tags)
10.1 Desired Composition Range of Synthetic
Field and Laboratory Audit Samples for
the National Stream Survey
Page
2 of 2
2 of 3
10 of 10
12 of 19
14 of 19
1 of 1
2 of 13
3 of 13
4 of 13
5 of 13
7 of 13
10 of 13
11 of 13
12 of 13
2 of 4
Section Tables
Revision 4
Date: 9/88
Page 1 of 2
Revision
4
4
4
4
4
4
4
-------
Section Tables
Revision 4
Date: 9/88
Page 2 of 2
Table
13.1
14.1
TABLES (Continued)
12.1 National Surface Water Survey Verification
Data Qualifers (Flags) for Raw Data Set .
Exception Generating and Data Review
Programs of AQUARIUS II
Some Physical Variables Subject to
Validation
14.2 Pairs of Variables Used to Check for
Random and Systematic Errors. .
14.3 Related Groups of Variables Used in
Multivariate Analyses
Paae
3 of 9
4 of 7
1 of 8
4 of 8
5 of 8
Revision
4
4
4
4
4
-------
10 INTRODUCTION
Data published in previous studies are
consistent with the hypothesis that certain
surface waters within the United States
have decreased in pH, alkalinity, or both
over time. Acidic deposition has been
suggested as a contributor to such decreases.
Also, numerous studies have led to the
conclusion that the effects of acidic depo-
sition on surface water chemistry are influ-
enced by variations among associated lake,
stream, and watershed characteristics (U.S.
EPA 1984a and 1984b). Attempts have been
made to extrapolate local studies to the
regional and national scale and thus to
provide a quantitative estimate of the risk
to aquatic resources (especially fish) from
acidic deposition. These assessments have
had only limited success because of problems
associated with (1) the comparability of
the sampling and analytical methodologies
used in different studies, (2) the possibility
of biased or nonrepresentative sampling
sites, and (3) small and incomplete data
bases.
The National Surface Water Survey
(NSWS) of the National Acid Precipitation
Assessment Program (NAPAP), Task Group
E (Aquatic Effects) is designed to overcome
some of these deficiencies. NSWS is a
three-phase project to evaluate the present
water chemistry of lakes and streams, deter-
mine the status of fisheries and other biotic
resources, and select regionally represen-
tative surface waters for a long-term moni-
toring program to study future changes in
aquatic resources.
Because of logistical and systematic
differences between lakes and streams, NSWS
was separated into lake and stream survey
components. The first phase of the National
Stream Survey (NSS) is a synoptic survey
of the chemistry of streams and includes
(1) a pilot survey in the southern Blue Ridge
Province, (2) a full-scale survey in the Mid-
Atlantic states, (3) a screening survey in the
Section 1.0
Revision 4
Date: 9/86
Page 1 of 2
Southeast, and (4) an episodes pilot survey
in the Mid-Atlantic states. This manual
delineates the quality assurance (QA) plan
for these four NSS components. A descrip-
tion of the project and its organization is
given in the following sections.
The QA policy of the Environmental
Protection Agency (EPA) requires that every
monitoring and measurement project have
a written and approved QA project plan
(Costle, 1979a and 1979b). This requirement
applies to all environmental monitoring and
measurement efforts authorized or supported
by EPA through regulations, grants, contracts,
or other formal means. The QA project
plan should specify the policies, organization,
objectives, functional activities, QA activities,
and quality control (QC) activities designed
to achieve the data quality goals of the
project. All project personnel should be
familiar with the policies and objectives
outlined in the QA project plan to ensure
proper interaction of the sampling operations,
laboratory operations, and data management.
EPA guidance states that the 16 items
shown in Table 1.1 should be addressed In
the QA project plan (U.S. EPA, 1980). Some
of these items are extensively addressed
in the analytical methods manual (Hillman
et al., 1986) for this project; therefore,
as allowed by the guidelines, method-specific
discussions are not repeated in this document.
-------
Section 1.0
Revision 4
Date: 9/86
Page 2 of 2
Table 1.1. Sections In this Report and In the Analytical Methods Manual that Address Quality
Assurance Subjects
Subject
Title Pace
Table of Contents
Project Description
Project Organization
and Responsibility
QA Objectives
Samplina Procedures
Sample Custody
Calibration Procedures
Analytical Procedures
Data Analysis, Validation,
and Reporting
Internal QC Checks
Performance and System Audits
Preventive Maintenance
Assessment of Precision, Accuracy
and Completeness
Corrective Actions
QA Reports to Management
Section
This Report
Tof C
2
3
4
6
6. 9
6, 7. 9
8
6, 9, 12,
13. 14
7, 9
10
6
4. 11
9. 11
9, 10
Methods Manual
1
2
2. 3
2-13
4-13
3
3
2.3
3
-------
Section 2.0
Revision 4
Date: 9/86
Page 1 of 4
2.0 PROJECT DESCRIPTION
Figure 2.1 shows the structure of NSWS
as presently planned.
This document defines the QA require-
ments for four segments of NSS: (1) the
Phase I - Pilot, a survey of approximately
60 streams; (2) the Mid-Atlantic Phase I, a
survey of 250 streams; (3) the Southeast
Screening, a survey of 200 streams; and
(4) the Episodes Pilot, a study of the mag-
nitude, frequency, duration, and causes of
episodic events in streams. The following
sections describe these four studies, deli-
neate the objectives to be achieved by each
study, and present the measures specified
by the NSS QA project plan.
2.1 Components of the National
Stream Survey
2.1.1 Phase I - Pilot Survey
The Phase I - Pilot Survey, which
was conducted in the Southern Blue Ridge
Province during spring 1985, had the follow-
ing primary objectives:
* Test the statistical sampling design.
• Test all methods that will be used
during the survey.
« Finalize the data quality objectives
(DQOs) for Phase I.
* Finalize the QA/QC guidelines for
Phase I,
• Train personnel for the field operations.
• Test the data analysis plan.
The Phase 1 - Pilot Survey is described in
detail in Messer et al., 1986.
2.1.2 Mid-Atlantic Phase / Survey
The Mid-Atlantic Phase I Survey in-
volves sampling approximately 250 stream
reaches in an area bounded by the Catskill
and Pocono Mountains to the north, the
North Carolina-Virginia state line to the
south, the approximate western boundaries
of Pennsylvania and West Virginia to the
west, and the Atlantic Ocean. This region
is expected to contain many areas of low
acid-neutralizing capacity (ANC) waters,
and it is subject to relatively high levels
of acidic deposition. Each stream reach is
sampled twice during spring baseflow condi-
tions, at its upstream and its downstream
nodes.
The primary objectives of the Mid-
Atlantic Phase 1 survey are those of the
overall NSS-Phase I:
• Determine the percentage, extent (miles,
drainage area), and location of streams
that are in potentially acid-sensitive
regions of the United States and that
are presently acidic.
• Determine the percentage, extent, and
location of such streams that are
presently characterized by low alkalinity
and that may become acidic in the
future.
• Determine which streams are represen-
tative of typical subpopulations of
streams in a particular region and
from these identify which streams should
be more intensively studied.
2.1,3 Southeast Screening Survey
It is necessary to decide whether
and where to extend Phase I sampling. The
-------
Section 2.0
Revision 4
Date: 9/86
Page 2 of 4
NATIONAL SURFACE WATER SURVEY (NSWS)
NATIONAL LAKE SURVEY (NLS)
Phase I - Synoptic Survey
Eastern Lakes (1984)
Western Lakes (1985)
Phase II
Temporal Variability (1986-87)
Biological Resources (1986)
NATIONAL STREAM SURVEY (NSS)
Phase I - Synoptic Survey
Pilot Survey (1985)
Synoptic Survey (1986)
Southeast Screening (1986)
Episodes Pilot (1986)
Episodic Effects (1988)
Biological Resources (1988)
Long-Term Monitoring (1988)
Figure 2.1. Organization of the National Surface Water Survey.
-------
Section 2.0
Revision 4
Date: 9/86
Page 3 of 4
National Lake Survey will be useful in
targeting potential areas of interest in the
Northeast, Midwest, and West. However,
virtually no data are available for the south-
eastern states other than historical data
depicted on the regional alkalinity maps
and from the Phase I - Pilot Survey.
The purpose of the Southeast Screening
Survey is to prioritize other potential NSS
regions in the Southeast for full Phase 1
study. The statistical design will allow
regional characterization, just as in the
Phase I study area. However, each stream
is sampled only once (at two sites); thus,
no temporal variance estimate will be avail-
able. The single sample is not expected to
provide enough information to allow robust
classification of the Phase I streams for
later study. Also only crude discharge
estimates are made in the screening area,
and streams are sampled only under baseflow
conditions.
The screening area comprises the
Southern Appalachian Mountains (exclusive
of the Phase I - Pilot Survey area), the
Piedmont, the Ouachita and Ozark Moun-
tains, and parts of Florida, Approximately
400 samples will be collected from the
screening area from upstream and down-
stream nodes on each of 200 streams.
The primary objectives of the Southeast
Screening Survey are the same as the objec-
tives of the Mid-Atlantic Phase 1 Survey
and of NSS-Phase I, given in section 2.1.2.
2.1.4 Episodes Pilot Survey
The Episodic Response Project (of
NAPAP Task Group E3) has as its primary
objective predicting the magnitude, fre-
quency, duration, and causes of episodic
events in lakes and streams. These events
cause marked shifts from baseflow conditions
in pH and associated parameters. In areas
dominated by snowpack, such episodes result
from snowmelt, often exacerbated by warm
days or rainfall. In warmer regions, epi-
sodes are usually associated with rain events.
The field work described here represents a
pilot effort aimed at providing a preliminary
assessment of the frequency, duration, and
causes of such episodes in the Mid-Atlantic
states. It also is expected to provide suffi-
cient information about design and logistics
aspects to allow a cost-effective, full-scale,
regional episodes survey (or surveys) to be
implemented.
There are several ways to estimate
the number and frequency of episodes, and
each method has its own set of conceptual
and logistical difficulties. A pilot survey
is valuable because it helps answer questions
about survey design. Thus, a pilot survey
helps produce an efficient design for a full-
scale episodic effects field effort. It is
anticipated that approximately five storm-
fronts wilt be suitable for event sampling
during the stream survey. Six episode teams
will sample each event, so 30 sets of event
data should be obtained. Each team will
collect 4 samples per event under ideal
conditions; thus, approximately 120 samples
should be collected for the Episodes Pilot
Survey. In addition to collection of water
samples, pH, specific conductance, and
dissolved oxygen concentration will be
determined at 30-minute intervals throughout
each event.
2.2 Data Quality Objectives
DQOs, in terms of anticipated value
range, detection limits, and precision, were
defined for each measurement parameter in
early 1985. These QA goals were originally
based on published literature, statistical
error propagation, and Eastern Lake Survey
findings. Equipment, sampling protocols,
and analytical methodologies were selected
and standardized in order to achieve the
DQOs. The Phase I - Pilot Survey provided
the opportunity to evaluate and revise
-------
Section 2.0
Revision 4
Date: 9/86
Page 4 of 4
methodologies, equipment, and DQOs. The
observed range of values from the Phase I
- Pilot Survey, required detection limits,
and relative intrataboratory precision goals •
for each variable are summarized in Table
4.1.
•
2.3 Specifications of the Quality
Assurance Project Plan
To ensure that procedures are per-
formed consistently and that the quality of •
the data generated can be determined, the
QA project plan for NSS specifies the fol- •
lowing measures:
• Provide detailed, written sampling
methodology (protocols are documented
in Hagley et al., 1986). •
• Simultaneously train all personnel who
will participate in field activities.
•
• Conduct site visits to each field opera-
tions base throughout the sampling
period to ensure that all methods are
being performed properly. «
• Perform extensive evaluation of analy-
tical laboratories before their selection
and throughout their participation.
• Assess variability introduced at each
level of activity in the mobile pro-
cessing and analytical laboratories by
preparing (if necessary) and analyz-
ing audit samples (synthetic samples
and natural lake samples), duplicates,
and blanks along with routine samples.
NOTE: "Mobile processing laboratory"
refers to the laboratory complex of
trailers located in Las Vegas, Nevada,
where sample processing and prelimi-
nary analyses are performed. For the
Phase I - Pilot Survey, the mobile
laboratory was located in the field.
"Analytical laboratory" refers to the
off-site contract laboratory that per-
forms the more sophisticated analyses.
Provide detailed, written analytical
methodology (Hillman et al., 1986).
Use internal QC procedures at the
analytical laboratory to detect potential
contamination and to verify established
detection limits.
Enforce holding time requirements.
Use protocols in the field, in the mobile
processing laboratory, and in the analy-
tical laboratories to confirm that
reported data are correct.
Enter data into the data base twice,
and scan for outlying values to detect
and eliminate transcription errors.
Verify data by means of range checks,
internal consistency checks, and QA
evaluations.
Validate verified data by analyzing
the reasonableness of data; base the
analysis on the values expected for
the particular region or subregion
involved.
-------
Section 3.0
Revision 4
Date: 9/86
Page 1 of 3
3.0 PROJECT ORGANIZATION
Figure 3.1 illustrates the NSWS man-
agement structure. The program director
is the EPA official who has overall respon-
sibility for the program. The responsibilities
of the program manager, technical director,
and administrative coordinator are as fol-
lows:
Program Manager
The program manager is the EPA
Headquarters representative for NSWS and
serves as the liaison among the headquarters
staff, the laboratory directors, and NAPAP.
Questions regarding general management
and resources should be forwarded to the
program manager through the technical
director.
Technical Director
The technical director performs program
responsibilities at the discretion of the
program manager. The primary role is to
see that the program objectives are satisfied,
that the components of the program are
well-integrated, and that deadlines are met.
The technical director coordinates and
integrates the activities of the Environ-
mental Research Laboratory at Corvallis,
Oregon (ERL-C), the Environmental Monitor-
ing Systems Laboratory at Las Vegas, Nevada
(EMSL-LV), and the Oak Ridge National
Laboratory (ORNL) at Oak Ridge, Tennessee.
The technical director also coordinates peer
review and resolves issues of responsibility.
The office is the focal point of general
public inquiry and distribution of report
information. The technical director repre-
sents the program manager as necessary
and keeps the program manager informed
of EPA laboratory activities, progress, and
performance.
Administrative Coordinator
The administrative coordinator reports
directly to the program manager. The
primary rote of this position is to monitor
the budget and personnel needs of the survey
staff. The administrative coordinator also
makes contractual arrangements at the head-
quarters level and provides special services
as needed.
The roles of the laboratories are as follows:
Environmental Research Laboratory, Corvaliis
In view of the role the Corvallis lab-
oratory plays in the Agency's acidic depo-
sition research program and the major roles
it must perform during the survey program,
it is appropriate that ERL-C becomes a
primary focal point for NSS. Its respon-
sibilities for al! phases of NSWS
include:
• developing the sampling design
• selecting the sampling sites
• preparing sampling protocols (jointly
with EMSL-LV)
• collecting supplemental historical and
other available data on each sampling
site (aquatic and terrestrial components)
• analyzing data (jointly with EMSL-LV)
• interpreting data and maps
• preparing reports (final and progress
reports, with contributions from the
other laboratories relative to their
responsibilities)
-------
Section 3.0
Revision 4
Date: 9/86
Page 2 of 3
Program Director
NAPAP Acid Deposition
Assessment Staff
Program Manager
NAPAP Task
Group E
Administrative
Coordinator
Peer Review
ERL-C
Sampling and Design
Site Selection
Site Description
Data Validation
Data Interpretation
Reporting
EMSL-LV
Operations and
Logistics
Analytical Methods
Data Verification
Data Validation
QA/QC (including
QA Manager)
ORNL
and
ERL-C
Data Management
Figure 3.1. National Surface Water Survey Internal management atrueture.
-------
Section 3.0
Revision 4
Date: 9/86
Page 3 of 3
assessing and resolving all science-
related issues other than QA/QC and
data management (jointly with other
laboratories as necessary)
coordinating survey activities
NAPAP management staff.
with
Environmental Monitoring Systems Labo-
ratory, Las Vegas
The Las Vegas laboratory has particular
expertise in matters relating to QA/QC,
logistics, analytical services, and sampling
protocols. The responsibilities of this labo-
ratory, for all phases of NSWS, include:
• developing QA/QC procedures for all
components of the program except data
management (ORNL and ERL-C)
• preparing all sampling protocols (jointly
with ERL-C)
• preparing the analytical methods manual
• preparing the field training and oper-
ations manual
• preparing the mobile processing labo-
ratory operations manual for those
component surveys for which such
information is not included in the
other manuals.
• preparing and implementing the QA
project plan
• coordinating logistical support and
equipment needs for all field operations
• training sampling personnel
• developing and implementing QA/QC
procedures for verifying all field meas-
urements and analytical laboratory data
• independently assessing field measure-
ments and laboratory data quality (bias
and variability)
• assessing and resolving all problems
pertaining to QA/QC, logistics, and
analytical services.
Oak Ridge National Laboratory
ORNL has considerable expertise in
managing large data bases, manipulating
data, and restructuring data bases to satisfy
data analysis needs. ERL-C oversees the
activities of ORNL, which has NSWS res-
ponsibilities for:
• developing and maintaining a data
management system
• entering all field, laboratory, and
support data into the data base, and
simultaneously assuring data quality
• preparing computer-generated summary
tables, statistics, and graphics for
reports.
distributing all samples to analytical
laboratories
-------
Section 4.0
Revision 4
Date: 9/86
Page 1 of 3
4.0 QUALITY ASSURANCE OBJECTIVES FOR PRECISION, ACCURACY,
COMPLETENESS, REPRESENTATIVENESS, AND COMPARABILITY
4.1 Precision and Accuracy
The QA objectives for precision and
accuracy of the parameters being measured
are given in Table 4.1. Precision and accu-
racy are determined in part by analyzing
data from QA/QC samples. QA samples
include the following:
• Field blank - A field blank is a deion-
ized water sample that meets specif-
ications for ASTM Type 1 reagent
water (ASTM, 1984). The blank sample
is carried to the stream and is
processed through the sampling device
as though it were a routine sample.
One field blank is collected for each
sampling region on each operating
day. Field blank data are used to
establish the estimated system decision
limit (the lowest instrument signal
that can be distinguished from the
background), the quantitation limit
(the lowest concentration of an analyte
that can be measured with reasonable
precision), and the system detection
limit (the lowest concentration that
can be measured above the system
decision limit) that can be expected
for each type of analysis. For data
interpretation, a data point for a field
blank above the expected value is
considered a positive response.
• Field duplicate - A field duplicate is
a second sample collected at the stream
site by the same team immediately
after the routine sample is collected.
Field duplicate data are used to esti-
mate the overall within-batch prec-
ision for the sampling and analysis
process. One field duplicate is col-
lected per field base per operating day.
• Audit samples - Audit samples are
materials with known characteristics,
used to determine the accuracy of
the measurement system. Two types
of audit samples serve as QA checks
for NSS: field audit samples (natural
and synthetic) help to check the overall
field and laboratory performance; lab-
oratory audit samples (natural and syn-
thetic) help to check the performance
of the analytical laboratory. Audit
samples are discussed in Section 10.0.
• Trailer duplicates; trailer blanks-
Trailer duplicates and trailer blanks
are used to check the precision of
mobile processing laboratory measure-
ments.
Field QA/QC samples are used primarily
by the sampling teams and mobile processing
laboratory staff to check the accuracy of
the measurement system in the field. Field
QA/QC samples include trailer duplicates,
trailer (calibration) blanks, quality control
check samples (QCCSs) for pH and specific
conductance at the site, and mobile proc-
essing laboratory QCCSs (for pH, DIG, and
turbidity). These samples are described
in Sections 6 and 7.
Analytical laboratory QA/QC samples
include calibration blanks, reagent blanks,
analytical laboratory duplicates, QCCSs,
and detection limit QCCSs. These are
described in Section 9.0.
4.2 Completeness
The objective for completeness of data
(the amount of valid data obtained from a
measurement system compared to the amount
expected to be obtained under correct normal
conditions) is 90 percent or better for
all parameters. This figure is based on
experience gained during previous studies
and is subject to change during the
survey.
-------
Table 4.1. Data
Parameter*
Al, Total
extractable
Al. Total
Al, Non-
exchangeable
and total
PCV reactive
ANC
BNC
Ca
cr
True Color
DIG
00
DOC
F', Total
dissolved
Fe
K
Mg
Mn
Na
NH4+
N03-
pH, Field
pH, Analytical
lab
P, Total
dissolved''
Quality Objectives' for Precision, Accuracy, and Oetectablllty
Units
mg/L
mg/L
mg/L
/jeq/L
/jeq/L
mg/L
mg/L
PCU'
rng/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
pH units
pH units
mg/L
Observed
Range"
0.0-0.3
0.005-75
—
-8-3,000
0.0-400
0.3-50
0.3-30
0-750
0.1-30
7.1-12.1
0.1-8
0.0-0.2
0.0-0.6
0.2-5
0.2-15
0.0-0.2
0.4-22
0.0-0.4
0.0-9
5-8.5
5-8.5
0.002-1.5
Required
Detection
Limits
0.005
0.005
0.010
e
e
0.01
0.01
0
0.05
_.
0.1
0.005
0.01
0.01
0.01
0.01
0.01
0.01
0.005
—
—
0.002
Precision
Relative
Standard Deviation (BSD)
Upper Limit (%}"
10 (if cone.
20 (if cone.
10 (if cone.
20 (if cone.
10 (if cone.
20 (if cone.
10
10
5
5
±sff
10
5
5 (M cone.
10 (if cone.
5
10
S
5
10
5
5
10
+0.1 pH^
±0.05 pH^
10 (if cone.
20 (if cone.
>0.01 mg/L)
<0.01 mg/L}
>0.01 mg/L)
0.01 mg/L)
.<0.01 mg/L)
>5 mg/L)
£5 mg/L)
>0.01 mg/L)
<0.01 mg/L)
Section 4.0
Revision 4
Date: 9/86
Page 2 of 3
Accuracy
Maximum Absolute
Bias (%)
10/20
10/20
10/20
10
10
10
10
..
10
5
10
10
10
10
10
10
10
10
10
±0.1 pH^
±0.1 pH^
10
20
(Continued)
-------
Table 4.1. (Continued)
Section 4.0
Revision 4
Date: 9/86
Page 3 of 3
Parameter*
Units
Observed
Range"
Required Precision Relative
Detection Standard Deviation (RSD)
Limits Upper Limit (%f
Conductance
Accuracy
Maximum Absolute
Bias (%)
SiO2
S042"
Specific
mg/L
mg/L
fJS/cm
3-25
0.3-20
7-300
0.05
0.05
'
5
5
2
10
10
5
Turbidity
NTU
0.1-1,800
10
10
* The objective for completeness of data is 90 percent or better for all parameters.
13 Dissolved ions and metals are being determined, except where noted.
0 Range of values observed in stream waters during Phase I - Pilot Survey.
d Unless otherwise noted, this is the RSD at concentrations greater than 10 times the required
detection limits.
* Absolute blank value must be <10 peq/L
' PCU = platinum-cobalt units.
9 Absolute precision/accuracy goal in terms of applicable units.
h For the Phase I - Pilot Survey, total P is determined (samples were unfiltered). For the Mid-Atlantic
Phase I, Southeast Screening, and Episodes Pilot Surveys, total dissolved P is determined (samples
are filtered).
1 The mean of six nonconsecutive blank measurements must not exceed 0.9 pS/cm.
4.3 Representativeness
The question of representativeness is
an important one. NSS is designed to
achieve the objectives outlined in Section
2. It is not intended to determine the
chemistry of any given reach in detail but
to obtain a good index of stream chemistry
for each reach, so that reaches can be
classified correctly. Therefore, achieving
survey objectives does not require that the
samples taken from a reach be completely
representative of those waters. Only two
samples from each of two sites per reach
(the upstream and downstream nodes) are
taken during the Mid-Atlantic Phase I Sur-
vey. For the Episodes Pilot Survey, a
sample is taken at the downstream node
during each of the following episode hydro)-
ogic stages: baseflow, rising, peak, and
falling. For the Southeast Screening Survey,
each reach is sampled only once at the
upstream node and once at the downstream
node. However, a determination of whether
this level of sampling is sufficient to achieve
the objectives of Phase I can be made only
when estimates of "within reach" and "among
reach" variances are obtained. Estimates
of these variances will be made using cur-
rently available data and the statistical
sampling design for Phase I, which will be
modified if necessary (see Section 5). In
later NSS phases, more intensive studies
of individual reaches will be performed.
4.4 Comparability
Comparability is assured by having a
uniform set of procedures for all sampling
teams and a uniform set of units for report-
ing the data. Furthermore, the QA pro-
cedures described in succeeding sections
allow for the determination of bias for each
sampling team and analytical laboratory
so that their results can be compared.
-------
5.0 SAMPLING STRATEGY
This section provides a summary of
the sampling design. More complete discus-
sions are presented in the draft research
plan (U.S. EPA, 1985) and in the draft
sampling plan (Overton, 1985).
5.1 Selection of Subregions
Sampling
for
The first process in designing the
sampling strategy is to determine what
geographic regions and subregions are to
be sampled. Highest priority was given to
regions in which a majority of "surface
waters were expected to have low ANC
based on current EPA alkalinity maps
(Omernik and Powers, 1983).
The Phase I • Pilot Survey viias con-
ducted in subregion 2A(S) - the (Southern
Blue Ridge Province.
The subregions selected for the Mid-
Atlantic Phase I Survey are the (following
(see Figure 5.1.):
• the Pocono and Catskill Mountains (10)
• the Pine Barrens and Chesapeake Bay
(3B)
• the northern portion of the Valjey
and Ridge Province (2B(N))
• the northern portion of the
Plateau (2C(N)).
alachian
Four stream populations are targeted
for the Southeastern Screening Suryey:
• the Piedmont (3A)
• the Ozark and Ouachita Mountains
(2D)
Section 5.0
Revision 4
Date: 9/86
Page 1 of 10
• the coastal plain in Florida (3C)
• the Southern Appalachian Plateau (2A(N),
2B(S), and 2C(S)).
Subregions selected for sampling during
the Episodes Pilot Survey are the same as
those selected for the Mid-Atlantic Phase
I Survey.
5.2 Selection of Reaches for
Potential Sampling (First
Stage Sample)
The stream population of interest drains
watersheds of several tenths to tens of
square miles (102 to 104 ha) with a maximum
of 60 square miles (1.55 x 104 ha).
A point-frame sampling design is used
to select reaches for possible sampling (see
Figure 5.2). A reach is defined as a blue-
line stream (as indicated on a 1:250,000-
scale topographic map) that lies between
the confluences of two tributaries. The
point frame is a rectangular grid of dots
on an acetate transparency that is placed
at random on a 1:250,000 scale topographic
map. A reach is included in the first stage
sample if it is intersected by a line drawn
perpendicular to the elevation contours,
proceeding from a grid point in a downslope
direction (Figure 5.2). This sampling design
produces a probability sample of reaches
with an expected frequency of inclusion
proportional to the direct drainage area,
a1t of each reach. Boundary and reservoir
reaches, large rivers, tidal reaches, urban
drainages, and severely polluted reaches
(e.g., reaches dominated by acid mine drain-
age or fossil fuel brines) are excluded from
the base of reaches for possible sampling.
These reaches were excluded using low pH
(less than 3.3) or high conductance (greater
-------
o. <
-------
Section 5.0
Revision 4
Date: 9/86
3 of 10
8
MILES
GEOGRAPHIC
SAMPLING
FRAME POINTS
SAMPLING
POINT
TOPOGRAPHIC
WATERSHED
BOUNDARY
LENGTH OF SELECTED
REACH. 1
31 ^DIRECT DRAINAGE
a2 = INDIRECT DRAINAGE
Figure 5.2. Description of sampling procedure for National Stream Survey atudy reaches. The aampllng-frarm
potato correapond to a uniform geographic grid. Th» tower laft point result* In Incluaton of tha
watershed ahown, provided a1 + a2 to toaa than 60 ml2 (1.55 x 10* ha).
-------
Section 5.0
Revision 4
Date: 9/86
Page 4 of 10
than 500 juS/cm for inland areas and greater
than 250 pS/cm for tidal areas) as criteria.
5.3 Selection of Sampling Sites
(Second Stage Sample)
The next task is to select a subpopu-
lation of the first stage sample upon which
to make physical measurements. A sample
size of 50 per stratum has proven to be
useful in earlier components of NSWS. To
achieve a more homogeneous inclusion pro-
bability, the selections are weighted with a
conditional inclusion probability inversely
proportional to the area of direct watershed.
An additional stratum of reaches with ANC
less than 50 peqfi is also included in the
survey, and 29 "special interest" reaches
will be sampled. These reaches include
selected sites in the ongoing long-term
monitoring project of NAPAP Task Group
E, in which storm events will also be mon-
itored, as well as other intensively studied
watersheds in the area. Second stage sam-
pling sites for the Phase I - Pilot Survey
are shown in Figure 5.3, those for the Mid-
Atlantic Phase I Survey are shown in Figure
5.4, and those for the Southeast Screening
Survey are shown in Figures 5.5, 5.6, and
5.7.
5.4 Selection of Types and
Locations of Measurements
The third stage sampling design
involves the designation of what physical
and chemical measurements should be taken
on each second stage reach, and when and
where to take them in order to best char-
acterize the reach. Table 5.1 lists the
parameters that will be measured for ail
NSS samples.
A point sample at the downstream node
may not represent the chemistry of the
reach lying above ft. Within-reach chemistry
may change along the reach as a result of
instream processes (e.g., primary produc-
tivity), contributions from other streams
that do not appear on 1:250,000-scale maps,
and inputs from springs and seeps that feed
the reach. For population description, it
is the way that a particular chemical value
characterizes the reach length (or some
transformation such as habitat area) that
is of interest. Therefore, variation in
chemical values along the reach must be
measured or inferred. For the Mid-Atlantic
Phase I Survey, samples are collected during
the spring sampling season on each of two
sampling dates at the upstream and down-
stream nodes of each reach. For the Phase
I surveys, streams are not sampled during
major hydrologic events (defined as flows
in excess of twice spring baseflow). Sam-
pling during these events is conducted as
part of the Episodes Pilot Survey.
Because the Southeast Screening Survey
is targeted primarily at producing population
estimates, only one sample is collected.
Upstream and downstream nodes are sampled
on each reach to provide population esti-
mates based on lengths of reaches in each
criterion range. In addition, sampling the
upstream node provides information about
watersheds too small to be included as direct
drainages in the NSS target population.
