United States Environmental Monitoring Systems
Environmental Protection Laboratory
Agency Research Triangle Park NC 27711
Research and Development EPA-600/4-82-042a & b March 1983
SEPA Quality Assurance
Handbook for Air Pollution
Measurement Systems:
Volume V. Manual for
Precipitation Measurement
Systems
Part I. Quality Assurance
Manual
Part II. Operations and
Maintenance Manual
-------
Parti ii March 1983
Disclaimer
This report has been reviewed by the Environmental Monitoring Systems
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or recommendation for
use.
-------
March 1983 Mi Parti
Foreword
Measurement and monitoring research efforts are designed to anticipate •
potential environmental problems, to support regulatory actions by developing an
in-depth understanding of processes that impact health and the ecology, to provide
innovative means of monitoring compliance with regulations, and to evaluate the
effectiveness of health and environmental protection efforts through the monitoring
of long-term trends. The Environmental Monitoring Systems Laboratory, Research
Triangle Park, North Carolina, is responsible for development of environmental
monitoring technology and systems; EPA-wide quality assurance programs for air
pollution measurement systems; and technical support to state agencies and to
EPA's Office of Air, Noise, and Radiation; Office of Pesticides and Toxic Substances;
and Office of Solid Waste and Emergency Response.
This quality assura nee manual has been developed to aid agencies, which plan to
measure precipitation, develop adequate quality assurance programs. This manual
covers both laboratory and field measurement systems currently used in precipita-
tion monitoring Requirements of the EPA-wide quality assurance program have
been incorporated into this manual. The main goal of the EPA quality assurance
program is to provide documented environmental monitoring data of adequate pre-
cision and accuracy
Thomas R. Mauser, Director-
Environmental Monitoring Systems Laboratory
Office of Research and Development
-------
March 1983 iv part |
Abstract
The quality assurance manual for precipitation measurement systems was
designed to assist agencies and personnel in obtaining high quality data in
monitoring and analyzing precipitation samples. Since precipitation samples
contain trace (micromolar) quantities of constituents, they are susceptable to
appreciable error due to contamination as well as chemical and biological
degradation Sections in this manual address requirements which should be
incorporated into every quality assurance project plan. The manual presents
guidelines for assessment of precipitation measurement data in terms of precision,
accuracy, representativeness, completeness, and comparability. The topics covered
are quality assurance objectives; organiation and planning; documentation; siting;
field operations; laboratory operations; data handling, validation and reporting; and
data quality assessment All or most of these guidelines and protocols should be
adapted and used for any precipitation monitoring program.
-------
Parti
March 1983
Contents
Section
1.0 Introduction
2.0 Quality Assurance Objectives
3.0 Planning and Organization
4.0 Documentation
5.0 Siting
6.0 Field Operations
7.0 Laboratory Operations
8.0 Data Handling, Validation, and Reporting
9.0 Data Quality Assessment
Appendix A Procedure Used for Hubaux and Vos
Detection Limit Calculations
Appendix B Methods Validation Study for Pb
Analysis by Flame Atomic Absorption
Pages
1
1
2
1
8
28
38
10
6
1
2
Date
1/1/81
1/1/81
1/1/81
1/1/81
1/1/81
1/1/81
1/1/81
1/1/81
1/1/81
1/1/81
1/1/81
-------
Part I vi March 1983
Acknowledgements
This quality assurance manual was prepared for the Environmental Monitoring
Systems Laboratory, Research Triangle Park, North Carolina, under the direction of
the project officer, John C. Puzak. Material in this manual has been based on
current EPA quality assurance procedures, on air and water monitoring methods,
and on procedures used in the Electric Power Research Institute (EPRI) Acid
Precipitation Study, the National Atmospheric Deposition Program (NADP), and the
Multi-State Atmospheric Power Production Pollution Study (MAP3S). The authors
extend gratef-ji appreciation to the staff at Rockwell International's Environmental
Monitoring & Services Center, Newbury Park, California, andto the many reviewers
of the drafts of this document for their contributions.
-------
Jan. 1981
Parti-Section 1.0
1.0 Introduction
Wet deposition is one of the chief
sinks for atmospheric pollutants; these
pollutants include substances benefi-
cial as well as harmful to the environ-
ment The harmful pollutants formed
primarily during dissolution and re-
moval of both gaseous and paniculate
nitrogen and sulfur oxides (NO* and
SOX) in rain and snow produce acid
precipitation with a pH less than 5.6—
the equilibrium value due to atmos-
pheric carbon dioxide (CCb) alone.
Increasing awareness of the poten-
tially harmful effects of acid precipita-
tion on soil, forest, and lake ecologies as
well as its corrosive effect on various
materials have made it the subject of
great scientific and social concern. One
result of these concerns has been an
increase in the number of programs for
monitoring acid precipitation. Monitor-
ing programs generally have one or
more of the following objectives:
1. To obtain a data base of variables
for evaluating acid precipitation
trends in remote areas generally
unaffected by pollutants.
2. To observe and to relate acid
precipitation trends throughout a
region, including urban and non-
urban areas, to meteorological or
seasonal conditions as well as to
emissions.
3. To provide a data base for evaluat-
ing effects, for developing and
evaluating abatement strategies,
and for developing and validating
models.
Because precipitation samples collected
during the monitoring contain trace
(micromolar) quantities of constituents,
they are susceptible to appreciable error
due to small changes from contamina-
tion as well as chemical and biological
degradation. Thus stringent procedures
and protocol are essential to preserve
the chemical integrity of the sample
before analysis.
Obtaining a high quality data base
requires a quality assurance program
which includes the following phases of
monitoring:
1. Organization and planning;
2. Site selection to obtain representa-
tive samples for the region;
3. Field operations;
4. Laboratory operations;
5. Data handling, validation, assess-
ment and reporting; and
6. Documentation.
Quality control and quality assurance
protocols for the above phases presented
herein can be used, or they can be
adapted for the quality assurance pro-
gram of any precipitation network. Pro-
cedures required to conduct the field
and laboratory operations are in the
companion O&M Manual (1).
The purpose of this quality assurance
manual is to provide guidelines and
protocols for operators, project officers,
and program managers carrying out a
quality assurance program in all phases
of precipitation monitoring. The mate-
rial m this manual is based primarily on
the quality assurance guidelines m the
EPRI (Electric Power Research Institute)
Acid Precipitation Study in the North-
eastern United States, the NADP
(National Atmospheric Deposition Pro-
gram), and MAP3S (Multi-State Atmos-
pheric Power Production Pollution
Study). The EPA handbooks for quality
assurance of air pollution measure-
ments (2,3) were used as guides for
format and content. To have this acid
precipitation manual stand alone with-
out requiring reference to the O&M
manual (1), some duplication of material
was necessary.
References
1. Operations and Maintenance
Manual for Precipitation Mea-
surement Systems, U.S. Environ-
mental Protection Agency, Re-
search Triangle Park, N.C., (in
press).
2. Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems - Vol. I - Principles, U.S.
Environmental Protection Agency,
Research Triangle Park, N C , EPA-
600/9-76-005 (March 1976).
3. Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems - Vol. II -Ambient Air Specific
Methods, U.S. Environmental Pro-
tection Agency, Research Triangle
Park, N.C., EPA-600/4-77-027a
(May 1977).
-------
Jan. 1981
Part I-Section 2.0
2.0 Quality Assurance Objectives
The agencywide EPA quality assur-
ance policy stipulates that every moni-
toring and measurement project must
have a written and approved quality
assurance (QA) project plan (1,2). This
requirement applies to all environ-
mental monitoring and measurement
efforts mandated or supported by EPA
through regulations, grants, contracts,
or other formal means not currently
covered. To generate and report moni-
toring data of the highest quality, no
precipitation monitoring project should
be initiated without a written, approved
QA project plan.
Each agency or laboratory generating
precipitation data has the responsibility
for implementing at least minimum
procedures which assure that precision,
accuracy, completeness, and repre-
sentativeness of the data are known and
documented. In addition, each should
specify the quality levels which data
must meet to be acceptable.
All project personnel should be
familiar with the policies and objectives
outlined in the QA project plan to assure
proper interaction between the field
operations, laboratory operations, and
data management.
2.1 QA Project Plan Elements
The QA project plan should specify
the policies, organization, objectives,
functional activities, and QA and QC
activities needed to achieve the" data
quality goals of the project The 16 items
listed below must be addressed in each
QA project plan:
1. Title page with provisions for
approval signatures,
2. Table of contents;
3. Project description;
4. Project organization and respon-
sibility;
5. QA objectives for precipitation
measurement data in terms of
precision, accuracy, complete-
ness, representativeness, and
comparability;
6. Sampling procedures;
7. Sample custody;
8. Calibration procedures and fre-
quencies;
9. Analytical procedures;
10. Data reduction, validation, and
reporting;
11. Internal quality control checks and
frequencies;
12. Performance and system audits
and frequencies;
13. Preventive maintenance proce-
dures and schedules;
14. Detailed routine procedures for
assessing data precision, accur-
acy, and completeness of specific
measurements;
15 Corrective actions; and
16. Quality assurance reports to man-
agement
All measurement results must be
representative of the media (air, precipi-
tation, etc.) and the conditions. Data
quality objectives for accuracy and
precision established for each mea-
surement parameter should be based
on prior knowledge of the measurement
system; on method or validation studies
using replicates, spikes, standards,
calibrations, recovery studies, and blind
samples, and on requirements of the
specific project.
2.2 Specific Objectives
The QA objective is to produce data
that meet the user's requirements and
the monitoring goal. The QA activities
described mthis manualaredesignedto
assist in generating data that are
complete, precise, accurate, repre-
sentative, and comparable.
Comp/eteness-\r\ general, precipita-
tion data can be considered complete if
90% or more of the total possible
observations of events are available.
The percentage of valid usable chemical
analysis data can be slightly lower due
to sample contamination m the field
before shipment and due to container
leakage in transit.
Precision and Accuracy -Interlabora-
tory studies conducted by EPA's En-
vironmental Monitoring Systems Lab-
oratory (EMSL) on water (not precipita-
tion) systems indicate about ±10%
precision and accuracy for conductivity
and ±0 1 pH unit. Precision and
accuracy for chemical analysis are
determined separately by each labora-
tory for each analytical technique and
range
Representativeness - Data must be
representative of the conditions being
measured This is addressed initially in
the siting criteria (Section 5.0) and later
in discussions of continuous observa-
tions of land use and development
which can impact the data generated by
the monitoring site.
Comparability - Data reported by
different networks should be in con-
sistent units to allow for data compari-
son; the recommended units are:
Sample weight gm
Precipitation mm
pH pH units
Conductivity micromho/cm or
microS/cm
Concentration micromoles/hter
Implementation of these QA objec-
tives will produce data that are complete,
accurate, precise, representative and
comparable. Attainment of these objec-
tives requires involvement of the QA
coordinator in the planning stages of the
measurement project. This allows the
identification of areas with potentially
large negative impact on data quality
and provides a mechanism for instituting
a system of internal QC, corrective
action, data validation, and external
assessment of precision a.nd accuracy
to avoid such weak areas. Each of the
objectives is discussed extensively in
the following sections.
2.3 References
1. D.M. Costle, Administrator's Mem-
orandum, EPA Quality Assurance
Policy Statement, U.S. Environ-
mental Protection Agency, Wash-
ington, D.C., May 30, 1979.
2 D.M Costle, Adminstrator's Policy
Statement, Quality Assurance
Requirements for all EPA Extra-
mural Projects Involving Environ-
mental Measurements, U.S. En-
vironmental Protection Agency,
Washington, D.C., June 14, 1979
3. U.S. EPA Quality Assurance Man-
agement Staff, "Interim Guide-
lines and Specifications for Prepar-
ing Quality Assurance Project
Plans", QAMS-005/80, U.S. En-
vironmental Protection Agency,
Washington, D.C., December 29,
1980
-------
Jan. 1981
Parti-Section 3.0
3.0 Planning and Organization
The agencywide EPA policy requires
participation by all EPA regional offices,
EPA program offices, EPA laboratories,
and States in a centrally managed QA
program, and the policy includes all
monitoring and measurement efforts.
Accordingly, a QA program plan pre-
pared by each program office, regional
office, and laboratory should include the
overall policies, organization, objec-
tives, and functional responsibilities
designed to achieve data quality goals
for monitoring and measurement activ-
ities. The 11 elements required in any
QA program plan (1) are:
Identification of organization;
Description of the organization re-
sponsibilities;
QA policy statement;
QA management;
Personnel qualifications;
Facilities, equipment and services;
Data generation;
Data processing;
Data quality assessment;
Corrective actions; and
Implementation requirements and
schedules.
Monitoring for precipitation chemical
analysis requires strong interfaces
between field operations, laboratory
operations, and data management.
Quality assurance management should
be a separate function in the organiza-
tion.
3.1 Personnel Assignments
and Responsibilities
3.1.1 Organization
A precipitation monitoring network
involves interdependent field monitor-
ing and laboratory operations. Each
operation has its own QA aspects. The
field monitoring sites and the laboratory
can be run by independent agencies or
organizations; however, the results are
reported to and are the responsibility of
the program manager. Reporting direct-
ly to the program manager are the field
manager and the laboratory manager or
supervisor. The QA coordinator should
also report directly to the program
manager. Brief discussions of the
qualifications and duties of the program
personnel are presented below.
3.1.2 Program Management
A program manager is required to
assure that high quality data are
generated within the time and funding
constraints of the program. The man-
ager must keep abreast of all develop-
ments, make necessary decisions,
review data and QA reports, issue
progress reports, and interpret the
results. The program manager should
have a degree in chemistry and some
experience in managing projects. Full
time employment is recommended. This
position can be combined with that of
the field manager
3.1.3 QA Coordination
A QA coordinator should report
directly to the program manager with
inputs into the other program functions
and with reports on the quality of the
data generated and on corrective
actions that need to be taken
3.1.4 Field Operations
The personnel needed to carry out the
field duties in a precipitation monitoring
network include a field manager and
station operators.
Field Manager - The field manager
may be a member of the agency or
organization operating the stations or a
member of the central laboratory staff;
should have a college degree, preferably
in chemistry; should be familiar with all
the procedures; and should have
experience in the operation of all
equipment. The duties are to solve field
problems, to notify the program man-
ager of such problems, to oversee, and
to tram, or coordinate the operators.
Station Operators - The operator's
duties at the station require only his part
time presence. 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
highquahtydata Ashort-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 O&M manual
(2), which will be given to all personnel;
the O&M manual contains the pro-
cedures necessary to generate a high
quality data base and a schedule of
tasks. If, at any time, an operator's
performance deteriorates, additional
training must be provided by the field
manager as soon as possible, by a
refresher course or by on-site guidance
(Section 6.7).
3.1.5 Laboratory Operations
Each analytical laboratory should
have at least the following types of
employees (1) director, (2) supervisor,
(3) analyst and (4) QC chemist. Informa-
tion on requirements, training, and
supervision of these people is presented
below.
Laboratory Director - This position is
recommended, but not required if the
laboratory has a supervisor; it may be
combined with the supervisor position
listed below. Either a laboratory director
or a laboratory supervisor is required.
The director should have a minimum of
one year's laboratory analytical experi-
ence in water/wastewater related
measurements; should have a degree in
chemistry; and should be employed full
time.
Laboratory Supervisor - This position
is not required if there is a laboratory
director, but it is recommended that a
laboratory supervisor be employed full
time, and have similar academic train-
ing and experience to that specified for
the laboratory director.
Laboratory Analyst - This person
should be employed full time, and
trained to perform with some supervi-
sion all routine chemical measure-
ments on water samples Academic
training should be completion of at least
1 year of college chemistry and/or a
laboratory-oriented vocational course; 1
year of experience in water/waste-
water analysis is recommended, but if
this experience is lacking, a minimum of
30 days of on-the-job training in
measurements performed by the agency
is highly recommended. The analyst
must be supervised by an experienced
professional scientist—the laboratory
director, the supervisor or a similarly
trained individual.
Quality Control Chemist - This re-
sponsibility is not necessarily a full time
position, it may be part time supple-
mented by other program duties. The
percentage of the chemist's time
dedicated to QC is dictated by the size
and complexity of the program.
This individual should have a mini-
mum of a bachelor's degree in chemis-
try, engineering, or mathematics with at
least 2 years of environmental or 1 year
of QC experience. This position is under
the general supervision of the labora-
tory director, but with access to the
program manager. Duties are to imple-
ment and monitor the routine applica-
tion of QC activities in the laboratory.
-------
Part l-Section 3.0
Jan. 1981
The individual can also participate in a
formal program to tram new employees
and to update skills of older employees.
Before analysts are allowed to ana-
lyze 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 (Section 4 3.3) Analyst perform-
ance can be evaluated by control charts
of critical QC parameters (Sections 5 5
and 5.8); if these charts indicate a
problem, the analyst should be given
further training.
3.1.6 Data Management
People involved in data acquisition,
reduction, and reporting are the field
operator, the analyst, the data entry
staff, the laboratory director, the
program manager, and the designated
QA coordinator. Presented here is an
outline of primary duties assigned to
different members of the data manage-
ment team-
Field Operator - Preparation of field
data forms, and QC of data form
preparation
Analyst - Reading or transcription of
strip charts, and entry of data to
computer
Laboratory Supervisor - QC check of
strip chart reading, preparation of data
forms, QC of data form preparation, and
review of computer-generated QC
information or control chart preparation
and interpretation.
Data Processing Personnel - Input of
data from data sheets, verification of
input for keypunch errors, and update of
computer files.
QA Coordinator (Officer) - Review of
data, preparation of QA reports and
submittal of audit data, and recommen-
dations to project manager.
3.2 Designation of QA Re-
sponsibilities and Duties
3.2.1 Quality Assurance Coordina-
tor
The QA coordinator should monitor
performance as follows.
1. Review the monthly QC plots
generated for each analysis to
verify that QC data are acceptable
and to identify any consistent bias
trend.
2. Evaluate the Monthly Field Audit
Report prepared by the QC chemist
to assess the accuracy of field
analyses and to identify needs for
corrective actions
3. Review all of QC information
presented with each set of analyti-
cal data to be reported — including
data from AC reports, ion summa-
tion values, and QC chemist's
independent internal audit report.
4. Evaluate laboratory and field
operations by conducting system
audits and reporting them to the
project manager.
5 Prepare quarterly reports to man-
agement that summarize QA activ-
ities and assess data quality in
terms of precision and accuracy
trends for both the field and the
laboratory operations.
3.2.2 The QC Chemist, the Analyst,
and the Data Clerk
Besides QA, routine QC activities are
carried out in laboratory operations by
the QC chemist, the analystand thedata
clerk
The QC chemist introduces blind
audit samples as an independent check
on data quality, and issues a monthly
report updating control limits for all
parameters. All data given to the
laboratory director should have been
evaluated by the analyst's supervisor
and the QC chemist
The analyst performs the analyses
and evaluates analytical performance in
real time, using readily available QC
information. As soon as possible after
analysis, the supervisor evaluates the
computer-generated QC information for
all QC parameters; reanalyzes the data
if necessary; and finally, when data are
"in control" (according to the supervisor)
releases the data for reporting to the
laboratory supervisor or director, who
submits them to the project manager
The data clerk inputs analytical and
field data into the computer, checks and
corrects the data input, and generates
reports and graphs of QC information.
3.3 References
1. U.S. EPA Quality Assurance Man-
agement Staff, Guidelines and
Specif/cations for Preparing QA
Program Plans, QAMS-004/80,
U S. Environmental Protection
Agency, Washington, D.C, Sep-
tember 20, 1980.
2. Operations and Maintenance Man-
ual for Precipitation Measurement
Systems, U.S. Environmental Pro-
tection Agency, Research Triangle
Park, N.C., (in pres.s).
-------
Jan. 1981
Part l-Section 4.0
4.0 Documentation
4.1 Document Control
A system of document control should
be established for documentation of all
precipitation monitoring field and
laboratory operations — siting, samp-
ling, analysis, QA data handling, and
validation. The system should be
patterned after the indexing format of
the QA handbook (7) for updating
operational procedures and adding
results of special studies and other
related documents. The indexing format
at the top of each page should include:
Section number.
Revision number
Date of revision
Page number.
The elements of a precipitation mea-
surement project for document control
should include:
1. Field operations and maintenance
procedures,
2. Analysis procedures,
3. Auditing procedures,
4. Computational and data validation
procedures,
5. Quality assurance plan, and
6. Quality assurance manual.
4.2 Internal Documentation
A central file of all data, reports,
correspondence, and so forth should be
maintained by the project manager. In
addition, a data file should be kept by the
laboratory. Records in the files should
meet the following requirements:
1 Records should have identification
numbers and must be kept for at
least 3 years by the agency in an
orderly, accessible form; records
should include all raw data, calcu-
lations, QC data, and reports.
2. Data in laboratory records must
include the following information:
A. Sample identification number,
B. Sample type,
C. Date sample received in labora-
tory,
D. Collection data (time, date,
volume, and so forth, if labora-
tory responsibility).
E. Date of analysis,
F. Name of analyst,
G. Results of analysis (including
all raw data), and
H. Name of person receiving the
analytical data.
3. If applicable, the laboratory should
follow chain-of-custody proce-
dures from receipt of sample
through completion of analysis;
the following are guidelines:
A. Computer printouts or report
forms verified against labora-
tory records before data re-
lease,
B. Bound notebooks with num-
bered pages,
C. Sampling information records
(e.g., field data forms) with
dates, time, site location, sample
amount, and so forth.
D. An example of the data hand-
ling and reduction system ex-
amined by legal counsel to
determine soundness in possi-
ble litigation.
4.3 Reports
Each organization must periodically
assess its QA program (2). The QA
project planprovidesthe mechanism for
reporting to management on QA activi-
ties, on the performance of measure-
ment systems and on data quality. As a
minimum these reports should be
submitted semiannual^ and they should
include (3).
Periodic assessments of measurement
data accuracy, precision, and complete-
ness;
Results of performance audits;
Results of system audits; and
Significant QA problems with recom-
mended solutions.
The designated QA coordinator should
be responsible for the reports, and
should provide a separate QA section to
the final report to summarize data
quality information contained in the
periodic reports.
4.4 References
1. Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems -Vol. /-Principles, U.S. En-
vironmental Protection Agency,
Research Triangle Park, N.C., EPA-
600/9-76-005 (March 1976).
2. U.S. EPA Quality Assurance Man-
agement Staff, Guidelines and
Specifications for Preparing Quality
Assurance Program Plans, QAMS-
004/80, U.S Environmental Pro-
tection Agency, Washington, D.C.,
September 20, 1980.
3. U.S. EPA Quality Assurance Man-
agement Staff, Interim Guidelines
and Specifications for Preparing
Quality Assurance Project Plans,
QAMS-005/80, U.S. Environ-
mental Protection Agency, Wash-
ington, D.C , December 29, 1980.
-------
Jan. 1981
Part l-Section 5.0
5.0 Siting
Criteria for site selection and evalua-
tion are similar to those proposed by the
World Meteorological Organization (1)
In the design of a monitoring program,
the program objectives and network
station density must be considered
Programs can be grouped as baseline
monitoring, long-term trend monitoring
and research. Stations may be charac-
terized as remote, rural, and urban,
however, they should yield representa-
tive results for the region although this
characterization is somewhat over-
simplified Precipitation composition,
long-range transport of acid precipita-
tion and precipitation comtammants,
and the impact of a source shown by its
distance or time scale are not well
known Many large point sources are in
rural areas. Thus, choosing optimum
site locations to monitor either the
impact of a specific source or regional
effects not related to any source is
difficult Instead of remote, regional,
urban, and so forth, a complete descrip-
tion of the site and its surroundings
could be given so the user of the data
could decide what the station is measur-
ing or what it represents However, to
aid the decision and for convenience,
site area categories have been used. In
all categories, the criteria are of prime
importance If no site in an area meets
the criteria, a site in an alternate area
must be selected
Network station density, which helps
define the spatial and temporal resolu-
tions of the data obtained, is determined
by the program objectives, the area
meteorology and topography, and the
budgetary constraints
5.1 Network Design Con-
siderations
In designing a precipitation monitor-
ing network, stations are located either
singly or in combination, 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 pollutantemission sources)
on precipitation
These three are generally differentiated
by concentration levels. The background
or remote station should show con-
tamination primarily due to natural
processes; the regional station would be
affected primarily by long-range trans-
port; and the urban, site would show
high concentrations due to a polluted
local environment To select station
locations, it is necessary to have
detailed information on the locations of
emission sources, the regional variabil-
ities of ambient pollutant concentra-
tions, the precipitation amounts, the
prevailing winds, and other meteorolog-
ical data Thus, the design of a network
needs to address details such as the
number, location, and type of sampling
stations and equipment to be used. Due
to the variation m acid precipitation
caused by terrain, meteorological
conditions and-demographic features,
each network should be designed
individually after considenmg physical
evidence, economic factors, and pro-
gram objectives
5.2 Site Selection Criteria
Selecting a precipitation monitoring
site is as important as selecting a
measurement technique to obtain
representative data The variabilities
and the long-range transports of pol-
lutants make it difficult to determine
whether a site is collecting precipitation
data representative of an area The
station should collect samples repre-
sentative of both the amount and the
composition of the precipitation in the
area
The transport and diffusion of air
pollutants and their resultant concentra-
tions in precipitation are complicated by
topography; minor topographical fea-
tures may exert small effects, but major
features such as valleys or mountain
ranges can affect large areas In
mountainous regions, precipitation
tends to be unevenly distributed due to
topographical lifting of clouds and
deflecting of airflows; such unevenness
should be recognized for site selection.
