United States Office of Research and EPA/600/R-94/038e
Environmental Protection Development April 1994 ,f
Agency Washington DC 20460 (~' '
v>EPA Quality Assurance
Handbook for
Air Pollution
Measurement
Systems
Volume V: Precipitation
Measurement Systems
(Interim Edition)
-------
EPA-600/R-94/038e
'"'v
j
QUALITY ASSURANCE HANDBOOK
FOR
AIR POLLUTION MEASUREMENT SYSTEMS
Volume V - Precipitation Measurement Systems
(Interim Edition)
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH and DEVELOPMENT
ATMOSPHERIC RESEARCH and ENVIRONMENTAL ASSESSMENT LABORATORY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
U.S F-,,,-.-, Crv
' ~" $7) Printed on Recycled Paper
^^ J^-' „ ^ -'-IIUI
ttfcW.' i. o.,0..^, J2tt
-------
CONTENTS
Section
1 OVERVIEW OF THE INTERIM EDITION OF
VOLUME V
2 PROGRAM PLANNING AND OBJECTIVES
2.1 DETERMINATION OF MONITORING AND
DATA QUALITY OBJECTIVES
2.2 DEVELOPMENT OF WORK PLAN
2.3 PREPARATION OF QUALITY ASSURANCE
PROJECT PLAN
3 PROGRAM ORGANIZATION AND RESPONSIBILITIES
3.1 PROGRAM OPERATIONS
3 . 2 FIELD OPERATIONS
3.3 LABORATORY OPERATIONS
3.4 DATA MANAGEMENT OPERATIONS
3.5 DESIGNATION OF QA RESPONSIBILITIES
AND DUTIES
3.6 REFERENCES
4 DOCUMENTATION
4.1 DOCUMENT CONTROL
4.2 INTERNAL DOCUMENTATION
4.3 QA DOCUMENTATION
5 SITING
5.1 NETWORK DESIGN CONSIDERATIONS
5.2 SITE SELECTION CRITERIA
5.2.1 Baseline Station
5.2.2 Regional Station
5.2.3 Urban or Local Station
5.3 SAMPLER AND RAIN GAUGE SITING CRITERIA
-------
CONTENTS (Continued)
Section
5.4 STATION IDENTIFICATION AND
CLASSIFICATION
5 . 5 REFERENCES
SITE DESCRIPTION REPORT
6 FIELD OPERATIONS
6.1 FACILITIES
6.2 METHOD SELECTION
6.2.1 Precipitation Collectors and
Rain Gauges
6.2.2 pH and Conductivity Apparatus,
"> Temperature Probe
6.2.3 Balance or Graduated Cylinders
6.3 ACCEPTANCE TESTING
6.3.1 Precipitation Collectors and
Rain Gauges
6.3.2 pH and Conductivity Meters
6.4 SAMPLER AND RAIN GAUGE INSTALLATION
AND OPERATION
6.4.1 Routine Checks on Collector,
Rain Gauge and Site
6.4.2 Corrective Action
6.5 SAMPLING METHODOLOGY
6.5.1 Sample Collection and Schedule
6.5.2 Handling of Plastic Containers
6.5.3 Sample Handling
6.5.4 Sample Preservation and Storage
6.6 FIELD MEASUREMENTS
6.6.1 pH Determination Method
6.6.2 Specific Conductance
Determination Method
6.6.3 Temperature Measurements
6.6.4 Gravimetric Measurements
6 . 7 DOCUMENTATION
6 .8 REFERENCES
-------
CONTENTS (Cont inued)
Section
7 LABORATORY OPERATIONS
7.1 ANALYTICAL REAGENTS
7.1.1 Purity Requirements
7.1.2 Storage Requirements
7.2 LABORATORY SUPPORT FOR THE FIELD
7.3 LABORATORY LOGISTICS
7.3.1 Sample Handling in
the Laboratory
7.3.2 Laboratory Documentation
7.3.3 Traceability of
Calibration Standards
7.3.4 Preparation of Analyst's Spikes
•> 7.3.5 Analytical Data Computations
7.4 QUALITY CONTROL PROGRAM
7.4.1 Real-Time Quality Control
Procedures
7.4.2 Analysis and Evaluation of
Quality Control Samples
7.4.3 Data Screening Tools
7.4.4 Control Limits Determination
7.4.5 Evaluation of QC Data
7.5 EVALUATION OF LABORATORY
PERFORMANCE
7.5.1 Independent Internal Quality
Control
7.5.2 Laboratory Audits
7.6 REFERENCES
8 DATA HANDLING, VALIDATION, AND REPORTING
8.1 DATA LOGISTICS
8.2 SOFTWARE REQUIREMENTS
8.2.1 Data Input
8.2.2 Data Storage and Indexing
8.2.3 Precipitation Data Bases
8.3 DATA HANDLING AND PRELIMINARY
SCREENING
8.3.1 Quality Control of Data Handling
8.3.2 Treatment of Outliers
-------
CONTENTS (Continued)
Section
8 8.4 DATA VALIDATION CRITERIA
8.4.1 Detection Limit Flag
8.4.2 Comparison of Sampler
and Rain Gauge Performance
8.4.3 Unusual Ion Ratios
8.4.4 Comparison of Anion and
Cation Equivalents
8.4.5 Comparison of Measured and
Calculated Conductances
8.5 DATA REPORTING
8.5.1 Average Concentrations and
Deposition
8.5.2 Median Concentrations
•> 8.5.3 Reporting and Treating
Below-Detection-Limit Data
8.5.4 Reporting of Out-of-Control
Data
8.6 QC CHECKS ON FINAL DATA
8.6.1 Time and Dates of Sampling
8.6.2 Codes, Flags and Identifiers
8.6.3 Overall Transcription Checks
8.6.4 Spotcheck/Recalculation of Data
8.6.5 QC Checks for Data Summaries
8.7 REFERENCES
9 DATA QUALITY ASSESSMENT
9.1 EVALUATION OF FIELD OPERATIONS
9.1.1 Measurement System Precision
9.1.2 Accuracy of pH and Conductivity
Measurements
9.1.3 Sampling Bucket Blanks
9.2 EVALUATION OF LABORATORY OPERATIONS
9.2.1 Analytical Precision
9.2.2 Accuracy of Chemical Analysis
9.3 DATA QUALITY REPORTING
9.4 DATA FORMS
9 . 5 REFERENCES
MONTHLY FIELD AUDIT REPORT
REPORT OF DUPLICATE ANALYSIS
-------
CONTENTS (Continued)
Section
10 ACID PRECIPITATION MONITORING PROGRAM
EVALUATION
10.1 PROGRAM AUDITS GUIDANCE
10.1.1 Support Material
10.1.2 Reporting
10.2 OVERALL PROGRAM OPERATION
QUESTIONNAIRE
10.3 SITE DOCUMENTATION EVALUATION
10.3.1 General Guidance
10.3.2 Site Evaluation Reporting
10.3.3 Site Documentation Review
10.4 LABORATORY OPERATIONS EVALUATION
10.4.1 Procedure
10.4.2 Analytical Laboratory
Questionnaire
10.5 PERFORMANCE AUDITS
10.5.1 Network Performance Audits
10.5.2 Performance Audit Reporting
10.5.3 Laboratory Performance Audits
10.6 DATA PROCESSING AUDITS
10.6.1 General Guidance
10.6.2 Estimating the Percent Errors
in a Data Base
A. OPERATION AND MAINTENANCE PROCEDURES FOR
PRECIPITATION MEASUREMENT SYSTEMS
-------
FIGURES
Figure
7-1 Analyst Spike Plot for S04" Analysis
TABLES
Table
2-1 USGS Survey of Synthetic Precipitation Samples
2-2 USGS Survey of Natural Precipitation Samples
2-3 Summary of Collocation Results
2.4 Recommended Reporting Units and Significant
Digits
•i
7-1 Sample Information to be Coded
7-2 Factors for Computing Control Limits
8-1 Suggested QC Spotcheck of Data Handling
8-2 Ion Ratios for Various Sources
8-3 Conversion Factors and Equivalent Weights
8-4 NADP Reanalysis Criteria
8-5 Equivalent Conductance at Infinite Dilution,
25°C
8-6 Confidence Bounds for Medians of Small Samples
8-7 NADP Minimum Detection Limit Criteria for
Laboratory Measurements
9-1 Network Summary of Upper and Lower Limits of the
SCDs for Daily Sampling
9-2 Acceptance Criteria for Test Sample Quality and
Field Analytical Accuracy
9-3 Absolute and Relative Blank Values
-------
Section No. 1
Date February 24, 1954
Page 1
1.0 OVERVIEW OF THE INTERIM EDITION OF VOLUME V
The Quality Assurance (QA) Handbook is comprised of five
volumes: Volume I (Principles), Volume II (Ambient Air Methods],
Volume III (Stationary Source Methods), Volume IV (Meteorological
Measurements), and Volume V (Precipitation Measurement Systems) .
Much of the material in Volumes II, III and V are out-of-date and
some portions of these volumes have long been out-of-print.
EPA is now preparing an updated version of the QA Handbook
series which will be available in September 1995. To meet the
needs of the user community until the updated version is
available, EPA has published Interim Editions of Volumes I, II,
III, IV and V. Each volume of the Interim Editions, is being
issued as a complete unit with out-of-date sections either
deleted or modified using addendum sheets and handwritten
notations in the text.
This volume and the other four volumes of the Interirr.
Edition of the QA Handbook are available at no charge frorr.:
USEPA/ORD
Center for Environmental Research Information
26 West Martin Luther King Drive
Cincinnati, Ohio 45268
The previous version of Volume V was published in twc parts:
Volume Va, Quality Assurance Manual for Precipitation Measurement
Systems and Volume Vb, Operation and Maintenance Manual for
Precipitation Measurement Systems. The appendices to Volume Vb
contained over 200 pages of Standard Operating Procedures (SOP'si
primarily concerned with the analysis of precipitation samples.
For the most part, these SOP's are now out-of-date technology and
were excluded from the Interim Version of Volume V. The reduced
Volume Vb is now an appendix to Volume V. The titles of the
excluded SOP's are:
• Aerochem Metrics Precipitation Collector Maintenance
Manual
• Instruction Book for Universal Recording Rain Gauge
• Method 150.6 -- pH of Wet Deposition by Electrometric
Determination
• Method 120-6 -- Specific Conductance in Wet Deposition
by Electrolytic Determination
-------
Section No. 1
Date February 24, ~_994
Page 2
Method 305.6 -- Acidity in Wet Deposition by
Titrimetric Determination
Method 305.2 -- Acidity (Tritrimetric)
Method 300.6 -- Chloride, Orthophosphate, Nitrate ar.a
Sulfate in Wet Deposition by Chemically Suppressed Ion
Chromatography
Method 300.7 -- Dissolved Sodium, Ammonium, Potassium,
Magnesium, and Calcium in Wet Deposition by Chemically
Suppressed Ion Chromatography
Method 375.6 -- Sulfate in Wet Deposition by Automated
Colorimetric Determination Using Barium-Methylthymcl
Blue
Method 353.6 -- Nitrate-Nitrite in Wet Deposition by
Automated Colorimetric Determination Using Cadmium
Reduction
Method 325.6 -- Chloride in Wet Deposition by Automated
Colorimetric Determination Using Thiocyanate
Method 365.6 -- Orthophosphate in Wet Deposition by
Automated Colorimetric Determination Using Ascorbic
Acid reduction
Method 340.6 -- Fluoride in Wet Deposition by
Potentiometric Determination Using an Ion-Selective
Electrode
Method 350.6 -- Ammonium in Wet Deposition by
Electrometric Determination Using Ion-Selective
Electrode
Method 350.7 -- Ammonium in Wet Deposition by Automated
Colorimetric Determination with Phenate
Method 200.6 -- Dissolved Calcium, Magnesium,
Potassium, and Sodium in Wet Deposition by Flame Atomic
Absorption Spectrophotometry
Method 200.6 -- Dissolved aluminum, Cadmium, Copper,
Iron, Lead, Manganese, and Zinc in Wet Deposition by
Graphite Furnace Atomic Absorption Spectrophotometry
-------
Section No. 1
Date February 24, 1554
Page 3
Copies of these SOP's can be obtained by writing tc:
QA Handbook Coordinator
US EPA/ORD/AREAL/MD-77B
Research Triangle Park, NC 27711
Many of the EPA contacts and organizational units identified
in Volume V are no longer correct and some of the reference
materials and procedures cited have been discontinued or
replaced. This type of out-of-date information is widely
dispersed throughout Volume V. Rather than change every affec~ed
section, for clarity and neatness sake, we have provided below a
listing of the original information and the corresponding updated
information.
1) NBS is now the National Institute of Standards and
Technology (NIST).
2) EMSL is now the Atmospheric Research and Exposure
Assessment Laboratory (AREAL).
3) QAD is now the Quality Assurance and Technical Support
Division (QATSD/AREAL).
The updated edition of Volume V which will be available ir.
September 1995 will also contain information on quality assuring
dry deposition measurement systems.
William J. Mitchell
Chief
Quality Assurance Support Branch
-------
Section No. 2
Revision No. 1
Date October 1, 1984
Page 1 of 8
2.0 PROGRAM PLANNING AND OBJECTIVES
Precipitation monitoring entails extensive chemical analyses and data
manipulations and requires strong interfaces between field operations,
laboratory operations and the data management functions. In a program of
this complexity, all elements must be carefully planned. A comprehensive
•top-down" approach to planning should be adopted, starting with an
experimental design and extending through to the preparation of detailed
procedures .
As described below, the planning process can be broken down into three
phases:
1. Determination of Monitoring and Data Quality Objectives
2. Development of Work Plan
3. Preparation of Quality Assurance Project Plan
2.1 Determination of Monitoring and Data Quality Objectives
Data quality objectives are usually defined in terms of precision,
accuracy, representativeness, comparability and completeness of the collected
data. To achieve these objectives, quality assurance/quality control
procedures are needed in all phases of the program from the initial planning
stages through final data reporting. This involvement helps identify areas
with potentially large negative impact on data quality and provides a
mechanism for instituting ongoing quality control, corrective action, data
validation, and external assessment of precision and accuracy.
In general, precipitation data may be considered to be complete if at
least B5% of the total possible field observations of either individual
events, daily, weekly, or longer term composite samples are captured. The
determination of completeness for the network is calculated as a percentage
-------
Section No. 2
Revision No. 1
Date October 1, 1984
Page 2 of 8
of the total ntmber of possible samples. The percentage of valid usable data
will usually be 5-10J lower due to sample contamination or loss and the
invalidation of some analytical results.
Tables 2-1, 2-2 and 2-3 show some precision estimates obtained from
interlaboratory studies by the 0. S. Geological Survey (Tables 2-1, 2-2) and
a collocated sampling study from the EPRI-Sure network (Table 2-3). These
results can be useful in establishing data quality objectives.
In order to establish data representativeness, stations should be sited
so that they collect samples representative of both the amount and the
composition of precipitation in the area. In general, monitoring sites
should be classified as: (a) baseline (remote area), (b) regional (rural),
or (c) urban or local area (for local impact emission sources). Also,
reporting data in consistent units permits easier data comparisons. A
listing of the recommended units for the most commonly sought monitoring
parameters are given in Table 2-4. To facilitate intercomparison of data
bases of various networks, data summaries should document, to the extent
possible, field laboratory, and computational procedures utilized in data
generation.
2.2 Development of Work Plan
Once the monitoring and data quality objectives have been determined, a
detailed work plan should be prepared. The Work Plan should address, as a
minimum, the following topics which are discussed in more detail in later
sections of this manual.
1. Program Organization - identifies the organization(s) and specific
personnel responsible for network operation, chemical analysis, data
management and quality assurance/ quality control (Section 3*0).
2. Experimental Design - addresses the spatial and temporal measurement
requirements in terms of program objectives (Section 2.0). Special
consideration should be given to proper network siting (Section 5.0), and
to thorough documentation of all detailed procedures to be used (Section
4.0).
3. Facilities, Equipment and Services - provides a list of resources required
to carry out the monitoring program. This listing should identify the
specific instruments and respective model numbers utilized in each facet
-------
Section No. 2
Revision No. 1
Date October 1, 1984
Page 3 of 8
TABLE 2-1. OSGS SURVEY OF SYNTHETIC PRECIPITATION SAMPLES
Observable
(Units)
Conductivity
(umho/cB)
PH
Sulfate
(ag/liter)
litrate
(ag/liter)
iamonia
(ag/liter)
Chloride
(ag/liter)
Sodium
(ag/liter)
Potassium
(ag/ liter)
Kagnesiun
(B«/ liter)
Calciiat
Cag/liter)
No. of
Labs
10
10
10
10
11
11
11
11
10
10
11
11
11
11
10
10
11
11
11
11
uses
Designated
43.5
13.2
4.35
5.03
2.35
0.779
2.15
0.826
0.660
0.214
3.27
0.908
0.558
0.284
0.506
0.156
0.226
0.056
0.654
0.248
Mean of
Determinations
40.7
11.6
4.22
4.83
2.38
0.702
1.80
0.732
0.630
0.203
3.18
0.891
0.586
0.326
0.549
0.198
0.27?
0.0752
0.652
0.253
Std.
Deviation
3-32
4.09
0.124
0.172
0.795
0.174
0.662
0.212
0.130
0.039
0.249
0.188
0.139
0.070
0.941
0.106
0.0872
0.0361
0.0911
0.0585
CV*
0.082
0.35
0.029
0.036
0.33
0.25
0.37
0.29
0.21
0.19
0.078
0.21
0.24
0.22
0.171
0.53
0.32
O.M8
0.14
0.23
a. Chemical Analysis and Precipitation Study, U.S.G.S., Spring 1981;
saaples designated B and N.
'=. CV =.Std. deviation/mean
-------
Section No. 2
Revision No. 1
Date October 1, 1984
Page 4 of 8
TABLE 2-2. USGS
Observable
(Units)
Conductivity
(y«ho/cm)
PH
Sulfate
(•g/liter) .
Hitrate
(•g/liter)
Aanonia
(•g/liter)
Chloride
(•g/liter)
Sodium
(•g/liter)
Potassium
(•g/liter)
Magnesium
(ng/liter)
Calcium
(ng/liter)
SURVEY OF
No. of
Labs
10
10
10
10
11
11
11
11
10
11
11
11
11
10
10
11
11
11
11
NATURAL PRECIPITATION
Mean of
Determinations
16.9
28.9
6.22
6.25
0.477
1.52
1.23
2.77
0.682
1.81
2.40
1.40
1.73
0.32
1.15
0.198
0.307
1.11
1.48
SAMPLES a
Std.
Deviation
1.56
1.93
0.219
0.258
0.162
0.606
0.436
1.05
0.138
0.3M9
0.391
0.981
0.249
0.0646
0.122
0.0399
0.0438
0.210
0.109
C?5
0.092
0.067
0.035
0.041
0.34
0.40
0.35
0.38
0.20
0.19
0.16
0.70
0.14
0.20
0.11
0.20
0.14
0.19
0.07
a. Chemical Analysis and Precipitation Study, U.S.G.S., Spring 1981;
samples designated AC and PL.
b. CV = Std. deviation/mean
-------
TABLE 2-3. SUMMARY OF COLLOCATION RESULTS3
Section No. 2
Revision Mo. 1
Date October 1, 1984
Page 5 of 8
Observable
Hydrogen Ion0
Total Acidityd
Conductivity
Sulfate
Nitrate
Chloride
Ammonia
Sodium
Potassium
Calcium
Magnesium
% Difference
9.0
24.3
6.0
5.0
5.3
11.5
7.8
17.6
43.3
16.1
12.5
a. From EPRI-SURE Acid Precipitation Study. The values reported
represent an average of at least 1000 collocations of samplers.
b. % Difference a 100x(Median Absolute Collocated Difference/Median All
Values).
G. From antilog of -pH.
d. Classical potentiometric titration
-------
Section No. 2
Revision No. 1
Date October 1, 1984
Page 6 of 8
TABLE 2-4. RECOWENDED REPORTING UNITS AND SIGNIFICANT DIGITS
Observable
Sample weight
Precipitation
PH
Conductivity
Deposi tion
Sulfate
Hitrate
Aanonia
Chloride
Sodium
Potassium
Magnesium
Calcium
Reporting Units
g
cm
pH units
y mho/on (yS/cm)
2
mg/m
mg/liter
yeq/liter
mg/liter
yeq/liter
mg/liter
yeq/liter
mg/liter
yeq/liter
mg/liter
yeq/liter
mg/liter
yeq/liter
mg/liter
yeq/liter
mg/liter
yeq/liter
Significant Digits
1.0
0.025
0.01
0.1
0.1
0.01
0.2
0.01
0.2
0.001
0.05
0.01
0.3
0.01
0.4
0.001
0.02
0.001
0.08
0.001
0.05
-------
Section No. 2
Revision No. 1
Date October 1, 1984
Page 7 of 8
of the measurement system (Sections 6.0 and 7.0).
4. Data Generation - specifies the measurement methods used. These include
calibration techniques, frequency of calibrations, acceptability
requirements for all calibration results, and action to be taken with
respect to data obtained previous to an unsatisfactory calibration
(Sections 6.0 and 7.0).
5. Data Processing, Validation and Reporting - describes the types of data
obtained for each measurement parameter and the overall data flow from
generation through reporting. Data validation methodologies, statistical
analysis techniques and reporting formats should also be addressed
(Section 8.0).
6.. Program Evaluation and Data Quality Assessment - provides specific details
of the planning and implementation of the independent quality assurance
activities associated with the monitoring program (Sections 9*0, 10.0 and
2.3 below).
2.3 Preparation of Quality Assurance Project Plan
To generate and report monitoring data of the highest quality, no
precipitation monitoring project should be initiated without a written, QA
project plan. The QA project plan should specify QA policies, organization,
objectives, functional activities, and QA/QC activities needed to achieve the
data quality goals of the project. Items that should be included are:
1. Quality Assurance Policy Statement - describes the organization's policy
and general approach to achieve quality results.
2. Quality Assurance Organization and Responsibility - assigns responsibility
to the organization's personnel for carrying out quality related
activities and defines the authority of the quality assurance officer
(QAO).
3. Sampling and Analysis Procedures - defines the methods to be used both in
the field and laboratory, in terms of their acceptability and
applicability, and references the appropriate Standard Operating
Procedures for method details.
4. Internal Quality Control Checks - describes the critical points in each
measurement system and establishes a system of control and/or checks for
those critical points together with a frequency for their performance.
5. Data Quality Assessment - addresses the implementation of independent
program evaluations and audits to assess data quality; also establishes a
schedule for these audits and a format for reporting precision, accuracy,
and completeness of specific measurements.
-------
Section No. 2
Revision No. 1
Date October 1, 1984
Page 8 of 8
Corrective Actions - indicates the methods for reporting problems,
responsibility for corrections, and the documentation of action taken.
Quality Assurance Reporting - outlines the data quality section to be
prepared as part of each data submission and final report. The report
will summarize data quality assessment, reliability of the measurement
system and corrective action pursued to correct the problems.
-------
Section No. 3
Revision No. 1
Date October 1, 1984
Page 1 of 6
3.0 PROGRAM ORGANIZATION AND RESPONSIBILITIES
A precipitation monitoring network involves interdependent field
monitoring and laboratory operations. Each operation has its own QA/QC
aspects. The field monitoring sites and the laboratory can be run by
independent organizations; however, the results reported are the
responsibility of the program manager. The QA officer should be at the same
reporting level as the program manager. Brief discussions of the
qualifications and duties of the program personnel are presented below.
3.1 Program Operations
In order to oversee, coordinate and review the program as a whole the
following functions are required.
Program Manager - A program manager should:
(a) Assure that data of acceptable precision and accuracy are generated within
the time and funding constraints of the program.
(b) Keep abreast of all program requirements and make necessary decisions.
(c) Review data and QA reports.
(d) Issue progress reports.
(e) Analyze and interpret the results.
The program manager should have a degree in science and some experience
in managing projects. Full time employment is recommended. This position
can be combined with that of the field manager.
QA Officer - The QA officer should:
(a) Report directly to the program manager on inputs from other program
functions.
(b) Assess the quality of the data generated.
(c) Recommend corrective actions that need to be taken.
-------
Section No. 3
Revision No. 1
Date October 1, 1984
Page 2 of 6
The QA officer might be assisted by one or more QA/QC • coordinators if
the magnitude of the program requires it. The QA officer should have
training in QA as it relates to experimental design, monitoring and data
validation.
3.2 Field Operations
The personnel needed to carry out the field duties in a precipitation
•onitoring network include a field manager and station operators.
Field Manager - The field manager may be a member of the organization
operating the stations or a member of the central laboratory staff. He
should have a college degree, preferably in chemistry; should be familiar
with all the field procedures; and should have experience in the operation
of all equipment. His duties are:
(a) Solve field problems.
(b) Notify the program manager of such problems.
(c) Oversee and train the operators.
(d) Coordinate between field and laboratory functions.
Station operators should have a technical background but need not have a
college degree. All operators should have the training to perform at the
necessary level of knowledge and the skill required to obtain and report
quality data. A short-term course in "hands-on" training is recommended.
This should be followed by on-the-job observation immediately after the
course and by a semi-annual inspection thereafter. The training should cover
all pertinent aspects of the Operations and Maintenance Manual (1), which
should be given to all personnel. If, at any time, an operator's performance
deteriorates, additional training must be provided as soon as possible by the
field manager, by a refresher course or by on-site guidance during network
evaluation visits.
Station Operators - The duties of a station operator are:
Ca) Operate and maintain samplers according to appropriate SOPs.
-------
Section No. 3
Revision No. 1
Date October 1, 1984
Page 3 of 6
(b) Operate and maintain rain gauges according to appropriate SOP.
(c) Maintain monitoring site and its surroundings free from obstruction and
dirt.
(d) Change sampling buckets at pre-assigned intervals and record activities in
field logs.
(e) Perform required calibrations of measurement instruments and analyses of
samples as described in SOPs.
(f} Preserve and store field samples as appropriate
(g) Ship samples to analytical laboratory at intervals specified in SOPs.
(h) Notify field manager on any problems with samplers, gauges, instruments,
standards and the like.
3.3 Laboratory Operations
Each analytical laboratory should have the following types of employees:
a director/supervisor, an analyst, and a QC chemist.
Laboratory Director/Supervisor - This person should have a minimum of one
year's analytical experience, a degree in chemistry and be a full-time
employee. His duties are:
(a) Schedule all analyses in the laboratory.
(b) Review all QC input to verify "in control" conditions.
(c) Release data to the program manager.
Laboratory Analyst - This person should be employed full-time and trained to
perform with minimum supervision all routine chemical measurements on water
samples. Academic training should include: completion of at least one year
of college chemistry or a laboratory-oriented vocational course. A minimum
of 30 days of on-the-job training in measurements performed by the
organization is also highly recommended. The analyst must be supervised by
an experienced professional scientist—the laboratory director, the
supervisor or a similarly trained individual. His duties are:
(a) Perform instrument calibrations according to SOPs.
-------
Section No. 3
Revision No. 1
Date October 1, 1984
Page 4 of 6
(b) Analyze field samples in the assigned order to meet required schedules.
(c) Notify laboratory supervisor on any problems either in the analysis or
ins trunentation.
Before analysts are allowed to analyze samples, they should demonstrate
their proficiency. Each new analyst should be instructed in instrument
operation and should be allowed to conduct an instrument performance study.
Analyst performance can be evaluated by control charts of critical QC
parameters. If these charts indicate a problem, the analyst should be given
further training.
Quality Control Chemist — This individual should have a minimum of a
bachelor's degree in chemistry, engineering, or mathematics with at least two
years of environmental and one year of QA/QC experience. This position is
under the general supervision of the laboratory director but with access to
the program manager. His duties are:
(a) Implement and monitor the routine application of QC activities in the
laboratory.
(b) Participate in a formal program to train new employees and to update
skills of older employees.
(c) Report to laboratory supervisor status of QC checks.
This is not necessarily a full time position; it may be part time
supplemented by other program duties. The percentage of the chemist's time
dedicated to QC is dictated by the size and complexity of the program.
3.4 Data Management Operations
Persons involved in data aco^iisition, reduction and reporting are the
field operator, the analyst, the data entry staff, the laboratory director,
the program manager, and the QA officer. Their primary duties are:
?ield Operator - His duties are:
(a) Preparation of field data forms.
(b) QC of field data form preparation before shipment to field manager.
-------
Section No. 3
Revision No. 1
Date October 1, 1984
Page 5 of 6
Analyst - His duties are:
(a) Reading or transcription of strip charts
(b) Entry of data to computer.
Laboratory Supervisor - He should perform:
(a) QC check of strip chart reading.
(b) Preparation of data forms.
(c) Review of computer-generated QC information.
(d) Preparation and interpretation of control charts.
Data Processing Personnel - Duties include:
(a) Input of data from data sheets.
(b) Verification of input for keypunch errors.
(c) Update of computer files.
QA Officer - He should perform:
(a) Review of data.
(b) Preparation of QA reports.
(c) Submittal of audit data and recommendations to program manager.
3.5 Designation of QA Responsibilities and Duties
QA Officer - The QA officer, or his designee should:
(a) Review the monthly QC plots generated for each analysis to verify that QC
data are acceptable and to identify any consistent bias trend.
(b) Evaluate the periodic Field Audit Report prepared by the QC chemist to
assess the accuracy of field pH and conductivity measurements of test
samples and to identify needs for corrective action.
(c) Review all QC information presented with each set of analytical data
reported — including data from QC reports.
-------
Section No. 3
Revision No. 1
Date October 1, 1984
Page 6 of 6
(d) Evaluate laboratory'and field operations by conducting program audits and
reporting results to the program manager.
(e) Prepare quarterly reports to management that summarize QA activities and
assess data quality in terms of precision and accuracy trends for both the
field and laboratory operations.
QC Chemist - The QC chemist should:
(a) Introduce blind samples to the laboratory as an independent check on data
quality.
(b) Issue a regularly-scheduled report updating control limits for all
observables.
(c) Evaluate all data prior to its submission to the Laboratory
Director/Supervisor.
Analyst - The analyst should:
(a) Perform all the analyses according to approved Standard Operating
Procedures (SOPs).
(b) Evaluate analytical performance in real time, using readily available QC
information.
(c) Reanalyze the sample if necessary.
(d) Submit data obtained "under control" conditions to the QC chemist.
Data Clerk - The data clerk should:
(a) Enter analytical and field data into the computer.
(b) Check and correct the data input.
(c) Generate reports and graphs of QC information.
3.6 Reference
". Quality Assurance Handbook for Air Pollution Measurement Systems. Vol. v
^ Manual for Precipitation Measurement Systems; Part IJ - Operations and
Maintenance Manual. U.S. Environmental Protection Agency, Research
Triangle Park, NC, EPA-600/4-82-042b (January 1981).
-------
Section No. U
Revision No. 1
Date October 1, 1984
Page 1 of 2
4.0 DOCUMENTATION
4.1 Document Control
A system of document control should be established for all precipitation
monitoring field and laboratory operations. Elements of a precipitation
measurement project subject to document control should include:
1. Field operations and maintenance procedures,
2. Analysis procedures,
3* Auditing procedures,
4. Computational and data validation procedures,
5. Work plan, and
6. Quality assurance plan (if a separate document).
4.2 Internal Documentation
A central file of all data, reports, correspondence, etc. should be
maintained by the program manager and a data file should be kept by the
laboratory. Records in the laboratory file should meet the following
requirements:
1. Records should have identification numbers and be kept in an orderly,
accessible form. Records should include all raw data, calculations, QC
data, and reports.
2. Data in laboratory records should include:
A. Sample identification number,
3. Sample type,
C. Date sample received in laboratory,
D. Collection data (time, date, volume, etc., if laboratory
responsibility),
E. Date of analysis,
-------
Section No. 4
Revision No. 1
Date October 1, 1984
Page 2 of 2
F. Name of analyst,
G. Results of analysis (including all raw data), and
H. Name of person receiving the analytical data.
3. The laboratory should have a sample-tracking system starting from receipt
of sample through to the completion of analysis; this should include:
A. Sampling information records (e.g., field data forms) with dates,
time, site location, sample amount, etc.,
B. Bound notebooks with numbered pages,
C. Computer printouts or report forms verified against laboratory records
before data release.
4.3 QA Documentation
QA reports should be submitted regularly to the program manager by the
QA officer. They should include:
1. Periodic assessments of measurement data accuracy, precision, and
completeness;
2. Results of performance audits;
3. Results of systems audits; and
*. Significant QA problems with documentation of remedial action taken or
recommended solutions.
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 1 of 16
5.0 SITING
In the design of a monitoring program, the program objectives and
network station density must be considered. Ideally, a complete description
of the site and its surroundings (out to 20 km) should be given so that the
user of the data could decide what the station is measuring or what the data
represent. However, for convenience, simplified site categories (urban,
remote, regional or rural) are used here.
Network station density, which helps define the spatial resolution of
the data obtained, is determined by program objectives, area meteorology and
topography, and budgetary constraints. Thus, it is difficult to design a
network for all users of this QA manual. However, some specific guidance can
be found in References 1 and 2. Points to consider for setting up a network
monitoring a large geographical area are described below.
5.1 Network Design Considerations
In the design of a precipitation monitoring network, stations are
located according to the objectives of the program:
1. Measurement of baseline (remote area) precipitation,
2. Measurement of representative regional (rural) precipitation, or
3. Measurement of urban area (local impact pollutant emission sources) on
precipitation.
These three are generally differentiated by expected concentration
levels. The background or remote station should show contamination primarily
due to natural processes; the regional station would be affected primarily
by long-range transport; and the urban site would show high concentrations
due to a polluted local environment. In the selection of station locations,
it is necessary to have detailed information on location of emission sources,
regional variabilities of ambient pollutant concentrations, precipitation
amounts, prevailing winds, and meteorological data. Thus, the design of a
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 2 of 16
network needs to address details such as the number, location, and type of
ipling stations and equipment to be used.
5-2 Site Selection Criteria
The variability and the long-range transport of pollutants make it
difficult to determine whether a site is collecting precipitation data
representative of any given area. In addition, the transport of air
pollutants and their resultant concentrations in precipitation are
complicated by topography. For example, in mountainous regions,
precipitation tends to be unevenly distributed due to topographical lifting
of clouds and deflection of air flows.
To optimize site locations for different station categories, the
following selection criteria are suggested.
5.2.1 Baseline Station - The station should be in a location where the
effects of human activities are negligible. Ideally, the station should meet
as many of the following criteria as possible:
1. The station should be in an area where no significant changes in land-use
practices within 100 to 1000 km (depending on prevailing wind direction)
from the station are anticipated during the study period.
2. The station should be at least 50 km from major population centers, major
highways, industries, air routes, and large natural sources (e.g.
geysers); it should be in remote, uninhabited, or sparsely inhabited
areas. If an isolated island is used, data corrections for sea salt
aerosols should be made.
3. The site should not have a history of frequent local natural phenomena
such as forest fires, dust and sand storms, or volcanic activities.
i. The site should have provisions, e.g. power, for setting up a
meteorological and aerometric monitoring station (3,^,5,6).
5. The site should be readily accessible on a flat or gently sloping terrain
(less than 20°) and sheltered from strong winds.
3.2.2 Regional Station - Ideally, site selection criteria for a regional
sonitoring network include criteria 3-5 for baseline stations as well as the
following:
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 3 of 16
1. The general area should be free from influences of large anthropogenic
sources such as cities or towns; industrial, sewage or power plants;
refineries, commercial areas, and airports; and large local natural
sources. Such pollutant sources should be distant enough for pollutants
not to unduly affect the precipitation chemistry. If the site location
must be near a large source (within 50 km), the station should be in the
prevailing upwind direction from the source.
2. The site should be at least on kilometer distant from local sources such
as houses, farmlands, orchards, marshes and swamps, landfills, and roads.
3. If stations are near pollutant sources, the site location should avoid
undue influence by a single pollutant source.
4. If an area is characterized by a common type of land use, the collector
can be near the common pollution source.
The selected site can be evaluated for representativeness and for local
contamination by installing a temporary grid of neighboring satellite
samplers around it. Sampling procedures for this temporary network should be
comparable to those at the original site. The permanent site should be
selected after evaluating the results of this temporary network.
5.2.3 Orban or Local Station - Ideally, to study urban areas or the effects
of point sources on precipitation, an array of monitoring stations should be
near the source in the area of interest. The stations should not be near
other sources. For studying the local effect due to a single point or area
source, the average interstation distance should be of the order of several
kilometers. However, actual station density and interstation distance should
be decided by the desired spatial resolutions,
5.3 Sampler and Rain Gauge Siting Criteria
Placements of precipitation samplers and rain gauges should assure that
the site collects unbiased samples. Samplers and rain gauges should stand
far enough from trees, hills, and other obstructions to minimize interference
with sampling. No object, even if smaller than the collector should be
within a few meters of the collector, and no object should shade the
collector. An open, flat, grassy area, surrounded by trees no closer than
IGOm would be an ideal site.
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page i» of 16
Ideal criteria for placement of samplers and rain gauges are:
1. The horizontal distance between a large obstruction and the collector
should be at least twice the obstruction height, or the top of an
obstruction as viewed from the collector should be less than 30° above the
horizon.
2. The horizontal distance between collocated samplers, or sampler and rain
gauge should be greater than two meters.
3- The collector should be far from mobile pollution sources. Routine air,
ground, or water traffic should not come within a 100m of the collector
site.
4. The distance between any overhead wires and the site must be great enough
not to affect the samples.
5. The collector should be at least 100m from open storage of agricultural
products, fuels, or other foreign materials.
6. The ground surface around the collector should be firm and have a grass
cover or gravel.
7. Wet/dry collectors should be oriented parallel to the prevailing wind
direction during precipitation events, with the wet bucket upwind of the
dry bucket (so that the dry bucket does not obstruct the wet bucket).
8. The rain gauge should be positioned parallel to both the collector and the
direction of the prevailing wind during precipitation events. If the
gauge has an access door to a recorder, weighing or drive mechanism, the
door should be kept closed, and the gauge should be mounted with the door
facing away from the wind.
The distance between obstructions such as growing trees and newly
erected structures and the collector should be checked periodically.
5.1 Station Identification and Classification
All stations must be identified by documentation of site characteristics
to facilitate evaluation of data generated from samples taken at that site.
Typically, the site identification record should contain:
1. Data acquisition objective (baseline, trend, or research monitoring).
2. Station location (address, map coordinates, elevation, etc.).
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 5 of 16
3. Type of station (remote, regional, or local type; and if it is primarily
an industrial, agricultural, forest, urban site, etc.).
4. Instrumentation checklist (manufacturer,1 model number, measurement
technique, etc.).
5. Important pollutant sources (point and area sources; their pollutants and
emission concentrations, proximities, etc.).
6. Topography description (trees, hills, valleys, bodies of water and type,
size, proximity, orientation, of water body, etc.). Photographs of the
monitoring site covering a 360 view from the precipitation collector
should also be taken.
7. Site diagram properly scaled (equipment configuration, trees, man made
structures, access road, electrical power lines, etc.).
All monitoring stations should be properly identified and classified as
described below using a clear, concise format. This can be accomplished
using forms similar to those at the end of this section or those developed by
other networks such as NADP and NTN.
Class I.
1. Station satisfies all siting criteria (Section 5.3),
2. On-site instrumentation includes
automatic precipitation collector,
recording rain gauge,
pH and conductivity meters,
meteorological sensors (windspeed and direction), and
aerometric analyzers (S02 and NO/NOX).
Class II
1. Station satisfies all siting criteria (Section 5.3).
2. On-site instrumentation includes
automatic precipitation collector,
a recording rain gauge, and
pH and conductivity meters.
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 6 of 16
Class III
1. Station satisfies all siting criteria (Section 5.3).
2. On-site instrumentation includes
automatic precipitation collector,
nonrecording rain gauge, and
pH and conductivity meters.
Class IV
1. Station does not satisfy all siting criteria (Section 5.3)*
2. On-site instrumentation identical to Class I stations.
Class V
1. Station does not satisfy all siting criteria (Section 5.3).
2. On-site instrumentation identical to Class II stations.
After initial classification, an on-site visit should be made by the QA
coordinator to evaluate and certify each monitoring station as soon as
possible after the start of operation to assure the quality of monitoring
data. Classes IV and V must be only temporary, since they are in
noncompliance with siting criteria. The program manager (or a designee) must
be sure that siting deficiencies are corrected within a reasonable time.
Most deficiencies should be corrected within 30 days, but for serious
deficiencies, a schedule should be established for compliance attainment.
•hen corrections are made, documentation should be provided to the QA officer
and program manager and the station classification should be changed by them.