These very small headwaters watersheds
may represent important "early warning"
indicators in regions that are not receiving
high inputs of atmospheric acids.
-------
Section 5.0
Revision 4
Date: 9/86
Page 5 of 10
STAGE I SITES
O SPECIAL INTEREST SITES
Figure 5.3. National Stream Survey Phase I - Pilot Survey eeeond stage sampling sites.
-------
Section 5.0
Revision 4
Date: 9/86
Page 6 of 10
r
Atlantic
Ocean
TN
4- LOW ANC SITES
O STAGE II SITES
Figure 5.4. National Stream Survey Mid-Atlantic Phase 1 Survey second stage sampling sites. Note: Low ANC
sites have ANC less than 50 jieq/U
-------
Section 5.0
Revision 4
Date: 9/86
Page 7 of 10
-------
Section 5.0
Revision 4
Date: 9/36
Page 8 of 10
O STAGE II
SITES
LA
Figure 5.6. National Stream Survey Southeast Screening Survey second stage sampling sites, Ozark and Ouachlta
Mountains subragton.
-------
• STAGE JL SITES
Figure 5.7. National Stream Survey Southeast Screening Survey second stage sampling sites, Florida subreglon.
-------
Section 5.0
Revision 4
Date: 9/86
Page 10 of 10
Table 5.1. Parameter* to be Measured During Phase I of the National Stream Survey
Site Measurement In Situ Measurement Laboratory Measurement
gauge height (stage- temperature pH (closed)*
Phase I surveys only) specific conductance pH (air equilibrated)
stream width dissolved oxygen pH (open system)
stream depth DIG*
land use DIG, air equilibrated
bank vegetation DOC
stream substrate true color5
stream substrate Streamside Measurement turbidity*
cloud cover specific conductance*
weather conditions pH (open system) ANC
watershed disturbances BNC
elevation aluminum (total)
aluminum (total extractable)
aluminum (nonexchangeable
and total PCV reactive)*
aluminum (organic extractable)
calcium
magnesium
potassium
sodium
nitrate
sulfate
chloride
fluoride
silica
iron
ammonium
manganese
phosphorus ( total dissolved* )
* Determined at the mobile processing laboratory only.
b Determined at the mobile processing laboratory and at the contract analytical laboratory.
0 For the Phase 1 - Pilot Survey, total phosphorus was determined (samples were unfiltered). For the
Mid-Atlantic Phase I, Southeast Screening and Episodes Pilot Surveys, total dissolved phosphorus was
determined (samples were filtered).
-------
6.0 FIELD OPERATIONS
Field operations are conducted at four
mobile field bases. Two of these field bases
are located in Phase I areas. Each of these
is staffed by a base coordinator, a logistics-
coordinator, and 5 two-person sampling
teams. The other two field bases are locat-
ed in Southeast Screening areas and are
each stalled by a base coordinator, a logis-
tics coordinator, and 2 two-person sampling
teams. The base coordinator is responsible
for the overall operation of the field base.
The logistics coordinator assists the base
coordinator. Each sampling team visits one
or two streams a day. Each stream is
sampled at its upstream and its downstream
node. The sampling teams from each Phase
I field base sample an average of seven
streams per day, anof the teams if cm each
Southeast Screening base station sample an
average of three streams per day. The
overall total sample load under these average
conditions is 44, including 2 duplicates and
2 blanks.
In the following sections, the activities
of the field crews are summarized. A more
detailed description of the field operations
is given in the field manual (Hagley et at.,
t988J.
6.1 Sampling Team Activities
Each sampling crew consists of two
scientists (except during episodics). It is
their responsibility to perform all sampling
operations correctly and to accurately record
all data. The scientists must be qualified
to operate all equipment and to follow
prescribed procedures.
For each sampling trip, the activities
of the sampling team are divided into (1) ac-
tivities at the field base conducted prior
to arrival at the stream site, (2) activities
at the stream site, and (3) activities
following sample collection. The following
Section 6.0
Revision 4
Date: 9/86
Page 1 of 19
subsections describe the activities in detail.
A flow scheme of sampling team activities
is shown in Figure 6.1
6.1.1 Field Base Activities Before
Sampling Trip
Prior to leaving the field base the
sampling team:
* Prepares a daily itinerary of sites to
be sampled, routes of travel to and
between sites, and personal identifi-
cation information. This itinerary is
given to the field base coordinator.
• Ensures that all necessary equipment
and supplies are present.
« Calibrates pH, and dissolved oxygen
(DO) meters used to obtain pH, temper-
ature, and DO measurements of each
stream; checks the calibration of the
conductance and How meters used to
measure specific conductance and f/ow
velocity. The QA/QC procedures are
described in detail in Section 7 of this
QA plan. QCCS, pH calibration buffers,
and any other reagents needed at the
site are prepared previously in a labor-
atory or are purchased. No reagent
preparation is performed in the field.
* Reviews a detailed reconnaissance sheet
that describes roads and trails and
gives distances to the sampling site.
6.1.2 Stream Site Activities
On the first visit to each stream site,
watershed characteristics 5 including elevation
location, stream width and mean depth,
disturbances, vegetative cover, and stream
substrate are noted on Form 7 (Watershed
Characteristics - Figure 6.2). Stream stage
(for Phase I streams), percent cloud cover,
-------
Section 6.0
Revision 4
Date: 9/86
Page 2 of 19
PERFORM CALIBRATION
AND QC CHECKS
. T
„, LOAD VEHICLES.
TRAVEL TO STREAM SITE
f
DESCRIBE SITE 1
t
INSTALL STEEL ROD.!.8
READ STAGE B
t
CONDUCT HYDROLOGY 2'B
MEASUREMENTS
^
SET UPINSTRUMENTS.PERFORM
ON-SITE CALIBRATIONS AND
QC CHECKS. PURGE PUMP
T
MEASURE pH
I »
f
MEASURE COLLf
SHtCIMC CONUUCI ANCt ~°
1 DUF
UlSSOLVtU OXYGtN ""p'ft.
1
t
READ STAGE.8
REMOVE STEEL ROD2.B
t
PACK EQUIPMENT
*
L CHECK DATA FORMS
^Jt^
CALL .. NO ^^.-r oiT^V. YFS
BASE COORDINATOR 'X. .S*
I^FIRST SITE VISIT
AMPLER 2
»
COLLECT
LANK SAMPLE
F NECESSARY)
*
:CT ROUTINE SAMPLE
t
COLLECT
'LICATE SAMPLE
F NECESSARY)
t
CK UP SAMPLES
,. -1
CALL BASE COORDINATOR,
RETURN TO BASE SITE
t
PERFORM FINAL QC CHECKS,
PACK AND SHIP SAMPLES.
PREPARE FOR NEXT DAY
2=SECOND SITE VISIT
A=UPSTREAM SITE
B=DOWNSTREAM SITE
Figure 6.1 Flowchart of sampling activities for the National Stream Survey.
-------
Section 6.0
Revision 4
Date: 9/86
Page 3 of 19
NATIONAL SURFACE WATER SURVEY
WATERSHED CHARACTERISTICS ... Y
FORM 7 D D M M M Y Y
DATE
STREAM ID
COUNTY STATE
U/L STREAM NAME LATITUDE
LONGITUDE
1 250.000 MAP NAME MAP DATE tctVAIIvW
STREAM WIDTH 7S*
Oeoduous Trees D D D D
Comlofous Trees DODO
Shrubs D D D D
Wetland Areas O D D O
Grasses and Forbs D D O O
Moss D D D D
Rocky/Bare Slopes D D D D
STREAM SUBSTRATE O
(Check all thai apply)
<»* n-if. >r>*
Boulders >:scm D D D D
Cobble 6-25 cm Q D D D
Gravel- 02-8 cm D D O O
Sand <02cm D D O D
Silt «nd Clay O O O D
Aufwuchi a O O O
PHOTOGRAPHS COMMENTS:
FRAME ID AZIMUTH
O LAP CARD
_ O •
O — •
FIELD CREW DATA DATA QUALIFIERS
rnpwi tn ft\
«AUPI CH 1
CAUPI f 0 9
•UUP!, ft* .•)
rMfnufp pv
»
/9> .
FORM DISTRIBUTION
White Copy - ORNl
Pink Copy - EMSL-LV
Yellow Copy - FIELD
Revised 1-M
GILL'S (702) 362-2100
Figure 6.2. National Surface Water Survey Form 7 - Watershed Characteristic*.
-------
recent or current rainfall, stream tempera-
ture, and visit number are noted on Form
4 (Stream Data Form, see Figure 6.3).
In addition, on the first visit, photo-
graphs are taken of the areas upstream and
downstream from the sampling site. Before
the set of stream photographs is taken, a
lap card is photographed that shows the
date and stream name, stream and team ID
numbers, and frame number. After the
photographs are taken, the photograph num-
bers are recorded on NSWS Form 7 (Water-
shed Characteristics). The exact sampling
site is marked on a 1:24,000-scale topo-
graphic map, and the site coordinates are
determined and are entered on Form 7.
For Southeast Screening streams,
hydrologic discharge components (width,
flow, and average channel depth) are esti-
mated at each downstream site and are
recorded on NSWS Form 4A (Hydrologic
Data - Figure 6.4). For Phase I streams,
discharge components are measured on the
second visit to each downstream site and
are recorded on Form 4A.
During Episodes Pilot sampling, precipi-
tation, pH, temperature, conductance, and
DO are measured periodically; water samples
are collected and discharge measurements
are taken at the appropriate times. Data
are recorded on NSWS Form 6 (Stream
Episode Data - Figure 6.5).
6.1.2.1 Sample Collection
The team member collecting water
samples sets up the peristaltic pump using
new Tygon tubing. If required, a field blank
sample is collected first. The pump tubing
is rinsed with deionized (DI) water for 2
minutes, and a clean Cubitainer is rinsed
three times with DI water from one of two
Cubitainers of water carried to the sampling
site for this purpose. A blank sample is
collected in the rinsed Cubitainer from the
second Cubitainer of water. Two syringe
Section 6.0
Revision 4
Date: 9/86
Page 4 of 19
samples taken from the blank are used for
the aluminum analyses. Field blank samples
are taken before routine and duplicate
samples. A new piece of tubing is attached
to a sampling boom and is submerged in
midstream. The tubing is rinsed with stream
water for 2 minutes (pump flow rate about
2 L/min). A clean Cubitainer is rinsed three
times with 100 to 200 ml stream water,
then is filled. It is imperative that the
tubing orifice does not come into contact
with the ground or other sources of con-
tamination. Next, four syringe samples are
collected and are sealed with syringe valves.
The syringes used are not acid-rinsed.
Before sample collection, the sample
containers are labeled, and appropriate sample
types are checked on the field sample label
(Figure 6.6). After collection, samples are
placed in coolers at 4 "C for shipment to
the mobile processing laboratory.
When required, duplicate samples are
collected from each sampling region. The
procedure used to collect duplicate samples
is the same procedure used to collect routine
samples.
While one sampler collects the samples
and aliquots and makes the stage and hydro-
logic measurements (if necessary), the other
sampler sets up instruments and begins cali-
bration procedures. The dissolved oxygen
(DO) probe is placed in a watertight, moist
container and is lowered into the stream
to allow temperature equilibration. The
calibration of the pH meter is checked by
measuring a QCCS solution. If the pH meter
does not meet QC limits, it is recal-
ibrated in the field using pH 4 and pH 7
buffers. An open pH measurement is made
as described in the field manual (Hagley
et al., 1986). The conductance meter is
checked using a QCCS, and in situ conduc-
tance and temperature measurements are
made. Next, the calibration procedure for
the DO meter is completed, and an in situ
DO measurement is made.
-------
Section 6.0
Revision 4
Date: 9/86
Page 5 of 19
NATIONAL SURFACE WATER SURVEY
STREAM DATA
FORM 4
STREAM ID
STREAM NAME
SAMPLE DATE
DOM
PROGRAM
D PHASE 1
D SCREENING
D EPISODE PILOT
TIME
START .
FINISH :
U L
PHASE 1 \
M M Y Y
SAMPLES COLLECTED
D ROUTINE
D DUPLICATE
D BLANK
GAUGE HEIGHT (II)
O
o
PH Y N
(FIELD RECALrBRATION'l Q O
OCCS pH 4 00
OCCS INITIAL . O
ROUTINE _
SAMPLE TEMP
DUPLICATE _
SAMPLE TEMP
OCCS FINAL _
O
-c 0
O
O
N . ,
IQIT •
EPISODE SAMPLE TYPE f*\
O BASE FLOW - EPISODE ONLY
D BASE FLOW • EPISODE AND PHASE 1
D RISING STAGE
QPEAK STAGE
D FALLING STAGE
RAIN ^^
CH£f H ONf (IN) >
DNO
D PREV D MOD
D LIGHT D HEAVY
CLOUD COVER
%o
UNCOMPENSATED
CONDUCTIVITY uScm-1
QCCS INITIAL O
OCCS TEMP
IN SITU
STREAM TEMP
OCCS FINAL
OCCS TEMP _
500 uS'cm
Q pH<3 30
n
FIELD CREW DATA
SAUPI fa i
<5AMP| fa 7
5AMPI fa 1
r.Hfr.xtn nv
DATA QUALIFIERS
A INSTRUMENT UNSTABLE
D SLOW STABILIZATION
a DID NOT MEET OCC
»
V
FORM DISTRIBUTION
WHITE COPY - OHNL
PINK COPY - EMSL-LV
YELLOW COPY - FIELD
ORANGE COPY - MOBILE LAB
Revised 1-6-86
GILL'S (702) 362-2100
Figure 6.3. Nations! Suidacc Water Survey Pom 4 - Stream Data.
-------
Section 6.0
Revision 4
Date: 9/86
Page 6 of 19
NATIONAL SURFACE WATER SURVEY
HYDROLOGIC DATA
FORM 4A
SHEET.
OF.
D D M M M
DATE ._
FLOW METER ID
STREAM ID L-
RTRfAM K|AMF
SAMPLE TYPE
D PHASE 1
D SCREENING
D EPISODE PILOT
EPISODE TYPE CHECK ONE
D BASE - EPISODE ONLY
D BASE • EPISODE AND PHASE 1
D RISING
DPEAK
D FALLING
ESTIMATED HYDROLOGY: _ EST MEAS ^x
DEPTH (max -It 1 . \J O D W
TIME START __: W1DTH (melers) Q D D O
TIME END : VELOCITY (m sec -') O D D O
MEASURED HYDROLOGY:
TIME
START :
FINISH :
R.Qht CH/jo rtf W9(Ar Im)
STAGE
— — . _
INTERVAL CENTER |m)
1 ,
2
3
4
5
6
7
8 (mm) .
9
10
11
12
13 ,_
14. .
15. .„
_o
_0
_o
o
_0
_o
o
_o
o
o
_o
o
o
_o
_o
(ft) STEEL ROD STAGE (ft ) WIDTH (m)
._o
. _o
Int^rupl Wi^f
DEPTH AT
CENTER (It)
O
o
o
. o
o
o
0
o
o
o
o
o
o
o
. o
o . o
o o
h (rmf
VELOCITY AT
CENTER (msec'1)
O
o
. o
. o
o
o
o
o
o
. _o
o
o
o
o
o
COMMENTS:
FIELD CREW DATA:
r.acvt in
S&MPI PB 1
CiMOl CO •>
DATA QUALIFIERS
(Si INSTRUMENT UNSTABLE
@ SLOW STABILIZATION
@ DID NOT MEET OCC
ff,
ff\
(2'j -,_
fORM DISTRIBUTION
WHITE COPY — ORNL
PINK COPY - EMSL-LV
YELLOW COPY • FIELD
Revised 1-86
Gtu S 17021 3*7 2100
Figure 6.4. National Surface Water Survey Form 4A - Hydrotogte Data.
-------
Section 6.0
Revision 4
Date: 9/86
Page 7 of 19
NATIONAL SURFACE WATER SURVEY
STREAM EPISODE DATA
FORMS
D D M M M Y
DATE BEGIN
DATE END „_ ,
V
TIME
ARRIVAI
STREAM ID
U.L
I
STREAM NAME
BASE FLOW SAMPLE
RISING SAMPLE
PEAK SAMPLE
FALLING SAMPLE
TIME ii&ewNVsi P«ECIP (in) STAGE III)
O O . O
o . o . o
o o o
o o . o
o o o
o o . o
0 0 _O_
o . o . o
o . o . o
0 0 0_
o o o_
o . o . o
o o o_
o o o_
o o . o
O 0 O-
o o o_
__o o o
o_; o o_
0 0 0-
o o o_
o o . o
__o o o_
__o o o_
pH UNCORRECTED
CONO luScm-i)
o . o
o
o_
o_
o_
o_
0.
o
o
o
o
o
o
o
o
o
o
o o
0 0
. O 0
O-
o.
o
0
o
o
0 0
o o
0 0
0 O
0 0
0 0
o o
o o
o o
TEMP CC)
o
o
o
o
o
o
0
o
o
o
0
o
o
o
o
o
o
o
o
o
_ o
o
__ o
o
DISS 0
(mg l|
o
o
o
o
o
o
o
o
o
o
0
o
o
__o
o
o
o
o
_ o
o
0
o
o
o
COMMENTS:
DATA QUALIFIERS
A'INSTRUMENT UNSTABLE
jySLOW STABILIZATION
S-DID NOT MEET OCC
(1 Base Flow
(2;Bising
(5;Peak
©Falling
IVl
(I*
FIELD CREW DATA
r.ucuii ip
SAUDI en i
SAMPI ea i
$AMP( F^ ^
f-HCrxcn nv
FORM DISTRIBUTION
WHITE — ORNL
PINK — EMSL-LV
Flgura 6.5. National Surfac* Water Survey Form 6 - Stream Episode Data.
-------
Section 6.0
Revision 4
Date: 9/86
Page 8 of 19
STREAM ID U/L
DATE SAMPLED
PROGRAM
D PHASE 1
n SCREENING
D EPISODE PiLOT
CREW
TIME SAMPLED
SAMPLE TYPE
n ROUTINE
D DUPLICATE
D BLANK
EPISODE TYPE
n BASE-EPISODE ONLY
n BASE-EPISODE AND PHASE 1
n RISING
n PEAK
D FALLING
BATCH ID
SAMPLE
ID
Revised 1-86
Figure €.6. Field sample label.
-------
If duplicate samples are collected, a
duplicate set of pH measurements is taken.
The sampler who performs pH measurements
records all data in the field logbook. Dupli-
cate DO and conductance readings are not
made.
After all measurements have been made,
the meters and probes are turned off and
are returned to their carrying cases for
transport to the vehicles. Final measure-
ments of stream-gauge height and time are
recorded, and the team returns to the
vehicle with all gear, samples, and trash.
On arrival at the vehicle, the equip-
ment is loaded for transport, and all data
in the field logbook are transferred to the
Stream Data Form, Form 4.
The sampling team proceeds to the
next stream site where the same activities
are performed.
6.1.3 Field Base Activities After
Sampling Trip
After returning from the field, the
sampling team:
• Checks the calibration of the DO meter
and enters the "final" theoretical mea-
sured DO QC value on Form 4.
• Checks all data forms to ensure accu-
racy and completeness and given them
to the base coordinator, along with
the samples.
• Checks the calibration of the pH and
conductance meters if they did not
meet calibration checks during the day.
• Records all post-sampling calibration
check information on the calibration
form.
• Performs maintenance on or trouble-
shoots problems with meters, if nee-
Section 6.0
Revision 4
Date: 9/86
Page 9 of 19
essary, according to manufacturer's
instructions.
• Stores the conductance and DO probes
in DI water and the pH electrode in
3 M KCI.
6.2 Field Base and Mobile
Processing Laboratory
Activities
The field base and mobile processing
laboratory activities are outlined in Figure
6.7.
The collected, labeled samples are
shipped as soon as possible from the field
base to the mobile processing laboratory
by Federal Express. The supervisor and
analysts at the mobile processing laboratory
are responsible for preliminary measurements
and sample processing. These activities
are described below.
6.2.1 Reagent Preparation
Reagents for total extractable aluminum
extractions and MIBK extractions, and for
specific conductance, DIC, and pH deter-
minations must be prepared before the
samples arrive. Detailed reagent preparation
procedures can be found in the methods
manual (Hillman et al., 1986) and in the
mobile processing laboratory operations
manual (Chaloud et al., in preparation).
6.2.2 Sample Processing
The following steps describe sample
processing operations. They are performed
in the order given.
6.2.2.1 Sample Description and
Identification
Samples are organized into batches
that are processed together. A batch con-
sists of all samples collected and processed
on the same day, from the same survey, that
-------
Section 6.0
Revision 4
Date: 9/88
Page 10 of 19
FIELD SAMPLING TEAMS
MOBILE PROCESSING
LABORATORY
Figure 6.7. Flowchart of dally sampling and mobile processing laboratory activities for the National
Stream Survey.
-------
Section 6.0
Revision 4
Date: 9/88
Page 11 of 19
are sent to one analytical laboratory. It
is expected that there will be approximately
44 samples or fewer in a batch, including
routine, duplicate, blank, and audit samples.
Each batch is assigned a unique batch ID
number, which is recorded on the labels of
all samples (and corresponding aliquots) in
the batch. Each sample is then randomly
assigned a sample ID number as follows:
Routine Samples - Five sample con-
tainers are filled at each stream: a syringe
for DIG determination, a syringe for labo-
ratory pH determination, a syringe for alu-
minum extraction (with MIBK), a syringe
for PCV aluminum determinations, and a
Cubitainer. One sample ID number is
assigned to the five containers and is
recorded on each container label.
Duplicate and Blank Samples - Sample
ID numbers are assigned in the same manner
as for the routine samples.
Field Audit Samples - Aliquots are
prepared from one 2-L field audit sample
(received each day from a central source)
as shown on Table 6.1 and are included in
each day's batch of samples. The label
for the field and laboratory audit containers
is shown in Figure 6.8. The code (Table
6.2) indicates the sample type and the con-
centrate lot number. A field audit sample
is assigned a sample ID number in the same
manner as a routine sample. The ID number
is recorded on the label.
Laboratory Audit Samples - One or
more lab audit sample(s) (received each
day from a central source already prepared
as aliquots) is included in each day's batch.
A single lab audit sample consists of a set
of seven aliquots (eight for the Phase I-
Pilot Survey). Each aliquot has a temporary
label like that in Figure 6.8b. An aliquot
label (Figure 6.8c) is attached to each
aliquot. The lab audit sample is then
assigned a sample ID and batch number in
the same manner as for a routine sample.
The batch and sample ID numbers are recor-
ded on each aliquot label. The date and the
amount of preservative are also recorded
on the label.
After the batch and sample ID numbers
are assigned and are recorded on each sample
label, the same information is entered on
Form 5, Batch/QC Field Data (Figure 6.9).
Also, the stream ID, the appropriate code
for each sample (from Table 6.2), and the
appropriate site or type ID (from Table
6.2) are entered on Form 5.
Note 1: The sample ID numbers are ran-
domly assigned to all samples in a batch.
Furthermore, sample ID numbers run con-
secutively from 1 to the number of samples
in the batch. Audit samples must not always
be assigned the same sample ID number.
Note 2: Field audit samples are processed
exactly like routine stream samples. After
the batch and sample ID numbers are
assigned, the temporary audit label is removed
from the audit sample and is placed in the
audit logbook.
Note 3: Seven different aliquots (numbered
as in Table 6.1) are prepared from each
sample (routine, duplicate, audit, and blank
samples). Each aliquot is assigned the same
batch and sample ID numbers as the sample
from which it is prepared. Aliquot 8 was
prepared for the Phase I - Pilot Survey only.
Note 4: As soon as the 2-L field audit
samples are received from the central
source, four syringes of sample are taken
(for the appropriate analyses) by the mobile
processing laboratory coordinator. Two of
the syringe samples are for aluminum deter-
mination using FIA and MIBK extraction,
the third is for DIG, and the fourth is for
pH determination. Any sample that remains
after DIG determination may be used for
pH determination of aliquots, if necessary.
-------
Section 6.0
Revision 4
Date: 9/88
Page 12 of 19
«
1
S
I
"c
i
I
o
1
1
1
n
a
i
S
•C"
.1
(0
cS
I
«f
u
GO jX
=•
"1
»_•
wIT
&
IT
10 ?
IT
1
3-
MH £
1
IT
?
1
|o
1 4 •§ ffl 2
OL S uj 5 S
£
w
IIiTO
3 ax
i
£ v 6*
iix
*
1
±!
_ *
I X °k
u. a x
4D
|
2 ffl S
•"• *"* X
z -s. &
£
? «
iT a X
1 *^
Pi
o "
H ^
O UJ <
«
S
F
- 1
"Jo o5
o £*n
H- o a.
(0
^ S
oo g ? o
X 22 Q. 0 K
a < CD co O o
Q 3D
J
o*, o" SN
O CO a. 2 w
1
1 S 5
(S *£ 5 iff
1
1
2
in
(0
a>
E
10
(0
5
a.
75
<5 "5
"0 S
re
•E 1
P 1
* 8
"D ^
S. O
2 _sj> 0-
m O "
•° >. 2
j_j Q> J^
Ji*
to , £
<0 " W
1O W ^
0*3 w t t
IN "~
-------
Section 6.0
Revision 4
Date: 9/88
Page 13 of 19
FIELD AUDIT SAMPLE
Radian ID No.
Date
Shipped
Code
Batch
Date
Received
ID
a. Field Audit Sample
Label
LAB AUDIT SAMPLE
Aliquot No.
Date
Shipped
Date
Received
Code
Preservative Amount
b. Lab Audit Sample Label
Aliquot
Batch ID
Sample ID
Date Sampled
Preservative
Amount
Parameters
Note: The aliquot
number, preservative,
and parameters are
preprinted on the
aliquot labels.
c. Aliquot Label
Figure 8.8. Aliquot and Audit Sample Label*.
-------
Section 6.0
Revision 4
Date: 9/88
Page 14 of 19
Table 6.2. Sample Codes
Sample Type
Code
Description
Normal*'*
Audit*
R
D
B
TB
TD
QCCS
F L 1 -001
I I l_
I !
I I
I
I
Routine Stream Sample
Duplicate Stream Sample
Field Blank Sample
Mobile Processing Laboratory (Trailer) Blank
Mobile Processing Laboratory (Trailer) Duplicate
Quality Control Check Sample
Radian I.D. Number
Concentrate lot number
Concentration Level
L = low, N = Natural
Type of Audit Sample (F = Field, L = Lab)
Episodic0
EB
ER
EP
EF
M1
M2
S
Episodic sample, base hydrograph
Episodic sample, rising hydrograph
Episodic sample, peak hydrograph
Episodic sample, falling hydrograph
Initial Mid-Atlantic Phase I Sample
Final Mid-Atlantic Phase I Sample
Southeast Screening Sample
e Normal samples require a stream ID, except trailer blank.
* Recorded in sample code column on Batch/QC Field Data Form.
" Recorded in site or type column on Batch/QC Field Data Form.
-------
Section 6.0
Revision 4
Date: 9/88
Page 15 of 19
NATIONAL SURFACE WATER SURVEY
BATCH/QC FIELD DATA FORM
RE-ENTERED
D FORM 2 LAKES
OR
O FORM 5 STREAMS
DATA ounmtn» 1.1 *^ i AMI MMLASU rcm UH ON TMI »
OUM.WMM COMtWHT
«.9.
National Surface Water Survy Form I - Batch/QC FMd Data (Nattena. Stream Survey.
-------
Section 6.0
Revision 4
Date: 9/88
Page 16 of 19
All must be analyzed within 48 hours. Batch
and sample ID numbers are assigned at the
mobile processing laboratory and are record-
ed on the shipping form (Form 3, Figure
6.10).
Note 5: Copies of ail field, laboratory,
and streamside data forms are sent to the
QA staff in Las Vegas, Nevada, daily.
6.2.2.2 DIG Determination
Immediately after assignment of batch
and sample ID numbers, one analyst begins
the DIC analyses. DIC is determined in
routine, duplicate, and audit samples. The
routine and duplicate samples are contained
in sealed syringes (filled at the stream site).
The results of the DIC determination are
recorded on Form 5. The QC procedures
are discussed in Section 7.
6.2.2.3 pH Determination (mobile
processing laboratory)
After DIC determinations, the pH of
the remaining sealed syringe sample is deter-
mined.
The QC procedures are discussed in
Section 7. The results are recorded on Form
5. Copies of all raw data are sent to the
QA staff in Las Vegas, Nevada, at survey
completion or when requested.
Note: Two pH measurements are made:
one at the stream site in an open beaker
and one at the mobile processing labora-
tory in a sample chamber.
6.2.2.4 Sample Filtration, Preservation,
and Aliquot Preparation
Eight aliquots from each Phase - I
Pilot Survey sample and seven aliquots from
each sample taken for other NSS surveys
are prepared as specified in Table 6.1.
Preparation of aliquots is described in the
methods manual (Hillman et al., 1986).
6.2.2.5 True Color Determination
After centrifugation to remove turbid-
ity, color is determined using Hach Model
CO-1 Color Test Kit following the manufac-
turer's instructions. Results are recorded
on Form 5. The QC procedures are dis-
cussed in Section 7.
6.2.2.6 Turbidity
A Monitek Model 21 laboratory neph-
elometer is used to determine the turbidity
of routine, duplicate, audit, and blank sam-
ples. Results are recorded on Form 5.
The QC procedures are discussed in Section
7.
6.2.2.7 Specific Conductance
A YSI Model 32 meter with YSI 3400
series probe is used to determine the spe-
cific conductance of routine, duplicate, audit,
and blank samples. The QC procedures are
discussed in Section 7.
6.2.2.8 Sample Shipment
When a batch is completely processed
and is ready for shipment, the samples are
assembled into groups according to the
analytical laboratory to which they are being
shipped.