Near a large body of water, sea or land
breeze motions may change direction;
in the daytime, the winds may come
from the water, but at night they may
come from land. Because of these
changes, area sampling representation
would be affected In such areas, more
than one station should be installed but
the number of sites should depend on
site availability and on funding level
To optimize site locations for different
station categories, the following selec-
tion criteria are listed in order of
importance In the baseline and regional
categories, the site must meet the first
three as well as all those in Section 5 3
5.2.1 Baseline Scale
The station should be in a location
where the effects of human activities
are negligible It is difficult to find
locations wh ich meet all of the following
criteria (2,3,4) but each should meet the
following as closely as possible:
1 The station should be in an area
where no significant changes in
land-use practices within a 100 to
1000 km (depending on prevailing
wind direct ion) from the station are
anticipated during the study period.
2 The station should be far away
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 phenom-
ena such as forest fires, dust and
sand storms, or volcanic activities
4 The site should have provisions for
setting up a meteorological and
aerometric monitoring station
(2,3,4,5).
5. The site should be readily accessi-
ble on a flat or gently sloping
terrain (less than 20°), and shel-
tered from strong winds
5.2.2 Regional Scale
Site selection criteria for a regional
monitoring network are (1,6,7,8,9):
1. The general area should be free
from influences of large anthro-
pogenic 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 effect the
precipitation chemistry. If, the site
location must be near a large
source (e.g , within 50 km), the
station should be in the prevailing
upwind direction from the source.
2. The immediate area of the site
should be separated from local
sources such as houses, active
farmlands or orchards, marshes
and swamps, landfills, roads; such
sources should be several kilo-
meters distant, and preferably
downwind.
3. If stations are near pollutant
sources, the site location should
-------
Part l-Section 5.0
Jan. 1981
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
5 The site should have provisions for
setting up meteorological and
aerometric monitoring equipment
(2,3,4,5).
6. The site should be readily accessi-
ble on a flat or gently sloping
terrain (less than 20°), and shel-
tered from strong winds.
The selected site should be evaluated
for representativeness and for local
contaminations by installing a temporary
grid of neighboring satellite samplers
around it; sampling procedures for this
temporary network should be compara-
ble to those in the original site, data
collection by the network is recom-
mended for a few months, not to exceed
1 year. The permanent site should be
selected after evaluating the site's
representativeness by analyzing data
from stations in the temporary network.
For evaluating the effects of long-
range transport on acid precipitation,
the average interstation distance can be
several to several hundred kilometers
5.2.3 Urban Scale (or Local Scale)
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 must not be near other sources
or at sites that do not follow the criteria
(Section 5.3). The network station
density depends on the objective of the
monitoring program and on budgetary
considerations For studying the local
effect due to a single point or area
source, the average interstation dist-
ance should be of the order of kilo-
meters. However, actual station density
and interstation distance should be
decided by the desired spatial resolu-
tions. The following criteria apply:
1. If an area is characterized by a type
of land use, the site can be near
such pollution sources.
2. The site should have provisions for
setting up meteorological and
aerometric monitoring equipment
(2,3,4,5).
3 The site should be readily accessi-
ble on a flat or gently sloping terrain
(less than approximately 20°), and
sheltered from the strong winds.
5.3 Sampler and Rain Gauge
Siting Criteria
Placements of precipitation samplers
and ram gauges should assure the
adequacy of the site to collect unbiased
samples Samplers and ram gauges
should stand far enough from trees,
hills, overhead power lines and other
obstructions to minimize interference
with sampling. Natural and manmade
obstructions may cause turbulence
and/or contamination, and may cause
nonrepresentative samples; thus no
object (even if smaller than the collec-
tor) should be within a few meters of the
collector, and no object should cast a
rain shadow on the collector. An open,
flat, grassy area, surrounded by trees at
least 100 m away, but near no sources
(unless a local source is to be studied)
would be an ideal site.
Criteria for placement of collectors
(both 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 viewing
angle of an obstruction from the
collector should be less than 30°
above the horizon
2 The horizontal distance between
an obstruction and a collector of
comparable height should be at
least one unit height of the taller of
the two.
3 The collector should be far from
mobile pollution sources; routine
air, ground, or water traffic should
not come within 100 m of the
collector site.
4. The distance between any over-
head wires and the site must be far
enough for the samples not be
affected; thus, criterion 1 (above)
should assure that there are no
overhead wires.
5. The collector should be at least
100m from open storage of agri-
cultural products, fuels, or other
foreign materials.
6. The ground surface around the
collector should not be loose soil; it
should have a grass cover or gravel
to minimize splash and airborne
ground surface particulates that
contaminate samples.
7. Wet/dry collectors should be
oriented parallel to the prevailing
wind direction 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 posi-
tioned parallel to the collector and
the prevailing wind direction. If the
gauge has an access door (to a
recorder, weighing or drive mech-
anism), the door should be kept
closed and the gauge should be
mounted with the door facing
away from the wind to prevent
precipitation entering the com-
partment.
The distance between certain ob-
struction (e g., growing trees and newly
erected structures) and the collector
should be checked periodically Criteria
2 should minimize the sampling area
required when more than one collector
or when a sampler and ram gauge are
on the site, and should minimize the
effect of turbulence caused by obstruc-
tions.
5.4
tion
Site Evaluation Descrip-
5.4.1 Station Identification
All stations must be identified by
documentation of site characteristics to
facilitate evaluation of data generated at
that site and interpretation of the
monitoring data. Typically, the site.
identification record should contain the
following:
1. Data acquisition objective (base-
line, trend, or research monitoring)
2 Station location (address, map
coordinates, elevation, etc.)
3 Scale of representativeness (re-
mote, regional, or area type; i.e.,
industrial, agricultural, forest,
urban, etc.)
4. Instrumentation checklist (manu-
facturer, model number, measure-
ment technique, etc.)
5 Important pollutant sources (point
and area sources, their emission
concentrations, proximities, pol-
lutants, etc.)
6 Topography description (trees,
hills, valleys, bodies of water; type,
size, proximity, orientation, etc.);
photographs of the monitoring site
covering a 360° view from the
precipitation collector are recom-
mended).
7. Site diagram properly scaled (equip-
ment configuration, vegetation,
manmade structures, access road,
electrical powerlmes, etc.).
The site identification should be docu-
mented by filling out a Site Description
Report (Section 5.5).
5.4.2 Stat ion Classification and Eval-
uation
Each station should be initially
classified by using the information in
the Site Description Report Section 5.5
gives a modified version of one devel-
oped by NADP(8). The five site classes
defined depend on the sophistication of
instrumentation in the station as well as
on the satisfaction of siting criteria.
Class I
1. Station satisfies all siting criteria
(Section 5.3),
2. On-site instrumentation includes
automatic precipitation collector,
a recording rain gauge,
pH and conductivity meters,
-------
Jan. 1981
Part l-Section 5.0
meteorological sensors (wind-
speed and direction), and aero-
metric analyzers (SO2 and NO/
NO,).
C/ass II
1. Station satisfies all siting criteria
(Section 5.3).
2. On-site instrumentation includes
automatic precipitation collector, a
recording rain gauge, pH and con-
ductivity meters, meteorological
sensors (wind-speed and direction),
and aerometricanalyzers (S02 and
NO/NO,)
Class II
1. Station satisfies all siting criteria
(Section 5.3).
2. On-site instrumentation includes
automatic precipitation collector,
a recording ram gauge, and pH
and conductivity meters.
Class III
1. Station satisfies all siting criteria
(Section 5.3).
2. On-site instrumentation includes
automatic precipitation collector,
nonrecording ram 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.
C/ass V
1 Station does not satisfy all siting
criteria (Section 5.3)
2. On-site instrumentation identical
to Class II stations.
5.5 Site Description Report
All monitoring stations should be
properly identified and classified using
the report forms on the next pages.
After initial classification, an on-site
visit should be made by the QA coordi-
nator 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 project manager (or a
designee) must be sure that siting
deficiencies are corrected within a
reasonable time. Reasonableness de-
pends on the deficiencies found; most
should be corrected within 30 days but
for deficiencies requiring longer time, a
schedule must be established for
compliance attainment. When correc-
tions are made, documentation should
be provided to the QA coordinator and
program manager and the station
classification should be changed by
them.
All sites should be evaluated again
once a year to establish that they
maintain compliance. All aerometric
and meteorological instrumentation
should conform to standard ambient
monitoring guidelines (4,10).
-------
Part l-Section 5.0
Jan. 1981
A. Data Acquisition Objective (Description)
B Site Category
1. Station Identification
4 Latitude*
7 Station environment. Remote _
Suburban
Industrial
5 Longitude'"
2 County.
Rural.
Urban
3 State
6 Elevation
(m)
Commercial _
8. Name of official
9. Mailing address _
position
(number and street)
10. Phone (
(city)
(state)
(zip)
C. Instrumentation
1. Precipitation Collector Type: Automatic
Diameter (I. D) of
Sample Bucket _
Manufacturer
Model
Non-automatic.
2. Ramgauge:
3. /VOX Monitor.
4. S02 Monitor:
5. Other Aerometric:
1 Recording
2 Type: Weighing
3. Manufacturer _
4. Funnel Size:
Tipping Bucket.
(cm) Serial No.
Non-recording _
Other.
Model
(cm) Serial No.
1. Recording
2. Type: Chemiluminescent
3. Manufacturer
-Non - recording
..Other
1. Recording
2. Type: Fluorescent
3. Manufacturer
Model/Ser. No.
Non -rec or ding
Other
1'. Sensor:
Non -recording
2. Type
Model/Ser. No.
Recording
3. Manufacturer
Serial No.
Model
6. Other Meteorological Instrument:
Serial No.
2.
3
4.
7. pH Meter:
1.
2.
8. Conductivity Meter:
1.
2.
Wind Direction Sensor:
Temperature Sensor:
Solar Radiation Sensor:
Type
Manufacturer
Type
Manufacturer
Manufacturer
Type
Manufacturer
Serial Nn
Type
Manufacturer
Serial Nn
Type
Manufacturer
Temp.
Compensated:
Mnrlf
Temp.
Compensated:
Mnfif
Mnrtvl
Mnrlfil
Mnrtel
Serial Nn /Model
>l/Ser Mr,
•l/Sfir Nn
*To be reported in xx.yy.zz format corresponding to deg., min., sec.
-------
Jan. 1981 6 Part I-Section 5.0
D Site Documentation:
1. Local topographic map (Scale
1:250,000)
Identify the site location and
major sources on the map.
-------
Jan. 1981
Part l-Section 5.0
2. Sketch a map to document the
environment within a 1/2 mile
radius of the site. Include the
following information on the draw-
ing where applicable.
Site Diagram and Equipment Con-
figuration at Center of Drawing
Roadways with Names (paved and
unpaved)
Parking Areas (paved and unpaved)
Stationary Sources (NEDS#)
Buildings (number of stories)
Undeveloped Land (ground cover)
Tree Lines or Clusters
High Power Lines
Topographical Features
(Valleys, Hills etc )
Bodies of Water
North Direction
-------
Part l-Section 6.0
Jan. 1981
3. Site photographs, labelled to in-
dicate the four compass directions.
-------
Jan. 1981
Part l-Section 5.0
5.6 References
1. WMO Operational Manual for
Sampling and Analysis Tech-
niques for Chemical Constituents
in Air and Precipitation, World
Meteorological Organization Pub.
No 299(1974).
2. Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems, Vol. II - Ambient Air Specific
Methods, EPA-600/4-77-027a,
ResearchTriangle Park, NC (1977).
3. Guide to Meteorological Instru-
ment and Observing Practices,
World Meteorological Organiza-
tion Pub. No. 8, TP8 (1971).
4. Ambient Monitoring Guidelines
for Prevention of Significant Deteri-
oration (PSD), EPA-450/1-78-019
(1978).
5 Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems, Vol. I - Principles, EPA-
600/9-76-005, Research Triangle
Park, NC (1976).
6. Galloway, J.N., C. Hall, and G E.
Likens, The Collection of Precipi-
tation for Chemical Analysis,
TellusSO, 71 (1978).
7 Granat, L, Principles in Network
Design for Precipitation Chemistry
Measurements, Proc. Symp. on
Atmospheric Contribution to the
Chemistry of Lake Waters, J. Great
Lakes Res. 2(1), 42(1976).
8. Site Selection and Certification,
National Atmospheric Deposition
Program (1970).
9. Bowersox, V.C., Acid Precipitation
at a Rural Central Pennsylvania
Site, MS Thesis, Dept. of Mete-
orology, Pennsylvania State Uni-
versity (1980).
10. EPA Ambient Air Quality Surveil-
lance Regulations, Code of Federal
Regulations, Title 40 Part 58 (May
10, 1979).
-------
Jan. 1981
Part l-Section 6.0
The primary aims of quality assurance
requirements for field operations are to
maximize sample collection and to
collect and handle samples in a way that
preserves their integrity and identity.
Precipitation samples are very dilute,
and thus very susceptible to contami-
nation. The goal of this section is to
present requisites for obtaining high
quality data; discussed are facilities,
method selection, acceptance testing,
equipment installation and operation,
sampling methodology, field measure-
ments, network evaluation, and docu-
mentation
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 1.5 micromho/
cm), and a sink or drain If deionized or
distilled water cannot be produced at
the site, it can be purchased at local
supermarkets or drug stores. A refriger-
ator and a 110v AC outlet are highly
desirable; the former is necessary for
event or sequential sampling to pre-
serve the samples until they are
shipped. For weekly sampling, the
samples should be shipped within 24 h
of collection so refrigeration is not
important. Since samples can be at
ambient temperature in the collector for
as long as 1 week, another few hours
after collection should not matter
The precipitation collector and the
recording rain gauge can be run on
either 12v DC by storage batteries or
110v AC. Both means have advantages
and disadvantages With battenesthere
can be no loss of operation due to a
power failure, and sites can be located
in areas with no power. However,
batteries must be periodically re-
charged, and power output drops in cold
weather can result in collector down-
time. Furthermore, if large current
usage is required (e g., for heating),
batteries are not recommended.
6.2 Method Selection
Similar equipment and supplies are
required at a station for event, daily, or
, weekly sampling, weekly sampling is
most common atpresent. For sequential
sampling, a different type of collector is
needed. For event, daily, or sequential
sampling, a refrigerator, polyethylene
bottles with caps, polyfoam insulated
shipping containers, andfreeze-gel cold
packs are recommended for sample
storage and shipment. A list of equip-
6.0 Field Operations
ment for typical weekly sampling is in
Table 6-1, for other types of sampling,
the list should be changed to include a
different type of collector, a more
sensitive balance (2.6 kg) capacity,
polyethylene bottles and caps, and
means for shipping the sample in a cold
state.
6.2.1 Precipitation Collectors and
Ram Gauges
The ram gauge and the precipitation
collector serve different functions. The
ram gauge measures the amount of
precipitation. The precipitation collector
collects a sample for chemical analysis
The two devices are not interchange-
able.
6.2.1.1 Precipitation Collectors—The
precipitation collectors must meet the
following criteria
1. Reliable automatic operation -
collector container opens at start
of precipitation and closes after
event ends
2. Prevention of contamination of
wet sample by dry deposition.
3. Minimization of evaporation
4 Container inertness to sample
constituents of interest
Collectors, which adequately meet
these criteria, are available. For event,
daily, or longer-time rain sampling, 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 less
efficient for snow collection.
The first three criteria are met by
means of a precipitation sensor and a
motor-driven tightfittmg 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 goes on to
melt snow or ice (on the sensor) when
the temperature is below 2°C, and the
other is activated when the lid lifts off
the sample bucket to heat the sensor to
about 55°C and 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 will
wobble and allow some contamination
to enter the sample bucket
To ensure inertness to major con-
stituents m acid precipitation, use
polyethylene sample buckets (1,2). Use
only plastic (or Teflon) containers for
inorganic constituents. For cost, dura-
bility, and availability, high-density
linear polyethylene containers are
generally employed. Glass or metal can
affect inorganic sample integrity, but
either should be used if organics are
studied
For subevent or sequential samplers,
the same requirements as above hold,
but mixing between sequential samples
should be minimized because these
samplers separate samples on either a
volume or time-of-collection basis. 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 cor-
responding 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 auto-
mated and electrically operated designs
(3-8). One of the latter is available
commercially (8).
Table 6-1. Field Equipment List for Weekly Sampling
Automatic precipitation sampler (Aerochem Metrics #301)
Collection containers (3.5 galj for collector
Recording rain gauge with event marker
Rain gauge mount
pH meter, electrode
Buffers, pH 3.0, 4.0, 6.0, 7.0 and 8.0 (1 liter each)
Conductivity meter and cell
Standard KCI solution, 74 micromhos/cm (500 ml)
Temperature probe
Balance (20 kg capacity) or graduated cylinder (2 liter)
Set of attachment weights for balance (1,2,2,5,10 kg)
No. /site
1
3
1
1
1
1
1
1
1
1
1
-------
Part l-Section 6.0
Jan. 1981
6.2.1.2 Ram Gauges—In case of
collector malfunction and to reference
all the precipitation amounts against a
standard, a ram gauge is used to record
the quantity of precipitation Recording
ram 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.) and they should be accurate to
a few percent Weighing gauges mea-
sure within about ±0 76 mm (0.03 in ),
but their accuracy of about 1% of full
scale is independent of precipitation
rate. The generally accepted accuracy
for tipping bucket gauges is 1% for
precipitation rates of 25 mm/h (1 in/h)
or less, 4% for 75 mm/h (3 in/h), and
6% up to 150 mm/h (6 m/h); rates are
either measured directly or derived from
the cumulative data. Data from the
tipping bucket are amenable to tele-
metry
The recording 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 and
the pen falls to its baseline position To
prevent an event marker pen from
interfering with a sample 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 not to use
the incorrect beginning or endmgtimes.
Since the operator is seldom present to
observe collector behavior during an
event, the event pen marker is invalua-
ble in indicating malfunction.
6.2.2 pH and Conductivity Apparatus.
Temperature Probe
The pH and conductivity of a 20 ml
aliquot of the sample should be mea-
sured at the station as soon as possible
after the sampler is collected and after it
has reached the same temperature as
the calibration solutions. The sample
and the aliquot should be protected from
contamination during this time.
6.2.2.1 pH Apparatus—The pH meter
and electrode must be capable of
measuring within ±003 unit precision
with an accuracy of 0.05 unit. Meters
should have an impedance of at least
1011 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 wash-
ings than 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 in a pH 4 or 7
buffer solution, as recommended by the
manufacturer, before it is used The
electrode should be stored m pure
water, m a 74 micromho/cm KCI
conductivity standard, or in a 0 0001 N
acid solution, these storage media may
cause the electrode to require longer
times for stabilization than concen-
trated buffer solutions do, but they
increase the life of the electrode.
6.2.2.2 Conductivity Apparatus—The
conductivity meter and cell must have a
measurement range of 0 to 1000
micromho/cm, a precision of ±0 5% of
range, and an accuracy of ±1.0% of
range The range most frequently used
is 10-100 micromho/cm A tempera-
ture-compensated cell with a cell
constant of 1.0 is preferred.
6.2.2.3 Temperature—a thermistor,
thermocouple, or thermometer can be
used to measure temperature. The
temperature probe must have an
accuracy of at least 1 °C and a precision
of ±0 5°C.
6.2.3 Balance or Graduated Cylinders
The amount of precipitation sample
collected can be measured with a
balance or with graduates. Since the
density of ram samples is approximately
1.0 g/ml at 20°C, the weight of the
sample can betaken toequal its volume.
The measurement of sample volume by
graduates increases the chance of
contamination, so use of a balance is
recommended
The p-recipitation volume can be
compared to that recorded by the ram
gauge (e g , with the Aerochem Metrics
collector, 16 g of sample = 0.013 in. =
0.25 mm) to calculate the collection
efficiency of the sampler (Section 8.5.4).
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
must have a capacity of 20 kg and a
precision of at least ±10 g. Triple beam
balances meeting these requirements
are readily available. The balance
should be kept on a sturdy level table,
and it should be zeroed 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
Graduates are not recommended, but
if they are used, they should be plastic,
and they should be cleaned and dried
before use .To measure within ±10 ml,
the graduate should not be more than
1000 ml in capacity. For large samples,
several graduates or one graduate plus
a clean auxiliary container (in which to
empty the graduate) will be needed The
graduate should be checked for accur-
acy in the laboratory by weighing
various volumes of water from the
vessel
6.3 Acceptance Testing
All precipitation collectors, rain
gauges, pH and conductivity meters,
and electrodes should be tested for
acceptance before they are used m the
field. Acceptance tests should cover the
functions of the instruments. Collectors
and rain gauges should be tested on
site However, it is more convenient to
test meters and electrodes in a central
support laboratory, where common
standards and procedures would be
used for all instrumentation in the
monitoring network. General pro-
cedures for acceptance testing are
detailed below; procedures for carrying
out these tests are m 0 & M manual (9).
6.3.1 Precipitation Collectors and
Rain Gauges
Collector tests should include: (a)
sensor heating and actuating the lid
when the sensor is shorted with water
drops, (b) sensor cooling and return of
the lid after removal of the shorting
material (water wiped dry), (c) sensor
temperature attainment when lid is off
of the wet bucket (50° to 60°C), (d)
sensor temperature when ambient
temperature is below freezing and (e)
observation of lid cycling and sealing. If
any of these tests indicate a malfunc-
tion, contact the manufacturer. For test
c above, the temperature can be varied
by a potentiometer identified by TH in
the sensor circuit. In general, the
malfunctions can "be rectified by re-
placing the sensor or the motor box.
Rain gauge tests should include a
check of: (a) sensitivity and accuracy, (b)
clock function, and (c) pen and recorder
functions For test a, add an amount of
water equivalent to 0.02 in. (0.51 mm),
and measure the response. For the
Belfort recording rain gauge (5-780
series), 0.01 in. =8.2gm.Thegaugecan
be calibrated at each inch level with a
set of weights from the manufacturer or
with known weights of water.
6.3.2. pH and Conductivity Meters
All pH and conductivity meters,
electrodes, and cells should be tested in
a laboratory for acceptance before being
shipped to the field stations. The meters
usually have a serial number affixed,
but the electrodes do not, so an
identification number should.be affixed
to each electrode. For each of the tests
indicated below, a total of 10 solutions
is measured, and an average value and
a standard deviation are calculated.
-------
Jan. 1981
Part l-Section 6.0
6.3.2.1 Evaluation of Conductivity
Meter and Cell—The conductivity meter
and cell are acceptable if the average
value for a 0.0003M KCI solution is
within 2% of the accepted value, 44.6
micromho/cm at 25°C, and if the
standard deviation is less than 2% The
results should be recorded on the
Conductivity Meter/Cell Acceptance
Test form and on the Conductivity Test
Summary form.
6.3.2.2 Evaluation of pH Meter—For
the tests presented here, a pH electrode
reference solution (Section 7.0) should
be used. The field pH meters should be
tested using a laboratory pH electrode
whose performance has been docu-
mented. The pH meter is acceptable if
the average pH and the standard
deviation are within 0.03 unit of the
documented values. The pH Meter/
Electrode Acceptance Test form, and
the pH Acceptance Test Summary form
(Section 6.8.5) should be completed.
6.3.2.3 Evaluation of pH Electrodes-
Each new electrode should be equili-
brated overnight in the storage solution
recommended by the manufacturer.
Before testing, the electrode should be
carefully rinsed with deionized water,
and then successively placed in deion-
ized water in a series of testtubes until a
constant pH reading is achieved The
electrode is acceptable if the average is
within 0.1 pH unit of the average
historical value and if the standard
deviation is less than 0.03 pH unit.
Results should be recorded on the pH
Meter/Electrode Acceptance Test form
and on the pH Acceptance Test Sum-
mary Form (Section 6.8.5)
6.4 Sampler and Rain Gauge
Installation and Operation
For placement of neighboring col-
lectors and rain gauges of equal or
smaller height, the distance between
rain gauge and collector or between
collector and collector should at least
equal the height of the taller object to
minimize interference as well as splash
effects. To ensure that the collector dry
bucket is not an obstruction to the wet
(or precipitation sample) bucket, the
collector should be aligned either
perpendicular to the prevailing winds or
with the dry bucket downwind of the
wet bucket. The ground surface around
the collector and rain gauge should be
natural vegetation or gravel; being
paved may cause splashing into the
collector or gauge.
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. The collector may be
shielded from the wind, but it should 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 In
the winter, loss of snow from the
collector can occur due to blow out.
The sampler installation and opera-
tion are described in the manufacturer's
instructions and in the 0 & M manual
(9) The precipitation collector requires
no calibration, but proper functioning
should be checked frequently.
The rain gauge operation isdiscussed
in the manufacturer's instructions The
rain gauge should be calibrated ac-
cording to the manufacturer's instruc-
tions after installation and at least
annually (Section 6.6). A calibration
check at two points, approximately half
and full scale, should be made monthly
The ram gauge should be mounted on
a firmly anchored support or base so
that the funnel rim is level and at about
the same height as the collector rim to
enable comparisons of collection a-
mounts between the two. The gauge
level can be checked with a carpenter's
level placed at two intersecting posi-
tions. The gauge mouth should be high
enough not to be covered by snow In
open areas, a wind shield (e.g., swing-
leaf like the one used by the U.S.
Weather Service) should be used. For
rain gauges which contain a recorder,
the access doortothechartdriveshould
be on the leeward side of the prevailing
winds, and should be kept closed to
minimize dirt and moisture affecting the
chart and the mechanism.
6.4.1 Routine Checks on Co/lector,
Rain Gauge, and Site
The following tests should be carried
out routinely on the precipitation
collector and the rain gauge The
detailed procedures for these tasks and
a checklist are in the O & M manual (9).
1. Collector Sensor Test - At intervals
which coincide with the sampling
schedule, the sensor should be
shorted with a piece of metal or
some water to check the lid
opening and sensor heating func-
tions. When the sensor short is
removed, the lid should close
immediately, and the sensor should
cool. The sensor should be cleaned
at least monthly, or as needed.