All sites should be reevaluated yearly to verify that they remain in
compliance with the siting criteria. All aerometric and meteorological
instrumentation should conform to standard ambient monitoring guidelines.
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 7 of 16
5.5 References
1. WHO Operational Manual for Sampling and Analysis Techniques for Chemical
Constituents in Air and Precipitation, World Meteorological Organization
Pub. No. 299 U97477
2. Site Selection and Certification, National Atmospheric Deposition Program
(1970).
3. Quality Assurance Handbook for Air Pollution Measurement Systems, Vol. II
- Ambient Air Specific Methods, EPA-600/U-77-027a, Research Triangle Park,
NC (1977).
4. Guide to Meteorological Instrument and Observing Practices, World
Meteorological Organization Pub. No.8, TP8 (1971).
5. Ambient Monitoring Guidelines for Prevention of Significant Deterioration
(PSD), EPA-450/1-78-019 (1978).
6. Quality Assurance Handbook for Air Pollution Measurement Systems, Vol. I -
Principles, EPA-600/9-76-005, Research Triangle Park, NC (1976).
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 8 of 16
SITE DESCRIPTION REPORT
Data Prepared
Reason (New site, change, revision)
A. DATA ACQUISITION OBJECTIVE (Description)
B. SITE CATEGORY
1. Station Identification 2. County 3. State_
4. Latitude 5. Longitude 6. Elevation
7. Station environment: remote rural
suburban urban commercial
Indus trial
8. Available USGS Topographical Map (Yes, No) (circle)
9. Revision year ..._.
10. Scale (1:24,000 preferred) __
C. SITE ADMINISTRATION
1. Name of official position.
Mailing address
(number and street)
(city) (state) (zip) (Phone)
2. Name of Program Manager Title
Mailing address
(number and street)
(city) (state) (zip) (Phone)
3. Name of Site Operator
Mailing address
(number and street)
(city) (state) (zip) (Phone)
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 9 of 16
D. SITE INSTRUMENTATION
1. Precipitation Collector Type: Automatic Non-automatic
Manufacturer
Model Serial No.
Diameter (I.D.) of
Sample Bucket (cm)
2. Raingauge:
Recording Nonrecording
Types Weighing Tipping Bucket Other
Manufacturer
Model Serial No.
Funnel Size (cm)
3. Nitrogen Oxide Monitor:
Automatic Data Acquisition Stripchart recording
Types Cbemiluninescent Other
Manufacturer
Model Serial No.
Sulfur Dioxide Monitors
Automatic Data Acquisition Stripchart recording
Types Fluorescent Other
Manufacturer
Model Serial No.
5. Other Aerometrlc Analyzers:
Sensors
Recording Nonrecording
Type Serial No.
Manufacturer Model
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 10 of 16
Other Meteorological Instrument:
a. Wind Speed Sensor: Type
Manufacturer
Model
Serial No.
Wind Direction Sensor: Type
Manufacturer
Model
Serial No.
Temperature Sensor: Type
Manufacturer
Model
Serial No.
d. Solar Radiation Sensor: Type
Manufacturer
Model
E. ANALYTICAL INSTRDMENTATION:
Serial No.
1. pH Meter: Type
Model
Temp Compensated
Serial No.
Conductivity Meter: Type
Manufacturer
Model
Balance:
Temp. Compensated
Serial No.
Type
Manufacturer
Model
Serial No.
Type of low conductivity water available:
dionized distilled
bottled
Laboratory Space: Location _
Good Fair
Special problems
Poor
-------
Section No. 5
Revision No. 1
Date October 1, 198U
Page 11 of 16
F. SITE DOCUMENTATION
1. Identify site location and major sources on local topographical
map (attach to report).
2. Sketch a map to document the environment within a 1/2 mile radius
of the site. Include the following information on the drawing where
applicable.
Site diagram and equipment configuration High power lines
at center of drawing Topographical features
Roadways with names (paved and unpaved) (valleys, hills, etc.)
Parking areas (paved and unpaved) Bodies of water
Stationary sources (NEDS*) North direction
Buildings (number of stories) Undeveloped land
Tree lines or clusters (ground cover)
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 12 of 16
3. Site photographs, labelled to indicate the four compass directions.
Looking Morth Looking South
Looking West Looking East
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 13 of 16
G. POTENTIAL SOURCES OF INTERFERENCE OR CONTAMINATION AT THE SITE
1. Within 30 m of the SAMPLER identify all objects that are taller
than the sampler.
a. Structures
Type Use
Height (m) Distance (m) Direction
Type Use
Height (m) Distance (m) Direction
b. Trees
Species M*** Height (m)
Distance (m) Direction
Species Max. Height (m)
Distance (m) Direction
Species Max. Height (m)
Distance (m) Direction
c. Other (e.g., overhead wires, masts, etc.)
(1) Object
Height (m) Direction Distance (m)
(2) Object
Height (m) Direction Distance (m)
(3) Object
Height (m) Direction Distance (m)
(4) Object
Height (m) Direction Distance (m)
(5) Object
Height (m) Direction Distance (m)
d. Is public road access to site in
summer Good Fair Poor
and winter Good Fair Poor
e. Type of public road surface?
f. How far from the public road will collector be sited? (m)
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 1U of 16
g. Is there other than public road access to the site? If so,
please describe.
h. How close can a vehicle approach the collector? (m)
i. How is site secured against vandalism, etc.?
j. Are there any special logistical problems? Please describe.
Within 200 m of the site identify:
a. Predominate land use in the area:
Use 1 , % Ose 2 , % All Others
(cultivated, orchard, lawn, pasture, forest, water, swamp,
residential)
b. Unpaved roads and parking areas:
Dnpaved road: Distance (km,m) Direction from sampler _
Traffic: Heavy , Medium , Light
Unpaved road: Distance (km,m) Direction from sampler _
Traffic: Heavy , Medium , Light
Parking lot: Distance (km,m) Direction from sampler _
Unpaved Surface material
Use: continuous intermittent car volume
large truck volume __
Parking lot: Distance (km,m) Direction from sampler _
Unpaved Surface material __
Use: continuous intermittent car volume
large truck volume
-------
Section No. 5
Revision No. 1
Date October 1, 1981
Page 15 of 16
Within 1 km of the site identify significant agricultural operations
such as feedlots, dairy barns, cultivated fields, etc.
Type
Distance (km,m) Direction from sampler
Type
Distance (km,m) Direction from sampler
4. Within 10 km of the site identify transportation related sources.
a. Main highways or expressways: Traffic volume
Direction from sampler Distance (km,m) Route
Main highways or expressways: Traffic volume
Direction from sampler Distance (km,m) Route #
b. Other paved roads: Distance (km,m) Direction from sampler
Traffic: Heavy , Medium , Light Traffic Volume
Other paved roads: Distance (km,m) Direction from sampler
Traffic: Heavy , Medium , Light Traffic Volume
c. Lake/river or rail traffic:
Distance (km,m) Direction from sampler
barge , lake steamer , ocean vessels , rail
Traffic: Heavy , Medium , Light Traffic Volume
Airports Distance (km,m) Direction from sampler
Traffic: Heavy , Medium , Light
Other transportation related sources
Within 60 km of the site identify stationary sources:
a. Power plant(s):
Name
Distance
Fuel
Name
Distance
Fuel
(km) Direction from sampler
Electrical capacity
(km) Direction from sampler
Electrical capacity
(KW_, MW_)
(KW_, MW_)
-------
Section No. 5
Revision No. 1
Date October 1, 1984
Page 16 of 16
Industry:
Type/product
Distance (km) Direction from sampler
Comment
Type/product
Distance (km) Direction from sampler
Comment
Other stationary sources:
Type/product
Distance (km) Direction from sampler
6. Within 60 km of the site identify significant area sources:
Type /product :
Distance (km) Direction from sampler
H. PERSON WHO FILLED OUT THIS FORM
Name
Position
Phone #__
Address
Affiliation
Signature __
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 1 of 14
6.0 FIELD OPERATIONS
6.1 Facilities
All field facilities should be accessible and should have a clean
shelter, a level table, deionized or distilled water (conductivity less than
2.0 ymho/cm), and a sink or drain. A refrigerator and a 110v AC outlet are
highly desirable. The former is necessary to preserve samples until they are
shipped. For weekly sampling, the samples should be shipped within 24 hours
of collection.
The precipitation collector and the recording rain gauge can be run on
either 12v DC storage batteries or 110v AC. Both means have advantages and
disadvantages. If large current usage is required (e.g., for heating),
batteries are not recommended.
6.2 Method Selection
6.2.1 Precipitation Collectors and Rain Gauges - The rain gauge and the
precipitation collector serve different functions. The rain gauge measures
the amount of precipitation. The precipitation collector collects a sample
for chemical analysis. The two devices are not interchangeable.
Precipitation Collectors — The precipitation collectors should have the
following characteristics:
(a) Reliable automatic operation - collector opens at start of precipitation
and closes after event ends.
(b) Prevent contamination of wet sample by dry deposition.
U) Minimize evaporation.
(d) Inert to sample constituents of interest.
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 2 of 14
Collectors, meeting these criteria are available. The most satisfactory
collector is one based on the design of the Department of Energy's Health and
Safety Laboratory (HASL)(1). However, the sampler is not very efficient for
collecting snow.
The first three criteria are met by means of a precipitation sensor and
a motor-driven, tight-fitting lid for the collector container. When the grid
and plate of the sensor are shorted by a drop of water, the motor is
activated, lifting the lid from the container. The sensor has two heating
circuits. One operates to melt snow or ice (on the sensor) when the
tanperature is below 2°C. When the lid lifts off the sample bucket, the
other circuit heats the sensor to about 55°C to increase the rate of
evaporation of water from the sensor. (Heat hastens the sealing of the
sample by the lid after precipitation ceases, thus minimizing the exposure
times to dry fallout and to snow blowout from the collector.) A seal between
the container and the lid is achieved by a plastic foam gasket under the lid
and by a spring load. However, in strong winds the lid can wobble and may
permit contaminants to enter the sample bucket.
To ensure inertness to major constituents in acid precipitation,
polyethylene sample buckets (1,2) are usually used, because of their low
cost, durability, and availability, high-density linear polyethylene
containers can be used for collecting and shipping samples. Glass or metal
can affect inorganic sample integrity, but should be used if organic
compounds are being monitored.
Subevent or sequential samplers separate samples on either a volume or
time-of-collection basis, but the same requirements as above hold. In
volume-based sampling, precautions must be taken to minimize mixing or carry
over of samples. For time-based sampling, there should be provision for
overflow during heavy rainfalls. For sequential sampling, the time
corresponding to each subevent specimen must be known for correlations with
other data. Sequential samplers vary greatly in sophistication from a series
of connected bottles to completely automated and electrically operated
designs (3-8). One of the latter is available commercially(8).
Hain Gauges — A standard, a rain gauge is used to record the quantity of
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 3 of 14'
precipitation. Recording rain gauges are of two basic operational
designs—the weighing type and the tipping bucket type. Both types of gauges
should be capable of measuring precipitation to approximately 0.25 mm (0.01
in). Weighing gauges measure within +0.76 mm (0.03 in), and their accuracy
(1f of full scale) is independent of precipitation rate. The accepted
accuracy for tipping bucket gauges is 1} for precipitation rates of 25 mm/h
(1 in/h) or less, 4J for 75 mm/h (3 in/h), and 6% up to 150 mm/h (6 in/h).
Rates are measured either directly or derived from the cumulative data.
The recording weighing rain gauge should have an event marker pen to
indicate when the collector is open and closed. Such a pen typically rises
from its baseline when the collector lid opens, and remains actuated until
the lid closes at which time the pen falls to its baseline position. To
prevent the event marker pen from interfering with the weighing trace pen,
the two must be offset on the time axis. Thus only one pen can be set at the
correct time, and care must be taken to use the correct beginning and ending
times. Since the operator is seldom present to observe collector operation
during an event, the event pen marker is an invaluable aid in indicating
sampler malfunction or a power outage.
For windy areas and especially where snowfall constitutes more than 20%
of the mean annual precipitation, an Alter-type windshield should be
installed around the rain gauge. The shield should be level and its top
should be 1.3 cm (0.5 in) above the level of the gauge collecting orifice.
In addition, the shield should be concentric with the gauge. Installation
instructions for the improved Alter-type windshield can be obtained from the
U.S. Weather Bureau (9). Alternatively, a Nipher-shielded snow gauge can be
used for snow depth measurement.
6.2.2 pJH and Conductivity Apparatus, Temperature Probe - The pH and
conductivity of a 20 ml aliquot of the precipitation sample should be
withdrawn and measured at the field station as soon as possible after the
sample is collected but only after it is at the temperature of the
calibration solutions. The sample and the aliquot should be protected from
contamination during this time.
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 4 of 14
pg Apparatus ~ The pH meter and electrode must'be capable of measuring with
a precision within + 0.03 unit and with an accuracy of 0.05 unit. Meters
should have an impedance of at least 10 ohms. A combination glass and
reference electrode of the nongel type with an unprotected membrane bulb is
preferred. The combination electrode requires less sample and fewer washings
Uan two separate electrodes; electrodes with shielded bulbs are more
difficult to clean, and are thus more prone to yield errors. When a new
electrode is obtained, it should be equilibrated overnight and stored in the
solution recommended by the manufacturer.
Conductivity Apparatus — The conductivity meter and cell must have a
measurement range of 0 to 1000 umho/cm, a precision of +0.5/1 of range, and an
accuracy of ±1.0% of range. The range most frequently used is 10-100
ysbo/cm. A temperature-compensated cell with a cell constant of 1.0 is
preferred.
Teaperature — A thermistor, thermocouple, or thermometer can be used to
measure solution temperature. The temperature probe must have an accuracy of
at least 1°C and a precision of +p.5°C.
6.2.3 Balance or Graduated Cylinders - The amount of precipitation sample
collected can be measured with a balance or with graduated cylinders. Since
tie density of rain samples is approximately 1.0 g/ml at 20°C, the weight of
tte sample (in grams) can be taken to equal its volume (in ml). The
measurement of sample volume by graduated cylinders increases the chance of
contamination, so a balance is preferred.
The precipitation sample volume can be compared to that recorded by the
rain gauge (e.g., with the Aerochem Metrics collector, 16.2.g of
sanple = 0.01 in. = 0.25 mm) to calculate the collection efficiency of the
sanpler (Section 8.4.2). Differences between the field and the laboratory
weight values (container plus sample shipped to the laboratory) indicate
either loss of sample during shipment or a weighing error.
For weekly sampling, the balance should have a capacity of 20 kg with a
accuracy of at least +10 g. Triple beam balances meeting these requirements
sre readily available. The balance should be kept on a sturdy, level table,
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 5 of 14
and it should be zeroed daily before weighing. The balance should be checked
initially in the laboratory with 1.0 and 5.0 kg test weights. For event and
sequential sampling, a 2.6 kg capacity triple beam balance with a sensitivity
of at least +1 g should be used.
Graduated cylinders are not recommended, but if they are used, they
should be plastic. To measure within +10 ml, the graduated cylinder should
not be more than 1000 ml in capacity. The graduated cylinder should be
checked before use for accuracy in the laboratory by weighing known volumes
of water and comparing the results to the volume measurement after converting
the weight to volume by multiplying by the density of the water.
6.3 Acceptance Testing
All precipitation collectors, rain gauges, pH and conductivity meters,
and electrodes should be functionally tested before they are used in the
field. Acceptance tests should cover the essential operations of the
instriments. Collectors and rain gauges should be tested on site. It is
convenient to test meters and electrodes in a central laboratory, where
common standards and procedures are available. Procedures for this
acceptance testing are detailed below. Procedures for carrying out these
tests are in the Operations and Maintenance Manual (10).
6.3.1 Precipitation Collectors and Rain Gauges - Collector tests should
include: •
(a) sensor heating and actuating of the lid,
(b) sensor cooling and return of the lid,
(c) sensor temperature attainment when the lid is raised,
(d) sensor temperature with lid closed when ambient temperature is below
freezing, and
(e) observation of lid cycling and sealing.
Rain gauge tests should include:
(a) sensitivity and accuracy,
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 6 of 14
(b) clock function, and
(c) pen and recorder functions.
6.3.2 pH and Conductivity Meters - Erroneous pH measurements may not be
revealed by conventional two-point calibration procedures. The majority of
standard buffer solutions have a similar total ionic strength. However,
precipitation samples differ in ionic strength from the standardizing
buffers, which might introduce a bias in the measurement.
For these reasons it is advisable for the electrode systems used in acid
precipitation .pH measurements to be used exclusively for this task and to be
•onitored with respect to their performance with known reference solutions
intended to simulate the unknown samples. However, for documentation
purposes, calibration with the certified buffer solution is essential
(11,12).
For each of the tests indicated below a total of ten solutions are
measured, and an average value and a standard deviation are calculated.
These teats should include:
(a) Evaluation of conductivity meter and cell using 0.0003M KC1. System is
acceptable if within 2f of 44.6 ymho/on at 25 C, and the standard
deviation is less than 2%.
(b) Evaluation of field pH meters using a certified laboratory pH electrode.
System is acceptable if the average pH and standard deviation are within
0.03 pH unit of the documented values obtained using two certified buffer
solutions.
(c) Evaluation of pH electrodes using pH electrode reference solution. System
is acceptable if the average pH is within 0.1 pH unit of the known value
and the standard deviation is less than 0.05 pH unit, when using reference
solutions which simulate the ionic strength of precipitation samples.
6.4 Sampler and Rain Gauge Installation and Operation
The precipitation collector should be mounted on the ground so that the
rim of the mouth or opening is level and at least 1 m above the ground, and
it should be properly anchored against strong winds (cement blocks can be
used as weights). The collector may be shielded from the wind, but it should
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 7 of 14
not be put in an area where there will be excessive turbulence caused by the
shield or where there are obstructive objects such as trees and buildings
close by. In the winter, loss of snow from the collector can occur due to
blow out. Also, in windy areas the bucket should be secured to the sampler
by means of a spring or elastic tie down cord hooked to the bucket handle and
collector table.
The sampler installation and operation are described in the
•anufacturer's instructions and in the Operations and Maintenance Manual
(to). The precipitation collector requires no calibration, but proper
functioning should be checked frequently. . The rain gauge should be
calibrated according to the manufacturer's instructions after installation
and checked at least annually.
The rain gauge should be mounted on a firmly anchored support or base,
e.g., cement blocks, so that the funnel rim is level and at about the same
height as the collector bucket rim to enable comparisons of collection
amounts between the two. The gauge level can be checked with a carpenter's
level placed at two intersecting positions. The gauge mouth should be high
enough not to be covered by snow.
6.4.1 Routine Checks on Collector, Rain Gauge and Site - Some tests should
be carried out routinely on the precipitation collector and the rain gauge.
The detailed procedures for these tasks and a checklist for conducting them
are provided in the Operations and Maintenance Manual (10).
6.4.2 Corrective Action - Any indication of a malfunction should be recorded
in the logbook, and the field manager should be notified. An attempt to
diagnose and correct the problem should be made with the aid of the
Operations and Maintenance Manual (10) as soon as possible. If the problem
cannot be corrected, the field manager or equipment manufacturer should be
asked for advice and direction. Any action taken should be recorded in the
logbook.
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 8 of 14
6.5 Sampling Methodology
This section gives the methodology for sample collection, handling,
measurement, and preservation. The procedures used to accomplish each of the
above-mentioned tasks are given in the Operations and Maintenance
Manual (10).
6.5.1 Sample Collection and Schedule - The choice of sampling schedule
depends on the program objective and the available funds. To correlate
precipitation data with aerometric and/or meteorological data, event or daily
sampling must be used. To measure the amount of deposition and/or its
effects* a weekly sample may be sufficient. Sampling periods longer than one
week are not advised because significant changes in sample chemistry can
occur as the sample stands in the collector.
6.5.2 Handling of Plastic Containers - Treatment of plastic containers
depends on the species to be measured, the container's previous use, and its
cleanliness. In most cases, the cleaning of the bucket, lid and gasket
should be done in the laboratory.
The container should be capped with a clean lid and kept in a plastic
bag until immediately before use, and it must be resealed immediately after
use. When a bucket is to be returned to the laboratory with or without a
sample, it should not be sealed with its original lid. Instead, the sample
should be covered with the lid from a new bucket which is replacing the old
bucket in the collector. Thus the chance of contamination from the stored
lid is eliminated.
An alternative approach used by the Canadians (13,14) that eliminates
the washing of buckets and lids is the use of laminated nylon-polyethylene
bags inserted into the buckets. When the bag is placed in the bucket,
plastic gloves are worn and no contact with the interior of the bag further
than 7.5 cm inside is made. A fold of 7-10 cm is extended down over the
outside surface of the bucket. When removing the bag containing a sample,
the bag is grasped by the fold. The bag is sealed below the fold with a
cable tie (13) or by heat (13,14). Up to 500 ml of sample is emptied from
-------
Section Ho. 6
Revision No. 1
Date October 1, 1984
Page 9 of 14
the bag into a polyethylene bottle by cutting a prewashed and dried corner
off the bag with a clean scissors and pouring the sample directly into the
bottle.
6.5.3 Sample Handling - The sample bucket must be checked for precipitation
at the time and frequency set by the schedule. However, if a snowfall
occurs, the sample should be removed as scon as possible and replaced with a
new bucket to minimize snow loss by overflow and blow out.
5.5.4 Sample Preservation and Storage - Sample degradation can occur due to
chemical interactions (e.g., with particulates or gases) or to biochemical
reactions. In addition, losses of potassium and some trace metals by
adsorption on polyethylene walls have been reported (15). Preservation of
sample integrity can be maximized by filtration, sealing, and storage in the
dark at about 4°C. After pH and conductivity measurements, filtration should
be done with a 0.45 urn organic membrane filter (16), if inorganic species are
to be analyzed. Although biocides such as toluene or chloroform might be
effective in stopping biochemical activity, they may interfere in the various
aeasurements or analyses (2) and thus should not be added to the sample. If
certain species must be preserved, an aliquot of the sample can be mixed with
a preservative in a separate container.
6.6 Field Measurements
Field measurement of pH, specific conductance, and temperature are
discussed in general terms in this section; detailed procedures are given in
the Operations and Maintenance Manual (10). Sample pH and conductivity are
aeasured in both the field and the laboratory to detect sample changes and
errors in measurement. Results of field measurements should be recorded on a
Field Data Form that will accompany the sample to the laboratory.
5.6.1 £H Determination Method - Since rain samples generally have pH values
between 3.0 and 6.0, the pH meter should be calibrated with pH 3.0 and 6.0
standard buffers. For other less acidic samples, pH 4.0 and 7.0 buffers
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 10 of 14
should be used, and for basic samples, pH 5.0 and 8.0 buffers should be used.
Each station should receive the needed calibration buffer solutions from the
central laboratory. The stations should notify the laboratory when the
buffer supply is nearly exhausted or if it appears to be contaminated by
algae growth.
The pH meter should be calibrated before and after each measurement or a
series of measurements at one time. If the initial and final calibrations
have changed by more than 0.02 unit, the measurements must be repeated. If
this change reoccurs, a problem exists with the apparatus, and it should be
remedied.
Each electrode should be assigned an identification number so that its
calibration can be traced. To check for electrode problems (aging and loss
of sensitivity), each site should periodically receive from the central
laboratory a polyethylene bottle of electrode reference solution with pH and
conductivity similar to those of rain samples. The pH of this sample is
aeasured and reported to the central laboratory. The field pH value should
agree within +0.10 pH unit of the assigned value if the electrode is
operating properly.
Electrode performance can also be determined by observing the time
needed to attain a stable reading, where a stable reading is defined as a
constant pH value (+0.02 units) for a period of 1 min. The time required to
attain stability should be less than 5 min. for a well-behaved electrode.
Results of these tests should be guides for the measurement technique and the
equilibration time to be used for precipitation sample measurement. If an
electrode test at any time exhibits out-of-control behavior (as indicated by
the above criteria), the electrode and/or solution should be replaced.
5.6.2 Specific Conductance Determination Method - The conductivity (or
resistance) of a solution varies with electrode area and spacing as well as
with temperature and ion concentration. Therefore, the measuring apparatus
has to be calibrated to obtain the cell constant or to adjust the meter. For
calibration, a KC1 solution of known conductivity should be used. The
temperature of the KC1 standard and the precipitation sample should be the
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 11 of 14
sane. For rain samples, a 0.00050M KC1 solution is ideal. All conductances
should be reported in umho/cm corrected to 25 C.
The conductivity apparatus should be calibrated before and after each
measurement or series of measurements at one time. If a change of more than
5$ occurs, the measurements should be repeated. If the drift persists, a
problem exists with the apparatus. In general, stable values occur in about
30 sec. Conductivity of the samples can be measured on the same aliquot used
for pH. IF THIS IS DONE, THE CONDUCTIVITY MUST BE MEASURED BEFORE pH TO
AVOID ERROR DUE TO SALT CONTAMINATION FROM THE ELECTRODE.
The conductivity cell generally has few problems. However, the working
conductivity standard (0.0005M KC1) may degrade slowly or become
contaminated. To minimize errors due to changes in the calibration standard,
the working solution should be replaced quarterly. When a new working
standard is received, it should be checked against the old working standard,
and the two values should agree within 4.0?. If they do not, the laboratory
which supplied the standard to the site should be notified.
Conductivity standards should be sealed and stored in a refrigerator to
minimize changes. Generally, changes of less than 2} monthly may be ignored;
if greater than 2%, the field values can be corrected for the larger changes
by prorating with time in a linear manner. Another means of evaluating the
working conductivity standard is the use of unknown quality control test
samples submitted periodically from the laboratory to determine the accuracy
and precision of the station's specific conductance measurements. If the
laboratory finds that the station's conductivity measurement differs from the
laboratory's by more than 5%, the laboratory should inform the field and
quality assurance personnel and should replace the old conductivity standard
or the meter.
5.6.3 Temperature Measurements - Each field thermometer or temperature probe
should be assigned an identification number so that it will be possible to
trace its certification. The temperature probe should be washed and dried
before the solution temperature is measured. It should never be placed in a
solution on which pH and conductivity measurements will subsequently be made.
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 12 of 14
The central support laboratory should maintain and store an
IBS-calibrated thermometer as a primary standard, and one field thermometer
should be certified against this as a secondary standard. All field
thermometers or temperature probes should be calibrated against the certified
(secondary) thermometer in a circulating water bath in the 0° to 25°C range.
After initial calibration, the temperature probes should be recalibrated at
least once per year.
6.6.4 Gravimetric Measurements - For weighing rain buckets, the balance
should be in a room free from drafts and on a table that minimizes
ribrations. The balance should be level.
Before being shipped to the field, each balance should be calibrated
with MBS traceable weights in the central support laboratory. Annually, a
full calibration should be performed by weighing two NBS traceable weights
(1.0 and 5.0 kg) on the field balance. The actual reference weight, measured
weight, and weight difference should be recorded. The rain gauges can be
calibrated using a set of weights generally available from its supplier.
6.7 Documentation
All data, observations, and changes or modifications must be dated and
documented on data forms and/or in logbooks in triplicate and duplicate,
respectively (carbon paper may be used). One copy of each should be kept in
the station records, and another shipped with the sample; the third copy (of
the data form) should be mailed to the laboratory separately from the sample
to help trace a missing sample.
Samples must be labeled so that they can be readily and correctly
aatched with their data forms. The label should contain station, date, and
sample weight marked with a pencil or a ball point pen so that it is legible
if it should gets wet.
Forms for use in documenting the data are provided in the Operations and
Maintenance Manual(10).
-------
Section No. 6
Revision No. 1
Date October 1, 1984
Page 13 of 14
6.8 References
1. J.N. Galloway and G.E. Likens, Water, Air and Soil Pollut. 6, 241 (1976).
2. J.N. Galloway and G.E. Likens, Tellus 30, 71 (1978).
3. H.B.H. Cooper, Jr., J.A. Lopez, and J.M. Demo, Water, Air Soil Pollut. 6,
351 (1976).
4. D.F. Gatz, R.F. Selznan, R.K. Langs, and R.B. Holtzman, J.
Appl. Meteorol. JO, 341 (1971).
5. J.J. Morgan and H.M. Liljestrand, "The Measurement and Interpretation of
Acid Rainfall in the Los Angeles Basin," California Institute of
Technology Report, No. AC-2-80 (February 20, 1980).
6. J.K. Robertson, T.W. Dolzine and R.C. Graham, "Chemistry and Precipitation
from Sequentially Sampled Storms," EPA report to be published.
7. G.S. Raynor and J.P. McNeil, "The Brookhaven Automatic Sequential
Precipitation Sampler," BNL-50818, Brookhaven National Laboratory (January
1978); Atmos. Environ. J3, 149 (1979).
8. P.B.S.K. Associates, P.O. Box 131» State College, PA. 16801, Bulletin
177.6801.
9. Installation Instructions for Improved Alter-Type Windshield,
U.S. Dept. of Commerce, Weather Bureau, Instrumental Engineering Division
(November 1957).
10. Quality Assurance Handbook for Air Pollution Measurement Systems - Vol. V
- Manual for Precipitation Measurement Systems. Part 13! - Operations and
Maintenance Manual. 0. S. Environmental Protection Agency, Research
Triangle Park, N.C., EPA-600/4-82-042b (January 1981).
11. G. Masinenko and W.F. Koch, "A Critical Review of Measurement Practices
for the Determination of pH and Acidity Atnospheric Precipitation,"
NBSIR 84-2866. To be published by Environment International (1984).
12. J.A. Illingworth, "A Common Source of Error in pH Measurements,"
Biochem. J., 195. 259-262 (1981).
13. Acidic Precipitation in Ontario Study, Technical and Operating Manual
APIOS Deposition Monitoring Program, W.S. Bardswick, Ed. Ontario Ministry
of the Environment, Toronto, Ontario, Canada (April 1983).
-------
Section No. 6
Revision No. 1
Date October 1, 1984
14 of 14
14. T. Jarv, "Acid Rain Studies at Ontario Hydro: Air, Aerosol and
Precipitation Chemistry Measurements for 1981," Report No. C82-81-K,
Ontario Hydro Research Division, Toronto, Ontario, Canada (Sept. 2, 1982).
15. W.H. Chan, F. Tomassini and B. Loescher, "The Evaluation' of Sorption
Properties of Precipitation Constituents on Polyethyelene Surfaces,"
Atmos. Environ. J7, 1779 (1983).
16. M.E. Peden and L.M. Skowron, "Ionic Stability of Precipitation Samples,"
Atmos. Environ. J2, 2343 (1978).
-------
Section No. 7
Revision No. 1
Date October 1, 1984
Page 1 of 20
7.0 LABORATORY OPERATIONS
7.1 Analytical Reagents
7.1.1 Purity Requirements - Water having a conductivity less than
2.0 umho/cm (resistivity greater than 0.5 megohm/cm) is acceptable for
analysis of major constituents in rainwater. In the past, high purity has
been obtained by distilling water; however, distillation systems have
several drawbacks. Even double- or triple-distilled water contains easily
detectable impurities. Stills require periodic shutdown and careful
cleaning, and water production is relatively low. If distilled water is used
then it must be passed through an ion exchange column before use.
Ion exchange systems, on the other hand, provide high quality water, are
relatively maintenance free, and provide water on demand. The only
maintenance required is to change cartridges periodically. It is preferable
to pretreat the feed water with a reverse osmosis system to remove a high
percentage of ionic impurities and to prolong the life of the ion exchange
beds. In the last stage of treatment, a 0.2-micron filter should be used to
remove microorganisms and particles. A meter to monitor the conductivity of
the water should be installed inline directly before the spigot, and the
system should be checked if the conductivity becomes greater than
2.0 ymho/cm. If trace organics are to be determined, an activated charcoal
filter should also be used for purification.
Reagents used for analyses must meet standards of quality denoted by the
terms "analytical reagent grade," "reagent grade," and "ACS analytical
reagent grade." All of these grades are equivalent, and they identify
reagents which conform to current specifications of the Committee on
Analytical Reagents of the American Chemical Society (1).
It may not be possible to obtain dyes of analytical reagent grade for
automated colorimetric ammonium and phosphate analyses. For these, a
statement of purity should be obtained from the manufacturer and, if
necessary, the weights of dye used in reagent preparation should be adjusted.
-------
Section No. 7
Revision No. 1
Date October 1, 1984
Page 2 of 20
Tie lanthanum nitrate used as a flame buffer in atomic absorption should be
•atomic absorption grade."
7.1.2 Storage Requirements - Analytical reagents have a finite shelf life,
so all chemicals should be dated by the receiving clerk and labeled "Do not
use after...". Unless otherwise specified by the manufacturer, inorganic
chemicals have a shelf life of 5 yr at room temperature.
Concentrations of reagents in solution may change due
to: (1) biological action, (2) chemical reaction (e.g., oxidation),
(3) evaporation, and (4) adsorption-desorption phenomena on container
surfaces. All of these effects can be slowed by refrigeration. Guidelines
for reagent storage can be found in references 2, 3 and 4.
f-2 Laboratory Support for the Field
The laboratory must prepare standards for calibrating field instruments
and for field testing the quality control samples. Clean sample containers
and shipping materials should be supplied as needed. This section discusses
reference solutions, laboratory evaluation of field equipment, and routine
supply of materials. Detailed procedures for preparation of solutions are in
the Operations and Maintenance Manual (5).
Accuracies of field conductivity and pH measurements should be evaluated
vith audit samples. At scheduled intervals, an audit sample should be
prepared using the procedures in the Operations and Maintenance Manual (5).
All meters and electrodes should be tested in the laboratory before they
are shipped to the field. Meters usually have a serial number affixed, but
electrodes do not, so an identification number should be affixed to each
electrode. Acceptance tests are described in the Operations and Maintenance
«anual (5).
-------
I •": S\ , —,
'-', -V--1: F:oc
-------
Section No. 7
Revision No. 1
Date October 1, 198U
Page 3 of 20
7.3 Laboratory Logistics
7.3.1 Sample Handling in the Laboratory - All samples received by the
laboratory should be checked in by a receiving clerk who: (1) records the
site, date, and other identification; (2) checks the field data form against
sample labels to identify discrepancies; (3) assigns a laboratory
identification number to the sample and records the number and the date of
arrival on the data form and in the logbook; and (4) examines the data form
and the sample for certain conditions and codes the information on the data
form. These codes, which may be useful later in interpreting the data,
should be stored with the sample data in the computer. Table 7-1 suggests
some information to be coded. After logging-in the samples, the receiving
clerk should refrigerate them immediately and retain the data forms received
with the samples.
The receiving clerk should replace the old sample bucket or containers
with clean, sealed, bagged ones, and should ship the new ones in shipping
cartons to the field sites with other required materials. These can be sent
by ground transport since each site should have a several-week supply on
hand.
After all analyses have been completed and the results checked, the
sample can be transferred into a 125 ml polyethylene bottle for storage in a
refrigerator or freezer for 6 mo to 1 yr for other tests or analyses.
Stability tests over several months indicate that storage at either 4°C (6,7)
or 0°C (7) will preserve the sample.
7.3«2 Laboratory Documentation - The following documents should be reviewed
regularly by the laboratory analyst and the supervisor to determine if the
documents are up-to-date and are being followed.
1. Laboratory Standard Operating Procedure - instructions on laboratory and
instrument operations.
2. The Laboratory Quality Assurance Plan - laboratory QA protocol, including
personnel responsibilities and use of QC samples.
-------
Section No. 7
Revision No. 1
Date October 1,
Page 4 of 20
TABLE 7-1. SAMPLE INFORMATION TO BE CODED
Snow/ice
Mixed: snow/rain; hall/rain
Sample contaminated
Possible sample leakage in shipping
Sampler inoperative - no sample
Insufficient sample for complete measurement
Bain gauge inoperative
Boticeable suspended particulates
Lid cycling
Field pH and conductivity measured x days after scheduled
iple removal or end of event
pR/conductivity/temperature meter inoperative
Sample partially frozen
Unusual condition in area
Collocated samples
Sequential samples
-------
Section Mo. 7
Revision No. 1
Date October 1, 1984
Page 5 of-20
3. List of_ In-Houae Samples - dates for completion of analysis to allow the
analyst to schedule further analyses.
4. Instrument Performance Study Information - 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.
5. Quality Control Charts - Once a month, update all control limits to
include data from analyses of the previous month; generate plots of all
QC samples and curve parameters.
6. Data Sheet Quality Control Report - generate a QC report after data for
each analysis are placed in the computer (preferably within 1 day of
analysis); present Information for all QC parameters; flag all data
which exceed the statistically established QC limits; have the supervisor
review this report to decide what is to be done for out-of-control
samples; and if necessary, reanalyze samples.
»
7. The Analyst's Spike Plot - daily when the analysis is set up, the first
sample analyzed should be the analyst spike; percentage recovery for this
sample should be calculated and plotted by the analyst in real time.
7.3.3 Traceability of Calibration Standards - For chemical traceability all
calibration standards must be prepared from ACS reagent grade salts, and the
accuracy of calibration standard preparations must be checked. With
procedures proposed here, accuracy is checked by running an independently
prepared analyst spike with each analysis and by checking each new set of
stock standards against the old.
For gravimetric measurements, NBS traceability is provided by daily
balance checks with weights traceable to NBS-certified weights, so each
laboratory should maintain a set of NBS-traceable weights. The recommended
procedure is to purchase weights traceable to NBS from a commercial supplier
and to have them certified by an NBS-approved laboratory; it is unnecessary
and expensive to have NBS calibrate a set of weights directly.
7.3.4 Preparation of Analyst's Spikes - When preparing calibration
standards, the analyst should prepare an analyst's spike from a different
stock solution. The concentration of the spike should be approximately at
tie midpoint of the calibration curve; however, if the majority of samples
-------
Section No. 7
Revision No. 1
Pate October 1, 1984
Page 6 of 20
nave concentrations below the midstandard, the spike should be prepared
within that range.
7.3.5 Analytical Data Computations - The concentrations of the various
constituents in each sample are based on calibration standards, which should
be run at the beginning and end of the analysis and periodically during the
analysis. Sample data are calculated from linear least squares fit
parameters of the bracketing calibration standards. The linear least squares
fit yields the following parameters: slope (m), intercept (b), error of fit
(e) and correlation coefficient (r). The slope and intercept predict a
relationship between concentration standards and instrument response:
pi -s.
b (r XJ.9990) 7-1
where y . is the predicted instrument response, based on the calibration
constants, and x is the concentration of standard i.
Equation 7-1 yields the preferred fit where major components of random
variance are assumed to occur primarily in instrument response. If the range
of concentrations observed is very large, it might not be possible to use a
single equation. Concentration ranges should then be defined with separate
linear least square fit.
Rearrangement of Equation 7-1 yields the sample concentration
corresponding to an instrumental measurement (Equation 7-2):
*j = (yaj - b)/m 7-2
vsere x . is the calculated concentration for a sample , y_ . is the actual
J aJ
instrument response for a sample, and m and b are the calculated slope and
intercept from the latest calibration standards run.
The error term is calculated from the difference between the predicted
instrument response, y ., and the actual instrument response, yai, for a
given calibration standard (Equation 7-3):
-------
Section No. 7
Revision No. 1
Date October 1, 1984
Page 7 of 20
where n is the number of calibration standards.
This term, which indicates how much random scatter is in the calibration, has
the same units as the y variable (instrument response) and thus should be
directly compared between calibrations only when all setup parameters (scale
factor, concentration range, etc.) are identical.
7.1* Quality Control Program
When analytical data are reported, it is essential to specify their
quality. Statements about quality should refer to the particular data set
being reported, not to laboratory analyses in general. To accomplish this,
an internal QC program should be implemented. Internal QC includes
calibration and real-time control by the analyst, preparation of special QC
samples by the QC chemist, analysis of those samples by the analyst, review
of the data by the laboratory supervisor and QA coordinator, scheduled data
checking for transcription errors by data processing personnel, and a final
review of all QC data by the QA coordinator before final reporting.
This section specifies QC samples to be analyzed, and discusses
responsibilities for evaluation of the QC data. Control charts for both
analyst review and managerial review are stressed. Procedures require
real-time review of analytical performance by the analyst, and QC review of
all data directly after input to the computer.
7.1*.! Real-Time Quality Control Procedures - These procedures are designed
to spot problems during the analysis so that corrections can be made
immediately. A brief description of the recommended procedures for each
measurement is provided below:
(a) Real-Time Plotting of Analyst Spike Data — After each instrument is
calibrated, the analyst should immediately run an analyst spike to ensure
that calibration standards were correctly prepared and that no degradation
of the standards has occurred. After the analyst spike sample has been
run, its value should be calculated by using the first calibration curve
-------
Section No. 7
Revision No. 1
Date October 1, 1984
Page 8 of 20
of the day. The percentage recovery should be calculated and plotted as
indicated in Figure 7-1. (For pH and conductivity, the absolute
magnitude, not the percentage recovery, should be plotted.) The horizontal
average recovery and control limit lines (Figure 7-1) are those calculated
in the most recent monthly QC report. If an out-of-limits data point is
noted, an explanation should be sought. If eight successive values fall
on one side of the average line, the indicated bias should be evaluated.