6.2.2.9 Data Distribution
Copies of all forms (except labels and
Form 3) are kept at the mobile processing
laboratory. Copies of Forms 3, 4, 4A, 5, 6,
and 7 are sent to the locations indicated in
Figure 6.11. One copy of Form 3 is sent
to the Sample Management Office (SMO)
and two copies are sent with the samples
to the analytical laboratory. Upon receipt
-------
Section 6.0
Revision 4
Date: 9/88
Page 17 of 19
NATIONAL SURFACE WATER SURVEY
SAMPLE MANAGEMENT OFFICE
P.O. BOX 818
ALEXANDRIA, VA 22314
NSWS
FORMS
SHIPPING
RECEIVED BY
IF INCOMPLETE IMMEDIATELY NOTIFY
SAMPLE MANAGEMENT OFFICE
(703) 557-2490
FROM
(STATION ID)
SAMPLE
ID
01
02
03
04
05
06
07
OB
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
1
TO
ILAB)
BATCH
ID
DATE PROCESSED
AUOUOTS SHIPPED
(FOR STATION USE ONLY)
2
3
4
5
6
7
8
DATE SHIPPED DATE RECEIVED
AIR-BILL NO
SPLITS
SAMPLE CONDITION UPON LAB RECEIPT
(FOR LAB USE ONLY)
QUALIFIERS
V
M
ALIQUOT SHIPPED
ALIQUOT MISSING DUE TO DESTROYED SAMPLE
WMITt- FIELD COPY
lt I FOB Ml 1MB
PINK - LAB COPY
YELLOW- SMOCOPY
GOLD- LAB COPY FOft RETURN TO SMO
Figure 6.10. National Surface Water Survey Form 3 - Shipping.
-------
Section 6.0
Revision 4
Date: 9/88
Page 18 of 19
FIELD BASE/
MOBILE
PROCESSING
LABORATORY
FORM 3 (2 Copies)
ANALYTICAL
LABORATORY
FORM 3
(1 Copy)
SAMPLE
MANAGEMENT
OFFICE
FORM 3
(ICopy)
FORMS 4, 4A,
5, 6, and 7
(2 Copies each)
and
FORM 3
(Original)
FORMS 4, 4A, 5, 6, and 7
(1 Copy each)
Figure 6.11. Flow scheme for field data forme.
-------
by the analytical laboratory, the condition
of the samples is noted on the two forms
and one is forwarded immediately to SMO.
After the laboratory analyses are
completed, the results are entered on the
data reporting forms (see Appendix A) and
are sent to the data base manager for entry
into the data base. In either case, a copy
of the results is sent to the QA staff for
data evaluation. (Data evaluation is dis-
cussed in Section 13.) QA for the data in
the data base is discussed in Section
12.
6.3 Training
Prior to the NSS field activities, all
personnel must be trained. Safety proce-
dures and regulations, field operations, and
mobile processing laboratory analytical
procedures are the chief subjects of instruc-
tion. Training includes classroom instruction
and realistic simulations of actual activities
to be performed. Where possible, these
simulations include sampling at a stream
site and sample preparations in a mobile
processing laboratory.
Section 6.0
Revision 4
Date: 9/88
Page 19 of 19
-------
Section 7.0
Revision 4
Date: 9/86
Page 1 of 5
7.0 FIELD MEASUREMENT QUALITY CONTROL CHECKS
Every sampling day, the sampling teams
check the calibration of the meters used in
the field. The check is performed at the
field base prior to departure for sampling
and again at the stream sites before and
after stream measurements are taken. The
mobile processing laboratory personnel
conduct QC checks of their instruments
after the samples are received at the
mobile processing laboratory. The QC
measurements are described in Sections 7.1
and 7.2.
7.1 Stream Site Measurements
Stream site measurements consist of
six determinations. Five of these, stream
temperature, pH, specific conductance, DO,
and staff gauge height, are recorded on
NSWS Stream Data Form 4. Flow velocity
is recorded on the Hydrology Form (4A).
7.1.1 Site Measurements of
Chemical Parameters
Portable pH, specific conductance, DO,
and flow velocity meters are utilized in
stream site measurements.
The QC procedures consist of calibra-
tion or calibration checks of the instruments
before and after each sampling trip and
of determining any change in instrument
response between calibrations. These proce-
dures are described in detail in Hagley et
al. (1986). The following is a summary of
the stream QC procedures.
Water temperature - Sample tempera-
ture readings are required in order to calcu-
late temperature-corrected values for pH,
specific conductance, and DO. Each meter
has its own temperature function, which is
checked every morning against an NBS-
traceable thermometer. The readings must
agree within 0.5 "C. There is no calibration
control for temperature, so the probe must
be replaced if manufacturer's troubleshooting
instructions do not resolve a discrepancy
in the readings. All data taken with the
defective probe must be qualified. The
stream temperature obtained using the con-
ductance meter is recorded on Form 4 as
the in situ temperature, unless the meter
temperature function was not within the
acceptable criteria. In that case, the DO
stream temperature is recorded, and a nota-
tion is made on Form 4.
pH - At each stream site, a QCCS that
has a theoretical pH value of 4.00 must
be analyzed prior to and following the stream
pH determinations. If any QCCS reading
deviates from the theoretical pH by more
than 0.1 pH unit, the instrument is recali-
brated and the pH of the QCCS is measured
again. If the reading still does not meet
the specifications and no functioning
back-up electrode or meter is available,
the appropriate data qualifier (listed in
Table 9.8) is recorded on the Stream Data
Form 4.
Spec/fie Conductance - The meter in
use has no conductance calibration controls.
Therefore, the operator must determine if
the manufacturer-set conductance calibration
is within specifications by measuring QCCSs
of 718, 147, and 74 //S/cm during the daily
presampling calibration check. The allowed
error on the QCCS is ±72 juS/cm, ±15 juS/cm,
and ±10 /^S/cm, respectively. If the reading
is not within these limits, the manufacturer's
troubleshooting guide should be consulted,
and the meter or probe should be replaced
if necessary. Before and after the in situ
specific conductance determinations, a QCCS
of 74 pS/cm is analyzed. The measured
QCCS must be within 10 juS/cm, or the data
must be qualified. The QCCS data are
recorded on Stream Data Form 4.
Dissolved O2 - No QCCS is analyzed
at the stream site because the meter is
-------
recalibrated at each base site. The QC
check is as follows: After the dissolved
oxygen meter is calibrated with water-
saturated air, air-saturated water is
measured. The readings must be within
0.5 mg/L of one another, or the appropriate
data qualifier must be recorded on Stream
Data Form 4. The QC check is performed
at the field base before and after each day's
sampling.
Stream Flow - The manufacturer-set
calibration of the flow meter is checked
daily by the sampling team before the team
leaves the field base for sampling and again
at streamside before entering the stream.
The meter reading should be 10.0 ± 0.2
ft/sec. Once a week, the zero value on
the meter is checked and adjusted; the
reading should be 0.0 ± 0.1 ft/sec.
There are no QC checks for staff gauge
measurements. 6.
7.2 Mobile Processing Laboratory 7.
Measurements
Measurements made at the mobile
processing laboratory include DIC, pH,
turbidity, specific conductance, PCV alumi-
num, and true color. The data are recorded
on Form 5. The QC procedures are des-
cribed in detail in Hillman et at. (1986)
and Chaloud et al. (in preparation). This
section contains a summary of the QC
procedures.
7.2.1 Dissolved Inorganic Carbon
DIC is measured in routine, duplicate,
and field audit samples using the Dohrman
Model DC-80 carbon analyzer. The measure-
ment procedure is as follows:
1. Initial calibration is performed using
the working standard (10.00 mg/L C).
2. Two QC standards (2.00 mg/L C and
Section 7.0
Revision 4
Date: 9/86
Page 2 of 5
20.00 mg/L C) are measured to verify
the initial calibration.
3. If both QC standards are within 10
percent of the theoretical concentra-
tion, the values are entered in the
DIC logbook and analysis proceeds.
If the standards are not within 10
percent, steps 1 and 2 are repeated.
4. A calibration blank is measured.
5. If the calibration blank is less than
0.1 mg/L C, the value is recorded and
sample analysts continues. If the
calibration blank is 0.1 mg/L C or
greater, the laboratory supervisor is
informed, corrective action is taken,
and steps 1 through 5 are repeated.
Normally, one calibration blank is
analyzed at the beginning of the batch.
8.
DIC is measured for eight samples.
A 2.0 mg/L C QCCS is analyzed to
check the calibration.
If the QCCS is within 10 percent of
the theoretical concentration, the value
is recorded in the logbook and sample
analysis continues. If the QCCS is
not within 10 percent, it should be
determined whether there is enough
left of the samples associated with
the unacceptable QCCS to reanalyze
them. If enough sample is left, steps
1 through 7 are repeated, including
analysis of all samples since the last
acceptable QCCS. If not enough sample
remains, the unacceptable QCCS value
is recorded in the DIC logbook, the
sample ID numbers associated with
the unacceptable QCCS are noted, and
the appropriate data qualifier is entered
on Form 5 for the affected samples.
Sample analysis must not continue
until acceptable QCCS values are
obtained.
-------
9. One sample is measured in duplicate
per batch. These duplicates are called
trailer duplicates. If the difference
between the two measurements is
greater than 10 percent, another sample
is analyzed in duplicate. If the differ-
ence is still greater than 10 percent,
the laboratory supervisor is notified,
and the problem is noted on Form 5
with a data qualifier.
10. When sample analysis is complete, a
final QC check is required, and the
relevant QC information is recorded
on Form 5, Batch QC Field Data.
7.2.2 pH
pH is determined in routine, duplicate,
and field audit samples using an Orion Model
611 pH meter with Orion Ross Model 8104
glass body combination pH electrode. The
measurement procedure is as follows:
1. The instrument is standardized accord-
ing to the manufacturer's instructions
and the methods manual (Hillman et
al., 1986).
2. The pH of pH 4 and pH 7 buffers is
measured and the results are recorded
in the logbook. If either measurement
differs from the certified value by
more than 0.02 pH units, steps 1 and
2 are repeated. If acceptable results
cannot be obtained, the electrode is
replaced and the above procedure is
repeated. (Failed electrodes should
be sent with a description of the
problems observed to the QA manager
at EMSL-LV where they will be tested
further.)
3. When satisfactory results are obtained
for the buffers, the pH of a pH 4.00
QC sample is measured and the result
is recorded in the logbook. If the
reading differs from 4.00 by more than
8.
Section 7.0
Revision 4
Date: 9/86
Page 3 of 5
0.1 pH unit, steps 1 and 2 are repeated,
and the pH of a fresh QCCS is meas-
ured. If acceptable results are still
not obtained, the laboratory manager
should be consulted. Stream samples
are not to be analyzed until an accept-
able value for the QCCS has been
obtained.
Samples are measured for pH. After
every five samples, a pH 4.00 QC
sample is measured and the result is
recorded in the logbook. If the meas-
ured pH of the QC sample is 4.0 ±
0.1 pH units, measurement of samples
proceeds.
If the QCCS is not acceptable, it should
be determined whether there is a
sufficient amount of sample remaining
in any of the other three syringes to
repeat the analysis. If so, steps 1
through 3 are repeated and all samples
analyzed since the last acceptable QCCS
are reanalyzed. If not enough sample
remains, the sample ID numbers asso-
ciated with the unacceptable QCCS are
recorded in the logbook.
One sample per batch is measured in
duplicate. If the difference between
the two measurements is greater than
0.1 pH unit, another sample is measured
in duplicate. If the difference is still
greater than 0.1 pH unit, the laboratory
supervisor is notified, and the problem
is noted on Form 5 with a data quali-
fier.
After the last sample in a batch has
been analyzed, a final QCCS is analyzed
and the value is recorded in the log-
book.
When this analysis is completed, the
relevant QC information is recorded
on Form 5.
-------
7.2.3 Turbidity
Turbidity is determined in routine,
duplicate, field audit, trailer duplicate, and
blank samples using the Monitek Model 21
laboratory nephelometer. The measurement
procedure is as follows:
1. The nephelometer, set on Range 20,
is zeroed and then is calibrated with
a 10.0 NTU standard, following the
manufacturer's recommendations.
2. Calibration linearity is verified by
analyzing 2.0, 5.0, and 20.0 NTU QC
samples. (The 20.0 NTU QC sample
is measured on Range 200.) The meas-
ured values must be 2.0 ± 0.2, 5.0 ±
0.5, and 20.0 ± 2.0. If the measured
values are unacceptable, step 1 is
repeated. Acceptable results must
be obtained prior to sample analy-
sis. Acceptable results for the 5.0
NTU QC sample are recorded on
Form 5.
3. For every eight samples, a 5.0 NTU
QC sample is measured. If the meas-
ured value is 5.0 ± 0.5 NTU, QC and
sample results are recorded on Form
5.
4. If the QC measurement is unacceptable,
the instrument must be recalibrated and
the previous eight samples must be
reanalyzed. Acceptable QC values are
recorded on Form 5 along with associ-
ated sample results.
Note: Some samples must be analyzed on
range 2 or on range 200. If the range 200
setting is used, the instrument must be
recalibrated and a different QCCS must be
analyzed.
7.2.4 True Color
The only QC check on true color is
that one sample per batch is measured in
Section 7.0
Revision 4
Date: 9/86
Page 4 of 5
duplicate. If the two measurements differ
by more than 10 units, another sample is
measured in duplicate. If acceptable results
are still not obtained, the laboratory super-
visor must be notified and a data quali-
fier must be recorded on Form 5 with the
results. Acceptable results are also recorded
on Form 5.
7.2.5 Nonexchangeable and Total
PCV-Reactive Aluminum
Nonexchangeable and total PCV-reactive
aluminum is determined in routine, duplicate,
audit, blank, and trailer duplicate samples
using a Lachat FIA. The measurement pro-
cedure is as follows:
1. Three calibration ranges may be defined.
For the low range, 0, 10, 25, 50, 65,
100, and 125 ppb calibration standards
are used; for the high range, 750, 1250,
1750, 2250, and 3000 ppb calibration
standards are used; and for samples
with very high levels of analyte, a
calibration range is defined by standards
with concentrations of 1000, 2000,
3000, and 5000 ppb.
2. After every 10 samples, a 75 ppb QCCS
is analyzed. The measured value of the
QCCS must be 75 ppb ±10 percent for
channel 1, and 75 ppb ±20 percent
for channel 2.
3. If the QCCS measurement is unac-
ceptable, the instrument must be recali-
brated and the previous 10 samples
must be reanalyzed. Acceptable QC
values are recorded on Form 5.
7.2.6 Specific Conductance
Specific conductance is measured for
routine, duplicate, audit, blank, and trailer
duplicate samples using a YSI conductance
meter and probe. The measurement procedure
is as follows:
-------
1. A calibration blank and a 147 pS/cm
calibration standard are measured and
the cell constant is calculated.
2. After every 10 samples, QCCS of 14.7,
72.8 and 147 /jS/cm are measured.
The QCCSs are prepared from a differ-
ent stock solution than are the calibra-
tion standards.
3. If the QCCS measurements are unac-
ceptable, the instrument must be recali-
brated and the previous 10 samples
must be reanalyzed. Acceptable values
for the 147 juS/cm QCCS are recorded
on Form 5.
4. A final 147 /uS/cm calibration standard
is measured and the cell constant is
recalculated.
5. The temperature is recorded in the
logbook. Values on Form 5 are cor-
rected for the cell constant but not for
temperature (25 *C). The temperature
log is photocopied and the copy is
attached to Form 5 so that temperature
corrections can be performed by EMSL-
LV QA staff.
Section 7.0
Revision 4
Date: 9/86
Page 5 of 5
-------
3.0 ANALYTICAL PROCEDURES
Section 8.0
Revision 4
Date: 9/86
Page 1 of 1
Table 8.1 lists the analytical procedures
that are used to determine each required
parameter. A detailed description of these
procedures is provided in the methods man-
ual (Hillman et al., 1986). Internal QC checks
on the analytical procedures are discussed
in the next section.
Table 8.1. Parameter* and Corresponding Measurement Methods
Parameter
Method*
1. ANC
2. BNC
3. Aluminum, total
4. Aluminum, total extractable
5. Aluminum, Nonexchangeabie and total PCV
reactive*
6. Ammonium, dissolved
7. Calcium, dissolved
8. Chloride, dissolved
9. Fluoride, total dissolved
10. Inorganic carbon, dissolved
11. Iron, dissolved
12. Magnesium, dissolved
13. Manganese, dissolved
14. Nitrate, dissolved
15. Organic carbon, dissolved
16. pH
17. Phosphorus, totat dissolved**
18. Potassium, dissolved
19. Silica, dissolved
20. Sodium, dissolved
21. Sulfate, dissolved
22. Specific conductance
Titration with Gran plot
Titration with Gran plot
EPA Method 202.2 AAS (furnace)
Extraction with 8-hydroxyquinoline into MIBK
followed by AAS (furnace)
Automated colorimetric pyrocatechol violet (PCV)C
EPA Method 350.1
EPA Method 215.1 - AAS (flame)
Ion chromatography
Ion selective electrode
Instrumental (Similar to DOC)
EPA Method 236.1 - AAS (furnace)
EPA Method 242.1 - AAS (flame)
EPA Method 243.1 - AAS (flame)
Ion chromatography
EPA Method 415.2
pH electrode and meter
USGS Method 1-4600-78 or Modified USGS Method
EPA Method 258.1 - AAS (flame)
USGS Method 1-2700-78
EPA Method 273.1 - AAS (flame)
Ion chromatography
EPA Method 120.1
a AAS methods are taken from U.S. EPA, 1983. Laboratories that have ICP instrumentation may use EPA
Method 200.7, reproduced in Appendix A of Hillman et al. (1986), for determining Ca, Fe, Mg, and Mn,
providing they can demonstrate the detection limits specified in Table 4.1. If the ICP instrumentation
cannot meet the required detection limits, it may still be used to analyze samples which contain the
analytes at concentrations greater than 10 times the ICP detection limit. Other samples must be
analyzed by furnace or flame AAS.
b Determined in the mobile processing laboratory.
c Nonexchangeabie and total PCV reactive aluminum extraction was not performed for the Phase I - Pilot
Survey but is performed for the other three surveys.
d For the Pilot Survey, total P was determined (samples were unfiltered). For the Mid-Atlantic - Phase I,
Southeast Screening, and Episodes Pilot Surveys, total dissolved P was determined (samples were filtered).
-------
Section 9.0
Revision 4
Date: 9/86
Page 1 of 13
9.0 ANALYTICAL INTERNAL QUALITY CONTROL
9.1 Sample Receipt
MI samples received by the analytical
laboratory should be checked in by a receiv-
ing clerk who (1) records the date received
on the shipping form, (2) checks the samples
to identify discrepancies with the shipping
form, (3) fills out the "sample condition"
portion of the shipping form, and (4) mails
a copy of the completed shipping form to
SMO. The "sample condition" column should
note such information as leakage in shipping,
insufficient sample, noticeable suspended
particulates, partially frozen samples, and
the temperature of the sample containers.
If there are any discrepancies, the field
base coordinator must be notified immedi-
ately. These data are kept on a computer
file by SMO and are available to interested
parties. The laboratory retains a copy of
the completed shipping form for the labo-
ratory records. The samples are refrigerated
as soon as possible.
Samples are received already preserved
and ready for analysis. Sample aliquots 2,
3, 4, 5, 6, and 8 are stored at 4 "C in the
dark while not in use. When an analysis is
to be performed, the analyst should remove
an aliquot from the sample and should return
the sample to the refrigerator as soon as
possible.
Even after all analyses have been
completed and the results have been
checked, samples remain stored in a refri-
gerator at 4*C for 6 to 12 months, or until
laboratory personnel are notified otherwise
by the QA manager, in case reanalysis is
necessary.
9.2 Sample Analysis
Procedures given in the methods manual
(Hiliman et a!., 1986) are to be followed
exactly. Table 8.1 is a list of all required
measurements and the associated methods.
Table 4.1 lists the required precision,
expected ranges, and detection limits for
each parameter. All analyses for each
parameter must be performed within the
specified holding times given in Table 9.1.
9.3 Analytical Laboratory
Documentation for Quality
Control
The following documents and information
must be updated constantly and must be
available to the analyst and to the supervisor
involved in the project:
« laboratory standard operational proce-
dures (SOPs) - detailed instructions
about the laboratory and instrument
operations.
« laboratory QA plan - clearly defined
laboratory protocol, including personnel
responsibilities and use of QC samples.
* list of in-house samples - including
projected dates for completion of
analyses; allows analyst to schedule
further analyses.
* instrument performance study informa-
tion - information on baseline noise,
calibration standard response, precision
as a function of concentration, and
detection limits; used by analyst and
supervisor to evaluate daily instrument
performance.
-------
Table 9.1. Maximum Holding Times
able and
organic
equilibrated
Section 9.0
Revision 4
Date: 9/86
Page 2 of 13
Holding
Time
Parameter
7 days 14 days
NO3'* ANC
BNC
Total extract- DIG, initial and
28 days
Total P
NH4+
SO/"
28 days*
Ca
Mg
Na
extractable
Al
DOC
PH"
Specific
Conductance
F"
Si02
cr
Total AI
Mn
Fe
K
a Although the EPA (U.S. EPA, 1983) recommends a 6-month holding time for these metals, this study
requires that all of the metals be determined within 23 days. This requirement ensures that
significant changes do not occur and that data are obtained in a timely manner.
b Although the EPA (U.S. EPA, 1983) recommends that nitrate in unpreserved samples (un-acidified)
be determined within 48 hours of collection, evidence exists (Peden, 1981 and APHA et al., 1985)
that nitrate is stable for 2 to 4 weeks if stored in the dark at 4 *C.
0 Although the EPA (U.S. EPA, 1983) recommends that pH be measured immediately after sample collection,
evidence exists (McQuaker et al., 1983) that it is stable for as long as 15 days if stored at 4 "C
and sealed from the atmosphere. The pH is also measured in a sealed sample at the mobile processing
laboratory within 48 hours of sample collection.
QC charts - the most recent QC charts
with 99 percent and 95 percent control
limits for all QCCS and detection limit
QCCS, generated and updated for each
batch. The same QCCS must be used
for ail QC charts to ensure the conti-
nuity of the charts. (Not©: The pur-
pose of preparing QCCS charts is to
ensure that the actual control limits
do not exceed the limits given in
Table 9.2.)
data sheet QC report - report by the
laboratory manager reviewing QC
results for each parameter and flagging
al! results outside statistically
established QC limits for reanalysis
before data are submitted to recipients.
9.4 Internal Quality Control
Within Each
Internal QC must be an integral part
of any measurement procedure to ensure that
results are reliable. Internal QC procedures
each method are summarized in Table 9.3.
These QC procedures are performed for
every sample batch, unless otherwise noted.
QC procedures for certain measure-merits
(pH, BNC, ANC, and specific conductance)
are detailed in the appropriate method
description in Hillman at al. (1986). Details
-------
Section 9.0
Revision 4
Date: 9/86
Page 3 of 13
Table 9.2. Maximum Control Umlts for Quality Control Check Samples
Parameter
Maximum Control Limit for QCCS (% Deviation from
Theoretical Concentration of QCCS)
Al, Nonexchangeable and total
PCV reactive
Al, total extractable
Al, total
Ca
cr
DIG
DOC
F", total dissolved
Fe
K
Mg
Mn
Na
NH4+
P, total dissolved
pH
SiO,
Specific Conductance
±20%
±20%
±20%
±5%
±5%
±10%
±10%
±5%
±10%
±5%
±5%
±10%
±5%
±10%
±10%
±20%
±0.05 pH unit
±5%
±5%
+2%
on internal QC procedures for automated
colorimetric analyses (total or total dissolved
P, NH4+ and SiO2), instrumental carbon
analyses (DIG and DOC), ion-selective elec-
trode analysis (F~), ion chromatography
analyses (NO3', Cl", and SO42'), and atomic
absorption or emission analyses (Ca, Mg,
K, Na, Mn, Fe, total Al, and total extract-
able Al) are described below.
1. Initial Calibration - An initial calibra-
tion is performed as required for each
analytical method. Next, the linear
dynamic range (LDR) is determined
for the initial calibration. The concen-
trations of the calibration standards
must bracket the expected sample
concentrations. (Occasionally the
standards suggested by a method must
be adjusted to meet this requirement.)
The low standard should not be greater
than 10 times the detection limit. If
during the analysis the concentration
of a sample is above the LDR, two
options are available. One option is
to dilute and reanalyze the sample. In
this case, the diluent should have for
a matrix similar to the sample matrix
with respect to all preservatives (acid
type and concentration) used. Alter-
natively, two concentration ranges
may be calibrated. Samples are first
analyzed on the lower concentration
range. Each sample whose concentra-
tion exceeds the upper end of the
LDR is then reanalyzed on the higher
concentration range. If this option is
taken, separate QC samples must be
analyzed and reported for each range.
-------
Table 9.3. Summary of Internal Quaiity Contro! Checks for Analysis Methods
Parameter or Method
QC Check
Control Limits
Corrective Action*
ANC, BNC, pH
1. Titrant standardization
crosscheck
1 Relative differences <5%
1. Restandardize titrants.
2. Electrode calibration
(Nemstian response check)
3. pH QCCS (pH 4 and 10)
analysis
4. Blank analysis (salt
spike)
5. Duplicate analysis
6. Protolyte comparison
2. Slope - 1.00 i 0.05
3. pH 4 = 4.00 i 0.05
pH 10 = 10.00 ± 0.05
4. | Blank | <10 //eq/L
5. RSD <10% (ANC and BNC)
+. O.05 pH units (pH)
6. See Section 13.2.4
2. Recalibrate or replace
electrode.
3. Recalibrate electrode.
4. Prepare fresh KCI spike
solution.
5. Refine analytical technique,
analyze another duplicate.
6. See Section 13.2.4
Ions (CI", total
dissolved F,
NH4+, NO3- ,SO42-)
Metals (total At,
total extractable
Al, Ca, Fe, K, Mg,
Mn, Na)
Si02,total or total
dissolved P, DIG,
DOC, spec. cond.
1a. Initial QCCS analysis
(calibration and
verification)
1b. Continuing QCCS analysis
(every 10 samples)
2a. Detection limit deter-
mination (weekly)
2b. DL QCCS analysis (daily;
for the parameters desig-
nated in Section 9.4
1a,b. The lesser of the 99% CI
or value given in Table 9.2
2a, Detection limits given in
Table 4.1
2b. % Recovery = 100 +. 20%
1a. Prepare new standards
and recalibrate.
1b. Recalibrate. Reanalyze
associated samples.
2a,b. Optimize instrumentation
and technique.
(Continued)
fi> fii
«>
a To be followed when QC check is outside control limits.
-------
Table 9.3. (Continued)
Parameter or Method QC Check
Ions (CI", total 3. Blank analysis
dissolved F,
NH4+, NO3', SOf)
Metals (total Al,
Control Limits
3a. Blank <2 x DL (except
sp. cond.)
3b. Blank <0.9 pS/cm
(sp. cond. only)
Corrective Action*
3a,b. Determine and eliminate
contamination source.
Prepare fresh blank
solution. Reanalyze
associated samples.
total extractable
Al, Ca, Fa, K, Mg,
Mn, Na)
iOg, total or total
dissolved P, DIC,
OOC, spec. cond.
4. Duplicate analysis
5. Matrix spike* (except ext.
Al, DIC, and spec, cond.)
4. Duplicate precision (%RSD)
limits given in Table 4.1
5. % Recovery = 100 ± 15%
4. Investigate and eliminate
source of imprecision.
Analyze another duplicate.
5. Analyze 2 additional spikes
If one or both outside
control limits, analyze
all samples in that batch by
method of standard additions.
6. Resolution test (1C only)
6. Resolution >60%
6. Clean or replace separator
column. Recalibrate.
a To be followed when QC check is outside control limits.
b Matrix spikes were performed for the Phase I - Pilot Survey only.
D) 0)
CO
u
-------
Section 9.0
Revision 4
Date: 9/86
Page 6 of 13
2. QCCS - Immediately after the instru-
ments are standardized, a QCCS con-
taining the analyte of interest at a
concentration in the mid-calibration
range must be analyzed. If a wider
range than necessary is calibrated (e.g.,
for analysis by inductively coupled
plasma emission spectroscopy), the QC
sample must be in the same concentra-
tion range as the samples. QCCS may
be obtained commercially or may be
prepared by the analyst from a source
which is independent from the calibra-
tion standards (i.e., the QCCS cannot
be made by diluting the same stock
solution used to make the calibration
standards). The calibration QC sample
must be analyzed to verify the calibra-
tion curve prior to any sample analysis,
after every 10 samples, and after the
last sample. The observed value for
the QC sample must not differ from
the theoretical value by more than the
limits given in Table 9.2. When an
unacceptable value for the calibration
QC sample is obtained, the instrument
must be recalibrated and ail samples
that were analyzed after the last
acceptable QC sample must be reana-
lyzed. Furthermore, the observed con-
centrations for the QC sample must
be plotted on a QC chart and 99 per-
cent and 95 percent confidence inter-
vals must be developed. To ensure
the continuity of QC charts, a QCCS
sample of the same theoretical con-
centration must be used throughout
the plotting process. The 99 percent
control limit must not differ from the
theoretical value by more than the
limits given in Table 9.2. If it does,
the QA manager must be consulted.
Weekly, QC charts should be updated,
cumulative means should be calculated,
and new warning and control limits
(95 percent and 39 percent, respec-
tively) should be determined. To indi-
cate bias for a given analysis, there
must be at least seven successive points
on one side of the theoretical mean.
If bias is indicated, analysis must be
stopped and an explanation must be
sought.
3. Detect/on Limit QCCS - This is a low-
level QC sample that contains the ana-
lyte of interest at a concentration
two to three times the required detec-
tion limit. This QC sample must be
analyzed once per batch for the follow-
ing parameters: total extractable Al,
total Al, dissolved metals (Ca, Fe, K,
Mg, Mn, Na), anions (Cl", SO42',
NO3-), NH4+, SiO2, DOC, air-equilibrated
DIC, initial DIG, and total or total
dissolved P. The results are reported
on Form 20, Blanks and QCCS Results
(see Table 9.4.). The purpose of the
detection limit QC sample is to elimi-
nate the necessity of formally deter-
mining the detection limit on a daily
basis. The measured value must be
within 20 percent of the theoretical
concentration. If it is not, the problem
must be identified and corrected, and
an acceptable result must be obtained
prior to sample analysis.
4. Calibration Blank - A calibration blank
must be analyzed once per batch.imme-
diately after the initial calibration,
to check for base-line drift and
low-level calibration-curve bias (y-
intercept). Rezero if necessary. The
calibration blank is defined as a "0"
mg/L standard and contains only the
matrix of the calibration standards.
The observed concentration of the
calibration blank must be less than or
equal to twice the required detection
limit. If it is not, the instrument
must be rezeroed and the calibration
must be rechecked.