2. Inspection of Dry Collector Bucket -
If the collector has a dry bucket, it
should be checked after an event
or a time period which has de-
posited more than 0.25 mm (0.01
in.) of precipitation to ascertain if
the bucket contains or has con-
tained any precipitation. Precipita-
tion in the dry bucket may be
evidence of a collector malfunc-
tion. Possible causes are (a) a dirty
or faulty sensor, (b) a high sensor
heating temperature and/or a low
precipitation rate, (c) a defective
magnetic mercury switch in the
motor box, (d) a lid arm is loose
and/or has moved too far out from
the magnetic switch to actuate it.
The above causes, except for the
dirty or faulty sensor, can result in
the lid cycling. This could cause
contamination and some loss of
sample.
3. Test of Dry Sample Bucket -
Weekly, if no event has occurred,
the sample bucket should be
returned to and/or tested in the
laboratory for contamination due
to poor initial cleaning, dry de-
position, and/or handling. If con-
tamination is frequent at a site,
poor collector sealing and/or an
operator handling problem are
likely occurrences.
4. Examination of the Event Pen
Marker Trace - Weekly, the event
trace should be inspected to see if
the lid cycled. The event trace
openings and closings should
correspond to the beginning and
ending of the event, as indicated by
the slopes of the sample weight
trace. Many upanddown markings
in a short time may indicate lid
cycling. Some cycling traces may
occur during light rain events or
heavy dew when no event is
apparent. No lid movement traces
when the sample weight trace
shows an event occurred indicate
a collector malfunction.
5. Cleaning Techniques and Schedule
The collector sensor should be
washed monthly with deionized
water to remove dirt, salt, and film
buildup. If a film persists, clean the
sensor grid and plate with deter-
gent and a toothbrush. The rim of
the dry bucket should be wiped
with clean tissues (e.g., Kimwipes)
to prevent carry over of dustfall to
the sealing gasket and to the wet
bucket.
6. Adjusting the Zero Setting of the
Rain Gauge - Daily or weekly as
needed, with no precipitation in
the rain gauge, adjust the zero
setting. The zero setting fluctuates
with temperature, but generally
not more than 0.25 mm (0.01 in.).
7. Checking the Rain Gauge Pail
Level - Whenever the rain gauge
pail is removed, be sure it is
replaced correctly so that it is level.
-------
Part I-Section 6.0
Jan. 1981
8 Adjusting and Winding the Ram
Gauge Clock - Weekly, 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.
9 Rain Gauge Check - Monthly, add
several known weights or water to
the rain gauge to be sure it is
measuring correctly at about the
75mm (3 in.) and 125mm (5 in )
levels. For the Belfort weighing
gauge, 25 mm = 1 in = 820 gm A
complete calibration at each inch
level should be made at least
annually
10 Inspection of Rain Gauge Pens and
Ink - Weekly, the pens should be
inspected 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.
11 Rain Gauge Chart Replacement -
At the appropriate interval, de-
pending on the chart range, general-
ly weekly, remove the old chart and
replace it with a new one.
12. Rain Gauge Level Check - At 6-
month intervals, measure the
gauge level to be sure it is still
horizontal.
13. Winterizing - For the precipitation
collector, encase the lid arms in
plastic and tape one end of the boot
to the table. If necessary, to
prevent the lid freezing to the
bucket, attach a heater (e.g., a 40W
light bulb) to the top of the lid.
Check the sensor when the am-
bient temperature is below freezing
to be sure the heater is working.
Allow collected samples to warm
up to room temperature before
measurement. For the weighing
rain gauge, remove the funnel
(generally in the inlet), and add
ethylene glycol antifreeze to the
pail. For a tipping bucket gauge,
turn on the heater.
14. Site Maintenance and Inspect/on
for Obstacles - Periodically, mow
the grass and inspect the site area
for new obstacles (e.g , a growing
tree).
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 O & M manual (9) as soon as
possible. If the problem cannot be
corrected, the field manager or equip-
ment manufacturer should be asked for
advice and direction. The field manager
or central laboratory should maintain a
supply of spare parts The diagnosis and
action taken should also be recorded in
the logbook.
6.5 Sampling Methodology
Since precipitation samples generally
contain microconcentrations of pol-
lutants, extreme care must be taken to
avoid contaminating the sample and to
preserve its integrity. This section gives
the methodology for sample collection,
handling, measurement and for pre-
serving sample integrity The pro-
cedures, used to accomplish each of the
above-mentioned tasks, are in the O &
M manual (9).
6.5.1 Sample Collection and Schedule
Sampling schedules commonly in-
clude weekly, daily, event and subevent,
or sequential. An event can be defined
as a storm separated from a second
storm by a dry interval of at least 6 h in
the winter or at least Shin the summer.
The choice of sampling schedule de-
pends on the program objective and the
available funds To correlate rain data
with aerometric and/or meteorological
data, a subevent, event, or at most a
daily schedule must be used. To mea-
sure the amount of deposition and/or
its effects, a weekly sample may be
sufficient. Sampling periods longer than
1 wk are not advised because important
changes in the sample can occur on
standing in the collector.
6.5.2 Hand/ing of Plastic Containers
Treatment of plastic containers de-
pends on the species to be measured,
the container's previous use, and its
cleanliness In most cases, the cleaning
should be done in the laboratory; only
routine bucket rinses (when the buckets
need no.t be returned to the laboratory)
should be done in the field The final
rinse water conductivity should be
measured, and should be below 2
micromho/cm Procedures for cleaning
plasticwareare in theO&M manual (9).
The container should be capped 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 lid is
minimized
6.5.3 Sample Hand/ing
The sample container must be checked
for precipitation at the time and fre-
quency set by the schedule. If a sample
is present, the container should be
removed and weighed; the bucket
swirled to help ensure a homogeneous
sample, and a sample aliquot should be
measured after assuring that the
containers are correctly labeled A
pencil or a ball point pen should be used
to inscribe the label If the sample is
frozen, it must be allowed to warm up to
room temperature before measure-
ment. Avoid breathing onto a sample to
prevent ammonia contamination.
6.5.3.1 Weekly Sampling—After a
week if no sample is present, the empty
bucket can be sealed and returned to the
laboratory or rinsed in the field and
reused. In either case, the bucket should
be rinsed with distilled/deionized
water, and the conductivity of the rinse
water measured to provide a blank
which reflects the previous cleaning,
operator handling, collector sealing,
and so forth. The deionized/distilled
water conductance should be checked
before it is used. After cleaning, the
container should be shaken dry, and
reused.
6.5.3.2 Event Sampling—To mini-
mize the number of buckets required as
well as storage and shipment space, the
sample should be weighed and trans-
fered from the bucket to a labeled 500
ml wide rnouth polyethylene bottle
Some sample is taken for measurement
and the bottle sealed. If sufficient
sample (e g., more than 300 ml) is pre-
sent, use about 50 ml to rinse out the
shipping bottle One 500 ml bottle per
event is sufficient for all measurements,
the rest of ihe sample maybe discarded
The sample transfer can be made direct-
ly from the bucket to the bottle. The
bucket should be rinsed with deionized/
distilled water before reuse.
6.5.3.3 Sequential Samples—When
samples are collected through a funnel
directly into prenumbered polyethylene
bottles, the bottles should be labeled
and sealed immediately after the
samples are measured.
6.5.4 Sample Preservation, Storage.
and Shipment
6.5.4.1 Sample Preservation—Sam-
ple degradation can occur due to
chemical interactions (e.g., with parti-
culates or gases) or to biochemical
reactions. Preservation of sample
integrity can be maximized by filtration,
sealing, and storage in the dark at about
4°C. After pH and conductivity mea-
surements, filtration should be done
with a 0.45 um organic membrane filter
(10), if inorganic species are to be
analyzed. Although biocides (e.g.,
toluene or chloroform) are effective,
they can interfere in the various
measurements or analyses (2) and thus
must not be added to the sample. If
certain species must be preserved, an
-------
Jan. 1981
Part l-Section 6.0
aliquot of the sample can be mixed with
a preservative in a separate container.
Procedures initiated in the field should
be continued in the laboratory; however,
to minimize contamination, filtrations
should be done in the laboratory as soon
as possible after sample arrival.
6.5.4.2 Storage—Samples should be
stored in a refrigerator after collection
After the laboratory analyses have been
completed and the results checked, the
sample should be transferred to a 125
ml polyethylene bottle to save space and
stored in a refrigerator or freezer for a
time period of at least six months. Thus
the samples are available for further
tests or analyses if desired. Availability
and location of storage should be
recorded in the chain of custody
documentation It is not known whether
the best method for longterm storage is
to freeze the samples or to keep them at
4°C Stability tests for several months
indicate that both the 4°C (10,11) and
the freezing (11) are satisfactory. Until
more evidence is available, 4°C storage
is recommended.
6.5.4.3 Shipment—If sampling week-
ly, ship samples to the laboratory
weekly, on a scheduled day and by the
method specified in the 0 & M manual
(9). If event or subevent sampling is
being performed, ship the sample as
soon as possible after collection and
measurement; for economic reasons,
the samples can be logged in, stored in a
refrigerator, and shipped weekly. Gener-
ally, the shipment should be made early
in the week by air (UPS Blue or air parcel
post) so that samples are received in the
laboratory before the weekend. Upon
arrival, the samples should be logged in
and stored in a refrigerator to minimize
degradation. Cumulative weekly samples
should be shipped with their collecting
buckets packed into cardboard cartons
or other protective boxes
Event and sequential samples col-
lected for special studies should be kept
cold during shipment, and should be
shipped in cardboard-enclosed Styro-
foam boxes (Polyfoam Packers Corp.,
Chicago, III.) with freeze-gel packs
which have been kept in the freezer
compartment of a refrigerator for about
24 h before shipping to ensure that they
are completely frozen. Gel-type packs
are preferred because they are less
likely to leak when unfrozen. Generally,
four packs per box are sufficient to keep
the samples cold for 4 to 5 days. The
temperature of the box interior should
be measured on arrival at the labora-
tory. The central laboratory should
replace all sample containers and
shipping materials weekly.
6.6 Field Measurements
Field measurement procedures for
pH, specific conductance, and tempera-
ture are discussed in this section
Procedures described in the 0 & M
manual (9) should be used mthecentral
laboratory as well as in the field Sample
pH and conductivity are measured in
both the field and the laboratory to
detect sample changes. The field values
representing the fresh sample are the
true values if no measurementerror has
occurred A warmup of the meters may
be needed before they are used (see
manufacturer's instructions) Results of
field measurements should be recorded
on the Field Data Form (Section 6 8 5).
6.6.1 pH Determination Method
6.6.1.1 General Description—The pH
of a solution is related to free-acid
activity.
PH = -logFT 6-1
where FT is the FT activity or free FT
concentration.
Thus pH does not measure the total acid
concentration The pH meter measures
the electrical potential difference
between a reference electrode and an
FTglass electrode; the glass electrode
potential varies with the activity (or
effective concentration) of hydrogen ion
(H+) in solution. Although the meter
measures electrical potential (volts), it is
calibrated to give data as pH.
6.6.1.2 Calibration—Since acid rain
samples generally have pH's between
3.0 and 6.0, the pH meter should be
calibrated with pH 3.0and 6.0 standard
buffers. For other less acid samples, pH
4.0 and 7.0 buffers should be used. 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.
The pH 6.0 (or 7.0) buffer is used to
adjust the calibration or standardization
setting; the slope setting is adjusted
against the pH 3.0 (or 4.0) buffer. After
the slope has been adjusted, the first
setting should be rechecked with the pH
6.0 (or 7.0) buffer; if it has changed by
more than ±0.02 pH unit, the calibra-
tion should be repeated. For these
buffer solutions, a stable reading
generally occurs in 30 s to 2 min. Since
pH is temperature dependent, calibra-
tion buffers and samples should always
be measured at the same temperature.
Before calibration, unsealed electrodes
should be topped off with filling solution
(available from the electrode manu-
facturer), and the exterior should be
carefully rinsed with deionized water.
During the measurements, the pH
electrode should not touch the bottom of
the solution vessel. If a metal ring stand
is used to hold the electrode, the stand
and the pH meter should be connected
to the same electrical ground.
The pH meter should be calibrated
before and after each measurement or a
series of measurements. 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.
6.6.1.3 Sample Analysis—Sample
measurement is performed directly
after the meter has been calibrated and
the electrode washed. Never insert the
electrode or any other object into any of
the bulk solutions, never pour a solution
back into its bulk container; and never
measure while the solution is being
stirred to eliminate errors due to
streaming potentials. For measure-
ments, a small vessel (vial or test tube)
should be used. A procedure for sample
analysis is in the O & M manual.
Precipitation samples usually yield a
stable potential in about 4 min. The pH
electrodes, after being rinsed in de-
ionized/distilled water, should be
stored in deionized water, in 10~4Nacid,
or in a 74 micromho/cm KCI standard.
6.6.1.4 Quality Control - Accuracy
and Precision — The main problem with
pH measurements is aging of the
electrode; diagnostic tests are pre-
sented here. The first test for electrode
or procedural problems should be with
unknown test samples sent monthly to
the field station by the laboratory; these
audit samples have pH and conductance
values similar to rain, and thus should
be measured for both before returning
them to the laboratory with the results
for recheck and evaluation. If the field
pH differs from the laboratory pH by
more than ±0.15 unit, the electrode
probably needs replacing; laboratory
values are assumed to be correct
because they are measured on several
samples of the unknown and because
the electrode is checked with a second
backup electrode. Thus the field results
are indicators of the accuracy of the pH
measurements; if a number of mea-
surements are made, the precision is
also obtained and should be better than
0.05 unit If the accuracy and/or
precision are poor, consultation with
the field operator on technique should
confirm the source of the problem, and it
should be corrected by the QA coordi-
nator as soon as possible.
-------
Part I-Section 6.0
Jan. 1981
A second test for electrode problems
should be a reference solution of known
pH to learn the precision of the station's
measurements. Each site should re-
ceive from the central laboratory a
polyethylene bottle of electrode refer-
ence solution with pH and conductivity
similar to those of ram samples, and
should check the electrode biweekly.
The measurement procedure should be
identical to that for rain samples. The
solution should be stored in the re-
frigerator and replaced when needed or
when the pH or conductivity appears to
have changed. The average value and
the standard deviation of a series of at
least five test sample measurements
taken consecutively and the time
required for electrode equilibration can
be used to evaluate the electrode
performance. The standard deviation (s)
is:
s = [Z(x,-x)/(n-1)]1/2 6-2
where x, = the measured pH for the ith
sample,
x = the average pH reading for the
series, and
n = the number of sample tubes mea-
sured.
For acceptable electrode behavior, s
should not be greater than 0 05 pH unit,
and the average pH should agree with
the previous month's value within
±0.10 unit. Frequently, the first mea-
surement differs by more than ±0. 1 unit
from the others, so this value can be
excluded from the average for six
sample tubes. If this behavior is
exhibited by an electrode, it is strongly
recommended that two tubes of each
precipitation sample be measured for
pH and that the second value be entered
on the data form. Performance is
determined by the time needed to attain
a stable reading — when the pH reading
becomes constant within ±0.02 unit for
1 min.; the time 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 should
be replaced. If the average pH value has
changed from that of the previous
month by more than 0. 1 0 unit, the solu-
tion conductivity should be checked. If
the conductivity has changed by more
than 10% from its original value, the
solution has degraded, and should be
replaced. Always return enough of the
solution so that it can be checked by the
laboratory.
6.6.2 Specific Conductance Deter-
mination Method
6.6.2.1 General Description—The
conductivity of a solution is the recipro-
cal of its resistance, and it is related to the
solution temperature and to the total
concentration and species of free ions
present.
6.6.2.2 Calibration—The conductiv-
ity (or resistance) varies with the
electrode area and spacing as well as
with the temperature and the ion
concentration; therefore, the mea-
suring apparatus has to be calibrated to
obtain the cell constant or to adjust the
meter. For calibration, a KCI solution of
known conductivity should be used, and
the temperature of the KCI standard and
the sample should be the same. For rain
samples, a 0.0005M KCI solution with a
specific conductance of 74 micromho/
cm at 25°C should be used in the
calibration with the measurement
procedure recommended by the instru-
ment manufacturer All conductances
should be reported in micromho/cm or
in microSiemen/cm corrected to 25°C.
If the apparatus has automatic tempera-
ture compensation or if the standard
and the unknown are at the same
temperature, no temperature correc-
tions are generally necessary if the
25°C KCI value is used for calibration.
(Temperature coefficients of the two
solutions are assumed comparable.) If
the apparatus does not have tempera-
ture compensation or if the KCI standard
and the unknown are at different
temperatures, corrections must be
applied; KCI values near 25°C, based on
a temperature coefficient of 2% per
degree, are in Table 6-2.
The conductivity apparatus should be
calibrated before and after each mea-
surement or series of measurements. If
a change of more than 5% occurs, the
measurements should be repeated; if
the drift reoccurs, a problem exists with
the apparatus, and it must be corrected.
In general, stable values (for minutes)
occur in about 30 s.
Table 6-2.
Temperature Specific
Conductance of
0.0005M KCI
T°C
20
21
22
23
24
25
26
27
28
Micromho/cm
66.8
68.2
69.5
71.0
72.4
73.9
75.4
76.9
78.4
6.6.2.3 Sample Analysis—Conductiv-
ity of the samples can be measured on
the same aliquot used for pH; if so,
conductivity must be measured before
pH to avoid error due to salt contamina-
tion from the electrode The conductivity
cell should be washed with distilled/
deionized water after calibration and
rinsed with sample solution The
procedure is in the 0 & M manual (9) If
the temperature of the sample is not the
same as that of the standard and if the
apparatus does not have automatic
temperature compensation, the mea-
sured conductance should be corrected
to 25°C by adding 2%/°C if below 25°C
or by subtracting 2%/°C if above 25°C.
6.6.2.4 Quality Control - Accuracy
and Precision—Unlike the pH electrode,
which has a limited life, the conductivity
cell generally has few problems How-
ever, in contrast to the pH buffers,
which are concentrated and quite
stable, the working conductivity stand-
ard is a very dilute, 0.0005M KCI, and
may either degrade slowly or become
contaminated To minimize errors due
to changes in the calibration standard,
the working 74 micromho/cm solution
should be replaced approximately
monthly. When a new working standard
is received, it should be checked against
the old working standard, and the two
values should agree within 10%. If they
do not, notify the laboratory. Always
return enough of the old standard to the
central laboratory for it to be remea-
sured. Never return the old working
standard before receiving a new one
Conductivity standards should be
sealed and stored in a refrigerator to
minimize changes Generally, changes
of lesslhan 5% monthly may be ignored;
if greater than 5%, the field values can
be corrected for the large changes by
prorating with time in a linear manner.
Such corrections should be duly noted.
If the conductivity meter has its own
builtin standardization circuit, it can be
used to check the KCI standard. If the
KCI standard has changed from its
original value by more than 5%, the
laboratory should be informed. The cells
should be stored as recommended by
the manufacturer.
Another means of evaluating the
working conductivity standard is to use
unknown test samples submitted month-
ly from the laboratory to determine the
accuracy and precision of the station's
specific conductance measurements.
These lest samples must be returned to
the central laboratory with the next
sample shipment for remeasurement to
ensure that they have not changed in
value. If the laboratory finds that the
station's conductivity differs from the
laboratory's by more than 10%, the
-------
Jan. 1981
Parti-Section 6.0
laboratory should inform the field and
quality assurance personnel, and should
replace the old conductivity standard.
6.6.3 Temperature Measurements
6.6.3.1 General—Each field ther-
mometer and temperature probe should
be assigned an identification number so
that it will be possible to trace its
certification and to document it proper-
ly. To minimize any chance of con-
tamination, the probeshould be washed
and dried before the temperature mea-
surement of a solution, and the probe
should never be placed in a solution
before pH and conductivity measure-
ments unless there are duplicate
vessels for each sample. The probe
should be placed only in the solution
used for rinsing, not for pH and
conductivity determinations.
6.6.3.2 Calibration andTraceability—
The central support laboratory should
maintain and store an NBS-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 cali-
brated against the certified (secondary)
thermometer in a circulating water bath
in the 0° to 25°C range; the procedure
should be similar to that used for
certifying the secondary standard. The
Thermometer Calibration Log (Section
6.8) should be completed, one copy
should be filed in the laboratory, and
another should be sent 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. Generally, calibra-
tion at two temperatures, near 0° and
25°C, is sufficient. Linear temperature
behavior may be assumed.
After initial calibration, the tempera-
ture behavior probes should be cali-
brated at least once a year in the labora-
tory or in the field by using a certified
thermometer and two temperatures
(i.e., 0°C and ambient).
6.6.4 Gravimetric Measurements
6.6.4.1 General—The volume of rain
is measured as mass of ram with a
density of 1 g/ml. The mass of ram is
measured in the field to determine the
rain collector efficiency (compared to
the rain gauge), and the mass of the
sample sent to the laboratory is mea-
sured to determine if leakage occurred
in shipment.
The field site should have a 20 kg
capacity balance for weighing rain
buckets; the balance should be in a
room free from drafts and on a table that
minimizes vibrations; and the legs of the
balance should be adjusted to level the
balance.
Before each weighing, the balance
pan should be brushed off with a soft
brush and the balance zeroed. After
each weighing, the balance should be
cleaned of all (potentially corrosive)
chemicals.
Periodic maintenance should be
according to the manufacturer's recom-
mendations; factory maintenance is
usually once a year, and a record of
maintenance should be kept for each
balance by completing the Balance
Factory Service form (Section 6.8.5).
6.6.4.2 Balance Calibration—Each
balance, before being shipped to the
field, should be calibrated with NBS
traceable weights in the central support
laboratory. In the field the balance
should be zeroed before each rain
sample is weighed. Annually, a full
calibration as part of the network
evaluation (Section 6.7) should be
performed by weighing two NBS trace-
able weights (1.0 and5.0kg)onthefield
balance. The auditor-recorded actual
reference weight, measured weight,
and weight difference should be re-
corded on the Audit Record form (Sec-
tion 6.8.5).
6.6.4.3 Rain Gauge Calibration—To
calibrate the weighing gauge, a set of
weights is generally available from the
manufacturer. With a dual traverse(0-6
and 6-12 in.) pen recorder such as the
Belfort, the range 5-7 in. has been
difficult to calibrate and to keep cali-
brated; this range is generally not
needed if the rain gauge bucket can be
emptied after each event or week of
events. However, antifreeze must be
added to the weighing gauge bucket in
the winter to melt captured snow, so
severe or prolonged storms can bring
the gauge to 5-7 in.; accordingly, it is
recommended that the bucket be
emptied when the 5 in. level is ap-
proached before adding new antifreeze.
The tipping bucket gauge iscalibrated
by using a slow drip technique to add a
controlled volume of water. The set
screws may need to be adjusted to limit
the travel of the tilting bucket.
6.7 Network 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 QA program. An audit
for a precipitation chemistry network
should include both qualitive and
quantitative evaluations. The quantita-
tive evaluation of precision and ac-
curacy is discussed in Section 9; the
remainder of this section discusses a
qualitative audit of performance.
6.7.1 System Audit
A system audit is an on-site inspec-
tion and review of the QA efforts used
for monitoring (sample collection,
sample analysis, data processing, etc.).
The QA project plan (Section 3) should
be the basis for a system audit.
For each monitoring network, a
system audit should be conducted as
soon as possible after the start of
monitoring; subsequent system audits
should be scheduled at least once per
year. About 4 to 6 weeks before the
audit, the QA coordinator should send a
questionnaire to the field operations
manager, and the questionnaire, (Sec-
tion 6.7.3) should be returned before the
auditor's visit to allow the auditor to
review and evaluate the recorded
information. The system audit should be
conducted by (a) interviewing the field
manager and the station operators, (b)
visiting the monitoring site, and (c)
writing a summary report with results,
observations, and recommendations
6.7.2 System Audit Questionnaire
The system audit questionnaire
(provided herein) is in the recommended
format, but it can be modified and
supplemented to fit the needs of each
specific monitoring network. In the
questionnaire, the R & G denote
"recommended strongly" and "guide-
line", respectively
-------
Part l-Section 6.0 8 Jan. 1981
Field Agency Questionnaire
Forms
General Information
Precipitation Monitoring Resources/Staff Size, Organization,
Qualification and Utilization
Precipitation Monitoring Resources/Staff Training
Precipitation Monitoring Resources/Equipment
Precipitation Monitoring Resources/Facilities
Monitoring Network/Network Design
Monitor Network/Network Status
Monitor Network/Network Operation
Monitor Network/Network Maintenance
Monitor Network/Network Calibration
Quality Assurance: Data Quality Assessment Requirements/General .
Quality Assurance: Data Quality Assessment Requirements/Methods.
Quality Assurance- Data Quality Assessment Requirements/
Required Calculations For Data Quality Assessment
Data Handling and Reduction/General
-------
Jan. 1981 9 Part l-Section 6.0
General Information
Questionnaire completion date.
Survey visit date:
Agency name and address.
Agency ma/ling address (if different from above)
Telephone number FTS' Commercial:! )-
Agency Director: _
Monitoring Supervisor.