(b) Gravimetric Measurements — The analytical balance should be calibrated
frequently against Class-S weights, and the Balance Calibration Log should
be completed. The balance should be zeroed before each use.
(c) £H Measurement — The pH meter should be calibrated as indicated in the
Operations and Maintenance Manual (5). The first sample analyzed after
calibration should be the pH electrode reference solution. The .analyst
should plot and evaluate the pH value. Backup electrodes should always be
in the laboratory to check the first electrode(s) if the reading differs
from the previous analysis of the reference solution by more than +0.03
units. Calibration drift should be evaluated after 20 samples, are
analyzed. If the drift is more than +0.02 pH unit, the analysis should be
stopped, and the meter and electrodes should be checked.
(d) Strong Acid and Acidity Measurements — For strong acid determination,
each day when sample measurements are begun, three conditioning solutions
(if applicable) and an analyst spike should be measured and calculated
using the linear least squares fit, as described in the procedure in the
Operations and Maintenance Manual (5). If correlation coefficients from
the calculation are less than 0.9990, the indicated problem should be
eliminated. The value of V (the equivalent volume of base added) for
each conditioning solution* and the analyst spike percentage recovery
should be plotted and obtained as indicated in the Operations and
Maintenance Manual (5). At the end of the day, an analyst spike and a
conditioning solution sample (if applicable) should be analyzed. The
initial conditioning solution potential for each sample (if applicable)
should be within 1.2 mv of the potential for the conditioning solution.
According to the Operations and Maintenance Manual (5), an analyst spike
should be analyzed before and after samples are determined and these
values should be plotted daily.
(e) Conductivity Measurement — Each day, the conductivity apparatus should be
calibrated before and after samples are analyzed. The first sample should
be an aliquot of a 0.0003M KC1 (conductivity - U4.6ymho/cm at 25° C) or
the pH electrode reference solution. The analyst should calculate the
conductivity and then plot and evaluate the conductivity of this reference
sample.
(f) Automated Colorimetric Analysis — This instrument should be set up, and
the baseline noise and instrument response should be evaluated by
comparison with data from the instrument performance study. Any problem
noted should be investigated. The instrument should be calibrated as
-------
CD
3<
o
CO
o
01
ft
*
01
O
o
»
«4
O
> N
O *
0 8
tu <>
3«
ID
01
10
N
sou
Section No. 7
Revision No. 1
Date October 1, 198i
Page 9 of 20
SPIKE DftTfl FSC= 1O.OO ug/mI
Upper Control
"- Limit
103,7
uwi_- Upper Warning
1O3.4 Limit
1OO.S
Lower Warning
l-Ct_» Lower Control
•38. 14 Limit
+
+
8 12 16 2O 24 28 32 36
DBTR POINT *
48 52
Figure 7-1. Analyst Spike Plot for SO* Analysis
-------
Section No. 7
Revision No. 1
Date October 1, 196-
Page 10 of 20
described in the Operations and Maintenance Manual (5). For real-time QC,
the first calibration curve should be checked for linear response and
adequate detection limit by using a linear least squares fit of the first
calibration curve and by determining a detection limit. Linearity should
not be less than 0.9990; the detection limit should be within the
statistically established control limits. The first sample analyzed
should be the analyst spike. Concentrations of this sample should be
calculated from the first calibration curve, and the value obtained should
be plotted and evaluated. In addition, the calibration response during
analysis should be checked to see that it is changing less than 5% from
one calibration curve to the next. If a greater change is noted, the
analysis should be stopped and an explanation sought.
(g) Ion Chroma to graphic Analysis — Calibration procedures are in the
Operations and Maintenance Manual (5). For real-time QC the baseline
noise and the response of the first standard at setup time should be
monitored. The first sample analyzed should be the analyst spike; it
should be calculated from the first calibration curve, and the value
should be plotted and evaluated. Linearity of the least square fit should
be no less than 0.995 (to allow for 1C non-linearity in calibration
curve).
(h) Atomic Absorption Analysis — Atomic absorption calibration procedures are
in the Operations and Maintenance Manual (5). For real-time QC, the first
calibration curve should be analyzed, and the linear least squares fit of
response vs concentration should be calculated. The correlation
coefficient should be 0.9995 or greater, and the detection limit should be
within the statistically established limits. The first sample analyzed
should be the analyst spike; it should be plotted in real time.
7.^.2 Analysis and Evaluation of Quality Control Samples - At least once each
analysis day, a reagent blank, an old sample, a duplicate sample, an analyst
spike, and a blind sample should be analyzed following the recommendations
below:
(a) Reagent Blank — This deionized water QC sample, which is subjected to the
same preparation procedure as the routine samples, should be analyzed to
check for random contamination which may have occurred in sample
preparation or analysis.
(b) Old Sample — This randomly chosen, previously analyzed QC sample (if no
sample degradation has occurred) provides information on analytical
precision for different days of analysis; it may provide information on
sample stability, but this is not its primary purpose. Sample degradation
can be corroborated by a repeat measurement at another time and the value
in the QC data noted accordingly.
-------
Section No. 7
Revision No. 1
Date October 1, 1984
Page 11 of 20
(c) Duplicate Sample — This randomly chosen QC sample is a reanalysis of a
sample analyzed during the same analytical run. If preparation is
necessary before analysis, it is prepared twice. The result may be used
to calculate analytical precision for the measurement method.
(d) Analyst Spike — This QC sample prepared by the analyst from a stock
solution independent of that used to prepare the calibration standards
provides information on the accuracy of the calibration standard and the
precision of analysis. The analyst spike should be analyzed at the
beginning of the run; results should be calculated and plotted in real
time.
(e) Blind Sample — This QC sample is a standard of known authority (NBS,
OSGS, or EPA). It is inserted into the analytical run as a blind sample
by the laboratory supervisor. The purpose of this QC sample is to assess
data quality independently of analyst judgment.
7.4.3 Data Screening Tools - For most precipitation measurements, 100% QC of
each sample cannot be attained. However, with QC procedures properly
implemented for each analysis batch, adequate screening of continued, proper
instrument functioning can be achieved. Recommended data checks for routine
screening of measurement data include the following:
1. Calibrations — provide statistics for evaluating the analytical method.
Duplicate calibrations, performed before and after analysis of field samples,
yield data on instrument reproducibility and drift. Statistics from routine
calibration data include:
Slope (m) and Intercept (b) - of the least squares fit of the data for a
technique and instrument range should be fairly constant. Visual inspection
of values from successive days or comparison with values obtained during
method validation can be used for quality control.
Correlation Coefficient (r) - is often used in a semi-quantitative way
to evaluate goodness of fit of the relation of one variable with respect to
another. Values near +1.0 and -1.0 are good and values near 0 are poor. The
actual range of good fit values will depend on the particular measurement or
test. For most instrumental techniques employed for analysis of
precipitation samples, rX).9990 would be acceptable.
Residual Error (e) - is a measure of the scatter of data points off the
regression line indicating "noise11 in the calibration. Such scatter is
related to the expected precision of the analysis. For analysis of
precipitation samples, e<2% of full scale of instrument response, for a given
concentration range, is acceptable.
-------
Section No. 7
Revision No. 1
Date October 1,
Page 12 of 20 .
2. Spikes, Duplicates and. Reagent Blanks — are the primary tools for
monitoring the integrity of the analysis. As a minimum, at least one QC
spike, one duplicate and one reagent blank should be included with every
sample batch. If calibrations are performed before and after each batch of
samples for optimum control, a QC spike of known value should be run at the
beginning, after calibration is completed. If a large batch of samples is
run, analysis of the QC spike should be repeated after every twenty samples.
The concentration values of the spikes, duplicates and reagent blanks should
be entered into the computer before analysis, so that control limits can be
checked as the analysis results are keyed into the computer. Detailed
discussion on computation of control limits is given in Section 7.4.4 below.
3. Old Samples — are reanalyzed to obtain information on sample stability,
which varies from sample to sample. Analytes such as H*, NH£, POT , and
NO." are susceptible to degradation, and if degradation is observed (i.e.,
lower analytical results are obtained for an ion), sample handling and
preservation techniques should be examined. Old sample data may be used to
calculate control limits vising a mathematical approach similar to that
discussed In 7.4.4 below.
7.4.4 Control Limits Determination - QC data are stored, tabulated and
sometimes plotted as a function of time. The abscissa is the chronological
order of analysis, and the ordlnate may be either the range, absolute
magnitude, value of the difference of replicates, or percentage recovery.
The data are used for the determination of expected values and the associated
control limits. These are plotted as the average-value line and the control
limit lines when constructing control charts (2) such as the one given in
Figure 7-1.
Computer calculation and data evaluation with or without plotting may be
substituted for manually plotting the data. Because the purpose of
establishing control limits is to flag potential problems as soon as
possible, a computer can be easily programmed to recognize control criteria
without having to plot the data manually. Even in those instances where a
-------
Section No. 7
Revision No. 1
Date October 1,
Page 13 of 20
laboratory has such computer capabilities, a visual inspection of QC data is
still a valuable tool, and any out of control situation should be examinee
without delay. Two types of limits are frequently used to identify such
situations. These are:
1 . Control Limits - corresponding to the 99% confidence interval for the
mean value of the control parameter. Thus the upper control limit (DCL) and
the lower control limit (LCL), given by the mean +3 S. (standard deviations),
imply that for normal data distributions, less than W of valid data is
flagged due to random error alone and that other flagged data may be assumed
to indicate nonrandom error, i.e., malfunction or contamination which
requires immediate action.
2. Warning Limits - corresponding to the 95% confidence interval for the
mean value of the control parameter. Thus it includes about 95$ of the
expected random variation about the mean u for normal distributions,
i.e., u+2S.. However, simple probability predicts that there is less than a
W probability that two independently chosen values for the controlled
parameter will exceed the upper warning limit (UWL) or the lower warning
limit (LWL) due to chance alone. Thus two successive values which exceed the
warning limits should be reason for investigation of analytical control.
Also, probability predicts that there is less than a U chance of eight
independent, consecutive values occurring on the same side of the mean value
line. Thus, eight or more such values would indicate a systematic bias in
the measurement.
For precipitation monitoring there are five key parameters that are
useful for controlling the field and the laboratory measurement processes.
Control and warning limits for each of those parameters should be determined
as follows (Equation
(a) Spike Recovery — best determined by utilizing reference standards (such
as NBS or EPA) . The analyst spike data are used to calculate an average
percent recovery (Equation 7-4):
-------
Section No. 7
Revision No. 1
Date October 1, 195-
Page 1* of 20
_ ! N X
ZR -Z ~X 100
where, JR is the average percent recovery, X is the found spike
concentration, C. is the known spike concentration, and N is the number of
spikes used for constructing the control limits.
The standard deviation of the percent recoveries will be given by:
7-5
and the appropriate warning and control limits are:
ZR — ** ZR 7-6
Table 7-2 presents the numerical values of the various factors for computing
control limits (i.e., control chart lines).
(b) Range of Duplicates - determined by utilizing the absolute differences
of duplicate analyses of selected samples. These data are used to calculate
an average range (r)
i K
7-2 'Xil -Xi2l 7"7
K i-1 i 1Z
where, X, 1 and Xi2 are the results of the duplicate analysis of the ith
sample, and K is the number of duplicates used to construct the average. The
corresponding warning and control limits for the range are:
7-8
When utilizing Table 7-2 for duplicate analysis, N=2; N>2 only if a larger
number of replicate analyses of the same sample are performed.
(c) Detection Limits — determined by running either a series of ten
non-consecutive reagent blanks (for AA analyses) or ten standards of the same
-------
Section No. 7
Revision No. 1
Date October 1, 198M
Page 15 of 20
TABLE 7-2. FACTORS
FOR COMPUTING
CONTROL LIMITS3
Number of
Observations, N
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Factor
D3
0
0
0
0
0
0.08
0.14
0.18
0.22
0.26
0.28
0.31
0.33
0.35
0.36
0.38
0.39
0.40
0.41
0.42
0.43
0.44
0.45
0.46
Factor
D4
3.27
2.57
2.28
2.11
2.00
1.92
1.86
1.82
1.78
1.74
1.72
1.69
1.67
1.65
1.64
1.62
«• 1.61
1.60
1.59
1.58
1.57
1.56
1.55
1.54
Factor
D5
2.51
2.05
1.85
1.74
1.67
1.62
1.58
1.55
1.52
1.50
1.48
1.46
1.45
1.43
1.42
1.41
1.40
1.40
1.39
1.38
1.38
1.37
1.36
1.36
Factor
D6
0
0
0.15
0.26
0.33
0.38
0.42
0.45
0.48
0.50
0.52
0.54
0.55
0.56
0.58
0.59
0.59
0.60
0.61
0.62
0.62
0.63
0.63
0.64
a. Based on EPA-APTD-1132, "Quality Control Practices in
Processing Air Pollution Samples," March 1973.
-------
Section No. 7
Revision No. 1
Date October 1, 195-
Page 16 of 20
concentration, at 10> of working curve (for ion chromatographic and
colorimetric analyses). For each series of runs, a standard deviation (S ]
a
is determined with the detection limit (DL) defined as DL=3S . When reagenc
blanks (B) are part of the analytical run, the control limit should be sec
such that HK2DL for each of the observables.
The notification that a sample is below the detection limit of the
analytical method for a given analyte range should be available immediately
after raw data entry. The analyst then has two options: to reanalyze the
sample on an instrument scale of greater sensitivity; or to accept and flag
data as being below the detection limit.
(d) (Field - Laboratory) pH — determined by comparing the measurements
performed in the field by the individual site operators to those performed at
the central network laboratory upon sample receipt. The warning limits for
this comparison are:
lpHfield " pHlab'-°*2 S^P16 PH<5.00 7-9
lpHfield " pHlab'-°'3 SamPle PH>5.00
If those limits are exceeded, the laboratory reanalyzes the sample and
the field measurement is investigated (i.e. pH electrode, buffer solutions,
etc.). The control limit for this difference has been set at 0.5 pH unit.
If this is exceeded the sample should be reanalyzed and the data flagged as
suspicious .
(e) (Field - Laboratory) Conductivity ~ determined by comparing the
conductivity (COND) measurements performed in the field by the individual
site operators to those performed at the central network laboratory upon
sample receipt. The control limits for this comparison are:
COND,. . . - COND. , |>3ymho/cm COND <15ymho/cm
rield lab — —
COND,.., - CONDlab| xlOO± 20% COND >15pmho/cm 7.10
COND- ,
lab
-------
Section No. 7
Revision No. 1
Date October 1,
Page 17 of 20
If those limits are exceeded, the laboratory reanalyzes the sample and
the field measurement is investigated (i.e. conductivity cell, KC1 standards,
etc.). If the laboratory reconfirms its initial analysis, the field value is
flagged as suspicious.
7.4.5 Evaluation of QC Data - Data for QC samples should be calculated and
compared to control limits established in the most recent monthly QC chart.
After input, a Data Sheet QC Report is printed by the computer. If no
computer is available, this procedure can be done manually. The report on
the performance of the QC samples should be given to the laboratory
supervisor for evaluation to see if the data are acceptable for reporting or
if reanalysis is necessary.
The analyst spike and the blind sample data are also calculated and the
results given to the QC chemist, who tabulates the data and calculates the
percent recoveries* The QC chemist routinely gives the blind sample data to
the laboratory director, who reviews this before reporting the analytical
data to the program manager. The QC chemist reviews the blind sample data
with the analyst monthly.
Once a month, the QC chemist should combine all QC sample data and
calibration curve parameters obtained during the month with all previous data
for the same parameters and plot the data to yield new control limits are
calculated. A copy of the monthly QC plot should be given to the laboratory
supervisor, the QC chemist, and each analyst.
A Data Quality Control Report should be given to the laboratory
supervisor on a regular basis. This report should flag any out-of-limits
conditions for the analysis. The analyst or the supervisor should check to
see that any out-of-limits conditions are not results of manual transcription
errors. If a transcription error is found, the transcription of all data on
the data form should be checked, a corrected data form should be input and a
new QC report issued. The problem should be carefully documented and the old
data form and QC report should be filed with the corrected forms. The new QC
data should replace the old in the computer QC data base (if applicable). If
the out-of-limits conditions are hot due to transcription errors, another
-------
Section No. 7
Revision No. 1
Date October 1,
Page 18 of 20
explanation should be sought.
The most critical parameters are the calibration curve parameters,
because any problem with them directly affects the data. If an explanation
cannot be found for out-of-limits calibration parameters, all samples
analyzed in the time interval between the questionable calibration curves
should be reanalyzed.
If only one of several QC samples is out-of-limits, an explanation
should be sought. However, if one cannot be found, no action is needed. The
supervisor may assume that the problem was with the particular QC sample
itself, but he may retain the out-of-limits data in the QC data base. If
several QC samples are out-of-limits and an explanation is not found, all
samples analyzed with the QC samples should be reanalyzed.
In any case, an out-of-limits QC sample requires evaluation and an
explanation by the supervisor. The explanation may be noted on the QC
report. If samples are to be reanalyzed, the supervisor should note this on
the data form and in the QC report. At the time of the reanalysis, it should
M~
be noted on the data form that the samples are being reanalyzed, and the date
of the first and the repeat analyses should be given.
The supervisor also evaluates the QC plots when they are periodically
generated. Since the supervisor has already evaluated out-of-limits
conditions in the QC reports, all out-of-limits conditions should have been
explained or eliminated (by reanalysis) before the monthly plots are
generated. Accordingly, the monthly plots should be examined primarily for
systematic bias. If no systematic bias is present, all plotted values should
be evenly distributed about the average-value line.
7.5 Evaluation of Laboratory Performance
The QC procedures (Section 7.3) stress the supervisor's role in
evaluating QC data and in scheduling reanalyzes until data are acceptable for
reporting. This section discusses the QC chemist's role in evaluating
laboratory performance by independent QC checks and by external audits.
-------
Section No. 7
Revision No. 1
Date October 1, 1984
Page 19 of 20
7.5.1 Independent Internal Quality Control - The QC chemist ensures that QC
procedures are implemented and provides independent judgment on the accuracy
and precision of the data generated in the laboratory. These are checked by
analyst spikes and synthetic rainwater samples which have been
inconspicuously added to the sample analysis stream. Blind samples can be
prepared by the QC chemist from NBS-SRMs, from reagent grade salts or from
reference samples obtained from EPA, USGS, or other reliable sources and
routinely submitted to the laboratory as blind samples to be analyzed in the
same manner as a routine sample. The data should be flagged in the data set
by the QC chemist ,_when checking for out-of-limits conditions before the
reports are reviewed by the laboratory director.
7.5.2 Laboratory Audits - Laboratories analyzing rainwater samples should
regularly conduct both systems and performance audits of their operations. A
complete description of the schedule, scope and methods for conducting both
types of audits is included in Section 10.0 of this document.
7.6 References
1. Reagent Chemicals, American Chemical Society Specifications, 5th Edition,
American Chemical Society, Washington, D.C. (1974).
2. "Standard Methods for Preparation, Standardization, and Storage of
Standard Solutions for Chemical Analysis", from Part 31 of 1976 Book of
ASTM Standards, American Society for Testing and Materials, Philadelphia
(1977).
3. Standard Methods for the Examination o£ Water and Wastewater, 13th
Edition, American Public Health Association, New York (1971).
4. M. Reichgott, "Organic Coatings and Plastic Chemistry," Vol. Ul, 1979,
Paper presented at 178th National Meeting of the American Chemical
Society, Washington, D.C. Sept. 9-14, 1979.
5. Quality Assurance Handbook for Air Pollution Measurement Systems - Vol.
V 2. Manual for Precipitation Measurement Systems. Part II ^ Operations
and Maintenance Manual. U.S. Environmental Protection Agency, Research
Triangle Park, N.C., EPA-600/4-82-042b (January 1981).
6. M.E. Peden, and L.M. Skowran, "Ionic Stability of Precipitation Samples,"
Atmos. Environ. _12, 2343 (1978).
-------
Section No. 7
Revision No. 1
Date October 1, 195-
Page 20 of 20
7. J.E. Rothert, Battelle Pacific Northwest Laboratories, Richlanc,
Washington, MAP3S Program, private communication.
-------
Section No. 8
Revision No. 1
Date October 1,
Page 1 of 23
8.0 DATA HANDLING, VALIDATION, AND REPORTING
All data collected by a monitoring program from both field and
laboratory activities must accurately represent the concentrations of
measured constituents, i.e., the data must be valid. Validity depends on
control of error a'nd bias, and such validity is assured only by careful
screening during all phases of data handling from field and .analytical
results to final reporting. Thus, the field operator, laboratory analyst,
data entry staff, laboratory director, program manager and QA officer all
contribute to data validity by screening data generated or processed by their
respective areas.
8.1 Data Logistics
Ideally, data should go directly from an instrument to a
machine-readable raw data base to avoid transcription errors. Not all
laboratories have computer facilities, so manual data-recording techniques
are discussed, but all calculations and data-processing steps in this section
can be performed by either manual or automated processing.
If output from the analytical instrument is not recorded automatically,
a data form must be prepared. Each analyst should keep a bound notebook to
record all analytical data, and-the notebook should have carbons so copies
can be pulled for data reduction. Typical data forms for each analytical
procedure are in the Operations and Maintenance Manual (1).
Whenever data forms are prepared from strip charts, transcriptions
should be checked by recalculating 5% of all values. If errors are found,
all data should be reprocessed. Manual data-recording practices for several
typical techniques used in precipitation analysis are described briefly
below:
*
(a) j>H and Conductivity — The pH and conductivity measurements read directly
from meters are recorded on a prepared data form. The baseline reading
category is ignored.
-------
Section No. 8
Revision No. 1
Date October 1, '-za
Page 2 of 23
(b) Ion Chromatography for Chloride, Phosphate, Nitrate and Sulfate — Peak
heights on strip charts are a measure of response. Baselines should be
carefully drawn and each peak height read from the baseline with a clear
plastic ruler. Each peak height is recorded on the strip chart and alsc
on the data form. Because this is a chromatographic technique, care must
be exercised in drawing the baseline, particularly where the peak of cr^
anlon is resting on the tail of another due to extreme differences ir
concentration. In all cases, the method used when drawing the baselines
should be identical for both standards and samples.
(c) Automated Colorimetry for Ammonium and Phosphate — The data on the strip
charts should be read at the midpoint of the flat-topped peaks and s.
straight line drawn between baseline points on the chart. Each peak
height should be read from the baseline using a clear plastic ruler, arc
the data on both the strip chart and the data form recorded.
(d) Atomic Absorption for Sodium, Potassium, Calcium, and Magnesium — Tbe
data should be processed the same as for automated colorimetry.
(e) Strong Acid b£ Gran Method ~ In the microtitration, the electrical
potential (mv) or pH is recorded as a function of the volume of base added
to the sample. As a QC procedure, the initial potential reading of tbe
conditioning solution (if applicable) before addition of sample Is
recorded on the data form. The final temperature of the sample before
titration should also be recorded.
(f) Acidity — The sample is titrated potentiometrically with a basic solution
to an end point of pH 8.3* The normality (N) of the base and the volute
(ml) required are recorded on the data form. As a QC measure, it is
suggested that the base be standardized at least monthly in an acceptable
manner (such as NaOH standardized with potassium biphthalate).
(g) Volume — In determining the amount of precipitation sample, a density cf
1.0 gm/ml is assumed; thus the ratio of mass (g) to volume (ml) is 1.3.
The mass is recorded directly on the data form.
8.2 Software Requirements
Data handling from raw data input through finished report should be
computerized as much as possible to facilitate data management. All software
used by an organization to process data should be well documented. Current
source code listings should be available to responsible personnel and soch
code should include sufficient comment statements to explain and describe
both algorithms and data transfer steps. As a further QC measure, it is
recommended that, during execution, the software version identifiers be
printed on each generated report. Typically these include date and time cf
-------
Section No. 8
Revision No. 1
Date October 1, 196*
Page 3 of 23
generation and the name of the program used to make the calculations. Such
identifiers will increase the efficiency of tracking when data problems
occur. The software employed should have the characteristics and
capabilities described below.
8.2.1 Data Input - Data can be input manually or automatically. For manual
data entry, the data clerk should screen all of the terminal input by
comparison of the computer printout with the original data forms or by
duplicate entry of the same data batch. For automated entry, errors can be
detected by monitoring the data display while the data are being taken and by
occasional spotchecks of the data acquisition apparatus. Such checks might
include the introduction of a dummy data set or the introduction of a known
voltage to the instrument from which data are being taken.
8.2.2 Data Storage and Indexing - At a minimum, acid precipitation data
bases should be stored on a computer readable medium (disk, tape or cassette)
which can be efficiently accessed. A duplicate backup file, stored in a
different location, should be maintained so that lost data files can be
retrieved or reconstructed. Special file attributes, such as random access,
keys, and indexing, can be useful for efficient data management. Acid
precipitation data bases may be handled efficiently by any one of a number of
commercially available data base management systems. Such systems decrease
the necessity of writing vast amounts of software and are available for any
size computer installation from a small microcomputer to mainframe. Most
include sufficient provisions for the development of audit trails.
8.2.3 Precipitation Data Bases - Acid precipitation measurement data froa
the following networks: NAOP, CANSAP, APIOS, MAP3S, EPRI SURE, and EPBI
OAPSP network are currently stored at Pacific Northwest Laboratory (PNL) in
Richland, WA. The use of this data base can be helpful when developing QA
procedures for an acid rain network's software.
Data are also available in hard copy format. Data are obtained from the
various networks in one-year increments, but input data are reorganized by
-------
Section No. 8
Revision No. 1
Date October 1,
Page U of 23
region and not maintained by network. Reports • may be obtained detailing
network inventories of precipitation-weighted concentrations by site, area,
analyte and the time period. Although not yet available at the time of this
writing, PNL is developing a spatial distribution map for precipitation data
and is working to standardize isopleth derivation from sparse amounts of
network data. Both of these will be available soon. The PNL data base
currently serves as a repository for all data collected from the National
Atmospheric Deposition Program (NADP)/National Trends Network (NTN). The PNL
data bank is not currently available by dial up (interactive mode). However,
interested contributor/user organizations may contact the individual noted
below for assistance in obtaining the proper data summaries. One extremely
useful feature of this data base is that PNL will prepare data sets for
organizations in a custom-tailored magnetic tape format. (Contact:
A.R. Olsen (Statistical or Data Inquiries) or C. Watson (Data base use and
tape formats), Pacific Northwest Laboratory, Rlchland, WA, 509/376-2227.)
8.3 Data Handling and Preliminary Screening
In an acid precipitation monitoring network, data handled both in the
field and in the laboratory must be critically reviewed to identify and
isolate errors. Data should be validated at each step of the measurement
process, beginning with sample validation in the field and followed by a
preliminary physical screening process when the sample is received at the
laboratory. Once data have been entered into the storage-retrieval system,
more critical screening procedures should be implemented.
In establishing statistical screening procedures, it is necessary to
recognize characteristics of the chemical analyses. For the most part,
analyses are done in a batch mode. A batch may contain several dozen samples
with multipoint calibrations performed before and after the samples. In an
efficiently run laboratory, most analyses are automatically sequenced with
data recorded by direct computer interface or continuously recorded onto
strip charts.
-------
Section No. 8
Revision No. 1
Date October 1,
Page 5 of 23
8.3.1 Quality Control of Data Handling - Table 8-1 summarizes the data
handling steps and the corresponding QC procedures to be utilized at each
step. A range check is sometimes effective as an additional screen against
keypunch errors. The data handling steps outlined are self explanatory and
are designed for computerized data reduction systems for analytical results.
If completely automatic data acquisition systems are utilized, the first two
data handling steps should be modified to verify the integrity of the
digitization of the analog signal output from the analytical instrumentation
and the computer interface.
TABLE 8-1. SUGGESTED QC SPOTCHECK OF DATA HANDLING
Data Handling Step
QC Procedure
Manual reading of strip chart
Transfer of analytical raw data
and results to data sheet
Input of data (field or analytical
form) into computer
Electronic digitization of strip chart
Field report record of event time
and amount.
Duplicate reading of
5% of the data
Spotcheck transfer
accuracy for 531
of data
Enter both data sheets from
above into separate files,
followed by computer
comparison of 100% of data
Check 5 to 8% for proper
baseline determination and
for accurately and properly
recorded retention times
Check 100% vs rain gauge strip
chart.
-------
Section No. 8
Revision No. 1
Date October 1,
Page 6 of 23
8.3-2 Treatment of Outliers - This section will discuss treating data which
fall outside of certain control limits. In doing this, it is important to
understand the difference between outliers and erroneous data and how each
should be handled. Outliers have two origins: random errors and operational
errors. Random errors can occur in the physical condition being measured and
in the measurement system itself. Operational errors originate in numerous
technical problems such as sample handling, difficulty in accurately weighing
chemicals for standard solutions, and shift of a decimal point in manual
transcription of numbers. Operational errors contribute systematic errors or
the bias component of the measurement data.
There are basically three ways of dealing with outliers:
(a) Reject them as invalid;
(b) Subject them to special scrutiny, and reject or adjust them only if
documentary evidence of errors is found; and
(c) Do nothing.
If outliers are rejected, a great deal of care should be exercised. It
is not recommended to directly reject the highest XJ of the data points,
since this would artificially bias the averages low, and quite possibly would
weaken the ability to detect correlations in the data by throwing out valid
but high-valued data. This danger of rejecting valid data is the motivation
behind statistical methods for detecting "true" outliers.
It is advisable to check outlying data points for data handling errors.
Since no data should be rejected without documentary evidence of handling
errors, there is no serious consequence if some good data are checked.
However, outlier-specific spot-checking should not be used to completely
replace all random-selection spot-checking. Outlier-specific spot checking
alone could fail to detect a procedural or computer software error. Fully
randomized spot-checking is discussed further in 8.6. Outlier-specific
screening methods can be based on examination of the most extreme values of a
variety of computed parameters, such as calibration slopes, collocation
differences, total anion - cation balance, etc., as discussed in Section 7.
-------
Section No. 8
Revision No. 1
Date October 1, 1984
Page 7 of 23
The moat commonly used outlier-detection statistics assume that the
normal fluctuations (in both the measurement error and the physical condition
being measured) are well-approximated by a Gaussian distribution (3). It is
also assumed that there is at most one outlier due to systematic bias. The
mean and standard deviation are calculated from the data, and the chance that
the most outlying value from that Gaussian distribution would be as large as
the most outlying value observed is determined. If this chance is less than
a selected statistical significance level, then the outlier is declared to be
caused by an operational error. Details on applying this statistical
procedure are given elsewhere (3).
These methods suffer from the problem of "masking11. If the sample of
data points has two outliers that are far from the majority of data points,
then the computed standard deviation is increased, and no outliers are
detected. There is another method called backward elimination, which
eliminates the masking problem. However, it still has the assumption that
the normal fluctuation fits a Gaussian distribution. In most cases a
log-normal or other transformation of the parent data distribution has proven
adequate. The chance of experiencing any particular extreme value is
under-predicted by all Gaussian-based methods. Therefore, rejection of data
points on the basis of those statistical outlier detection methods should be
undertaken with great care and only after examining in detail the
distributional characteristics of the entire measurement data.
8.4 Data Validation Criteria
Data validation based on a set of criteria is the process of evaluating
the data after its preliminary screening, and either accepting or rejecting
them. Validation in this sense also includes the investigation of anomalies.
Procedures in this section are used after the preliminary screening of the
analytical runs has been completed. They are designed to flag questionable
data for subsequent investigation based on some key physical properties of
precipitation samples.
-------
Section No. 8
Revision No. 1
Date October 1,
Page 8 of 23
8.4.1 Detection Limit Flag - If a sample concentration is below detection
limits, a flag should be entered into the data base indicating that the
result is below the detection limit. With a fully automated system, this
flagging could serve as an additional QC procedure because the analyst cannct
mistakenly enter data which are below the detection limit if the computer
program is written to question and flag all such entries. Statistical
handling and reporting of below-detection-limit data are discussed below in
Section 8.5.3.
8.1.2 Comparison of Sampler and Rain Gauge Performance - At stations
equipped with duplicate samplers or with a sampler and a rain gauge, sampler
performance can be evaluated by comparing the quantity of precipitation
measured by the two instruments. In most stations a sampler and a rain gauge
are present. The rain gauge is used as the reference to measure both the
sampler capture ratio and Its collection efficiency. The percent capture
(JCAP), which indicates sampler down time, is given by
JCAP s (No. of samples collected/Total No. of events)x 100 8-1
The total number of events are counted from the rain gauge charts, and the
failure of the sampler to capture an event is detected either by direct
observation of the sample bucket or by counting the event pen markings for a
given collection period. In the 1982 Utility Acid Precipitation Study
Program (UAPSP), a percent capture of 98J was found (5) for over 2200 events
at 20 sites.
The sampler collection efficiency (COL.EFF.) is given by the ratio of
the sample amounts collected by the sampler and the rain gauge:
COL.EFF. s Sample Depth (cm)/Rain Gauge Reading (cm) 8-2
The Aerochem Metrics bucket sample weight can be converted to cm(in) by the
relation, 1 cm = 6MO g (1 in s 1625 g). When a sample is not collected due
to either sampler or rain gauge down time, the event is not included in the
-------
Section No. 8
Revision No. 1
Date October 1, * 95-
Page 9 of 23
efficiency average. An overall sampler efficiency of 1.OH was found for
OAPSP (5) and indicates that the samplers, on the average, collected MJ mere
precipitation than the rain gauges. For rain events greater than 0.25 cs.
(0.1 in), amounts which differ by more than 15S between the collector asc
rain gauge should be flagged for investigation of faulty collector function.
Light rainfalls generally yield high rain gauge capture vs. the sampler, and
heavy rainfalls (and wind) yield high sampler capture vs. tipping bucket
gauges. For snow with no windshield present, differences of 30* or more
between the sampler and the rain gauge are frequent. It is recommended that
the sample be rejected if
Ratio - Rain Gauge Reading (cm) - Sample Depth (cm) > Q ,
Rain Gauge Reading (cm) 8'3
The Ratio in Equation 8-3 applies directly to both rain and snow when the
rain gauge is of the weighing type; if another rain gauge is used the snow
should be melted before rain gauge reading is utilized. If the sampler
volume is greater than the rain gauge volume, the reason for the difference
in sample amount captured by the rain gauge and sampler should be adequately
resolved before a decision is made as to which quantity will be used in
reporting precipitation amounts.
As an additional check on data validity, the stripe hart record of the
precipitation gauges should be compared with the field data form. If
discrepancies in time of event or precipitation amount are noted, they must
be resolved before the data are reported.
8.U.3 Unusual Ion Ratios - Another check on data validity is to inspect the
ratios of ions in individual samples. Table 8-2 shows average and typical
ion weight ratios for different geographical locations. Average values for
sea water (6) should apply to those areas within 50 km of a sea coast, but
those for the earth's crust (7) cannot be assumed to represent any specific
region. The range data (last column) are based on a 1968 national study (8).
-------
Section No. 8
Revision No. 1
Date October 1, 195-
Page 10 of 23
The seawater ratio of SOr/Na* in the table is sometimes used in coastal
areas to correct measured SOj! values for seawater contribution.
4 corrected
(S°4)measured ' °'25 measured
8-4
However, the ratio of sulfate to sodium in spray has been shown to exceed
that in seawater by 10$ to 30* (9). The SOjj/Na* ratio is preferred to
S0jj/Cl~ because there are non-sea sources of Cl~ and because loss of
atmospheric Cl~ occurs by oxidation.
TABLE 8-2. ION RATIOS FOR VARIOUS SOURCES
Geo chemical Primary
Ratio Source
Cl"/Na* a seawater
earth's crust
Na*/K* a seawater
earth's crust
Mg^/Ca** a seawater
earth's crust
SOi^/Na* seawater
Average
Value (6,
(yg/g)
1.8
0.01
27.8
1.1
3.2
0.6
0.25
Precipitation
7) Area
industrial area
seacoast
arid region
(soil particles)
seacoast
inland
seacoast
inland
seacoast
Acceptable
Range (8)
(yg/g)
1.8 -
1.5 -
0.8 -
6
1.2 -
0.1 -
0.03 -
0.25
3.5
1.8
1.0
13
4
1.0
0.3
a. Do not test if_either number is at detection limit.
b. To correct SO,.8 for seawater S0^s contribution.
-------
Section No. 8
Revision No. 1
Date October 1, 1984
Page 11 of 23
8.4.U Comparison of_ Anion and Cation Equivalents - The principle of
electroneutrality requires that total anion equivalents equal total cation
equivalents. The Anion Equivalents (AE) are
- ** CAi
where CAI denotes the concentration of the ith anion in mg/liter,
(Eq.Wt.). is the equivalent weight for the corresponding species as given in
Table 8-3, ai
Equivalents.
Table 8-3, and N. is the nunber of anions used to compute the Anion
The cation equivalents (CE) are:
Nc C
_ 1Q(3-pH) z ci
CE 10 + i-1 (Eq.Wt.)1
where C is the concentration of the ith cation in mg/liter, (Eq.Wt.). is
Cl 1
the same as in Equation 8-5, and N is the nunber of cations, excluding E*,
used in computing the Cation Equivalents.
Using Equations 8-5 and 8-6 above, we obtain the % difference for the
discrepancy between the cations and anions.
Ion I Difference - * 100 8-7
For the EPRI Utility Acid Precipitation Study Program (5) the observed Ion %
Difference (Equation 8-7) had a median of 4.6} for 3061 samples (HCO~ was
accounted for). The greatest spread in data occurred at low concentrations.
If the % discrepancy significantly exceeds 0, it suggests that errors exist
in the data and/or that important constituents, such as bicarbonate HCO~ or
organic anions, have not been analyzed. For solutions in equilibrium with
atmospheric C02 at 25°C, the bicarbonate concentration is given by (7):
(HCO~) = KtHgCO^/df) = 4.M5 x 1
-------
Section No. 8
Revision No. 1
Date October 1,
Page 12 of 23
TABLE 8-3. CONVERSION FACTORS and EQUIVALENT WEIGHTS
Analyte
Cl"
N03-
sou2
P0,j~3 as H2POj
HCO:
3
H*
MH4*
Na+
K*
Mg~
Ca**
Equivalent Weight
35.46
62.01
48.03
96.98
61.0
1.01
18.04
22.99
39.10
12.15
20.04
Factor3
28.2
16.1
20.8
10.3
16.4
990.1
55.4
43.5
25.6
82.3
49.9
a. (peq./liter) = (mg/liter) x Factor
-------
Section No. 8
Revision No. 1
Date October 1, 1984
Page 13 of 23
Expressing the concentration of HCO~ and H+ in ueq. /liter, we find (HCO~) =
5.1/(H*). Thus, bicarbonate is negligible for solutions more acidic than a
pH of about 5.0, i.e. for H* = 10 yequiv. /liter, HCO~ = 0.031 mg/liter. It
can
also be shown that carbonate
follows: (CO*) = K., K2
+2
(CO*)
can be expressed as
C03)/(H). If we substitute the
concentration from the solubility of C02 at 25°c, and K.. and K2 respectively
for the two dissociation constants of carbonic acid we would obtain
(Cop a 2.9 x10~22/(H*)2 moles/liter. This shows that CO* is negligible
for solutions more acidic than a pH of about 7.7 (H* = 0.02 yeq. /liter).
The anion-cation balance criteria used for reanalysis by the National
Atmospheric Deposition Program are presented in Table 8-4. This information
may prove to be a useful starting point for those organizations wishing to
use this information in screening and validating data.
TABLE 8-4. NADP REANALYSIS CRITERIA
1. Ion Balance
Anions + Cations (peq/liter)
<50
>50 < 100
>100
2. Specific Conductance Balance
Measured Conductance (ymho/cm)
<5
>5 < 30
>30
Ion % Difference
>+ 60
>+ 30
>± 15
Conductance
% Difference
> 50
> 30
> 20
Frequency of Check
All samples at
monthly intervals
Frequency of
Check
All samples at
monthly intervals
From Illinois State Water Survey for National Atmospheric Deposition
Program
-------
Section No. 5
Revision No. 1
Date October 1, 1Sc
Page 14 of 23
8.4.5 Comparison of Measured and Calculated Conductances - For dilute
solutions (below 10~^M) of known composition, the equivalent conductance is
the sum of the equivalent ionic conductances at infinite dilution (Table
8-5) . From the relation between equivalent and specific conductance
A = 1000K/N 8-9
the conductance in mho/cm can be calculated as follows:
NT
K - N. r A./IOOO . 8-10
1 i-1
where N. s z^M, with M, - g. moles of ion I/liter, z^ = valence or charge of
ion i, and N_ is the total number of species used in the computation.