-------
Section 9.0
Revision 4
Date: 9/86
Page 7 of 13
Table 9.4. Data Forms Used by the Analytical
Laboratory4
Data
Form
Description
11 Summary of Sample Results
13 ANC and BNC Analyses Results
14* QC Data for ANC and BNC Analyses
15* Specific Conductance (Measured and
Calculated)
16* Anion-Cation Balance Calculations
17 Ion Chromatography Resolution Test
18 Detection Limits
19 Sample Holding Time Summary
20 Blanks and QCCS Results
21* Dilution Factors
22 Duplicates Results
a These forms are shown in Appendix A.
* Form not required to be submitted with
data package but recommended for internal
QC requirements.
5. Reagent Blank - A reagent blank must
be prepared and analyzed for each
batch of samples for methods which
require sample preparation (dissolved
SiO2 and total Al). A reagent blank
is defined as a sample composed of
all the reagents (in the same quantities)
used in preparing a real sample for
analysis. It is also carried through
the same digestion and extraction
procedure as a real sample. The con-
centration of the reagent blank must
be less than or equal to twice the
required detection limit. If the con-
centration exceeds this limit, the source
of contamination must be investigated
and eliminated. A new reagent blank
must be prepared and analyzed for
each sample in which the high reagent-
blank value contributed significantly
(>10 percent) to the value of the
parameter in question. If a high reagent
blank problem cannot be corrected,
the QA manager must be contacted.
Reagent blank results are reported on
Form 20 but are not subtracted from
sample results.
Preliminary Sample Analysis - Approx-
imately seven samples and a reagent
blank must be analyzed prior to dupli-
cate analyses to determine approximate
endogenous sample concentrations.
Duplicate Sample Analysis - One sample
per batch must be prepared and analyzed
in duplicate for each parameter. The
relative standard deviation is plotted
on a QC chart and 99 percent and 95
percent confidence intervals are estab-
lished. Initial control limits are set
at the precision levels given in Table
4.1. The control limits should not
exceed these values. If they do, the
QA manager must be notified imme-
diately. If duplicate values fall outside
the control limits, an explanation must
be sought (such as instrument mal-
function, calibration drift, etc.). A
second, different sample must then
be analyzed in duplicate. No further
samples may be analyzed until duplicate
sample results are within the control
limits, unless approval is given by the
QA manager. The percent relative
standard deviation (%RSD) is calculated
as described below:
%RSD
x 100
/ Z(X-X)2\
\ n- 1 /
2V1/2
-------
Section 9.0
Revision 4
Date: 9/86
Page 8 of 13
where: s is the standard deviation
X is the mean
n is the routine (or other sample)
/duplicate pair (-2)
8. Matrix Spike Analysis - Matrix spike
analysis was performed ONLY for the
Phase I - Pilot Survey. A matrix spike
was not required for total extractabfe
At analyses. The procedure used is
as follows: One matrix spike is prepared
for each batch by spiking an aliquot
of a sample* with a known quantity
of analyte prior to analysis. The spike
concentration must be twice the endo-
genous level or 10 times the required
detection limit, whichever is larger.
Also, the volume of the spike added
must b© negligible (less than or equal
to 1 percent of the sample aliquot
volume). The spike recovery must be
100 +. 15 percent to be acceptable. If
the recovery is not acceptable for all
parameters, two additional, different
samples must be spiked with the ana-
lyte in question, must then be ana-
lyzed, and recoveries must be calcu-
lated. If one or both recoveries arenot
100 +. 15 percent, the entire batch
must be analyzed by standard addi-
tions for the parameter in question.
The standard addition is performed by
analyzing the sample, the sample plus
a spike at about the endogenous level,
and th© sample plus a spike at about
twice the endogenous level. The con-
centration of the matrix spike must
DA analysis on a foil sampia is
recommended. If sufficient sample
volume is not available, QA ana -
lysis may b® performed on a par
aliquot basis.
not exceed the linear range of the
method. For this reason, the matrix
spike for graphite furnace analyses,
which determine low levels of analyte,
must be chosen judiciously and may
be different than suggested above.
The samples may be diluted or the
spike levels may be adjusted so that
the linear range is not exceeded when
performing standard additions for
furnace AA analyses. The percent
recovery of spikes is calculated as
described below:
% spike recovery =
value of sample value of unspiked
plus spik© - sample
value of spike added
x 100
9. Ion Chromatography Resolution Test-
An ion chromatography resolution test
must be performed once per analytical
run (day) by analyzing a standard that
contains approximately equa! concen-
trations of nitrate and sulfate ions (1
mg/L). If the resolution does not
exceed 60 percent, the column should
be replaced and the resolution test
should be repeated.
10. Continuing Sample Analysis - The
remaining samples are analyzed if the
reagent biank, duplicate, and QC
samples are within limits. After every
10 (or fewer) samples and after the
last sample, a QC sample must be
analyzed to continually verify the
calibration curve. If the measured
value differs from the theoretical value
by more than the limits given in Table
9,2, the instrument is recalibrated and
the previous 10 samples are reanalyzed.
-------
Section 9.0
Revision 4
Date: 9/86
Page 9 of 13
9.5 Overall Internal Quality
Once the value of each parameter in
a sample is determined, there are
procedures for cheeking the correctness of
analyses. These procedures are outlined in
the following subsections,
9.5,1 Aniou-Cation Balance
Theoretically, the ANC of a sample
equals the difference, expressed as micro-
equivalents per liter (jneq/t,), between the
concentration of cations and the concentra-
tion of anions in a sample (Kramer, 1982).
In practice, this is rarely the case; devia-
tions are caused by analytical variability
and the presence of ions (protolytes) that
are not measured (e.g., organic ions). The
concentrations of these unmeasured ions
can be significant in natural lake samples.
The percent ion difference (%ID) calculation
below utilizes the ANC value to take these
ions into account; as a result, the calcu-
lation is more accurate. For each sample,
%ID is calculated as follows:
ANC + I anions - I cations
x100
TI
where:
I anions - JCI1 + IF"] -f [N031
+ [SQ421
I cations - [Na+3 + [K+] + £Caa+
+ EMg2+] + [NH4+]
TI • total ton strength
» ANC + I anlons
+ X eatlone + 2[H">"S
ANC » [HC03'1
+ [OH"] + [titrated
organic bases] - [H+]
(10'pH) x 10® j»q/L
All concentrations are expressed as
microaquivaients per liter. A list of factors
for converting mg/L to /Lieq/L for each
parameter is given in the methods manual
(Hillman at a!., 1988). Samples which have
a poor ton balance are reanalyzed.
Table 9.5 lists the criteria for reanalysis.
Prior to reanaiysis, the data should
be checked for possible causes of poor ion
balance. This check may indicate which
analytical results give rise to poor ion
balance and, hence, the parameters for which
the sample should be reanalyzed. Also,
careful examination of the data may explain
the poor ion balance. For example, if the
%ID is negative and the DOC is large enough
to account for the difference, unmeasured
organic anions are probably responsible;
thus, reanalysis is unnecessary. The QA
manager must be contacted when questions
arise regarding reanalysis.
9.5.2 Conductance Balance
An approximation of the conductance
of a sample can be calculated by adding
together the equivalent conductances for
each measured ion at infinite dilution. The
calculated conductances are determined
by multiplying the concentration of each
ion by the appropriate factor given in
Table 9.6. The percent conductance dif-
ference (%CD) is calculated as follows:
calculated cond . - measured eond .
measured conductance
Samples which have %CD§ that exceed
the limits listed in Table 9.5 are reanalyzed.
As with the %ID calculation, an unacceptable
valu© for %CD indicates either the presence
of unmeasured ions or an analytical error
in the measurement. For the surface waters
sampled, the ions included in the %CD cal-
culation are expected to account for 90 to
-------
Section 9.0
Revision 4
Date: 9/86
Page 10 of 13
Table 9.5. Chemical Reanalysls Criteria for Sample Ion Balance Difference and Percent
Specific Conductance Difference
Anion-Cation Balance
Total ion strength Cueq/L)
<50
>50 and IOO
Maximum % ion balance difference^
60
30
15
B. Specific Conductance
Measured specific
conductance (uS/cm)
<5
>5 and <30
>30
Maximum % specific
conductance difference*
50
30
20
a If the absolute value of the percent difference exceeds these values, the sample is reanalyzed.
When reanalysis is indicated, the data for each parameter are examined for possible analytical
error. Suspect results are then redetermined and the above percent differences are recalculated
(Peden, I98I). If the percent differences for reanalyzed samples are still unacceptable or no
suspect data are identified, the QA manager must be contacted for guidance.
100 percent of the ions in a sample. How-
ever, in contrast to the percent ion differ-
ence calculation, there is no term in the
%CD calculation to account for protolytes
that are not specifically determined. The
QA manager must be contacted when ques-
tions arise regarding reanalysis.
9.6 Instrumental Detection Limits
Instrumental Detection Limits (IDLs)
must be determined and reported weekly for
each parameter except pH, specific conduc-
tance, ANC, and BNC. For this study, the
IDL is defined as three times the standard
deviation of 10 nonconsecutive replicate
reagent or calibration blank analyses.
Calibration blanks are analyzed when a
method does not require a reagent blank.
In some analyses, such as those using ion
chromatography and Technicon AutoAnaly-
zers, a signal may or may not be obtained
for a blank analysis. If a signal is not
obtained for a blank analysis, the IDL is
defined as three times the standard deviation
of 10 nonconsecutive replicate analyses of
a standard whose concentration is three to
four times the required detection limit.
Detection limits must not exceed the limits
listed in Table 4.1.
9.7 Data Reporting
Results from each method are recorded
on the appropriate data forms (Table 9.4).
After a sample (all aliquots) is completely
analyzed, the results are summarized on
Form 11 (Summary of Sample Results) and
are reported to the number of decimal places
listed in Table 9.7. Results are annotated
by the data qualifiers (tags) listed in Table
9.8, where applicable. After a form is
completed, the analytical laboratory super-
visor must sign it to indicate that he or
-------
Section 9.0
Revision 4
Date: 9/86
Page 11 of 13
Tabto 9.6. Conductance Factors of
Ion
Ca2+
cr
cog2-
H+
HC03-
Mg2+
Specific Concuctance
(j/S/cm at 25 "C)
per mg/L
2.60
2.14
2.82
3.5 x 105
(per mole/L)
0.715
3.82
Specific Conductance
(pS/cm at 25 *C)
Ion per mg/L
Na+ 2.13
NH4+ 4.13
SO42" 154
NO3" 1.15
K+ 1.84
OH' 1.92 x 105
(per mole/L)
[H+] moles/L = 10'^
ph = initial pH measured before acidity titration
Kw
[Orf]
HCO " fmo/U =
HC03 (mg/L) =
CO * (ma/Li
C03 (mg/L)
4.4463 x 10'7
5.080 (DIC(mg/L))
4'996
4.6881 x 10
'11
Kw = 10'
-13.80
Taken from American Public Health Association et al. (1985) and from Weast (1972).
Conductance factors are not given for ionic aluminum, iron, or manganese because these ions are rarely
present in concentrations great enough to affect the percent conductance difference.
-------
Table 9.7. List of Decimal Place Reporting
Requirements
Recommended Number of Decimal
Parameter Places in Reported Results*
Al, total extractable 4
Al, total 4
ANC 1
BNC 1
Ca 3
CI" 3
DIG 3
DOC 2
F", total dissolved 4
Fe 3
K 3
Mg 3
Mn 3
Na 3
NH4+ 3
NO/ 4
pH 2
P, total 4
Si02 3
SO/' 3
Specific conductance 1
3 Report to the number of decimal places
in the actual IDL plus one.
she has reviewed the data and that the
samples were analyzed exactly as described
in the methods manual (HHIman et al., 1986),
All deviations from the manual require the
authorization of the QA manager prior to
sample analysis.
Section 9.0
Revision 4
Date: 9/86
Page 12 of 13
Copies of raw data must be submitted
as requested by the QA manager. All orig-
inal raw data must be retained by the lab
until notified otherwise by the QA manager.
Raw data include data system print-
outs, chromatograms, notebooks, QC charts,
standard preparation data, and all information
pertinent to sample analysis.
9.8 Daily Evaluation of Quality
Control Data
Each laboratory should make a daily
sample status report by telephone to the
EMSL-LV QA staff as directed. The objec-
tive of these reports is to keep the QA
manager informed of the status of the
internal and external QC checks in the
laboratory in order to identify and solve
problems that may arise. The reports also
allow the QA manager to obtain preliminary
results for the blanks, duplicates, and labor-
atory and field audit samples that are dou-
ble-blind to the laboratories. (A discussion
of blind and double-blind samples is present-
ed in Section 10.) Otherwise, these data
would not be available for QA/QC data
evaluation until they were reported by the
laboratories, which may be as long as 35
days after the samples are received. During
the daily telephone contact, the EMSL-LV
QA staff record all communications into a
bound notebook to track and resolve all
problems encountered during analyses.
Each week QC charts are updated and
new control and warning limits are deter-
mined. The QA chemist then performs a
QC audit in which all the data are reviewed.
Any values that lie outside the control or
warning limits are checked to verify that
they are not the result of a transcription
error. If bias is indicated (seven successive
points on one side of the theoretical mean),
analyses are stopped and an explanation is
sought. Copies of the plots are given to
the analytical laboratory supervisor, the
QA chemist, and each analyst.
-------
Section 9.0
Revision 4
Date: 9/86
Page 13 of 13
Table 9.8. National Surface Water Survey Lab/Field Data Qualifiers (Tags)
Qualifier
Indicates
A Instrument unstable.
B Redone, first reading not acceptable.
C Instruments, sampling gear not vertical in water column.
0 Slow stabilization.
E Result not available; sample destroyed during shipment.
F Result outside QA criteria (with consent of QA manager).
G Atypical result; already reanalyzed and confirmed by the laboratory manager.
H Holding time exceeded criteria.
J Result not available; insufficient sample volume shipped to analytical
laboratory from the mobile processing laboratory.
K Result not available; entire aliquot not shipped.
L Not analyzed because of interference.
M Result not available; sample lost or destroyed by laboratory.
N Not required.
P Result outside QA criteria, but insufficient volume for reanalysis.
Q Result outside QA criteria.
R Result from reanalysis.
S Contamination suspected.
T Leaking container.
U Result not required by procedure; unnecessary.
V % ion balance difference (%IBD) value (Form 16) outside criteria because of high DOC.
W % difference (%D) calculation for calculated ANC (Form 14) outside criteria because of
high DOC.
X, Y, Z Available for miscellaneous comments in the field and mobile processing laboratory only.
-------
Section 10.0
Revision 4
Date: 9/86
Page 1 of 4
10.0 PERFORMANCE AND SYSTEM AUDITS
10.1 Performance Audit Samples
Field and laboratory audit samples
are used as part of the QA activities for
NSS. The audit samples are shipped to
the analytical laboratories from the mobile
processing laboratory as though they were
routine stream samples. Every attempt is
made to ensure that the analytical laboratory
does not recognize the audit samples as
different from the routine samples. As a
result, the audit samples are double-blind
to the analytical laboratory. That is, the
laboratory neither recognizes them as audit
samples nor knows their compositions.
10.1.1 Audit Samples
The purpose of fiefd audit samples is
to identify problems that affect data quality
and that may occur during sample process-
ing, shipment, or analysis. These problems
could include sample contamination, sample
degradation, solvent evaporation, and impro-
per or inaccurate sample analysis. When
used in conjunction with laboratory audit
samples, the analysis of these samples pro-
vides data that can be used to distinguish
mobile processing laboratory problems from
analytical laboratory problems. There are
two types of field audit samples: synthetic
field audit samples and natural field audit
The synthetic field audit samples are
prepared at a central laboratory and are sent
to the mobile processing laboratory to
undergo ail the filtration and preservation
steps and to be labeled as though they were
authentic stream samples. Thus, they are
single-blind samples to the field laboratory
(i.e., recognized as audit samples but of
unknown composition) and, concurrently,
double-blind samples to the analytical lab-
oratory. The desired composition of the
synthetic field audit samples is shown in
Table 10.1.
Waters collected from Big Moos© Lake
in the Adirondack Mountains and Bagley Lake
in the state of Washington are available to
be utilized as natural audit samples for the
survey. The waters of Big Moose Lake are
low in alkalinity and thus are susceptible to
acidic deposition; the Bagley Lake waters
represent a medium level of alkalinity. These
natural samples are passed through a 0.45/.'
filter and are maintained at 4 °C to minimize
changes in composition. Aliquots are pre-
pared in the mobile processing laboratory
from 2-liter portions of these waters and
are included as part of a batch.
10.1.2 Laboratory Audit Samples
The purpose of these samples is to
identify problems that affect data quality and
that may occur during the analytical process.
Thus, lab audit samples help verify the
accuracy of analytical procedures and ensure
that the laboratory continues to properly
analyze samples.
The synthetic laboratory audit samples
are sent to the mobile processing laboratory
from a central laboratory, already split into
seven aliquots (eight aliquots for the Phase
I - Pilot Survey). The audit samples are
labeled by the mobile processing laboratory
personnel, are included in a batch with
routine stream samples processed on the same
day, and are shipped to the analytical lab-
oratory for analysis.
The desired composition of the synthetic
laboratory audit samples is given in Table
10.1. Only low-concentration synthetic samples
are used for NSS because the stream samples
are not expected to contain analytes are
higher levels.
-------
Section 10.0
Revision 4
Date: 9/86
Page 2 of 4
Table 10.1. D«*lred Composition Rang*
National Stream Sumy
Parameter
me*
At (total and total extractable)
BNC*
Ca
cr
Die
DOC
F", total dissolved
Fe
K
Mg
Mr?
Na
P, total dissolved
SiO2
Specific Conductance*
of Synthetic Field and Laboratory Audit Samples
Concentration Range
10-50
0.01-0.10
10-50
0.1-1.0
0.1-1.0
0.1-1.0
0.1-1.0
0.01-0.05
0.02-1.0
0.1-1.0
0.1-1.0
0.02-1.0
0.5-3.0
0.01-0.50
0.01-0.50
0.005-0.030
4-5
1-5
1-5
1-50
for the
Units
jjeq/L
mg/L
peq/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
pH
mg/L
mg/L
IjQ/cm
* These parameters are related and affeet the analytical results of one another.
h To be determined by concentration of other parameters.
Not§: M@@» balance (unions vs. cations) must be maintained. Nitrogen/phosphorus ratio must be
reasonable (10/20),
-------
Section 10.0
Revision 4
Date: 9/86
Page 3 of 4
10.1.3 Application of Audit
Sample Data
Data are obtained from the analyses
of the audit samples for the following pur-
poses:
• to judge the performance of the mobile
processing laboratory in the preparation
and shipment of samples
• to judge the continued capability of
the analytical laboratories to properly
analyze the samples
• to establish a statistically valid esti-
mate of the overall bias and precision
of the analyses
• to establish a statistically valid esti-
mate of the stability of a typical
stream sample when stored at 4 *C
by evaluating the natural lake sample
over the period of the study.
Acceptance windows are established
for the measurement of each parameter in
the audit samples. The size of the windows
is based on the information available for
each analytical method at the time the study
is initiated. If the analytical results for a
measurement fall outside the acceptance
window, the EMSL-LV QA staff reviews
the data to determine the cause of the
problem and immediately contacts the anal-
ytical laboratory, mobile processing labora-
tory, or field base, whichever Is appropriate,
to seek corrective action. Data for routine
samples analyzed with the audit samples
are also checked to determine if they were
also affected by the problem. If they were
affected, reanalysis of the samples in ques-
tion is requested from the analytical
laboratories. The establishment of the
acceptance windows is described in Sec-
tion 11.
Approximately ninety audit samples
are scheduled to be processed during NSS.
A statistical evaluation of the audit sample
data should provide a good estimate of the
bias and precision of the analytical methods
for each required measurement. Furthermore,
any changes over time in analytical results
for the natural-water audit samples without
corresponding change in the other audit
samples can be attributed to lack of analyte
stability.
The findings of a comparative study
between audit sample types will provide a
statistically valid estimate of the true maxi-
mum holding times allowable for each type
of analysis.
The audit samples are a key factor in
the NSS QA program. It is intended that
every effort be made to provide high-quality
audit samples.
10.2 Quality Assurance System
Audits (On-Site Evaluations)
The systems audit consists of qualitative
evaluation of field and analytical laboratory
facilities, equipment, and operations such
as record keeping, data reporting, and QC
procedures.
10.2.1 Field and Mobile Processing
Laboratory Operations On-Site
Evaluation
Each NSS field base and sampling team
can expect at least one on-site evaluation
during the course of the sampling effort.
This is an on-site inspection to review the
sampling procedures, field base operations,
sample processing, sample analyses, and QA
efforts.
For each field base, the corresponding
sampling team, and the mobile processing
-------
Section 10.0
Revision 4
Date: 9/86
Page 4 of 4
laboratory, the on-site evaluation should
be conducted as soon as possible after the
start of monitoring. The questionnaire given
in Appendix B is used to assist in the eva-
luation.
The field auditor conducts an in-depth
review of all sampling and processing opera-
tions. This includes but is not limited to
(a) interviewing the field base coordinator,
(b) interviewing each sampling team, (c)
accompanying one or more of the sampling
teams during a sample excursion, (d) inter-
viewing the supervisor of the mobile pro-
cessing laboratory, (e) observing operations
at the mobile processing laboratory, and (f)
writing a summary report that includes
results, observations, and recommendations.
If any problems are found, the evaluator
must either correct them or must bring
them to the attention of the field base coor-
dinator or mobile processing laboratory
supervisor.
10.2.2 Analytical Laboratory
On-Site Evaluation
Each analytical laboratory participating
in NSS can expect a minimum of two in-
depth, on-site evaluations conducted by
the EPA QA manager or the QA manager's
authorized representative. The questionnaire
in Appendix C is used to assist in the on-
site laboratory evaluation.
The first on-site laboratory evaluation
is performed after the laboratory has suc-
cessfully analyzed a set of Pre-Award Per-
formance Evaluation (PE) samples for the
contract-required parameters and before
the actual survey analytical work begins.
The PE samples may contain some or all
of the analytes for which determination is
required, in the expected concentration
ranges. The PE sample results are scored
using the NSWS Pre-Award Audit Sample
Score Sheet given in Appendix D. Grading
emphasizes analytical accuracy, but a sub-
stantial portion of the grade depends on
meeting the QA, internal QC, reporting, and
deliverable requirements.
The auditor summarizes all observations
in an on-site laboratory evaluation report
and brings ail problems that occur to the
attention of the laboratory manager for
corrective action.
The second on-site laboratory evaluation
is conducted after approximately one-third
of the NSS analyses have been completed.
During the second on-site evaluation, QA
sample (audit, duplicate, and blank) data
and QC data received to date are reviewed.
The laboratory questionnaire is updated, if
necessary, to note all changes that have
been made since the first on-site evaluation.
An on-site laboratory evaluation report is
written for this and for each additional
on-site laboratory evaluation.
-------
Section 11
Revision 4
Date: 9/86
Page 1 of 2
710 ACCEPTANCE CRITERIA
11.1 Audit Sample Results
Acceptance windows for single values
from audit samples are based on previous
interlaboratory analyses of the same sample
material. The objective of creating windows
is to predict intervals for acceptable single
future values based on a sample mean (X)
and sample standard deviation (s) computed
from n previously observed values. The
limits of the windows are determined by
using a t-statistic (t).
is a Student's t
where:
Z-is the standard normal variate, having
a normal distribution with a mean of 0
and a variance of 1
p is a variable with a chi-square distri-
bution with r degrees of freedom, and
Z and fj are independent
The observed values X-,, X2. X3, ... X,,
are independent and have a normal distri-
bution (N) with a population mean (/u) and
variance (a2). A (1 - a) prediction interval
or a single future value y is needed. Let
X = sample mean and s = sample standard
deviation. It is known that:
y~ N fa, a2) and X~N \/i,
Therefore,
y-X~N 0, a2 1 +
n
y-x
N (0, 1)
n
n-1 — ~X2
(n-1)
and
r =n-1.
Substituting,
y -
y-x
(n - 1)a2
n
The upper and lower limits of the window
can be formalize^ as follows:
X + (t)(s)
X - (t)(s)
= upper limit of
n the window
1 + = lower limit
n the window
of
The Student's t-value (t) has n-1
degrees of freedom. The t-value is for a
2-tailed test with a cumulative probability
of 0.95 (i.e., 2.5 percent probability on either
side).
For predicting future values, wider
windows than the standard 95 percent con-
fidence interval about the mean are desir-
able. As the number of observed values
increases, more variance occurs because of
chance alone.
-------
Section 11
Revision 4
Date: 9/86
Page 2 of 2
Initially, there may not be sufficient are still unacceptable, further corrective
data (n < 10) available to provide good action must be initiated.
interval estimates. Arbitrary criteria may
be used until 10 or more values are avail-
able. The windows should be periodically
updated as more data are accumulated.
Grubbs' test (Grubbs, 1969) is applied
to the data before interval estimation to
detect outliers. The outliers are excluded
from the computation of the windows.
11.2 Duplicate Anatysis Results
Acceptance criteria for the RSD are
based on the upper 95th percentile of ob-
served values of RSD. Because the RSD is
affected by concentration, these criteria
are applied only when the mean of the
duplicate analyses exceeds the contract-
required detection limit (CRDL) by a factor
of 10.
Arbitrary acceptance criteria may be
used until sufficient (at least 10) RSD values
have been observed.
The distribution of the RSD values
cannot be estimated accurately until suf-
ficient RSD values have been observed. It
is recommended that no outlier test be
applied until the distribution has been esti-
mated.
11.3 Blank Analysis Results
Windows for blank analysis results are
computationally identical to those for dup-
licate sample results. Historical data will
be used to calculate these windows.
11.4 Corrective Action
Laboratories which fail to meet the
acceptance criteria for analysis of audit
samples or duplicates are required to repeat
the analysis that produced the erroneous
results. If results from the second analysis
-------
Section 12.0
Revision 4
Date: 9/86
Page 1 of 9
12.0 DATA BASE MANAGEMENT SYSTEM
The purpose of the data base manage-
ment system is to assemble and store data
generated as part of NSWS, to provide basic
reports of the survey results, to perform
statistical analyses describing target popu-
lations, and to provide data security. A
detailed description of the system is given
in the Data Management Proposal (ORNL,
1984; Sale et al., in preparation). The
relationship of data base management to
the overall NSWS is shown in Figure 12.1.
The data are stored in four major
data sets, namely (1) a raw data set, (2) a
verified data set, (3) a validated data set,
and (4) a final data set. All data sets are
protected from unauthorized or accidental
access by individual, system, and file pass-
word protection.
12.1 Raw Data Set
At ORNL, the Statistical Analysis
System (SAS) is used to enter the field
and laboratory data (analytical results and
data qualifiers - see Table 9.8) reported on
Data Forms 4, 5, 11, 13, 18, 19, 20, and 22
into the raw data set. The data package
consisting of these forms is also sent to
the EMSL-LV QA staff for concurrent data
analysis. Data receipt is acknowledged,
and field and laboratory personnel verify
that all forms are received by the data
base management personnel.
The SAS full-screen editor procedure
is used to provide initial error checking as
data are entered. All data are entered into
two separate data sets by two different
operators. For the NSWS data base, a
custom program (COMPARE) has been dev-
eloped in SAS to compare the two data
sets and to identify any inconsistencies in
numeric and alphabetic variables. The
advantage of this double entry and com-
parison process is that typographical errors
are identified and are removed from the
system.
12.2 Verified Data Set
The raw field and laboratory data are
transmitted on magnetic tapes to the EMSL-
LV QA group. All data are then evaluated
and verified, and appropriate flags (see Table
12.1) are applied to the raw data as described
in Section 13.0. The data are processed
using the "Automated Quality Assurance
Review, Interactive Users System" (AQUA-
RIUS II), an online QA system developed
by the EMSL-LV QA staff. Reports gen-
erated by AQUARIUS II range in subject
from complex protolyte analysis to simple
external and internal blank checks for QA
purposes (see Table 13.1).
For the Pilot Survey, as for the East-
ern Lake Survey - Phase I, AQUARIUS was
used to generate exception tuples (change
records) that were sent to ORNL for imple-
mentation. For the other three NSS surveys,
a modified system called the Aquatics Anal-
ysis System (AQUARIUS II) was developed
based on transaction processing. Rather
than generating and applying tuples, the
modified system generates data changes in
the form of transaction records from ex-
ception programs and from manually edited
records copied from a local master data
base (LMD). These transaction records
can be used by the EMSL-LV QA group
and by ORNL to update the LMD and the
official raw data set, respectively. Results
of EMSL-LV QA group verification consist
of a copy of the updated LMD and a history
file that contains the transaction records
that were used to update the LMD. A
detailed flow scheme of the Aquatics Analy-
sis System is presented in Figure 12.2.
In addition to the standard QA ana-
lyses, AQUARIUS II is used to generate
-------
/FIELD BASE SITES/X
(MOBILE PROCESSING)
f ANALYTICAL A
V LABORATORIES J
Section 12.0
Revision 4
Date: 9/86
Page 2 of 9
Figure 12.1. Data ba*e management for the National Surface Water Survey.
-------
Section 12.0
Revision 4
Date: 9/86
Page 3 of 9
Table 12-1. National Surface Water Survey Verification Data Qualifiers
(Flags) for Raw Data Set
FLAGS USED WITH ANION/CATION BALANCE CHECK PROGRAM:
AO Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
unknown cause.
A1 Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
unmeasured anions/cations (other anions/cations not considered in % ion
balance difference calculation).
A2 Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
anion (flag suspect anion) contamination.
A3 Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to to
cation contamination.
A4 Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
unmeasured organic protoiytes (fits Oliver Model).
AS Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
possible analytical error - anion concentration too high (flag suspect anion).
A6 Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
possible analytical error - cation concentration too low (flag suspect
cation).
A7 Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
possible analytical error - anion concentration too low (flag suspect anion).
AS Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
possible analytical error - cation concentration too high (flag suspect
cation).
A9 Anion/Cation % Ion Balance Difference (%IBD) is outside criteria due to
^gSlbje_jnajytjcjaj_e_rror - alkalinity (ANC) measurement.
FLAGS GENERATED BY APPROPRIATE BLANK EXCEPTION PROGRAM:
BO External (field) blank is above expected criteria for pH, DIG, DOC, specific
conductance, ANC, and BNC determinations.
B1 Internal (laboratory) blank is >2 x CRDL for DIC, DOC, and specific
conductance determinations.
(Continued)
-------
Section 12.0
Revision 4
Date: 9/86
Page 4 of 9
Table 12-1. (Continued.)