Quality Assurance Coordinator:
Survey Conducted by
Affiliation of Auditor:
Persons present during entrance interview:
Persons present during exit interview:
-------
Part l-Section 6.0 10 Jan. 1981
Precipitation Monitoring Resources
A. Staff Size, Organization, Qualifications and Utilization
Yes A/o Comments
1. Provide a current organizational chart
showing precipitation monitoring and
data handling personnel. ( )
2. Are the following items adequate to
demonstrate that the equipment can
be operated, calibrated and maintained:
a. Staff size? (R)
b. Program organization? (R)
c. Staff qualifications? (R)
d. Staff utilization? fR)
B. Staff Training
1. Do staff members receive periodic training
to upgrade employee's skills? (G)
a. At least once a year? (G)
2. Are the manufacturer instrument man-
uals, and quality assurance guideline
documents for precipitation monitoring
available to the operators? (R)
3. Are the following references available?
a. Atmospheric Environment (G)
b. Journal of the Air Pollution Control
Association (G)
c. Environmental Science and Techno/ogyfG)
-------
Jan. 1981 11 Part l-Section 6.0
Precipitation Monitoring Resources
C. Equipment
Yes No Comments
1. Does the agency have the necessary hand
tools, electrical testing, and calibration
weights to operate, and maintain the in-
struments, to calibrate the rain gauges
and assure the data quality in the
network? (R)
2. Are the following types of equipment
available or on hand?
a. Hand tools-screwdrivers, wrenches,
etc. (G)
b. Multimeter b-i analog (G)
bz digital (G)
c. Soldering iron, gun, and accessories (G)
d. Electric drill, saws, etc. (G)
e. Tubing cutters and accessories (G)
3. For precipitation collection are the
following types of equipment used?
a. Automatic precipitation collectors (R) .
b. Recording rain gauge sensitive to
0.01 in (0.25 mm). (R) .
c. Meteorological gear:
(1) wind speed (G)
(2) wind direction (G)
4. Are the means to calibrate the rain
gauges available? (R) .
D. Facilities
1. Is adequate space available to operate
and maintain the network? (R)
2. Is the work space used for the sample
measurements maintained at 25°C ±5°C?(R)
3. Are the instruments operated at a
normal line voltage between 105
and 125V? (G)
-------
Part l-Section 6.0
12
Jan. 1981
A. Network Design
Monitoring Network
Yes
No
Comments
1. Is the network designed in accordance
with the program objective? (R)
a. Is there a written plan describ-
ing the network prescribing:
(J) The basis for design of the
network, selection of instru-
ments and siting? (R)
(2) The locations of the instru-
ments (site locations! by UTM? (R)
(3) The sampling schedules? (R)
(4) The methods of sampling
and analysis? (R)
(5) The method of data hand/ing
and analysis procedures? (R)
b. Have the site description ques-
tionnaires been submitted? (R)
2. Are instruments installed at a site
in accordance with:
a. Manufacturer's specifications? (R)
b. EPA guidelines? (R)
c. Sound scientific principles? (R)
3. At the sites:
a. Are there any obstacles the
height of which subtends an
angle greater than 30°? (R]
b. Are the precipitation collec-
tor and rain gauge at least 10
feet apart? (R)
c. Are the rain gauge and collec-
tor placed in a line perpendi-
cular to the prevailing winds? (G).
d. Is the rain gauge level? (R)
e. Is the access door of the rain
gauge on the leeward side of
the wind path? (G)
f. Can the rain gauge measure O.OJ
inches of precipitation? (R)
-------
Jan. 1981
13
Part l-Section 6.0
A. Network Design
Monitoring Network
Yes
Comments
4. Does network design consider:
a. Access? (R)
b. Power availability? (Ft)
c. Localized interferences? (R)
5. Is the precipitation fall to the
sites unobstructed? IR)
6. Are sites located in accordance with
current EPA site location criteria
and ambient precipitation monitor-
ing siting guidelines? (R)
7. a. Does each site ha ve a written
identifying purpose? (Rj
b. Are the samplers located at the
optimum site to meet the purpose? (R)
B. Network Status
1. Does the agency have the following
records identifying the history and
status of each monitoring site?
(R)
a. Completed site identification
form? fR)
b. Photographs or slides. One
photograph or slide toward each
of the four compass directions
and one closeup photograph of the
instrumentation at the site? (R)
c. Date site was started up and date
site was shut down, as appropriate? (Rj
d. Model, manufacturer and serial
number of instruments at the site
and dates each instrument operate? (R)
e. Reasons ior periods of missing date? (R)
C. Network Operation
1. Are the instruments in the network
operated in strict accordance with the
agency's written Standard Operating
Procedures? (R)
-------
Part l-Section 6.0
14
Jan. 1981
Monitor Network
C Network Operation
Yes
No
Comments
2. Are these procedures compatible with:
a. The Federal reference or equivalent
method if available? (R)
b. The manufacturer's instruction manual?(R)
c. The Agency Quality Assurance plan? (R)
3. Is a formal schedule used for operating
the instruments? (Attach a copy of the
schedule.) (R) _
4. Does the agency maintain an adequate
supply of expendables necessary to
service the instruments? (G)
5. Precipitation samples are picked up:
a. Every 24 hours? (G)
b. Every 7 days? (G)
c. Other (specify) hours (G)
6. In the winter:
a. Are the necessary precautions taken
with the rain gauge (antifreeze,
funnel removal, heater)? (R)
b. Are precautions taken with the collec-
tor (boots for lid arms, seal removal)? (G)
D. Network Maintenance
1. Is preventive maintenance performed in
strict accordance with the agency's
Standard Operating Procedure? (R)
2. Is a formal written schedule used for
performing preventive maintenance on
the instruments? (Attach a copy.)
3. Does the agency have access to a qualified
instrument repairman?
a. Is it a staff position?
b. Is it a manufacturer's serviceman?
c. Does the agency have a service con-
tract on any instruments? (Specify
on back of this page.)
4. Are the recommended spare parts ade-
quate and available to maintain instru-
ment downtime?
5. Are sample buckets, containers and lids
kept clean?
6. Is the collector sensor cleaned period-
ically?
(R)
(R)
(R) -
(R)
(R) .
(G)
(R) _
(R) -
-------
Jan. 1981
15
Part l-Section 6.0
Monitor Network
D. Network Maintenance
Yes
7. Is the collector rim of the dry bucket
wiped clean with Kimwipes weekly? (G)
8. Is the rain sensor tested at every site visit?(R)
9. Is the collector dry bucket inspected
for moisture whenever a sample is
removed? (R) .
10. Are the rain gauge pens checked for
ink weekly? (GJ
11. Is the rain gauge clock wound at the
prescribed intervals? (R)
12. Is the time clock on the rain gauge
recorder accurate (to '/? hr/week)? (G)
13. If not, is it corrected weekly? (R]
E, Network Calibration
Comments
1. a. Does the agency conduct frequent rain
gauge, pH and conductivity apparatus
calibrations? (R) -
b. Is the frequency adequate to demonstrate
the accuracy, precision, and completeness
for all data submitted? (RJ _
c. Are the rain gauges calibrated:
(1) Upon installation?
(2) On a semi-annual basis?
(3) When major maintenance is
performed?
(4) Before removal from operation?
(R)
(Ft)
(R)
(R)
(5) When the daily service and periodic
audits fail to meet the prescribed
performance specifications? (R)
d. Are new calibration or standard
solutions for pH and conductivity
checked against the old ones?
(R)
e. Is there a statistically significant
quantity of calibration data routinely
generated to determine the accuracy
and precision, at the 95% confidence
level, of all data? (R)
2. Is the conductivity working standard
replaced mofithly? (R)
-------
Pert l-Section 6.0 16 Jan. 1981
Monitor Network
E. Network Calibration
Yes No Comments
3. Does the agency own or have access to
calibration capability for each:
a. Rain gauge (R)
b. Recorder (R)
c. pH and conductivity apparatus (R)
4. a. Do all calibration standards pur-
chased have a maximum analytical
error of ±2%? (R) _
b. Is the accuracy of all calibration
standards specified and documented? (R)
5. Are all instruments calibrated under the
same conditions as they are operated? (R) .
6. Are instruments calibrated at the same
site as they are operated? (R) .
7. Are the pH and conductivity meters cali-
brated daily before sample measurement?^} .
8. Are records kept documenting all
a. Audits? (R) .
b. Calibrations per (R)
-------
Jan. 1981 17 Part I-Section 6.0
Quality Assurance: Data Quality Assessment Requirements
A. General
Yes No Comments
1 Are data quality "reporting organiza-
tions" properly defined? (R)
a. Is operation by common team of field
operators? (G)
b. Does organization have common
calibration facilities? (G)
c. Is organization supported by a
common laboratory or headquarters? (R)
2. Does the agency have a formal quality
assurance program, developed and
implemented, as evidenced by a
written plan? (R)
3. Are confidence limits assigned to all
data collected? (R)
4. Are all the data submitted documented
to show that they are acceptable? (R)
5. Have written Standard Operating Pro-
cedures been developed and imple-
mented for each routine task performed;
i.e. operation, maintenance, calibration,
etc.? (R)
6. Are all quality assurance data docu-
mented and available for inspection? (R)
7. Does thr agency ha ve a quality assurance
coordinator to insure that a maximum
quantity and quality are generated? (R)
8. Does the agency participate in the EPA
quality assurance audit program? (R)
9. Is the chain of custody maintained on
all samples? (Ft)
JO. Is complete traceability maintained
on all data? (R)
11. Does the agency maintain sufficient
quality assurance information related
to data collection and analysis to
demonstrate that the data are accept-
able? (R)
-------
Part l-Section 6.0 18 Jan. 1981
Quality Assurance: Data Quality Assessment Requirements
B. Methods
Yes No Comments
1. Is a one point (minimal) check on the rain
gauge performed monthly? (R)
2. Is the balance zeroed before each
weighing? (R)
3. Is the outside of the wet bucket wiped
dry before weighing? (G)
4. Does the pH electrode have an identifi-
cation number? (G)
5. Is the pH electrode stored in the
proper pH buffer? (R)
6. Is the pH electrode rinsed well with
distilled/deionized water after removal
from the buffer? (R)
7. Are the samples allowed to come to room
temperature before they are measured
for pH and conductivity? (R)
8. For the pH measurement is 4 M KCIadded
to the sample aliquot to yield a 0.04 M
solution? (Rj
3. Are the sample aliquots measured for pH
and conductivity discarded after meas-
urement? (R)
10. Is the electrode test run monthly? (G)
7 /. Are the electrode test results sub-
mitted to the laboratory? (R)
12. Are the conductivity standards and
electrode test solution refrigerated? (G)
13. Is the conductivity of the rinse water
measured and recorded? (R)
14. Are the data properly recorded in a log
book and on the data? (R)
a. Are the container labels made out
in pencit.or indelible ink? (Rj
b. Are the samples sealed securely
before shipment? (R)
c. Are shipments made on the
prescribed schedule? (R)
15. Are the stations audited semi-annually?(R)
16. As a check on precision, have at least
two pairs of collocated samplers been
operated during the calendar quarter? (R)
17. Do the sites meet the Class A require-
ments? (R)
18. Are calibration, sampling, & analysis
the same for collocated samplers? (R)
C. Required Calculations for Data Quality Assessment
1. Are the average deviations of the pH elec-
trode from the test solution calculated
monthly? (R)
2. Have the following quality assurance
reporting requirements been met? (Ask
to see copy of a typical report.)
a. A list of all monitoring sites and
their identification codes in each
reporting organization (Regional
Office and EPA-EMSL/RTP)? ( )
b. Changes in assignments of monitoring
site (Region and EPA/EMSL)? ( )
c. Quarterly reports? ( )
d. Quality assurance portions of the
annual report? ( )
-------
Jan. 1981 19 Part l-Section 6.0
Data Handling and Reduction
A. General
Yes No Comments
1. Have standard data forms (bound or
loose) been developed and implemented
to document the results as contained in
the Standard Operating Procedures!' (R)
2. Do the standard data forms document
complete data traceability? (Rj
3. Are all standard data forms completed,
dated, and signed by the person per-
forming the task? (R)
4. Are the station data record sheets for
each sample made out in duplicate? (R)
5. Is one copy of the data sheet sent with
each sample? (R)
6. Is the rain gauge chart sent weekly
with the samples? (Rl
7. Is a log book maintained? (Rl
8. Are problems, equipment changes, new
standards, new operators, etc. docu-
mented in the log book and reported to
the central laboratory? (R) .
9. Is the sample container labelled with
station identification, weight, date,
sample number? (R) .
10. Are duplicate or sequential samples
identified properly? (R) .
11. Do standard data forms show a com-
plete chain of custody and traceability
for all data from generation to sub-
mission to the data bank? (R) .
-------
Part l-Section 6.0
20
Jan. 1981
6.7.3 Auditor Site Visit
The site visitation should cover all
aspects of site operation, so the auditor
can evaluate the operator's overall
capabilities The auditor should bring a
test sample and add it to a clean bucket
at the station; observe the operator's
weighing and measurements of its pH
and conductivity; record the data on the
Audit Record form (Section 6.8.5);
watch the operator transfer the sample
back to the test sample bottle, rinse the
bucket, and test the rinse water con-
ductivity to ensure that the bucket is
clean; and finally inspect all the equip-
ment, calibrate the rain gauge, offer
advice, and ask questions while the
operator goes through his rounds and
tests. If there are any problems, the
auditor must attempt to correct them or
bring them to the attention of the field
manager.
6.8 Documentation
All data, observations, changes, or
modifications must be dated and docu-
mented on data forms and/or in
logbooks in triplicate and duplicate,
respectively (carbon paper may be
used) One copy of each should be kept
m the station records, and another
shipped with the sample, the third copy
(the data form) should be mailed
separately from the sample to the
laboratory to help trace a missing
sample.
6.8.1 Logbook
A bound notebook with perforated
pages that can be easily torn out should
be used 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 should be recorded in the logbook,
and a copy should be sent to the central
laboratory
6.8.2 Rain Gauge Chart
The rain gauge charts, marked with
station identification, date and nota-
tions for tests, problems, and so forth
should be submitted weekly to the
central laboratory.
6.8.3 Field Data Form
Field Data forms should contain all
information required for identification
of the sample—station, date, weight,
sampling times, pH, conductivity, prob-
lems, and supply orders. Triplicate
forms should be made out for each
sample.
A sample of a data form used by NADP
for event sampling is in Section 6.8.5.
Instructions for filling out the form are in
the O & M manual(9). Under item 11
(Remarks) should be included unusual
occurrences—plowing, harvesting,
burning, increased atmospheric pollu-
tion or dust, and soforth—andanyother
information, comments, or observa-
tions—power outage, smell, moisture in
the dry bucket, new electrode, new
conductance cell or standard.
The importance of the information
requested in item 11 cannot be over-
emphasized. Careful observations of the
sample and the occurrences m the
surrounding environment can aid in
evaluating the validity of the sample and
the collected data.
6.8.4 Sample Identification
Samples must be labeled so that they
can be readily and correctly matched
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 gets
wet.
For duplicate (collocated) samplers or
sequential samplers, each sample
container should be coded and logged m
separately so they can be easily identi-
fied in the data base. The station can
distinguish between duplicate samples
by adding a -1 and -2 in the site number
space on the data form; for sequential
samples, add -11, -12, -13, and soforth
to denote chronological order of collec-
tion for each event.
6.8.5 Data Forms
For the convenience of the manual
user, blank data forms are listed and
then provided on the following pages;
many of these were adapted from EPA
and other forms. No documentation is
on these forms The titles are at the top,
as customary for a data form. To relate
the form to the text, a nur .ber is in the
lower right-hand cornt —for example,
1 1/6.3.2 indicates form 1/version 1 as
discussed in Section 6.3.2. A revision of
the form may be documented with
1 2/6.3.2, form 1/version 2 and so
forth.
Form Number Title
1.1 / 6.3 2 Conductivity Meter/Cell
A cceptance Test Form
21/6.3.2 Conductivity Test Sum-
mary Form
3.1/6.32 pH Meter/Electrode Ac-
ceptance Test Form
4.1/6.3.2 pH Acceptance Test Sum-
mary Form
5.1/6.6.4 Balance Factory Service
Form
6.1/6.6.4 Thermometer Calibration
Log
7.1/6.7.2 A udit Record Form
8.1/6.8.3 Field Data Form
6.9 References
1. Galloway, J.N., and G.E. Likens,
Water, Air and Soil Pollut. 6, 241
(1976).
2. Galloway, J.N , and G.E. Likens
Tellus30, 71 (1978)
3. Cooper, Jr, H.B H.,J.A. Lopez, am
J.M. Demo, Water, Air Soil Pollul
6, 351 (1976).
4. Gatz, D.F., R.F. Selman, R.K
Langs, and R.B. Holtzman, J. Appl
Meteorol. 10, 341 (1971).
5. Morgan, J.J., and H.M. LiljestranC
The Measurement and Interpreta
tion of Acid Rainfall in the Loi
Angeles Basin, California Instituti
of Technology Report, No. AC-2
80, February 20, 1980.
6. Robertson, J.K., T.W. Dolzine am
R.C. Graham, Chemistry and Pre
cipitation from Sequentially Sam
pled Storms, EPA report to b<
published.
7. Raynor, G.S., and J.P. McNeil, Th(
Brookhaven Automatic Sequentia
Precipitation Sampler, BNL-50818
Brookhaven National Laboratory
January 1978; Atmos. Environ
73,149(1979).
8. P.B.S.K. Associates, P.O. Box 131
State College, PA. 16801, Bulletir
177.6801.
9. Operations and Maintenance Man
ual for Precipitation Measure
ment Systems, U.S. Environ
mental Protection Agency, Re
search Triangle Park, N.C., (ir
press).
10. Peden, M.E., and L.M. Skowron
Atmos. Environ. 12, 2343 (1978).
11. Rothert, J.E., Battelle Pacifii
Northwest Laboratories, Richland
Washington, MAP3S Program
telephone communication.
-------
Jan. 1981
21
Part l-Section 6.0
Conductivity Meter/Cell Acceptance Test Form
Date of Test:
Preparation date of a 0.0003 M KCI reference solution:
Meter type/serial no. [_
Conductivity cell type/serial no. _
Indicate whether test of
.meter or
cell
(Analyst Signature)
Conductivity Values Obtained:
Aliauot 1:
Aliauot 2:
A/iauot 3:
Aliquot 4:
Aliauot 5:
Aliquot 6:
Aliquot 7:
Aliquot 8:
Aliquot 9:
Aliauot 10:
Test Solution
Average conductivity
± standard deviation:
Accepted_
Rejected.
QA Manual for Precipitation
1.1/6/3/2
-------
Part I-Section 6.0
22
Jan. 1981
Conductivity Acceptance Test Summary Form
Meter Type/
Serial #
Cell Type/Serial^
Date of Ref
So/n Prep.
Date of
Check
Conductivity Value
Average ± Standard
Deviation (umho/cm)
Number
of
Values
Analyst
Initials
QA Manual for Precipitation
2.1/6.3.2
-------
Jan. 1981
23
Part I-Section 6.0
pH Meter/Electrode Acceptance Test Form
Date of Test:
Preparation date of pH electrode reference solution: _
Meter type/serial no. Z
pH electrode type/serial No..
Indicate whether test of.
meter or
electrode
(Analyst Signature)
pH 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:
pH Electrode Test Solution
A verage pH + standard deviation-
Accepted _
Rejected^
QA Manual for Precipitation Measurement
3.1/6.3.2
-------
Part l-Section 6.0
24
Jan. 1981
pH Acceptance Test Summary Form
Meter Type/
Serial #
Electrode Type/
Serial #
Date of Ref.
Soln. Prep.
Date of
Check
pH Value
Average ±Standard
Deviation
Number
of
Values
Analyst
Initials
QA Manual for Precipitation
4.1/6.3.2
-------
Jan. 1981
25
Part l-Section 6.0
Balance Factory Service Form
Balance
Identification
Number
Date of
Factory
Service
Name of Factory Service
Representative
Date Next
Service
Due
QA Manual for Precipitation
5.1/6.64
-------
Part l-Section 6.0
26
Jan. 1981
Thermometer Calibration Log
Date
Therm.
Serial #
Temperature #1
NBS
Value
Lab
Value
Temperature #2
NBS
Value
Lab
Value
Temperature #3
NBS
Value
Lab
Value
Slope3
Intercept^
Anal.
Init.
a Calculated from a linear least squares fit with the lab value on the X-axis and the NBS value on the Y-axis. The
laboratory thermometer correction to NBS is, then: NBS value - slope x(lab value) + intercept.
QA Manual for Precipitation
6.1/6.6.4
-------
Jan. 1981
27
Part l-Section 6.0
Site Identification:
Operator.
Date:
Auditor:
A udit Record Sheet
Lab Anal. Before
pH Cond
Audit Sample No.-
Field Anal.
pH Cond
Lab Anal. After
pH Cond
Difference Between Initial and Final Lab Values:
pH-
Difference Between Lab and Field Values:
pH
initial final
Cond -
Cond
initial final
Comments'
QA Manual for Precipitation Measurement
7 1/6.7.2
-------
Field Data Form
1 Station
Name,
ID .[
4 Bucket
On
Bucket
Off
Date
Mo
I
Day
I
Yr
I
Dale
Mo
I
Day
I
Yr
I
EOT 1-1)
ESTCD710)
CST/MDT 111
Circle Time Zone
MST/PDT (2)
PST 131
AKDT (4)
AKST MST 151
American (6)
Samoa
7 Sample Weight - Grams
Only for Buckets with water ice. or snow
Bucket &
Samp'e & Lid
Bucket & Lid
Sample
weight
CAL/NREL USE ONLY
BULK
DA
QA
NS/Exclude
2 Observer
Signature
LD
NO
NN
3 Sample Bucket
Check || ||
One Dry Side Wet Side
5 Site Operations
Check Yes or No for each item for wet-side samples only
if No. explain in remarks
/ Collector appears to have operated properly and sampled all precipitation
events during entire sampling period
2 Rain gauge appears to have operated properly during the week
3 Collector opened and closed at least once during the week
6 Sample Condition Complete for all samples containing precipitation
1 Bird Droppings
2 Cloudy or Discolored
3 Lots of Soot or Dirt
Describe all else in remarks
8 Precipitation Record
For wet-side samples
only
Type
Circle one for each
day of precipitation
Amount In inches
or circle
one
R-Ram only
z-zero
Bucket On
s-Snow only
t-Trace
To
m-Mixture
mm-Miss ing
Bucket Off
R S M U
Z T MM
Wed
R S M U
T MM
Thur
R S M U
T MM
R S M U
Z T MM
Sat
R S M U
Sun
R S M U
Z T MM
Won
R S M U
Total sampling period precipitation from rain gauge
Total precipitation from sampler = sample weight x 0 00058
inches/gram
9 Sample Chemistry
Only for wet-side buckets with precipitation
Aliquot Removed Standard Certified Standard Measured Correction Factor
Correction Factor Sample Measured Sample Corrected
pH
urn
pH 4 Observed
Sample
Tues
R S M U
Z T MM
inches
inches
10 Supplies
Circle if needed
pH 4
pH 7
75 fjS/cm
Field Forms
11 Remarks
For example' Contamination by operator, equipment malfunction, harvesting in
area
TJ
B
CO
o
o
r*
5'
3
O>
b
10
00
c_
B
3
to
00
QA Manual for Precipitation
Revised 10/1/80
81/683
-------
Jan. 1981
Part l-Section 7.0
7.0 Laboratory Operations
Quality data generated by an analyt-
ical laboratory requires detailed infor-
mation on needed equipment, control of
instrument performance, and a formal
QC program.
7.1 Facilities
Minimum facilities which must be
available to laboratories analyzing
rainwater samples include adequate
benchspace, lighting, electrical power,
temperature control, compressed gases,
exhaust hoods, reagent storage space,
refrigerated sample storage, glassware,
analytical equipment, and an area for
washing glassware (1).
7.1.1 Benchspace
Benchspace should be adequate for
equipment, and should allow each
analyst a minimum of 4 ft2 for prepara-
tion of reagents. A minimum of 8 ft2
each is needed for the atomic absorp-
tion spectrophotometer, ion chroma-
tograph, and automated colorimeter
system; 4 ft2 each is needed for the
analytical balance, pH meter, and
conductivity meter Bench tops should
be 36 to 38 in. above the floor.
7.1.2 Lighting
Lighting should be sufficient for
easy reading of glassware gradations,
balance verniers, and other measuring
lines. It is recommended that the
lighting of benchtops be at least 100
footcandles.
7.1.3 Electrical
The electrical system should provide
both 115 v and 230 v, and should have
the capacity to accommodate the
wattage requirements of all the electri-
cal devices in the laboratory Voltage
must be stable to protect instrument
components and to avoid instrumental
response changes. Most modern instru-
ments have builtin voltage regulators,
so that the only requirement of the
voltage source is that "spikes" greater
than ±10% do not exist; if necessary,
voltage regulators can be readily
obtained. AM electrical equipment
should be carefully grounded, prefer-
ably using three-pronged plugs.
7.1.4 Laboratory Temperature
The laboratory temperature should be
maintained at 22±3°C, without exces-
sive drafts or temperature changes. It is
especially important to isolate the ion
chromatograph, which has a conductiv-
ity detector, from drafts and tempera-
ture fluctuations. Dilute solutions have
conductivity temperature coefficients of
approximately 2%/°C.
7.1.5 Compressed Gases
Compressed gases are needed for
the ion chromatograph and for atomic
absorption analyzers. Compressed air
for the ion chromatograph provides a
constant 80 psi to the pneumatically
actuated valves; any inert gas is
acceptable, and purity is not critical
although particulates should not be
present
Since gases used for atomic absorp-
tion spectroscopy eventually pass the
light path of the atomic absorption
detection system, standards of quality
must be met. For the analysis of calcium,
magnesium, sodium, and potassium,
acetylene is the fuel for the atomic
absorption burner and air is used as the
oxidant; both gases should be free of oil,
water droplets, and particulates. A filter
in the line will remove these impurities.
Air flow to the atomic absorption
burner is approximately 20 liter/min;
a standard #1 air cylinder lasts about
5 h at this rate. If a cylinder is used, it
should be compressed air, not "breath-
ing" air or any other type. Instead of
cylinder air if a compressor is used, the
flow should be at least 30 liter/min. at
40 psi. A surge tank will help provide a
smooth flow.
Acetylene for atomic absorption is
burned at 5 liter/min. when air is the
oxidant; at this rate, a standard sized 1 B
cylinder will last about 30 h. Acetylene
in cylinders is dissolved in acetone, and
the ratio of acetone vapor to acetylene
vapor changes as the acetylene is
depleted. Because atomic absorption of
some elements is affected by this ratio,
it is good practice to renew acetylene
cylinders when the pressure is below
100 psi.