For the major ionic constituents in precipitation this will become:
[10'pH(350) + 2(SOj)(79.0) + (NO^HTO.S) + (CDC75.5) +
.S) + (Na*)(50.9) + (K*)(7U.5) + 2(Ca*2)(60)] x 10"3
where the parentheses denote the ionic concentrations in aoles/liter.
With Equation 8-10 and Table 8-5, the calculated specific conductance of
solution containing
H* = 7x10~5 mol/liter, NH* s Mx10"5 mol/liter,
SO,.* = MxICT5 mol/liter, and NO ~ = 3x10"5 mol/liter.
is
1000K = 7x10'5 (350Mx10-5(7M.5)+2(4x10-5)(79.0)+3x10-5(70.6)
= 3592x1O"5 mho/cm
or
< * 35.9 umho/cm.
-------
Section No. 8
Revision No. 1
Date October 1,
Page 15 of 23
The calculated specific conductance (COND) values can be compared with the
measured values for precipitation samples, which are generally 10" M or less,
by the relation
Conductance % Difference
COND , , . - COND
calc d meas
COND
meas
x 100
8-11
For the EPRI-OAPSP study (5), the median value of the Conductance J
Difference was -0.8% and, as expected, the greatest spread occurred at low
conductivities. The H* ion is the chief contributor to the specific
conductance of a solution, and any significant error in the H concentration
will generally be evident in a comparison of the specific conductances and
the anion/cation equivalents. The conductivity criteria used by the National
Atmospheric Deposition Program for reanalysis of samples are given in Table
8-4.
TABLE 8-5. EQUIVALENT CONDUCTANCE AT INFINITE DILUTION, 25°C (7)
Ion
H*
NH/
Na*
K*
1/2Mg *2a
X . (mho/cm)
350.0
74.5
50.9
74.5
53.1
Ion XA (mho/cm)
1/2 Ca+2a 60.0
1/2 SO^ 79.0
NO " 70.6
Cl~ 75.5
HCO ~ 41.5
a. 1/2 = value for 1 gram equivalent
-------
Section No. 8
Revision No. 1
Date October 1, 155-
Page 16 of 23
8.5 Data Reporting
Directly measured observables typically reported by an organization are
pH, specific conductance, concentrations of major constituents and the amount
of precipitation. Recommended reporting units are
pH - unitless;
conductivity - yS/cm or ymho/cm;
concentration of ion - mg/liter or yeq./literj and
precipitation - cm (1 in = 2.5% cm).
Table 8-3 provides the relation and the transformation factors for converting
mg/liter to yeq./liter for the ions of interest. Specifics on the
computational procedures and/or flagging practices used in precipitation data
reporting are provided below.
8.5.1 Average Concentrations and Deposition - Precipitation-weighted mean
concentrations, C. for various ions are generally reported for a given time
interval (month, quarter, or year).
s • i pj v)i PJ 8"12
where P. is the amount of precipitation in event j (cm), C^j is the
concentration of constituent i for event J (mg/liter), and ME is the number
of events.
For pH or H* concentration, the cumulative value is calculated, and the
precipitation weighted pH value is obtained by converting the cumulative H*
back to pH.
pH = -log jf-i \3 7 8-13
?E
-------
Section No. 8
Revision No. 1
Date October 1, 1SE-
Page 17 of 23
The average concentration at each site can be used to study the
distribution of the constituents as a function of time (month or season}
and/or geographic location. It might also be useful to report the
site-specific, precipitation-weighted standard deviation of the means (S )
as an indicator of constituent variability for different seasons anc
geographical locations. A proposed equation for S , (10,11) is
8-1*
Total deposition of the ith analyte D. per unit area (mg/m ) in a
precipitation event or time interval is calculated as:
Di ' 10 l J ij
where P is the precipitation amount of event J in cm, C. . is the
J * J
concentration of the ith analyte in mg/ liter, K is the total nunber of
11 2 2
events, and the conversion factor is 10 cm /m .
8.5.2 Median Concentrations - Average values are also reported in terms cf
the median (50th percentile value). The medians and their respective
confidence intervals are estimated from the ordered observations obtained
from the independent precipitation samples. The advantages of medians over
means is that extreme values (outliers) have much less effect on medians than
on means, and it is possible to readily compute confidence intervals for
medians. In computing confidence intervals for the medians, let
xd)< x(2)<...
-------
Section No. 8
Revision No. 1
Date October 1, 19
Page 18 of 23
confidence level, P, can be computed from the appropriate a given in the
table i.e., P = 1-a, where a'denotes the significance level.
(b) For n > 20, the approximations below may be used to obtain the confidence
intervals at the 0.95 confidence interval.
r« s 1/2 (n - 1.96/rT) 8-16a
s» = 1/2 (n + 1.96/n~) 8-16b
In general r* and s* will not be integers. Let r and s be the integers
obtained by rounding r* and s* upward to the next higher integer.
For example, for a set of 30 precipitation samples we obtain 29 pH, 22
S0j.s, and 19 NH^* values. In order to compute the appropriate medians and
confidence intervals for each of the three observable parameters,- the
following procedure is followed:
(a) Order the observed values for each of the three observables starting with
the lowest and going to the highest.
(b) For pH: the median will be the value of the 15th element. The lower and
upper bounds of the confidence interval (P s 0.95), obtained from
equations 8-l6a'and 8-16b, are the values of the 10th (r*s 9.22) and 20th
(s*= 19*77) elements, respectively.
(c) For S0j.s: the median will be the average of the 11th and 12th elements.
The lower and upper bounds of the confidence interval (P s 0.95), obtained
from equation 8-l6a and 8-16b, are the values of the 7th (r*= 6.4) and
16th (s*s 15.59) elements, respectively.
(d) For NH^*: the median will be the value of the 10th element. The lower
and upper bounds of the confidence interval (P = 0.94), obtained from
Table 8-6, are the values of the 6th and 14th elements, respectively.
-------
Section No. 8
Revision No. 1
Date October 1,
Page 19 of 23
TABLE 8-6. CONFIDENCE BOUNDS FOR MEDIANS OF SMALL SAMPLES
(n<20)a
Confidence Interval Bounds
Sample Size
(n)
5
6
7
8
9
10
11
12
12
14
15
16
17
18
19
20
Lower
(r)
1
1
1
2
2
2
3
3
4
4
4
5
5
5
6
6
(element number)
Upper
(s)
5
6
7
7 '
8
9
9
10
10
11
12
12
13
14
14
15
a"
0.062
0.031
0.016
0.070
0.039
0.021
0.065
0.039
0.092
0.057
0.035
0.077
0.049
0.031
0.064
0.041
a. Based on Table A3 in W.
J. Conover, Practical Nonparametric
Statistics. 2nd Edition, John Wiley 4 Sons, New York (1980).
b. Significance Level
8.5.3 Reporting and Treating Below-Detection-Lindt Data - Data below the
detection limit (BDL) or the minimal detectable limit for the analytical
method used should be flagged with a code both in the printouts and in all
computer readable data forms. Table 8-7 provides as an example the minima
detection limit criteria used in the NADP program for Laboratory
measurements.
-------
Section No. 8
Revision No. 1
Date October 1, 198U
Page 20 of 23
TABLE 8-7. NADP MINIMUM DETECTION LIMIT CRITERIA FOR LABORATORY MEASUREMENTS
Chemical
Na
K
Ca
Mg
NH,
soj
NO!
POJ
cr
Measurement
Technique
Flame AA
Flame AA
Flame AA
Flame AA
Colorimetry
Ion Chromatography
Ion Chromatography
Ion Chromatography
Ion Chromatography
Detection limit
(mg/liter)
0.01
0.01
0.03
0.004
0.025
0.1
0.02
0.02
0.02
8.5.4 Reporting Out-of-Control Data - Analytical data obtained from a sample
batch in which an out-of-control QC "sample is found should be flagged in the
data base as suspect and should not be used unless the laboratory supervisor
determines that the cause of the out-of-control condition did not affect the
analytical results. Data reported as invalid should not be used in
summaries, statistics, analysed, or other interpretation. For data stored or
reported on computer readable media, it is often best to use a character or
code to indicate out-of-control, missing, BDL, and other such data.
8.6 QC Checks on Final Data
As part of the ongoing QC program, data intended for use in such
summaries should be checked for the parameters indicated below and any
discrepancies brought immediately to the attention of the person with overall
program responsibilities.
8.6.1 Time and Dates of Sampling - Since acid precipitation networks may be
scattered over a large geographic area containing sites in different time
zones, it is useful to reference all event times and dates to one time zone,
-------
Section No. 8
Revision No. 1
Date October 1, 1984
Page 21 of 23
e.g. Eastern Standard Time. Such necessary changes from the local times
usually recorded on field data sheets to this standard reporting time zone
should be checked.
8.6.2 Codes, Flags and Identifiers - Where a code such as that described in
Section 8. 1 is employed for field observations and where a similar coding
system has been employed for laboratory operations, the accurate
transcription of subcodes should be checked at the 10? level.
8.6.3 Overall Transcription Checks - Other observables including rain, gauge
vs. sample precipitation amounts, field measurements (pH, cond.) and other
transcribed information should be checked against field data sheets at the
level.
8.6.4 Spotcheck/Recalculation of Data - The laboratory supervisor should
regularly check and recalculate 5% of the data points which can be spread
over the total sites in the monitoring network and should include the
specially difficult analyses and extreme values.
8.6.5 QC Checks for Data Summaries - Where quarterly or less frequent data
summaries are prepared, the calculations (made by the person with program
responsibility or at his direction) should be verified as being correct from
previously screened, clean, raw data. Usually, one or two sites within the
summary period are selected at random and for those sites one or two analytes
are checked. This is equally important for both manually and computer
derived summaries. Quality control checks on such summaries should include
verification of
1. calculation of precipitation-weighted means
2. calculation of depositions
3- derivation of medians
4. 95* confidence bounds on the medians
-------
Section No. 8
Revision No. 1
Date October 1, 1984
Page 22 of 23
5. reporting of maximum and minimum values.
If yearly summaries derived from quarterly or monthly data are reported,
checks should be made for consistency between the four quarterly summaries or
the twelve monthly summaries and the data reported for the year.
8.7 References
1. Quality Assurance Handbook for Air Pollution Measurement Systems, Vol. V
- Manual for Precipitation Measurement Systems; Part II - Operations and
Maintenance Manual. U.S. Environmental Protection Agency, Research
Triangle Park, NC, EPA-600/4-82-042b (January 1981).
2. Handbook for Analytical Quality Control in Water and Wastewater
Laboratories, United States Environmental Protection Agency, Cincinnati,
OH, EPA-600/4-79-019 (1979).
3* Quality Assurance Handbook for Air Pollution Measurement Systems- Vol. I-
Principlea, United States Environmental Protection Agency, Research
Triangle Park, NC, EPA-600/9-76-005 (1976) Appendix F.
4. T.R. Fitz-Simons and D.M. Holland, The Maximum Likelihood to Probabilistic
Modeling of Air Quality Data, United States Environmental Protection
Agency, Research Triangle Park, NC, EPA-600/4-79-044 (1979).
5. Electric Power Research Institute, "The Utility Acid Precipitation Study
Program: Annual Report on Operations and Results for 1982." Contract
U101-1, Rockwell International Environmental Monitoring and Services
Center (1983).
6. J.L. Mero, The Mineral Resources of the Sea, Elsevler Publ., New lork,
NY (1964) p. 25.
7. Handbook of Chemistry and Physics, U8th ed. Edited by R.C. Weast, The
Chemical Rubber Co., Cleveland, OH 44128, 1967-68.
8. J.P. Lodge, Jr., J.B. Pate, W. Basbergill, G.S. Swanson, K.C. Hill,
E. Lorange and A.L. Lazrus, "Chemistry of United States Precipitation,"
Final Report on the National Precipitation Sampling Network, National
Center for Atmospheric Research, Boulder CO, (August 1968).
9. J.A. Garland, "Enrichment of Sulphate in Maritime Aerosols,"
Atmos. Environ. _15, 787 (1981).
-------
Section No. 8
Revision No. 1
Date October 1,
Page 23 of 23
10. J.M. Miller, "A Statistical Evaluation of the U.S. Precipitation Chemistry
Network," in Precipitation Scavenging, P.B. Semonin and R.W. Beadle,
eds., CONF-741003 (1971*); pp. 639-659. Technical Information Center,
Energy Research and Development Administration, Springfield, VA.
11. H.M. Liljestrand and J.J. Morgan, "Error Analysis Applied to Indirect
Methods for Precipitation Acidity," Tellus, 31, M21-431 (1979).
12. G. Mehls and G. Akland, J. Air Pollut. Control. Assoc., 23, 180 (1973).
-------
Section No. 9
Revision No. 1
Date October 1, 1984
Page 1 of 15
9.0 DATA QUALITY ASSESSMENT
The determination of the precision and accuracy of precipitation data is
the principal means employed to quantitatively assess data quality. Specific
procedures developed for determining precision and accuracy for the
measurement methods used in precipitation monitoring are presented below. In
general, data generation is a three-step process; the first two are
functions of field operations, and the third one is part of laboratory
operations. These are;
1. Collection of the sample in a suitable sampler.
2. Initial field analyses for pH, specific conductance, and weight.
3. Expanded chemical analyses in the laboratory*
Data quality should be routinely assessed for each step in the
measurement system. Precision of .the measurement system is estimated from
data obtained by using collocated samplers and employing identical sampling,
handling and analysis protocols for all samples. Field measurement accuracy
is determined by regular test sample audits. Precision of field and
laboratory analytical methods can be derived from multiple analyses of
certain samples. Accuracy of laboratory analyses is derived from laboratory
analysis of internal and/or external blind audit samples.
9.1 Evaluation of Field Operations
9-1.1 Measurement System Precision - Precision is estimated by duplicate
sampling with collocated precipitation samplers. Each network of sites
operated by an organization should have a minimum of one duplicate sampler of
the type used for routine monitoring. The collocated sampler(s) should be
operated during routine sampling and be installed consistent with the siting
criteria (Section 5.0).
Sampling precision is known to vary among sites and during different
seasons (1). Each organization should therefore develop a schedule for the
-------
Section No. 9
Revision No. 1
Date October 1, 19E-
Page 2 of 15
deployment of the collocated sampler (s) such that all the possible
site/season combinations are satisfied (2).
In order to derive a precision estimate, data from the collocated
sampler(s) and data from the station sampler are compared. The measured
differences in pH (pH units), conductivity (pmho/cm), total mass captured and
concentrations of the various analytes (mg/liter oryeq/liter) are then used
to calculate precision. For each pair of measurements, the signed (SCD) and
absolute (ACD) collocation differences are calculated:
(SCD)
(ACD).
ik
- xik, and
,
9-1
'ik - ljrik -
where y.. is the observable determined for the duplicate sampler for the ith
species and kth collocation event, and xik is the observable determined for
the corresponding station sampler for the same species and collocation event.
Quarterly, the mean, SCD. ., and the standard deviation, S,.., are
estimated for the SCDs of each observable (i) obtained from the jth site
having collocated samplers:
SCDy -
k-1
(SCD)
k.
r
- (SCD).
9-2
9-3
where k. is the number of collocation events at the jth site during a
calendar quarter.
Equations 9-2 and 9-3 could also be used for summarizing the ACDs, if
appropriate. If the network contains more than one site with collocated
samplers, a mean, (SCD)., and a pooled standard deviation, S«,4f for each
observable should be computed quarterly:
(SCD).
(SCD).
-------
S
ai
_ n
Section No. 9
Revision No. 1
Date October 1, 19£J
Page 3 of 15
9-5
where n is the number of sites having collocated measurements within the
network. Equations 9-1* and 9-5 can also be used for summarizing the ACDs, if
appropriate.
An analysis of the distributions of SCDs for the different observables
in precipitation monitoring reveals that they are symmetrical but may not be
derived from a Gaussian (normal) population. Therefore, statistical
prediction methods based solely on the assumption of a normal distribution
may lead to erroneous predictions. To illustrate, consider the summary of
upper and lower SCD limits presented in Table 9-1 for an event sampler
network. The entries in the table correspond to the observed (0) 95% SCD
population limits, i.e. the 97.5th percentile is the upper limit (UL) and the
2.5th percentile is the lower limit (LL). The table also presents the
computed ratio (P/0) where the predicted value (P) was derived for a normal
distribution. The range of P/0 is 0.96 to 23.9 with the majority between
1.00 and 3*00. This indicates that using a normal distribution to estimate
the upper and lower limits for the SCDs would not be very accurate in this
case.
The usual Gaussian approach, which brackets 95} of the SCD population,
ls: UL - (SOftj. + 1.96 S.. for a site
- (SCD). + 1.96 S , for a network
i ai
LL - (SCD)^ - 1.96 Sjj for a site
9-7
- (SCD). - 1.96 S , for a network
Each network should maintain at least one pair of collocated samplers.
For large national networks it might prove more useful to form regional
clusters of sites for the purpose of precision estimation. The collected
samples should be tagged in a consistent manner so their respective
measurements in the field and the laboratory can be easily traced and used to
quantitatively assess monitoring precision.
-------
Section No. 9
Revision No. 1
Date October 1, 1984
Page 4 of 15
TABLE 9-1. NETWORK SUMMARY OF UPPER AND LOWER LIMITS OF THE
SCDs FOR DAILY SAMPLING
Observables
(units)
pH
(pH units)
Hydrogen Ion
(mg/liter)
Total Acidity
(mg/liter)
Conductivity
(ymho/cm)
Sulfate
(mg/liter)
Nitrate
(mg/liter)
Chloride
(mg/liter)
AniniiOTi ^i TO
(mg/liter)
Sodium
(mg/liter)
Potassium
(mg/liter)
Calcium
(mg/liter)
Magnesium
(mg/liter)
Phosphate
(mg/liter)
No. of
Collocation
Events
936
936
847
935
912
912
910
897
886
887
-
885
886
904
Observed 95Z*
Population Limits
LL OL
-0.44
-0.07
-0.06
-13.9
-1.07
A.
-0.72
-0.44
-0.22
-0.33
-0.21
-0.20
-0.03
-0.04
0.39
0.06
0.06
13.4
1.09
0.62
0.58
0.26
0.44
0.30
0.30
0.05
0.05
Predicted /Observed
(P/O)**
LL
1.12
0.96
1.09
1.43
1.05
1.04
2.20
10.6
1.60
2.87
3.78
2.52
23.9
UL
1.21
1.13
0/94
1.50
1.03
1.18
1.75
8.74
1.21
1.97
2.29
1.60
17.8
* Obser id (0) values based on the 2.5 and 97.5 percentiles of the SCDs for Che lover
(LL) and upper (UL) limits, respectively.
* ' P/O ratxo obtained from predicted values based on 95Z probability limits of a normal
distribution, i.e. P - mean ±1.96 std. dev.
-------
Section No. 9
Revision No. 1
Date October 1, 1984
Page 5 of 15
A quarterly precision assessment report containing the descriptors below
should be prepared. These summaries should pertain to entire networks or
regional clusters of sites, as appropriate:
1. SCDs, ACDs and associated standard deviations for each observable
following Equations 9-1 through 9-3 •
2. Network averages (SCD), (ACD), and the corresponding pooled standard
deviations for each observable, following Equations 9-4 and 9-5.
3. Upper (UL) and lower (LL) precision probability limits for each
observable, either for a given site or for the entire network, following
Equations 9-6 and 9-7.
These quarterly summaries should be reviewed by program management, and
out of control conditions identified and corrected. The precision summaries
should be submitted together with the monitoring data to the accountable
organization.
9.1.2 Accuracy of £H and Conductivity Measurements - To assess the accuracy
of field measurements of pH and conductivity, audits should be conducted
using test samples prepared and sent from the central laboratory each month.
The samples should be measured at the site as soon as possible after receipt,
and the results should be returned to the laboratory. The laboratory should
analyze the test sample before the sample is sent to the field. The results
of the laboratory analyses should be included with the field analysis
results. The QA officer should record all the data on the Monthly Field
Audit Report (Section 9*4). The acceptance criteria to be used as control
limits on potential sample degradation are given in Table 9-2.
Accuracy for each measured variable is estimated by computing the
differences:
dj ' fj - (1ij+ Xf j)/2 9-8
where d. is the difference in measurements for jth site (appropriate units),
f j is the field analysis of observable for jth site, Ijj is the initial
laboratory analysis of observable before shipment to the jth site, and 1_ . is
the final laboratory analysis of the observable after return from the jth
-------
Section No. 9
Revision No. 1
Date October 1, 155-
Page 6 of 15
site. Equation 9-8 should be used only if |1 - lf ,| is less than the
acceptable control limits (Table 9-2). If a site value of d. in equation 9-c
exceeds the acceptance criteria in part b of Table 9-2, this should alert tbe
program manager that the pH electrode or conductivity standard solution ma 7
require replacing.
TABLE 9-2. ACCEPTANCE CRITERIA FOR TEST SAMPLE QUALITY AND FIELD
ANALYTICAL ACCURACY.
(a) Laboratory Before and After Audit Measurements (one-asterisk flag if exceeded)
pH £ 5.00 I Before - After I £0.1 pH unit
pH > 5.00 I Before - After I £0.2 pH unit
Cond. £15 umho/cm I Before - After I £2 umho/cm
Cond. > 15 umho/cm I Before - After I ,.n < 15*
After * 1UU
(b) Lab vs. Field Measurement Comparison (two-asterisk flag if exceeded)
pH £ 5.0 lavg. field - average lab I £0.2
pH > 5*00 lavg. field - average lab I £0.3
Cond. £ 15 umho/cm (average field - avg. lab I £ 3 umho/cm
Cond. > 15 umho/cm lavg. field - avg. lab I ,nn < 20%
—•^ ~avgT TaF^ x 100 -
Data should be summarized in the Monthly Field Audit Report by the QA
er, who also computes
observed pH and conductivity
officer, who also computes an average monthly network difference 3 for the
m
m
C d. 9-9
and the monthly variance, S~, or standard deviation, Sm.
-------
3m " I Cm-
Section No. 9
Revision No. 1
Date October 1, 195-
Page 7 of 15
m •
9-10
1) J-l
where 3 is the average monthly network difference for a given variable, d, is
as defined in equation 9-8 above, and m is the number of sites audited during
the month. Monthly results should be summarized in a quarterly report (Section
9.3).
9.1.3 Sampling Bucket Blanks - Sampling bucket blanks should be obtained to
ascertain the levels of constituents that contaminate the bucket over a period
of time when there is no precipitation. This contamination can be due to poor
techniques used for bucket cleaning and handling in the laboratory and field, or
may be due to dry deposition entering the bucket around the lid seal. A bucket
blank is obtained by cleaning a sample bucket that has been in the collector for
seven consecutive days without an event occurring and analyzing the rinse
solution. A known amount of distilled or deionized water, e.g., 100 g, is used
to rinse the bucket, and this rinse is analyzed for the constituents SO]!, N(L,
Cl", NH^, Na*, K*, Ca* and Mg* . The ionic quantities measured should be
corrected for the deionized water blank.
The bucket blanks serve as a quality control check in two respects. First,
any increase in bucket blank contamination over time nay indicate the presence
of a problem. Second, data from bucket blanks may be used as a correction for
event concentrations or deposition averages. Bucket blanks are expected to vary
from site to site and season to season with the largest values occurring at
arid, windy sites with poor lid seals. Thus blank averages are most useful if
they have been derived for a site over a significant time period, such as a
quarter of the year.
The blank deposition can be used as an indicator of a problem. The blank
deposition is calculated from equation 9-11 with P given by W/k(cm), where ¥ is
the weight of the deionized water rinse and k is the conversion factor for
sample weight to centimeters, e.g., 640 g/cm for the Aerochem Metrics bucket.
-------
Section No. 9
Revision No. 1
Date October 1, 19c-
Page 8 of 15
The blank deposition then is
Db (mg/m2) = 10 P C = 10 (W/k) C S-H
where C is the analyzed concentration of the ion in mg/liter.
Since C and W are inversely proportional, the deposition is independent of
the amount of rinse water. In contrast, concentration varies with the amount of
water used and would have to be calculated as a function of W for comparisons.
After a data base of blank values has been accumulated for each site and
quarter, a mean or a median can be obtained and updated. The median ion
deposition in blanks is generally less than 8% of the median precipitation
values (1). The occurrence of extreme blank values, e.g., those greater than
the lowest 5% of event sample values, or, alternatively, greater than 3 to 10
times the blank median, can be used as an indication of unduly high
contamination. The actual upper limit will vary with the analyte, site and
season.
The blank values can also be used to correct event value averages.
However, this is not a straightforward procedure since the conditions under
which the blanks and the event samples were obtained may be quite different.
Corrections should be made only on an individual site and seasonal basis. To
simplify matters, the median or average blank values for the analyte for a
particular quarter is subtracted from the median or average event value for the
same quarter and site. For concentration, the median bucket blank value for the
ith analyte C., must be corrected for the rinse water blank and calculated for
the median precipitation quantity P, i.e:
Ci(mg/liter) s (Cg-C^W/P 9-!2a
or
C^ijeq/1) = lO^dng/literVEq.Wt. 9-12b
where CB and Cy are the measured concentrations in mg/1 for the analyte in the
-------
Section No. 9
Revision No. 1
Date October 1, 1984
Page 9 of 15
bucket rinse and the deionized water, respectively, P denotes the median
precipitation volume (ml) or weight (g) for the quarter, Eq. Wt. is the
equivalent weight of the analyte, and W has been defined above.
If a daily sampling schedule is followed, the above C, which is for a
weekly time interval, should be divided by the average number of daily samples
per week. (It is assumed here that the chief source of contamination is
deposition of dust in the bucket while in the sampler and that deposition
increases uniformly with time.) Table 9-3 shows the medians of both the
absolute analyte blank values and the relative magnitude of these blanks
expressed as percentage of the median precipitation values, for the 1982-1983
OAPSP network (3). The blank data are for 224 weekly samples taken in the last
quarter of 1982 through October 1983 from most of the sites, whereas the
precipitation medians are for approximately 1600 daily samples from the entire
network for 1982. The blank data have been normalized for a median
precipitation volume of approximately 500 ml but have not been corrected for the
average time interval between samples. Since the number of samples per week
averaged slightly less than two, the concentrations listed should be halved.
The data indicate that the cations generally have larger blank
contributions than sulfate and nitrate, and are consistent with soil dust being
the chief source of contamination. However, even without the time correction,
the largest bucket blank median value (for potassium) is only 6% of its median
precipitation concentration. Thus, these blank contributions to the
precipitation averages are small and can be neglected. For specific sites and
time intervals the blank concentration may not always be negligible, in which
case the blank concentration should be subtracted from the average.
For those sites which frequently show large blank values, the sampler lid
seals should be checked and/or replaced as necessary. If the seal is
functioning properly, the problem may be caused by excessive winds. The
pressure of the seal can be increased as indicated by the sampler manufacturer.
If neither of the above corrects the problem, bucket preparation and handling
techniques should be evaluated and changed as necessary.
-------
Section No. 9
Revision No. 1
Date October 1, 1984
Page 10 of 15
TABLE 9-3. ABSOLUTE AND RELATIVE BLANK VALUES.
Observable
: so,
N03
Cl
MHU
Na
K
Ca
Mg
Median Blank Values* 'b
(mg/liter)
0.0086
0.0068
0.0043
0.0068
0.0009
0.0012
0.0048
0.0010
Relative Median Blank Values0
(*)
0.5
0.5
3.0
3.0
1.5
6.0
5.0
5.0
a. (Blank - D.I. H_0)Concn. x Rinse vol. /500
(Median precipitation volume ~ 500 ml)
b. For 224 samples
c. expressed as median blank value as a percent of median
concentration in approximately 1600 precipitation samples.
9.2 Evaluation of_ Laboratory Operations
The QA officer should routinely assess the reproducibility, precision, and
accuracy of laboratory chemical analyses of precipitation samples in order to
obtain a representative indication of overall data quality.
9.2.1 Analytical Precision - To estimate the contribution of analytical
variability to total variability, duplicate analyses should be performed on
approximately 10% of all the precipitation samples routinely analyzed by the
laboratory. Samples randomly selected for replicate analyses by the QC chemist
should contain a large quantity of precipitation. The split samples should be
-------
Section No. 9
Revision No. 1
Date October 1, "934
Page 11 of 15
properly identified, and the results of such duplicate analyses recorder and
reported to the QA officer and program manager in the Report of Duplicate
Analyses (Section 9.1*). Analytical precision is commonly defined and reported
as the standard deviation of duplicate analyses. An additional factor of 2 is
included in the denominator to take into account the random error associated
with both measurements. Thus, for precipitation measurements analytical
precision for the ith analyte is defined as
S-13
where d. denotes the difference for the duplicate analysis of the ith analyte,
and N. is the number of sample pairs for the ith analyte in the reporting
period.
9.2.2 Accuracy of Chemical Analysis - Accuracy of chemical analyses should be
determined monthly from results of the analyses of blind samples submitted to
the laboratory by randomly selected field sites. These samples, prepared by
diluting various precipitation standards (NBS or EPA), should be shipped in
sealed plastic bottles to the field sites. Each sample should be identified and
accompanied by two postcards. On arrival at the field site, the sample should
be refrigerated at JJ°C until it is forwarded to the laboratory. At the end of
the first week in which no event has occurred, the sample should be transferred
to a clean weighed container generally used for shipment (plastic bottle,
plastic bag or sampling bucket). The sample should be weighed, and an aliquot
should be measured at the site for pH and for specific conductance. A field
data form should be filled out with the measured values and the other required
data. To do this, a simulation of a normal precipitation record will have to be
used. Sample and site identification, sampling date(s), pH, conductivity and
weight should be recorded on postcards. The sample and the data form should be
sent to the laboratory (as for an event sample), and the postcards mailed to the
QA officers for both the monitoring network and for the blind sample preparation
laboratory; thus, if one card is lost, the information will be available from
the second. Once received at the laboratory* such a sample is treated as a
-------
Section No. 9
Revision No. 1
Date October 1, 1984
Page 12 of 15
routine sample and is analyzed in the normal manner for all constituents of
interest.
At regularly scheduled intervals, when the analytical laboratory sends a
printout of sample results to the QA officer, he should identify the blind field
samples and transfer their data to the QA data file. He should also notify the
analytical laboratory to delete these results from the precipitation data file.
The laboratory is then informed on its performance analyzing the QA samples.
Data should be summarized by the same computational procedure indicated by
Equations 9-1 through 9-7. The QA officer should also obtain QA samples from
EPA's EMSL Quality Assurance Division and/or from the U.S. Geological Survey
semiannually as an external check on analytical accuracy.
9.3 Data Quality Reporting
The QA officer should provide the following reports to the program manager
on a routinely scheduled basis:
1. Measurement System Precision — Quarterly reports, summarizing all data on
collocated samples and duplicate analyses performed during the three months,
should include differences In analytical results for the split samples and
the average difference for each analyte. The report should compare average
differences with the QC ranges typical for the laboratory and the measurement
method.
2. Analytical Accuracy — Quarterly reports, summarizing data on at least three
blind QA samples analyzed by the laboratory in the three months, should
include the different constituent concentrations and an average difference
and standard deviation computed for each. Data for the comparison obtained
by the QA officer should include also a summary of all spike recovery data
for the reporting period.
3. Field Measurements Accuracy ~ Quarterly reports should give the accuracy of
the measurements performed by the field operator using audit or test samples
sent from the laboratory on a monthly basis. Audit results should include
the laboratory analyses of the sample before it was sent to the field and the
field analysis. The data summary should give the mean and standard deviation
for each station and for the network as a whole. Typical acceptance criteria
for pH and conductivity are presented in Table 9-2. These data should be
used to indicate when a bad pH electrode or conductivity standard needs
replacing.
-------
Section No. 9
Revision No. 1
Date October 1,
Page 13 of 15
4. Site Evaluations — Semiannual reports should be made providing a qualitative
evaluation of the site and operational procedures. The auditor visiting the
field sites summarizes his observations and provides recommendations for
corrective action, if appropriate. The auditor observes and documents the
operator's performance in the analysis of unknown performance evaluation
samples for pH and conductivity. The report should also summarize instrument
performance checks conducted on the precipitation sampler and the rain gauge
and any recalibration that has been performed on the latter.
9.4 Data Forms
Blank data forms on the following pages were taken or adapted from EPA
forms and from other references. The titles are at the top of the figures as is
customary. The two forms included here are:
1. Monthly Field Audit Report
2. Report of Duplicate Analyses
9.5 References
1. L. Topol, "Precision of Precipitation Chemistry Measurements11, Proceedings
APCA Specialty Conference on Atmospheric Deposition, SP-49, pp. 197-209,
November 1982.
2. R.J. Schwall, M. Lev-On and L. Topol, "Guide for Collocation Planning" Final
Technical Report to Work Assignment No. 51, EPA Contract No. 68-02-3767,
Rockwell International Environmental Monitoring and Services Center
(EMSC8391. FR), in preparation.
3> Electric Power Research Institute, "Acid Precipitation in the Eastern United
States (1978-1980)", Contract Noa. HP 1376-1 and RP 1630-2, Rockwell
International Environmental Monitoring and Services Center, Final Report.
-------
Section No. 9
Revision No. 1
Date October 1,
Page 11 of 15
MONTHLY 'FIELD AUDIT REPORT
Sample #:
Date of Preparation of Field Audit Sample:
(Analyst Signature)
LABORATORY ANALYSIS BEFORE SHIPMENTa
TO THE FIELD
LABORATORY ANALYSIS AFTER RETURK2
FROM THE FIELD
Date:
Conductivity
PH
Date:
Conductivity
PH
1.
2.
3.
+Average
Std. Dev.
1.
2.
3.
^Average
Std. Dev.
LABORATORY ANALYSIS OF AUDIT SAMPLES VS. FIELD ANALYSIS
Conductivity (mho/cm)
PH
Site
Field
Date
Field
Analysis
Lab"
Analysis
Dirt.
Field
Analysis
Lab
Analysis
Diff/
a) These data are for three laboratory aliquots which are analyzed before
shipment of samples, are then refrigerated, and are reanalyzed with tie
samples returned from the field.
b) Values after return from the field.
c) Diff. = Field-Lab
-------
Section No. 9
Revision No. 1
Date October 1, 1984
Page 15 of 15
REPORT OF DUPLICATE ANALYSIS
Analyte
Routine^ Duplicate"
Analysis (ID ) (ID )
Technique Date Result Date Result
Diff.c
PH
Conductivity
Sulfate
Nitrate
Chloride
Phosphate
Carbonate
Bicarbonate
Acidity
Strong Acid
Ammonium
Sodium
Potassium
Magnesium
Calcium
a. Value reported as routine sample
b. Duplicate sample, might have different ID
c. Diff. = Duplicate - Routine
-------
Section No. 10
Revision No. 1
Date October 1, 1£6i
Page ' of 58
10.0 ACID PRECIPITATION MONITORING PROGRAM EVALUATION
Establishing a schedule for audits and independent checks to evaluate
the quality of data provided by the total measurement system is an important
part of an overall Quality Assurance Program. A complete audit for a
precipitation chemistry network should include both qualitative and
quantitative evaluations. The quantitative evaluation, derivation cf
estimates of precision and accuracy, has been discussed in Section 9-0-
Overall program audits, site evaluations, laboratory systen audits and
performance audits will be discussed here.
10.1 Program Audits Guidance
A program audit is a qualitative on-site inspection and appraisal of the
QA efforts used for the program. The topics that should be reviewed by
audits and the sections of this manual pertaining to them are as follows:
1. Overall Program Operation (Management):
Background material - Sections 2.0 and 3*0
Review questionnaire - Section 10.2, items A through E;
2. QA Project Plan and Its Implementation:
Background material - Sections 2.2 and 2.3
Review questionnaire - Section 10.2, item F;
3. Site Documentation:
Background material - Sections 4.0 and 5.0
Review questionnaire - Section 10.3;
4. Site and/or Field Operation:
Background material - Section 6.0
Review questionnaire - Section 10.2, item G;
5. Support/Analytical Laboratory Operation:
Background material - Section 7.0
Review questionnaire - Section 10.4;
6. Performance Audits:
Background material - Sections 6.0, 7.0 and 9.0
Specific guidance - Section 10.5;
-------
Section No. 10
Revision No. 1
Date October 1, 196-;
Page 2 of 58
7. Data Handling, Analysis, Validation and Reporting:
Background material - Section 8.0
Specific guidance - Section 10.6.
For each monitoring network, a program audit should be conducted as socr
as possible after the start of monitoring; subsequent audits should be
conducted at least once each year thereafter. In order to conduct a program
audit effectively, the auditor or audit team should:
(a) Decide on the audit scope - this should depend on the size of the program
and the use of the data derived from the program.
(b) Select appropriate questionnaires and have organization personnel complete
and return them prior to the on-site visit. Several questionnaires and
checklists suggested for this application have been prepared and included
in this manual.
(c) Review the completed questionnaires and supporting documentation prior to
the on-site visit in order to become as familiar as possible with Program
Operations and to discover potential problem areas ahead of time.
(d) Set-up and carry out on-site evaluation and interviews - the auditor
should, at a minimum, visit the program operations headquarters, a field
site, and field laboratory. If the support analytical laboratory
operations are to be evaluated, then the .auditor should visit the
laboratory also. He should interview responsible individuals and members
of the technical staff with respect to the material presented in the
completed questionnaire.
(e) Conduct an exit briefing with program personnel to give a preliminary
report of audit findings and to set some expected time frame in which the
organization can expect to receive the audit report.
(f) Prepare, edit and deliver a written report of audit results and program
assessment in DRAFT form; receive and incorporate comments made by the
organization and issue as a completed document.
10.1.1 Support Material - For a program audit the following material should
be used:
1* Overall Program Operation; Field and Field/Lab Operations Questions - The
auditor should use questionnaires similar to or identical with the two
questionnaires given in Section 10.2. The first of these questionnaires
is intended to cover management and organizational activities while the
second addresses site operation and field measurement activities.
-------
Section No. 10
Revision No. 1
Date October 1,
Page 3 of 58
2. QA Project Plan - This document should be checked for completeness against
the material found in Section 2.0 and throughout this manual.
3« Site Documentation - A completed checklist/form such as that given in
Section 5.0 should be obtained for at least one site in the monitoring
network. The auditor should visit that site and verify the accuracy of
the information written on the form. The questionnaire given in Section
10.3 of this chapter may be used for this purpose.
**• Analytical Laboratory Operations - The auditors should use a questionnaire
based on or identical to that contained in Section 10.4.
10.1.2. Reporting - At the conclusion of any program/laboratory/site audit
assignment the auditor should prepare a complete evaluation report. A
possible format for such a report would include, as appropriate:
1. General Background - This section identifies who was interviewed and who
was present for the audits and their affiliation. A general overview of
audit procedures should also be given.
2. Field Laboratory - This may be included where such a laboratory is
maintained within, or is supported by, the central analytical laboratory.
The field laboratory and site operation may be evaluated as a separate
entity or during an audit of overall operations.
3« Analytical Laboratory - This section • should describe current ongoing
operations in sufficient detail as to present a complete understanding of
the level to which both QC and QA have been implemented. Since this is
the major portion of the report, this section will necessarily include
examples of documentation which the auditor feels are critical to an
understanding of operations. Since support to a precipitation monitoring
program requires many different types of analyses, the auditor may choose
to describe in detail two or three analyses which he considers
representative of laboratory operations.
4. Data Management - This section should describe data management practices
and the levels of data screening, QC checks and independent data
processing audits employed.
5. Conclusions and Recommendations - In this section the auditor should
discuss the audit findings with a view to potential data impact, and where
possible, indicate potential courses of corrective action.
The report should be prepared in DRAFT format and submitted to both the
audit requestor and the audited organization. This allows for the early
clearing of potential misunderstandings and points of contention and offers a
-------
Section No. 10
Revision No. 1
Date October 1,
Page 4 of 58
mechanism for the inclusion of additional information necessary to complete
the report. Once the comments to the DRAFT report have been received, they
should be reviewed and incorporated. The report can then be issued in final
form and copies sent to responsible parties for appropriate distribution.