FLAGS GENERATED BY APPROPRIATE BLANK EXCEPTION PROGRAM (continued):
B2 External (field) blank is above expected criteria and contributed >20% to
sample concentrations. (This flag is not used for pH, DIG, DOC, specific
conductance, ANC, and BNC determinations.)
B3 Internal (laboratory) blank is 2 x CRDL and contributes >10% to the
sample concentrations. (This flag is not used for DIG, DOC, and specific
conductance determinations.)
B4 Potential negative sample bias based on internal (laboratory) blank data.
B5 Potential negative sample bias based on external (field) blank data.
FLAGS USED WITH CONDUCTANCE BALANCE CHECK PROGRAM:
CO % Conductance Difference (%CD) s outside criteria due to unknown cause.
C1 % Conductance Difference (%CD) is outside criteria due to possible
analytical error-anion concentration too high (flag suspect anion).
C2 % Conductance Difference (%CD) is outside criteria due to anion
contamination.
C3 % Conductance Difference (%CD) is outside criteria due to cation
contamination.
C4 % Conductance Difference (%CD) is outside criteria due to unmeasured
organic ions (fits Oliver Model).
C5 % Conductance Difference (%CD) is outside criteria due to Qossjbje
analytical error in secific conductance
C6 % Conductance Difference (%CD) is outside criteria dua to Qcjsjbje
ajTaj£tijsL§!rfir--anion concentration too low (flag suspect anion).
C7 % Conductance Diffarence (%CD) Is outside criteria due to ynmeasyrecj
anigns/cations (other anions/cations not measured in % conductance
difference calculation).
08 % Conductance Difference (%CO) is outside criteria due to Qgjsjbje
anal^tjca,Lerror--cation concentration too low (flag suspect cation).
(Continued)
-------
Section 12.0
Revision 4
Date: 9/86
Page 5 of 9
Table 12-1. (Continued.)
FLAGS USED WITH CONDUCTANCE BALANCE CHECK PROGRAM (continued):
C9 % Conductance Difference (%CD) is outside criteria due to possible
analytical error-cation concentration too high (flag suspect cation).
FLAGS GENERATED BY DUPLICATE PRECISION EXCEPTION PROGRAM:
D2 External (field) duplicate precision exceeded the maximum expected %
relative standard deviation (%RSD), and both the routine and duplicate
sample concentrations were >.10 x contract required detection limit (CRDL).
D3 Internal (laboratory) duplicate precision exceeded the maximum required %
relative standard deviation (%RSD), and both the routine and duplicate
sample concentrations were >.10 x contract required detection limit (CRDL).
FLAGS USED WHEN FIELD DATA ARE OUTSIDE CRITERIA:
FO % Conductance difference (%CD) exceeded criteria when in situ field
conductance value was substituted.
F1 Hillman/Kramer protolyte analysis program indicated field pH problem when
stream site pH value was substituted.
F2 Hiiiman/Kramer protolyte analysis program indicated unexplained problem
with stream site pH or processing laboratory PIC values when stream site
pH value was substituted.
F3 Hillman/Krarner protolyte analysis program indicated field problem-mobile
F4 Hiliman/Kramer protolyte analysis program indicated field problem-mobile
processing laboratory PIC.
F5 Hillman/Kramer protolyte analysis program indicated unexplained problem
with mobile processing laboratory pH or DIG values when mobile processing
laboratory phi value was substituted.
F6 % Conductance Difference (%CD) exceeded criteria when processing
laboratory (trailer) specific conductance value was substituted.
FLAGS GENERATED BY HOLDING TIME EXCEPTION PROGRAM:
HO The maximum hoiding time criteria were not met.
H1 No "Date Analyzed" data were submitted for reanalysis data.
(Continued)
-------
Section 12.0
Revision 4
Date: 9/86
Page 6 of 9
Table 12-1. (Continued.)
FLAG GENERATED BY DETECTION LIMIT EXCEPTION PROGRAM:
L1 Instrumental Detection Limit (IDL) exceeded contract required detection
limit (CRDL) and sample concentration was <10 x instrumental detection
limit.
MISCELLANEOUS FLAGS:
MO Value obtained using a method which is unacceptable as specified in the
Invitation for Bid contract.
FLAGS GENERATED BY AUDIT CHECK PROGRAM:
NO Audit sample value exceeded upper control limit.
N1 Audit sample value was below control limit.
FLAGS GENERATED BY HILLMAN/KRAMER PROTOLYTE ANALYSIS PROGRAM:
PO Laboratory problem-initial pH from alkalinity (ANC) titration.
P1 Laboratory problem-initial pH from acidity (BNC) titration.
P2 Laboratory problem-unexplained - initial pH from ANC or BNC titration.
P3 Laboratory problem-initial DIG determination.
P4 Laboratory problem-air-equiiibrated pH or DIC determinations.
P5 Laboratory problem-unexplained - initial pH from ANC or BNC titrations or
initial DIC determinations.
P6 Laboratory problem-alkalinity (ANC) determination.
P7 Laboratory problem-CO2-acidity (BNC) determination.
FLAGS GENERATED BY QCCS EXCEPTION PROGRAM(S):
Q1 Quality Control Check Sample was above contractual criteria.
Q2 Quality Control Check Sample was below contractual criteria.
Q3 Insufficient number of Quality Control Check Samples were measured.
Q4 No Quality Control Check Sample was analyzed.
(Continued)
-------
Section 12.0
Revision 4
Date: 9/86
Page 7 of 9
Table 12-1. (Continued.)
FLAGS GENERATED BY QCCS EXCEPTION PROGRAM(S):
Q5 Detection Limit Quality Control Check Sample was not 2 to 3 x Contract
Required Detection Limit and measured value was not within 20% of the
theoretical concentration.
MISCELLANEOUS FLAGS:
MO Value obtained using a method which is unacceptable as specified in the
Invitation for Bid contract.
M1 Value reported is questionable due to limitations of the laboratory
methodology.
XO Invalid but confirmed data based on QA review.
X1 Extractable aluminum concentration is greater than total aluminum
concentration by 0.010 mg/L where extractable aluminum >. 0.015 mg/L
X2 Invalid but confirmed data-potential aliquot switch.
X3 Invalid but confirmed data-potential gross contamination of aliquot or
parameter.
X4 Invalid but confirmed data-potential sample (all aliquots) switch.
Values for flags XO through X4 should not be included in any statistical
analysis.
MISSING CODE VALUE
"." Value never reported.
(Note: This code appears in numeric fields only.)
-------
LOCAL
MASTER
DATA BASE
I - Data checked after first program runs
2 - Data checked before manual edits are entered
3 - Data checked after transactions are generated
4 - Data checked before update
5 - Data checked for correctness and completeness
6 - Data checked to make sure update was
completed correctly
TRANSACTION
FILE
0) D) CD CD
tn ri < O
* Data are reviewed by at least two individuals for each process.
-*
-------
Section 12.0
Revision 4
Date: 9/86
Page 9 of 9
various printouts supplied to the QA manager
to point out intralab, interlab, and interfield
bias, as well as discrepancies in blanks,
audits, or other QA samples. The overall
outcome is a verified data set in which all
questionable values are qualified. The QA
personnel coordinate with the field bases,
the mobile processing laboratory, and the
analytical laboratories to make all appro-
priate corrections in the data.
12.3 Validated Data Set
The verified data set is provided to
the ERL-C staff on a magnetic tape, and the
staff initiates the validation process. The
validation process increases the overall
integrity of the data base by evaluating all
data for internal and regional consistency
using all QA and QC information available.
The validation process compares data
for a set of variables against a more re-
stricted range utilizing knowledge of rela-
tionships in aquatic chemistry and limnology
to identify intersite sample inconsistencies.
Intersite validation consists of comparing
single site values with values from adjacent
sites within a region. Data for groups of
sites are compared to check for consistency.
The validation process is discussed further
in Section 14.0. After undergoing this
review process, the data, site by site, are
transferred to the validated data set.
12.4 Final Data Set
Calculating population estimates is
difficult if values are missing from the data
set. A final data set (Data Set 4) is pre-
pared to resolve such problems by inserting
reliable values where ones are missing in
the validated data set (Data Set 3). Data
Set 4 also is modified from Data Set 3 by
averaging field duplicate values (if QA
precision criteria are met) and by replacing
analytical values determined during validation
to be erroneous.
In those cases where a value is missing
or incorrect (i.e., the value is identified as
an outlier during validation and the aber-
ration is not a result of an episode or some
site condition) and a new value must be
incorporated, the new value is obtained
from one of the following sources, listed
from most to least desirable:
1. Value from the duplicate of an R/D
pair for the routine value.
2. Value from an alternate sample.
3. Value of a redundant measurement on
the same sample.
4. Value predicted from the best regression
of related variables (e.g., Ca vs. ANC,
pH vs. ANC) or from the best regression
of the same variables; these regressions
include all comparisons and combinations.
If there are four or more variables
identified as outliers for the same sample,
the site is considered unusual.
Another modification is that negative
values for parameters other than ANC that
resulted from analytical calibration bias
are set equal to zero. The bias in the
estimate of variance due to this adjustment
is not expected to affect data analyses.
Ail values modified in the final data set
are flagged. After the final data set is
completed, the data will be released by EPA
and will be made available to all data users.
-------
Section 13.0
Revision 4
Date: 9/86
Page 1 of 7
13.0 DATA EVALUATION AND VERIFICATION
As the field and analytical laboratory
data are received by the EMSL-LV QA staff,
all data are evaluated based on the available
QA and QC information, using the estab-
lished and organized review process describ-
ed here. The objective of the data verifi-
cation process is to identify, correct, or
flag data of unacceptable quality. Computer
programs have been developed to automate
this process as much as possible. Each
batch of data is evaluated on a sampie-by-
sample basis, as described in the following
sections. Figure 13.1 is a summary of the
verification process.
13.1 Field Data Review
Each field data form is reviewed to
check for the following items:
1. Stream ID. Forms 4, 4A, 6, and 7
are compared to Form 5 to identify
and correct transcription errors.
2. Trailer Duplicate. On Form 5, a dupli-
cate stream sample ID should match
a routine stream sample ID.
3. pH. The streamside pH reading record-
ed on Form 4 is compared to the
mobile processing laboratory pH reading
on Form 5. The difference must be
< 0.3 pH units.
4. Mobile Processing Laboratory pH, DIG,
Specific Conductance, and Nonex-
changeable and Total PCV Reactive
Aluminum. Form 5 measurements for
field audit samples are evaluated in
accordance with the associated accept-
ance criteria. Routine-duplicate pairs
and trailer duplicate pairs are also
evaluated for precision.
5. DIG, pH, Specific Conductance, and
Nonexchangeable and Total PGV Reac-
6.
tive Aluminum QCCS Data. Form 5
QCCS data are evaluated to ensure
that QCCS criteria are met.
Data Qualifiers. Comments and data
qualifiers are reviewed for correct
use and consistency.
Data anomalies are reported to the
mobile processing laboratory coordinator for
review, and data reporting errors are reported
to ORNL to be corrected before entry into
the raw data set. All telephone communi-
cations are recorded in bound notebooks,
and data corrections (e.g., transcription
errors, missing data, incorrect units, and
incorrect use of data qualifers) are annotated
on the appropriate forms before they are
sent to ORNL for data entry.
13.2 Analytical Data Review
13.2.1 Daily Quality Assurance
Communications
Daily calls are made to each field base,
to the mobile processing laboratory, and to
each analytical laboratory to ensure that
QA and QC guidelines are being followed and
that samples are being handled and analyzed
properly, to obtain current sample data,
and to discuss problems that may occur
during analyses.
The primary objective of these calls
is to identify and resolve issues quickly,
before they affect data quality or interfere
with the completion of the survey. Pre-
liminary sample data are obtained verbally
or by computer, depending on the capabilities
of the analytical laboratory. The preliminary
data are evaluated by comparing the QA
sample data against acceptance criteria.
Responsible parties are notified of problems
and all interactions are recorded in bound
notebooks.
-------
Section 13.0
Revision 4
Date: 9/86
Page 2 of 7
Enttr
Raw DBU
(DouMt Entry)
ORNL
• \
Compta.
Raw Dm Sit
on Tip*
ORNL
*
Conduct
Automitxf
QA RMim
LEMSCO
ComptoU
NSWS
Vwificotion (Upon
LEMSCO
Review
Exception Racordi
(Vllut Ch»ng»)
LEMSCO
t
Encut*
Exoption
GwMriting Progrtmt
LEMSCO
(No RxMlyM if rtoqulrad from FMd Cram)
'XBWortaffFll.x-''
^\? ./^
jVt«
VvMM
Dro S«
Complete
ORNL
CortHxw. Chintid
Dm SM
With Rtw Ora
ORNL
G«Mr>t«
VvitM Dm. SM
Tw
LEMSCO
|
Sind
VwifM Ctau S«
Tip.
to ORNL
Figure 13.1. Flowchart for data verification process.
-------
13.2.2 Preliminary Review of Sample
Data Package
The sample data packages are reviewed
for completeness, internal QC compliance,
and appropriate use of data qualifiers. The
Data Package Completeness Checklist in the
verfication report (given in Appendix E) is
used to assure consistency in the review of
all data packages. Any discrepancies related
to analytical data are reported to the appro-
priate analytical laboratory manager for
corrective action. If discrepancies affect
billing or data entry, then SMO or ORNL
is notified. Comments provided in the cover
letter are also reviewed to determine their
impact on data quality and the need for
any follow-up action by the laboratory.
This data review process is aiso important
in verifying that the contractual require-
ments are met for the purpose of payment.
13.2.3 Review of Quality Assurance
and Quality Control Data
The analytical data reported on data
forms are entered into the raw data set by
ORNL as the data packages are received.
A magnetic tape containing raw data is sent
to the National Computer Center (NCC),
Research Triangle Park, North Carolina,
for use on the EPA IBM 3081 computer.
Each tape received by the NCC tape library
is given a volume serial number and a BIN
number that indicates the physical location
of the tape. The tape is loaded remotely
by the EMSL-IV QA staff, and exception
programs, listed in Table 13.1, are generated
by AQUARIUS II.
The NSS Verification Report (Appendix
E) is completed with the use of outputs from
exception reports (along with the original
data, mobile processing laboratory data, and
field notebooks). The verification report
is a worksheet designed to systematically
guide the auditor through the verification
process by explaining how to flag data,
tracking data resubmissions, tracking reana-
Section 13.0
Revision 4
Date: 9/86
Page 3 of 7
lysis and confirmation requests, listing the
steps to help explain the QA exceptions,
summarizing all modifications to the raw
data set, and listing all flagged sample data.
One hundred percent of the analytical
data are verified, sample by sample. Stream
sample analytical data that meet the anion/
cation %IBD, the %CD, and the internal and
external QA and QC criteria can be regarded
with a high degree of confidence. If the
%IBD or %CD results are outside the specified
limits but the discrepancy can be explain-
ed by either the presence of organic species
(as indicated by the protolyte analysis pro-
gram) or an obvious correctable reporting
error, the data are still verified by the
EMSL-LV QA staff.
Additional flags are applied to a given
parameter, even though the verification is
on a "per sample" basis, when the batch
QA sample data do not meet the acceptance
criteria for QA sampies such as field blanks,
field duplicates, or audit samples. Each
parameter is also flagged if internal QC
checks such as calibration and reagent blank
analytical results, internal duplicate precision,
instrumental detection limits, QCCS analytical
results, and required holding times do not
meet specifications. The final source of
flags is the protolyte analysis program. A
detailed description of the evaluation of
DIG, pH, MIC, and BNC data by the protolyte
analysis program is given in Section 13.2.4.
In all cases, the flags that are generated
by the computer programs are reviewed by
the auditor for reasonabieness and consistency
before they are entered into the verified
data set.
13.2.4 Computer Evaluation of DIC,
pH, ANC, and BNC Data by
Protoiyte Analysis
An evaluative computer program per-
forms data checks and uses carbonate equil-
ibria and DOC data to identify analytical
error and the source of protolytes (acidic
-------
Section 13.0
Revision 4
Date: 9/86
Page 4 of 7
Table 13.1. Exception Generating and Data Review Programs of Aquarius II
Program
Type
Exception Generating Programs:
1 = Audit Sample Summary
2 = Field Blank Summary
3 = Field Duplicate Precision Summary
4 = Instrumental Detection Limit Summary
5 = Holding Time Summary
8 = % Conductance Difference Calculations
7 = Anion/Cation Balance Calculations
8 = Internal Lab Duplicates
9 = Protolyte Analysis (DIC, pH, ANC, and BNC Data Evaluation)
10 = Reagent/Calibration Blanks, QCCS, and Detection Limit QCCS
Data Review Programs:
1 = Raw Data Listing - Format for QA Manager
2 = Complete Raw Data Listing - Format for Audit Staff
3 - Comparison of Form 4 and Form 5
4 = Comparison of Form 5 and Form 11
5 = QA/QC Flag Summary
6 = Modified Gran Analysis Program
(FL,LL,FN,LN)
(B)
(R/D Pairs)
(All Species)
(All Species)
(All Species)
(All Species)
(pH and DIC)
(pH and DIC)
or basic species) in the sample. Thus, the
DIG, pH, ANC, and BNC data are rigorously
evaluated in light of known characteristics
of carbonate equilibria. The overall process
of data evaluation based on carbonate equil-
ibria is summarized below.
13.2.4.1 Redundant Alkalinity
Checks for pH and DIC
Evaluations of carbonate equilibria
indicate that alkalinity is not affected by
changes in dissolved CO2 concentration.
Furthermore, alkalinity can be calculated
from carbonate equilibria if the DIC and
pH are known. A theoretical alkalinity, C,
is calculated from each of the three pH/DIC
pairs:
C1 - pH/DIG of "closed system" syringe
samples (mobile processing laboratory)
Cfl-
pH/DIC of "open system" samples
(analytical laboratory)
pH/DIC of "air-equilibrated system"
samples (analytical laboratory)
The third data pair is obtained on an
aliquot that has been equilibrated with stand-
ard air (300 ppm CO2). If there is no
analytical error, the three calculated aSka-
linities should agree within the limits of
experimental error. The precision for cal-
culated alkalinity values of less than or
equal to 100 peq/L should be within 10 ^eq/L
and within 10 percent for calculated alkalinity
values greater than 100 jueq/L. The precision
windows are based on the estimated precision
of the pH and DIC measurements used in
the calculations. If this comparison indicates
a potential analytical error (i.e., the precision
limit is exceeded), the redundant pH and
-------
DIG values are compared to identify the
source of error. Further evaluation of the
QA and QC information for the individual
data pairs usually identifies one of the pH
or DIG measurements within the outlier
pair as the source of error. Because the
measurement is redundant, an acceptable
pH or DIG value from one of the data pairs
should be available to the data user for
every sample that is analyzed.
13.2.4.2 Verification of Measured ANC
The measured ANC is evaluated by
comparing it to the average of the ac-
ceptable calculated values for alkalinity
determined during the evaluation of pH and
DIG.
Carbonate Systems - For a true carbonate
system, the measured ANC should equal
(within the limits of experimental error)
the calculated alkalinity. The difference
between measured ANC and the calculated
alkalinity should be within 15 /ueq/L for
calculated alkalinities less than or equal to
100 A*eq/L, and within 10 percent for larger
values. If the measured ANC differs from
the calculated alkalinity, an analytical error
is indicated in the titration or in the pH
or DIC measurements.
Mixed Systems - Mixed systems are those
represented by samples that have significant
concentrations of other protolytes in addi-
tion to the carbonate species. In naturai
waters, weak conjugate bases of natural
humic and fulvic acids are often present
and can contribute significantly to the ANC.
The acidic functional groups of natural
humic substances contribute to the BNC of
natural waters as well. Two empirical
relationships among DOC, pH, and organic
protolytes have been proposed by Oliver et
al. (1983). The first relates the total or-
ganic protolyte to DOC, and the second
relates the mass action quotient (pKo) of
the organics present to the sample pH.
Section 13.0
Revision 4
Date: 9/86
Page 5 of 7
DOC and pH are measured in each
sample. The empirical relationships (defined
by the Oliver model) and the measured pH
and DOC values are used to estimate the
contribution of organic protolytes to the
measured ANC. The measured ANC should
equal, within experimental error, the sum
of the calculated alkalinity and the estimated
organic protolyte contribution, assuming
that significant concentrations of other
non-organic protolytes are not present and
there is no analytical error. The precision
should be within 15 A/eq/L for calculated
ANC less than or equal to 100 A/eq/L and
within 10 percent for larger values.
13.2.4.3 Verification of Measured BNC
BNC, unlike ANC, is affected by chan-
ges in dissolved CO2 concentration. There-
fore, evaluation and verification of those
data cannot utilize as much redundancy as
that of ANC data. Only the initial pH and
DIC values determined in the analytical
laboratory (data pair C2) can be used to
calculate BNC for comparison with the
measured value. As with ANC, other proto-
lytes can contribute to the measured BNC.
An estimate of CO2-acidity is calculated
from data pairs and carbonate equilibria.
If no other protolytes are present, the
calculated acidity should equal, within the
limits of experimental error, the measured
BNC. Precision for calculated acidity values
less than or equal to 100 /jeq/L should be
within 10 fjeq/L and within 10 percent for
larger values. If the calculated acidity is
greater than the measured BNC, an analytical
error in the pH, DIC, or BNC determination
is indicated.
The pH and DIC measurements are
verified by the previous tests (QA/QC redun-
dancy and alkalinity checks). If the cal-
culated acidity is less than the measured
BNC, the difference may be due to the
presence of other protolytes or to an analy-
tical measurement error. The Oliver model
-------
is used to evaluate the contribution from
organic acids.
13.2.4.4 System Check for Total
Carbonate
For a carbonate system, it can be
shown that the sum of alkalinity and acidity
equals total carbonate concentration in the
sample. For a mixed system, it can be
shown that the sum of ANC and BNC equals
the total protolyte concentration in the
sample. Thus, the calculated values of
alkalinity and acidity can be combined and
compared to the sum of the measured ANC
and BNC, as an additional check of the
data. For a carbonate system, the sum of
ANC and BNC should equal, within the limits
of experimental error, the total carbonate
concentration or the sum of calculated
acidity and alkalinity. If this sum is less
than the calculated total carbonate, an
analytical error is indicated because the
two titrations must account for all carbonate
species present in the sample. Other proto-
lytes or analytical error is indicated if the
sum of ANC and BNC exceeds the calculated
total carbonate. Again, the Oliver model
is used to evaluate the data.
The precision for this evaluation should
be within 15 //mole/L for total carbonate
concentrations less than or equal to 100
pmole/L, and within 10 percent for higher
concentrations. The protolyte analysis
program generates flags (Table 12.1), based
on the data checks described above, to
indicate the source of problems. Flowcharts
that demonstrate these data checks are
available from EMSL-LV.
13.2.5 Follow-up with Analytical
Laboratories
After the review of all data is com-
pleted, the analytical laboratories are re-
quested to resubmit data reporting forms
that are incomplete, to submit corrections
Section 13.0
Revision 4
Date: 9/86
Page 6 of 7
of previously reported data, to confirm
previous results, and to reanalyze certain
samples that do not meet QA and QC cri-
teria. In certain cases, the EMSL-LV QA
staff may request that the analytical lab-
oratory submit the raw data for a particular
sample or batch. These raw data are used
(1) to evaluate data anomalies not easily
explained or corrected during the data review
process and (2) to support requests for sample
reanalysis or value confirmation. The analy-
tical laboratories are required to submit
confirmation and reanalysis data on Form
26 (see Figure 13.2). The analytical labo-
ratories are directed to respond within a
reasonable time so that the results are
evaluated in time for them to be useful
to the survey.
13.2.6 Evaluation of Outliers
Generated by Contains
Staff
During the verification process, outliers
(defined in Section 14.2) identified by the
ERL-C staff are examined further by the
EMSL-LV QA staff. For any of these out-
liers not identified previously, confirmation
of the value is requested from the contract
analytical laboratory. Any value changes
are incorporated into the changed data set
before it is sent to ORNL.
13.2.7 Preparation and Delivery
of Verification Tapes
The steps identified in Sections 13.2.2
through 13.2.6 are followed to identify suspect
data and to correct erroneous data. The
information obtained by this process is ac-
cumulated by the EMSL-LV QA staff and
is placed on magnetic tapes, which are sent
to ORNL There, the new data and qualifiers
are entered into the raw data set to correct
and flag the original data. The identification
and transfer of corrected data for entry
into the verified data set are described
more fully in Section 12.
-------
Section 13.0
Revision 4
Date: 9/86
Page 7 of 7
DATE RECEIVED
NATIONAL SURFACE WXIER SURVEY
FORM JS
Dili Confirmation/Rejnalysis Request Form
Bitch f
The following values
rariable
NSUS
Form •
ontracl Analytic;
•enuire: (
Smile l.D.
) Laboratory Laboratory Supervisor
onfimation (See
Suspect
Original
Value
I) _ . Reanaly
Reconfirmed/
New
Value
:is (See II)
Explanation
Contract
Analytical
laboratory
LEKSCC
I. Confirmation Request: Did ANT values change: _ Yes _ No
If yes, reason (note above in explanation column):
(A) Reporting Error (C) Original reported value did not change
(6) Calculation Error ID) Data Previously Omitted
(E) Other - Explain
If values changed, submit supporting ran data AS REQUIRED.
Additional Comnents Regarding Confirmation:
II. Reanalysis Requested Due to:*
External OA Data
Internal OC Data Indicated Below:
• 1C Resolution
IDL > CRDL
Blank > 2 x CRDL (Reagent; Calibration)
QCCS Outside Criteria (OL; Low; High)
Sample Concentration Outside Calibration Range
QCCS Not In Mid-Range of Calibration Range
Duplicate Precision (X RSO) Outside Criteria; Insufficient Hunter of Duplicates
Analyzed
Additional Contents Regarding Reanalysis:
• An abbreviated version of NSUS Form II, 18, !». and 20 mist bt submitted for all reanalyzed data.
NSUS Form 13, II, and 22 must be submitted when applicable.
FOR LEI6CO USE ONLY: INITIAL REVIEW
VERIFICATION
OF VALUES SUBMITTED .
MMBER DF VALUES CHANGED _
Figure 13.2 National Surface Water Survey Form 28 - Data Conflrmatlon/Reanaly&l* Request
-------
14.0 DATA VALIDATION
14.1 Overview
Validation, in the context of data
bases, is the process by which data are
evaluated for quality consistent with the
intended use of the information. Because
validation is a process linked to the goals
and methods of a project, the process must
be defined for each data base. Con-
sequently, no single set of criteria can be
applied to all data bases to judge their
validity. Validation is, therefore, a func-
tional term for describing the continuing
process of defining the quality of the data
with each step resulting in increased know-
ledge of, and presumably confidence in,
the data. This is accomplished by reviewing
the data for errors; data known to be er-
roneous are identified so that correct data
can be substituted, and possible errors are
flagged to alert the user to their question-
able status.
In the verification step, which precedes
validation, the quality of the analytical
chemical data is determined through a rig-
orous protocol based on known principles of
chemistry. However, not all potential errors
in the data are evaluated in the verification
process. Verification scrutinizes the internal
consistency of chemical concentrations
within a sample; the validation process seeks
to determine the plausibility of sample
physical and chemical data in the context
of a subregional set of samples. Therefore,
the purpose of the validation process for
NSWS is to investigate errors in the chem-
ical analyses not detected in verification
and to provide a review of the quality of
the nonchemical variables. The list of some
physical variables subject to validation is
shown in Table 14.1. Two aspects of the
data validation process are the identification
of outliers and the evaluation of possible
systematic error in the measurement process.
Both of these aspects are exploratory, as
opposed to test-oriented, and as such, the
Section 14
Revision 4
Date: 9/86
Page 1 of 8
methods stress visual presentations and
subjective, though conservative, selection
procedures. The objective is to attract
attention to certain data values or sets of
values so that special thought and caution
will be applied to them during data analysis
and model building. The methods selected
for detection of outliers and systematic
errors were chosen for their simplicity of
implementation from a computational stand-
point and for the degree of their ability
to use pre-existing software.
Table 14.1. Somo Phyaleal Variables Subject
to Validation
Variable
General Description of
Validation Checks
1. Latitude
2. Longitude
3. Elevation
4. Watershed Area
5. Stream Inlets
and Outlets
6. Streambank Land
Use
Stream location is
compared to location
measured on USGS maps
Stream characteristics
are checked against
state records, where
available, to confirm
stream identification.
Data are compared to
aerial photographs.
7. Water Temperature Recorded temperature is
checked to see if it
falls in appropriate
range.
8. True Color
9. Turbidity
Data are checked for
internal consistency.
The techniques to be used in validating
the Mid-Atlantic Phase I data are essentially
the same as those used for the Phase I-
Pilot Survey. The major difference lies in
the type of special univariate techniques
to be employed. In the Phase I - Pilot
-------
Survey, univariaie fences were used to
evaluate individual streams; this was possible
because large numbers of observations were
available for each site. In the Mid-Atlantic
Phase I study, there is a maximum of four
routine samples per site, and this precludes
use of the fence technique. In place of
the fences, univariate ratio comparisons
are made. These ratios are computed as
follows:
1.
3.
Downstream-Upstream Observation
Upstream Observation
Downstream1-Downstream2 Observation
Downstream 1 Observation
Upstream1-Upstream2 Observation
Upstreaml Observation
For Southeast Screening Survey sites,
only ratio 1 is computed.
Following ratio computations, each
ratio is subjected to univariate analysis
to determine underlying distributions. Paired
t-tests are made to discern if there is a
significant difference between the test
subpopulations. If there is a significant
difference, the actual ratio values are in-
spected for high (and low) extremes, which
represent outliers. This technique identifies
outlier pairs in actuality, specific to th©
ratio formula employed. In practice, only
those values above 2X the system detection
limit or 1/2 the quantitation limit are passed
into these analyses.
The data are divided Into 10 subsets
that correspond to the discrete Mid-Atlantic
and Southeast Screening subregions (or parts
thereof). Each subset is evaluated indiv-
idually via principal component analysis
(PGA). (Missing values at this level are
substituted by subset means.) Muitivariate
suites are extracted from the scoping PGA
Section 14
Revision 4
Date: 9/86
Page 2 of 8
where at least 95 percent of the modeS
variance is displayed; i.e., the number of
principal components that demonstrate 95
percent of the cumulative variance of the
model is the cutoff number for multivariate
suites. The PCA cross correlation matrix
is used to define one bi- or multiple-linear
regression model for each variable. There-
fore, there are 38 suites for regression
analyses plus n suites for restricted PGA
and cluster analyses (in 20,14 PCA/GLUS
suites were defined).