7.1.6 Exhaust Vent for Atomic Absorp-
tion Burner Gases
An exhaust vent is necessary to
protect personnel from toxic vapors, to
minimize the effects of room drafts on
the flame, and to protect the atomic
absorption instrument from corrosive
vapors. The venting system should
provide a flow of 5400 to 8400 liter/
min. The vent system has an "inverted
funnel" intake over the burner, tubing to
the blower intake, the blower itself, and
tubing from the blower exhaust. Stain-
less steel sheets and flexible stainless
steel tubing are recommended for
construction of the vent system. All
connections should be made with
screws or rivets because temperatures
up to 310°C may occur at the inlet.
Atomic absorption instrument manu-
facturers provide details of construction
for vent systems.
7.1.7 Exhaust Hood
Acids are used to stabilize samples
for metal analysis and to wash glass-
ware and plasticware. It is recom-
mended that these operations be done
in an exhaust hood. The air flowthrough
the hood should achieve a linear face
velocity of 100 ft/mm with the sash
fully open.
7.1.8 Vacuum
Rainwater samples are filtered to
remove particulates. Vacuum or inert
gas pressure filtration should be used to
minimize the time of exposure of the
sample to laboratory air and dust.
7.1.9 Sink
A sink with both hot and cold water
should be provided for washing glass-
ware and plasticware. An area (about
16 ft2) should be provided for air drying
glassware and plasticware.
7.1.10 Storage Space
Adequate shelf space should be
available for storing reagents, glass-
ware, and plasticware. Closed cabinets
are preferred to minimize chances of
dust settling in containers.
7.1.11 Distilled or Deionized Water
Distilled or deionized water is used
in an analytical laboratory to prepare
reagents, to make dilutions, and to
rinse glassware and plasticware. Table
7-1 offers criteria for water purity (2,3).
Water having a conductivity less
than 1.0 micromho/cm (resistivity
greater than 1 .0 megohm/cm) is
acceptable for analysis of major constit-
uents in rainwater. In the past, high
purity has been obtained by distilling
water; however, distillation systems
have several drawbacks. Even water
double or triple distilled contains easily
detectable impurities (3). Stills require
periodic shutdown and careful cleaning
and water production is relatively low.
Ion exchange systems, on the other
hand, provide high quality water, are
relatively maintenance free, and pro-
vide water on demand. The only main-
tenance 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
-------
Part l-Section 7.0
Jan. 1981
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 1.0 micromho/cm. If trace
organics are to be determined, an
activated charcoal filter should also be
used for purification
Water treatment equipment can be
leased from companies such as Culli-
gan; complete systems, such as those
listed in Table 7-2, can be purchased.
High-purity water should be pur-
chased in polyethylene bottles except
when organics are to be analyzed; then
the storage container should be glass
High-purity water should be periodically
checked with a conductivity meter
independent of the mime meter. If the
laboratory conductivity meter is cali-
brated (Section 6.3), the conductivity of
the laboratory deionized water should
be checked with each set of rainwater
samples analyzed.
7.1.12 Equipment
Major equipment needs for rain-
water analysis include an analytical
balance, a pH meter, a conductivity
meter, an atomic absorption spectro-
photometer, an ion chromatograph, and
an automated colorimetric system. In
addition .the laboratory should have a
desiccator, a drying oven that heats to
200°C, an autoclave capable of main-
taining 121±0.5°C, an NBS-certified
thermometer, a thermometer which can
be calibrated against the NBS-certified
thermometer, a set of Class-S weights,
and other weights which can be
calibrated against the Class-S weights.
7.2 Analytical Reagents
7.2.1 Reagent Quality
Reagents used for analyses must
meet standards of quality denoted by
Table 7-1. Water Purity
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 Rea-
gents of the American Chemical Society
(4).
It may not be possible to obtain dyes of
analytical reagent grade for automated
colorimetric ammonium and phosphate
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.
The lanthanum nitrate used as a flame
buffer in atomic absorption should be
"atomic absorption grade."
7.2.2 Drying of Reagents
Hydrated salts should be stored in a
desiccator overnight before weighing.
Anhydrous reagent chemicals should
be dried overnight in an oven at 105° to
110°C, allowed to cool in a desiccator,
and promptly weighed before dissolu-
tion
7.2.3 Traceability to NBS
Traceability to NBS is usually pro-
vided by analysis of an NBS standard
reference material (SRM). As of May
1980, NBS had not provided a rainwater
SRM, so the only way to obtain NBS
traceability is to reference calibration
standards to NBS-SRM's. Table 7-3 lists
NBS SRM's applicable to rainwater.
These materials may be ordered from
NBS.
Office of Standard Reference Materials
Room B311, Chemistry Building
National Bureau of Standards
Washington, D C. 20234
Telephone: 301/921-2045
This NBS traceability information is
provided mainly as a service. Labora-
tories are not required to trace all
calibration standards to NBS. Calibra-
tion standards and independently
Degree of Purity
Pure
Very Pure
Ultrapure
Theoretically Pure
Maximum
Conductivity
(micromho /cm)
10
1
0.1
0.055
Approximate
Concentration
of Electrolyte
(mg/l)
2-5
0.2-0.5
0.01-0.02
0.00
Table7.2. Commercially Available Water Purification Systems
Manufacturer Reverse Osmosis System Ion Exchange System
Barnstead
Culligan
Millipore
RO pure
Milli-RO
NANO pure
Milli-Q, Super-Q
prepared "analyst spikes" should be
prepared from ACS reagent grade
chemicals. When a new stock calibra-
tion standard is prepared, it should be
checked against the old standard.
7.2.4 Storage of Reagents and Re-
agent Solutions
Analytical reagents have a finite shelf
life, so all chemicals received by the
laboratory should be dated by the
receipt clerk and labeled "Do not use
after . . ." Shelf lives of chemicals vary
with manufacturer, so the manufac-
turer should be consulted. Unless
otherwise specified by the manufac-
turer, inorganic chemicals have a shelf
life of 5 yr at room temperature. The
analyst should become familiar with
publications (5,6) related to preparation,
standardization, and storage of rea-
gents.
Concentrations of reagents in solu-
tion may change due to: (1) biological
action, (2) chemical reaction (e g., oxida-
tion), (3) evaporation, and (4) adsorp-
tion-desorption phenomena on solution
container surfaces. All of these effects
can be slowed by refrigeration Guide-
lines for reagent storage are in applic-
able analytical methods and in the
publications (5,6,7) noted above As
discussed below (Section 7.3), poly-
ethylene or Teflon bottles are the
preferred storage vessels for all rea-
gents for inorganic analyses. Chemical-
ly resistant borosilicate glass such as
Kimax or Pyrex are acceptable for
storage except for reagents to be used
for alkali and alkaline earth metal
analyses.
7.3 Glassware and
Plasticware
This section discusses recommended
types of and cleaning procedures for
glassware and plasticware, and Federal
specifications for volumetric glassware.
7.3.1 Choice of Vessels for Prepara-
tion and Storage of Reagents and
Samples
Potential problems include: (1) con-
tamination of the solution by leaching of
species from container surfaces, (2) loss
of species from solution due to adsorp-
tion on container walls, and (3) break-
age of containers. The second is most
important for trace metal analysis,
which is not addressed in this docu-
ment; several plastic materials are
acceptable for. containers to be used in
trace metal analysis if the solution is
acid stabilized. To minimize the other
two problems, plastic vessels should be
used when possible. Most laboratory
vessels are available in plastic.
The type of plastic depends on the
application. A recent article(7) dis-
-------
Jan. 1981
Part l-Section 7.0
cusses trace contamination, durability,
and recommended used and cleaning
procedures for plastics. Polystyrene
(PS), Teflon (TFE) and conventional or
low-density polyethylene (CPE) were
found to be the cleanest plastics,
presumably due to the relatively simple
manufacturing processes. Polypro-
pylene (PP) and linear polyethylene
(LPE) were much more contaminated.
The Teflon CPE and FEP fluorocarbon
resins were recommended because
they are relatively clean chemically
resistant tough materials suitable for
field applications. Plastics such as LPE
are acceptable for major constituent
analysis of rainwater.
All glassware used in the laboratory
should be borosilicate, and should be
thoroughly cleaned as detailed in the 0
& M manual(8).
7.3.2 Volumetric Glassware
All glassware calibrated to contain
(TC) or to deliver (TD)—a precise mea-
sure of volume—must meet the NBS
specifications for Class A volumetncs(9,
10), as given in Table 7-4. The volume of
solution and the internal volume of the
glass container itself change with
temperature. The temperature (usually
20°C) at which the volumetric glass-
ware was calibrated is indicated on the
glassware, solutions should be ±5°C of
Table 7-3. NBS-SRM's Applicable to Rainwater Analysis
Measurement SRMU Type Value
Unit
pH
pH
Conductivity,
1C. Cl
Ca++
Mg~
iC.NOs
NHA\ PO*'3
185e
186 Ic
186 He
999
915
929
193
194
Potassium Acid
Phthalate
Potassium Dihydrogen
Phosphate
Disodium Hydrogen
Phosphate
Potassium Chloride
Calcium Carbonate
Magnesium Gluconate
Potassium Nitrate
4.004 pH(25°C)
(6.863)a pH(25°CJ
(7.4 15} pH(25°C)
Ammonium Dihydrogen
Phosphate
Na\Cr 919 Sodium Chloride
SO*" None Available
Instructions are included for preparation of each pH value using different
ratios of 186 Ic and 186 lie.
Table 7-4.
Tolerances for Class A Volumetric Glassware"
Type of
Glassware
Graduated flask
Transfer pipet
Capacity"
(ml)
25
50
100
200
250
300
500
1000
2000
2
5
10
25
30
50
100
200
Limit of Error
(ml)
±0.03
±0.05
±0.08
±0.10
±0.11
±0.12
±0.15
±0.30
±0.50
±0.006
±0.01
±0.02
±0.025
±0.03
±0.05
±0.08
±0.10
RAbridged reference 10
hLess than and including.
the calibration temperature for accurate
volume measurements
Disposable glassware (e.g , Pasteur
pipets; culture tubes) should not be used
because they may be of questionable
cleanliness and because sodium may
readily leach from the soft glass used m
the manufacture of such disposable
items.
Disposable plasticware (e.g., poly-
propylene test tubes) may be used, but
they should be thoroughly cleaned
before use.
7.3.3 Cleaning of Glassware and
Plasticware
Glassware and plasticware include
the borosilicate volumetrics, test tubes,
and flasks and the polyethylene con-
tainers All should be rigorously cleaned,
segregated, and dedicated to various
analyses. Glassware for trace metals
analyses should be routinely acid
cleaned; other glassware should be
rinsed in deionized/distilled water.
Detailed cleaning procedures are in the
O & M manual(8)
7.4 Laboratory Support for
the Field
The laboratory must prepare stan-
dards for calibrating field instruments
and for field testing the quality control
samples. Clean sample containers and
shipping materials should be supplied
weekly. This section discusses refer-
ence solutions, laboratory evaluation of
field equipment, and routine materials
supplies. Detailed procedures for prepara-
tions of solutions are in the O & M
manual(8).
7.4.1 Conductivity Standards
Field calibration for conductivity mea-
surement is a single point calibration
•with the standard 0.0005M KCI. Each
month, sufficient quantity is prepared
for each field site to receive 1 liter of this
dilute standard. It is recommended that
this standard be prepared in a large
polyethylene carboy dedicated to this
purpose.
The conductivity of the field standard
is established by comparison with a
calibration curve, then the Field Con-
ductivity Standard form (Section 7.8) is
filled out, and a 1-liter bottle for each
site is filled and labeled for shipment.
After 1 mo, a new standard is sentto the
field, and the old field conductivity
standard is returned to the laboratory,
and rechecked, and the data recorded
again on the Field Conductivity Stan-
dard form so that the before and after
values can be compared.
-------
Part l-Section 7.0
Jan. 1981
7.4.2 pH Electrode Reference Solu-
tion
A pH electrode may perform well in a
pH buffer, but it may be inaccurate for a
dilute sample; therefore, it is important
to check pH electrodes with -a dilute
solution. The pH electrode reference
solution is a dilute solution used to
check all laboratory and field pH
electrodes before the latter are sent to
the field The pH electrode reference
solution should be protected from air in
a sealed container, and should be
refrigerated. The data should be re-
corded on the Field pH Electrode Test
Solution form (Section 7.8).
7.4.4 Evaluation of Equipment
All meters and electrodes are 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
given to each electrode. Acceptance
tests are in Section 6 3 and detailed
laboratory procedures are in the 0 & M
manual(8)."
7.4.5 Supplier of Material
All sample containers sent to the field
must be washed and tested in the
laboratory for cleanliness. All sample
containers and shipping materials sent
to the laboratory from the field sites
must be replaced routinely by the
central laboratory as soon as possible
after arrival to avoid exhausting the
field's supply, thus minimizing down-
time. All other supplies required by the
field should be shipped when requested
on the data form (Sections 6.8.3 and
6.8.5; O & M manual. Section 2.8.3).
7.4.3 Preparation of Audit or Test
Samples
Accuracies of field conductivity and
pH measurements are evaluated with
field quality control audit samples.
Monthly, an audit sample should be
prepared using the procedures in the O
& M manual(8). Aliquots(about 60ml) of
the diluted audit sample should be sent
to each site. Three aliquots in bottles
should be kept by the laboratory; when
the samples are sent to the field, the
laboratory should determine the initial
pH and conductivity of the three
aliquots, and then the aliquots should
be refrigerated.
When the field audit samples from all
sites have been returned to the labora-
tory, the samples are reanalyzed along
with the three aliquots. The laboratory's
pH electrode is checked for accuracy
against a backup electrode. The data are
reported to the QA coordinator (Section
9.2).
7.5 Analytical Methodology
To obtain acceptable data, one must
choose an analytical procedure and
range appropriate for the sample The
next step is to gam experience with the
instrument and its performance charac-
teristics—its precision and accuracy.
One way to gain such experience is to
document performance before analyz-
ing samples
7.5.1 Selection of Analytical Methods
Discussed here are state-of-the-art
analytical procedures applicable for
rainwater analysis, in Table 7-5; others
may be used if similar sensitivities are
achieved.
Existing data(12) demonstrate that
procedures are not sensitive enough to
provide meaningful data for most
phosphate (PCU~3) analyses and that
results of potassium (K+), calcium (Ca++),
and magnesium (Mg++) analyses are
frequently at the detection limits so
careful analytical procedures must be
used. Recommendations for optimal
instrument sensitivity and performance
are in Sections 7 5.3 and 7 6. Apossible
alternate is to obtain greater sensitivi-
ties for potassium (K+), calcium (Ca++),
and magnesium (Mg++) with graphite
furnace atomic absorption
7.5.1.1 Gravimetric Measurements—
In the field and the laboratory, the
volume of ram is determined by mea-
suring the mass and by assuming a
density of 1 gm/cm3 Mass of rain is
measured in the field to compare
collector efficiency with the rain gauge;
the sample mass sent to the laboratory
is checked to determine if leakage
occurred in shipment.
In addition to a 20 kg capacity balance
for weighing ram buckets, the analytical
laboratory should have an analytical
balance which achieves the precision
necessary for the preparation of rea-
gents and salts for calibration stan-
dards. Balances should be in a tempera-
ture-controlled room free from drafts
and on a rigid table to minimize vibra-
tions. The legs of the balance should be
adjusted so that it is level. Electronic
balances should be electrically grounded
Before weighing, the balance pan is
brushed off with a soft brush, and then
the balance zero is set After the
weighing, the balance should be cleaned
of all (potentially corrosive)chemicals. If
analytical balances are not in use, the
beam should be raised above the knife
edges.
Analytical balances should be cali-
brated daily using either NBS Class-S
weights or weights traceable to NBS
Class-S; NBS Class-S weights should
be stored as primary references in the
laboratory, and daily calibration weights
should be certified against these; and
the Certification of Weights to NBS form
(Section 7 8) should be completed. All
certified weights should be recertified
every 6 mo.
The procedure used to certify weights
should be repeated five times- the
balance should be zeroed according to
manufacturer's recommendations, the
NBS-certified 1.0- and 5.0-g weights
should measured, and the test 1.0- and
5.0-g weights should be weighed. Each
balance should be calibrated using
weights close to those actually mea-
sured High-capacity balances should
' j calibrated similarly, with 1.0-and
5.0-kg weights. The Balance Calibration
Log (Section 7 8) should be completed
each time the balance is calibrated.
The 1.0- and 5.0-kg weights should
be shipped to the field in wooden
storage boxes suitable to protect the
weights when they are not being used.
Explicit instructions on the care of
calibration weights should be given by
the QA coordinator or field manager
Balances should be maintained peri-
odically according to manufacturer's
recommendations. For the laboratory
analytical balances, factory mainten-
ance is usually once a year. A record of
maintenance should be kept for each
balance; (complete the Balance Factory
Service form (Section 7.8)).
7.5.1.2 pH Measurement—Labora-
tory pH measurement is a check on field
measurement and sample degradation
Table 7-5. Procedures Recommended for Rainwater Major Constituent
Analysis
Analysis
Instrumental Method
Volume (weight)
pH
Conductivity
Strong acid
Total acid
Chloride, phosphate,
nitrate & sulfate
Ammonium, phosphate
Sodium, potassium,
calcium, magnesium
Analytical balance
pH meter
Conductivity meter
Gran plot using pH meter
Potentiometric titration with pH meter
Ion chromatography (preferably automated)
Technicon automated colorimetry
Atomic absorption spectrophotometry (flame)
-------
Jan. 1981
Part l-Section 7.0
The procedure is given in the 0 & M
manual(8).
The pH meter should be calibrated
before and after each measurement or
each series of 20 measurements. If
initial and final calibrations have
changed more than 0 02 unit, the mea-
surements should be repeated; if
changes reoccur, a problem with the
apparatus should be remedied
Each day the meter is used, it should
be calibrated, and real-time quality
control procedures (Section 7 6) should
be applied The quality control sample
(Section 7.6) should be included with
each set of samples analyzed
Samples are measured directly after
the meter has been calibrated and the
electrode has been washed. Neither the
electrode nor any other object should be
inserted into any of the bulk solutions,
and no solution should ever be poured
back into a bulkcontamer For measure-
ment, use a small vessel (vial or test
tube) rinsed first with distilled/deion-
ized water and then (if sufficient solu-
tion permits) with sample
The working pH electrode should be
checked weekly, if performance is
unacceptable, replace it with a new
tested electrode
7.5.1.3 Strong Acid (by the Gran
Method)—Data from a recent study(11)
indicate that 95% of rainwater samples
from the eastern United States contain
strong acid at a concentration at 1 6 x
10~4N or less (pH greater than 3 8) and
that the median rainwater concentra-
tion of strong acid is 4.6 x 10~5N (pH of
4.3) Strong acid is determined using
the Gran(12) plot, based on the Nerns-
tian response of the pH electrode. The
procedure is in the O & M manual(8)
At setup time, conditioning solution
intercepts and spike recovery data
should be evaluated. Real-time evalua-
tion of this information along with
quality control samples to be analyzed
are discussed in Section 7.6.
Key performance indicators for Gran
titration are the solution temperature,
the conditioning solution's initial po-
tential, time for potential stabilization
after each addition of base, the correla-
tion coefficient of the linear least
squares fit of a function
-------
Parti-Section 7.0
Jan. 1981
defined (scale expansion)to indicate the
most sensitive analysis and to serve as
guides to evaluate daily performance.
As part of real-time QC, thefirst calibra-
tion curve linear least squares fit should
be determined, and the analyst spike
should be calculated and compared with
the known value. After the analytical
data are measured, performance is
evaluated using results of QC samples
and curve parameters (Section 7 6).
7.5.1.8 Atomic Absorption Measure-
ments—Atomic absorption is used to
determine sodium (Na+), potassium (K+),
calcium (Ca++), magnesium (Mgt+) ca-
tions. Recent data(11) show that 96% of
rainwater samples from the eastern
United States contain Na+ in the range
of 0.02 to 1.0 microgram/ml; 94%
contain K+at Iessthan0.25 microgram/
ml; 68% contain Ca++ at less than 0.25
microgram/ml, and 29% contain Ca++ in
range of 0.25 to 1 25 microgram/ml;
and 66% contain Mg++ at less than 0.04
microgram/ml, and 30% contain Mg++
in the range of 0.04 toO.16 microgram/
ml. The median concentrations are 0.25
microgram/ml for Na+, 0.06 micro-
gram/ml for K+, 0.13 microgram/ml for
Ca++, and 0.03 microgram/ml for Mg++.
These low concentrations are detection
limit analyses for K+, Ca++, and Mg++, so
very sensitive procedures must be used.
Procedures for the analyses are in the 0
& M manual(8), with recommendations
for maximum atomic absorption sensi-
tivity. Since the detection limit is
determined to a large extent by light
source noise, the most stable light
source available should be used.
Electrodeless discharge lamps (EDL's)
are more intense than the conventional
hollow cathodes; an EDL is not available
for K+ analysis.
Key performance indicators for atom-
ic absorption are the baseline noise, the
calibration standard response, and the
curve linearity and reproducibility. After
instrument performance is documented,
baseline noise and calibration standard
response are defined (scale expansion)
to give the most sensitive analysis as
guides to evaluate daily performance.
Real-time QC control procedures as
well as information on analysis of QC
samples are in Section 7.6
7.5.2 Choice of Analytical Ranges
For pH and conductivity measure-
ments, one specifies the instrument
precision. Instruments such as the ion
chromatograph, the Technicon auto-
mated colonmetric analyzer, and atomic
absorption analyzers have wide analyt-
ical ranges. Because very dilute sam-
ples are analyzed, the lowest range
should be the instrument's most sensi-
tive range; higher ranges are also used
so that a high off-scale response on the
lowest range can be measured to better
than ±10% on the higher range
7.5.3 Documentation of Instrumental
Performance
Instrument performance should be
studied before rainwater samples are
analyzed to give the analyst experience,
to provide useful QC information, and to
evaluate laboratory analytical tech-
niques used at the concentrations of
rainwater samples.
The instrument performance study
for each constituent to be determined
should consist of repetitive analyses of
calibration standards at different con-
centrations. For pH, the Gran strong
acid, the acidity and conductivity, and
the instrument performance proce-
dures are different from those for the
other measurements. For pH, strong/
total acid and conductivity, solutions of
10"3, 10"4, 10"5,10"6N HCI are prepared
in C02-free water, and carefully pro-
tected from air until they are analyzed,
five aliquots of each of these solutions
are poured, and five aliquots of the pH
electrode reference solution are poured.
The meter is calibrated for each mea-
surement, as indicated by the specific
procedure and the samples analyzed.
The samples should be analyzed in the
following order—electrode reference
solution, 10~6IM HCI, 10"5N HCI, 10"4N
HCI, 10"3N HCI—repeated for five
sample aliquots. For each analysis, the
pH, Conductivity, and Gran Strong Acid
Instrument Performance form (Section
7.8) is completed
For all other analytes, the instrument
is set up for the most sensitive analysis,
and the calibration standards contain-
ing concentrations to yield responses of
5%, 10%, 20%, 40%, 60% and 80% of
full scale are analyzed. A synthetic
rainwater sample (Section 7.6) con-
taining concentrations close to the
median rainwater concentrations should
also be analyzed. The six-point calibra-
tion curve and the synthetic rainwater
sample should be analyzed consecu-
tively five times. A linear least squares
fit of expected concentration versus
response is calculated for the calibra-
tion standards; the curve parameters
from this fit are used to calculate
regression concentrations of each data
point. The average or standard deviation
of the regression concentrations should
be calculated at each concentration,
and the percentage relative standard
deviation and the percentage accuracy
should be compared to the known
concentration.
Data for each calibration curve should
be considered separately to determine
detection limits. A linear least squares
fit of each calibration curve should be
calculated; one of the curve's param-
eters which can be calculated from this
fit is the Hubaux and Vos( 15) detection
limit (Appendix A), which is much
higher than the one calculated as two
times the baseline noise; thus it could
be considered the lowest level for which
a precise analysis can be done. After a
sufficient number of values are avail-
able, an upper control limit can be used
to reject poor calibration curves Until
control limits are established using
range chart techniques (Section 8 0),
the average detection limit can be taken
as the control limit for evaluation of
daily instrument performance.
All data from the instrument per-
formance study are tabulated and
plotted as QC data on the Precision-
Accuracy Instrument Performance form
and its corresponding plot (Section 7 8),
and the key data should be summarized
on the Instrument Performance Sum-
mary form
The si rip charts, the reduced data and
the Precision-Accuracy forms from this
study are filed to be readily available to
the analysts and the laboratory director.
If, at any time, degradation of instru-
ment performance is suspected, then
the original study can be reviewed to
evaluate instrument performance. An
example of a instrument performance
study is in Appendix B.
7.6 Quality Control Program
for Chemical Analysis of
Precipitation Samples
When analytical data are reported, it
is essential to specify the quality. State-
ments about quality should refer to the
particular data set being reported, notto
laboratory analyses in general, so it is
necessary to implement a formal QC
program which indicates the accuracy
and precision of each data point
reported.
Both an internal and an external QC
program should be implemented. In-
ternal QC includes calibration and real-
time control by the analyst, analysis of
special QC samples by the analyst,
review of the data by the laboratory
supervisor and QA coordinator, sched-
uled data checking for transcription
errors by data processing personnel,
and a final review of all QC data by the
QA coordinator before reporting. The
external QA includes analysis of blind
samples received from an agency
external to-the central laboratory.
This section specifies QC samples to
be analyzed, and discusses responsibil-
ities 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.
-------
Jan. 1981
Part l-Section 7.0
and QC review of all data directly after
input to the computer.