-------
Section No. 10
Revision No. 1
Date October 1, 198M
Page 5 of 58
10.2 Overall Program Operation Questionnaire
GENERAL INFORMATION
Questionnaire Completion Date:
On-Site Visit Date:
Organization Name and Address:
Telephone No. FTS: Commercial: ( )_
Person Completing Questionnaire
Position:
Telephone No. ( )_
Organization Director:
Monitoring Supervisor:
Quality Assurance Officer:
On-Site Audit Conducted By:
Affiliation of Auditor(s):
Persons Present During Entrance Interview:
Persons Present During Exit Interview:
-------
Section No. 10
Revision No. 1
Date October 1,
Page 6 of 58
A. GENERAL
1. How long has the program been operational?
Number of Sites
Operating Since
2. What is the objective of the monitoring program?
Baseline
Trends
Other
3. Provide a current organizational chart indicating each person's
participation in the current program.
4. Have the following been prepared, approved, issued, revised?
QA Project Plan Date
Documentation on Sites and Network Date
Standard Operating Procedures for Field Sampling Date
Standard Operating Procedures for Analytical Lab Date_
Does the program operate in compliance with
EPA Protocol? Yes No Comments
NADP Protocol? Yes No Comments
NTN Protocol? Yes No Comments
Other Protocol? Yes No Comments
6. Was the operating protocol derived from any of the above and modified to
meet network needs?
Explain
-------
Section No. 10
Revision No. 1
Date October 1,
Page 7 of 58
7. Indicate number of sites currently operational as part of network?
8. How many of these sites have collocated instrumentation for precipitation
measurements?
-------
Section No. 1C
Revision No. 1
Date October 1,
Page 8 of 58
B. STAFFING
Please include a list of educational background, experience and training
for each responsible person identified in the program organization charr.
Are the following adequate to current and proposed program operation?
Staff Site? Yes No Comment
Organization? Yes No Comment
Staff Qualifications? Yes No Comment
Staff Utilization Yes No Comment
3. Do staff members receive regular and periodic training to maintain
and upgrade Job skills? Please indicate examples of each responsible
individual's training including period and training method (course?,
on-the-job?, etc.)
. Are staff members adequately conversant with appropriate standard
operating procedures to carry out job duties?
Yes No Comment
5. Has a staff member been identified as a Quality Assurance Officer?
Yes No Comment
If not, who handles this responsibility?_
(Name)
Who does he report to?
(Name and Title)
6. Which of the following references are available to staff members?
(a) Atmospheric Environment
(b) Journal of the Air Pollution Control Association
(c) Environmental Science and Technology
(d) (Other)
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page'9 of 58
C. NETWORK DESIGN
1. Are all sites documented according to specified criteria?
Yes No Comment
(Please attach an example of the documentation for one site)
2. Has the network been designed in accordance with stated program
objectives? Include a brief description of any siting compromises.
Yes No Comment
3. Is there a written plan describing the overall network?
Yes Title Date
No
4. Does the organization have records identifying the status and history
of each site? Does it include
(a) Some site identification?
Yes No Comment
(b) Site coordinates and elevation?
Yes No Comment
(c) Photos or slides, taken to adequately show siting?
Yes No Comment
(d) Date monitoring initiated?
(e) Model, manufacturer and serial numbers of equipment at the site
and sampling schedule?
(f) Reason for periods of missing data?
Is equipment installed at site in accordance with
(a) Manufacturer's specifications? Yes No Comment
(b) Network guidelines? Yes No Comment
(c) Sound scientific principles? Yes No Comment
-------
Section No. 10
Revision No. 1
Date October 1,
Page 10 of 58
6. Does the network design consider
(a) Access? Yes No Comment
(b) Power availability? Tes No Comment
(c) Potential localized interferences such as closely located sources?
Tes No Comment
7. How often are sites visited?
8. How often are samples removed?
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 11 of 58
D. NETWORK OPERATION
1. Is equipment in the network operated in accordance with
organization's standard operating procedures (where such exist)?
Yes No Comment
2. Are the operating procedures compatible with
(a) Agency QA plan? Yes No Comment __
(b) EPA guidelines? Yes No Comment _
(c) Manufacturer's.recommendations? Yes No Comment
3. Is equipment operated on a (documented) schedule? (Please attach
a copy for one site.) Yes No Comment
4. Are an adequate supply of spare parts and expendables maintained at the
site by the network to minimize downtime?
Yes No Comment ^
5. Are all sites operated year round? Yes No
Explain Schedule
6. Is a bound logbook maintained at the site? Yes No
containing records of site visits?
problems?
data?
7. Is routine minor maintenance performed regularly at the site?
Yes No Comment
By Whom:
(Name - Position)
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 12 of 58
8. Does the person performing such maintenance possess (or have access
to):
(a) Standard troubleshooting/maintenance procedures? Yes No
Comment
(b) Instrument manuals? Yes No Comment
(c) Other guidance? Explain
9. Indicate which tasks (if any) are included as part of site operation
duties:
Task Frequency
10. Are any measurements made on samples at sites? Yes No_
Comment
At Site Freq. Measurement Device
(Yes/No) (Times per wk)
Cond.
PH
Precip. depth
Wt/Vol.
Other
(Attach pages as necessary.)
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 13 of 58
11. How are samples shipped to analytical lab? (Please circle
appropriate response.)
in buckets without field measurements by truck
in bottles with field measurements hand delivered
other1 by mail
other*
•Explain
-------
Section No. 10
Revision No. 1
Date October 1,
Page 14 of 58
E. DATA AND RECORDKEEPING
1. Please indicate data sources and, as necessary, attach examples of,
or briefly describe, the data format?
(a) field site/field lab data include:
rain gauge charts
copies of data sheets
copies of logbooks
other
(b) analytical lab data include:
analytical results
calibration data
separate QC data
(c) other data source used in conjunction with acid precipitation:
meteorological data fc
aerometric data
source emission data
2. Are field data checked for reasonableness?
Yes No Indicate what is checked
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 15 of 58
3. Are analytical lab data checked for reasonableness?
Yes No Indicate what is checked
1. Are a portion of data from field reverified by lab (such as
duplicate pH, conductivity or weight measurements)?
Yea No Specify
5. Are such crosschecks used to validate or flag data?
Yes No Indicate any cutoff points
6. How are data finally reported? How often?
7. Where and how are data archived? For how long?
-------
Section No. 10
Revision No. 1
Date October 1, 198M
Page 16 of 58
F. QUALITY ASSURANCE
1. Is there a defined QA function ongoing within the network?
Yes No Comment
2. Is this function independent of all routine operations?
Yes No Comment
3. Does the individual, responsible for this function regularly evaluate
or audit the following operations?
(a) Site operations (Performance Audits)? Yes No
Comment on Frequency
(b) Site data? Yes No
Indicate % of data recalculated ..
(c) Analytical laboratory operations? Yes No
Indicate dates of last audits of lab
(d) Analytical laboratory data? Yes No
Indicate % data recalculated
Does QA maintain (and/or prepare) independent check solutions or
standards specifically used to monitor accuracy?
Yes No Comment on types, concentrations and uses
Does the QA function include the use of EPA supplied check samples
such as from interlaceratory surveys?
Yes No Comment on frequency and analytes checked
-------
Section No. 10
Revision Mo. 1
Date October 1, 1984
Page 17 of 57
6. Does the analytical support lab participate in EPA, USGS and other
interlaboratory round robin test programs?
Yes No Comment on frequency and attach last results summary
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 18 of 57
G. FIELD SITE EVALUATION
This portion of the questionnaire concerns measurements made by field or
support personnel independent of any measurements made by the analytical
support laboratory. If measurements are made only by the analytical lab,
please mark only those applicable.
This part of the questionnaire is to be repeated for each site visited
during a program audit.
1. Site address
Designation (no./identifier)
2. Does the agency have the necessary hand tools, electrical testing
and calibration equipment to operate and maintain equipment,
calibrate rain gauges and repair samplers at the site?
No Comment
3. For precipitation collection are the following types of equipment used?
(a) Automatic precipitation collectors?
(b) Bucket manual-type collectors?
(c) Recording rain gauges (sensitive to +0.01 in. (0.25mm)?
(d) Event pen markers on rain gauges? •_
Are buckets cleaned at site? Yes No
Identify responsible person ~
5. Does site have adequate supply of deionized water? Yes No_
(indicate source) and average conductivity (yS/cm)
6. Please indicate the types, make and model of field measurement
equipment (attach pages as necessary - not necessary if site
documentation has been attached).
7. Are there an adequate number of clean buckets kept at the site?
Yes No Indicate number usually on hand
8. Is the collector sensor cleaned periodically with deionized water?
Yes No How often?
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 19 of 57
9. For wet/dry collectors, is the rim of the dryfall bucket wiped
clean regularly?
Yes No How Often?
10. Is the rain sensor tested regularly?
Yes No Each Site Visit?
How Often?
By what method?
11. Is the dryfall bucket inspected for moisture at each site visit?
Yes No How Often?
12. Are rain gauge pens (weight trace and event) checked for ink?
Yes No How Often?
13. Is the rain gauge clock wound at prescribed intervals?
Yes No Indicate Interval?
14. Is the clock accurate to +1/2hr per week?
Yes No Comment -_
15. Indicate the frequency of calibrations for
(a) Rain gauge per
(b) pH meter per
(c) Conductivity meter per
(d) Balance for precipitation weighing per
16. Are rain gauges calibrated
(a) upon installation? Yes No_
(b) at least semi-annually? Yes No_
(c) after major maintenance? Yes No
(d) when performance audits indicate the need? Yes No_
17. Is a conductivity standard solution kept at site?
Yes No Indicate Source
Indicate cone. umho/cm. Kept for how long?
18. Are the shelf life and accuracy of conductivity standards documented?
Yes No Comment
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 20 of 57
19. Is the conductivity meter calibrated under the same conditions as
used for the samples?
Yes No Indicate Discrepancies
20. Are standard pH buffers kept at site?
pH 4
PH 7
pH 8
Other
Yes
(source)
(source)
(source)
(source)
No
21. Is the pH meter calibrated with simulated precipitation reference
solutions in addition to standard buffers?
Yes No Frequency
Source of reference solutions
Briefly outline procedure used
22. Is conductivity standard kept refrigerated when not in use?
Yes No Comment
23. Are pH and conductivity meter calibrations checked at least at one point
immediately prior to sample measurement?
Yes No Briefly outline procedure
24. Are measurements made at known temperature?
Yes No Indicate Temp.°C
25. Where are measurement data recorded?
(a) Site Logbook?
(b) Data Sheet?
(c) Other?
-------
26. Are other check solutions maintained at site?
Conductivi ty
level
source
frequency
of use
Section No. 10
Revision No. 1
Date October 1, 1984
Page 21 of 57
replacement
interval
level
source
use
shelf life
27. Is a check on the rain gauge calibration made regularly?
Yes No Indicate Frequency
28. Is the outside of wet bucket wiped dry before weighing?
Yes No Comment
29. Is precipitation measured by weight? Yes_
by volume? Yes No
No
30. Has the balance used to weigh precipitation been calibrated?
Yes No Frequency
31. Has the balance calibration been performed with traceable weights?
Yes No Indicate traceability of weights
32. Is balance zeroed before each use?
Yes No Comment
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 22 of 57
33. How are pH and conductivity cells/electrodes stored between use?
p_H Conductivity
buffer (indicate pH) conductivity standard
deionized water deionized water
other other
3*. Are samples allowed to come to room temp, before measurements are
made? Yes No Comment
35. Are separate sample aliquots used for pH and conductivity?
Yes No If no, indicate which measurement is
made first
36. Are aliquots discarded after use?
Yes No Comment
37. How are samples shipped to laboratory? (Circle appropriate response)
In buckets/In bottles/In plastic bags
By air/By surface mail/By truck
With cold-packs/At ambient temperature
Comments
38. Are samples shipped within 2U hours of collection?
Yes No Briefly describe sample storage and
treatment prior to shipment.
39. Do copies of field measurements accompany sample? Yes_
No Are any additional copies made? How many __
Purpose
-------
Section No. 10
Revision No. 1
Date October 1, 198M
Page 23 of 57
*0. Do these records indicate (please attach an example data sheet if
possible)
date of event Yes No
beginning and ending dates
for emulative sampling period Yes No
amount of precipitation Yes No
temperature Yes No
pH Yes No
conductivity Yes No
signature Yes No;
additional comments Yes No
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 24 of 57
10.3 Site Documentation Evaluation
10.3.1 General Guidance - During the initial phase of network installation,
each site should be documented using a site documentation form such as that
included in Section 5.0 or that used by the NADP. This form should be
conpleted by organization personnel to record station location, site
classification, station instrumentation, topography and important pollutant
sources. This documentation should be repeated at least annually thereafter.
It is important that the information contained on such site
documentation be verified as accurate. While it does not fall within the
scope of the quality assurance function to prepare these site documents, the
Quality Assurance officer should verify, for a small number of sites, that
the information contained in such documents is accurate and complete. He
should note any changes which may affect data quality and notify organization
management of such problems. Of particular importance in this regard are
sites where collocated instrumentation has been placed; such data may be
used to estimate measurement or data precision.
The suggested questions in 10.3.3 should only be used as guidance and
should be modified as necessary to fit the exact documentation used in the
Site Description Report. In general, the site evaluation auditor will
perform the following tasks:
(a) obtain from organization management completed copies of site description
do cumentation;
(b) schedule on-site visits for all or a representative number of sites;
(c) evaluate site and seek information to answer those questions given in the
questionnaire (10.3-3 or Appendix B); and
(d) prepare a site evaluation report.
Tiis site documentation evaluation may be performed concurrently with the
site visitations required for program audits.
10.3.2 Site Evaluation Reporting - At the conclusion of a site evaluation or
evaluation of a group of sites for a single organization, the auditor should
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 25 of 57
prepare a brief written report. This report should include at least a
discussion of observations made during the site visit as noted in the
questionnaire and a copy of the site documentation used for the evaluation.
Where major discrepancies are noted) additional information needs to be
included. If further documentation has been provided by the auditor, a newly
completed accurate site description document should be attached.
Recommendations to improve siting and thus the data quality obtained from the
respective sites should be included.
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 26 of 57
10.3.3 Site Documentation Review - to be completed for each site reviewed
1. Site Address
Designation (Number/Identifier)
2. Has the data acquisition objective changed?
les No Comment
3. Verify the longitude and latitude by independently obtaining
maps of the area. OK Problem
4. Are the names, addresses and identification of responsible
individuals still valid? If not, note changes.
5. Verify that all instrumentation is present and note any that are
not operational. Give reason for_non-operation and estimate of
down-time. Is this a potential data impact or? Comment
6. Has additional equipment been added since the site documentation
was prepared or equipment removed or changed? Add any changes to the
equipment list.
7. If measurements are made at the site (or closely located site
laboratory), verify the indicated information on type, model,
description, etc., of pH and conductivity meters, balance, etc.
OK Problem
3. Is there a map indicating location and distances to the major
sources which may affect data gathered at the site?
Yes No
-------
Section No. 10
Revision Mo. 1
Date October 1, 1984
Page 27 of 57
9. Is it still valid? Yes No Or have the number and/or
location of sources changed?
Note problem areas
10. Review sketch of map. Is it complete with respect to indication of
roadways, parking areas, buildings (including number of stories),
tree lines, power lines, bodies of water, and fences?
Complete Incomplete (note problem areas)
11. Verify all distances using a tape measure or rule? Indicate
significant discrepancies.
12. Walk around the site and compare view in the four cardinal
directions with that as given in the site photos. If photos
have not been included with the site documentation, the
auditor should take at least one in each of the four cardinal
directions (N,S,E&W) looking outward from the main sampler.
13. Are there any obstacles with a height that subtends an angle of 30°
with the ground horizontal?
1U. Are the precipitation collectors and/or rain gauges at least
7 feet (2 meters) apart and no further than 15 meters apart?
15. Are rain gauge and precipitation collector placed in a line
perpendicular or parallel to the prevailing wind, or in the
direction specified for network sites? If parallel, is the
wet bucket end upwind of the rest of the collector?
-------
Section No. 10
Revision No. 1
Date October 1, 1981
Page 28 of 57
16. Is the rain gauge level? Yes No Comment
17* Is the access door to the rain gauge on the leeward side of the wind
path? Yes No Comment
18. Is the rain gauge capable of measuring 0.01" (.025cm) of precipitation?
Yes No Comment
19. Is the precipitation fall to the sites unobstructed? (The auditor
should comment on vegetative obstructions such as trees which do
not now pose any problem but which may impact precipitation within
the next few years.)
20. For collocated precipitation collectors is the distance between
then 7-45 ft (2-15 meters)?
Yes Ho Comment
21. Will there be any changes made to the site or site equipment in
the near future? Note the intended changes and schedule, and
estimate any potential data impact. (Attach sheets as necessary.)
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 29 of 57
10.4 Laboratory Operations Evaluation
Each analytical support laboratory should be evaluated at least once
each year to qualitatively assess the laboratory's ability to produce
analytical data of high quality. Such an evaluation is commonly referred to
as a systems audit and is performed in a manner very similar to that
described in Section 10.2 for systems audits of overall operations.
10.4.1 Procedure - A laboratory systems audit is normally conducted in three
steps. First, a questionnaire, such as that included in Section 10.4.2, is
sent to the analytical laboratory prior to the audit visit. The laboratory
should then fill out the questionnaire as completely as possible and return
it with sufficient documentation through the use of attachments. Second, the
questionnaire is reviewed by the auditor to become familiar with the system
operations and to determine any weaknesses and potential problem areas.
Third, after the questionnaire has been reviewed, the onsite interviews are
scheduled. The preliminary review of-the questionnaire serves the purpose of
allowing a greater amount of time to be spent onsite examining potential
problem areas.
The auditor should interview the laboratory manager, any person who has
direct analytical responsibility for precipitation sample analysis, personnel
associated with data validation, analysis and reporting, and the person
identified by the laboratory manager who has responsibility- for quality
assurance. The information gathered from these interviews should be complete
and up to date and should present an adequate picture of the current and
proposed levels of implementation of all quality assurance activities,
including internal quality control.
At the conclusion of the series of interviews, the auditor should inform
the laboratory manager of the audit interview results and discuss any
potential data impacting problems uncovered. This is commonly referred to as
an exit interview. During this activity, the auditor also explains the
reporting procedures and schedule.
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 30 of 57
10.4.2 Analytical Laboratory Questionnaire
GENERAL INFORMATION
Questionnaire Completion Date
On-Site Visit Date
Laboratory:
Street Address:
City: State: Zip:
Laboratory Phone No. (Area CodeX )
Organization Director:
Laboratory Director:
Quality Assurance Officer:
(Quality Control Chemist)
Questionnaire completed by (if more than one, please indicate which
aection(s) of the questionnaire completed):
On-Site Audit Conducted by:
Affiliation of Auditor(s):
Persons Present During Entrance Interview:
Persons Present During Exit Interview:
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 31 of 57
A. GENERAL
1. Please use a simple block diagram to show the organization structure
and how the laboratory functions within it.
2. Standard Operating Procedures (SOP)
a. Has the organization written and implemented official
Standard Operating Procedures?
Yes No Comment
Implementation Date:
b. Is the SOP Manual followed in detail?
Yes No Comment
c. Does it contain all quality control steps practiced?
Yes No Comment
d. Does each analyst have a copy-at his/her disposal?
Yes No Comment
e. Has an instrument performance study been completed for each
analysis?
Yes No Comment
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 32 of 57
3. Please provide a complete list of laboratory personnel, their
educational background, analytical experience in general and
specific experience in precipitation sample analysis.
4. Laboratory Staff Training
a. Is a formal training program used? Yes No
If yes, is it:
Organization wide Yes No
In-house Yes No
On-the-job training Yes No
b. Training ouside local organization (courses attended).
Course
description Course Course Year of
or title Who attended ~ length type* attendance
•State, Federal, College, University or other
c. Publications routinely received and used by the organization.
-------
Section Mo. 10
Revision No. 1
Date October 1, 1984
Page 33 of 57
5. Laboratory Facilities
Available Comments
Item Yes No (adequacy of facility and/or space)
1. Support Gas
2. Lighting
3. Compressed Air
4. Vacuum Systems
5. Electrical Services
6. Hot and Cold Water _
7. Laboratory Sink
8. Ventilation System
9. Hood Space
10. Cabinet Space
11. Bench-top Area
12. Lab Space
13. Lab Space Utilized
for Offices
14. Office Space
15. Storage Space
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 34 of 57'
6. Laboratory Equipment
I of
Item units
Balance
analytical
Vacuum Filtration
apparatus
BBS traceable
calibrated
thermometer
Desiccator
Ion
Chroma tograph
Technicon
Atomic Absorption
Balance, top
loader
Class "S"
weights
Balance table
Distilled water
or deionized water
Conductivity meter
Glassware
pH meter
Drying oven
Hot plates
Refrigerator
Equipment
Make Model
Condition/Age
Good-Fai r-Poor
-
Ownershi p
Air Water
> of time used
in Rainwater
Programs
-------
B. SAMPLE LOGISTICS
Receiving Clerk
Section No. 10
Revision No. 1
Date October 1, 1984
Page 35 of 57
Initials
(Name)
Yes
No
1. Are all chemicals dated on receipt and
discarded when shelf life is exceeded?
2. Are all samples received by the laboratory
logged into a bound notebook?
3. Are all samples filtered before analysis?
4. Are all samples stored in the refrigerator
between analyses?
5. Are all containers washed before they are sent
to the field?
6. Is the conductivity of the last rinse water
measured for 10% of the washed containers?
7. If the conductivity of the rinse is greater
than 2 vimhos/cm, is the container rinsed further?
8. After the containers and lids are dried are the
containers capped immediately?
9. Are precautions taken not to touch the inside
of the containers and lids?
10. Are all samples stored in a refrigerator when
not being analyzed?
11. Are precautions taken not to breathe on
the sample?
12. After completion of the analyses, are the
samples stored in a refrigerator for at
least six months?
-------
C. GRAVIMETRIC MEASUREMENTS
1. Is the analytical balance calibrated daily
with weights traceable to NBS?
2. Is a Balance Calibration Log kept up
to date?
3. Is routine factory service scheduled?
Date next service is due:
Section No. 10
Revision No. 1
Date October 1, 1984
Page 36 of 57
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 37 of 57
D. £H MEASUREMENT
Analyst Initials
(Name)
Yes No
1. Does the analyst have his/her own copy of the
standard operating procedures?
2. Does the analyst have his/her own copy of
Instrunent performance data?
3. Does the analyst have his/her own copy of
safety instructions?
4. Does the analyst have his/her own copy of the
latest monthly QC plots?
5. Is the analyst aware of the most recent
control limits?
6. Does the analyst have a copy of -the most recent
list of samples in-house to be analyzed?
Date of list
7. Are all solutions properly labelled?
8. Has a pH Meter /Electrode Acceptance Test been
completed and documented for the meter and
electrode currently in use?
9. Is the pH electrode rinsed well before and after
buffer and sample measurements?
10. Before and after samples are analyzed, is the
pH meter and electrode calibration checked
with simulated precipitation reference
samples (low ionic strength)?
11. Is the pH meter recalibrated after every set
of twenty samples with simulated precipitation
reference samples (low ionic strength)?
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 38 of 57
12. After the initial calibration of the day, when the
meter is recalibrated after a series of measurements,
is the old calibration information written down
before the meter settings are changed?
13* Is the pH electrode reference solution analyzed
first and are the results compared to the
pre-established control or warming limits?
14. Are the following control samples analyzed with
each run?
Distilled Water Blanks
Old Samples
QC Spike
15. Are electrodes stored as recommended by the
manufacturer?
16. Are electrodes checked and filled, if necessary,
before each analysis?
-------
E. TOTAL ACIDITY ANALYSIS
Section No. 10
Revision No. 1
Date October 1, 1984
Page 39 of 57
Analyst -
(Name)
Initials
Yes
No
1. Does the analyst have his/her own copy of
the standard operating procedures?
2. Does the analyst have his/her own copy of
instrument performance data?
3. Does the analyst have his/her own copy of
safety instructions?
4. Does the analyst have his/her own copy of
the latest monthly QC plots?
5. Is the analyst aware of the most recent
control limits?
6. Does the analyst have a copy of the most recent
list of samples in-house to be analyzed?
7. Are all solutions properly labelled?
8. Has a pH Meter/Electrode Acceptance Test been
completed and documented for the meter and
electrode currently in use?
9. Are micropipets calibrated on at least a quarterly
basis or whenever the tip breaks?
10. Are repipets calibrated on a quarterly basis?
11. Is the stock 1.0 N NaOH standardized each
month against potassium acid phthalate and
protected from C02 absorptions?
12. Is solution temperature carefully monitored
during analysis to see that it changes by
less than 0.1°C?
-------
13* Are conditioning solution data and analyst
spike data calculated and plotted real time?
14. Are the i|» function correlation coefficients of
these data examined to ensure that they are
greater than 0.9990?
15. Are the following analyzed each day?
Three conditioning solutions and an
analyst spike initially.
An analyst spike and a conditioning
solution at the end of the analysis.
A QC spike.
16. Are electrodes stored as recommended by the
manufacturer?
17. Are electrodes checked and filled if necessary
before each analysis?
Section No. 10
Revision No. 1
Date October 1, 1984
Page 40 of 57
-------
F. AUTOMATED COLORIMETRY MEASUREMENTS
Analyst
Section No. 10
Revision No. 1
Date October 1, 198U
Page U1 of 57
Initials
(Name)
Yes
No
1. Does the analyst have his/her own copy of the
standard operating procedures?
2. Does the analyst have his/her own copy of
instrument performance data?
3. Does the analyst have his/her own copy of
safety instructions?
4. Does the analyst have his/her own copy of the
latest monthly QC plots?
5. Is the analyst aware of the most recent
control limits?
6. Does the analyst have a copy of-the most recent
list of samples in-house to be analyzed?
7. Are all solutions properly labelled?
8. Is a Standard Preparation Form completed
when new stock standards are prepared?
9. Are dilute calibration standards prepared
fresh daily?
10. Is the analyst spike prepared fresh daily
from an independent stock?
11. Is the calibration curve at least a five
point curve?
12. Is the first calibration curve of the day
checked for detection limit and linearity?
13. Are the analyst spike data calculated and
plotted real time?
-------
14. Is each new calibration curve checked to see that
instrumental response changed less than
15. Are the following control samples analyzed with
each run?
Blanks
Old Samples
Analyst Spikes
QC Spike
16. Is water pumped through all lines daily before
and after analysis?
17. Are pump tubes changed at least once per
month?
18. Is the pump cleaned when the pump tubes are
changed?
19. Is soap solution pumped through all lines
once per week?
20. Is the flowcell cleaned with a sulfuric acid-
potassium dichromate solution once per month?
21. Is the pump oiled once per three months?
Date of last service _
22. Is the colorimeter mirror assembly and color
filter cleaned and the alignment optimized
once per three months?
Date of last service
Section No. 10
Revision No. 1
Date October 1, 1984
Page 42 of 57
-------
G. ION CHROMATOGRAPHY ANALYSIS
Analyst
(Name)
Section No. 10
Revision No. 1
Date October 1, 1984
Page U3 of 57
Initials
Yes
No
1. Does the analyst have his/her own copy of the
standard operating procedures?
2. Does the analyst have his/her own copy of
instrument performance data?
3. Does' the analyst have his/her own copy of
safety instructions?
4. Does the analyst have his/her own copy of the
latest monthly QC plots?
5. Is the analyst aware of the most recent
control limits?
6. Does the analyst have a copy of-the most recent
list of samples in-house to be analyzed?
Date of list
7. Are all solutions properly labelled?
8. Is a Standard Preparation Form completed
when new stock standards are prepared?
9. Are dilute calibration standards prepared
fresh weekly?
10. If manual techniques are used, are samples
and eluent prepared fresh daily from the same
concentrated stock buffer?
11. Is the analyst spike prepared from an
independent stock?
12. Is the calibration curve at least a four point
curve for each analytical range?
-------
Section No. 10
Revision No* 1
Date October 1, 1984
Page 44 of'57
13» la the first calibration curve of the day
checked for detection limit and linearity?
14. Are the percent recoveries for the analyst
spike data calculated in real time and
compared to pre-established warming and
control limits?
15. Are the following control samples analyzed
with each run?
Blanks
Old Samples
Analyst Spikes
QC Spike
16. Is the drip tray examined daily for reagent
spills, and are spills cleaned up daily?
17. Are pumps oiled once per week?
18. Is the anion precolumn cleaned once per month
with 0.1 M Na2C03?
19. Is the Br", NO" resolution checked once
a month and documented?
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 45 of 57
H. ATOMIC ABSORPTION ANALYSIS
Analyst
(Name)
Initials
Yes
No
1. Does the analyst have his/her own copy of the
standard operating procedures?
2. Does the analyst have his/her own copy of
instrument performance data?
3. Does the analyst have his/her own copy of
safety instructions?
4. Does the analyst have his/her own copy of the
latest monthly QC plots?
5. Is the analyst aware of the most recent
control limits?
6. Does the analyst have a copy of the most recent
list of samples in-house to be analyzed?
Date of list
7. Are all solutions properly labelled?
3. Is a Standard Preparation Form completed
when new stock standards are prepared?
9. Are dilute calibration standards prepared
fresh monthly?
10. Is the analyst spike prepared from an
independent stock?
11. Is the instrument allowed to warm up at
least 15 minutes with the flame on before
the final wavelength adjustment is made?
12. Is the calibration curve at least a five
point curve?
-------
13. Is the first calibration curve .of the day
checked for detection limit and linearity?
14. Are the analyst spike data calculated and
plotted real time?
15. Is each new calibration curve checked to
see that instrumental response changed less
than 5*7
16. Are the following control samples analyzed
with each run?
Blanks
Old Samples
Analyst Spikes
QC Spike
Section No. 10
Revision No. 1
Date October 1, 1984
Page 46 of 57
-------
Section Mo. 10
Revision No. 1
Date October 1, 1984
Page 47 of 57
I. DATA MANAGEMENT
Data Clerk
(Name)
Initials
Yes
No
1. Are field data sheets filled in an organized
manner?
2. Does the data cleric do a 100% QC check for
accuracy of data input to the computer?
3. Is output from computer checked with input
data?
4. Does strip chart reduction by on-line electronic
digitization receive at least 5% manual spot
checking?
5. Are control charts or equivalent checks (e.g.,
computer calculated range limits" or regression
charts) current and available for inspection?
6. Do laboratory records include the following
information?
a. Sample identification nunber
b. Station identification
c. Sample type
d. Date sample received in laboratory
e. Time, date and volume of collection
f . Date of analysis
g. Analyst
h. Results of analysis (including raw
analytical data)
i. Recipient of the analytical data
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 48 of 57
7. Are rain gauge chart data for event times
and amount checked?
8. Does laboratory follow chain-of-custody
procedures from sample receipt to discard?
9* Doea the data clerk routinely repot quality
control data sheet information to the analyst?
10. Does the data clerk submit quality control data
sheet information to the lab manager along with
the analytical data to be reported?
11. Are computer printouts and reports routinely
spotchecked against laboratory records before
data are released?
12. Are manually interpreted strip chart data
spotchecked after initial entry?
13. Are minimum detection limits calculated by an
approved method or baseline standard deviation?
14. Are calibration curve coefficients tabulated
and regularly reviewed as evidence for instrumental
control? (An alternative is to use Regression-
Hypothesis testing in lieu of control charting.)
15. Are control charts, regression charts or computer
QC data bases up to date and accessible?
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page U9 of 57
- LABORATORY QUALITY CONTROL
Quality Control Chemist
(Name)
Initials
Yes
No
1. Does the QC chemist have his/her own copy of the
standard operating procedures?
2. Does the QC chemist have his/her own copy of
instrument performance data?
3. Does the QC chemist have his/her own copy of
safety instructions?
4. Does the QC chemist have his/her own copy of the
latest monthly QC plots?
5. Is the QC chemist aware of the most recent
control limits for each analytical method?
6. Does the QC chemist prepare and submit a blind QC
spike once per month for each analytical method?
7. Does the QC chemist routinely review and report
blind QC spike data to the laboratory manager?
8. Does the QC chemist update control limits and
obtain new control chart plots once per month?
9. Does the QC chemist review the quality control
data sheet provided by the data clerk, and then
decide whether or not to release data for
reporting?
10. Does the QC chemist prepare monthly check
samples for the field sites?
11. Does the QC chemist compare the laboratory and
field data to the monthly check sample's value
and report this to the laboratory manager?
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 50 of 57
K. LABORATORY MANAGEMENT
Laboratory Manager Initials
(Name)
Yes No
1. Does the laboratory manager have his/her
own copy of the standard operating procedures?
2. Does the laboratory manager have his/her
own copy of instrument performance data?
3. Does the laboratory manager have his/her
own copy of safety instructions?
4. Does the laboratory manager have his/her
own copy of the latest monthly QC plots?
5. Is the laboratory manager aware of the most
recent control limits?
6. Does the laboratory manager review the
following before reporting data?
a. The data itself
b. The quality control data sheet with
analyst notes
c. The quality control chemist QC reports
d. The ion summation ratios for the data
e. The calculated vs. measured sample
conductivity
7. Does the laboratory manager ensure that at least
5% of the data have been checked indepedently
by the QA officer? (p. 1 of 6, Section 3)
8. Does the laboratory manager ensure that all the
necessary corrections have been implemented
in the data base before release?
-------
Section No. 10
Revision No. 1
Date October 1,
Page 51 of 57
10.5 Performance Audits
10.5.1 Network Performance Audits - A performance audit for a precipitation
monitoring network should be made at least once per year on all sites. The
purpose of the audit is to quantitatively assess site operations which are
usually performed during the site and/or field laboratory audit visit. This
on-site visit may be combined with a site evaluation or other types of audits
such as program systems audits or laboratory systems audits.
Detailed and complete protocols for performance audits of an
organization's precipitation network should be developed by the organization
operating the network. A complete performance audit should be designed to
include, as a minimum, the following activities:
1. Check Sample Analysis — The auditor takes to the site a 150-ml check
sample of known pH and conductivity. Such samples are prepared by the
auditor's support laboratory to closely resemble that site's ambient
samples in terms of pH and conductivity. The sample can then be used as
follows:
(a) The station operator is requested to provide a precleaned bucket.
(b) The auditor adds the sample to the clean bucket.
(c) The station operator is- asked to treat this sample as though it were a
routine precipitation sample.
(d) The auditor observes, asks questions, and takes notes on sample
treatment and field measurements (usually sample volume or weight, pH
and conductivity).
(e) The auditor records the results of the pH, conductivity and
volume (wt) measurements and, based on a comparison of the support
laboratory value with that obtained by the station operator, an
assessment of the accuracy of site measurements is obtained.
2. Calibration of Weighing Bucket Rain Gauge — The auditor brings to the
site a set of calibrated weights. Such weights may be obtained directly
from the rain gauge manufacturer but should be certified at the auditor's
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 52 of 57
support lab prior to the audit. At the site the auditor proceeds as
follows:
(a) The rain gauge stripchart in use is marked, checked for the correct
time setting, removed and a new chart put in place. The new chart is
used for the audit.
(b) The gauge funnel and collection bucket are removed.
(c) The weight-handle assembly is placed on the platform normally occupied
by the collection bucket and the gauge is zeroed.
(d) Weight plates representing 1" depths (approximately 825 grams each for
the Belfort gauge) are then stacked onto this assembly and the
indicated inches of precipitation are read from the stripchart
recorder.
(e) If the gauge is properly calibrated, each weight should cause the
stripchart to indicate 1" of rain. Since added weights are known, the
found (read from the stripchart) and audit (calculated from known
weight) precipitation amounts may be compared.
(f) The gauge is calibrated to at least twice the maximum precipitation
expected for that site or to 8", whichever is greater. If the gauge
is out of calibration by more than 0.05 in. at any depth, it must be
adjusted according to the manufacturer's directions.
(g) The stripchart is removed and the original chart replaced and the pen
set to the correct time. The audit chart is annotated and kept with
the records of the audit.
3» Operation of Wet/Dry Automatic Sampler (e.g. Aerochem Metrics) — This is
performed to check proper sampler operation and to make certain that the
sampler would collect an entire precipitation event if one occurred. To
check this, the auditor should add 1-2 drops of distilled or deionized
water to the precipitation sampler sensor. The sampler is judged to be
operating normally if, within a few seconds, the lid covering the wet
bucket moves to cover the dry bucket.
After the wet bucket has been open for several minutes, the auditor
should touch the sensor plate to check that it is heating. If so, the
moisture is removed from the sensor plate by blowing. The sensor plate
then dries more quickly and the lid should cover the wet bucket again.
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 53 of 57
Lid opening and closing cycles may occur during an event. Where the
rain gauge at the site has been interfaced with the sampler and equipped
with an event pen, the auditor should check if marks, indicating a lid
cycle, are recorded on the rain gauge strip chart.
4. Review of Site Procedures and Data Documentation — The auditor should
observe site personnel performing all the routine site operation duties.
This should include handling of samples and sampling containers, checks on
instrumentation and data recording. If a check sample has been analyzed,
as described above, part of this check has already been performed.
If an analysis check sample has not been used, the auditor should ask
site personnel to describe in detail how samples, buckets and measurements
are handled. After determining (or observing) sample handling, the
auditor should then interview site personnel to gather detailed
information on standards ( pH and conductivity), sample treatment after
analysis, water supply and data recording. This interview is used to
assess operator training and performance and to establish sample and data
integrity up to the point where both leave the site.
10.5.2 Performance Audit Reporting - At the conclusion of the audit, a short
report should be prepared summarizing audit results and recommendations. The
audit report should serve to establish that a precipitation sample from the
audited site is collected, handled, measured, and shipped to a support
laboratory for further analysis in a proper manner, and that all data are
properly documented. A format for this report is suggested below together
with material to be included under each topic.
1. Introduction - summary of data, time, place of audit, site and
organization identification; and identification of people present during
the audit.
2. Audit procedures - summary of audit procedures employed together with
notation of any deviations from previous accepted procedures. Included
are references to traceability to establish credibility of all standards
used for the audit.
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 54 of 57
3* Site operations and audit results - summary of data obtained both from
observations and interviews of site personnel; included in this section
should be copies of data forms used at the site and/or other pertinent
documentation; included at the end of this section are summary tables of
the audit results themselves.
4. Discussion of audit results and recommendations - brief discussion and
interpretation of the results together with a discussion of any problem's
impact on data integrity and quality. Recommendations should also be
included to remedy such problems.
10,5.3 Laboratory Performance Audits - As a part of the independent checks
to evaluate the quality of data provided by the total measurement system,
performance audits of the central analytical laboratory provide an assessment
of .measurement accuracy. These audits are carried out by having the
laboratory analyze well-characterized, independently-prepared simulated
precipitation solutions.
Performance audits should be conducted at least quarterly by analyzing
rain-type solutions for all the .pbservables reported in precipitation
samples. It is acceptable to send the samples to the laboratory and request
that they be processed as "blind" samples. Performance audits of field pH
and conductivity should be performed during an on-site program audit.
Samples useful for performance assessment are available from at least
four organizations - Environmental Protection Agency/Environmental Monitoring
Support Laboratory (Cincinnati), United States Geological Survey,
Environmental Protection Agency/Environmental Monitoring Systems Laboratory
(Research Triangle Park), and the Canadian National Water Research Institute.
The first three organizations conduct interlaboratory surveys every six
"months; the Canadian tests are more frequent. Each analytical laboratory
supporting an acid precipitation monitoring program is encouraged to
participate in these round-robin surveys.
1. EPA Cincinnati - (Quality Assurance Branch, EMSL - Cincinnati, EPA,
Cincinnati, Ohio 45268, 513/684-7327). QC water/wastewater samples are
available without cost. All samples are prepared from ACS reagent grade
chemicals and are sent as concentrates in sealed glass ampoules. When
diluted to the indicated volume, values obtained by the laboratory should
agree with those given by the EPA. Samples are available for:
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 55 of 57
Minerals/Physical Analyses - Na*, K"1", Ca*2, Mg*2, SO'2, Cl", F", pH,
alkalinity/acidity, total hardness, total dissolved solids, and
specific conductance (available in two concentrations).
Nutrients - Nitrate-N ammonium,-N, Kjeldahl-N, orthophosphate and
total phosphorus (available in two concentrations).
2. United States Geological Survey - (Leroy Schroder, Water Resources
Division, USGS National Water Quality Laboratory, 5293 Ward Road, Arvada,
CO 80002). These samples approximate precipitation samples. They are
available in a limited supply and may be requested only by government
laboratories.