14.2 Defection of Outliers
Outliers are defined as observations
that are not typical of the population from
which the sample is drawn. They are iden-
tified using univariate, bivariate, and multi-
variate analyses. These procedures assist
in identifying outliers that require further
scrutiny. However, observations that are
atypical with respect to the population may
result from analytical error or heterogeneity
in chemistry among streams. It is essential
to separate analytical errors from abnormal
stream chemistry to avoid the undesirable
effect of purging analytically correct values
from the data base (discussed in Section
14.4).
14.2.1 Frequency Analyses
A SAS procedure is used to produce a
1-, 2-, 3-, 4-, and 5-way frequency and
crosstabulation (PROC FREQ) in order to
determine completeness of the data set.
Outputs may be organized:
• by stream ID to determine duplicate
entries, entry errors, and missing values
• by batch ID, sample ID, and sample
code to determine duplicate entries,
invalid entries, and missing values
® by variable (e.g., mobile laboratory
pH measurement) to determine invalid
entries and missing values.
-------
1 4.2.2 Univariate Analyses
An initial approach to outlier detection is
to consider each variable individually and
to search for values that are extreme with
respect to the sample. The method used
here is the box plot (Tokay, 1077) as imp-
lemented in SAS (SAS Institute, 1982).
The box plot summarizes the data for one
variable based on the median and upper
(Fu) and lower (Fl) fourths or quartiles.
The difference between the upper and lower
quartiles is known as the inter-quartile
range (Fu - Fl = dF); any value greater
than the absolute value of 3 dF is identified
as an outlier.
Summary statistics with plots are used
to identify five high and low extreme values
to determine underlying distributions, to
flag extreme values, and to assess data
variability. These statistics are performed
on the entire data base, the spring data
set, the spring upper and lower node mea-
surements, the summer data measurements,
the summer upper and lower node measure
merits, and individual streams.
Univariate windows for each variable
determine "unusual" values for a given
stream. This method involves using by-
stream quartiles computed by SAS under
Definition No. 1 (SAS, 1985, p. 1136):
weighted average of xnp
y
where:
np
and
x
y
| + g
is taken as xt
a value for the variable
the tth percent!!© where j
is the fractional part of np.
Section 14
Revision 4
Date: 9/86
Page 3 of 8
The difference between the values for
the first and third quartiles is used to
compute inner and outer, upper and lower
windows for each stream and variable as
follows:
--—Q DIFF----
! !
01 Q3
II ! |..-Median---!-----I-----II
outer inner 25th 75th inner outer
tower percentite percentiie upper
window (first quart lie) (third quartile) window
where:
inner lower
outer lower
inner upper
outer upper
Q1 - (1.5 x Q DIFF)
Q1 - (3.0 x Q DIFF)
Q3 + (1.5 x Q DIFF)
Q3 + (3.0 x Q DIFF)
Following window computations, all
data are compared to their appropriate
windows. Ons data set is prepared that
contains the by-stream window statistics.
Other data sets are also prepared, one for
each inner and outer, lower and upper window
so that values that fail outside the inner,
inner and outer, lower and upper windows
are identified. These data are then screened
to determine whether they are "outliers"
(from the traditional definition) or episodes,
polluted, ate, Outliers are reported to the
EMSL-LV QA staff for further verification.
14,2,3 Principal Components Analysis
The objective of this analysis is to
determine multivariate associations to be used
in establishing bivariate, multiple linear
regression, and sets of multivariate statistical
tests for advanced statistical analysis.
Relationships among variables are presented
in Table 14.2.
-------
Section 14
Revision 4
Date: 9/86
Page 4 of
Table 14.2. Pairs of Variables Used to Check for Random and Systematic Errors
AND
vs.
Aluminum (total)
Calcium
vs.
vs.
Chloride
Specific conductance
vs.
vs.
Aluminum (organic ext.)
vs.
Potassium
Ammonium
Silica
Turbidity
pH (mobile processing lab}
OIC (mobile processing lab)
vs.
vs.
vs.
vs.
vs.
vs.
Calcium
Specific conductance
Magnesium
Silica
pH (Mobile processing lab)
Ammonium
Turbidity
True color
Specific conductance
Fluoride (total dissolved)
Sulfate
Silica
Specific conductance
Sodium
Fluoride (total dissolved)
Potassium
Magnesium
Sodium
Silica
Sulfate
Potassium
Magnesium
Silica
Aluminum (total ext.)
Magnesium
Turbidity
True color
BNC
Mobile processing lab pH
Magnesium
True color
pH (initial and air equilibrated)
DIC (initial and air equilibrated)
-------
14.2.4 Bivariate Analyses
Although values of two variables may
not be outliers within their respective uni-
variate distributions, the pair may be con-
sidered extreme relative to some expected
or typical relationship. Scatter plots are
useful for examining expected theoretical
or empirical relationships between variables.
The bivariate relationships examined in this
process are shown in Table 14.2. Outliers
are identified by visual inspection of the
plots and by listing of residuals based on
a least-squares regression analysis where a
linear relationship exists.
Observations are identified as outliers
if the absolute value of the standardized
residual [(actual-predicted)/residual standard
deviation] is generally greater than 3. Be-
cause the least-squares analysis can be
strongly biased by certain types of outliers,
the residuals from resistant line fits, lines
fit through the medians of partitions of
the data, are examined for DOC, true color,
and turbidity (Velleman and Hoaglin, 1981).
Other variables are treated by use of an
iterative process of linear regression, iden-
tification and removal of outliers, and re-
peated linear regression to identify addi-
tional outliers that would not have necess-
arily been identified had major outliers not
been removed first.
14.2.5 Multivariate Analyses
Although examination of scatter plots is an
important and necessary step for evaluating
possible errors in the data, bivariate ana-
lyses must be limited to those variables that
have obvious associations. The magnitude
of the data set precludes examination of all
possible distributions and bivariate plots.
For example, the number of bivariate plots
required for all combinations of the analy-
tical variables exceeds 4,600. Although many
of these combinations of variables are of
no interest, many combinations remain.
Section 14
Revision 4
Date: 9/86
Page 5 of 8
Clearly, this is not a practical or efficient
method for examining all the data.
An alternative method of examining
data for systematic and random errors is
through multivariate analysis in which sev-
eral variables are examined simultaneously.
Because theoretical relationships are ex-
pected to exist among certain chemical
variables, it is useful to examine these sets
of variables as groups (Table 14.3).
Table 14.3. Related Groups of Variables Used In
Multivariate Analyses
Group
Variables
1. Major anions and cations
2. pH, ANC, DOC, true color
3. Turbidity, true color
4. Nitrate, phosphorus, ammonium, turbidity
5. Anion deficit, DOC, true color
6. pH; total extractable, organic extractable,
and total aluminum; fluoride; DOC
7. Silica, major cations
8. Iron, manganese, total extractable and
organic extractable aluminum, DOC
9. ANC, DIG, pH
10. pH, sulfate, DOC
Two primary multivariate techniques
are used to identify outliers: cluster ana-
lysis and principal component analysis (PCA).
Cluster analysis is a classification technique
for identifying similarities (or, conversely,
dissimilarities) among observations. Each
observation is compared to other observa-
tions in the set and is assigned to a group
or cluster using a measure of similarity.
The primary clustering technique used
in the validation process is the FASTCLUS
procedure in SAS (SAS Institute, 1982).
This method is a non-hierarchical divisive
method that is sensitive to outliers. A
less formal clustering technique also used
for selected samples is the Trilinear Diagram
-------
(Hem, 1970). The Trilinear Diagram is useful
for examination of possible errors associated
with the major cations and anions. The
other clustering techniques are used for
related sets of variables such as those shown
in Table 14.3.
Principal component analysis is ' a
technique that also is commonly used to
reduce large data matrices into manageable
dimensions. New variables called principal
components are formed from linear combi-
nations of the original variables such that
the first principal component reflects most
of the variance or dispersion in the data.
Each successive principal component ex-
plains less variance, and examination of
the first several components is generally
sufficient to describe the data. If the
original data are approximately normally
distributed, the resulting principal compon-
ents are also approximately normal. Thus,
a plot of any two components typically
results in an elliptical cluster with outliers
displaced from the ellipse.
Where appropriate, least-squares mul-
tiple linear regression techniques also are
used to identify observations with high
absolute values of the standardized residual.
14.3 Detection of Systematic Error
Methods for evaluating systematic error
are less exploratory because they require a
source of external comparison. Here the
tests are similar to comparison with stan-
dards (such as audits or split samples), with
one major difference. The external refer-
ences consist of data sets obtained from
other investigators and cannot be viewed
as "standards." Hence, a difference
between data from NSS and another data
source does not necessarily imply that the
NSS data are in error. However, compari-
sons with external data sources serve as
aids for evaluating the quality of the data
by bringing attention to data that may
require additional scrutiny. Clearly, existence
Section 14
Revision 4
Date: 9/86
Page 6 of 8
of systematic differences between the NSS
data and several external data sources would
be cause for careful reevaluation of the
data in question. Two types of systematic
errors are investigated in the NSS data
base: a constant additive effect (resulting
in a nonzero intercept) and an effect that
is dependent on the magnitude of the variable
being measured (resulting in a slope / 1 or
nonlinearity in the relationship).
14.4 Treatment of Outliers
and Systematic Differences
Data identified as outliers through
the procedures described above may be
acceptable when evaluated in the context
of other variables or when considering
limitations of the methods used in NSWS.
Therefore, before the original data sources
are rechecked, the outliers and systematic
differences identified in the validating process
are reviewed for plausibility by the staff
at ERL-C. Data that remain suspect follow-
ing screening by staff scientists are sent
to the appropriate organization for reexa-
mination.
Outliers and systematic differences
for all chemical variables are checked
against reported values by staff at EMSL-LV;
site location and watershed related variables
are reviewed by the Geographic Research
Team at ERL-C; and remaining variables
are checked by staff at ORNL. When the
data are rechecked for suspect values that
were identified during validation of the
chemical variables, the following possible
conditions may be revealed. These cond-
itions may require the associated response
listed:
Condition
(1) Suspect value in data set number 2
(verified data set) is found to be a
transcription or transposition error.
-------
Response
(1) Correct value is placed in data set
number 3 (validated data set).
Condition
(2) Suspect value in data set number 2
agrees with reported value, and value
was flagged in verification.
Response
(2) Value is flagged in data set number 3.
Condition
(3) Suspect value in data set number 2
agrees with reported value, but value
was not flagged in verification.
Response
(3) Value may be flagged in data set
number 3 depending on evidence for
possible error.
Values flagged in data set number 2
but not identified as aberrant in data vali-
dation remain unchanged and flagged except
in cases where the flag is not required for
interpretation of the data; in these cases,
the flag is removed. The protocol for reso-
lution of outliers for the non-chemical
variables is similar, with the exception that
response (2) is omitted.
Resolution of systematic differences
between NSS and external reference data
involves reexamination of the methods used
to collect the NSS data. The effort involved
in evaluating systematic differences depends
on the evidence available to suggest that a
bias may exist in the NSS data and the
variable under consideration.
In most cases, sufficient information
to perform an appropriate correction for bias
in the NSS data is not likely to be available.
Section 14
Revision 4
Date: 9/86
Page 7 of 8
However, the identification of a possible
bias is provided to assist the user in inter-
preting such data. The process of validation
is summarized in Figure 14.1.
-------
Section 14
Revision 4
Date: 9/86
Page 8 of 8
(Data Set No.
v Verified
UNIVARIATE
*Box Plots
*Probability Plots
MULTIYARIATE
*PCA
*Cluster Analysis
*Trilinear Plots
*MLR
Outliers
BIVARIATE
*Scatter Plots
*Regression
Modify or\ Yes
J<
Delete Value>
Relational
i
Comparative
'
System
Differ
fFlag or \
(Modify Values/-
Yes
.
Data Set
No. 3
Validated
Flguro 14.1. Flowchart for data validation proceM.
-------
Section 15.0
Revision 4
Date: 9/86
Page 1 of 3
15.0 REFERENCES
American Public Health Association, American Water Works Association, and Water Pollution
Control Federation, 1985. Standard Methods for the Examination of Water and
Wastewater, 16th Ed. APHA, Washington, D.C.
American Society for Testing and Materials, 1984. Annual Book of ASTM Standards, Vol.
11.01, Standard Specification for Reagent Water, 01193-77 (reapproved 1983). ASTM,
Philadelphia, Pennsylvania.
Chaloud, D. J., L J. Arent, B. B. Dickes, J. 0. Nitterauer, M. D. Morison, and D. V. Peck,
in preparation. National Surface Water Survey, Eastern Lake Survey - Phase II,
National Stream Survey - Phase I Processing Laboratory Training and Operations
Manual. Internal Report. U.S. Environmental Protection Agency, Las Vegas, Nevada.
Costle, D. M., May 30, 1979(a). Administrator's Memorandum, EPA Quality Assurance Policy
Statement. U.S. Environmental Protection Agency, Washington, D.C.
Costle, D. M., June 14, 1979(b). Administrator's Policy Statement, Quality Assurance
Requirements for All EPA Extramural Projects Involving Environmental Measurements.
U.S. Environmental Protection Agency, Washington, D.C.
Grubbs, F. E., 1969. Procedures for Detecting Outlying Observations in Samples. Tech-
nometrics, TCMTA, v. 11, n. 4, pp. 1-21.
Hagley, C. H., C. M. Knapp, C. L Mayer, and F. A Morris, 1986. National Surface Water
Survey - Stream Survey (Middle-Atlantic Phase I, Southeast Screening, and Middle-
Atlantic Episode Pilot). Field Training and Operations Manual. Internal report.
U.S. Environmental Protection Agency, Las Vegas, Nevada.
Hem, J. D., 1970. Study and Interpretation of the Chemical Characteristics of Natural
Water, 2nd Ed. U.S. Geological Survey Water Supply Paper 1473. U.S.G.S., Washing-
ton, D.C.
Hillman, D. C., S. H. Pia, and S. J. Simon, 1986. National Surface Water Survey * Stream
Survey (Pilot, Middle-Atlantic Phase I, Southeast Screening, and Middle-Atlantic
Episode Pilot). Analytical Methods Manual. U.S. Environmental Protection Agency,
Las Vegas, Nevada.
Kramer, J. R., 1982. Alkalinity and Acidity. In: R. A. Minear and L.H. Keith (eds.),
Water Analysis, Vol. 1. Inorganic Species, Part 1. Academic Press, Orlando, Florida.
McQuaker, N. R., P. D. Kluckner, and D. K Sandberg, 1983. Chemical Analysis of Acid
Precipitation: pH and Acidity Determinations. Environ. Sci. Technol., v. 17, n. 7, pp.
431-435.
-------
Section 15.0
Revision 4
Date: 9/86
Page 2 of 3
Messer, J. J., C. W. Ariss, J. R. Baker, S. K. Drouse, K. N. Eshleman, P. R. Kaufmann, R.
A Linthurst. J. M. Omernik, W. S. Overton, M. J. Sale, R. D. Schonbrod, S. M.
Stambaugh, and J. R. Tuschall, Jr., 1886. National Surface Water Survey National
Stream Survey Phase I - Pilot Survey: A Contribution to the National Acid Preci-
pitation Assessment Program. EPA 600/4-86/026. U. S. Environmental Protection
Agency, Office of Research and Development, Washington, D.C., Environmental Re-
search Laboratory, Corvaiiis, Oregon, and Environmental Monitoring Systems Labor-
atory, Las Vegas, Nevada.
Oak Ridge National Laboratory, 1984. National Surface Water Survey Project - Data Man-
agement Proposal. Environmental Sciences Division and Computer Sciences, UCC-ND,
ORNL, Oak Ridge, Tennessee.
Oliver, B. G., E. M. Thurman, and R. K. Malcolm, 1983. The Contribution of Humic Sub-
stances to the Acidity of Colored Natural Waters. Geochlm. Cosmochim. Acta, v. 47,
pp. 2031-2035.
Omernik, J. M., and C. F. Powers, 1983. Total Alkalinity of Surface Waters - a National
Map. Ann. Assoc. Amer. Geographers, v. 73, pp. 133-136.
Overton, S., 1985. Draft Sampling Plan for Phase I of the NSS. U.S. Environmental Pro-
tection Agency, Corvallis, Oregon.
Peden, M. E., 1981. Sampling Analytical and Quality Assurance Protocols for the National
Atmospheric Deposition Program. ASTM D-22 Symposium and Workshop on Sampling
and Analysis of Rain, American Society for Testing and Materials, Philadelphia,
Pennsylvania.
Sale, M. J., J. M. Coe, M. I. Jager, and M. A, Faulkner, in preparation. Data Management
and Analysis Procedures for the National Stream Survey. To be published January
1987, is an ORNL Technical Mamorandum, Oak Ridge National Laboratory, Oak Ridge,
Tennessee.
SAS Institute, Inc., 1985. SAS User's Guide: Basic®, Version 5, Ed, SAS, Cary, North
Carolina.
SAS Institute, inc., 1932. SAS User's Guide: Statistics. SAS, Cary, North Carolina.
Tukey, J. W., 1977. Exploratory Data Analysis. Addison-Wesley Publishing, Reading,
Massachusetts.
U.S. Environmental Protection Agency, 1980. Interim Guidelines and Specifications for
Preparing Quality Assurance Project Plans. QAMS-005/80. U.S. Environmental
Protection Agency, Washington, D.C.
U.S. Environmental Protection Agency, 1983 (revised). Methods for Chemical Analysis of
Water and Wastes. EPA-600/4-79-020. U.S. EPA, Cincinnati, Ohio.
-------
Section 15.0
Revision 4
Date: 9/86
Page 3 of 3
U.S. Environmental Protection Agency, 1984(a). National Surface Water Survey, Phase I.
U.S. EPA, Office of Research and Development, Washington, D.C.
U.S. Environmental Protection Agency, 1984(b). National Surface Water Survey, Phase I.
Research Plan, A Summary of Contents. U.S. EPA, Corvallis, Oregon.
U.S. Environmental Protection Agency, 1985. National Surface Water Survey: National
Stream Survey - Mid-Atlantic Phase I and Southern Screening Draft Research Plan.
U.S. EPA, Washington, D.C. Velleman, P. F. and D. C. Hoaglin, 1981. Applications,
Basics, and Computing of Exploratory Data Analysis. Duxbury Press, Boston, Massa-
chusetts.
Weast, R. C. (ed.), 1972. CRC Handbook of Chemistry and Physics, 53rd Ed., CRC Press,
Cleveland, Ohio.
-------
Appendix A
Revision 4
Date: 9/86
Page 1 of 16
APPENDIX A
Data Forms for Reporting Analytical Results
LAB NAME
BATCH ID
NATIONAL SURFACE WATER SURVEY
FORM 11*
SUMMARY OF SAMPLE RESULTS
*.
LAB MANAGER'S SIGNATURE
Page 1 of 2
SAM-
PLE
ID:
01
_°1_J
03
04
OS
66
07
08
09
10
11
12
13
JlT™r"1
15
16
17
18
19
20
~2T
22 ""
23
24
" 25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
tote : I
AtllJUDT ID
1
Ca
rcg/L
Mg
mg/L
Approved data quail f
K
mg/L
Na
mg/L
iers and instruct!
Mn
mg/L
ons for t
Fe
mg/L
2
Total
Extr. Al
mg/L
3
cr
mg/L
sof
mg/L
NO,"
mg/L
Si02
mg/L
ISE
Total F"
mg/L
heir use are listed in Table 9.8 of the QA plan.
aThis form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening Survey, and the Episodes Pilot Survey.
For the Phase I - Pilot Survey, the form includes a column for aliquot 8.
-------
Appendix A
Revision 4
Date: 9/86
Page 2 of 16
NATIONAL SURFACE WATER SURVEY
FORM lla
Page 2 of 2
SUMMARY OF SAMPLE RESULTS
LAB NAME
BATCH ID
LAB MANAGER'S SIGNATURE
SAM-
PLE
ID:
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35 '
36
"37 """
38
39
40
Note:
ALIQUOT ID ' ' ~ " " •
4
DOC
mg/L
NH4+
mg/L
5
Measured
Eq.
pH
Alk
Init, pH
Acy
Inft. pH
ANC
ueq/L
Approved data qua! if ers and Instructions for 1
BNC
Meq/L
Spec.
Cond.
uS/cm
Eq.
DIC
mg/L
Init.
DIC
mg/L
6
Total
Dissolved
P
mg/L
r '
Total
Al
mg/L
-heir use are listed in Table 9.8 of the QA plan.
aThis form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening Survey, and the Episodes
Pilot Survey. For The Phase I - Pilot Survey, the form includes a column for aliquot 8.
-------
Appendix A
Revision 4
Date: 9/86
Page 3 of 16
Lab Name
NATIONAL SURFACE WATER SURVEY
Form 13
ANC AND BNC RESULTS
Batch ID
Page 1 of 1
Lab Manager's Signature
Analyst
Sample ID
RESULTS
[ANC] =
[BNC] =
DATA
CB =
peq/L
peq/L
eq/L
eq/L
INITIAL SAMPLE VOLUME
BLANK ANC
mL
Meq/L
DATE STANDARDIZED
DATE STANDARDIZED
ACID TITRATION
BASE TITRATION
VOLUME HC1
(mL)
0.00
0.00 (with KC1) j
MEASURED
pH1
CALCULATED
PH
VOLUME NaOH
(mL)
0.00
0.00 (with KC1)
MEASURED
PH1
"CALCULATED
pH
-------
NATIONAL SURFACE WATER SURVEY
Form 14a
QC DATA FOR ANC
AND BNC ANALYSES
Appendix A
Revision 4
Date: 9/86
Page 4 of 16
Page 1 of 1
LAB NAME
BATCH ID
LAB MANAGER'S SIGNATURE
SAMPLE
ID
01
02
03
04
05
06
07
08
09
10
" 11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
ANC
ueq/L
BNC
peq/L
CALCULATED ANC
RESULT
DIFFERENCE15
3DC
aForm not required in data package but recommended for internal QC requirements.
boifference = Calculated ANC-Measured ANC
cRefer to methods manual
-------
Appendix A
Revision 4
Date: 9/86
Page 5 of 16
LAB NAME
NATIONAL SURFACE WATER SURVEY
Form 15a
SPECIFIC CONDUCTANCE
BATCH ID LAB MANAGER'S SIGNATURE
Page 1 of 1
Sample
ID
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
SPECIFIC
CONDUCTANCE (yS/cm)
Calculated
Measured
J!CDU
Specific Conductance
Factors of Ions
[(uS/cm at 25 °O per mg/L]
CALCULATED SPECIFIC CONDUCTANCE FOR EACH ION • us/cm" - - • - •
HCOj
0.715
Ca+2
2.60
Note: Reanalysis criteria are given 1n Table 9.5 o1
C03"2
2.82
cr
2.14
M9+2
3.82
N03"
1.15
K+
1.84
Na+
2.13
so4-'
1.54
NH4*
4.13
H*
3.5xlOs
(per
mole/L)
OH~
1.92x105
(per
mole/L)
the QA plan. "" ' " " ~
aForm not required in data package but recommended for internal QC requirements.
Calculated Specific Conductance + Measured Specific Conductance
Measured Specific Conductance
x 100
-------
Appendix A
Revision 4
Date: 9/86
Page 6 of 16
LAB NAME
NATIONAL SURFACE WATER SURVEY
Form 16a
ANION-CATION BALANCE CALCULATION
BATCH ID LAB MANAGER'S SIGNATURE
Page 1 of 1
Sample
ID
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
23
2$
3D'
1 "31
32
33
34
"35
36
37
38
3$
40
% Ion
Difference^
Factors to Convert
mg/L to peq/L
Note: Reanalysis crite
aForra not required in d
b% Ion Difference UID)
C[H+] = (10-PH) x 106
Ions - (ueq/L)
Ca+2
49.9
cr
28.2
Mg+2
82.3
N03-
16.1
K+
25.6
Na+
43.5
S04"2
20.8
-fa are given in Tab e 9.5 of the QA p an,
ata package but recommended for internal QC r
ANC + r Anions - £ Cations (except H*)
ANC + E Anions +• l Cation + 2[H+]
F-
52.6
NH4"1"
55.4
ANC
....
H+c
_.«_
equirements.
100
-------
Appendix A
Revision 4
Date: 9/86
Page 7 of 16
NATIONAL SURFACE WATER SURVEY
Form 17 Page 1 of 1
1C RESOLUTION TEST
LAB NAME
BATCH 10
LAB MANAGER'S SIGNATURE
1C Resolution Test
1C Make and Model:
Date :
Concentration: S042" ug/mL, N03 ug/mL
Column Back Pressure (at max. of stroke): psi
Flow Rate: mL/min
Column Model: Date of Purchase:
Column Manufacturer:
Column Serial No:
Is precolumn in system Yes No
(a) cm (b) cm
Percentage Resolution: 100 x (1-a/b)
The resolution must be greater than 60%
Test Chromatogram:
so
N03-
-------
LAB NAME
NATIONAL SURFACE WATER SURVEY
Form 18a
DETECTION LIMITS
BATCH ID
Appendix A
Revision 4
Date: 9/86
Page 8 of 16
Page 1 of 1
LAB MANAGER'S SIGNATURE
Parameter
Ca
Mg
K
Na
Mn
Fe
Al, Total
Extractable
cr
so42"
N03~
Si02
F-, Total
Dissolved
NH4+
DOC
Specific
Conductance
DIC
P, Total
Dissolved
Al, Total
Mote 1: Report with
Units
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
.uS/cm
mg/L
mg/L
mg/L
Required
Detection
Limit
0.01
0.01
0.01
0.01
0.01
0.01
0.005
0.01
0.05
0.005
0.05
0.005
0.01
0.1
c
0.05
0.002
0.005
Instrumental
Detection
Limitb
Date Determined
(DO MMM YY)
four significant figures or down to IDL
Note 2: Indicate the instrument for which the IDL applies with an "F" (for Fur-
nace AA), a "P" (for ICP) or an "L" (for Flame AA) after the IDL value.
aThis form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening
Survey, and the Episodes Pilot Survey. For the Phase I - Pilot Survey, the
form includes a column for aliquot 8.
"To be calculated as required in Section 9.6 of the QA plan and filled out by
the analyti£al laboratory.
cReport the X, which must not exceed 0.9 uS/cm, of six (6) nonconsecutive blanks.
-------
Appendix A
Revision 4
Date: 9/86
Page 9 of 16
LAB NAME
DATE SAMPLEOb
BATCH ID
DATE RECEIVEDb
NATIONAL SURFACE WATER SURVEY
FORM 19 a
SAMPLE HOLDING TIME SUMMARY
LAB MANAGER'S SIGNATURE
Page 1 of 2
Parameter
Holding
Time
Holding Time
Plus Date
Processed
Sample ID:
01
02
03
04
05
d<5
07
•08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
"24
25
26
27
28
29
30
31
32
'33 '
34
3S ' "
36
37
38
39
40
Ca
28
Mg
28
K
28
Na
28
Mn
28
Dateb
Fe
28
Total
Extr. Al
7
cr
28
S042
28
N03
7
Analyzed0
Si02
28
ISE
Total F-
28
aThis form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening Survey, and the Episodes Pilot
Survey. For the Phase I - Pilot Survey, the form included a column for aliquot 8.
bReport these dates as Julian dates (i.e., March 26, 1984 » 4086).
clf parameter was reanalyzed because of QA problems, report the last date analyzed.
-------
Appendix A
Revision 4
Date: 9/86
Page 10 of 16
LAB NAME
BATCH ID
NATIONAL SURFACE WATER SURVEY
FORM 19a
SAMPLE HOLDING TIME SUMMARY
LAB MANAGER'S SIGNATURE
Page 2 of 2
DATE SAMPLED13
DATE RECEIVEDb
'arameter
folding
Time
Holding Time
Plus Date
Processed
Sample ID:
"01
'tit
•03
04
05
06
"07
08
09
10
' 11
12
13
14 ' " '
15
16 " '
17
18
19 " '
' 20
21
22
23
24
25
26
27
28
29
30 "
31
32
33
34
35
36
37
38
39
40
DOC
14
NH4+
28
Eq. pH
14
ANC
14
BNC
14
Specific
Conductance
14
Eq. DIC
14
Iiiit. DIC
14
Total
Dissolved
P
28
Total Al
28
Dateb Analyzedc
aThis form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening Survey, and the Episodes Pilot
Survey. For the Phase I - Pilot Survey, the form includes a column for aliquot 8.
^Report these dates as Julian dates (i.e., March 26, 1984 = 4086).
=If parameter was reanalyzed because of OA problems, report the last date analyzed.
-------
Appendix A
Revision 4
Date: 9/86
11 of 16
NATIONAL SURFACE WATER SURVEY
FORM 20a
BLANKS AND QCCS
Page 1 of 2
LAB NAME BATCH ID LAB MANAGER'S SIGNATURE
Parameter
Calibration
Blank
Reagent Blank
DL [Theoretical
QCCSJMeasured
LOW ices
True Value
Low QCCS Upper
Control Limit
Low QCCS Lower
Control Limit
Initial
Continuing
Continuing
Continuing
Continuing
Continuing
Final
High QCCS
True Value
High QCCS Upper
Control Limit
High QCCS Lpwer
Control Limit
Initial
Continuing
Continuing
Continuing
Continuing
Continuing
Final
ALIQUOT ID
1
Ca
mg/L
• IT "
Mg
mg/L
N
K
mg/L
" N
Ma
mg/L
N
lote: Approved data qualifiers and instruction 1
Mn
mg/L
N
Fe
mg/L
N
2
Total
Extr.Al
mg/L
N
3
cr
mg/L
N
or their use are listed in Tab!
SO/-
mg/L
N
e 9.8 o
NO,
mg7L
N
Si02
mg/C
ISE
Total F"
mg/L
N
N
N
f the QA plan.
*This form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening Survey, and the Episodes Pilot
Survey. For the Phase I - Pilot Survey, the form includes a column for aliquot 8.