7.6.1 Sample Handling in the Labora-
tory
7.6.1.1 Sample Logistics—All samples
received by the laboratory should be
checked in by a receiving clerk who
records the site, date, and other
identification; checks the field data form
against sample labels to identify dis-
crepancies; assigns a laboratory identi-
fication number to the sample, records
the number and the date of arrival on
the data form and in the logbook;
examines the data form and the sample
for certain conditions; and codes the
information on the data form. The
codes, which are basically footnotes
that may be useful later in interpreting
the data, should be stored with the
sample data in the computer (Section
8.0). Table 7-6 suggests information to
be coded. After logging in the samples,
the receiving clerk should refrigerate
them as soon as possible Data forms
received with the samples are kept by
the data clerk.
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
with other required materials to thefield
sites as soon as possible These can be
sent by ground transport since each site
should have a 3-week supply on hand.
As soon as possible after receipt of
samples, aliquots should be poured into
clean 35 ml vials or 10 ml plastic test
tubes for pH and conductivity analyses,
and the vessels should be sealed with
paraffin film. Then the remainder of the
sample should be filtered, as indicated
below (Section 7.6.1.2), and the filtrate
should be sealed in a labeled plastic
Table 7-6. Sample Information to
be Coded
Snow/ice
Mixed: snow/rain; hail/rain
Sample contaminated
Possible sample leakage in shipping
Sampler inoperative — no sample
Insufficient sample for complete
measurement
Rain gauge inoperative
Noticeable suspended particulates
Lid cycling
Field pH and conductivity measured x
days after scheduled sample removal
or end of event
pH/conductivity/temperature meter
inoperative
Sample partially frozen
Unusual condition in area
Duplicate samples
Sequential samples
bottle of appropriate size and refriger-
ated until it is analyzed. Each analyst
should take an aliquot for analysis, and
return the sample to the refrigerator as
soon as possible. If a volume of a sample
is low, the analysis priorities should be
in the following order- pH, conductivity,
SO4=, N03~, CI", NlV, Na+, K", Ca"2,
Mg+2, PQ4~3, and acidity If there is
adequate sample after pH and conduc-
tivity are determined, the sample should
be analyzed for NH4+ before the other
constituents.
After all analyses have been com-
pleted and the results checked, the
sample may be transfered into a 125 ml
polyethylene bottle for storage in a
refrigerator or freezer for 6 mo to 1 yr for
other tests or analyses. The best method
for long-term storage may be to freeze
the samples or to keep them at 4°C
Stability tests for several months
indicate that both the 4°C(16,17) and
the freeze(17) methods are satisfactory.
7.6.1.2 Filtering the Samples—Par-
ticulates in rainwater can contaminate
the sample over a period of time(18);
thus it is necessary to remove them by
filtration as soon as possible after the
sample is received All parameters,
except pH and conductivity, should be
determined on the filtered sample. A
procedure for sample filtering is in the 0
& M manual(8).
7.6.1.3 Scheduling the Analyses—
Once a week, a list of all received
samples for which analysis results have
not been reported should be prepared
from the receiving clerk's logbook, and
copies given to laboratory personnel to
alert everyone to the analyses that are
to be done and to the data yet to be
reported.
7.6.2 Laboratory Documentation
The following documents should be
available to the analyst and the super-
visor, and should be constantly updated.
1. Laboratory Standard Operational
Procedure - detailed instructions
on laboratory and instrument
operations.
2 The Laboratory Quality Assurance
Plan - clearly defined laboratory
QA protocol, including personnel
responsibilities and use of QC
samples.
3. List of In-House Samples - dates
for completion of analysis to allow
the analyst to schedule further
analyses.
4 Instrument Performance Study
Information - information on base-
line noise, calibration standard
response, precision as a function
of concentration, and detection
limits used by analyst and super-
visor to evaluate daily instrument
performance (Section 7.5.3).
5 Quality Control Charts - most
recent QC charts with control
limits for all calibration curve QC
parameters and for aII QC samples;
once a month, update all control
limits to include data from analy-
ses of the previous month; gener-
ate plots of all QC samples and
curve parameters.
6 Data Sheet Quality Control Report
- Generate a QC report after data
for each analysis are input to 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; if necessary,
the supervisor will schedule repeat
analysis of 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 calcu-
lated and plotted by the analyst in
real time (Section 7.6.5).
7.6.3 Traceability of Calibration
Standards
7.6.3.1 NBS Traceability —As com-
mon reference points for laboratories,
NBS traceabilities of chemical gravi-
metric, and thermometric calibration
standards are desirable. As of May
1980, NBS had not provided a rainwater
SRM, however, preparation of an
analyst spike or an audit spike (Section
7.6.4) from NBS-SRM's is not required,
but is recommended
For chemical traceability all calibra-
tion standards must be prepared from
ASC reagent grade salts, and the
accuracy of calibration standard prep-
arations must be checked. With proce-
dures proposed here, accuracy is
checked by running an independently
prepared analyst spike (Section 7.6.4)
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-certified
weights (Section 7.5.1.1). The recom-
mended procedure is to purchase
weights traceable to NBS from a
commercial supplier and to have them
certified by an NBS-approved labora-
tory; it is much more expensive to have
NBS calibrate a set of weights directly.
-------
Part l-Section 7.0
Jan. 1981
For temperature measurements, NBS
traceability is provided by thermometers
calibrated against an NBS-calibrated
thermometer, so each laboratory should
maintain an NBS-calibrated thermom-
eter, and should calibrate thermom-
eters for daily use against the NBS
thermometer (Section 6.3). It is time
consuming but not expensive to have
NBS calibrate a thermometer, NBS will
calibrate at $30.00/pomt (calibration
needed at two points only, 0°C and
25°C); and NBS will visually inspect for
flaws and reject if flaws are found For
NBS calibration, thermometers should
be shipped (not mailed) to:
National Bureau of Standards
Route 270
Quince Orchard Road
Gaithersburg, MD 20760
7.6.3.2 Reference Water Samples—
The QC chemist may dilute and mix
reference water samples to prepare
synthetic rainwater samples for in-
dependent internal QC for an audit spike
sample. The two sources of reference
water/wastewater samples are(19)-
1. EPA Cincinnati - (Quality Assur-
ance Branch, EMSL-Cincmnati
EPA, Cincinnati, Ohio 45268,
513/684-7327 ) QC water/waste-
water samples without charge to
be used as "secondary checks
within laboratory quality control"
prepared solely for internal QC
samples, data not reported to EPA
Samples prepared from ACS rea-
gent grade chemicals are sent as
concentrates in sealed-glass am-
pules; when diluted according to
instructions, should give the calcu-
lated EPA values which are sent in
a separate envelope. Available
samples for rainwater analyses:
MINERAL/PHYSICAL ANALYSES
Na+, K+, Ca++, Mg*\ pH, SO4=, CI",
F", alkalinity/acidity, total hard-
ness, total dissolved solids, and
specific conductance, two concen-
trations
NUTRIENTS
nitrate-N, ammonia-N, Kjeldahl-N,
orthophosphate, and total phos-
phorus, two concentrations
2. Environmental Resource Associ-
ates (ERA) - (120 East Sauk Trail,
Suite 150, South Chicago Heights,
Illinois 60411, 312/755-6060.)
Commercially prepared reference
water/wastewater samples of
known composition from reagent
grade salts; selected samples from
the batch analyzed by ERA and by
three independent laboratories.
7.6.4 Preparation of Analyst's Spikes
and QC A udit Spikes
7.6.4.1 Analyst's Spike—When pre-
paring calibration standards, the analyst
should prepare an analyst's spike from a
different stock solution; concentration
of the spike should be approximately at
the midpoint of the calibration curve,
however, if the majority of samples have
concentrations below the midstandard,
the spike should be prepared within that
range
7.6.4.2 Quality Control Audit Spikes—
For management of laboratory analyses,
the QC chemist should prepare a set of
audit spike samples once a month;
these are simulated rainwater samples
containing all of the major constituents
of rainwater Table 7-7 is a suggested
scheme for the preparation of an audit
spikS sample. In this scheme, stock
standards are prepared by weighing
reagent grade chemicals, and diluting
them to 1 liter. Each stock solution
undergoes a preliminary dilution, dilu-
tion A; each aliquot of dilution A is
diluted to 1 liter (in a volumetric flask) to
obtain dilution B The resulting rain-
water concentrations (last column,
Table 7-7), are approximately the
medians obtained in a recent study of
Eastern United States precipitation
samples(11) Each month, the QC may
vary the final rainwater concentrations
of dilution B by varying the volumes of
dilution A delivered. Dilution A should
be kept refrigerated, especially if nitrate
and ammonium are present
If a QC chemist weighs several
reagents into the same volumetric flask
for the stock standard, the amount of
pipetting would be decreased, but the
possible ratios of the mixed species
would be limited. The level of pipetting
could be reduced by mixing several
species at the dilution A level without
loss of possible ratios of the mixed
species
7.6.5 Real-Time Quality Control Pro-
cedures
Real-time QC procedures, which
stress analyst evaluation of the calibra-
tion curve during analysis and plotting
of one QC data point when each
analysis is set up, are designed to spot
problems during the analysis so that
corrections can be made immediately.
QC data obtained by analysis of special
QC samples (Section 7.6.6) are not
evaluated in real time. Data from
analyses with the automated colori-
meter, ion chromatograph, and atomic
absorption spectrophotometer yield
calibration curve correlation coeffi-
cients (linearities) and detection limits
which should be evaluated in real time.
The control limits for detection limits
and correlation coefficients are to be
statistically established (Section 8.0),
but until sufficient data are available to
calculate these limits, the limits estab-
lished in the instrument performance
study may be used (Section 7.5.3)
7.6.5. / Real-Time Plotting of Analyst
Spike Data—After each instrument is
calibrated, the analyst should immedi-
ately run an analyst spike (Section
7.6.4) as the first QC sample to ensure
that calibration standards were cor-
rectly 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 of the day The
percentage recovery should be calcu-
lated 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.
7.6.5.2 Balance—The analytical bal-
ance should be calibrated daily against
Class-S weights, and the Balance
Calibration Log (Section 6 8.5) should
be completed. The balance should be
zeroed before each use
7.6.5.3 pH Measurement—The pH
meter should be calibrated as indicated
in the O & M manual(8) After the first
calibration, the first sample analyzed
should be the pH electrode reference
solution. The analyst should plot and
evaluate the pH value (Figure 7-1).
Backup electrodes should always be in
the laboratory to check the first elec-
trode(s) if the reading differs from the
previous one by more than ±0.03 unit.
The pH calibration drift should be
evaluated after 20 samples are analyzed;
if the drift is more than ±0.02 pH units,
the analysis should be stopped, and the
meter and electrodes should be checked.
7.6.5.4 Strong Acid and Acidity
Measurements—For strong acid deter-
mination, each day when sample
measurements are begun, three condi-
tioning solutions and an analyst spike
should be measured and calculated
using the linear least squares fit, as
described in the procedure in the O & M
manual(8) If correlation coefficients
from the calculation are less than
0.9990, the indicated problem should
be eliminated. The value of Ve (the
equivalent volume of base added) for
each conditioning solution and the
-------
Jan. 1981
Part l-Section 7.0
Table 7-7. Preparation of a Synthetic Rainwater Sample
Stock3
Dilution A
Primary
Salt
NaCI
KH2PO,
KNO3
CaSO, • 2H20
MgSO, • 7H20
NHtNOz
NaNO3
HiSOJO. 1Nf
Weight g
01648
0.0904
0.2586
0. 1074
0.2536
0.4437
0.3697
—
Species
Cl
P0<
K
Ca
Mg
NH,
Na
H
fjg/m/
1000
2501
1000
25.00
25.00
10000
1000
101 0
Secondary
Species
Na
K
NO3
SO4
SO,
/VO3
N03
SO,
fjg/m/
64.83
10.27
58.80
59.92
98.83
343.8
269.7
9606.0
MLb
20
4
10
—
12
20
10
10
Primary
Species
Cl
PO,
K
—
Mg
NHt
Na
H
fjg/ml
2000
1 000
WOO
—
3000
2000
WOO
10 10
Secondary
Species
Na
K
NO3
—
SO,
NO3
NO3
SO*
fjg/m/
12.97
04108
5880
—
11 86
68.76
26.97
480.3
Dilution B*
Final Ion Concentrations
Salt
NaCI
KH2POi
KNO3
CaSO, • 2H2O
MgSO, • 7H20
NH^NOa
NaNO3
H2SO,fO INjf
Vol Oil A
(ml)
13
8
6
5C
10
14
10
5
Primary
Species fjg/ml
Cl
PO,
K
Ca
Mg
NH,
Na
H
0260
0008
0060
0.125
0030
0280
0.100
00505
Secondary
Species fjg/ml
Na
K
NO3
SO,
SO4
/VO3
N03
SO,
0 169
0004
0035
0300
0 119
0963
0270
2.402
"Dilution B"
Species Cone fog/ml)
H
Cl
PO,
NO 3
SO4
Na
K
Ca
NH,
0.050
.260
008
1.27
2.82
0269
0.064
.125
.280
a Concentrations are based on a final dilution volume of 1 liter
bML gives the volume to take to dilute to 100 ml
c Volume (ml) of stock
"Either standardized in the laboratory or purchased from a commercial supplier
analyst spike percentage recovery
should be plotted and obtained as
indicated in theO &M manual(8). Atthe
end of the day, an analyst spike and a
conditioning solution sample should be
analyzed. The initial conditioning solu-
tion potential for each sample should be
within 1 2 mv of the potential for the
conditioning solution. According to the
0 & M manual(8), an analyst spike
should be analyzed before and after
samples are determined and these
values should be plotted daily (Section
7.6.5.1).
7.6.5.5 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 3x10~4M KCI or
the pH electrode reference solution. The
analyst should calculate the conductiv-
ity using the first calibration curve, and
then plot and evaluate the value of the
conductivity of this reference sample
(Figure 7-1).
7.6.5.6 Automated Colorimetric Anal-
ysis (Technicon)—This instrument should
be set up, and the baseline noise and
instrument response should be evalu-
ated by comparison with data from the
instrument performance study. Any
problem noted should be investigated
The instrument should be calibrated as
described in the 0 & M manual(8). For
real-time QC, the first calibration curve
should be checked for linear response
and adequate detection limit by using a
least squares fit of the first calibration
curve and by calculating 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 (Figure 7-1) 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
7.6.5.7 Ion Chromatographic Analy-
sis (Dionex)—Calibration procedures
are in the 0 & M manual(8). 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 (Figure
7-1).
7.6.5.8 Atomic Absorption Analy-
sis—Atomic absorption calibration
procedures are in the O & M manual(8)
For real-time QC, the first calibration
curve should be analyzed, and the linear
least square fit of response vs concen-
tration should be calculated. The
correlation coefficient should beO.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.
7.6.6 Analysis and Evaluation of
Quality Control Samples
Each day (each analytical run) or at
least after every 50 samples, a reagent
blank, an old sample, an analyst spike,
and an audit spike should be analyzed.
For most procedures discussed (except
the ion Chromatographic, the Gran
strong acid and the acidity analyses), an
analytical run may include more than 50
samples so QC is about 8%. For ion
Chromatographic analysis, one analyt-
ical run may include 20 samples; 4 QC
samples and 20 samples make the QC
16.7%. The Gran strong acid procedure
is relatively slow, so fewer than 25
samples/day can be determined; recom-
mended are only two analyst spikes and
one audit spike as QC samples for a QC
of 1 2%.
7.6.6.1 Reagent Blank—This deion-
ized water QC sample, which is sub-
jected to the same preparation procedure
as the routine samples being analyzed,
should be analyzed to check for random
contamination which may have oc-
curred in sample preparation or analysis.
7.6.6.2 Old Sample—This randomly
chosen, previously analyzed QC sample
(if no sample degradation has occurred)
-------
Part I-Section 7.0
10
Jan. 1981
100
Percent
Recovery
90
80
Upper Control Limit
Average % Recovery Monthly QC
Report Mar 80
Lower Control Limit
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Order of Analysis
Figure 7-1. Analyst spike plot for SO^analysis spike stock solution prepared
4 Jan 1980; expected cone. 2.00 fjg/ml.
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
7.6.6.3 Analyst Spike—This QC sam-
ple 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 preci-
sion of analysis. The analyst spike
should be analyzed at the beginning of
the run; results should be calculated
and plotted in real time (Section 7.6 5)
7.6.6.4 Audit Spike—This QC sample
is prepared by the quality control
chemist (Section 7.6.8) and it is ana-
lyzed as an unknown (blind) by the
analyst. The data are reported to the QC
chemist who calculates percentage
recovery and reports the value to the
analyst and the laboratory director. The
purpose of this QC sample is to assess
data quality independently of analyst
judgment.
7.6.7 Evaluation of QC Data
7.6.7.1 QC Data Handling—If a com-
puter is available, data for QC samples
are calculated as they are being input
and compared to control limits estab-
lished in the most recent monthly QC
chart (Section 7.6.2). After input, the
Data Sheet QC Report (Section 7.6.2) is
printed by the computer If no computer
is available, this procedure can be done
manually The report on the perform-
ance of the QC samples should be given
to the laboratory supervisor for evalua-
tion (Section 7 6.7 2) to see if the data
are acceptable for reporting or if re-
analysis is necessary
When the data form is input to the
computer, the audit spike data are also
calculated, these spike data should be
given to the QC chemist, who tabulates
the data and calculates the percentage
recoveries The QC chemist routinely
gives the audit spike data to the
laboratory director, who reviews this
independent audit data before reporting
the analytical data to the project
manager Once a month, the QC
chemist performs a QC audit to review
the audit data with the analyst.
Once a month, all QC sample data and
calibration curve parameters obtained
during the month are combined with all
previous data for the same parameters;
the data are plotted, and the 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.
7.6.7.2 Quality Control Reports and
Guidelines for Initiating Corrective
Action—After input of data to the
computer, the Data Form Quality
Control Report (Section 7.6.2) is given to
the laboratory supervisor; this report
flags 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
transcription errors from stripchart to
data form or from data form to computer,
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 new The new QC should replace the
old in the computer QC data base If the
out-of-limits conditions are not due to
transcription errors, another explana-
tion 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 be-
tween the questionable calibration
curves should be reanalyzed.
If only one of several QC samples is
out-of-limits, an explanation should be
sought but if one cannot be found, no
action is needed The supervisor may
assume that the problem was with the
particular QC sample itself, but may
retain the out-of-limits data in the QC
data base
If several QC samples are out-of-
hmits but an explanation is notfound, all
samples analyzed with the QC samples
(between bracketing calibration curves)
should be reanalyzed.
In any case, an out-of-limits QC
sample requires evaluation and an
explanation by the supervisor The
explanation may be notes 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 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 gener-
ated Since the supervisor has already
evaluated out-of-limits conditions in the
QC reports, all out-of-limits conditions
should have been explained or elimi-
nated (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 If at any
time, eight successive values appear on
one side of the line, a bias in the data
exists, the analyses should be stopped,
an explanation sought, and any action
taken should be noted on the plot
7.7 Evaluation of Laboratory
Performance
The QC procedures (Section 7.6)
stress the supervisor's roles in evaluat-
ing QC and in scheduling reanalyses
until data are acceptable for reporting.
This section discusses the QC chemists'
-------
Jan. 1981
11
Part l-Section 7.0
roles in evaluating laboratory perform-
ance by independent QC checks and by
external audits
7.7.1 Independent Internal Quality
Control
The QC chemist is to ensure that QC
procedures are implemented and to
provide independent judgement on the
quality of the data generated in the
laboratory. Quality of data is checked by
introducing audit spikes and by synthet-
ic rainwater samples inconspicuously
added to the sample stream (Section
7.6.64). Audit spikes may be prepared
by the QC chemist from NBS-SRM's
(Section 7.2.3), from reagent grade salts
or from reference waste/wastewater
samples (Section 7.6 3.2). Whateverthe
source, a reference rainwater with a
known composition in the range of
typical rainwater samples is submitted
to the laboratory as a blind audit sample
to be analyzed routinely, and the data
reduction is done by the individual who
handles the data Finally, the data are
flagged in the data set by the QC
chemist who checks for out-of-limits
conditions and then reports the data to
the laboratory director
7.7.2 External Performance Audit
Laboratories analyzing rainwater
samples are recommended to partici-
pate in an interlaboratory comparison at
least two times a year. An external
agency (e g , EPA or USGS) should
provide'to each laboratory audit samples
which contain constituents of interest
in concentrations typically found m
rainwater samples, these samples
should arrive in the laboratory as blind
samples, and exactly the same proce-
dures should be applied to them as in
routine analysis When audit data are
reported with regular sample data, the
external agency informs the laboratory
of the blind sample identities These
audits can be used by the QA coordina-
tor to assess analytical accuracy (Sec-
tion 9 2)
7.7.3 System Audit
Once a year, the project manager or a
designee should document current
standard operating and QA procedures
by completing the Laboratory Ques-
tionnaire (Section 7.7.4) in cooperation
with the laboratory supervisor and the
QC chemist.
The questionnaire will be submitted
to the laboratory director 4 to 6 weeks
before the interview is conducted to
amplify responses to various questions.
During the interview, written proce-
dures, QC charts, and audit spike
recovery data may be examined. The
auditor summarizes all findings in a
System Audit Report.
7.7 A Laboratory Questionnaire
In the laboratory questionnaires, the
R and G denote "recommended strong-
ly" and "guidelines", respectively
-------
Part l-Section 7.0 12 Jan. 1981
Preliminary Questionnaire
Questionnaire Completion Date
Laboratory
Street Address:
Mailing Address (if different from above):
City:
State. . Zip
Laboratory Telephone No.'Area Code No
Agency Director. „__
Quality Assurance Officer
(Quality Control Chemist)
Questionnaire Completed By (If more than one, please indicate which section(s) of the questionnaire completed):
-------
Jan. 1981 13 Part l-Section 7.0
Introduction-Laboratory Operations
Standard Operating Procedures /SOP) Yes No
1 Has an off/cat agency Standard Operating Procedures Manual been written?
2. Is the SOP Manual followed in detail?
3. Does it contain all quality control steps practiced?
4. Does each analyst have a copy at his/her disposal?
5. Has a methods validation study been completed for each analysis?
6 Are plots of instrumental accuracy and precision available for every analysis?
7. Are detection limit data tabulated for each analysis?
-------
Part l-Section 7.0 14 Jan. 1981
Laboratory Personnel
% time
Years presently
Academic Training experience spent
ESS MS in rainwater in rainwater
Position Name HS BA MA Ph.D Special Training analysis activities
-------
Jan. 1981
15
Part l-Section 7.0
Laboratory Staff Training
1. Is a formal training program used? Yes No_
If yes, is it •
Agency-wide Yes No.
In-house Yes No.
On-the-job training Yes No_
2 Training outside local agency (courses attended).
Course
description
or title
Who attended
Course
length
Course
type*
Year of
attendance
*State, Federal, College, University, Other.
3 Publications routinely received and used by agency.
4. Do you feel that your agency training is adequate?
If not, what would be required to make it adequate?
Yes
No
-------
Part l-Section 7.0
16
Jan. 1981
Laboratory Facilities
Item
A vailable
Yes No
Comments
{where applicable, cite system,
QC check, adequacy of space)
1. Filter room or desiccator
(either required for TSP)
maintained at 15°-35° C
and 50% relative humidity
2 Gas
3 Lighting
4. Compressed air
5 Vacuum system
6 Electrical services
7. Hot and cold water
8. Laboratory sink
9. Ventilation system
10. Hood space
11 Cabinet space
12 Bench-top area
(cite linear ft.)
13 Lab space
(cite linear ft.)
14 Lab space utilized for
offices (cite sq ft)
15 Office space
(cite sq. ft.)