3. EPA Research Triangle Park - (Performance Evaluation Branch - MD77B
EMSL-RTP, EPA, Research Triangle Park, NC 27711, 919/541-4531). These are
commercially prepared, reference, concentrated synthetic rain samples of
well-characterized composition. They are prepared from reagent grade
salts, following a procedure developed by the National Bureau -,of
Standards. Samples are characterized by analysis by EPA and three (3)
independent laboratories. Dilution instructions are supplied with the
samples, together with appropriate reporting forms. They are available
free of charge to both private and governmental organizations.
4. Canadian National Water Research Institute - (Mr. K.I. Aspila or Ms. Susan
Todd, National Water Research Institute, P.O. Box 5050, 867 Lakeshore Rd.,
Burlington, Ontario L7R4A6, 416/637-4638 or 637-4653). Samples consist of
soft water, precipitation and synthetic rain and have all the constituents
present in precipitation. Samples available only during test periods.
For ongoing accuracy assessment during those quarters when interlaboratory
or round-robin surveys are not available, the laboratory is encouraged to
obtain and use additional samples from the suppliers given above. It is
important that all samples used for such accuracy assessment be submitted
"blind" in order to obtain a representative picture of ongoing operations. .. -
For performance audits, it is not recommended that the laboratory prepare
its own independent standards since no easy means exists to characterize them
well before use. In all cases the known integrity of the purchased standards
is higher than that which can easily be achieved by a single laboratory, and
the turnaround time on results for such samples should be sufficiently short to
prove useful.
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 56 of 57
10.6 Data Processing Audita
10.6.1 General Guidance - Data processing may involve reading a strip chart or
other instrument output tape, calculating the concentrations and transcribing
these results to a data form. In a data processing audit a certain percentage
of results is recalculated. In a system which has a large proportion of manual
data processing activity, the audit should be regularly performed by an
individual other than the one who originally reduced the data. In a
computerized data acquisition environment the difference in personnel may not
be as critical but, in general, the audit should be performed as independently
and objectively as possible.
The data processing audit should be performed on each of the groups of
data as they are reported. Thus if an agency is reporting summaries by
quarter, each quarter's data should be checked. Similar strategies should be
employed for semiannual or annual reporting.
The auditor should obtain the complete data base for that period. At
least 5} of the total number of events in the audit data base should be
selected at random for checking. Since calculation of data is critical for a
proper derivation of precision summaries for collocated sites in the network,
at least one site with collocated samples should be included in the audited
data. For each site, at least two analytes other than pH or conductivity
should be recalculated. Note that the same two analytes may be used for all
sites.
If- the recalculated results do not agree to within rounding errors with
the reported values, the entire data base should be rejected and returned to
the laboratory supervisor for further quality control measures and data
validation before being released.
10.6.2 Estimating the Percent Error in a Data Base - The following can be
applied to estimate the percent error in data bases. Each organization will
differ in the total number of values in its data base and the acceptable error
level. For large data bases (N>3000) and for the simple case where no errors
are found in n values tested, the following table gives the approximate n
-------
Section No. 10
Revision No. 1
Date October 1, 1984
Page 57 of 57
values required to be checked to yield the listed error probability £ at the
significance level x:
x I 0.05 I 0.05 I 0.05 I 0.05 I 0.10 I 0.10
p | 0.01 I 0.02 I 0.05 I 0.10 I 0.01 I 0.10
n I 298 I 148 I 58 I 28 I 229 I 22
Thus for a large data base, if 148 values are checked and no errors found, the
probability of errors in the base is less than 2% at the 5% significance level
(959 probability).
For small to moderate size data bases (N^3000), the above binomial
approximation is inadequate. Therefore, an approximation to a non-replacement
sampling model is used. For testing that the data base has less than 2% errors
(at the 5% significance level), the following table yields the number of test
values n needed, for N values in the data base:
N I 100 I 1000 I 3000
n I 78 | 140 I 146 -
If errors are found in the tests, the upper and lower confidence bounds t>
of the error fraction (f) of a data base of size n can be estimated from
Equation 10-1.
f * nb ± 1.96[n(l-b)] 1/2 10-1
where the test is applied to n=300 measurements (n - 0.1N) and f failures
(errors) are found. The larger the parent base N, the more accurate the
estimate. Thus if a random check of 706 values in a data base of over 7200
values yields 14 errors, the error bounds are approximately 3.3 to 1.29.
-------
APPENDIX A
Operation and Maintenance Procedures for
Precepitation Measurement Systems
(Originally Volume Vb of the Quality Assurance Handbook,
EPA 600/4-82-042b, Revised 1986)
-------
CONTENTS OF APPENDIX A
Section
1 INTRODUCTION
1.1 COLLECTION SITES
1.2 PARAMETERS AND ANALYTES GENERALLY
MEASURED
1.3 SAMPLING PERIODS, DEFINITION OF EVENT
1.4 REFERENCES
2 FIELD OPERATIONS
2.1 EQUIPMENT AND SUPPLIES
2.1.1 Station Supplies
2.1.2 Spare Parts
2.1.3 Precipitation Collector
Description
2.1.4 Rain Gauge Description
2.2 INSTALLATION AND ACCEPTANCE TESTS
2.2.1 Precipitation Collector
2.2.1.1 Installation
2.2.1.2 Acceptance Tests
2.2.2 Rain Gauge
2.2.2.1 Installation
2.2.2.2 Acceptance Tests
2.3 EQUIPMENT CHECKS, MAINTENANCE AND
TROUBLESHOOTING
2.3.1 Precipitation Collector
2.3.1.1 Routine Checks
2.3.1.2 Special Calibration/
Maintenance
2.3.1.3 Winter Maintenance
2.3.2 Weighing Bucket Rain Gauge
2.3.2.1 Routine Checks
2.3.2.2 Calibrat ions
2.3.2.3 Winter Maintenance
-------
CONTENTS (Continued)
Section
2 2.8 QUALITY CONTROL
2.8.1 Unknown or Quality Control Test
wSamples for the Field
2.8.2 "site Visits/Audits
2.8.3 Blind Samples for the Laboratory
2.9 FIELD PROCEDURE SUMMARY
2.10 REFERENCES
3 CENTRAL LABORATORY SUPPORT OPERATIONS
FOR THE FIELD
3.1 CLEANING AND SUPPLYING OF GLASSWARE
AND PLASTICWARE
3.1.1 Cleaning of New or Used
Plasticware
3.1.2 Cleaning of Glassware
3.1.2.1 Glassware Used for
Metal Analyses
3.1.2.2 Glassware Used for
Anions and NH%
3.1.3 Supplying Containers to the Field
3.2 PREPARATION OF STANDARDS FOR THE FIELD
3.2.1 Preparation and Measurement of
Conductivity Standards
3.2.2 Preparation and Measurement of
pH Reference Solution
3.2.3 Preparation of Quality Control
Samples
3.3 INITIAL EVALUATION OF FIELD EQUIPMENT
3.3.1 Evaluation of Conductivity
Meters and Cells
3.3.1.1 Evaluation of Accuracy
and Precision of Meter
3.3.1.2 Evaluation of Linearity
of Meter
3.3.2 Evaluation of pH Meters
3.3.3 Evaluation of pH Electrodes
3.3.4 Evaluation of Field Balance
and Thermometers
-------
CONTENTS (Continued)
Section
2 2.4 SAMPLE COLLECTION AND HANDLING
2.4.1 Avoiding Contaminat ion
2.4.2,, Sampling Schedules
2.4.3 Collection and Handling
Procedures
2.4.3.1 Wet Buckets
2.4.3.2 Plastic Bag Liners
2.4.3.3 Bottles
2.5 FIELD MEASUREMENTS
2.5.1 Weighing Sample Containers
2.5.1.1 Balance Specifications
2.5.1.2 Procedure
2.5.2 Specific Conductance Measurement
2.5.2.1 Apparatus Requirements
2.5.2.2 Procedure
2.5.2.3 Conductivity Measurement
Problems and Tests
2.5.3 pH Measurement
2.5.3.1 Apparatus and Equipment
2.5.3.2 Procedure
2.5.3.3 Electrode Problems
and Tests
2.5.4 Temperature
2.5.4.1 Requirements
2.5.4.2 Procedure
2.6 SAMPLE IDENTIFICATION, PRESERVATION,
STORAGE, AND SHIPMENT
2.6.1 Background
2.6.2 Procedure
2.6.2.1 Weekly Cumulative
Samples
2.6.2.2 Daily, Event or
Sequential Samples
2.6.3 Field Blanks
2.6.3.1 Buckets
2.6.3.2 Bottles
2.7 DOCUMENTATION
2.7.1 Logbook
2.7.2 Rain Gauge Charts
2.7.3 Field Data Forms
-------
CONTENTS (Continued)
Section
3 3.4 MONITORING OF FIELD OPERATION
3.4.1 Evaluation of Field Conductivity
and pH Measurement Systems
3.4.2 Evaluation of Field Precipita-
tion Collector, Rain Gauge, and
Balance
3.5 REPORT FORMS
3.6 REFERENCES
4 LABORATORY PROCEDURES
4.1 GRAVIMETRIC MEASUREMENTS
4.1.1 Apparatus
4.1.2 Calibrat ion
4.1.3 Procedure
4.2 PH MEASUREMENT
4 . 3 CONDUCTANCE MEASUREMENT
4.4 SAMPLE FILTRATION
4.5 ACIDITY MEASUREMENTS
4 . 6 DETERMINATION OF SULFATE
4 . 7 DETERMINATION OF NITRATE
4.8 DETERMINATION OF CHLORIDE
4.9 DETERMINATION OF ORTHOPHOSPHATE
4.10 DETERMINATION OF FLUORIDE
4.11 DETERMINATION OF AMMONIUM
4.12 DETERMINATION OF SODIUM, POTASSIUM,
MAGNESIUM AND CALCIUM
4.13 REFERENCES
-------
Section No. 1
Revision No. 1
Date July 31, 1986
Page 1 of 4
1.0 INTRODUCTION
The increasing national awareness of the harmful effects of acid
deposition on the ecology and materials has led to a significant increase
in the number of deposition monitoring networks and related effects
studies. It has become necessary to provide uniform, systematic and
approved precipitation monitoring procedures so that the acquired data
are accurate and comparable among all monitoring networks. The purpose
of this operations and maintenance (0 & M) manual is to describe in
detail the currently recommended procedures for conducting precipitation
monitoring. However, it is essential that these procedures be
supplemented by the quality assurance tasks which are presented in the
Quality Assurance Manual for Precipitation Measurement Systems (1). While
these two manuals contain the procedures recommended by the Environmental
Protection Agency, it must be emphasized that network protocols take
precedence if conflicts occur.
The basic goals of this manual are to instruct how to collect
representative samples without contamination and to preserve sample
integrity for analysis. The possibility of sample contamination,
degradation, or reaction must be minimized. The sample collector, the
first object that contacts the sample, must meet these basic goals and
must be reliable. The Department of Energy (DOE) Health and Safety
Laboratory (HASL) type automatic, wet/dry collector (2), (3), (4) has
been tested and accepted by most U.S. monitoring networks. Discussion
is limited to this type of collector. A reliable rain gauge, pH and
conductivity meter, a balance, and other accoutrements are also needed in
a monitoring station.
The material in this manual is based primarily on the procedures
used in the Electric Power Research Institute (EPRI) precipitation
network and the Utility Acid Precipitation Study Program (UAPSP) in the
Eastern United States, in the National Atmospheric Deposition Program
(NADP), and in the Multi-State Atmospheric Power Production Pollution
Study (MAP3S).
-------
Section No. 1
Revision No. 1
Date July 31, 1986
Page 2 of 4
The EPA handbooks for air pollution measurements (5,6) and for water
measurements (7) were used as guides for format and content. The
analytical procedures are based on those in the manual: Development of
Standard Methods for the Collection and Analysis of Precipitation (March,
1986) (7). To have this 0 & M manual stand alone without requiring
referrals to the other EPA handbooks, some duplication of material was
required; this material is referenced.
1.1 COLLECTION SITES
Collection sites must be located to meet the objectives of the
monitoring program—for example, baseline, regional or urban, and siting
criteria is given in Section 5.0 of the quality assurance manual (1).
The quality assurance manual also contains the general rules for the
placement of precipitation collectors, and the appropriate siting
documentation. In addition, siting characteristics may be quantified (8)
if desired. In essence, the site must yield representative samples—thus
must not have obstructions which may affect the results.
1.2 PARAMETERS AND ANALYTES GENERALLY MEASURED
The constituents and/or indicators commonly measured are listed
below. All are measured in the laboratory; items 8, 9, and 10 are also
measured in the field.
1. Sulfate (SO?) - Concentrations above the baseline values are caused
mainly by human activities, principally by the release of S0« during
the burning of fossil fuels and during refining processes; tne SO™ is
oxidized to sulfate in the atmosphere.
2. Nitrogen Compounds (NOZ, NH* and - NO - essentially NO + N02)
concentrations above the baseline values are caused primarily by tne
burning of fossil fuels, such as for transportation purposes; NH^
occurs chiefly from biochemical reactions.
-------
Section No. 1
Revision No. 1
Date July 31, 1986
Page 3 of 4
3* Chloride Ion (Cl~) - Originates chiefly from sea salt aerosols.
_o
4. Phosphate (orthotribasic PO, ) - Source is soil, rock, and
fertilizers; an important nutrient.
5* Metal Ions (Na+, K+, Ca++, Mg+*) - Na+ originates mainly from sea
salt aerosols, but all of these ions can originate from soil dust in
desert, semiarid and intensively cultivated areas.
6. Acidity - both SO- and N02 form the strong acids found in
precipitation; organic acids are frequently also present.
7. Alkalinity - Calcareous material (e.g., soil carbonate (C0~)), can
make precipitation alkaline, and can neutralize the effects or acids.
8. £fl - A quantitative measure of precipitation acidity or alkalinity.
In a theoretically clean atmosphere, a water sample in equilibrium
with atmospheric C0» would measure pH 5.6; the acidity increases as
the pH decreases from 5.6 to zero. Alkaline samples have pH 7 to 14.
9. Specific Conductance - The reciprocal of the resistance of a
solution; its magnitude depends on the concentrations and types of
dissolved salts.
10. Precipitation Amount - Value required both to calculate the weighted
mean values of theconstituents and to derive the total amount of
materials deposited over a time period.
1.3 SAMPLING PERIODS, DEFINITION OF EVENT
Precipitation sampling schedules that are commonly used include
weekly, daily, event, and subevent (sequential). An event can be defined
as a storm separated from a second storm by a dry interval, commonly at
least six hours in the winter or at least three hours in the summer. The
sampling schedule depends on the objectives of the program and on the
available funds. Aerometric and/or meteorological studies such as
transport modeling often require daily or hourly sampling. Studies of
long-term trends, and spatial and temporal variability generally use
longer sampling intervals. Sampling periods longer than one week are not
advisable because significant changes may occur to the sample while it
remains in the collector.
-------
Section No. 1
Revision No. 1
Date July 31, 1986
Page 4 of 4
1.4 REFERENCES
1. Quality Assurance Handbook for Air Pollution Measurement Systems,
Vol. V - Manual for Precipitation Measurement Systems, Part I -
Quality Assurance Manual.U.S. EnvironmentalProtectionAgency,
Research Triangle Park, NC. EPA-600/4-82-042a (January 1985).
2. Volchok, H.L., and R.T. Graveson, Proc. Second Fed. Conf. on Great
Lakes., pp. 259-264 (1976).
3. Galloway, J., Water, Air and Soil Pollution 6, p. 241 (1976).
4. Bogen, D.C., Water, Air and Soil Pollution 13, p. 453 (1980).
5. Quality Assurance Handbook for Air Pollution Measurement Systems -
Vol. I - Principles, U.S.Environmental Protection Agency,Research
Triangle Park, N.C., EPA-600/9-76-005 (December 1984).
6. Quality Assurance Handbook for Air Pollution Measurement Systems -
Vol. II - Ambient Air Specific Methods^U.S.Environmental
ProtectionAgency,Research Triangle Park,N.C., EPA-600/4-77-027a
(May 1977).
7. Development of Standard Methods for the Collection and Analysis of
Precipitation, U.S.Environmental ProtectionAgency, Environmental
>pj
>it
Monitoring and Support Laboratory, Cincinnati, OH (March 1986).
8. Eaton, tf.C. and E.L. Tew, "Site Evaluation Assistance to New and
Existing Acid Precipitation Collection Sites in the State-Operated
Network," Research Triangle Institute, RTP, NC, under EPA Contract
68-02-4125 (August 1985).
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 1 of 49
2.0 FIELD OPERATIONS
Precipitation collection field operations are covered in the
following major areas:
1. Equipment operation and maintenance;
2. Sample collection, handling, measurement, preservation, storage,
and shipment;
3. Documentation of field activities;
A. Quality control procedures.
Precipitation samples are very dilute, thus large measurement errors can
occur due to contamination or degradation. Field procedures must be
accomplished in a way that ensures measurement accuracy.
2.1 EQUIPMENT AND SUPPLIES
This section contains a list of the field equipment required for
typical precipitation collection stations, followed by a list of spare
parts for support of these stations. The section concludes with detailed
descriptions of the precipitation collectors and rain gauges most
commonly in use. The rain gauge measures the amount of precipitation,
and the precipitation collector collects the sample for chemical
analysis. The two devices are not interchangeable.
2.1.1 Station Supplies
The equipment and supplies required depend upon sampling objectives.
Equipment and supplies for a weekly precipitation sampling station are
listed in Table 2-1. If plastic bag bucket liners are used, the number
-------
Section No. 2
Revision No. 1
Date July 31, 1986
Page 2 of 49
TABLE 2-1. FIELD EQUIPMENT LIST FOR EACH STATION
Equipment/Material Min. Quantity/Site
Automatic precipitation collector 1
Collection buckets (3.5 gal) for sampler and lids 5
Fuses for sampler 2
Recording rain gauge with event marker 1
Rain gauge mount 1
pfl neter, electrode 1
Buffer, pH 4.0, and 7.0 (1 liter) 1
Conductivity meter and cell 1
Standard KCl solution, 74 yS/ca (500 mL) 1
Temperature probe 1
Pipette, syringe (20 mL capacity) 1
Tips, disposable (pkg. of 100) 1
Balance (20 kg capacity) or graduated cylinder (2 liter) 1
Set attachment weights for balance (1,2,2,5,10 kg) 1
Hailing cartons 3
¥ash bottle 1
Test tubes, plastic (17x100 mm) disposable, or vials (35 mL) 375
Test tube rack 1
Rain gauge charts (package of 100) 1
Self-adhesive labels 300
Envelopes 300
Logbook (bound with perforated pages) 1
Data forms 300
Kinvipes or other tissues (boxes) 15
Shipping tape (rolls) 3
Kallet, rubber 1
Deionized vater
Saran wrap (roll) 1
Bucket tie down 1
-------
Section No. 2
Revision No. 1
Date July 31, 1986
Page 3 of 49
TABLE 2-1. FIELD EQUIPMENT LIST FOR EACH STATION (cont.)
Equipment/Material Min. Quantity/Site
Additional Requirements for Bags:
Bucket modified for use with bags 2
Bucket lids 2
Plastic bucket liners (bags) 50
Strap with Velcro fasteners 2
Polyethylene gloves (box of 100) 1
500 nL polyethylene bottles 50
Indelible marking pen (black) 2
Scissors 1
Plastic cable ties (pkg. of 100) 1
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 4 of 49
of vet buckets required are reduced from five to two and a supply of
bucket liners, plastic gloves, plastic bottles, cable ties, two retaining
straps, and scissors vould be added to the list. Equipment required for
event or daily precipitation sampling is similar. Similar supplies as
listed in the table are also recommended for event or daily sampling. For
sequential sampling, the list in Table 2-1 should include a different
type of collector, a lover capacity (2.6 kg) more sensitive balance,
polyethylene bottles vith caps, and possibly a means for storing and
shipping the sample in a cold state (insulated containers and freeze-gel
packs), and few, if any, buckets. If meteorological and/or aerometric
•easurenents are made, the appropriate instruments must be included in
the list; hovever, these instruments are not discussed in this manual.
All sites require deionized or distilled vater. If this cannot be
produced at the site, it can be purchased locally. It is advisable to
use only vater vhich has the analysis (or conductivity) printed on the
label. The specific conductance of the water should be 3 yS/cm or less
and should be measured before the water is used.
2.1.2 Spare Parts
Precipitation collector fuses should be kept at each station along
vith spare parts and supplies. For larger networks, these items are more
conveniently supplied through the field manager or the central laboratory
vhen needed. Supplies for a network of 10 to 12 stations are listed in
Table 2-2. The polyethylene bottles are for special sampling studies
and/or for sample storage in the laboratory.
Electrodes in contact with solution have a limited life because the
vet glass membrane ages. Only electrodes that can be stored in a dry
state have a long shelf life. However, electrodes should not be emptied,
cleaned and filled with electrolyte solution by the station operators.
Vhen an electrode breaks or becomes suspect (Section 2.5.3.3), it should
-------
Section No. 2
Revision No. 1
Date July 31, 1986
Page 5 of 49
TABLE 2-2. SUPPLIES LIST FOR A NETWORK OF 10 TO 12 STATIONS
EQUIPMENT/MATERIAL NO.
Precipitation collector fuses (12) 1
Precipitation collector sensor and motor box 2
Rain gauge clock 2
Rain gauge chart clip 3
Rain gauge chart paper (package of 100) 3
Rain gauge pens and ink (set) 1
pH meter 1
pH electrode 3
Buffer, pH 3.0, 4.0, 6.0, and 7.0 (1 gal) 2
Conductivity meter 1
Conductivity cell (cell constant ~1) 2
Standard KCl solution, 74 yS/c« (a)
Syringe (20 «L) 20
Pipette, disposable tips 100
Shipping cartons and collection containers 36
Polyethylene sample bottles
16 02 (500 mL) 600
8 oz (250 mL) 600
4 oz (100 mL) 600
2 oz ( 50 mL) 200
Wash bottle 12
Temperature probe 3
Test tubes, plastic (17x100 mm) disposable, 35 mL vials 1000
Test tube racks 12
Self-adhesive labels 1000
Envelopes 1000
Logbooks 12
Data forms 300
Kimwipes or other tissues (boxes) 36
Shipping tape (rolls) 12
Plastic bucket liners (bags) 500
Polyethylene gloves (pkg. of 100) 12
Bucket modified (for use with plastic bucket liners if
netvork uses bags) 24
Collection buckets and lids (3.5 gal) 24
Saran wrap (roll) 12
Strap with fastener 24
Marking pen (black) 24
Scissors 12
Plastic cable ties (pkg. of 100) 12
Bucket tie down 24
(a) Make up as needed
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 6 of 49
be replaced vith a new tested electrode from the central laboratory.
Regardless whether the electrode is used or stored, it has a finite
useful life.
2.1.3 Precipitation Collector Description
The HASL-type precipitation collector (see Figure 2-1 and Appendix
A) has tvo containers and a common lid. The lid seals the wet sample
bucket when precipitation is not occurring, and thus minimizes
evaporation and contamination by dry deposition or dustfall. When
precipitation occurs, the lid moves off the wet bucket and covers the dry
deposition bucket. Two polyethylene buckets (1,2) are generally used to
collect wet and dry deposition, respectively, for inorganic species. For
organic constituents, glass or stainless steel containers should be used.
The conmon lid is driven by a motor that is controlled by a rain sensor.
The sensor contains a face plate with a grid closely spaced above it;
when the grid and plate are shorted by a drop of water (precipitation),
the motor is actuated to lift the lid from the collection bucket. The
sensor contains two heating circuits: one goes on when the temperature
falls below approximately 4°C to melt snow or ice on the sensor plate,
and the second goes on when the lid lifts off the sample bucket to heat
the sensor to about 55°C. Heating increases the rate of water
evaporation from the sensor, and hastens the closing of the wet bucket by
the lid after precipitation ceases to minimize the exposure time to dry
fallout. A seal between the bucket and the lid is achieved by a plastic
foam gasket under the lid and by a spring load; however, with strong
winds the lid may wobble, and some contamination may enter the wet
bucket. A bucket tie-down is useful in windy weather.
-------
Section No. 2
Revision No. 1
Date July 31, 1986
Page 7 of 49
plata activates
ncraeable lid when
bucket to another
Airmirfljpt Table
*
Motor Bex
(under table top)
Figure 2-1. Wee/Dry Precipitation Collector
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 8 of 49
2.1.4 Rain Gauge Description
To reference all the precipitation amounts against a standard, a
recording rain gauge is used to measure the quantity of precipitation.
Recording rain gauges are of two basic designs (identified by the
principle on which they operate)—the weighing-type gauge and the tipping
bucket-type gauge. Recording rain gauges should be capable of measuring
precipitation to approximately 0.25 mm (0.01 in.), and be accurate to a
few percent. For the weighing gauges, the sensitivity is a few
hundredths of an inch (less than 1 mm), and the accuracy is independent
of precipitation rate and is about 1% of full scale. For the tipping
bucket gauges, the generally accepted accuracy is 1% for precipitation
rates of 25 mm/h (1 in./h) or less, 4Z for rates of 75 nun/h (3 in./h),
and 6Z for rates up to 150 mm/h (6 in./h). The precipitation rates are
either measured directly or derived from the cumulative precipitation
data. The weighing gauges generally have 8-day clocks and charts, and a
0 to 30 cm (0 to 12 in.) dual traverse weighing range. See Appendix B
for a typical weighing rain gauge manual.
The recording rain gauge should have an event marker pen to indicate
vhen the wet-side collector bucket is open or closed. The times can be
read off the 8-day chart. The Aerochem Metrics collector will interface
vith the Belfort 5-780 series rain gauge. The event pen typically is
actuated, rising from its baseline, when the collector lid opens; and
remains actuated until the lid closes, whereupon the pen falls to its
baseline position again. To prevent the event marker pen from
interfering with the sample trace pen on the weighing gauge, the two are
offset on the time axis. Thus only one pen can be set at the correct
rime. Care must be taken to use the correct event beginning or ending
time. Since the operator is seldom present to observe the collector
behavior during an event, the event marker pen is invaluable for
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 9 of 49
indicating a collector malfunction. For precipitation collector
assembly, operation, installation, and servicing, see the manufacturer's
instructions.
2.2 INSTALLATION AND ACCEPTANCE TESTS
After a suitable site location is chosen, the precipitation
collector and rain gauge must be properly installed and system acceptance
tests performed before actual precipitation data can be collected.
2.2.1 Precipitation Collector
2.2.1.1 Installation—
The precipitation collector should be mounted on the ground so that
the rims of the buckets are level and at least 1 m above the ground. The
collector should be properly anchored against strong winds, and the
bucket should be secured to the precipitation collector by means of a
spring or elastic cord (bungee cord) hooked to the bucket handle and
collector table edge. The precipitation collector may be shielded from
the wind, but it should not be put in an area where excessive turbulence
vill be caused by the shield or where there are obstructions such as
trees and buildings (Reference 3, Section 5). For the placement of any
neighboring collectors and rain gauges of equal or smaller height, the
distance between rain gauge and collector, or between collector and
collector, should be at least equal to the height of the taller object.
Correct spacing should minimize interference as well as splash effects.
To ensure that the collector dry bucket does not act as an obstruction
for the wet bucket (or precipitation sample), the collector should be
aligned either perpendicularly to the prevailing winds or with the dry
bucket downwind of the wet bucket. The ground surface around the
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 10 of 49
collector and rain gauge should consist of natural vegetation or gravel.
It should not be paved because a hard surface may cause contamination
from dust settling and water splashing into the collector or gauge.
2.2.1.2 Acceptance Tests-
Precipitation collector acceptance tests should be carried out
before the collector is used in the field. These tests include: (1)
sensor heating and actuating the lid when the sensor is shorted with
water drops, (2) sensor cooling and return of the lid to the wet-side
bucket upon removal of the shorting material (water may be wiped dry),
(3) sensor temperature attainment (50°-60°C) when the lid is off the wet
bucket, (4) sensor temperature (1°-2°C) when ambient temperature falls
below freezing and (5) lid cycling and sealing observation. The
procedures to be used for these acceptance tests are outlined below:
a) With the collector lid in its normal position over the wet
bucket, add several drops of water to the sensor. The lid should
move off the wet bucket within seconds, and should cover the dry
bucket. After the water evaporates, the lid should return to
cover the wet bucket. If there is no response, check to see that
the sensor is connected to the motor box and that the power is
on. If neither is the problem, the sensor or motor box is
probably faulty and should be replaced. To remove the box, see
the manufacturer's instructions.
b) Affix a temperature probe (thermistor, thermometer, or
thermocouple) to the sensor plate near the screw head in the
plate. Make sure good contact occurs, and cover the probe with
an insulating material. Short the grid and plate together with a
paper clip or coin. The temperature should start to climb in a
few minutes, and should level off at 50° to 60°C. If the
temperature setting is incorrect, it can be adjusted by turning
the potentiometer screw inside the sensor box. Directions are
given in the manufacturer's instructions, reprinted in Appendix
A.
c) Remove the shorting object. The lid should close within a few
seconds and the temperature should fall to ambient.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 11 of 49
d) During steps b and c, check that the lid does not cycle. Also
check the lid seals.
e) If the lid does not seal the wet bucket, check to see if the
plastic foam gasket is secured in the correct position. To
remove the seal, see the manufacturer's instructions and Section
2.3.1, step 5. If this is not the problem, contact the
manufacturer.
f) If the lid cycles while the sensor is shorted, the cause is
probably a bad magnetic switch in the motor box or the lid arm
that actuates the switch. The arm may be loose or may have moved
too far out (more than 1 mm (1/32 in.)) from the switch as it
passed the switch during lid movement. If the latter is the
case, the lid arm can be adjusted and secured by tightening the
1/4 x 20 head screw in the bronze collar that secures the arm and
the clutch to the motor shaft.
g) Check the sensor heating circuit at freezing temperatures. The
Aerochem Metrics collector has a standard heater/ammeter test
plug which connects the sensor and the table cannon plugs. When
the heater goes on, 0.6 to 0.7 A of current flows through the
heater. The sensor can be cooled at warm temperatures by
unscrewing the sensor probe from the collector table and by
placing it in a refrigerator freezer compartment. A temperature
probe on the sensor will give its temperature. Current should
flow when the temperature falls to 0° to 2°C. The temperature
setting of this circuit cannot be altered except by changing the
resistor in the circuit.
If any of the above tests indicate a malfunction, either the problem
oust be remedied or the apparatus returned to the manufacturer. In
general, the problem can be rectified by the operator replacing the
sensor or the motor box. Do not replace any switches.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 12 of 49
2.2.2 Rain Gauge
2.2.2.1 Installation—
The rain gauge should be mounted on a firmly anchored support or
base so that its funnel rim is level and at about the same height as the
collector rim to enable comparisons of collection amounts between the
tvo. The Belfort gauge can be mounted with three bolts to a level
platform of 30.5 x 30.5 x 0.48 cm (12 x 12 x 3/16 in.) hot-rolled steel,
velded to a 5.1 cm (2 in.) diameter 1.0 m (3.5 ft.) pipe. The pipe is
sunk in cement for stability, and it should extend above ground about
0.53 m (21 in.) to bring the gauge to the same height as the
precipitation collector. Alternatively, the gauge can be mounted (bolted)
to cinder blocks. Boles can be drilled in the cinder block with a
masonry bit. The level of the gauge can be adjusted by the addition of
washers to the bolts. The gauge level can be checked with a carpenter's
level placed at two intersecting positions. The gauge mouth should be
high enough so that it will not be covered by snow.
In open, windy areas, a wind shield (e.g., swingleaf wind shields
such as the Alter used by the U.S. Weather Service) should be used with
the rain gauge.
For rain gauges which contain a clock (recorder), the access door to
the chart drive should be on the leeward side of the gauge and should be
kept closed to minimize dirt and moisture affecting the chart and the
clock mechanism. Never oil any part of the gauge except for the chart
drive, and oil this only when necessary with a light machine oil.
2.2.2.2 Acceptance Tests-
Rain gauge acceptance tests should include checks on the following:
1) sensitivity and accuracy, 2) clock function, 3) pen and recorder
function, and 4) event pen function. The procedures to be used for these
acceptance tests are outlined below.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 13 of 49
a) With the weighing rain gauge level and zeroed, add water
equivalent to several inches. For the Belfort rain gauge 5-780
series, 1 in. - 824 g.
b) If the rain gauge does not read correctly, adjust it according to
the manufacturer's instructions (Appendix B, Instruction Book for
Universal Recording Rain Gauge).
c) With the pens inked and a chart in place, turn the drum to
produce a zero-level trace; add water equivalent to 0.51 mm (0.02
in.), and measure the response. (For the Belfort recording rain
gauge 5-780 series, 0.51 mm « 16.4 g (0.02 in.). If there is no
response or if the response is more than 1.0 mm (0.04 in.),
contact the manufacturer. Check the turnover point on dual
traverse gauges. For tipping bucket gauges, add water in 0.25 mm
(0.01 in.) increments, and note when the bucket empties.
d) Wind the chart drive (or clock) until it is fully wound, and set
it for the correct time. Let the clock run for at least 24 h,
and check the pen traces and the clock time. The time should be
correct to within 0.5 h/24 h of running. If the clock does not
meet this specification, it should be replaced. If any other
problems are evident but are not addressed in the manufacturer's
instructions, call the manufacturer.
e) Note that the event and weight trace pens are offset about 4 h so
that they cannot interfere with each other. Set the weight pen
for the correct time. Make sure that the pens (weight and event)
are writing. If contact between the pen tips and the chart paper
is made but writing does not occur, draw some ink with a
toothpick down the pen tip to form a. small pool at the contact
point.
f) Connect the wires to the proper terminals on the collector and
the rain gauge. Short the collector rain sensor, and observe if
the event pen moves up about 3 mm (1/8 in.) from its baseline.
Remove the short, and note if the event pen falls back to its
baseline position. If problems with the event pen occur, notify
the manufacturer.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 14 of 49
2.3 EQUIPMENT CHECKS, MAINTENANCE AND TROUBLESHOOTING
This section contains the checks or maintenance that should be
conducted on a routine basis on the precipitation collector and rain
gauge. In addition, equipment problems that commonly occur are
discussed, and troubleshooting remedies are presented. Records of all
equipment checks and maintenance should be clearly documented in the
station logbook. If malfunctions occur, attempt to diagnose and correct
the problem as soon as possible. If the problem cannot be corrected, ask
the field manager and/or the equipment manufacturer for advice and
direction. Record the diagnosis and corrective action taken in the
logbook.
2.3.1 Precipitation Collector
The precipitation collector does not require calibration, but to
ensure proper functioning of the collector the following checks and
Maintenance should be conducted. The tasks are divided into routine
checks, special calibration/maintenance and winter maintenance.
2.3.1.1 Routine Checks—
These checks should be performed at daily or weekly intervals in
accordance with network procedures.
1. Collector Sensor Test - Short the sensor with a piece of metal or some
water to check the lid opening and the sensor heating functions. When
the sensor short is removed, the lid should close within a few
seconds, and the sensor should ce*>l. If an event pen is used, mark
its traces on the rain gauge chart for these tests. Clean the sensor
at monthly intervals or as needed.
2. Inspection of Dry Collector Bucket - If the collector has a dry bucket
(as the wet/dry collector has), check the bucket after an event or a
time period in which an event depositing more than 0.25 mm (0.01 in.)
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 15 of 49
of precipitation has occurred. Ascertain if the dry bucket contains
or did contain any precipitation. Precipitation in the dry bucket is
possible evidence of a collector malfunction. Possible causes of such
a malfunction are (1) a dirty or faulty sensor, (2) a too high sensor
heating temperature and/or a low precipitation rate, (3) a defective
magnetic mercury switch in the motor box, or (4) the lid arm too far
out from the magnetic switch to actuate it. All of the above reasons,
except for the dirty or faulty sensor, can cause lid cycling.
3. Test of Wet-Side Bucket or Plastic Liner - At weekly intervals, if no
event has occurred, test the wet-side bucket for cleanliness. Add 250
mL of deionized or distilled water, swirl the bucket so. that its
interior is washed, and measure the specific conductance of the
solution. If the conductance is over 3 pS/cm, rinse the bucket until
the rinse water conductance is less than 3 yS/cm. Conductivities
greater than 3 uS/cm indicate that the bucket is contaminated due to
poor initial cleaning, dry deposition and/or handling. If high
conductivities are frequent at a site, poor collector sealing and/or
an operator handling problem are probably present and must be
corrected.
4. Examination of the Event Pen Marker Trace - At weekly intervals,
inspect the event marker trace to see Ti the lid cycled. The event
trace openings and closings should correspond to the beginning and end
of the event as indicated by the slopes of the sample weight trace.
Numerous up and down markings in short time intervals indicate lid
cycling. Some cycling traces may occur when no event is apparent;
this can occur during short, light rain events. Cycling during a
heavy rainfall is symptomatic of a collector problem. No lid movement
traces when the sample weight trace shows that an event occurred
indicates either a collector or sensor malfunction.
2.3.1.2 Special Calibration/Maintenance—
These special maintenance and troubleshooting tasks should be
undertaken as needed. Any other maintenance advised by the equipment
manufacturer should be carried out at the recommended time periods.
1. Minimizing Lid Lifting by Strong Winds - Where strong winds are
common, check the lid to be sureitdoes not wobble or is not lifted
off the bucket by the wind. If either is a common occurrence, replace
the springs on the collector with stronger ones, and readjust the lid
arm and the counterweight balance.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 16 of 49
2. Lid Cycling - As a common occurrence, lid cycling can be due to
several causes. First, cycling during low rainfall can take place if
the hot sensor plate dries the sensor rapidly. If this is a frequent
occurrence, lower the temperature by turning the potentiometer screw
(see manufacturer's instructions). Second, the lid arm can be loose
or too far out from the magnetic switch in the motor box. Third, the
svitch may be defective. (For the last two, see Section 2.1.2.2, step
6).
3. Lid Malfunctioning - Another common source of collector problems is a
faulty sensor. The lid may remain open, not open or open
intermittently. The lid staying open indicates a shorted rain sensor.
A short can be verified by unscrewing the sensor cannon connector at
the BO tor box. The lid should then close over the wet bucket; if the
lid does close, check if dirt is shorting the sensor plate and grid.
If so, clean with a toothbrush or by passing a card between the grid
and plate. For the other problems, the simplest remedy is to replace
the sensor.
4. Replacement of Collector Lid Seal - Replace the plastic foam underseal
on the lid annually or as soon as needed. It will deteriorate with
ti»e, especially in hot, dry climates. The collector lid seal is
removed using the following procedure:
a) With the collector power disconnected, place the collector lid in
the middle position.
b) Remove the two (2) screws on the edge of the lid.
c) Remove the two (2) L-brackets into which the screws were threaded.
d) Remove the lid pad by prying it gently along its edge with a coin
or a screwdriver.
5. Cleaning Techniques and Schedule - Wash the collector rain sensor
monthly with deionized water toremove dirt, salt, and film buildup.
If a film persists, clean the sensor grid and plate with detergent and
a toothbrush. Wipe the rim of the dry bucket weekly with clean tissues
(e.g., Kimwipes) to prevent carryover of dustfall to the sealing
gasket and then to the vet bucket.
6. Site Maintenance and Inspection for Obstacles - Periodically mow the
grass and inspect the site area for new obstacles (e.g., a growing
bush or tree) that may become an obstacle even though not one
initially.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 17 of 49
2.3.1.3 Winter Maintenance-
Check the sensor temperature if the ambient temperature falls below
freezing to ensure that the heater is working. This may be done by
adding snow to the sensor and observing if the snow melts (the lid will
open). If necessary, the following may be conducted to prevent the
freezing of equipment:
1. Prevention of Lid Freezing—To prevent the lid from freezing to the
bucket, the following is recommended (4):
a) Attach a peaked roof (available from Aerochem Metrics) to the lid
to prevent buildup of snow on the lid and to help insulate the lid.
b) Cut a small notch in one corner of the roof to insert a power cord.
c) Attach the power cord inside the roof to an air thermostat
(Honeywell or WRAP-ON) set for about 2°C (36°F); tape the cord to
the roof arms.
d) Use a 60-tf or 75-W light bulb as a heater; set the bulb on a piece
of 9 mm (3/8 in.) Styrofoam on the lid top to prevent a hot spot.
e) Install a piece of 18 mm (3/4 in.) Styrofoam under the slope of the
roof to minimize heat loss.
f) To compensate for the additional weight on the lid, add two large
U-bolts to the counterweight shaft (approximately 200 gms).