-------
Appendix A
Revision 4
Date: 9/86
Page 12 of 16
LAB NAME
NATIONAL SURFACE WATER SURVEY
FORM 20a
BLANKS AND QCCS
BATCH ID LAB MANAGER'S SIGNATURE
Page 2 of 2
Parameter
Calibration
Blank
Reagent Blank
OL theoretical
QCCS measured
Low QCCS
True Value
Low QCCS Upper
Control Limit
Low QCCS Lower
Control Limit
Initial
Continuing
Continuing
Continuing
Continuing
Continuing
Final
High QCCS
True Value
High QCCS Upper
Control Limit
High OCCS Lower
Control Limit
Initial
Continuing
Continuing
Continuina
Continuing
Continuing
Final
4
DOC
mg/L
N
NH/
mg/L
N
' ALIQUOT ID "
Measured
Eq
PH
N
N
N
N
ANC
pH
N
N
N
N
BNC
pH
N
N
N
N
tote: Approved data qualifiers and instruction for
of the QA plan.
5
Spec.
Cond.
uS/cm
N
N
N
Eq.
DIC
mg/L
N
Init.
OIC
mg/L
N
6
Total
Dissolved
P
mg/L
N
7
Total
Al
mg/L
heir use are listed in Table 9.8
aThis form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening Survey,
and the Episodes Pilot Survey. For the Phase I - Pilot Survey, the form includes a
column for aliquot 8.
-------
Appendix A
Revision 4
Date: 9/86
Page 13 of 16
LAB NAME
BATCH ID
NATIONAL SURFACE WATER SURVEY
FORM 21a«b
DILUTION FACTORS
LAB MANAGER'S SIGNATURE
Page 1 of 2
SAM-
PLE
10:
61 '
02
03
04
06
'66 '
W '
08
09
10
il
12
13
14
15
'16
17
18
19
20
21
22
23
24
"25 "
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
ALIQUOT ID
1
Ca
Mg
K
Ha
Mn
Fe
2
Total
Extr. Al
3
CT
so,2'
NO,"
SiO?
ISE
Total F~
tote: Indicate samples analyzed on hfgher concentration range by using a check mark for each parameter.
aForm not required in the data package but recommended for QA purposes.
bThis form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening Survey, and the Episodes Pilot Survey.
For the Phase I - Pilot Survey, the form includes a column for aliquot 8.
-------
Appendix A
Revision 4
Date: 9/86
Page 14 of 16
NATIONAL SURFACE WATER SURVEY
FORK 21«.b
Page 2 of 2
DILUTION FACTORS
LAB NAME
BATCH ID
LAB MANAGER'S SIGNATURE
SAM-
PLE
ID:
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
AL1CUOT ID
4
DOC
NH/
Mote: Indicate sampl
parameter.
5
Measured
Eq.
pH
Alk
Init. pH
Acy
Init. pH
_
ANC
BNC
Com).
Eq.
DIC
Init.
DIC
es analyzed on higher concentration range by using a chec*
6
Total
Dissolved
P
7
Total
Al
mark for each
not required in the data package but recommended for QA purposes.
°This form is used for the Mid-Atlantic Phase I Survey, the Southeast Screening Survey
Episodes Pilot Survey. For the Phase I - Pilot Survey
the form includes a column
and the
for aliquot
-------
Appendix A
Revision 4
Date: 9/86
Page 15 of 16
LAB NAME
NATIONAL SURFACE WATER SURVEY
Form 22a
Page 1 of 2
DUPLICATES
BATCH ID
LAB MANAGER'S SIGNATURE
Parameter
Oupl icate
Sample ID
Sample Result
Duplicate
Result
t RSD
Second Duplicate
Sample ID
Sample Result
Duplicate
Result
* RSD
Third Duplicate
Sample ID
Sample Result
Duplicate
Result
Z RSD
ALIQUOT ID
1
Ca
mg/L
Mg
mg/L
K
mg/L
Na
mg/L
Mn
mg/L
tote: Approved Data Qualifiers and instructions for thei
Fe
mg/L
2
Total
Extr.Al
mg/L
r use are listed
3
CT
mg/L
in Tab!
S042-
mg/L
e 9.8 of
NOT
mg/L
Si02
mg/L
1SE
Total F"
mg/L
the QA plan.
aThis form Is used for tne Mid-Atlantic Phase I Survey, the Southeast Screening Survey, and the Episodes Pilot
Survey. For the Phase I - Pilot Survey, the form includes a column for aliquot 8.
-------
Appendix A
Revision 4
Date: 9/86
Page 16 of 16
LAB NAME
BATCH ID
NATIONAL SURFACE WATER SURVEY
Form 22a
DUPLICATES
LAB MANAGER'S SIGNATURE
Page 2 of 2
Parameter
Duplicate
Sample ID
Sample Result
Duplicate
Result
% RSDb
Second
Duplicate
Sample ID
Sample Result
Duplicate
Result
Z RSDb
Third Duplicate
Sample ID
Sample Result
Duplicate
Result
Z RSDb
•" ' ALIQUOT ID ' '
4
DOC
mg/L
NH/
mg/L
Measured
Eq.
PH
Alk
Initial
PH
Acy
Initial
PH
5
ANC
peq/L
BNC
ueq/L
Spec.
Cond.
US/cm
Eq.
DIG
ma/L
Init.
DIC
mg/L
6
Total
Dissolved
P
ma/L
7
Total
Al
mg/L
Note: Approved Data Qualifiers and instructions for their use are listed in Table 9.8 of the
QA plan.
aThis form is used for tne Mid-Atlantic Phase I Survey, the Southeast Screening Survey, and the
Episodes Pilot Survey. For the Phase I - Pilot Survey, the form includes a column for aliquot
"Report the absolute difference instead of %RSD for pH determinations.
-------
Appendix B
Revision 4
Date: 9/86
Page 1 of 17
APPENDIX B
FIELD SAMPLING AND MOBILE PROCESSING LABORATORY
ON-SITE EVALUATION QUESTIONNAIRE
GENERAL (Page 1 of 1)
Questionnaire Completion Date
Field Base
Location
Mobile Processing Laboratory Supervisor
Questionnaire Completed By (If more than one auditor, indicate sections
completed by each auditor.)
-------
MOBILE PROCESSING LABORATORY PERSONNEL (Page 1 of 1}
Position
Mobile Processing Laboratory
Coordinator
Mobile Processing Laboratory
Supervisor
Analyst
Analyst
Analyst
Name
Academic
Training*
Special
Training*
Years
Experience**
*List highest degree obtained and specialty. Also list years toward a degree.
**List only experience directly relevant to task to be performed.
Ql en CD ^j
(Q £ <.T3
CD ® o> (D
IN)
-------
Appendix B
Revision 4
Date: 9/86
Page 3 of 17
MOBILE PROCESSING LABORATORY - STANDARD OPERATING PROCEDURES (Paye 1 of 1)
Item
Is the training manual followed in detail?
Are copies available to the personnel?
Are analysis logbooks kept up to date?
Are all on-site changes in procedures clearly documented and
justified in mobile processing laboratory supervisor's logbook and
approved by appropriate personnel?
Yes
No
Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 4 of 17
MOBILE PROCESSING LABORATORY FACILITIES (Page 1 of 1)
Item
Is the laboratory kept clean and organized?
Are the refrigerator and freezer temperatures
monitored on a daily basis and recorded in a
logbook?
Are waste disposal containers available and
clearly labeled, and is waste disposed of
properly?
Are chemicals stored properly?
Is balance calibration checked daily and
recorded in a logbook?
Is water supply purity monitored daily and
recorded in a logbook?
Yes
No
Comment
General Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 5 of 17
MOBILE PROCESSING LABORATORY EQUIPMENT (Page 1 of 7)
Color Test Kits
Item
Is manufacturer's operating manual readily
available?
Is kit cleaned and stored properly?
Are viewing tubes kept clean?
Is logbook kept up to date and signed daily?
Is centrifuge maintained and kept clean?
Yes
No
Comment
Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 6 of 17
MOBILE PROCESSING LABORATORY EQUIPMENT (Page 2 of 7)
Nephelometer
Item
Is manufacturer's operating manual available?
Is instrument kept clean?
Are cuvettes kept clean and scratch-free?
Is logbook kept up to date and signed daily?
Is calibration checked before and after every
eight samples?
Are standards kept refrigerated when not in use?
Yes
No
Comment
Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 7 of 17
MOBILE PROCESSING LABORATORY EQUIPMENT (Page 3 of 7)
Carbon Analyzer
Item
Is manufacturer's operating manual available?
Is instrument kept clean?
Is the injection valve flushed with deionized
water daily after use? j
Is logbook kept up to date and signed daily?
Is IR analyzer power left on at all times?
Is standard stock solution prepared biweekly,
and is QC stock solution prepared weekly;
are they stored at 4°C?
Are working standards prepared daily?
Is exposure of samples and standards to the
atmosphere minimized?
Is required QC followed?
Are pump tubes checked for wear and
rejDjaced on a regular basis_Jjibout every 2 weeks)?
Are syringes and glassware cleaned
properly_after use?
Are C0£ and moisture scrubbers on
standard bottles replaced when exhausted?
Is tin scrubber in IR analyzer checked
dailjf and refilled when necessary?
Yes
No
Comment
Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 8 of 17
MOBILE PROCESSING LABORATORY EQUIPMENT (Page 4 of 7)
pH Apparatus
Item
Are meter and electrode operating manuals
available?
Is logbook kept up to date and signed daily?
Is pH QC sample prepared daily?
Is electrode stored in 3M KC1?
Is required QC followed?
Are electrodes checked and filled (if
necessary) prior to use?
Are sample chambers cleaned after use?
Are buffers capped tightly after use?
Yes
No
Comment
Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 9 of 17
MOBILE PROCESSING LABORATORY EQUIPMENT (Page 5 of 7)
Filtration and Preservation Apparatus
Item
Is hood kept neat and clean?
Is contamination evident?
Is hood sealed when not in use?
Is filtration apparatus kept ultraclean as
specified?
Are precautions taken to prevent contamination
of filtrators, filter funnels, filters, sample
bottles, and reagents?
Is a water trap used with the vacuum pump?
Are micropipets kept in an upright position
at all times?
Is the calibration of micropipets checked
weekly?
Are sample aliquots properly labeled?
Is vacuum maintained at 10 to 12 inches Hg while
filtering?
Yes
No
Comment
Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 10 of 17
MOBILE PROCESSING LABORATORY EQUIPMENT (Page 6 of 7)
MIBK Extraction
Item
Is the centrifuge operating manual available?
Is the extraction logbook kept up to date and
signed daily?
Is leakage of sample volume (> 8.5 mL) noted in
the logbook?
Are reagents (NaOAc and hydroxyquinoline) made
fresh daily?
Is NffyOH made fresh weekly and is the preparation
recorded in the logbook?
Are pipets calibrated weekly?
Is the 25 mL of standard measured accurately?
Is the sample buffered to pH 8?
Is the buffer/MIBK solution shaken vigorously
for 10 seconds?
Is disposal of solid and liquid wastes conducted
properly?
Yes
No
Comment
Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 11 of 17
MOBILE PROCESSING LABORATORY EQUIPMENT (Page 7 of 7)
Conductance Meter and Cell
Item
Is manufacturer's operating manual available?
Is instrument kept clean?
Is logbook kept up to date and signed daily?
Is standard stock solution prepared biweekly,
and is QC stock solution prepared weekly;
are solutions stored at 4°C?
Are working standards prepared daily?
Is exposure of samples and standards to the
atmosphere minimized?
Is required QC followed?
Are syringes and glassware cleaned
properly after use?
Yes
No
Comment
Comments:
-------
Appendix B
Revision 4
Date: 9/86
Page 12 of 17
SAMPLE PROCESSING (Page 1 of 1)
Item
Are all station documents kept in an orderly
fashion?
Are all forms completed as required and signed
by supervisor?
Is lab audit notebook kept up to date (labels
inserted, etc.)?
Are audit samples assigned different ID numbers
from day to day?
Are samples kept at 4°C when not being used?
Are coolers containing sample kept shut?
Are freeze-gel packs kept frozen?
Are samples properly packed for shipping
(sealed, cooled to 4°C, individually
wrapped, etc.)?
Are two copies of the completed shipping form
included with each batch of samples?
Yes
No
Comment
Comments:
-------
FIELD PERSONNEL (Page 1 of 1)
Position
A. Base Coordinator
B. Logistics Coordinator
C. Team A
1.
2.
D. Team B
1.
2.
E. Team C
1.
2.
F. Team D
1.
2.
6. Team E
1.
2.
Name
Agency
Academic
Training*
Special
Training*
Years
Experience**
*List highest degree obtained and specialty. Also list years toward a degree.
**List only experience directly relevant to task to be performed.
0> (O
o
—»•
^4
CD
-------
Appendix B
Revision 4
Date: 9/86
Page 14 of 17
FIELD BASE FACILITIES (Page 1 of 1)
Item
Has adequate space been provided for
predeparture activities?
Are facilities clean and organized?
Is equipment clean and organized?
Yes
No
Comment
-------
Appendix B
Revision 4
Date: 9/86
Page 15 of 17
FIELD SAMPLING-PREPARATION (Page 1 of 1)
Item
Are checklists followed for loading equipment?
Was sampling ever aborted because of forgotten
items?
Is equipment organized and easily accessible on
sampling craft/ vehicle?
Is equipment stored properly to prevent injury
or damage during transport?
Are adequate plans for the excursion made and
understood by personnel?
Does the field base coordinator know where all
teams are at any given time?
Are all meters properly calibrated or checked
for calibration?
Is calibration information completely and
correct!/ recorded?
Has an itinerary form been filled out
completely?
Yes
No
Comment
-------
Appendix B
Revision 4
Date: 9/86
Page 16 of 17
FIELD SAMPLING-EN ROUTE (Page 1 of 1)
Item
Are the maps adequate?
Are there problems locating streams?
Are the Stream Data Forms understood and
correctly filled out?
Yes
No
Comment
-------
Appendix B
Revision 4
Date: 9/86
Page 17 of 17
FIELD SAMPLING-ON SITE (Page 1 of 1)
Item
Are procedures clear and easily followed?
Is required QC followed?
Are required safety procedures followed?
Are adequate volumes of sample being taken?
Are rinse procedures followed carefully?
Are samples stored correctly?
Are all forms filled out correctly?
Yes
No
Comment
-------
-------
Appendix C
Revision 4
Date: 9/86
Page 1 of 53
APPENDIX C
ANALYTICAL LABORATORY ON-SITE EVALUATION QUESTIONNAIRE
GENERAL (Page 1 of 2)
Questionnaire Completion Date
Laboratory:
Street Address:
Mailing Address (if different from above):
City:
State: Zip:
Laboratory Telephone Number: Area Code: No.:
Laboratory Director:
Quality Assurance Officer:
(Quality Control Chemist)
Type of Evaluation:
Contract Number:
Contract Title:
-------
Appendix C
Revision 4
Date: 9/86
Page 2 of 53
GENERAL (Page 2 of 2}
Personnel Contacted:
Name Title
Laboratory Evaluation Team:
Name Title
-------
ORGANIZATION AND PERSONNEL (Page 2 of 3)
Analytical Laboratory Personnel
Position
Name
Academic Training*
Special Training
Years Experience**
7 O 3)>
&> 0) (D o
w *0
*List highest degree obtained and specialty. Also list years toward a degree.
**List only experience directly relevant to task to be performed.
-------
ORGANIZATION AND PERSONNEL (Page 1 of 3)
LABORATORY ORGANIZATIONAL CHART
* <0 §:
^5j" "^ *^
O 00 "
-------
Appendix C
Revision 4
Date: 9/86
Page 5 of 53
ORGANIZATION AND PERSONNEL (Page 3 of 3)
Item
Do personnel assigned to this project have the
appropriate educational background to success-
fully accomplish the objectives of the program?
Do personnel assigned to this project have the
appropriate level and type of experience to
successfully accomplish the objectives of this
program?
Is the organization adequately staffed to meet
project commitments in a timely manner?
Does the QA officer report to senior
management levels?
Was the QA manager available during the
evaluation?
Was the QA officer available during the
evaluation?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 6 of 53
LABORATORY MANAGER (Page 1 of 1)
Item
Does the laboratory manager have his/her own
copy of the standard operating procedures?
Does the laboratory manager have his/her own
copy of the instrument performance data?
Does the laboratory manager have his/her own
copy of the latest monthly QC plots?
Is the laboratory manager aware of the most
recent control limits?
Does the laboratory manager review the
following before reporting data:
a. The data?
b. The OC data sheet with analyst's
notes?
c. The QC chemist's blind audit
data report?
d. The calculated vs. measured
sample specific conductance?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 7 of 53
STANDARD OPERATING PROCEDURES (SOP) (Page 1 of 1)
Item
Has SOP Manual been written?
Is the SOP Manual followed in detail?
Does it contain all QC steps practiced?
Does each analyst have a copy at his/her
disposal?
Are plots of instrumental accuracy and
precision available for every analysis?
Are detection limit data tabulated for
each analysis?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 8 of 53
LABORATORY FACILITIES (Page 1 of 4)
When touring the facilities, special attention should be given to: (a) the
overall appearance of organization and neatness, (b) the proper maintenance of
facilities and instrumentation, and (c) the general adequacy of the facilites
to accomplish the required work.
Item
Does the laboratory appear to have adequate
workspace (12 sq. feet, 6 linear feet of
unencumbered bench space per analyst)?
Does the laboratory have a source of
distilled/demineralized water?
Is the specific conductance of distilled/
demineralized water routinely checked and
recorded?
Is the analytical balance located away from
drafts and areas subject to rapid
temperature changes?
Has the balance been calibrated in the past
year by a certified technician?
Is the balance checked with a class S
standard before each use, and is the check
recorded in a logbook?
Are exhaust hoods provided to allow effi-
cient work with volatile materials?
Is the laboratory clean and organized?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 9 of 53
LABORATORY FACILITIES (Page 2 of 4)
Item
Are contamination-free work areas provided
for the handling of toxic materials?
Are adequate facilities provided for
separate storage of samples, extracts, and
standards, including cold storage?
Is the temperature of the cold storage units
recorded daily in logbooks?
Are chemical waste disposal policies/
procedures adequate?
Are contamination-free areas provided for
trace-level analytical work?
Can the laboratory supervisor document that
water free of trace contaminants is avail-
able for preparing standards and blanks?
Do adequate procedures exist for disposal
of waste liquids from the ICP and AA
spectrometers?
Is the laboratory secure?
Are all chemicals dated on receipt and
thrown away when shelf life is exceeded?
Are all samples stored in the refrigerator
between analyses? I
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 10 of 53
LABORATORY FACILITIES (Page 3 of 4)
Item
Filter room or desiccator
(either is acceptable)
maintained at 15° to 35°C and
50% relative humidity
Gas
Lighting
Compressed air
Vacuum system
Electrical services
Hot arid cold water
Laboratory sink
Ventilation system
Hood space
Cabinet space
Storage space
(cite sq. ft.)
Shared space
Available
Yes
No
Comments
(where applicable, cite system,
QC check, adequacy of space)
-------
Appendix C
Revision 4
Date: 9/86
Page 11 of 53
LABORATORY FACILITIES (page 4 of 4)
COMMENTS ON LABORATORY FACILITIES
-------
EQUIPMENT-GENERAL (Page 1 of 1)
Appendix C
Revision 4
Date: 9/86
Page 12 of 53
Item
Balance, analytical
NBS-calibrated
thermometer
Desiccator
Balance, top
loader
Class "S"
weights
Balance table
Distilled water or
deionized water -
to meet Type I
Reagent Grade
specifications
Glassware
Drying oven
Hot plates
Equipment
# of
Units
Make
Model
Condition/age
Good
Fair
Poor
% of Time
Used in
Survey
-------
Appendix C
Revision 4
Date: 9/86
Page 13 of 53
ICPES/AAS (Page 1 of 4)
Item
ICP
Flame AAS
Flame AAS
Graphite Furnace AAS
Data System
Data System
Manufacturer
Model
Installation
Date
Comments on ICP/AAS Instrumentation:
-------
Appendix C
Revision 4
Date: 9/86
Page 14 of 53
ICPES/AAS (Page 2 of 4)
Item
Has the instrument been modified in any way?
Is a permanent service record maintained in
a logbook?
Is service maintenance by contract?
Is preventive maintenance applied?
Does the analyst have his/her own copy of
the standard operating procedures?
Are manufacturer's operating manuals readily
available to the analyst?
Is there a calibration protocol
available to analyst?
Are calibration results kept in a permanent
record?
Does the analyst have his/her own copy of
the instrument performance data?
Does the analyst have his/her own copy of
the latest weekly QC plots?
Is the analyst aware of the most recent
control limits?
Is a permanently bound notebook with
preprinted, consecutively numbered pages
being used?
Is the type of work clearly displayed on
the notebook?
Are the entries in the notebook legible?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 15 of 53
ICPES/AAS (Page 3 of 4)
Item
Are anomalies routinely recorded?
Has the analyst avoided obliterating
entries?
Are inserts (e.g., chromatograms, computer
printouts, etc.) permanently affixed in
notebook and signed across insert edge and
page?
Has the supervisor of the individual
maintaining the notebook personally
examined and reviewed the notebook period-
ically, and signed his/her name therein,
together with the date and appropriate com-
ments as to whether or not the notebook is
being maintained in an appropriate manner?
Does the analyst have a copy of the most
recent list of in-house samples to be
analyzed?
Date of list
Are all solutions properly labeled?
Is the instrument properly vented?
Is the interference correction automatically
performed?
Are dilute calibration standards
prepared fresh weekly?
Source
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 16 of 53
ICPES/AAS (Page 4 of 4)
Item
Is the QC check sample prepared from an
independent stock?
Source
Is the instrument allowed to warm up at
least 15 minutes with the flame on before
the final wavelength adjustment is made?
Is the calibration curve at least a five-
point curve?
Is the first calibration curve of the day
checked for detection limit and linearity?
Are the matrix spike data calculated
and plotted immediately after determination?
Is each new calibration curve checked to see
that the change in instrumental response is
less than 5%?
Are the following control samples analyzed
with each run?
Blanks
QC Sample
Spiked Sample
Duplicates
Does the analyst review the QC data sheet
prepared by the data clerk before the
analyst decides whether or not to release
Yes
No
Comment
the data for reporting?
-------
Appendix C
Revision 4
Date: 9/86
Page 17 of 53
ION CHROMATOGRAPH (Page 1 of 5)
Item
1C
1C
1C
Autosampler
Data System
Precolumn
Separator Column
Manufacturer
Model
Installation
Date
Supressor Column
Comments:
-------
Appendix C
Revision 4
Date: 9/86
Page 18 of 53
ION CHROMATOGRAPH (Page 2 of 5)
Item
Has the instrument been modified in any way?
Is a permanent service record maintained in
a logbook?
Is service maintenance by contract?
Is preventive maintenance applied?
Does the analyst have his/her own copy of
the standard operating procedures?
Are manufacturer's operating manuals readily
available to the analyst?
Is there a calibration protocol available
to the analyst?
Are calibration results kept in a permanent
record?
Does the analyst have his/her own copy of
the instrument performance data?
Does the analyst have his/her own copy of
the latest weekly QC plots?
Is the analyst aware of the most recent
control limits?
Is a permanently bound notebook with
preprinted, consecutively numbered pages
being used?
Is the type of work clearly displayed on
the notebook?
Are the entries in the notebook legible?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 19 of 53
ION CHROMATOGRAPH .(Page 3 of 5)
Item
Yes
No
Comment
Are anomalies routinely recorded?
Has the analyst avoided obliterating
entries?
Are inserts (e.g., chromatograms, computer
printouts, etc.) permanently affixed in
notebook and signed across insert edge and
page?
Has the supervisor of the individual
maintaining the notebook personally
examined and reviewed the notebook period-
ically, and signed his/her name therein,
together with entering the date and
appropriate comments as to whether or not
the notebook is being maintained in an
appropriate manner?
Does the analyst have a copy of the most
recent list of in-house samples to be
analyzed?
Date of list
Are all solutions properly labeled?
-------
Appendix C
Revision 4
Date: 9/86
Page 20 of 53
ION CHROMATOGRAPH (Page 4 of 5)
Item
Are dilute calibration standards
prepared fresh weekly?
Source
If manual techniques are used, is eluant
prepared fresh daily from the same
concentrated stock buffer?
Is the QC check sample prepared from an
independent stock?
Source
Is the calibration curve at least a four-
point curve for each analytical range?
Is the first calibration curve of the day
checked for detection limit and recovery?
Are the analyst's spike data calculated and
plotted immediately following determination?
Are the following control samples analyzed
with each run?
Blanks
QC Sample
Spiked 5ampJ[e_
Duplicates
Does the analyst review the QC data sheet
output by the data clerk and then decide
whether or not to release the data for
reporting?
Is the drip tray examined daily for
reagent spills, and are spills cleaned up
daily?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 21 of 53
ION CHROMATOGRAPH (Page 5 of 5)
Item
Are pumps oiled once per week?
Is the anion precolumn cleaned as necessary?
Is the S042~/N03~ resolution checked once
per batch and documented?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 22 of 53
ACIDITY AND ALKALINITY (Page 1 of 4)
A. Manual System
Item
pH Meter
Electrodes
Manufacturer
Model
Installation
Date
Data System
Titration Apparatus (burets, etc.):
B. Automated System
Item
System
Meter
Electrodes
Manufacturer
Model
Installation
Date
Auto Titration Specifications:
Comments:
-------
Appendix C
Revision 4
Date: 9/86
Page 23 of 53
ACIDITY AND ALKALINITY (Page 2 of 4)
Item
Has the instrument been modified in any way?
Is a permanent service record maintained in
a logbook?
Is service maintenance by contract?
Is preventive maintenance applied?
Does the analyst have his/her own copy of
the standard operating procedures?
Are manufacturer's operating manuals readily
available to the analyst?
Is there a calibration protocol available
to the analyst?
Are calibration results kept in a permanent
record?
Does the analyst have his/her own copy of
the instrument performance data?
Does the analyst have his/her own copy of
the latest weekly QC plots?
Is the analyst aware of the most recent
control limits?
Is a permanently bound notebook with
preprinted, consecutively numbered pages
being used?
Is the type of work clearly displayed on
the notebook?
Are entries in the notebook legible?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 24 of 53
ACIDITY AND ALKALINITY (Page 3 of 4)
Item
Are anomalies routinely recorded?
Has the analyst avoided obliterating
entries?
Are inserts (e.g., chromatograms, computer
printout, etc.) permanently affixed in note-
book and signed across insert edge and page?
Has the supervisor of the individual
maintaining the notebook personally
examined and reviewed the notebook period-
ically, and signed his/her name therein,
together with entering the date and
appropriate comments as to whether or not
the notebook is being maintained in an
appropriate manner?
Does the analyst have a copy of the most
recent list of in-house samples to be
analyzed?
Date of list
Are all solutions properly labeled?
Are burets and micropipets calibrated
weekly or more often?
Is the stock 1.0 and 0.01 N NaOH
standardized as required in methods manual?
Are the correlation coefficients of the
data examined to ensure that they are
Yes
No
Comment
greater than 0.9990?
-------
Appendix C
Revision 4
Date: 9/86
Page 25 of 53
ACIDITY AND ALKALINITY (Page 4 of 4)
Item
Yes
No|
Comment
Does the analyst review the QC data sheet
prepared by the data clerk before the
analyst decides whether or not to release
the data for reporting?
Are electrodes stored as recommended
by the manufacturer?
Are electrodes checked and filled, if
necessary, before each analysis?
-------
Appendix C
Revision 4
Date: 9/86
Page 26 of 53
pH (Page 1 of 4)
Item
Meter
Electrodes
Manufacturer
Model
Installation
Date
Type of Temperature Compensation
Standard Gas Supplier
Standard Gas Specifications
Comments:
-------
Appendix C
Revision 4
Date: 9/86
Page 27 of 53
pH (Page 2 of 4)
Item
Has the instrument been modified in any way?
Is a permanent service record maintained in
a logbook?
Is service maintenance by contract?
Is preventive maintenance applied?
Does the analyst have his/her own copy of
the standard operating procedures?
Are manufacturer's operating manuals readily
available to the analyst?
Is there a calibration protocol available
to the analyst?
Are calibration results kept in a permanent
record?
Does the analyst have his/her own copy of
the instrument performance data?
Does the analyst have his/her own copy of
the latest weekly QC plots?
Is the analyst aware of the most recent
control limits?
Is a permanently bound notebook with
preprinted, consecutively numbered pages
being used?
Is the type of work clearly displayed on
the notebook?
Are the entries in the notebook legible?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 28 of 53
pH (Page 3 of 4)
Item
Are anomalies routinely recorded?
Has the analyst avoided obliterating
entries?
Are inserts (e.g., chromatograms, computer
printouts, etc.) permanently affixed in
notebook and signed across insert edge and
page?
Has the supervisor of the individual
maintaining the notebook personally
examined and reviewed the notebook period-
ically, and signed his/her name therein,
together with entering the date and
appropriate comments as to whether or not
the notebook is being maintained in an
appropriate manner?
Does the analyst have a copy of the most
recent list of in-house samples to be
analyzed?
Date of list
Are all solutions properly labeled?
Is the pH meter calibrated before samples
are analyzed?
Is the pH meter calibration checked every
batch as required in the methods manual?
Is the pH electrode QC solution analyzed
first and as specified, and are the results
plotted immediately after determination?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 29 of 53
pH (Page 4 of 4)
Item
Does the material used as a QC sample meet
specifications?
Source of QCCS:
Are the following control samples analyzed
with each run:
QCCS
Duplicate
Does the analyst review the QC data sheet
prepared by the data clerk before the
analyst decides whether or not to release
the data for reporting?
Are electrodes stored as recommended by
the manufacturer?
Are electrodes checked and filled, if
necessary, before each analysis?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 30 of 53
FLUORIDE ION SELECTIVE ELECTRODE (Page 1 of 3)
Item
Meter
Electrodes
Manufacturer
Model
Installation
Date
Comments:
-------
Appendix C
Revision 4
Date: 9/86
Page 31 of 53
FLUORIDE ION SELECTIVE ELECTRODE (Page 2 of 3)
Item
Has the instrument been modified in any way?
Is a permanent service record maintained in
a logbook?
Is service maintenance by contract?
Is preventive maintenance applied?
Does the analyst have his/her own copy of
the standard operating procedures?
Are manufacturer's operating manuals readily
available to the analyst?
Is there a calibration protocol available
to the analyst?
Are calibration results kept in a permanent
record?
Does the analyst have his/her own copy of
the instrument performance data?
Does the analyst have his/her own copy of
the latest weekly QC plots?
Is the analyst aware of the most recent
control limits?
Is a permanently bound notebook with
preprinted, consecutively numbered pages
being used?
Is the type of work clearly displayed on
the notebook?
Are the entries in the notebook legible?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 32 of 53
FLUORIDE ION SELECTIVE ELECTRODE (Page 3 of 3)
Item
Are entries noting anomalies routinely
recorded?