16. Storage space
(cite sq. ft J
17. Shared space
-------
Jan. 1981 17 Part l-Section 7.0
Laboratory Equipment
Items
n of Equipment
units Make Model
Condition/age
Good-fair-poor
Ownership
Air Water
% of use time
used in rainwater
programs
Balance
analytical
Vacuum Filtration
apparatus
NBS Calibrated
thermometer
Dessicator
Ion
Chromatograph
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
-------
Part l-Section 7.0 18 Jan. 1981
Laboratory Operation
A. RECEIVING CLERK
(Name)
Yes No
1 Are all chemicals dated on receipt and thrown away when shelf life is exceeded? (G)
2 Are all samples received by the laboratory logged into a bound notebook? (R)
3 Are all samples filtered before analysis7 (R)
4. Are all samples stored in the refrigerator between analyses? (Rj
B. GRA VIMETRIC MEASUREMENTS
1 Is the analytical balance calibrated daily with weights traceable to NBS? (R)
2 Is the "Balance Calibration Log" kept up to date? (R)
3 Is factory service scheduled and the "Balance Factory Service Form" completed? {R)
Date next service is due
C. ANALYST - pH,
(Name)
Yes No
1. Does the analyst have his/her own copy of the standard operating procedures? (R)
2. Does the analyst have his/her own copy of instrument performance data? (R)
3. Does the analyst have his/her own copy of safety instructions? (R)
4. Does the analyst have his/her own copy of the latest monthly QC plots? fR)
5. Is the analyst aware of the most recent control limits? (R)
6. Does the analyst have a copy of the most recent list of samples in-house to be analyzed? -(R)
Date of list
7. Are all solutions properly labelled? (R)
8. Has a "pH Meter/Electrode Acceptance Test Form" been completed for the meter and electrode currently
in use? (RJ
9. Is the "pH Acceptance Test Summary Form" kept up-to-date? (R)
10. Is the "Field Quality Control Audit Sample Report" completed on a monthly basis? (R)
//. Are rinse and measurement tubes poured for buffers and samples? (R)
72 Is the pH meter calibrated before and after samples are analyzed? (R)
13 Is the pH meter recalibrated after every set of 20 samples? fR)
14. 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? (G)
15 Is the pH electrode reference solution analyzed first and are the results plotted real time? (G)
16 Are the following control samples analyzed with each run?
Blanks (R)
Old Samples (R)
Audit Spike (R)
17. Does the analyst review the quality control data sheet output by the data clerk, and then decide whether
or not to re/ease data for reporting? (G)
18 Are electrodes stored as recommended by the manufacturer? (R) .
19. Are electrodes checked and filled, if necessary, before each analysis? (R)
-------
Jan. 1981 19 Part l-Section 7.0
Laboratory Operation
D. ANALYST - GRAN STRONG/TOTAL ACID
(Name)
Yes No
1 Does the analyst have his/her own copy of the standard operating procedures? (R)
2. Does the analyst have his/her own copy of instrument performance data? (R)
3. Does the analyst have his/her own copy of safety instructions? (R)
4 Does the analyst have his/her own copy of the latest monthly QC plots? (R)
5. Is the analyst aware of the most recent control limits? (R)
6. Does the analyst have a copy of the most recent list of samples in-house to be analyzed? (R)
Date of list
7. Are all solutions properly labelled? (R) —
5. Has a "pH Meter/Electrode Acceptance Test Form" been completed for the meter and electrode currently
in use7 (R)
9. Are micropipets calibrated on at least a weekly basis or whenever the tip breaks? (R)
10 Are repipets calibrated on a weekly basis? (R)
// Is the stock 1 0 N NaOH standardized each month against potassium acid phthalate? (R)
12. Is solution temperature carefully monitored during analysis to see that it changes by less than 0.1°C? (R)
13 Are conditioning solution data and analyst spike data calculated and plotted real time? (G)
14. Are the H* function correlation coefficients of these data examined to ensure that they are greater than
0.9990? (G)
15. Are the following analyzed each day?
Three conditioning solutions and an analyst spike initially. (R)
An analyst spike and a conditioning solution at the end of the analysis. (R}
An audit spike. (R)
16. Does the analyst review the quality control data sheet output by the data clerk, and then decide whether
or not to re/ease data for reporting? (G)
77. Are electrodes stored as recommended by the manufacturer? (R)
18. Are electrodes checked and filled if necessary before each analysis? (R)
-------
Part I-Section 7.0 20 Jan. 1981
Laboratory Operation
F. ANALYST - TECHNICON
(Name)
Yes No
1 Does the analyst have his/her own copy of the standard operating procedures? (Rj
2. Does the analyst have his/her own copy of instrument performance data? (R)
3. Does the analyst have his/her own copy of safety instructions? (R)
4. Does the analyst have his/her own copy of the latest monthly QC plots? (R)
5. Is the analyst aware of the most recent control limits? (R)
6. Does the analyst have a copy of the most recent list of samples in-house to be analyzed? (Ft)
Dafe of list
7. Are all solutions properly labelled? (R)
8. Is the "Standard Preparation Form" completed when new stock standards are prepared? (R)
9 Are dilute calibration standards prepared fresh daily? (R)
10. Is the analyst spike prepared fresh daily from an independent stock? (R)
11 Is the calibration curve at least a five point curve? (R)
12 Is the first calibration curve of the day checked for detection limit and linearity? (R)
13. Are the analyst spike data calculated and plotted real time? (G)
14. Is each new calibration curve checked to see that instrumental response changed less than 5%? (R)
15. Are the following control samples analyzed with each run?
Blanks (R)
Old Samples (R)
Analyst Spikes (R)
Audit Spikes (R)
16 Does the analyst review the quality control data sheet output by the data clerk, and then decide whether
or not to release the data for reporting? (G)
17 Is water pumped through all lines daily before and after analysis? (G)
18 Are pump tubes changed at least once per three days? (G)
19 Is the pump cleaned when the pump tubes are changed? (G)
2O Is soap solution pumped through all lines once per week? (G)
21 Is the flowcell cleaned with a sulfunc acid-potassium dichromate solution once per month? (G)
22. Is the pump oiled once per three months? (G)
Date of last service
23. Is the colorimeter mirror assembly and color fitter cleaned and the alignment optimized once per
three months? (Gl
Dafe of last service
-------
Jan. 1981 21 Part l-Section 7.0
Laboratory Operation
G. ANALYST - DIONEX
(Name)
Yes No
1. Does the analyst have his/her own copy of the standard operating procedures'' (R)
2. Does the analyst have his/her own copy of instrument performance data? /Ft)
3. Does the analyst have his/her own copy of safety instructions? (Ft)
4. Does the analyst have his/her own copy of the latest monthly QC plots? (Ft)
5. Is the analyst aware of the most recent control limits? (R)
6. Does the analyst have a copy of the most recent list of samples in-house to be analyzed? (R)
Date of list
7. Are all solutions properly labelled? (Ft)
8. Is the "Standard Preparation Form" completed when new stock standards are prepared? (Ft)
9 Are dilute calibration standards prepared fresh weekly? (R)
10. If manual techniques are used, are samples and eluent prepared fresh daily from the same
concentrated stock buffer? W
77. Is the analyst spike prepared from an independent stock? (R)
12. Is the calibration curve at least a four point curve for each analytical range? (R)
13. Is the first calibration curve of the day checked for detection limit and linearity? (R)
14. Are the analyst spike data calculated and plotted real time? (G)
15. Are the following control samples analyzed with each run?
Blanks (R)
Old Samples {R)
Analyst Spikes (R)
Audit Spikes (R)
16 Does the analyst review the quality control data sheet output by the data clerk, and then decide whether
or not to release the data for reporting? (G)
17. Is the drip tray examined daily for reagent spills, and are spills cleaned up daily? (G)
18. Are pumps oiled once per week? (G)
19. Is the anion precolumn cleaned once per month with 0.1 M NazCOy? (G)
2O. Is the Br~, NO3 resolution checked once a month and documented with a "Dionex Resolution Test Form"?(R)
-------
Part l-Section 7.0 22 Jan. 1981
Laboratory Operation
H. ANALYST -AA
(Name)
Yes No
1. Does the analyst have his/her own copy of the standard operating procedures? (R)
2. Does the analyst have his/her own copy of instrument performance data? (R)
3. Does the analyst have his/her own copy of safety instructions7 (R)
4. Does the ana/-,it have his/her own copy of the latest monthly QC plots? (R)
5 Is the analyst aware of the most recent control limits? (R)
6. Does the analyst have a copy of the most recent list of samples in-house to be analyzed? (R)
Date of list
7. Are all solutions properly labelled? (R)
8. Is the "Standard Preparation Form" completed when new standards are prepared? (R)
3. Are dilute calibration standards prepared fresh monthly? (R)
10. Is the analyst spike prepared from an independent stock? (R)
11. Is the instrument allowed to warm up at least 15 minutes with the flame on before the final
wavelength adjustment is made? (RJ
12. Is the calibration curve at least a five point curve? (R)
13. Is the first calibration curve of the day checked for detection limit and linearity? (R)
14 Are the analyst spike data calculated and plotted real time? (G)
15 Is each new calibration curve checked to see that instrumental response changed less than 5%? (R)
16. Are the following control samples analyzed with each run?
Blanks (R)
Old Samples (Rl
Analyst Spikes (R)
Audit Spikes (R>
1 7. Does the analyst review the quality control data sheet output by the data clerk, and then decide whether
or not to re/ease the data for reporting? (G)
-------
Jan. 1981 23 Part l-Section 7.0
Laboratory Operation
I. DA TA CLERK
(Name)
Yes No
1. Does the data clerk do a 100% QC check for accuracy of data input to the computer? (R) —
2. Does the data clerk routinely report quality control data sheet information to the analyst? (R) —
3. Does the data clerk submit quality control data sheet information to the lab manager along with the analytical
data to be reported? W —
4. Does strip chart reduction by on-line electronic digitization receive at least 5% manual spot checking? (R) —
5 Are control charts or equivalent checks (e.g., computer calculated range limits or regression charts) current
and available for inspection? (R) —
6. Are provisions made for data storage of at least 3 years for all raw data, calculations, quality control data
and reports? (R) —
7 Do laboratory records include the following information:
a. Sample identification number (R) —
b. Station identification IR) —
c. Sample type (R) —
d. Date sample received in laboratory (Ft) —
e. Time, date and volume of collection (R) —
f. Date of analysis (R) —
g. Analyst (R> —
h. Results of analysis (including raw analytical data) (R) —
/. Receptor of the analytical data (R) —
8. Are ram gauge chart data for event times and amount checked? (G) —
9. Does laboratory follow cham-of-custody procedures from sample receival to discard? (G) —
10. Are computer printouts and reports routinely spot checked against laboratory records before data
are released? (R) —
11. Are manually interpreted strip chart data spot checked after initial entry? (R) —
12. Are minimum detection limits calculated by an approved method such as either Hubaux- Vox (Appendix A) or
baseline standard deviation? (R) —
13 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 ) (R) —
14 Are control charts, regression charts or computer QC data bases up-to-date and accessible? (R)
15. Has the data handling and reduction system been examined by legal counsel to determine the soundness of
the system in possible litigation? (G) —
-------
Part l-Section 7.0 24 Jan. 1981
Laboratory Operation
J QUALITY CONTROL CHEMISTRY
(Name)
Yes No
1. Does the QC chemist have his/her own copy of the standard operating procedures? (Ft)
2 Does the QC chemist have his/her own copy of instrument performance data? (R)
3 Does the QC chemist have his/her own copy of safety instructions? (R)
4. Does the QC chemist have his/her own copy of the latest monthly QC plots? (R)
5. Is the QC chemist aware of the most recent control limits? (R)
6 Does the QC chemist prepare a blind audit spike once per month? (R)
7 Does the QC chemist routinely review and report blind audit spike data to the laboratory manager? (R)
8 Does the QC chemist update control limits and obtain new control chart plots once per month? (R)
K LABOR A TOR Y TECH NIC I A N
(Name)
1 Are all containers washed before they are sent to the field? (R)
2. Is the conductivity of the last rinse water measured for 70% of the washed containers? (R)
3 If the conductivity of the rinse is greater than 2 /jmhos/cm, is the container rinsed further? (R)
4. After the containers and lids are dried are the containers capped immediately? (R)
5 Are precautions taken not to touch the inside of the containers and lids? (R)
6. Are all samples stored in a refrigerator when not being analyzed? (R)
7 Are precautions taken not to breath on sample? (R)
8 After completion of the analyses, are the samples stored in a refrigerator for a time period of at least
six months? (R)
LABORATORY MANAGER _.
{Name)
1 Does the laboratory manager have his/her own copy of the standard operating procedures? (R)
2 Does the laboratory manager have his/her own copy of instrument performance data? (ft)
3 Does the laboratory manager have his her/own copy of safety instructions? (R)
4 Does the laboratory manager have his/her own copy of the latest monthly QC plots? (R)
5 Is the laboratory manager aware of the most recent control limits? (R)
6 Does the laboratory manager review the following before reporting data-
a. The data itself? (R)
b. The quality control data sheet with analyst notes? (G)
c. The quality control chemist blind audit data report? (R)
d The ion summation ratios for the data? (R)
e The calculated vs measured sample conductivity? (G)
-------
Jan. 1981
25
Part l-Section 7.0
7.8 Data Forms
Forms for recording laboratory
activities, including calibrations and
QC procedures, are mentioned
throughout Section 7 On the
following pages, blank data forms are
provided for the convenience of the
manual user. Many of these forms are
taken or adapted from EPA forms and
from other sources. The title is at the
top of each, as is customary for a data
form To relate the form to the text, a
form number (e g., 1.1 77.3 2) is in the
lower right-hand corner; the 1.1 means
form number 1, version 1, and the 7.3.2
means Section 7.3.2. A revision of the
form would change the number to
1 2/7.3 2, or form 1, version 2, and so
forth. The numbers and titles of the
forms listed herein are listed below:
Form Number
1.1/7.4.1
2 1/7.4.2
3.1/7.5 1 1
4 1/75.1.1
5.1/7.53
6.1 /753
71/753
8 1 /7.53
9 1/76
10 1/7.6
Title
Field Conductivity
Standard Form
Field pH Electrode
Test Solution
Certification of
Weights to NBS Form
Balance Calibration
Log
pH, Conductivity and
Gran Strong Acid
Instrument Perform-
ance Form
Precision Accuracy
Instrument Perform-
ance Form
Precision Accuracy
Instrument Perform-
ance Plot
Instrument Perform-
ance Summary Form
Most Recent Monthly
Control Limit
Audit Spike Recovery
Data
7.9 References
1 Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems, Vol. II - Ambient Air Specific
Methods. EPA - 600/4-77-027a,
May 1 977,. p 3-4 of 2.0.
2. Handbook for Analytical Quality
Control in Water and Wastewater
Laboratories, EPA - 600/4-79-
019, March 1979, p 2-1 to 2-6.
3. Applebaum, S.B., and G.J. Crits,
"Producing High Purity Water,"
Industrial Water Engineering (Sept./
Oct. 1964); Karamina, N.A., Amer.
Lab. 8(3), 24(1976).
4. Reagent Chemicals, American
Chemical Society Specifications,
5th Edition, American Chemical
Society, Washington, DC (1974).
5. "Standard Methods for Prepara-
tion, 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)
6. Standard Methods for the Exam-
ination of Water and Wastewater,
13th Edition, American Public
Health Association, New York
(1971)
7. Reichgott, M., "Organic Coatings
and Plastic Chemistry," Vol. 41,
1979, Paper presented at 178th
National Meeting of the American
Chemical Society, Washington,
DC. Sept 9-14, 1979.
8. Operations and Maintenance
Manual for Precipitation Mea-
surement Systems, United States
Environmental Protection Agency,
Research Triangle Park, N.C , (in
press).
9 United States Pharmacopeia, United
States Pharmacopeial Convention
Inc , Rockville, MD., XIX, 644
(1975)
10. Peffer, E.L. and G.C. Mulligan,
"Testing of Glass Volumetric
Apparatus," NBS Circular 434,
National Bureau of Standards
(1941).
11. Electric Power Research Institute,
"Acid Precipitation in the Eastern
United States," Contract Nos.
RP1376-1 and RP1630-2, Rock-
well International Environmental
Monitoring and Services Center
(1978-1980).
12. Gran, G., Analyst 77, 661 (1952).
13. Method 350.1 from Methods for
Chemical Analysis of Water and
Wastes, EPA-600/4-79-020, March
1979.
14. Method 365.4 from Methods for
Chemical Analysis of Water and
Wastes, EPA-600/4-79-020, March
1979.
15. Hubaux, A., and G. Vos, Anal.
Chem., 42, 849(1970).
16 Peden, M.E., and LM Skowron,
"Ionic Stability of Precipitation
Samples, Atmos Environ., 12,
2343(1978).
17 Rothert, J.E., Battelle Pacific
Northwest Laboratories, Richland,
Washington, MAP3S Program,
private communication.
18. Peden, M.E., LM. Skowron and
F.F. McGurk, "Precipitation Sample
Handling, Analysis, and Storage
Procedures," United States DOE,
Pollutant Characterization and
Safety Research Division, Contract
No. EY-76-S-02-1199, Research
Report 4, COO-1199-57 (1979).
19. Manual on Industrial Water and
Industrial Wastewater, 2nd Ed..
ASTM Special Publication 148-H,
American Society for Testing and
Materials (1965), p869.
-------
Part l-Section 7.0 26 Jan. 1981
Field Conductivity Standard Form
Date of Preparation of
O.IM KCI Stock Solution __
Date of Preparation of
Dilute Field Standard .
(Analyst Signature)
Laboratory Analysis Before Shipment to the Field
1
2
3
Average Cond
± Std Dev
Laboratory Values After Use In The Field.
Date of Lab Lab Value
Field Site # Analysis (/jmhos/cm)
QA Manual for Precipitation Measurement 1.1/7.4.1
-------
Jan. 1981 27 Part l-Section 7.0
Field pH Electrode Test Solution
Date of Preparation of
Stock Solution
(Analyst Signature)
Volume of Stock Taken
Final Dilution Volume of
Field pH Electrode Test Solution
Laboratory Analysis Before Shipment to the Field
Conductivity pH Ave. ± Std. Dev.
1.
2.
3.
Laboratory Analysis of Aliquots Returned From the Field
Field pH Values Lab Values After Return
Site # Ave. ± Std. Dev. (data) Date of Analysis pH Cond. Anal. I nit.
QA Manual for Precipitation Measurement 2.1/7.4.2
-------
Part l-Section 7.0
28
Jan. 1981
Date of Certification
Weight Set Serial #
Certification of Working Weights to NBS Form
Balance 0
NBS Jka
NBS 5kg
Test 1 kq
Test Ska
•#• * *
Balance 0
NBS 1kg
NBS Skg
Test Ika
Test 5kg
** *
Balance 0
NBS 1kg
NBS Skg
Test 1kq
Test Skg
Summary
Average = Standard Deviation
Balance Cl
NBS 1kg
NBS Skg
Test 1 kg
Test Skg
(Analyst Signature)
Balance Cl
NBS 1kg
NRS Skg
Test 1kg
Test Skg
* **
Balance 0
NBS 1kg
NBS Skg
Test 1kg
Test Skg
Balance 0
NBS 1kg
NBS Skg
Test 1 kg
Test Skg
QA Manual for Precipitation Measurement
3.1/7.5.1.1
-------
Jan. 1981
29
Part l-Section 7.0
Balance Calibration Log
Balance
ID
Number
Date
Test Weight #/
Known
Mass
Mass
Found
Test Weight #2
Known
Mass-
Mass
Found
Test Weight #3
Known
Mass
Mass
Found
Analyst
Initials
QA Manual for Precipitation Measurement
4.1/7.5.1.1
-------
Part l-Section 7.0
30
Jan. 1981
pH. Conductivity, and Gran Strong/Total Acid Methods Validation Form
Conductivity:
Date of Test.
Date of Solution Prep.
Concentration
HCI (Ml
HzCr
W6
W'5
10~*
W'3
Expected
Conductivity 1
limhos/cm
1
0.43b
4.26"
42.44"
421.0"
Conductivity Found Qf//
2 3 4 5 Avg Dev. Recov. Std. Dev.
pH
Date of Test
Date of Solution Prep.
pH Found
Date of4M KCI Solution Prep.
Concentration
HCI (M)
10~e
10~5
10-'
Expected
pH
7.08
6.08
5.08
4.08
3.08
Std.
12345 Avg. Dev.
Calculated Acid Concentration From pH(above):
Date of Test
Date of Solution Prep.
Concentration
HCI IN)
ro~6
1O~5
Acid Cone. Obtained from pH (N)
1 2 3 4 5 Avg
Dev.
%
Recov
Std. Dev.
J0~
Gran Strong Acid:
Date of Test
Date of Solution Prep.
Concentration
HCI (N)
Concentration Obtained (N)
Avg
Std.
Dev.
%
Recov.
Std. Dev.
J0~6
JO'5
10~*
10~3
"An aliquot prepared as the samples, but containing no acid.
b Corrected for conductivity of water
QA Manual for Precipitation Measurement
5.1/7.5.3
-------
Jan. 1981
31
Part l-Section 7.0
Precision-Accuracy Instrument Performance
Cone.
Taken
(/jg/ml)
Curve"
Parameters
Slope
Intercept
Error
Corr. Coef
Det. Limit
Instrumental Response
Curve 1
Curve 2
Curve 3
Curve 4
Curve 5
Spurlfx:-
Hate-
Scale*' .,
(Analyst -Signature)
Curve Parameters Overall0
Slope:
Intercept.
Error:
Correlati
Detect/or
v> C.np.f:
) Limit:
Average +
Standard Deviation
Control Limif
Regress/on Concentration11
Cone.
Taken
f/jg/ml)
Curve
1
Curve
2
Curve
3
Curve
4
Curve
5
A verage
Cone.
Found
Standard
Deviation
%
Recovery
% Std.
Dev.
B For Dionex, indicate approximate fjmho/cm full scale f/jmho/cm scale x volts on recorder); ForAA, indicate scale expansion;
For Technicon, indicate method and flowcell length.
b The curve parameters are calculated using the linear squares fit of Appendix C
c From a linear least squares fit of all data points
a This is the average value + 3x(standard deviation)
e This is calculated from the instrumental responses given above, using the "Curve Parameters Overall".c
QA Manual for Precipitation Measurement
6.1/7.5.3
-------
Part l-Section 7.0
32
Jan. 1981
Percentage ~
Recovery '•
Figure 1.
Percentage
Standard
Deviation
Precision -Accuracy Instrument Performance Plot
Species plot
I • j I ' ' • — . [ I >——=*=
--••• • ' : " ;•(••"!• • ;•"•;•: i " ;•• I" '• !. . : ' r ' ; ! : . I .• -\ < • • .i . I
= r~~ — T ! ' ! i i li! ' : : =F= i !
! | . . j • i : ' : !; i ; i—:.. ;.:.:. i.' . : , ===£
Concentration ([ig/ml)
Percentage recovery as a function of concentration
, j_ . 1 — -i 1 — ;• I : — i — | • i . -\ — |
i ' ' - , i i • i i , i i ! ' i . ' -i
! : ! : ; ! : 1 •.;..•, I !—;....! ; ! ' -i— i • . • I
1 1 -T 1 . - : -J— . . - -^==^~EEEl
Concentration (fig/ml)
Figure 2. Percentage standard deviation as a function of concentration
QA Manual for Precipitation Measurement
7.1/7.5.3
-------
Jan. 1981
33
Part (-Section 7.0
Instrument Performance Summary Form
Species
Date of
Test
Concentration Above
Which 10% Precision
is Obtained fag/ml)
A verage Detect/on
Limit ± Standard
Deviation f/jg/m/J
A verage Detect/on
Limit x 3x (Standard
Deviation) (fjg/ml)
Analyst
Initials
QA Manual for Precipitation Measurement
8.1/7.5.3
-------
Part l-Section 7.0
34
Jan. 1981
Most Recent Monthly Control Limits
Full Date Curve Dup Old Analyst Spike
Anal. Scale Control Curve Del Sample* Sample Blank
Instrument Species Rangett Std. Cone Limits Est. Error Limit Diff. Diff. Magnitude LCL UCL
Dionex C/~ 1
NO3
Technicon
Atomic
Absorption
Gran Strong
Acid
Gran Total
Acid
pH
Conductivity
A/a+
AT
If
NHS 1
PO
Mg" 1
2
H+
a/f duplicate field samples are taken
QA Manual for Precipitation Measurement
9.1/7.C
-------
Jan. 1981 35 Part l-Saction 7.0
B Audit Spike Recovery Data* Month-
Cone Date of Cone. % Date of Cone. % Date of Cone %
Instrument Species Taken Analysis Found Rec Analysis Found Bee Analysis Found Rec
Dionex Cl
/V03
Dionex SOI
Technicon
Technicon PCT*
B Indicate data points out of limits by * in % Rec. Column.
QA Manual for Precipitation Measurement 10.1/7.6
-------
Part l-Section 7.0 36
Jan. 1981
Audit Spike Recovery Data3 (Cont'd) Month.
Cone. Date of Cone. % Date of Cone % Date of Cone %
Instrument Species Taken Analysis Found Rec Analysis Found Rec Analysis Found Rec
Atomic /Va*
Absorption
Atomic K*
Absorption
BIndicate data points out of limits by * in % Rec. Column
QA Manual for Precipitation Measurement 10.1/7
-------
Jan. 1981 37 Part l-Section 7.0
Audit Spike Recovery Data* (Cont'dj Month:
Cone. Date of Cone. % Date of Cone. % Date of Cone. %
Instrument Species Taken Analysis Found Rec. Analysis Found Rec. Analysis Found Rec.
Atomic Mg"
Absorption
pH Meter pH
ph Meter Gran
Strong
Acid
aIndicate data points out of limits by* in % Rec. Column.
QA Manual for Precipitation Measurement JQ. 1/7.6
-------
Part l-Section 7.0
38
Jan. 1981
Audit Spike Recovery Data3 (Cont'd)
Month
Cone. Date of Cone. % Date of Cone %
Instrument Species Taken Analysis Found Rec Analysis Found Rec
Date of Cone %
Analysis Found Rec
Gran
Total
Acid
Conductivity
a Indicate data points out of limits by* in % Rec. Column.
QA Manual for Precipitation Measurement
10.1/7.6
-------
Jan. 1981
Part l-Section 8.0
8.0 Handling, Validation, and Reporting
Quality control of data handling and
other aspects of analysis and reporting
are part of data management activities,
and require discussions of the follow-
ing: calculating detection limits, spot-
checking and screening for accuracy of
data entry, reporting formats, calculat-
ing of averages and statistics for
cumulative reports, and classifying and
handling outliers.
Data obtained in the field and in the
analytical laboratory must be validated
and quality controlled to assure that
they accurately represent the concen-
trations of the the sample constituents
Validity depends on control of error and
bias during all phases of data handling
from analytical results to final reporting.
8.1 Data Logistics
Involved with data acquisition, reduc-
tion, and reporting are the field operator,
the analyst, the data entry staff, the
laboratory director, the program man-
ager (or project engineer), and the
designated QA coordinator.
Ideally, laboratory data should go
directly from an instrument to a machine
readable raw data base to avoid human
transcription errors. Because not all
laboratories have computer facilities,
manual data-recording techniques are
discussed, but all calculations and data-
processing steps in this section can be
performed by automated processing.
If voltages from the instrument are
not recorded automatically, a data form
must be prepared. Each analyst should
keep a bound notebook to record all
analytical data; this notebook should
have carbons so copies can be pulled for
data reduction. Typical data forms are in
the 0 & M manual(1) for each analytical
procedure and blank forms are in
Section 7.8 and 8.8
Whenever data forms are prepared
from strip charts, transcriptions should
be checked by reprocessing 10% of all
values; if errors are found, all data
should be reprocessed.
The sections below briefly describe
manual data-recording practices for
several typical techniques used in
precipitation analysis.
8.1.1 pH and Conductivity
The pH and conductivity measure-
ments read directly from meters are
recorded on a prepared data form. Meter
readings are recorded instead of "peak
height." Baseline readings are ignored.