2. Prevention of Lid Arms Freezing to Table—To prevent freezing of the
lid arms to the table, insulate one from the other.
a) Wrap and tape a plastic sheet around-each lid arm to make a boot.
b) Tape one end of the boot to the table and the other end to the arm.
c) Check to see that the boot is secure and does not tear loose when
the lid arms move between the closed and the open bucket positions.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 18 of 49
2.3.2 Weighing Bucket Rain Gauge
The weighing bucket rain gauge must be calibrated upon installation
and at least at annual intervals thereafter. To ensure proper
functioning of the gauge, the following routine checks, calibrations and
Maintenance should be conducted. Any other maintenance recommended by
the manufacturer should be carried out.
2.3.2.1 Routine Checks—
These checks should be performed at daily, weekly or monthly
intervals as appropriate.
1. Adjusting the Zero Setting - At daily or weekly intervals with no
precipitation in the rain gauge, adjust the zero setting if necessary
with the (red) fine adjust screw. The zero setting will fluctuate
with temperature, but generally not more than +0.75 mm (0.03 in.).
2. Checking the Pail Level - When the rain gauge pail is removed, be sure
it is replaced correctly so that it is level.
3. Adjusting and Winding the Clock - Weekly, for an eight day clock, wind
the clock (or chart drum) on the weighing gauge, and correct the time
setting if necessary. Be sure to correct for backlash and to set the
time correctly with respect to a.m. and p.m. Note the event pen
times are offset from the weighing pen by about 4 h; set the weighing
pen to the correct time.
4. Inspection of Pens and Ink - Weekly, inspect the pens to see if they
have ink and are writing. If they are not writing, clean the pens,
refill them, and be sure they are working. To help start the pens
writing, use a flat toothpick to make the ink from the pen reservoir
form a small pool at the point of contact between the pen and the
chart.
5. Chart Replacement - At the prescribed interval, generally weekly,
remove the old chart and replace i": with a new one. Close the access
door to the chart.
6. Level Check - At bimonthly intervals, measure the gauge level to
ensure that it is still horizontal.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 19 of 49
2.3.2.2 Calibrations-
Two types of calibrations are recommended. A single point check to
be performed monthly and a multi-point calibration to be conducted at
least annually.
1. Rain Gauge Check - Once a month, add several known weights of tap
water to the rain gauge to see that it is measuring correctly. For
the Belfort weighing gauge, 25.4 mm » 1 in. * 824 g.
2. Rain Gauge Calibration - At 12-month intervals (unless test. 1 shows it
is necessary sooner), calibrate and adjust the weighing bucket rain
gauge at each 25 mm (1 in.) level according to the manufacturer's
instructions. A set of weights and a linearity tool can be obtained
from the manufacturer for the calibration. Alternatively, weighed
quantities of tap water can be used. For the Belfort gauge, 25.4 mm »
1 in. =» 824 g. With a dual traverse pen recorder such as the Belfort
(0 to 6 in. and 6 to 12 in. traverses), the range 127 to 178 mm (5 to
7 in.) has been found difficult to calibrate and to keep calibrated.
To minimize use of this range interval, the rain gauge bucket can be
emptied after each event or week of events. In the winter, antifreeze
must be added to the weighing gauge bucket to help melt the captured
snow. Thus a severe or prolonged storm can bring the gauge to the 127
to 178 mm (5 to 7 in.) level. If it is found that a calibration
problem exists in the 127 to 178 mm (5 to 7 in.) range, it is
recommended that the bucket be emptied whenever the 127 mm (5 in.)
range is approached and that new antifreeze be added. If a tipping
bucket gauge is used, it can be calibrated by adding a measured volume
of water, using a slow drip technique, as specified in the
manufacturer's instructions.
2.3.2.3 Winter Maintenance—
In the winter, rain gauge problems can be caused by (1) snow filling
or drifting out of the gauge, (2) freezing of the collected precipitation
vhich can damage the gauge bucket, and (3) the cold affecting the clock
and/or ink.
Therefore, the following actions should be taken:
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 20 of 49
a) Remove the funnel in the inlet mouth.
b) Add approximately 1600 g (2 in.) of an ethylene glycol-methyl
alcohol (40:60) antifreeze mixture (Belfort manual, Section 4.3).
To retard evaporation, add 180 mL (6 oz) of a 10V motor oil. Do
not adjust the gauge reading after adding the antifreeze. The
gauge vill indicate rainfall of approximately 50 mm (2 in.). The
ethylene glycol-methanol with precipitation added to yield 150 mm
(6 in.) of solution will freeze below -40°C (-40°F). For less
severe conditions, use approximately 50 mm (2 in.) of ethylene
glycol antifreeze alone. When enough precipitation has been
collected to yield 127 mm (5 in.), the mixture will be liquid at
-24°C (-12°F). Since the weighing gauge is most difficult to keep
in calibration in the 127 to 178 mm (5 to 7 in.) range, empty the
gauge when the 127 mm (5 in.) level is reached, and add new
antifreeze. The antifreeze will not only aid in melting the
snow, but will prevent freezing of collected precipitation and
resulting damage to the container.
c) In extremely cold periods, the clock (if not new) may run slowly,
and/or the ink may not flow. Low-temperature ink is available
from the rain gauge manufacturer. Use of a low-temperature
lubricant may be helpful if the clock runs too slowly.
2.4 SAMPLE COLLECTION AND HANDLING
2.4.1 Avoiding Contamination
Careful handling of equipment and samples to prevent contamination
is extremely important. The dissolved substances have very low
concentrations, so any contamination will result in large errors. Thus
all articles that contact the samples must be clean. All the Buckets and
containers have been cleaned at the central laboratory. Only the
materials (e.g., sample buckets, electrodes, cells, and probes) that are
used and not returned to the laboratory must be rinsed. All rinses
should be done with deionized or distilled water.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 21 of 49
2.4.2 Sampling Schedules
Sampling schedules generally used include weekly, daily, event, and
subevent. Daily and weekly samples should be removed at the same time of
day for each sampling period unless precipitation is occurring at that
time; then a delay of up to 12 h for daily sampling and 24 h for weekly
is permissible. For event schedules, remove the sample immediately after
the event or at a set time daily.
For sequential schedules, treat the samples similarly to the event
samples. Since one event may produce several samples, it is important to
identify each sample in the chronological order of occurrence. For
sequential sampling it is important that the time corresponding to each
subevent specimen be known so that correlations with other data can be
nade.
The samples are identified and measured for amount, pH, and
conductivity. They are then sealed in plastic bottles, if event or daily
samples* and stored in a refrigerator until shipment (Sections 2.5 and
2.6).
2.4.3 Collection and Handling Procedures
Precipitation samples are collected in wet buckets, plastic bag
bucket liners, or plastic bottles. The methodology for each is given
below.
The containers for the wet samples should have been cleaned prior to
shipment to the field and do not require rinsing in the field before use.
Never substitute a precipitation collector dry bucket for a wet bucket.
At all times, take care not to contact the inside wall of a container, a
lid or a cap with any object—especially one's finger which can leave a
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 22 of 49
deposit of salt and oil. The container should be capped until
inmediately before use, and must be resealed immediately after use.
Since human breath contains ammonia, do not exhale into a container.
2.4.3.1 Wet Buckets—
I mediately before use, label the new precipitation collector vet
bucket (or for sequential sampling, the capture bottles). Veigh the
bucket after the label is affixed. The label should contain the station
identification, the date placed in use, and the bucket weight (without
lid). After the sampling period, the final weight is added. Use a
pencil or ball point pen to inscribe the labels. Below are standard
operational procedures to be adopted when handling precipitation
collection buckets.
a) Do not remove a clean bucket from the plastic bag in which it is
shipped until it is to be placed in the collector.
b) Check the collector bucket for precipitation at the scheduled
times. Move the lid from the wet bucket by contacting a coin or
metal object to the sensor grid-plate to activate the motor and
move the lid to the dry sample bucket. The lid will remain open
until the metal object bridging the sensor is removed. NEVER TRY
TO FORCE THE LID OPEN BY HAND.
c) Remove the bucket from the collector at the scheduled time, and
replace it with a clean, weighed, labeled bucket.
d) Remove the lid from the new bucket after it has been placed in
the precipitation collector, and cover the removed sample bucket
with the new lid to minimize the chance of contamination. Fasten
lid on old bucket with masking tape.
e) If no sample is present, seal the empty bucket and return it to
the laboratory, or, depending on the protocol, rinse it at the
field station for reuse (see Section 2.6.3).
f) Remove and replace the rain gauge chart. Record readings (times
of start and end of precipitation) on data form. For the final
amount of precipitation reading, use the maximum value on the
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 23 of 49
rain chart at end of event because loss of water by evaporation
will occur on standing.
g) If there is no antifreeze in the rain gauge, empty its bucket. If
there is antifreeze, do not empty the bucket until the reading is
127 mm (5.0 + 0.3 in.); then empty the bucket and add new
antifreeze.
h) Veigh bucket and sample (see Section 2.5.1).
2.4.3.2 Plastic Bag Liners—
When plastic bags are used for bucket liners, the buckets are
modified by drilling a vent hole near the bucket rim to allow air to
escape (5). A new bag is inserted in the bucket each week and removed at
the end of the weekly collection period. Avoid touching the inside of
the bucket or bag below its top except when wearing polyethylene gloves.
a) To insert the plastic bag (5), use a clean pair of plastic
disposable gloves. Open the bag to fill it with air.
b) About 4 to 5 in. down from the opening, squeeze the bag closed to
capture the air in the bag.
c) Push the inflated bag into the bucket until it touches the
bottom.
d) Fold the upper 4 or 5 in. of the bag over the bucket rim and
adjust to minimize creases on the rim.
e) Secure the bag flap to the outside of the bucket just above the
first ridge (and above the handle) with a retaining strap (see
Figure 2-2).
f) The bag can be opened more fully inside the bucket by smoothing
the bag against the inside wall. Always wear clean gloves while
doing this.
g) Weigh the bucket + bag assembly and record the weight.
h) Cover bucket with its lid until it is to be placed in the
precipitation collector.
-------
Section No. 2
Revision No. 1
Date July 31, 1986
Page 24 of 49
Ridgtin
Molded Budwt
Vent Hole
Figure 2-2. Plastic Bag Liner Assembly
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 25 of 49
To remove the sample (e), carry the covered bucket containing the
liner vith or without a sample to the site laboratory.
a) After the bucket containing the liner, retaining strap, and
sample is weighed, remove the retaining strap without passing it
over the open bucket (to avoid any contamination falling in).
b) Push up some of the liner flap outside the bucket to enable the'
top of the liner to be grasped.
c) Lift the bag up out of the bucket several inches.
d) Holding the bag with one hand, squeeze the bag shut about one in.
below the bucket riot and close the bag with a cable tie. If
sample is frozen, allow it to melt completely before closing the
bag with the cable tie.
e) Remove the bag from the bucket, swirl to mix contents, and wash
one of its bottom corners with deionized water.
f) Elevate the cleaned, dried corner so that it is not in contact
with solution and cut off about 1/2 in. using cleaned scissors.
g) Lower the cut corner, carefully pouring an aliquot of the sample
into a clean, 500-mL, wide mouth plastic bottle. Label the
bottle with an indelible marking pen. Discard any collected
sample remaining in the bag.
2.4.3.3 Bottles—
For event or daily sampling, the number of buckets required, as well
as storage and shipment space, are minimized by transferring the sample
from the bucket after it is weighed (Section 2.5.1.2) to a 500-mL
labeled, wide-mouth polyethylene bottle. If frozen, the sample must be
completely melted and mixed before transferring. If sufficient sample is
present (e.g., more than 300 mL), use 50 mL to rinse the shipping bottle.
One 500-mL bottle per event is a sufficient sample for all measurements;
the rest of the sample may be discarded.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 26 of 49
For sequential samples, which are collected through a funnel
directly into prenunbered, prelabeled polyethylene bottles, seal the
bottles immediately after the samples are collected.
Wash the sample bucket or, for sequential precipitation collectors,
the funnel and tubing with deionized water until the rinse water has a
specific conductance below 2 uS/cm (step 3, Section 2.3.1).
The samples are now ready for field measurement; check that the
containers are correctly labeled.
2.5 FIELD MEASUREMENTS
The field measurement procedures for weighing, conductivity, pH, and
reaper a ture should be identical to those used by the central laboratory.
Each bucket is weighed both before and after sampling. If sufficient
sample (more than 70 g) is available, its pH and conductivity are
aeasured both in the field laboratory and at the central laboratory.
These measurements are used as a check to detect sample changes. If less
than 70 g of sample are collected, the sample is sent to the central
laboratory without measuring conductivity and pH.
2.5.1 Weighing Sample Containers
2.5.1.1 Balance Specifications—
The balance should have a capacity of 20 kg and a precision of at
Isast +10 g. The mass of precipitation collected by the precipitation
collector is measured to determine the rain collector efficiency compared
:o the rain gauge.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 27 of 49
2.5.1.2 Procedure—
a) Before each weighing, brush off the balance pan with a soft
brush.
b) With the balance level, adjust to zero (see manufacturer's
instructions).
c) Before sampling, place a new bucket, or a bucket containing a new
plastic bag liner, without its lid (and/or bottle with its lid)
on the balance, and weigh to the nearest gram. Do not allow the
bucket lid's inner surface to contact any object. Record the
weight on the data form and on the container label (Section
2.7.3).
d) Before weighing the bucket containing a sample, tap the covered
bucket to knock any water drops off the inside lid surface into
the bucket, wipe off the outside of the bucket, remove the tape
and the lid from the bucket, and place lid with its outer surface
on the table.
e) After the balance has been zeroed, place the bucket without its
lid on the balance pan, cover the open bucket with Saran Wrap,
and weigh to the nearest gram.
f) Record the weight on the bucket label and on the field data form.
g) Subtract the initial weight of the empty container from the final
weight of container plus sample to obtain the sample weight.
Record on field data form.
h) Avoid breathing onto the sample to prevent ammonia contamination.
i) If sample is more than 70 g, remove an aliquot of about 20 g for
conductance and pfl measurements. For this, a disposable syringe
can be used. If sample is frozen, allow it to melt completely in
its closed container, and swirl the container to assure
homogeneity before removing the aliquot. Reweigh sample plus
container to obtain aliquot weight by difference.
j) Seal container with lid; obtain and record total weight to be
shipped to the central laboratory. If sample is shipped in its
bucket, secure the lid with a rubber mallet.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 28 of 49
2.5.2 Specific Conductance Measurement
2.5.2.1 Apparatus Requirements—
The conductivity meter should permit selection of several different
Measurement ranges between 0 to 10 and 0 to 1000 pS/cm, and have a
precision of +0.5Z of range and an accuracy of at least +1.0% full scale.
The range most frequently used is 0 to 100 pS/cm. A temperature-
compensated cell with a cell constant of 1.0/cm is preferred.
For calibration, use a KCl solution of known specific conductance
and be sure the temperature of the KCl standard and the sample are the
same. For rain samples, a 0.0005M KCl solution which has a specific
conductance of 74 uS/cm at 25°C is appropriate.
The specific conductance of the sample can be measured on the same
aliquot as used for pH. If this is to be done, measure the specific
conductance before measuring the pH to avoid any possible error due to
electrolyte contamination from the pH electrode.
2.5.2.2 Procedure—
Measure the specific conductance for all samples over 70 g, using
the procedure in Method 120.1 (Specific conductance) (6).
1. Summary of Method
a) Measure the specific conductance of a sample by using a
self-contained conductivity meter, Vheatstone bridge-type or
equivalent.
b) Analyze samples preferably at 25°C. If not and if the meter does
not have automatic temperature compensation, measure at 20° to
28°C, and correct to 25°C.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 29 of 49
2. Sample Handling and Preservation
a) Perform analyses in the field laboratory and/or the central
laboratory.
b) If analysis is not completed within 24 h of sample collection,
store sample at 4°C for preservation. Vash the apparatus vith high
quality distilled/deiqnized water, and prerinse with sample before
use.
c) Remove sample aliquot for measurement, and seal the bulk sample.
Allow sample aliquot to come to ambient temperature before
proceeding with conductance measurement.
3. Specific Conductance Meter Standardization
a) Follow the manufacturer's instructions for the operation of the
instrument.
b) Allow sample aliquot to come to room temperature (23°-27°C), if
possible.
c) Use 74 pS/cm standard. For dip tube cell, rinse and shake test
tube or vial three times with deionized or distilled water.
d) Add 1.3 cm (0.5 in.) of 74 yS/cm solution to test tube; swirl to
coat walls; drain. Add 20 mL of solution or enough to cover
electrodes; insert rinsed conductivity cell. Remove and shake;
repeat two times.
e) Insert cell; set meter to conductivity, and control knob to read
74. Readjust after 1 min. Move cell up and down to remove
bubbles, readjust to 74 if required.
f) Discard solution; shake cell and tube dry. Put a second aliquot of
74 pS/cm solution in same tube; check reading. Readjust meter if
necessary. Discard solution.
g) For closed bottom type cell, use above instructions omitting the
test tube, and add sufficient water or solution to cover
electrodes.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 30 of 49
h) Determine the temperature of the sample to +0.5°C. If the
temperature of the sample is not 25°C, make the" correction (as
shown belov) to convert reading to 25°C. After measurement, either
discard the solution or save it for pH measurement; never pour the
solution back into its container. Record the data in logbook.
The following temperature corrections are based on the standard KCl
solution, and are used with instruments with no automatic temperature
compensation. -
(1) If the temperature of the sample is below 25°C, add 2% of the
reading per degree.
(2) If the temperature is above 25°C, subtract 2% of the reading per
degree.
Report results as conductivity (yS/cm) at 25°C on the data form.
4. Conductance Test
a) For dip tube cell, rinse and shake a new test tube or vial five
times with deionized or distilled water. Rinse and shake
conductivity cell three times.
b) Pour deionized or distilled water into test tube or vial. Dip and
shake cell three times before reading. Let solution stand until
quiescent. If the conductance exceeds that of the deionized water,
repeat rinses until it is equal to that of the water. Record latter
of two readings on the field data form for conductivity of
distilled water.
c) Drain and shake tube; shake cell dry.
d) Rinse test tube or vial with sample. Pour sample into tube to
cover electrodes. Reseal bulk sample container. Dip and shake
cell three times, then, after solution has come to rest, measure
conductance and record.
e) For closed bottom type cell, use similar procedure, and add
sufficient deionized water or sample to cover electrodes.
f) Save this sample for pH test.
g) Rinse cell with deionized water; drain, shake, blot and store.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 31 of 49
2.5.2.3 Conductivity Measurement Problems and Tests—
The conductivity cell generally has few problems; store the cells as
recommended by the manufacturer. The working conductivity standard is
0.0005H KC1, which will slowly degrade and is easily contaminated. To
•inimize errors due to changes in the calibration standard, replace the
74 uS/ca working solution at approximately quarterly intervals.
tfhen a nev working standard is received, correlate it against the
old working standard. Report the measured value of the old working
standard to the central laboratory, and always return enough of the old
standard to the central laboratory so that it can be remeasured. Never
return the old working standard before checking it against the new
solution.
Store the conductance standards in a refrigerator to minimize
changes but always bring them to room temperature before use. Changes of
less than 32 «ay be ignored. If the change is more than 3%, order a new
standard from the central laboratory.
If the conductance meter has an internal standardization circuit,
use it to check the KCl standard by following the manufacturer's
instructions. If the KC1 standard has changed from its original value by
•ore than 52, inform the central laboratory immediately. Since the
internal meter calibration is not a traceable standard, it must not be
substituted for the KCl solution.
Another means of evaluating the working conductance standard is to
compare it against the Q.A. samples received periodically from the
central laboratory. Return the test samples to the central laboratory
vith the next sample shipment for remeasurement. If the laboratory finds
that the field conductance differs from the laboratory value by more than
102, the central laboratory will replace the old conductance standard.
Store the cells as recommended by the manufacturer.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 32 of 49
2.5.3 pH Measurement
2.5.3.1 Apparatus and Equipment (7)—
LABORATORY pH METER — The meter may have either an analog or
digital display with a readability of 0.01 pH unit. A meter that has
separate calibration and slope adjustment features (8) and is
electrically shielded to avoid interferences from stray currents or
static charge is necessary. It may be powered by battery or 110 V AC
line; if battery povered, the meter must have a battery check feature. A
temperature compensator control to allow accurate measurements at
temperatures other than 258C is desirable.
SENSING ELECTRODE — Select a sensing electrode constructed of
general-purpose glass. This electrode generates lower resistance, faster
response, and has a reliable range of 0-14 pH units. Refer to the manual
accompanying the probe for the manufacturer's recommendations on
electrode storage.
REFERENCE ELECTRODE — The reference electrode recommended for wet
deposition analysis is one equipped with a ceramic junction. The ceramic
construction minimizes differences in potential between high ionic
strength buffers and low ionic strength samples thus reducing errors from
residual junction potentials. A reference probe equipped with a ceramic
junction in an annular ring configuration generates a more stable
potential in less time due to a higher flow of internal electrolyte into
the solution. Single pore ceramic frit junctions also provide adequate
electrolyte flow.
COMBINATION ELECTRODE — The combination electrode combines the
indicating and reference elements in a single unit. Since sample volume
requirements are a consideration when analyzing wet deposition samples,
combination electrodes are more convenient than separate glass and
reference electrodes.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 33 of 49
-4
Before use, the electrode should be equilibrated in 1 x 10 mol/L
hydrochloric acid (7) and stored in the same solution. Refer also to
Appendix C.
THERMOMETER — The thermometer should be readable to 0.5°C in the
ambient range.
Stations may receive the required calibration buffer and storage
solutions from the central laboratory, according to network protocol.
The stations should notify the laboratory when the buffers need to be
replaced.
2.5.3.2 Procedure—
The pH is measured for all samples weighing over 70 g. If the
measurement is made on the same aliquot as that used for conductivity,
the pH must be measured after the conductivity (6). An alternate
procedure using dilute buffers is presented in Appendix C.
1. Scope and Application—This method is applicable to precipitation
samples'.
2. Summary of Method—The pfl of a sample is determined electrometrically
by using a glass electrode with a reference electrode.
3. Sample Handling and Preservation—
a) Perform the analyses on site immediately after sample collection.
b) After removal of a sample aliquot, seal the bulk sample container;
if the container is a bucket use a rubber mallet to secure the lid.
4. Reagents—Standard buffer solutions may be available from the central
laboratory, according to network protocol. Store buffer solutions at
room temperature.
5. pH Measurement—Always, determine the conductance first. Rinse water
should be the best grade of deionized or distilled water available. A
combination electrode is recommended. Rinse the pH electrode prior to
each measurement. Report the pH to the nearest 0.01 unit and the
temperature to the nearest 1.0°C on the data form.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 34 of 49
CALIBRATION FUNCTION:
a) Adjust the temperature control on the meter to room temperature.
Rinse the electrode(s) with three changes of water or with a
flowing stream from a wash bottle. Dispense two aliquots of the
buffer with the higher pH into separate, clean sample cups. Insert
the electrode(s) into one aliquot for 30 seconds.
b) Remove the electrode(s) from the first aliquot and insert directly
into the second. Allow either two minutes for equilibration or
allow sufficient time for the reading to remain steady within ±0.01
pH unit for 30 seconds.
c) Adjust the calibration control until the reading corresponds to the
temperature corrected value of the reference buffer solution.
SLOPE FUNCTION:
a) Rinse the electrode(s) with three changes of water or with a
flowing stream from a wash bottle. Dispense two aliquots of the
second reference buffer solution into separate, clean sample cups.
Insert the electrode(s) into one aliquot for 30 seconds.
b) Remove the electrode(s) from the first aliquot and insert directly
into the second. Allow the system to equilibrate.
c) Adjust the slope function until the reading corresponds to the
temperature corrected value of the reference buffer solution.
CALIBRATION CHECK:
a) Remove the electrode(s), rinse thoroughly, and place into the first
reference buffer solution. If the pH does not read within ±0.01
unit of the temperature corrected value, repeat the calibration
procedure until the buffers agree.
SAMPLE MEASUREMENT:
a) Again, remove the electrodes from the buffer and rinse them
thoroughly, using multiple rinsings (wash bottles are recommended).
Use 30 mL of water and be sure to rinse the tip. Gently blot the
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 35 of 49
electrode tip dry. Determine the pH of the pH CHECK SAMPLE. Allow
the electrode to equilibrate without stirring for 2-4 minutes, or
allow sufficient time for the reading to remain steady within +0.01
pH unit for 30 seconds, and read and record the pH to the nearest
0.01 pH unit.
b) Repeat above step only using the same 20 mL aliquot used to measure
conductivity. Record the pH of the PRECIPITATION SAMPLE to the
nearest 0.01 pH unit.
c) Discard the 20 «L aliquot used for conductivity and pH measurements
(do not return the aliquot to the bucket) and rinse the electrodes
one last time. Store the pH electrode in the KC1-HC1 solution (see
Section 2.5.3.1). Change the storage solution weekly.
2.5.3.3 Electrode Problems and Tests—
Two diagnostic tests which indicate the aging of the electrode are
presented here.
The first test uses periodic test samples sent out to the field
stations by the central laboratory. The samples should have pH and
specific conductance values typical of precipitation. They should be
unknown to the site operator, and are measured for both variables. These
are measures of the station's accuracy if the laboratory value is assumed
to be correct and if no solution change occurs in shipment.
The test solution is poured into five test tubes and the
conductivity and then pH of each tube are measured as if they were five
different samples. That is, the conductivity cell and pH electrode are
rinsed before and after each tube is measured. The readings are
:abulated and the average value and standard deviation calculated. From
:hese results and from the time required to attain stable readings,
electrode performance and precision can be evaluated. The standard
deviation(s) is calculated from the relation:
s =
5
E (x. - x)2/n
1/2
2-2
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 36 of 49
where x. and x are the measured and the average pH readings of the series
and where n is the number of sample aliquots measured.
If the average of five field pH measurements differs from the
laboratory by more than ±0.15 unit, or the standard deviation is greater
than O.05 pH unit, the pH electrode may need replacing. The samples are
returned to the laboratory vith the results for recheck and evaluation.
Consultation with the station operator on the technique may identify the
source of the problem.
The second test uses a reference solution, which has a known pH and
a conductivity similar to those of rain samples, to check the pH
electrode at the station at weekly intervals. The measurement procedure
is identical to that used for the rain sample. Store the solution in a
refrigerator, and replace it when needed or when the solution pH or
conductivity appears to have changed. For the reference solution, the pH
value should agree vith the value assigned by the central laboratory to
better than +0.10 unit.
If the first measurement differs by more than +0.1 pH unit from the
others for the same solution, this value should be excluded. Thus, for
the test sample, a sixth sample tube should be measured. If this poor
first reading behavior is exhibited by an electrode, and if sufficient
precipitation sample exists, two tubes of each precipitation sample
should be measured for pH, and the second value entered on the data form.
The ti«e to attain a stable reading, i.e., when pH is constant to +0.01
unit for 30 sec., should be no more than 2 minutes for a properly working
electrode. Results of these tests serve as guides for both the
measurement technique and the equilibration time to be used for
precipitation sample measurements. If an electrode consistently shows
behavior poorer than that given by the above time and pH difference
criteria, the electrode should be replaced. If the reference solution pH
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 37 of 49
•easureaent has changed from the previous month's value by more than 0.10
unit, check the solution's conductivity. If the conductivity has changed
by more than 10% from its original value, the solution and not the
electrode has probably degraded and should be replaced. Always return
enough of the solution so that it can be checked by the central
laboratory.
2.5.4 Temperature
2.5.4.1 Requirements —
The temperature probe must display at least 0.5°C increments. A
thermistor, thermocouple, or thermometer can be used. The probe should be
calibrated by the central laboratory.
2.5.4.2 Procedure—
a) Before measuring a solution, rinse the temperature probe with
deionized water, and shake it dry.
b) To minimize contamination, do not insert the probe into any
solution until after the other measurements, i.e., conductivity
and pH, have been made.
c) Read and record the temperature to the nearest 0.5°C.
2.6 SAMPLE IDENTIFICATION, PRESERVATION, STORAGE, AND SHIPMENT
2.6.1 Background
Samples must be adequately identified so that they can be readily
and correctly matched up with their data forms. The sample label should
contain station identification, sampling date, and sample weight. Use a
pencil or a ball-point pen to mark the label so that it is still legible
if it gets wet.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 38 of 49
Sample degradation can be caused by chemical interactions—for
example, with particles or gases, or biochemical reactions. Preservation
of sample integrity after removal from the collector can be improved by
filtration, sealing the sample, and storage in the dark. Freezing is not
recommended. Refrigeration is typically used for daily or event samples
but not for weekly samples. To minimize contamination, sample filtration
is performed in the central laboratory.
In the case of duplicate (collocated) or sequential collectors,
treat each sample container as a separate sample. For duplicate
collectors, distinguish the samples by adding a -1 and -2 beside the
station identification space on the data form. For sequential samples,
add -11, -12, -13, and so forth to denote the chronological order of
collection for each event.
Sample shipments are made weekly to the central laboratory for
economic reasons for weekly, daily, event, or sequential samples.
Generally, the shipment should be made early in the week (preferably on
Mondays or Tuesdays) and shipped to the central lab by the U.S. Postal
Service first class mail, or by other carriers that will guarantee
delivery within three days. The method used should be prescribed by the
network protocol. All samples must be well-identified, and should be
accompanied by the appropriate data forms.
The central laboratory, upon receipt of the shipment, will replace
the used sample buckets or containers with clean ones by return mail or
other delivery mode.
0
2.6.2 Procedure
Label each sample with station identification, date of sampling
period, and sample weight (Section 2.4.3.1).
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 39 of 49
2.6.2.1 Weekly Cumulative Samples—
a) Be sure the sample is sealed, identified, and accompanied by its
data form.
b) Pack the weekly sample collection bucket or plastic bottle (if a
liner is used) into a cardboard carton or other protective box.
c) Seal the carton, and ship it to the central laboratory by first
class mail or other method if prescribed in the program protocol.
2.6.2.2 Daily, Event or Sequential Samples—
Refrigerate event and sequential samples until they are shipped, and
keep them cold during shipment. Ship by first class mail or other method
if prescribed in the program protocol.
a) Be sure the samples are sealed, properly identified, and
accompanied by their data forms.
b) Pack the samples in cardboard-enclosed Styrofoam boxes with gel
freeze-packs. Keep the freeze-packs in the freezer compartment
of the refrigerator for about 24 h before shipping to ensure that
they are completely frozen. The gel packs are preferred because
they are less likely to leak when thawed. Generally, four packs
per box is sufficient to keep the samples cold for 4 or 5 days.
Seal the cartons, and ship to the central laboratory.
2.6.3 Field Blanks
Field blanks are used to determine the contamination of the sample
bucket or plastic bag bucket liner when there has been no measurable
precipitation for a week. The blanks levels are measured from a thorough
rinse of the bucket or its plastic liner and yield information on the
previous bucket cleaning, operator handling, contamination while in the
collector, and so forth.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 40 of 49
2.6.3.1 Buckets—
For weekly samples, when there has been no precipitation, the empty
wet-side bucket is removed from the precipitation collector. Depending on
network protocol, the bucket is either sealed and shipped to the central
laboratory or treated at the field lab as follows. At the field lab, 100
mL of deionized water is poured into the bucket, and the bucket is
swirled and tipped for the water to reach as much of the interior surface
as possible. Allow the water to stand for about 5 minutes and then
measure the conductance as in Section 2.9. Also measure the conductance
of the deionized water. Record both values in the site logbook and on
the field report form. Record under "Remarks" that the report is for a
field blank. Mail the field report form to the central laboratory.
Rinse out the bucket several more times using sufficient deionized
water (approximately 100 mL) to reach all the surfaces. Collect the
third rinse and measure the conductance. Continue the rinses until there
is no difference between the conductance of the deionized water and the
rinse sample. Cap the empty bucket securely and save for reuse on the
precipitation collector.
2.6.3.2 Bottles—
For daily or event samples which are shipped to the central
laboratory in plastic bottles, if no precipitation has been collected in
a week, prepare a field blank following the procedure described in
2.6.3.1. Perform the rinses that are required for buckets and record the
conductances on the field report form and note under "Remarks" that the
report is for a field blank. If the protocol calls for analysis of the
blank, the first rinse is shipped to the central laboratory in a labeled,
sealed plastic bottle using a similar procedure as for a sample.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 41 of 49
2.7 DOCUMENTATION
All data, observations, and changes or modifications must be
documented with dates on the proper data forms and/or in logbooks. The
data forms should be in duplicate (NCR paper) with one copy kept in the
station records and one shipped with the sample. The logbook entries are
•ade out in duplicate. One copy of the logbook entry is kept at the
station and the other is mailed with the data form and the rain chart to
the central laboratory.
2.7.1 Logbook
Use a bound logbook with perforated pages that can be torn out
easily. Record all problems and actions; dates and times of visits;
equipment changes; procedural changes or modifications; standard solution
changes; electrode replacement; operator changes, and so forth, in the
logbook. Submit one copy to the central laboratory and keep the other at
the station.
2.7.2 Rain Gauge Charts
Mark the rain gauge chart with station identification, dates and
notations for tests, and any problems encountered, and submit weekly to
the central laboratory.
2.7.3 Field Data Forms
The field data form must contain all of the following information:
station, operator, date, sample weight, sampling times, field values of
?H and conductance, and any problems encountered.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 42 of 49
Start a new form when a clean bucket is installed in the collector.
(For daily, event or sequential sampling use a new form for each sample
collected.) Complete the form when the sample is removed from the
collector. An example of a data form used by the State-Operated Network
for weekly sampling is in Figure 2-3. The items below refer to Figure
2-3.
o STATION NAME and ID supplied by the project coordinator.
o OBSERVER'S signature and printed initials; person completing the
form even if substituting for regular observer.
o Enter DATE ON and OFF (mo/day/yr) and the local TIME when sample
buckets are installed and removed; specify 24-h time.
o Check appropriate boxes for the three SITE OPERATIONS. Diagnose
items 1 and 3 from the event pen trace on the rain chart. Add
evidence for item 1, for example, the lack or presence of
moisture in the dry bucket and the reasonableness between the
collector and rain gauge amounts in the PRECIPITATION RECORD
below. Be sure the weight trace is complete for the sampling time
period.
o SAMPLE CONDITION is a qualitative observation of precipitation
quality. Note any comment on obvious causes of the condition
under REMARKS.
o Complete the form for SAMPLE WEIGHT by entering weight of SAMPLE
BUCKET with BAG (if one is used). Include total weight of sealed
bucket, bag, and sample, beneath SAMPLE WEIGHT designation.
Start a new form for newly installed bucket by entering BUCKET
WEIGHT of bucket with a bag. Obtain the weight of precipitation
in exposed bucket by subtracting BUCKET WEIGHT + BAG from BUCKET
+ BAG -t- SAMPLE WEIGHT, and entering it as SAMPLE WEIGHT.
o The PRECIPITATION RECORD gives daily TYPE (if known) and the
AMOUNT (in.). Circle proper type (R,S,M, or U) under each day.
The M denotes a mixture of rain plus snow/sleet/hail. Obtain the
daily AMOUNT from recording rain gauge, and record it in the
squares. Trace (T) indicates precipitation of 0.25mm (0.01 in.)
or less. If rain gauge, chart, or pen malfunctioned and if no
amount can be observed, circle MM. For cumulative weekly
-------
Section No. 2
Revision No. 1
Date July 31, 1986
Page 43 of 49
UJ
cc
o
ULJ
CC
O
s
CC
O
o
O
LU
o
03
•g
a
E
o
IX
o
0
9
U)
OBSERVER
—
e
m —
z S
z
o
•t
-
cc
*" , ,
So
UJ Z
S £
>
5
S;
;= i
a
>
55
Q C
C
2
1 Collector appaa'S 10 have operated properly and sampled all precipitation events during entire sampling period
;>
r
-4
\
5
c
J
)
UJ
a ac
00
CD
3 Collector opened and closed at least once during the week
u
T
H
ION COMPLETE FOR ALL SAMPLES CONTAINING PRECIPITATION
SAMPLE CONOIT
1 Bud droppings
O
Z
(/>
Ul
>
*s
• o
0
Ud
1-
C
>-
u<
X
CJ
=3
CO
ac
>
>
0
o
i
u.
o
2 Cloudy or discolored
3 Soot or dm (Describe All Else in Remarks)
1-
la
u.
C O _.
§h-
UJ
| Si
5 s5
z
I/I
» o>
5 i -
X IA
fi -
25 M
2
>. »-
O «
?« *
§5
W (_ ~
WH S?
3
Z
c _ z o
— O m IAJ
£r5 °5
3
?2
2
o J
3
2
(/) e
K
3
2
(r) e
ac
3
2
I/I e
ac
3
2
e >
a: i
D
2 :
(O • t
OC r
3
2 :
(/) e i-
K r
3 .
2 -
u> e h
OC IV
3 -
2 2
Ul e v
ac rs
Z T MM
Inches (gauge
Inches (wt )
2
2
t- e e
N
5
2
h-
M
i 1
w»
— w
c
u
* S
Gauge
Bight i 0 00058
anon From Rair
ei = Sample W
2 Period Precipif
lion From Samp
Total Samplm
Total Precipita
a
o
UJ
0 s"3 2 ° «
i Jii IH
E £ "5 € s 2 =0
u H-UUIO. < i o
UJ
cc
OL.
6
<%
Ul
? i i
9-
S. > u
l-2o
/> >•*
c •
0 .
I
•o
u
Measur
T3
w
(O
" 13
C
CO
73
V
0
0
"O
re
^
c
re
co
I!
<<2
>-
Q
O
5
J
_«MM
•
•
•
^•H—
II
•
e
)
)
»
>
O
\a
u
at
8
SAMPLE
le Corrected
u
F
re
T3
I/I
ID
;
w>
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 44 of 49
samples, add all daily rain gauge amounts, and record TOTAL
SAMPLING PERIOD PRECIPITATION (in.). Do not merely subtract
initial reading for week from final reading because errors occur
due to evaporation. Convert TOTAL COLLECTOR PRECIPITATION amount
collected from grams to inches by multiplying SAMPLE WEIGHT by
0.00058 in./g, and record in appropriate boxes.
o Space is provided for ONE measurement of sample conductivity and
pH. Only one is necessary, but if the quality of the measurement
is suspect and if over 70 g of sample remain, repeat the
measurement, and record only the final value. Mention only
problems in the remarks section. (Instructions in Sections 2.5.2
and 2.5.3.) Enter DATE of determination as veil as volume in mL
of sample ALIQUOT REMOVED. Record CONDUCTANCE of DISTILLED (or
DKEONIZED) WATER used for rinses and SAMPLE MEASURED conductance
corrected to 25°C. If resistance bridge cannot be adjusted,
insert measured value of 74 yS/cm standard in STANDARD MEASURED
to calculate CORRECTION FACTOR; then calculate and record the
SAMPLE CORRECTED value. For conductivity meters adjusted to 74
US/cm value using KC1 standard, the correction factor is 1.0.
The sample aliquot used for the conductance measurement can also
be used for pH measurement. Never return any aliquot to the bulk
saJiple. Avoid contaminating bulk sample or aliquot. Measure pH
of SAMPLE aliquot (Section 2.5.3). After the measurement is
completed, recheck the pH 4 buffer value, and enter it if it
differs by less than +0.03 from 4.00. If the difference is more
than +0.03, repeat calibration, and remeasure the pH of the
sample aliquot and the pH 4 buffer.
o Obtain SUPPLIES by circling the appropriate material. If pH
standards are needed, write it in this section. To avoid running
out, request new material when about one-fourth of original
supply remains.
o The REMARKS space is provided for the observer to record any
unusual problems, weather, or other occurrences at the field site
or in the laboratory. Unusual occurrences in site area may
include contamination by the operator, moisture in the dry
bucket, plowing, harvesting, burning, increased atmospheric
pollution or dust, or power outage. The importance of the
information requested in the remarks section cannot be
overemphasized. Careful observation of the sample and occur-
rences in the surrounding environment can aid greatly in
evaluating the validity of the sample and in the interpretation
of the data collected.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 45 of 49
2.8 QUALITY CONTROL
Quality control procedures are used on a routine basis to help
assure the collection of high quality data. Complete documentation of
all observations and measurements, the use of known test solutions for pH
tests, and the recheck of the pH calibration after sample measurements
are exaaples of quality controls. In addition, these types of audits are
used to test the field operators and the central laboratory. The first
type, performance audits, uses blind samples made up by the Network
Quality Assurance Manager or central laboratory for pH and specific
conductance measurements to test the measurement capability at the sites.
The second type is field systems audits by an experienced observer. The
third type requires the field personnel to forward a sample received from
quality assurance personnel to the central laboratory disguised as a
regular precipitation sample to test both field and laboratory sample
handling and analysis.
2.8.1 Unknown or Quality Control Test Samples for the Field
To evaluate the quality of each station's pH and conductivity
Measurements as well as to detect problems with these measurements, test
samples of rain-type composition should be received from the central
laboratory on a regular (e.g., quarterly) basis.