Has the analyst avoided obliterating
entries?
Are inserts (e.g., chroma to grams, computer
printouts, etc.) permanently affixed in
notebook and signed across insert edge and
page?
Has the supervisor of the individual
maintaining the notebook personally
examined and reviewed the notebook period-
ically, and signed his/her name therein,
together with entering the date and
appropriate comments as to whether or not
the notebook is being maintained in an
appropriate manner?
Does the analyst have a copy of the most
recent list of in-house samples to be
analyzed?
Date of list
Are all solutions properly labeled?
Is there an electrode dedicated to low-
level F~ analysis?
Is all labware that comes in contact with
standards and samples made of plastic?
Is the temperature regulated?
Yes
No
Comment
Is a multipoint calibration used?
-------
CARBON ANALYZER (Page 1 of 3)
Appendix C
Revision 4
Date: 9/86
Page 33 of 53
Make and Model:
Specifications:
Comments:
-------
Appendix C
Revision 4
Date: 9/86
Page 34 of 53
CARBON ANALYZER (Page 2 of 3)
Item
Has the instrument been modified in any way?
Is a permanent service record maintained in
a logbook?
Is service maintenance by contract?
1
Is preventive maintenance applied?
Does the analyst have his/her own copy of
the standard operating procedures?
Are manufacturer's operating manuals readily
available to the analyst?
Is there a calibration protocol available
to the analyst?
Are calibration results kept in a permanent
record?
Does the analyst have his/her own copy of
the instrument performance data?
Does the analyst have his/her own copy of
the latest weekly QC plots?
Is the analyst aware of the most recent
control limits?
Is a permanently bound notebook with
preprinted, consecutively numbered pages
being used?
Is the type of work clearly displayed on
the notebook?
Are the entries in the notebook legible?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 35 of 53
CARBON ANALYZER (Page 3 of 3)
Item
Are anomalies routinely recorded?
Has the analyst avoided obliterating
entries?
Are inserts (e.g., chromatograms, computer
printouts, etc.) permanently affixed in
notebook and signed across insert edge and
page?
Has the supervisor of the individual
maintaining the notebook personally
examined and reviewed the notebook period-
ically, and signed his/her name therein,
together with entering the date and
appropriate comments as to whether or not
the notebook is being maintained in an
appropriate manner?
Does the analyst have a copy of the most
recent list of in-house samples to be
analyzed?
Date of list
Are all solutions properly labeled?
Is C02-free water used to prepare
standards?
Are precautions taken to prevent COg
contamination of samples and standards?
Is instrument designed to determine
both DOC and DIG? If not, what
modifications are necessary?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 36 of 53
AUTOMATED ANALYZER (Page 1 of 5)
Item
Automated Analyzer
Electrodes
Data System
Manufacturer
Model
Installation
Date
Comments:
-------
Appendix C
Revision 4
Date: 9/86
Page 37 of 53
AUTOMATED ANALYZER (Page 2 of 5)
Item
Has the instrument been modified in any way?
Is a permanent service record maintained in
a logbook?
Is service maintenance by contract?
Is preventive maintenance applied?
Does the analyst have his/her own copy of
the standard operating procedures?
Are manufacturer's operating manuals readily
available to the analyst?
Is there a calibration protocol available
to the analyst?
Are calibration results kept in a permanent
record?
Does the analyst have his own copy of the
instrument performance data?
Does the analyst have his own copy of the
latest weekly QC plots?
Is the analyst aware of the most recent
control limits?
Is a permanently bound notebook with
preprinted, consecutively numbered pages
being used?
Is the type of work clearly displayed on
the notebook?
Are the entries in the notebook legible?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 38 of 53
AUTOMATED ANALYZER (Page 3 of 5}
Item
Are anomalies routinely recorded?
Has the analyst avoided obliterating
entries?
Are inserts (e.g., chromatograms, computer
printouts, etc.) permanently affixed in
notebook and signed across insert edge and
page?
Has the supervisor of the individual
maintaining the notebook personally
examined and reviewed the notebook period-
ically, and signed his/her name therein,
together with entering the date and
appropriate comments as to whether or not
the notebook is being maintained in an
appropriate manner?
Does the analyst have a copy of the most
recent list of in-house samples to be
analyzed?
Date of list
Are all solutions properly labeled?
Are dilute calibration standards
prepared fresh daily?
Source
Is the QC check sample prepared fresh
daily from an independent stock?
Source
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 39 of 53
AUTOMATED ANALYZER (Page 4 of 5)
Item
Is the calibration curve at least a five-
point curve?
Is the first calibration curve of the day
checked for detection limit(s) and
linearity?
Are the analyst QC sample data calculated
and plotted real time?
Is there an automated analyzer dedicated to
each analysis (Total P, NH4+ Si02)?
Is each new calibration curve checked to
see that the change in instrumental response
is less than 5%?
Are the following control samples analyzed
with each run?
Reagent Blanks
QC Sample
^piked Sample
Duplicates
Does the analyst review the QC data sheet
prepared by the data clerk and then decide
whether or not to release the data for
reporting?
Is the water pumped through all lines
daily before and after analysis?
Are pump tubes changed at least once
per three days?
Is the pump cleaned when the pump tubes
are changed?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 40 of 53
AUTOMATED ANALYZER (Page 5 of 5)
Item
Is soap solution that does not contain
phosphorus pumped through all lines once
per week?
Is the flowcell cleaned with a sulfuric
acid-potassium dichromate solution
once per month?
Is the pump oiled once every three months?
Date of last service
Is the colorimeter mirror assembly and
color filter cleaned and the alignment
optimized once every three months?
Date of last service
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 41 of 53
SPECIFIC CONDUCTANCE (Page 1 of 3)
Item
Meter
Conductance Cell
Manufacturer
Model
Installation
Date
Is temperature compensated to 25°C?
What Is the cell constant?
Comments:
-------
Appendix C
Revision 4
Date: 9/86
Page 42 of 53
SPECIFIC CONDUCTANCE (Page 2 of 3)
Item
Has the instrument been modified in any way?
Is a permanent service record maintained in
a logbook?
Is service maintenance by contract?
Is preventive maintenance applied?
Does the analyst have his/her own copy of
the standard operating procedures?
Are manufacturer's operating manuals readily
available to the analyst?
Is there a calibration protocol available to
the analyst?
Are calibration results kept in a
permanent record?
Does the analyst have his/her own copy of
the instrument performance data?
Does the analyst have his/her own copy of
the latest weekly QC plots?
Is the analyst aware of the most recent
control limits?
Is a permanently bound notebook with
preprinted, consecutively numbered pages
being used?
Is the type of work clearly displayed on
the notebook?
Are the entries in the notebook legible?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 43 of 53
SPECIFIC CONDUCTANCE (Page 3 of 3)
Item
Are anomalies routinely recorded?
Has the analyst avoided obliterating
entries?
Are inserts (e.g., chromatograms, computer
printouts, etc.) permanently affixed in
notebook and signed across insert edge and
page?
Has the supervisor of the individual
maintaining the notebook personally
examined and reviewed the notebook period-
ically, and signed his/her name therein,
together with entering the date and
appropriate comments as to whether or not
the notebook is being maintained in an
appropriate manner?
Does the analyst have a copy of the most
recent list of in-house samples to be
analyzed?
Date of list
Are all solutions properly labeled?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 44 of 53
DOCUMENTATION/TRACKING (Page 1 of 1)
Item
Is a sample custodian designated? If yes, what
is the name of the sample custodian?
Name
Are the sample custodian's procedures and
responsibilities documented? If yes, where
are they documented?
Are written Standard Operating Procedures (SOPs)
developed for receipt of samples? If yes,
where are the SOPs documented (laboratory
manual, written instructions, etc.)?
Are written Standard Operating Procedures (SOPs)
developed for compiling and maintaining sample
document files? If yes, where are the SOPs
documented (laboratory manual, written
instructions, etc.)?
Are samples that require preservation stored in
such a way as to maintain their preservation?
If yes, how are the samples stored?
After completion of the analysis, are the
samples properly stored for 6 months or
until laboratory personnel are told otherwise?
Are the magnetic tapes stored in a secure area?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 45 of 53
ANALYTICAL METHODOLOGY (Page 1 of 2)
Item
Are the required methods used?
Is there any unauthorized deviation from contract
methodology?
Are written analytical procedures provided to
the analyst?
Are reagent-grade or higher purity chemicals
used to prepare standards?
Are fresh analytical standards prepared at a
frequency consistent with good QA?
Are reference materials properly labeled with
concentrations, date of preparations, and the
identity of the person preparing the sample?
Is a standard preparation and tracking logbook
maintained?
Do the analysts record bench data in a neat and
accurate manner?
Is the appropriate instrumentation used in
accordance with the required protocol (s)?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 46 of 53
ANALYTICAL METHODOLOGY (Page 2 of 2)
COMMENTS ON ANALYTICAL METHODS AND PRACTICES
-------
Appendix C
Revision 4
Date: 9/86
Page 47 of 53
QUALITY CONTROL (Page 1 of 3)
Item
Does
Does
of a
the
the
QC
a.
b.
c.
d.
e.
f.
g.
h.
i.
j.
k.
1.
laboratory maintain a QC manual?
manual address the important elements
program, including the following:
Personnel?
Facilities and equipment?
Operation of instruments?
Documentation of procedures?
Procurement and inventory practices?
Preventive maintenance?
Reliability of data?
Data validation?
Feedback and corrective action?
Instrument calibration?
Recordkeeping?
Internal audits?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 48 of 53
QUALITY CONTROL (Page 2 of 3)
Item
Are QC responsibilities and reporting
relationships clearly defined?
Have standard curves been adequately
documented?
Are laboratory standards traceable?
Are QC charts maintained for each routine
analysis?
Do QC records show corrective action
when analytical results fail to meet
QC criteria?
Do supervisory personnel review the data and
QC results?
Does the QC chemist have his/her own copy
of the standard operating procedures?
Does the QC chemist have his/her own copy
of the instrument performance data?
Does the QC chemist have his/her own copy
of the latest QC plots?
Is the QC chemist aware of the most recent
control limits?
Does the QC chemist prepare a blind audit
sample once per week?
Does the QC chemist routinely review and
report blank audit data to the laboratory
manager?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 49 of 53
QUALITY CONTROL (Page 3 of 3)
Item
Does the QC chemist update control limits
and obtain new control chart plots once
per day of analysis?
Are all QC data (control charts,
regression charts, QC data bases, etc.)
up to date and accessible?
Are minimum detection limits calculated
as specified?
Is QC data sheet information reported to
the analyst?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 50 of 53
DATA HANDLING (Page 1 of 2)
Item
Does the data clerk do a 100% check for
accuracy of data input to the computer?
Are data calculations checked by another
person?
Are data calculations documented?
Does strip chart reduction by on-line
electronic digitizing receive at least
5% manual spot checking?
Are manually interpreted strip chart
data spot-checked after initial entry?
Do laboratory records include the following
information?
Sample identification number
Station identification
Sample type
Date sample received in laboratory
Time, date, and volume of collection
Date of analysis
Analyst
Result of analysis (including raw
analytical data)
Receptor of the analytical data
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 51 of 53
DATA HANDLING (Page 2 of 2)
Item
Does laboratory follow required sample
tracking procedures from sample receipt
until discard?
Does the data clerk routinely report
QC data sheet information to the analyst?
Does the data clerk submit QC data sheet
information to the lab manager along with
the analytical data to be reported?
Do records indicate corrective action
taken?
Are provisions made for data storage for
all raw data, calculations, QC data, and
reports?
Are all data and records retained the
required amount of time?
Are computer printouts and reports
routinely spot-checked against laboratory
records before data are released?
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 52 of 53
SUMMARY (Page 1 of 2)
Item
Do responses to the evaluation indicate that
project and supervisory personnel are aware of
QA and its application to the project?
Do project and supervisory .personnel place
positive emphasis on QA/QC?
Have responses with respect to QA/QC aspects of
the project been open and direct?
Has a cooperative attitude been displayed by all
project and supervisory personnel?
Does the organization place the proper emphasis
on QA?
Have any QA/QC deficiencies been discussed before
leaving?
Is the overall QA adequate to accomplish the
objectives of the project?
Have corrective actions recommended during
previous evaluations been implemented?
Are any corrective actions required? If so,
list the necessary actions below.
Yes
No
Comment
-------
Appendix C
Revision 4
Date: 9/86
Page 53 of 53
SUMMARY (Page 2 of 2)
Summary Comments and Corrective Actions
-------
Appendix D
Revision 4
Date: 9/86
Page 1 of 3
Laboratory:
Sample Set 1
Sample Set 2
Quantitation:
APPENDIX D
NATIONAL STREAM SURVEY
PREAWARD AUDIT SAMPLE SCORING SHEET
NATIONAL STREAM SURVEY
Preaward Audit Sample Scoring Sheet
Date:
QA/QC
Deliverables
Total Score (Maximum = 200 points)
PART I. QUANTITATION
Aliquot 1
(Number of parameters within
acceptance criteria x 12/6*)
Aliquot 2
(Number of parameters within
acceptance criteria x 8/1*)
Aliquot 3
(Number of parameters within
acceptance criteria x 14/5*)
Aliquot 4
(Number of parameters within
acceptance criteria x 8/2*)
Aliquot 5
(Number of parameters within
acceptance criteria x 30/5*)
Aliquot 6
(Number of parameters within
acceptance criteria x 4/1*)
Aliquot 7
(Number of parameters within
acceptance criteria x 4/1*)
Points Awarded
Sample Set 1
(Low Cone.)
Points Awarded
Sample Set 2
(High Cone.)
Total
Score
*Number of parameters present in aliquot. The scoring for pH and DIC is based
on the air-equilibrated values.
-------
Appendix D
Revision 4
Date: 9/86
Page 2 of 3
Laboratory:
Date:
PART II. QUALITY ASSURANCE
A. Calibration/Reagent
Blank Analyses:
1. All parameters at less than
2 x CRDL.
2. One parameter at greater
than 2 x CRDL.
3. Two parameters at greater
than 2 x CRDL.
4. Three or more parameters
at greater than 2 x CRDL.
B. Quality Control Check
Samp!e
1. All verifications within
acceptance criteria.
2. One or more verifications
outside acceptance criteria.
C. Duplicate Sample Analyses:
1. All 2RSD values within
acceptance criteria.
2. 1-2 outside acceptance
criteria.
3. 3-4 outside acceptance
criteria.
4. 5 or more outside
acceptance criteria.
Possible
Points
6
4
2
0
10
0
6
4
2
0
Points
Awarded
-------
Appendix D
Revision 4
Date: 9/86
Page 3 of 3
Laboratory:
Date:
PART II. QUALITY ASSURANCE
(Continued)
Anion-Cation Balance
Calculation:
1. Within acceptance criteria.
2. Outside acceptance criteria.
Detection Limits:
1. All instrumental detection
limits within acceptance
criteria.
2. One or more outside
acceptance criteria.
Possible
Points
4
0
4
0
Points
Awarded
PART III. REPORTING AND DELIVERABLES
A. Data results submitted in acceptable format
on standard forms.
B. Quality assurance/quality control data
supplied in acceptable format.
C. Raw data supplied.
D. Tabulated instrument detection limits and associated
blank data supplied.
E. Validation of results with signature
of laboratory manager supplied.
Possible Points
1
5
-------
Appendix E
Revision 4
Date: 9/86
Page 1 of 16
APPENDIX £
NATIONAL STREAM SURVEY
VERIFICATION REPORT
1.0 NATIONAL STREAM SURVEY VERIFICATION REPORT
The NSS Verification Report is used as a guideline to evaluate and verify
National Stream Survey data.
The verification report is completed for every batch of data by deleting
the inappropriate verb of a verb pair (i.e., were/were not, was/was not,
etc.) and listing the affected samples and analyses. Those sections which
do not apply are crossed out. Explanations of the reasons for flagging
the data are necessary.
NSS VERIFICATION REPORT
BATCH NO.
LABORATORY
SAMPLING
SITE(S)
DATE AUDITED _
DATE REVIEWED
TOTAL NO. OF SAMPLES
BY
BY
DATE VERIFIED (FIRST PASS)
DATE TAPE (FIRST PASS)
SENT TO ORNL
DATE VERIFIED (FINAL)
DATE VERIFIED TAPE (FINAL)
SENT TO ORNL
I. OUTSTANDING ISSUES - ANALYTICAL LABORATORY
A. The Sample Data Package (was/was not) complete as submitted. The following
items that are identified as missing should be resubmitted before verifica-
tion process can begin.
1. a. Required forms (11, 13, 17, 18, 19, 20,
and 22) submitted.
b. Lab name, batch ID, and lab manager's
signature submitted on all forms.
Yes
Par-
tial
No
Comments
(continued)
-------
Appendix E
Revision 4
Date: 9/86
Page 2 of 16
c. Sample ID reported on Forms 13, 21, and 22.
d. Analyst's signature on Form 13.
e. Correct units indicated on all forms.
Form 11:
a. Correct number of samples analyzed and the
results for each parameter tabulated.
b. Correct data qualifiers (see Table 9.8)
reported as needed.
c. Alk. Initial pH and Acy. Initial pH are
within 0.1 pH unit.
d. For all sample data, pH (initial/equilibrated)
increases as DIC (initial/equilibrated)
decreases and vice versa.
e. Total extractable aluminum < total aluminum
for all samples.
Form 17:
a. 1C Resolution data reported for each batch
of analyses.
Form 18:
a. Instrumental detection limits and associated
dates of determination tabulated.
Form 19:
a. Date sampled, date received, holding time
plus date sampled and dates of analyses for
the correct number of samples are tab-
ulated.
b. Date analyzed is less than or equal to the
reported holding time plus date processed.
c. pH measurements are performed on the same
day as the DIC analysis for each sample.
Form 20:
a. Calibration blanks, reagent blanks, correct
number of QCCS runs, and DL QCCS reported
where required.
b. If high QCCS true values are reported, the
samples analyzed on high range are discussed
in the cover letter or reported on Form 21
(Dilution Factors).
c. QCCS true values are in the midrange of
linear dynamic range, otherwise DL QCCS
data are used to verify the low end of
the dynamic range.
Yes
Par-
tial
No
Comments
-------
Appendix E
Revision 4
Date: 9/86
Page 3 of 16
d. Calibration blank data are indicative of
instrument drift (greater than 2X CRDL for
positive values or less than [-] CRDL for
negative values).
e. Calibration blank data do not indicate
any trends throughout all batches.
Form 21:
a. Dilution Factors (if any) are reported
for each required parameter.
Form 22:
a. Duplicate precision results are reported for
each parameter.
b. Correct standard deviation formula (using
n-1) is used to calculate 2RSD.
c. Samples selected for duplicate analysis
contained sufficient amounts of analytes
(10 times the CRDL if possible) to yield
reliable precision.
d. If £RSD criterion is not met, another sample
is selected to be analyzed in duplicate.
e. Sample results on Form 22 match sample
results on Form 11.
Any information pertinent to sample analyses
are noted on the cover letter.
Yes
Par-
tial
No
Comments
-------
Appendix E
Revision 4
Date: 9/86
Page 4 of 16
B. The Sample Data Package (was/was not) complete as submitted, but the
following sample results should be confirmed by the analytical laboratory:
Sample Form Date Date
ID Number Parameter Requested Confirmed Reason for Confirmation
C. Sample analysis (was/was not) complete based on data submitted. Reanalysis
is recommended for the following samples:
Sample Reported Date Date
ID Parameter Value Requested Submitted Reason for Reanalysis
-------
— »
o
5
*JL|
o
t-H
u_
u.
o
t—
z
UJ
s
ii
r— II
II
CD H
Z »
t-4 II
Q II
0 H
— [— ||
— ^ H
ii
ii
n
n
H
H
n
n
n
n
it
ii
n
ii
n
II
n
II
it
u
II
n
u
II
II
II
II
II
„ Appendix E
"»«
CO ^^ *^
•g ££ TQ
Q
U—
o
^
C d)
i_ 0 >
0 -r- -I-
4- 4-> »
(/) ^ -t->
Q.
CJ
u
X
UJ
M-
O
01
3
-
1
X C7) O)
UJ C C
•r— 'r— d)
, <1> I— .1-
3
•o s o
r— T- t.
0 h- CD
Z
,
(w ^_
S- +J
(O QJ
Q.
UJ
ex. o
<£
CO
ii Revision 4
II
i; Date: 9/86
i; Page 5 of 16
n
n
n
n
n
n
n
I!
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II
II >-.
II 4-> L.
II r—
M C -i-
II
-------
Date Changes Applied:
Batch
ID
III. NUMERIC AND FLAG MODIFICATIONS (ADDITIONS
Form
No.
Sample
ID
Stream ID
Variable
Name
(Field ID)
Old
Flag
OR DELE
New
Flag
:TIONS) i
Old
Value
rO BE MA[
New
Value
)E TO THE RAi
Date
Edited
By
By:
4 DATA SET
Date
Revi ewed
By
====================================================================================================:::====
m
o>
-------
Appendix E
Revision 4
Date: 9/86
Page 7 of 16
IV. ANION/CATION BALANCE CHECK
Note: The flagged samples and parameters listed in the following
sections must be consistent with the most current computer-
generated exceptions.
A. Based on Anion/Cation balance check program, all samples submitted for this
batch (were/were not) within criteria. The following samples were listed
as exceptions:
Sample Sample Reported % Ion Bal. Required % Ion Bal.
ID Type Diff. (IBP) Diff. (IBP) Explanation
Samples listed above should be flagged appropriately as outlined in the
following sections:
1. Contamination (was/was not) indicated in the field or laboratory blanks for
the above exceptions. Contamination was apparent in the following samples:
Field/Lab
Sample IP Parameter Blank Cone. Explanation
The sample(s) listed above should be flagged using the appropriate sample
flag "A2" or "A3."
2. Unmeasured organic protolytes (were/were not) indicated by the Protolyte
Analysis Program for the exceptions listed in Section IV-A. The following
samples appear to have %IBP outside criteria because of unmeasured organic
protolytes:
Sample Reported POC Non-titrated Recalculated %IBO
IP (mg/L) Organic Ions (Organic Ions Included) Explanation
The samples listed above should be flagged using the sample flag "A4."
-------
Appendix E
Revision 4
Date: 9/86
Page 8 of 16
3. Analytical Error (was/was not) indicated in measurement of one or more
of the anions or cations contributing to the anion/cation balance check
calculation. Analytical error was apparent in the following parameters
and samples:
Sample ID Parameter Reported Cone. Explanation
The samples listed above should be flagged using the appropriate sample
flags "AS," "A6," "A7," or "A8."
4. Other unmeasured anions or cations not considered in %IBD calculation (were/
were not) suspected to contribute to anion/cation balance. The following
samples were suspected to contain unmeasured anions or cations:
Suspect
Unmeasured
Sample ID anion/cation Reported Cone. Explanation
The sample(s) listed above should be flagged for the suspect anion or cation
using the sample flag "Al."
5. Analytical error (was/was not) indicated in measurement of acid neutralizing
capacity (alkalinity) that affects the %IBD calculation. Analytical error
was apparent in the following samples:
Recalculated
Sample ID Reported Value, ueq/L value, ueq/L Explanation
The sample(s) listed above should be flagged using the sample flag "A9."
-------
Appendix E
Revision 4
Date: 9/86
Page 9 of 16
V. CONDUCTANCE BALANCE CHECK
A. Based on conductance check program, all samples submitted for this batch
(were/were not) within criteria. Using the conductance check program,
the following conclusions were made:
1. The Form 11 measured conductance (agreed/disagreed) with the calculated
conductance. The following samples had a Form 11 measured conductance
(greater/less) than the calculated conductance:
Calculated Required
Sample Form 11 Calculated Laboratory Maximum
ID Conductance Conductance % CD %CD Explanation
2. The mobile processing laboratory (trailer) conductance (agreed/disagreed)
with the Form 11 measured conductance. The following samples had a
mobile processing laboratory conductance (greater/less) than the Form
11 measured conductance:
Form 11 Calculated Required
Sample Trailer Measured Trailer Maximum
ID Conductance Conductance % CD %CD Explanation
The sample(s) listed above should be flagged using sample flag "F6."
The field in situ conductance (agreed/disagreed) with the Form 11
measured conductance. The following samples had a field in situ
conductance (greater/less) than the Form 11 measured conductance.
-------
Appendix E
Revision 4
Date: 9/86
Page 10 of 16
Field Calculated Required
Sample In situ Form 11 Field Maximum
—12— Conductance Conductance % CD %CD Explanation
The sample(s) listed above should be flagged using sample flag "FU."
4. Contamination (was/was not) indicated by the field or lab blanks for
the above exceptions. Contamination was apparent in the following
samples:
Sample Contaminated Field/Lab
ID Parameters Blank Concentration Explanation
The sample(s) listed above should be flagged using the appropriate
sample flag "C2" or "C3."
5. The % Conductance Difference (%CD) indicates possible analytical error
in the analytical laboratory conductance measurement for the following
samples:
Sample Contract Required
ID ICD Max %CD Explanation
Samples listed above should be flagged using the sample flag "Cb.
-------
Appendix E
Revision 4
Date: 9/86
Page 11 of 16
6. The % Conductance Difference UCD) indicates analytical error in the
trailer and/or field conductance measurements for the following
samples:
Sample Trailer Contract Required Field Contract Required
ID %CD Max %CD (Trailer) £CD Max %CD (Field) Explanation
Samples listed above should be flagged using the sample flags "FU"
(field) and/or "F6" (trailer).
7. Based on review of the data, unmeasured organic ions (were/were not)
suspected in the samples. The following samples "are" suspected to con-
tain unmeasured organic ions:
Sample ID Reported DOC (mg/L) Explanation
All samples listed above should be flagged unmeasured organic ions
using the sample flag "C4."
-------
Appendix E
Revision 4
Date: 9/86
Page 12 of 16
8. Analytical error (was/was not) indicated in the calculated conductance
value. Analytical error was apparent in the following parameters and
samples.
Sample Contract Required
ID Parameter %CD Max %CD Explanation
The sample(s) listed above should be flagged using the appropriate
sample flags "Cl," "C6," "C8," or "C9."
Other unmeasured anions or cations not considered in the %CD calculation
(were/were not) suspected to contribute to conductance balance. The
following samples were suspected to contain unmeasured anions or cations:
Suspect
Unmeasured
Sample ID Am'on/Cation Reported Cone. Explanation
The samples listed above should be flagged using the sample flag "C7."
-------
Appendix E
Revision 4
Date: 9/86
Page 13 of 16
VI. INTERNAL AND EXTERNAL QA/QC DATA REVIEW
A. All data for the following parameters and samples were not acceptable based
on the following:
1. The field blank (did/did not) exceed expected values and (did/did not)
contribute greater than 20% to the other samples in the batch (except
for other blanks). The contaminated samples follow:
Sample
ID Parameter % Contribution Explanation
All samples for the parameters listed above should be flagged using the
appropriate flags "BO," "B2," or "B5."
2. The calibration and/or reagent blank (was, was not) greater than 2 X
CRDL and (did/did not) contribute greater than 10$ to the other samples
in the batch. The contaminated samples follow:
Sample
ID Parameter % Contribution Explanation
All samples for the parameters listed above should be flagged using the
appropriate flags "81," "B3," "B4."
3. For a routine-field duplicate sample pair with both concentrations
greater than 10 times the CRDL, the field duplicate precision (was/
was not) within expected criteria. The maximum expected %RSD was
exceeded for the following parameters:
Contract Required
Parameter Reported %RSD Maximum %RSD Explanation
The parameters listed above should be flagged using the flag "D2."
-------
Appendix E
Revision 4
Date: 9/86
Page 14 of 16
4. The contract laboratory duplicate precision (was/was not) met. If
initial precision was outside criteria, two additional duplicates
(were/were not) run as required by the contract.
Program
Calculated Contract
Parameter Reported %RSD %RSD Required %RSD Explanation
The parameters listed above should be flagged using the flag "03."
5. Audit sample data (were/were not) within the expected performance range.
The following audit samples were outside of the expected range:
Audit Reported Expected
Parameter Sample Type Value Range Explanation
All samples in the batch for the parameters listed above should be flagged
using the appropriate flags "NO" or "Ml."
6. Internal Quality Control Check Sample (QCCS) analyses (were/were not)
within contractual requirements and the number of runs (were/were not)
complete.
Reported Required No. of No. of QCCS
Parameter Value Range QCCS Runs Runs Required Explanation
All samples in the batch for the parameters listed above should be flagged
using the appropriate flags "Ql" or "Q2" or if appropriate "Q3" or "g4."
-------
Appendix E
Revision 4
Date: 9/86
Page 15 of 16
7. Detection Limit Quality Control Check Sample (DL QCCS) analyses (were/
were not) within 20% of the theoretical concentration and the theoretical
concentration of the QCCS (was/was not) 2 to 3 times the CRDL.
Required
Parameter Reported Value Range Explanation
All samples listed above should be flagged using the flag "Q5."
8. Instrumental detection limit (did/did not) exceed the CRDL. The
following sample values reported at less than 10 times the IDL could
be in question:
Sample Reported Reported
ID Parameter Cone. IDL CRDL Explanation
All samples with concentrations <10 x IDL for the parameters listed above are
in question and should be flagged using the flag "LI."
-------
Appendix E
Revision 4
Date: 9/86
Page 16 of 16
VII. SUMMARY OF FLAGGED DATA
All internal QC data (calibration blanks, reagent blanks, QCCS, duplicate
precision) and external QA data (audits, field blanks, and field dupli-
cates) were not within contractual or expected criteria for all the
samples and the associated parameters listed below:
(Parameter Flags: BO, 81, B3-B5, D1-D3, NO, Nl, Q1-Q5)
(Sample Flags: AO-A8, B2, CO-C9, FO-F6, HO-H1, LI, PO-P7, XO-X4)
Parameter Flags and Parameters:
Sample Flags, Parameters, and Sample IDs:
•tt U.S. GOVERNMENT PRINTING OFFICE: 1988- 5 it 8-1 5 8 / 870 t *
-------
SUBREGIONS OF THE NATIONAL STREAM SURVEY-PHASE I
Northern
Appalachians (2Cn)
Valley and Ridge (2Bn)
Southern Blue Ridge (2As)
(Pilot Study)
Poconos/Catskills (ID)
NY\
Mid-Atlantic
Coastal Plain (3B)
Ozarks/Ouachitas (2D)
Southern Appalachians (2X)
------- |