1. Measure these samples for pH and conductance as soon as possible after
receipt. Use the same procedure as for precipitation samples
(Sections 2.5.2 and 2.5.3).
2. Fill out a data form; record the data and the results in SAMPLE
CHEMISTRY, and identify the sample in REMARKS.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 46 of 49
3. Return the results on a data form, and the remainder of the sample to
the central laboratory.
At the central laboratory, the sample is remeasured to be sure it
has not changed during shipment to and from the station. Comparisons
between the site and laboratory results will assist in the validation of
routine field data. If the comparison results are poor, the Network
Coordinator should initiate troubleshooting with the field operator to
determine the cause of the problem and take the appropriate corrective
action.
2.8.2 Site Visits/Audits
To review the quality of the measurement system and to evaluate each
station's performance firsthand, a site visit should be conducted once a
year or at least once every two years by experienced personnel. The
audit covers all aspects of site operation.
1. About 4 to 6 weeks before the audit, a questionnaire should be sent
from the Network Coordinator's office to the field personnel. They
fill in the questionnaire (Section 10.0, quality assurance handbook
(3)) and return it prior to the auditor's visit so that the auditor
can assess the operator's overall capabilities and prepare pertinent
questions.
2. The auditor will either add a test sample to a clean bucket at the
station, or give the sample to the operator for measurement. The
operator will weigh the sample, measure its pH and conductivity, and
record the data on a data form.
3. The auditor will inspect all equipment, check the calibration of the
rain gauge, offer advice, and ask questions while the operator goes
through the rounds and tests.
4. If there are any problems, the auditor attempts to correct them; and
failing to do so, the auditor will bring them to the attention of the
site supervisor.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 47 of 49
5. The site personnel will be informed of the results at the end of the
audit. A written audit report will be submitted to the site
supervisor, site sponsor, network coordinator and other officials who
are concerned with operation of the site. A follow-up on corrective
action will be made in 60 days. The follow-up can be by letter,
telephone or revisit.
2.8.3 Blind Samples for the Laboratory
Blind samples will be sent to each station at various times for
testing and forwarding to the central laboratory for analysis as part of
the quality assurance program. The following are instructions and
guidelines to be followed:
1. A reference sample can be shipped in a 500-mL polyethylene bottle with
tvo preaddressed postcards, a mailing label, and a set of data.
2. Refrigerate the sample at 4°C until it can be submitted to the central
laboratory during a week in which your site had no wet deposition.
3. If your precipitation samples are submitted in buckets, pour the
contents of the bottle into a clean sample bucket (one which has not
been used in the field) when you are ready to submit the reference
sample; weigh, and record as usual on a data form.
If your precipitation samples are sent in bottles, transfer the sample
to one of your bottles, weigh, and record as usual on a data form.
4. Remove your normal aliquot and measure its pH and specific
conductance; record these values as usual on the data form.
5. Fill out the rest of the regular field report form, and ship it with
the sample to the central laboratory as a normal precipitation sample.
Make up the information for PRECIPITATION RECORD, and so forth, on the
data form.
6. Furnish the information requested on the two postcards which
accompanied the sample and mail the self-addressed cards.
7. Place a clean bucket in the collector, and proceed as usual.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 48 of 49
2.9 FIELD PROCEDURE SUMMARY
To serve as an outline, an operating procedure summary is given
belov. It includes the basic steps, but it is not complete. The
conductivity and pH procedures are those applicable to most of the
current instruments, but they may not pertain to a specific instrument
being used. Check the manufacturer's instructions, and adapt this
summary to your network as necessary.
Site Visits
a) Daily: check rain gauge for event occurrence. If an event
occurred, record its date and time, number of lid openings, and
amount of precipitation from gauge. Note the weather. Veekly:
change chart, fill pens, and wind clock. Monthly: check rain
gauge calibration, and clean collector sensor. Check the
condition of the lid pad.
b) Check dry side buckets for moisture and other unusual
occurrences. If not interested in the dry bucket analysis, wipe
off the rim and leave the dry bucket in place. The dry side
buckets should be cleaned weekly and replaced semi-annually.
c) If event occurred, replace wet bucket with a newly weighed one.
Put new weighed lid firmly on removed sample bucket. Record
observations on data form and in logbook.
d) Check collector, sensor, and rain gauge for problems.
Saaple Handling
a) Wipe outside of bucket dry; tap lid to knock off drops; remove,
weigh bucket (to nearest 1.0 g). Record on sample data form.
b) If sample is frozen, allow it to melt completely.
-------
Section No. 2
Revision No. 2
Date July 31, 1986
Page 49 of 49
c) If sample is sent to the laboratory in a bottle rather than
bucket, pour the sample into a 125, 250 or 500 mL sample bottle.
Discard any sample over bottle capacity. Mark total sample
veight on bottle.
d) Rinse sample bucket with deionized or distilled vater, shake, and
drain.
e) For sample in sealed bucket or bottle, allow at least 1 h for
sample to reach rooa temperature before performing measurements.
2.10 REFERENCES
1. Galloway, J.N., and G.E. Likens, Vater, Air and Soil Pollut. 6, 241
(1976).
2. Galloway, J.N., and G.E. Likens, Tellus 30, 71 (1978).
3. Quality Assurance Handbook for Air Pollution Measurement Systems, Vol.
V - Manual for Precipitation Measurement Systems,Part I- Quality
Assurance Manual'. UTjT Environmental Protection Agency, Research
Triangle Park, NC. EPA-600/4-82-042a (January 1981). October 1984
Revision in print.
4. Martin, C.W. NADP Winter Operation of Sampler, Hubbard Brook
Experimental Forest, West Thornton, NH; letter to V.C. Bowersox, March
25, 1980.
5. Eaton, W.C., and E.D. Estes, "Use of Plastic Bags as Bucket Liners For
the Aerochem Metrics Precipitation Collector", Research Triangle
Institute, Research Triangle Park, NC, RTI-2474-86 (May 1984).
6. Methods for Chemical Analysis of Water and Wastes, U.S. Environmental
Protection Agency, Cincinnati, OH, EPA-600/4-79-020 (March 1979).
7. Koch, W.F., and G. Marinenko, Guidelines for the Measurement of pH in
Acidic Rainwater. National Bureau of Standards, for EPA.
8. Galloway, J.N., B.J. Cosby, G.E. Likens, and J. Limnol. Oceanogr.
24, 1161 (1979).
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 1 of 20
3.0 CENTRAL LABORATORY SUPPORT OPERATIONS FOR THE FIELD
The central laboratory will supply clean containers to the field
sites, prepare reference standards to be used to calibrate field
instruments, and furnish quality control samples for use in the field.
This section discusses the care of glass and plasticvare, the preparation
of reference solutions, and the evaluation of field equipment by the
central laboratory. The referred data forms are in Section 3.7.
The central laboratory will serve as a focal point for solving field
operational problems involving equipment malfunctions. It will serve as
a central distributor of replacement parts for the collector, rain gauge,
pfl meter, conductivity meter, balance, thermometers or thermistor probes,
and ancillary supplies. The central laboratory will provide consultation
service to field personnel on any technical question involving siting,
sample collection, analysis, data quality and transport of the collected
saaple. The central laboratory will work with site personnel as
intensively as necessary to assure that data quality meets standards set
by the monitoring program.
3.1 CLEANING AND SUPPLYING OF GLASSWARE AND PLASTICVARE
3.1.1 Cleaning of New or Used Plasticware
a) Rinse with deionized water 6 to 10 times. NOTE: If the plastic
needs to be rubbed to remove a film, use a natural sponge.
b) Let stand, filled with deionized or distilled water for 48 h.
Empty and dry in an oven at 70°C.
c) After initial cleaning (steps 1 and 2), check a portion (~10£) of
the containers to ensure that rinsing has been adequate. To do
this, add 50 mL of deionized water to the cleaned container, seal
the container with a cap or with Parafilm, and slowly rotate it
so that the water touches all inner surfaces. DO NOT SHAKE.
Check the conductivity of the water (Section 4.3); it should be
less than 2.0 yS/cm. If any of the containers fail the check,
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 2 of 20
rerinse all of the containers cleaned for the checked samples and
retest 10*.
d) After the plasticvare is clean and dry, cap the containers and
place them in a plastic bag to be sealed for shipment or storage.
3.1.2 Cleaning of Glassware
3.1.2.1 Glassvare Used for Metal Analyses—
a) Rinse vith deionized vater twice and with 10* HNO~ once.
b) Rinse 6 to 10 times vith deionized vater.
3.1.2.2 Glassvare Used for Anions and NH*—
a) Rinse vith deionized vater tvice and vith 10% KOH solution once.
b) Rinse 6 to 10 times vith deionized vater.
c) If vater beads on the inner surface, the glassware needs to be
cleaned more thoroughly. Wash vith detergent, and then clean
vith 10% KOH solution. If vater still beads, soak the glassware
overnight in 10% KOH, and rinse 6 to 10 times vith deionized
water.
3.1.3 Supplying Containers to the Field
After a sample shipment has been logged in at the central
laboratory, replace the bucket or other sample container with a clean
one. The clean, sealed containers are shipped to the field site in
plastic bags and shipping cartons on an as-needed basis to maintain their
supply. If cold packs and insulated containers are used, these are
returned also. Check that the Styrofoam boxes are intact and not
cracked; if cracked, replace with a new one. The shipment can be made by
ground transport since each site should have a 3-week supply of these
aaterials on hand.
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 3 of 20
3.2 PREPARATION OP STANDARDS FOR THE FIELD
3.2.1 Preparation and Measurement of Conductivity Standards
a) Weigh out 7.456 g of predried (2 h at 105°C) KC1 and dissolve it
in 1 liter of deionized water (0.10M KC1).
b) Dilute 20 mL of the 0.1M KC1 to 4 liters with deionized water
(0.0005M KC1).
c) Fill washed 0.5-liter plastic bottles with the 0.0005M KC1
solution to be sent to the field. Label the bottles with the
preparation date and keep the solutions refrigerated.
d) Measure the conductivity of the solution in each bottle (Section
4.3).
e) Fill out the Field Conductivity Standard form and label the
bottle with the measured conductivity.
f) Send new standards to the field monthly. Vhen old standards are
returned to the laboratory, remeasure the conductivity. Complete
the Field Conductivity Standard form.
3.2.2 Preparation and Measurement of pH Reference Solution
a) Prepare 4 liters of a 10" to 10 N H-SO, solution for the pH
reference solution by diluting 4 mL or 0.4 mL of commercially
available 0.100N sulfuric acid stock solution.
b) Fill washed 500-raL bottles with the pH reference solution. Label
the bottles with the preparation date. Keep the solutions
refrigerated.
c) Measure the pH of the solution in each bottle (Sections 4.2).
d) Fill out the Field pH Test Solution form, and label the bottle
with the measured pH.
e) Remeasure the pH of these solutions after they are returned from
the field. Complete the Field pH Test Solution form.
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 4 of 20
3.2.3 Preparation of Quality Control Samples
a) Monthly, prepare a mixed solution of 10~ to 10~ N H,jSO, from
commercially available 0.100N solution and 0.0001M to 0.0003M KCl
to be used as a quality control sample. A 0.10M KCl stock
solution is prepared by dissolving 7.456 g of predried (2 h at
105°C) KCl in 1 L of deionized water at 25°C. For 0.0001N KCl,
dilute 1 mL of the 0.10 M stock solution to 1000 mL with
deionized water.
b) Fill clean 60-raL polyethylene bottles with the mixed audit
sample, and send each site one sample. Three bottles should be
retained by the laboratory.
c) Immediately measure the three samples kept by the laboratory for
pH (Section 4.2) and conductivity (Section 4.3). Check the
laboratory electrode against another backup electrode for one
sample. Fill out the appropriate section of the Field Quality
Control Audit Sample Report (Section 3.6.). Refrigerate the
laboratory samples.
d) When the field quality control audit samples from all sites have
been returned to the laboratory, reanalyze the samples along with
the laboratory's three aliquots. Check the laboratory electrode
against another backup electrode for one sample. Complete the
Summary Field Quality Control Audit Sample Report.
3.3 INITIAL EVALUATION OF FIELD EQUIPMENT
All meters and electrodes should be tested before they are shipped
to the field. The meters have a serial number affixed, but the
electrodes do not. A unique identification number should therefore be
taped to each electrode.
3.3.1 Evaluation of Conductance Meters and Cells
3.3.1.1 Evaluation of Accuracy and Precision of Meter—
a) Prepare a 0.0003M KCl Test Solution—Dilute 3 mL of the stock
0.10M KCl solution (Section 3.2) to 1 liter with deionized water.
(Prepare daily.)
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 5 of 20
b) Calibrate the Field Conductance Meter—Calibrate
the manufacturer or as described in Section 4.3.
as indicated by
c) Pill 11 Vials or Plastic (17x100 mm) Tubes—Pill to a depth of
3 cm (or to cover the electrode) with the 0.0003M KCl. The first
tube is to be used as a rinse tube.
d) Measure the Conductance of the 10 Solutions—Between each
measurement, rinse the conductivity cell thoroughly with
distilled water, carefully shake it dry, and dip it in the rinse
solution three times.
e) Calculate an Average Value and the Standard Deviation—Use the
following relationships. Programmed calculators make this a
simple operation.
10
E
x.
3-1
and
10
I
1/2
3-2
where
xi
x
s
n
the measured value (in pS/cra or pH units),
the average value,
standard deviation, and
the number of values.
f) Record the Results—Record results on the Conductance Meter/Cell
Acceptance Test form and the Conductance Acceptance Test Summary
Form. The conductivity meter and cell are acceptable if the
average value is within 2% of the theoretical value of 44.6 uS/cm
(25°C) and if the relative standard deviation is less than 2%.
-------
Section No. 3
Revision No. I
Date July 31, 1986
Page 6 of 20
3.3.1.2 Evaluation of Linearity of Meter--
a) Prepare three of the following standards so that each range of
the meter has at least one standard
Standard
147.0 pS/cm
75.0 pS/cm
44.6 pS/cm
14.9 pS/cm
7.5 pS/cm
Preparation
Dilute 1 mL of 0.1N
KCl to 100 mL
Dilute 500 pL of 0.1N
KC1 to 100 mL
Dilute 300 PL of 0.1N
KCl to 100 mL
Dilute 100 pL of 0.1N
KCl to 100 mL
Dilute 50 pL of 0.1N
KCl to 100 mL
Normality
KCl
0.001
0.0005
0.0003
0.0001
0.00005
b) Calibrate the field conductivity meter as indicated by the
manufacturer.
c) Measure the conductivity of each standard as described in Section
4.3.1.5.
d) Determine the linearity of the meter by performing a linear least
squares fit on the data. Record the results on the Conductance
Meter/Cell Acceptance Test Form and the Conductance Acceptance
Test Summary Form. The coefficient of correlation should be 0.999
or better. If it is less than 0.999, the meter should be
adjusted. Some manufacturers provide procedures for these
adjustments. Otherwise, the meter should be returned to the
manufacturer for calibration.
3.3.2 Evaluation of pH Meters
a) Calibrate the Field pH Meter—Calibrate as indicated by the
manufacturer or as described in Section 4.2.4. A laboratory pH
electrode of documented performance should be used.
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 7 of 20
b) Pill 11 Vials or Plastic (17x100 mm) Tubes—-Fill to a depth of
3 cm with fresh pH electrode reference solution (Section 3.2.2).
The first tube is to be used as a rinse tube.
c) Measure the pH of the 10 Solutions—Between each measurement,
rinse the pH electrode thoroughly with deionized or distilled
water, carefully drain or shake it dry, and dip it in the rinse
solution three times.
d) Calculate an Average Value and the Standard Deviation—See
Section 3.3.1.1, (e).
e) Record the Results—Record on the pH Meter/Electrode Acceptance
Test Form and the pH Acceptance Test Summary Form. The pH meter
is acceptable if the average pH is within 0.1 pH unit of the
calculated value and the standard deviation is less than 0.03
units. (Calculated pH =» -log (Normality H^SO^).
3.3.3 Evaluation of pH Electrodes
a) Assign Each New pH Electrode an Identification Number—Allow it
to equilibrate overnight in 1 x 10~ mol/L HCl.
b) Rinse the Electrode Carefully with Deionized Water—Prior to
testing, and then place it successively in deionized water in
different test tubes until a constant pH reading is achieved.
c) Calibrate the Laboratory pH Meter—Calibrate as indicated by the
manufacturer or as described in Section 4.2.
d) Measure the pH of 10 Tubes—Measure pH reference solution as
described in Section 3.3.2 and note drift, noise and response
time.
e) Calculate an Average Value and the Standard Deviation—See
Section 3.3.1.1, (e').
f) Record the Results—Record on the pH Meter/Electrode Acceptance
Test Form and the pH Acceptance Test Summary Form. The pH
electrode is acceptable if the average pH is within 0.1 pH unit
of the calculated value and if the standard deviation is less
than 0.03 pH unit.
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 8 of 20
3.3.4 Evaluation of Field Balance and Thermometers
Reference weights traceable to MBS are necessary for balance
calibration. Each laboratory should purchase a set of NBS-traceable
weights to be used to certify a set of working weights used in the field
and laboratory. Semi-annual calibration checks are recommended. The
procedure used to certify weights is as follows:
a) zero the balance according to manufacturer's recommendations,
b) veigh the working certified 1.0 and 5.0 Kg weights,
c) weigh reference 1.0 and 5.0 Kg weights,
d) repeat this procedure five times, and
e) complete the Certification of Vorking Weights to MBS form.
Working reference weights should be certified by this procedure once
a year. The NBS-traceable weights are kept as primary standards. All
working reference weights should weigh within 0.1% of the NBS-traceable
weights.
Each laboratory should have an NBS-traceable thermometer. One
thermometer in the laboratory should be certified against the
NBS-traceable standard. Keep the NBS-traceable thermometer as a primary
standard. Assign all laboratory and field thermometers (or temperature
probes) identification numbers, and then calibrate them against the
(secondary) certified thermometer. Calibrate the temperature probes in a
circulating water bath in the 0° to 25°C range against the certified
thermometer, and complete the Thermometer Calibration Log Form. File one
copy in the laboratory, and send another to the field with the
temperature probe. If the water bath does not have a cooling coil, an
ice-water mixture can be used to achieve the low-temperature reading.
Calibration at two temperatures, near 0° and 25°C, is sufficient and a
linear temperature behavior may be assumed. Temperature differences
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 9 of 20
between the certified thermometer and the test thermometer should be no
greater than 0.5 degree.
3.4 MONITORING OF FIELD OPERATION
It is the function of the central laboratory to monitor field
results to determine if a site is operating properly. The laboratory
receives a copy of the site log book and sample data sheets veekly. From
these and the results of the monthly Quality Control Sample (Section
3.2.3) the central laboratory can determine if the site is functioning
properly.
3.4.1 Evaluation of Field Conductivity and pH Measurement Systems
As needed, the central laboratory sends Quality Control (QC) samples
(Section 3.2.3) to each field site. After all samples have been measured
for conductivity and pH by the site operators, returned to the central
laboratory and remeasured by the central laboratory, the results are
recorded on the Monthly Field Audit Report (Section 9.1.2, QA Manual
(1)). Accuracy is estimated and the acceptance criteria are applied.
If a site's QC sample result is outside the pH acceptance criteria,
the pH reference solution (Section 3.2.2) values recorded by the site
operator on recent field sample data sheets' are checked. If these
results are also questionable, a new pH electrode which has been checked
by the central laboratory (Section 3.3.3) is sent to the field with a new
QC sample. If the pH system still does not function properly, the pH
meter is replaced.
If a QC conductance value measured by the site operator is outside
the conductivity acceptance criteria, a new conductance standard is sent
to the site. If this does not solve the problem, the conductivity meter
and cell are replaced. In all cases the central laboratory personnel
communicate with the site operator to determine if there is an outside
cause of malfunctioning.
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 10 of 20
3.4.2 Evaluation of Field Precipitation Collector, Rain Gauge and
Balance
All measurements in the field, including sample weight and rain
gauge reading, are recorded by the site operator on field data forms and
all observations are noted in a bound log book vith perforated second
copies that can be torn out easily. These pages are sent to the central
laboratory for review along vith the data forms. Central laboratory
personnel evaluate the performance of the precipitation collector, rain
gauge, and balance from these sources, the weekly telephone call, and the
precipitation collector collection efficiency. In addition, the
performance of the rain gauge and balance are evaluated by reviewing the
field records when the site operators check the rain gauge and balance
with known weights. During field audits, the auditor evaluates all site
equipment.
3.5 REPORT FORMS
Blank data forms are included in this section for the convenience of
the manual user. Use of the forms is discussed throughout Section 3. The
forms included are listed below:
Title
Field Conductivity Standard Report
Field pH Reference Solution
Field Quality Control Audit Sample Report - Laboratory Data
Conductivity Meter/Cell Acceptance Test Report
Conductivity Acceptance Test Summary
pH Meter/Electrode Acceptance Test Report
pH Acceptance Test Summary
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 11 of 20
Certification of Working Weights to NBS-Traceable Standards
Thermometer Calibration Log
3.6 REFERENCES
1. Quality Assurance Handbook for Air Pollution Measurement Systems, Vol.
V - Manual for Precipitation measurement Systems, Part I - Quality
Assurance ManualJU.S.EnvironmentalProtectionAgency,Research
TriangleParTcT NC. EPA-600/4-82-042a (January 1981). Revised
January 1985.
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 12 of 20
FIELD CONDUCTIVITY STANDARD REPORT
DATE OP PREPARATION OF
0.1M KC1 STOCK SOLUTION:
DATE OF PREPARATION OF
DILUTE FIELD STANDARD: (Analyst's Signature)
LABORATORY ANALYSIS BEFORE SHIPMENT TO THE FIELD (yS/cm)
s -
Laboratory Values After Use In The Field:
Lab Value Date of Lab Analyst's
Field Site! (uS/cm) Analysis Initials
Operations & Maintenance Manual for Precipitation Measurement Systems
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 13 of 20
FIELD pH REFERENCE SOLUTION
DATE OF PREPARATION OF
0.1N H2SOA STOCK SOLUTION
VOLUME OF STOCK TAKEN (Analyst's Signature)
FINAL DILUTION VOLUME OF DATE OF PREPARATION
FIELD pfl ELECTRODE TEST SOLUTION OF TEST SOLUTION
LABORATORY ANALYSIS BEFORE SHIPMENT TO THE FIELD (pH)
pH
LABORATORY ANALYSIS OF ALIQUOTS RETURNED FROM THE FIELD
Field Lab Values After Return
Site ft Date of Analysis pH Anal. Init.
Operations & Maintenance Manual for Precipitation Measurement Systems
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 14 of 20
FIELD QUALITY CONTROL AUDIT SAMPLE REPORT
SAMPLE #
Date of Preparation of Field Audit Sample:
Vol. of H2SO^ Stock Used: mL; Normality of HjSO^ Stock:
Date of H-SO, Stock Preparation:
Vol. of KC1 Stock Used: mL; Normality of KC1 Stock:
Date of KC1 Stock Preparation:
Final Dilution Volume of
Field Audit Samples: mL; (Analyst's Signature)
LABORATORY ANALYSIS BEFORE SHIPMENT*
TO THE FIELD
Conductivity pH
Average and
Std. Dev.
LABORATORY ANALYSIS OF ALIQUOTS RETURNED FROM THE FIELD
DATE
OF LAB CONDUCTIVITY pH ANALYST'S
SITE * ANALYSIS VALUE VALUE INITIALS
* These data are for three laboratory aliquots which are analyzed before
shipment of samples, are then refrigerated, and are reanalyzed with the
samples returned from the field.
Operations & Maintenance Manual for Precipitation Measurement Systems
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 15 of 20
CONDUCTIVITY METER/CELL ACCEPTANCE TEST REPORT
DATE OF TEST: ___,
(Analyst's Signature)
PREPARATION DATE OP
KC1 REFERENCE SOLUTIONS:
METER TYPE/SERIAL NO. /
CONDUCTIVITY CELL TYPE/SERIAL NO.
INDICATE WHETHER TEST OP METER OR CELL
CONDUCTIVITY VALUES OBTAINED FOR PRECISION TEST (0.0003M KC1 SOLUTION)
(Section 3.5.1.1)
Aliquot 1: Aliquot 6:
Aliquot 2: Aliquot 7:
Aliquot 3: Aliquot 8:
Aliquot 4: Aliquot 9:
Aliquot 5: Aliquot 10:
Average conductivity and
standard deviation:
CONDUCTIVITY VALUES OBTAINED FOR LINEARITY TEST (Section 3.5.1.2)
Normality Expected Found
KC1 Conductivity (uS/cm) Conductivity (yS/cm)
Slope _
Intercept
Linearity
Accepted Rejected
Operations & Maintenance Manual for Precipitation Measurement Systems
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 16 of 20
CONDUCTIVITY ACCEPTANCE TEST SUMMARY
Meter Type/
Serial f
Cell Type/
Serial *
Date of
Check
Linearity
Conductivity Value
Average and Standard
Deviation( yS/cm)
Number
of
Values
Anal.
Initials
Operations & Maintenance Manual for Precipitation Measurement Systems
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 17 of 20
pH METER/ELECTRODE ACCEPTANCE TEST REPORT
DATE OP TEST:
(Analyst's Signature)
PREPARATION DATE OP
pfl ELECTRODE REFERENCE SOLUTION:
SORMALITY OF pH ELECTRODE REFERENCE SOLUTION:
Heter Type/Serial No. /
pH Electrode Type/Serial No. /
Indicate whether test of meter or
electrode
pfl VALUES OBTAINED;
4.0(3.0) Buffer before:
7.0(6.0) Buffer before:
Aliquot 1:
Aliquot 2:
Aliquot 3:
Aliquot 4:
Aliquot 5:
Aliquot 6:
Aliquot 7:
Aliquot 8:
Aliquot 9:
Aliquot 10:
4.0(3.0) Buffer after:
7.0(6.0) Buffer after:
?5 ELECTRODE REFERENCE SOLUTION; (Section 3.5.2)
Calculated pH of reference solution:
Average pH and standard deviation:
Check One: Accepted
Rejected
Operations & Maintenance Manual for Precipitation Measurement Systems
-------
Section No. 3
Revision No. 1
Date July 31, 1986
Page 18 of 20
pH ACCEPTANCE TEST SUMMARY
Meter Type/
Serial t
Electrode
Type/
Serial*
Date of Ref
Soln. Prep.
Date
of
Check
pH Value
Average and Standard
Deviation
Number
of
Values
Anal.
Initials
Operations & Maintenance Manual for Precipitation Measurement Systems
-------
Section No. 3
Revision No. 1
Date July 31,-1986
Page 19 of 20
CERTIFICATION OF WORKING WEIGHTS TO NBS-TRACEABLE STANDARDS
DATE OP CERTIFICATION:
WEIGHT SET SERIAL #:
BALANCE 0
MBS 1kg
NBS 5kg
(Analyst's Signature)
BALANCE 0
NBS 1kg
NBS 5kg
TEST 1kg
TEST 5kg
BALANCE 0
SBS 1kg
BBS 5kg
* * *
TEST 1kg
TEST 5kg
BALANCE 0
NBS 1kg
NBS 5kg
* * *
TEST 1kg
TEST 5kg
BALANCE 0
SBS 1kg
KBS 5kg
TEST 1kg
TEST 5kg
* * *
TEST 1kg
TEST 5kg
BALANCE 0
NBS 1kg
NBS 5kg
TEST 1kg
TEST 5kg
* * *
SUMMARY: (Section 3.5.4)
Average and Standard Deviation
BALANCE 0
NBS 1kg
NBS 5kg
Check One:
Accepted
TEST 1kg
TEST 5kg
Rejected
Operation & Maintenance Manual for Precipitation Measurement Systems
-------
THERMOMETER CALIBRATION REPORT
Section No. 3
Revision No. 1
Date July 31, 1986
Page 20 of 20
DATE OF CALIBRATION:
IDENTIFICATION NUMBER:
(Analyst's Signature)
Certified Thermometer (°C) Test Thermometer (°C) Correction (°C)
(Section 3.3.4)
Check one: Accepted
Rejected
Operation & Maintenance Manual for Precipitation Measurement Systems
-------
Section No. 4
Revision No. 1
Date July 31, 1986
Page 1 of 9
4.0 LABORATORY PROCEDURES
The laboratory procedures herein are for chemical measurements and
analyses of precipitation samples. Methods include analyses for pH,
specific conductance, acidity, NH*, P0~3, S0~2, N0~, Cl~, F~, Na+, K+,
Ca"1"1', Mg++, and dissolved Al, Cd, Cu, Fe, Pb, Mn and Zn. Detection
limits for these procedures vill vary with instruments and conditions,
but representative detection limits, concentration ranges, precision and
1
bias are presented in each method. Brief discussions of the methods are
presented in this section vith the full text of the procedures in the
appendices.
4.1 GRAVIMETRIC MEASUREMENTS
In both the field and the central laboratory the volume of rainwater
is determined by measuring the mass of the rain and multiplying the mass
3
by 1 g/cm to obtain the volume. The mass of rain is measured in the
field to determine the rain collector efficiency (compared to that of the
rain gauge), and the mass of the sample sent to the central laboratory is
measured as a check to determine if leakage occurred in shipment. The
sample should be sent to the central laboratory in plastic buckets or
plastic bottles. If sent in the bucket, the sample is weighed at the
field station and by the central laboratory. If transferred to
polyethylene bottles, the sample is weighed in the bucket at the field
station and the weight entered on the Field Data Form. The volume
received is estimated by the central laboratory. If collected in bottles,
the sample is weighed by both the field station and the central
laboratory.
4.1.1 Apparatus
The balance should have a capacity of 20 kg and a precision of at
least +10 g (for bucket weighing) or 1 kg with a precision of +0.5 g (for
bottle weighing).
-------
Section No. 4
Revision No. 1
Date July 31, 1986
Page 2 of 9
4.1.2 Calibration
Calibrate the balance monthly, using weights traceable to
NBS-traceable weights. Store the NBS-traceable weights (primary
references) in the laboratory, certify the working calibration weights
against these, and complete the Certification of Vorking Weights to NBS
report (Section 3.4). Recertify all working calibration weights against
the NBS-traceable weights every six months. The procedure for weight
certification is the same as for field balances (Section 3.3.4).
Calibrate each balance using weights close to those actually measured.
Calibrate high-capacity balances, using 1.0 and 5.0 kg weights.
4.1.3 Procedure
To obtain the weight of the sample received in the laboratory, use
the following procedure. During the procedure, avoid breathing on the
sample to avoid NH~ contamination.
a) Be sure that the balance is level, and then adjust its zero knob
so that the balance zeroes (see manufacturer's instructions).
b) Place the bucket without its lid or the plastic bottle on the
balance pan, and weigh it to the nearest 10 grams or place the
bottle on the balance and weigh it to the nearest gram.
c) Record the weight on the bucket or bottle label and on a data
sheet.
d) Subtract the initial weight of the empty container (recorded on
the Field Data form) from the final weight of container plus the
sample to obtain the sample weight shipped.
4.2 pH MEASUREMENT
pH is measured in precipitation samples electrometrically by using
either a pH half cell with a reference electrode or a combination
electrode. The pH meter/electrode(s) measurement system is calibrated
-------
Section No. 4
Revision No. 1
Date July 31, 1986
Page 3 of 9
vith tvo reference buffer solutions that bracket the expected sample pH.
The acceptable method for the measurement of pH is presented in Appendix
C.
4.3 CONDUCTANCE MEASUREMENT
Specific conductance is measured in precipitation samples
electrolytically using a conductance cell. The conductance meter/cell
system is calibrated using potassium chloride solutions of known specific
conductances in the range of precipitation samples. The acceptable
•ethod for the measurement of specific conductance can be found in
Appendix D.
4.4 SAMPLE FILTRATION
After measuring the pH and specific conductance, but before
•easuring the other analytes, filter the rainwater sample. Use vacuum or
pressure filtration to minimize exposure of the sample to laboratory air.
The vacuum apparatus can be a bell jar (ground-glass plate) of sufficient
size to contain a 250-mL (8-oz) bottle, or it can be the apparatus used
by the Illinois State Water Survey for the National Atmospheric
Deposition Program (NADP). The recommended filter material is a 0.45- m
membrane filter (Millipore HA); the filter funnel should be plastic.
Before each filtration, thoroughly rinse the apparatus, including the
filter, with 200 mL deionized water and a portion of sample if there is a
sufficient amount. Filter the sample as quickly as possible, and cap the
labeled sample bottle containing the filtrate to minimize contact with
laboratory air. If the filtered particulates are to be analyzed, they
should be oven-dried at 60°C for one hour and stored in glass vials.
-------
Section No. 4
Revision No. 1
Date July 31, 1986
Page 4 of 9
4.5 ACIDITY MEASUREMENTS
Two methods for measuring acidity are acceptable. The first method
measures both strong and total acidity, while the second method only
measures total acidity.
In the first method, strong and total acidity are measured in the
precipitation samples by titrating the sample with dilute sodium
hydroxide solution and following the titration electrometrically with a
pH half cell and reference electrode or a combination electrode. The
titration is continued until a pH of 10.4 is reached. A method first
introduced by Gran (1) is used to calculate the strong and total acidity.
The Gran functions are plotted versus the volume of titrant added. The
total and strong acidity are obtained by extrapolating the linear
portions of the curve to zero. Weak acidity is obtained by subtracting
the strong acidity from the total acidity. This method can be found in
Appendix E.
In the second method, total acidity is measured by titrating the
sample elect rometrically vith a combination pH electrode to a pB of 8.3.
The total acidity is calculated from the volume and concentraiton of
titrant. This method is included with the above method in Appendix E.
4.6 DETERMINATION OF SULFATE
Sulfate is measured in the precipitation samples by one of two
methods; ion chromatography or automated colorimetry using barium-
methylthymol blue. The ion chromatographic method utilizes ion exchange
resins for separation and conductivity for detection. After a sample is
injected onto the separator column containing the ion exchange resin, an
eluent is used to pump the sample through the column. The anions are
separated depending on their radius and valence. After eluting from the
-------
Section No. 4
Revision No. 1
. Date July 31, 1986
Page 5 of 9
separator column, the ions are converted to the corresponding acids which
are detected in the conductance cell against a background of neutralized
eluent.
The automated colormetric method of analysis for sulfate uses the
blue colored barium-methylthymol blue complex to determine the
concentration of sulfate in the sample. After interfering cations are
removed by an ion exchange column, the sulfate in the sample reacts with
barium chloride to form barium sulfate. Excess barium ions react with
the methyl thymol blue to form the chelate. Thus, the concentration of
the sulfate in the sample is inversely proportional to the intensity of
the blue-colored chelate which is measured colorimetrically at 460 nm.
The ion chromatographic method is presented in Appendix F, while the
colorimetric method can be found in Appendix G.
4.7 DETERMINATION OF NITRATE
Nitrate is measured in the precipitation samples either by ion
chromatography or automated colorimetry using cadmium reduction. The ion
chromatographic method is identical to the method described in Section
4.6. It can be found in Appendix P. '•
The colorimetric method uses a color reagent made from NEDA
(n-(l-naphthyl)-ethylene-diamine dihydrochloride), phosphoric acid and
sulfanilamide to develop a color that can be used to measure the nitrate
present in a sample. After mixing with ammonium chloride, the nitrate in
the sample is reduced in a copper-cadmium column to nitrite. The nitrite
is mixed with the color reagent and forms a reddish-purple complex which
is measured colorimetrically at a wavelength of 520 nm. The
concentration of the original nitrate in the sample is directly
proportional to the intensity of the color complex formed by the nitrite.
This method can be found in Appendix H.
-------
Section No. 4
Revision No. 1
Date July 31, 1986
Page 6 of 9
4.8 DETERMINATION OF CHLORIDE
Chloride is measured in precipitation samples by chemically
suppressed ion chromatography or automated colorimetry using thiocyanate.
The ion chromatographic method is identical to the method described in
Section 4.6. It can be found in Appendix P.
The colorimetric method of analysis for chloride uses a colored
ferric thiocyanate complex to determine the concentration of chloride in
the sample. The chloride ions react with mercuric thiocyanate liberating
thiocyanate ions which reacts with ferric ions. The concentration of the
original chloride ions in the sample is directly proportional to the
intensity of the colored ferric thiocyanate complex. This method is
included vith this manual as Appendix I.
4.9 DETERMINATION OP ORTHOPHOSPHATE
Orthophosphate is measured in the precipitation samples either by
ion chromatography or automated colorimetry using the phosphomolybdenura
blue complex. The ion chromatographic method is identical to the method
discussed in Section 4.6. It can be found in Appendix P.
The colorimetric method involves developing the phosphomolybdenum
blue complex by mixing the sample vith an acidified solution of ammonium
aolybdate, ascorbic acid and antimony potassium tartrate, and passing the
mixture through a 37°C temperature bath. The concentration of
orthophosphate is proportional to the intensity of the phosphomolybdenum
blue complex measured coloriraetrically at 880 nm. This method is
presented in Appendix J.
-------
Section No. 4
Revision No. 1
Date July 31, 1986
Page 7 of 9
4.10 DETERMINATION OF FLUORIDE
Fluoride is determined in the precipitation samples using an
ion-selective electrode with a reference electrode. The meter/electrode
system is calibrated vith fluoride solutions of known concentrations. The
specified method is presented in Appendix K.
4.11 DETERMINATION OF AMMONIUM
Ammonium is determined in the precipitation samples by one of three
methods; ion chromatography, ion-selective electrode or automated
colorimetry using the indophenol blue complex.
The ion chromatographic method utilizes ion exchange resins for
separation and conductance for detection. After a sample is injected
onto the separator column containing the ion exchange resin, an eluent is
used to pump the sample through the column. The cations are separated
depending on their radius and valence. After eluting from the separator
column, the ions are converted to the corresponding bases which are
detected in a conductance cell against a background of neutralized
eluent. This method can be found in Appendix F.
The second acceptable method for determining ammonium is to use a
gas sensing ion-selective electrode with a reference electrode. Ammonium
ion is converted to ammonia gas when the pH of the sample is adjusted to
pH 11-14. An electrode potential develops across the sensing membrane in
proportion to the ammonia concentration in solution. The meter/electrode
system is calibrated with ammonium solutions of known concentrations.
This method is included as Appendix L of this document.
The third method for determining ammonium in precipitation samples
is the automated colorimetric method using the indophenol blue complex.
After removing cations that could form hydroxide complexes, the sample is
-------
Section No. 4
Revision No. 1
Date July 31, 1986
Page 8 of 9
•ixed with alkaline phenol and hypochlorite to form the indophenol blue
complex. Sodium nitroprusside is added to precipitation samples to
intensify the color. The concentration of the ammonium is proportional
to the intensity of the indophenol blue complex measured colorimetrically
at 630 n». This method can be found in Appendix L with the second method
described above.
4.12 DETERMINATION OF SODIUM, POTASSIUM, MAGNESIUM AND CALCIUM
Sodium, potassium, magnesium and calcium are measured in
precipitation samples either by chemically suppressed ion chromatography
or flame atomic absorption spec tropho tome try. The ion chroma tographic
procedure is identical to the one described in Section 4.11 for ammonium.
It can be found in Appendix F of this document.
The flame atomic absorption spectrophotometric method of analysis
for these metals involves aspirating the sample into a flame where the
cations are converted to ground state atoms. A light beam from a hollow
cathode lamp which emits light specific to the metal of interest is
passed through the flame, isolated by a monochromator and measured by a
photodetector. The ground state atoms of the metal of interest absorb
the light. The concentration of the metal in the sample is proportional
to the amount of light absorbed in the flame. This method is presented
in Appendix M.
4.13 DETERMINATION OF ALUMINUM, CADMIUM, COPPER, IRON, LEAD, MANGANESE
AND ZINC
Aluminum, cadmium, copper, iron, lead, manganese and zinc are measured in
precipitation samples by graphite furnace atomic absorption
spectrophotometry. Microliter quantities of sample are deposited into a
graphite tube vhere it is electro thermally dried, pyrolyzed and atomized.
-------
Section No. 4
Revision No. 1
Date July 31, 1986
Page 9 of 9
A light bean from a hollow cathode lamp which emits a spectrally pure
line source of light specific to the metal of interest is passed through
the atotts which are at ground state. The atraos of the specific metal
absorb the light. The concentration of the metal is propertinal to the
amount of light absorbed in the flame. This method can be found in
Appendix N.
4.14 REFERENCE
1. Gran, G., "Determination of the Equivalent Point in Potentiometric
Titrations," Acta Chemica Scandinavica, 4, 1950, p. 559.
------- |