8.1.2 Anion Chromatography for
Chloride, Phosphate, Nitrate, and
Sulfate
Peak heights on strip charts are a
measure of response. Draw baselines,
read each peak heightfromthe baseline
with a clear plastic ruler, and record
each peak height on the strip chart and
also on the data form Because this is a
chromatographic technique, care must
be exercised in drawing the baseline.
8.1.3 Automated Colonmetry for
Ammonium and Phosphate
Read the data on the strip charts as
steady-state voltage peaks, draw a
straight line between baseline points on
the chart, read each peak height from
the baseline using a clear plastic ruler,
and record the peak heights on the strip
chart and on the data form
8.1.4 Atomic Absorption for Sodium,
Potassium, Calcium, and Magnesium
Process the data the same as for
automated colorimeter (Section 8.1.3).
8.1.5 Strong Acid by Gran Method
In the microtitration, electrical poten-
tial (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 the conditioning
solution, before addition of sample, is
recorded on the data form (Section 8 8)
and the final temperature before
titration should also be recorded.
8.1.6 Acidity
The sample is titrated potentio-
metrically with a basic solution to an
end point of pH 8 3 The normality (N) of
the base and the volume (ml) required
are recorded on the data form.
8.1.7 Volume
In determining the amount of sample
precipitation, assume a density of 1.0
gm/ml; thus the ratio of mass (gm) to
volume (ml) is 1.0. Record the mass
directly on the data form.
8.2 Field Observation Coding
Encode field observations of weather,
sample condition, sampler, and other
relevant data into computer files or onto
summary forms in a standard format for
ease of interpretation. A six-digit code is
most convenient to summarize weather
conditions, sample conditions, and
equipment conditions:
Weather codes aa include.
00 no event
01 rain
02 snow
03 mixed snow/ram
Samples codes bb include-
00 sample intact
01 sample contaminated (unspeci-
fied cause)
02 Noticeable suspended
particles
03 sample leakage in shipment
04 insufficient sample for complete
analysis
Equipment codes cc include-
00 no equipment problems
01 sampler inoperative or
malfunctioned (no/or negligible
sample)
02 ram gauge inoperative or mal-
functioned
03 field pH and conductivity
measured late (x days after
scheduled sample removal)
04 pH/conductivity meters
inoperative
For example, a mixed rain/snow sample
with noticeable particles and no equip-
ment problems would be coded 030200.
8.3 Software Requirements
Data handling from raw data input
through finished report, should be
computerized as much as possible.
Figure 8-1 shows typical data flow from
chemical analysis through final QC
reporting The software has several
basic functions:
Data input (Section 8.3.1)
Calculation of concentrations from raw
data and calibrations (Section 8.4.2)
Data storage and indexing (Section
8.32)
Control charts and related tests (Section
8.4.4)
Data output and reporting (Section 8.6)
8.3.1 Data Input
Data can be input in two ways,
manually or on line. For manual data
entry, the data clerk should screen all of
the terminal input by manual compari-
sons of the computer printout with the
original data forms or by duplicate
entries. For automated on line entries,
errors can usually be detected by
software reasonableness checks, by
monitoring data display while data are
being taken, and by occasional spot-
checks or audits of data acquisition
apparatus.
-------
Part l-Section 8.0
Jan. 1981
( Start J
I
Initial
Calibration
(Analyst)
I
Analysis
of Samples
with Spikes, etc
f Analyst)
}
Final
Calibration
(Analyst]
I
Manual
Data Entry
(Data Clerk)
_ ,_
f
1
Initial QC
Checks
(Immediate Results
to Analyst)
i
f
^
Error
Recovery
_^
Calibration
Constants
Flag Out-
of-Range
QC Analyses.
"•
i1 (After all
Analyses
for Sample)
Cumulative QC Report
1 Cahb. constant
summary and plots
2. Graphs of accumulated
QC data on range
charts.
3 Summary of duplicate
analyses
4 For all samples:
a) Calculated vs meas.
conductivity.
b) Charge balance of
ions
c) Lab vs field
pH & conductivity.
1. £J • , .
ions' Na/CI, etc.
e) Collected sample
amount vs ram gauge
Figure 8-1.
Typical scheme of analysis and QC functions.
8.3.2 Data Storage and Indexing
At a minimum, large data bases
should be stored on a computer read-
able medium (e g , disk, tape or cassette)
which can be accessed efficiently A
duplicate backup file stored in a
different location should be maintained.
Special file attributes (e g , random
access, keys, and indexing) can be
useful for efficient data management.
Where elaborate file organization with
high capacity on-line storage devices
are available, the following data organ-
ization is recommended
Mam Data File(s) - raw (concentration)
data with appropriate retrieval keys.
Duplicate File - for backup
Index File(s) - data for cross-referencing
site, date of sampling, etc , with data in
main data file.
QC File - calibration constants, control
limits, and QC sample data (spikes,
blanks and duplicates); indexing, by site
and date with cross-reference to the
raw data to which the QC data pertain, is
desirable.
8.3.3 QC Functions
A computer enables extensive auto-
mation of QC functions One valuable
function is immediate flagging of error
conditions spike data which are out of
control, data which are below detection
limits; and calibration constants which
are out of tolerance. Software avoids
tedious control chart plotting and much
manual mathematical labor Equations
and criteria for data flagging have been
described (4, Appendixes E & H).
8.4 Data Handling and Sta-
tistical Analysis of Data
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 from the field. After data enter
the storage-retrieval system, a more
detailed screening process is taken All
procedures should be well documented
so that new personnel can easily
understand them
In establishing statistical screening
procedures, it is necessary to recognize
characteristics of the chemical analy-
ses For the most part, analyses are
done in a batch mode with multipoint
calibrations run before and after the
samples, which may total several dozen
In an efficiently run laboratory, most
analyses are automatically sequenced
with data recorded continuously on strip
charts or by computer As discussed in
Sections 7 5 and 7 6 on chemical
analyses, ion chromatography mea-
sures several constituents in the same
analysis
8.4.1 Quality Control of Data Hand-
ling
Table 8-1 summarizes the data
handling steps and the QC to be applied
at each step A range check is some-
times effective as an additional screen
against keypunch errors
8.4.2 Calculations
Calibration standards should be run
as spikes and blanks 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, equations used for these
calculations are available in most
elementary statistics books(2)
The linear least squares fit yields the
following parameters slope (m), inter-
cept (b), error of fit (e), correlation
coefficient (r), and the detection limit
(dl) The slope and intercept define a
relationship between concentration
standards and instrument response
p, = mx, + b
8-1
where Yp, = predicted instrument re-
sponse (not ya,, actual
instrument response),
and
x, = concentration of stan-
dard i.
Equation 8-1 is the preferred fit where
major components of random variance
are assumed to be in instrument
response. Rearrangement of Equation
8-1 yields concentration corresponding
loan instrumental measurement
, - b)/m
8-2
where x, = calculated concentration for
a sample,
Ya| = actual instrument response
for a sample, and
-------
Jan. 1981
Part l-Section 8.0
Table 8-1. Suggested QC Spotcheck of Data Handling
Data Hand/ing Step QC Procedure
Manual reading of strip chart
Transfer of analysis results to data sheet.
Input of data (field or analytical form)
into computer
Electronic digitization of strip chart
Field report records of event time and
amount
Check 10% of data
Check 1OO% of data
Check data printout 100% vs raw
data form
Check 5 to 8% for baseline and
timing
Check 100% vs rain gauge strip
chart
mandb = calculated slope and inter-
cept from the latest cali-
bration standards run.
The error term is calculated from the
difference between the predicted in-
strument response yp, and the actual
instrument Yai response for a given
calibration standard.
e = [(Ya, - mx, - b)2/(n-2)]12
8-3
where n = number of calibration stan-
dards.
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 calibra-
tions only when all setup parameters
(scale factor, concentration range, etc )
are identical.
The correlation coefficient, a measure
of how much of the variation in
instrument response is explained by
differences in concentration (as opposed
to pure random variation), is discussed
in elementary statistics texts(2)
8.4.3 Statistical Evaluation Tech-
niques
Sampling is assumed to collect single
(not duplicate) samples for most pre-
cipitation events so 100% QC of each
sample cannot be attained; however,
with QC procedures addressed to each
analysis batch, good screening of proper
instrument functioning can be achieved.
The following discussions present
statistical checks available under the
constraints of non-duplicate sample
collection. Table 8-2 summarizes avail-
able QC statistics.
Table 8-2. 'Quality Control Data
Type of Information
8.4.3.1 Calibrations—In addition to a
mathematical estimate of the relation-
ship between instrument output signal
and concentration (by regression curve),
calibration provides statistics for eval-
uating the analytical method. Duplicate
calibrations, before and after analysis of
field samples, yield data on instrument
reproducibility and drift Statistics from
routine calibration data include:
Mimimum Detection Limit - A conser-
vative method (e.g., Hubaux and Vos,
Appendix B (3))yieldsthedetection limit
as a function of the quality of the linear
least squares fit of the calibration data.
Comparison of the detection limit
calculated from each day's standardiza-
tions with the one for the technique and
range determined during instrument
documentation may indicate problems
with technique.
Slope and Intercept - The slope and
intercept of the least squares fit of the
data for a technique and instrument
range are fairly constant Visual inspec-
tion of values from successive days or
comparison with values obtained during
the method validation can be used for
quality control
Correlation Coefficient - This easily
calculated statistic 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. The
value can be used as a more quantita-
tive measure in two ways. First, r2 is
equal to the fraction of total variance s2
(s = standard deviation) due to the
Data Evaluated
1. Calibration Curve
Slope (m)
Intercept (b)
Error (el
Correlation coefficient (r)
Detection limit (dl)
2. Blank
3. Spike sample
4. Old sample
Not charted
Not charted
Calculated magnitude
Non-Gaussian, not used for QC
Calculated magnitude
Calculated magnitude
Percent recovery
Difference between old and new values
correlation of the tested variables; the
balance is due to random error. Second,
r together with sample size n can be
used to calculate the 95% confidence
limits of the derived slope of the linear
relation between the tested variables
(2).
Residual Error - The scatter of data
points off the regression line is a
measure of "noise" in the calibration,
and it is related to expected precision of
the analysis
8.4.3.2 Spikes and Blanks—If the
experimental design calls for only one
analysis of nonduplicate field samples,
the primary tool for monitoring the
integrity of the analysis is the occa-
sional insertion of spikes of known
concentrations. If calibrations are done
before and after each batch of samples,
the QC samples of known value should
be near the middle of the batch for
optimum control. If the final calibration
is omitted, a QC spike and blank should
be the last samples analyzed. The
concentration values of the spikes
should be entered into the computer
before analysis, so that control limits
can be checked as the analysis results
are keyed into the computer If control
limits are exceeded, there should be an
immediate report by the QC coordinator
to the laboratory director If no computer
is available, the laboratory director, QC
coordinator, or analyst should check the
spike results immediatley to determine
if an out-of-control condition exists.
8.4.3.3 Old Samples—Another con-
trol check is reanalysis of old samples to
obtain information on sample stability,
which varies from sample to sample the
r-T, NH/, and N03~ are susceptible to
degradation; if degradation is observed,
the sample handling and preservation
techniques should be examined. If
degradation is not observed (i.e. agree-
ment is good), old sample data can be
used to calculate the precision of
analysis. The precision may vary from
day to day.
8.4.4 Control Charts and Related
Tests
With control charts, QC data are
tabulated and plotted as functions of
time; the abscissa is the chronological
order of analysis, and the ordinate may
be either the range, absolute magni-
tude, value of the difference of replicates,
or percentage recovery. All data are
plotted. An expected value and control
limits are calculated, and these are
plotted as the average-value line and
the control limit lines. Table 8-3 shows
types of data which are quality controlled
using the control chart. Other EPA
publications (4-Appendix H,5) give
details of chart construction.
-------
Part l-Section 8.0
Jan. 1981
Table 8-3. Types of Data Controlled by Controlled Chart Methods
Type of Data
Duplicate Difference
Value Plotted
Absolute Value
of Difference
It/,1
Expectation Value
Mean Difference
from method vali-
dation or other
historical data
base, \3~\
Control Limits
Lower limit - 0
Upper limit - \d\ + ZS
QC spike
QC Blank
Percentage
Recovery, X
Calculated
Magnitude
100%
0
lower limit - O
or 100-(1-ZS)
Upper limit
(1+ZS1 • 100
Lower limit - none
Upper limit - dl
d( - Individual duplicate difference
IcTi- Mean of absolute value of duplicate differences
Z - Coefficient from distribution tables for any specified level of precision
{(2) and Appendix H of (4)).
S - Estimate of standard deviation from the historical data base
dl- Detection limit
Computer calculation and data evalu-
ation with or without plotting may be
substituted for manually plotting the
data. Because the purpose of control
charts 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. In either case, the control
chart results should be correlated as
soon as practical after an analysis to
expedite corrective action.
Any control situation unlikely tooccur
statistically should be examined with-
out delay. Many of these occurrences
may be due to chance rather than to
technical problems with the analysis
Two types of limits are frequently
defined:
Control Limits - the 3s deviation from
the mean value, where s is estimated
standard deviation for technique and
concentration range; thus the upper
control limit (UCL) is u + 3s (u is mean or
expected value of the charted parameter),
and the lower control limit (LCL) is u - 3s
(or 0 if the calculated value is negative)
Setting limits at ±3s implies that less
than 2% of valid data are flagged due to
random error alone and that other
flagged data may be assumed to
indicate nonrandom error, i.e., malfunc-
tion or contamination.
Warning Limits - The ±2s deviations
from the expected value of the control
parameter include about 95% of the
expected random variation about the
mean However, simple probability
would predict less than 1 % probability of
two independently chosen values for
the charted parameter to exceed the
upper (UWL) or the lower (LWL) warning
limits due to chance alone; thus two
successive values which exceed one or
the other (not both) warning limits
should be reason for investigation of
analytical control.
A final control applied in control
charting is testing for a number of
successive points which lie uniformly
on one side of the expected value.
Probability predicts less than 1% chance
of eight independent consecutive values
occurring on the same side of the mean
value line, so eight or more would
indicate a small systematic bias in the
technique that may not be observed
until after many days of analysis
because of the relatively large number
of points required.
8.5 Data Validation Criteria
Data validation based on a set of
criteria is the process of filtering the
data and either accepting or rejecting
them. All procedures in this section are
applied after data are first obtained to
identify and flag questionable data for
subsequent investigation. Validation
includes investigation of apparent
anomalies.
8.5.1 Least Squares Fit of Calibration
Curves
As a QC function, calibrations are
analyzed for excessive scatter from the
fit line and for anomalous intercepts
Cumulative records on each analyte
concentration range and method allow
the laboratory director to examine
trends or discrepancies in coefficients
and standard deviations of fit
8.5.2 Control Limit Checks
To know if inputs from an analysis
batch are valid, it is necessary to know
as soon as possible after analysis if
control limits were exceeded on the
known QC samples that were analyzed
with the unknown samples
8.5.3 Detection Limit Checks
Of immediate value to the analyst is
notification that a sample is below the
detection limit of the measurement
range. Ideally, this knowledge should be
available immediately after raw data
entry, then the analyst has two options.
1. Reanalyze the sample on an in-
strument scale of greater sensitivity
2 Improve the sensitivity (as calculated
by Hubaux-Vos(3)) by multiple
measurements of the unknown, by
judicious choices of concentration
values for the standard curve, or by
use of another analytical tech-
nique If a sample is below all
detection limits and cannot be
reanalyzed because of lack of
material or cost of reanalysis, a
flag should be entered into the data
base indicating that the datum is
below the detection limit. With a
fully automated system, this flag-
ging could serve as an additional
QA procedure because the analyst
cannot mistakenly enter data
below the detection limit if the
computer program is written to
reject all such data Statistical
handling and reporting of below-
detection-limit data are discussed
in Section 8.6.4
8.5.4 Inconsistencies in Precipitation
Collection Results
Stations equipped with duplicate
samplers or with a sampler and a rain
gauge can test data quality by checking
the quantity of precipitation measured
by the two instruments Amounts of rain
which differ by more than 1 5% from a
duplicate should be flagged for investi-
gation of laulty function. To compare the
amounts collected by samplers and by
rain gauges, convert the sampler weight
to inches by multiplying the weight
(grams) by 0.00058 m./gm. However,
compare carefully because light rain-
falls generally yield high weighing rain
gauge capture vs the sampler, and
heavy rainfalls (and wind) yield high
sampler capture vs. tipping bucket
gauge. Differences of 30% between the
sampler and the rain gauge are fre-
quent, for snow, much greater differ-
ences can occur. It has been suggesed
by MAP3S personnel, if the ratio (rain
gauge - sampler volume) /rain gauge is
greater than 0 5, reject the sample; if
the sampler volume is greater than the
rain gauge volume, use the sampler
volume for calculations. The reason for
the difference in sample amount cap-
tured by the rain gauge and sampler
must be resolved before a recom-
mendation is made.
-------
Jan. 1981
Part l-Section 8.0
As an additional check on data
validity, the stripchart record of the
precipitation gauges should be com-
pared with the field data form. If
discrepancies in time of event or
precipitation amount are noted, they
must be resolved before the data can be
reported.
8.5.5 Unusual Ion Ratios
Another check on data validity is
subjective analysis of the ratios of ions
in individual samples. Table 8-4 shows
average and typical ion weight ratios for
terrains and locations. Average values
for sea water(6) should apply to most
areas(7), but those for the earth's
crust(6) cannot be assumedto represent
any specific region. The range data (last
column) were taken from a 1968
national study(8). Significant depar-
tures from these or from typical ratios
for the area are reasons for laboratory
director to investigate an individual
sample or an entire analytical proce-
dure. The seawater ratio of SO
-------
Part l-Section 8.0
Jan. 1981
SITE tt
11
21
:;1
41
'-I'/
61
72
•JANUARY
2 4 6 8 10 1
1 35 7 9 11
R
R
M S R
S R
M R R
R R R
X R
1 9y'.>
2 14 16
13 15 17
R
R R
R
P
R R
PI R
18 20 22 24 26 26 30
1? 21 23 25 27 29 31
R
X
R R
S
S
R R R 1-1 S
S
M
R R
•f'l
R
R
M M
R
SITE
11
21
FEBRUARY 19yO
4 6 y 10 12 14 16 IS 20 22 24 26 28
5 7 v 11 r-: 15 17 19 21 23 25 27 29
X\~- i "•
I1 _'
SI
41
R
M
R rl
R
61
11
M M
R.
R
R
R
72
O ""/
R
M
R R
91
S = Snow
M = Mixed (rain & snow)
X = Missed or lost sample
Figure 8-2. Calendar of events.
time periods, two other important Table 8-6.
terms, are discussed m Section 8.6.3.
Reporting units are typically:
pH - pH units,
conductivity - microS/cm or micromho/cm,
concentration - micromol/liter, and
precipitation - mm (1 in. = 25.4 mm).
Transformation from microgram/ml
(ppm), the generally used analytical
term, to micromol/liter is:
micromoles/liter = microgram/ml
x conversion factor
8-6
The conversion factor for each ion is in
Table 8-6.
Factors to Convert
Microgram/ml to Micro-
mol/liter (micromol/l =
microgram/ml x factor)
Analyte
cr
NO3~
SO4=
P04~3
rf
NHS
/Va+
r
Mg"
Ca++
Molecule
Weight
35.46
62.01
96.07
9498
1.01
18.04
22.99
39.10
24.31
40.08
Factor
28.2
16.1
10.4
10.5
990.1
55.4
43.5
25.6
41.1
25.0
8.6.3 Weighted Mean Values and
Deposition
Generally reported on a cumulative
basis for a given time interval are
precipitation weighted mean concen-
tration C, for various ions.
C^ZP.C./IP,
8-7
where
P, = amount of precipitation
m event j (mm or ml),
C,|= concentration of con-
stituent i for
event j (micromol/l).
For pH or I-T concentration, the cumula-
tive value is calculated, and the final pH
-------
Jan. 1981
Part l-Section 8.0
pH = -log
I, (Pj10
"pH|
"I
J
8-8
The cumulative average concentration
at each site can be used m studies of
distribution of the constituents as a
function of time (month or season) and
geographic location.
Total deposition of an analyte per unit
area (mg/m2) in a precipitation event or
time interval is calculated as:
2 = microgram/ml x 25.4xP
8-9
where
P = precipitation (in.)
and 25.4 = 2.54 cm/in.
x 10~3 mg/microgram
Treatment of missing data in calculating
the above values isdiscussed in Section
8.7.3.
8.6.4 Reporting and Treating of Be-
low'-Detection-Limit Data
Data below the detection limit (BDL)
for the analytical method used should
be flagged with a code both in the
printouts and in all computer readable,
data forms. Because it is important for
data interpreters to know the detection
limit, a suggested code is the negative
DL. The treatment of these data m data
processing programs depends on the
analysis being performed; examples are
replacing all BDL data with 0 if an
arithmetic mean is being taken or with
Vi DL(10) if a geometric mean is being
computed. Any computed summary
value which has 25% or more BDL
contributing data points should be
coded with a unique flag.
8.6.5 Reporting of Out-of-Control
Data
Analytical data in the same batch with
an out-of-control QC sample must be
flagged in the data base as suspect. All
such data must be either reanalyzed or
reported as invalid, unless the labora-
tory supervisor decides that the cause of
the out-of-control condition could not
have affected the analytical results.
Data reported as invalid should not be
used in statistics, analysis, or other
interpretation. For data stored or
reported on computer readable media, it
is often best to use a cha racter or code to
indicate out-of-control, missing, BDL,
and so forth.
8.6.6 Treatment of Outliers
Any data point that lies far beyond the
normal range of values can be considered
an outlier. Responsibilities should be
divided between invalidations by experi-
mental and data interpretation person-
nel. Experimental personnel should
invalidate data only for experimental
reasons such as known contamination
or violation of QC criteria for a spike,
blank, or duplicate; they should not
invalidate outliers Outliers in data
interpretation is discussed in Section
8.7.2.
8.6.7 EPA Rain Water Data Base
The EPA is establishing a precipita-
tion data base for all United States and
Canadian deposition programs. This
System of Atmospheric Deposition
(SAD) data will be coordinated through
the National Computer Center (NCC).
Prospective contributors/users should
contact Gardner Evans, United States
EPA, Environmental Research Center
(ERC), (MD 075), Research Triangle Park,
NC 27711.
8.7 Statistics in Data Inter-
pretation
This section discusses briefly the
types of statistics that should be used in
precipitation data interpretation and the
subjects of outliers and missing data.
8.7.1 Parametric and Nonparametric
Methods and Tests
Many procedures routinely used in
statistical analysis involve the assump-
tion that the parent population from
which a sample is drawn has a normal
(gaussian) distribution. Normal distribu-
tion is not needed to compute means,
weighted means, and variances; the
most common formulas for computing
means and variances converge to
values for the parent population for
large sample sizes, regardless of
distribution. Many common hypothesis
tests normal distribution; these are
called parametric tests because they are
concerned with parameters (mean and
standard deviations) of a single func-
tional form (the gaussian curve). For
example, the Student t-test for deter-
mining whether two samples come
from populations with the same mean
assumes that the samples come from
normally distributed populations with
the same variance; thus it isimporantto
know whether the data set one is
working with comes from a normally
distributed population.
Precipitation data frequently show
skewed distributions sometimes inter-
preted as log-normal(11). The distribu-
tions 'are inevitably nonnormal (for all
analytes except pH) because negative
values of concentration and rainfall are
not allowed. The resulting distortions
could be small if the mean is two or
more times the standard deviation, but
this has not been the case in precipita-
tion for analytes other than pH.
To test whether an analyte has a
normal distribution, use an optimized
chi-square test(12) for sample popula-
tions of over 100 with the number of
class intervals chosen according to the
algorithm of Mann and Wald(13). If the
number of data points in the set is below
100, use the Lilhefors(14) test for
normality; this test may also be used for
sample sizes greater than 100. If an
analyte fails the test for normality,
repeat the test with the logarithm of
each data point to determine if the
parent distribution is log-normal.
All data below the DL of the analytical
procedure must be set to positive
nonzero values An acceptable value is
1/2 DL. A significant number of BDL data
points results in a distribution which is
neither normal nor log-normal. If the
distribution is log-normal and the
interpreter wishes to proceed with
parametric hypothesis testing, log-
transformed data must be used. It is
unacceptable to use ordinary para-
metric statistics on a distribution which
is skewed enough to appear log-normal.
If the distribution fails both the
normal and the log-normal tests, do the
hypothesis testing with nonparametric
tests, also known as distribution-free
statistics; these tests do not assume
that the parent distribution is normal so
they work whether it is or not. If distribu-
tion is normal, these tests are only
slightly less powerful than the para-
metrics, so they can serve as a viable
alternative if the data analyst does not
wish to test for normality. Another
advantage of nonparametrics is its
being less sensitive to outliers than
parametrics; this msensitivity to outliers
is due to dealing with ranks and
categories of the data rather than
original data points. A good reference
on nonparametric hypothesis testing is
Conover(15).
8.7.2 Outliers
Data interpreters may delete outliers,
depending on the computations being
performed. For example, computations
of mean values are generally not
sensitive to deletion of the relatively few
data from both ends of the distribution,
but computations of skewness may be
very sensitive to the loss of data at the
extremes of the normal distribution.
A large concentration event which is
an outlier m the concentration distribu-
tion may not be an outlier m the deposi-
tion distribution if the total rainfall
amount is small. The decision to delete
data from the ends of the distribution
must be made with the full understand-
ing of the interpretations being con-
sidered; this understanding is not
possible if data are deleted by experi-
mental personnel before interpreters
see the data.
-------
Part l-Section 8.0
Jan. 1981
In precipitation chemistry, selection
of outliers is difficult due to the
following considerations1
1. Spatial distributions of pollutants
vary, so one site or more than one
site in similar locales must be
considered separately from sites in
different locales
2. Seasonal amounts of pollutants at
each site generally vary.
3. Concentrations of pollutants can
reach unusually high or low values
in precipitation events.
Considering the above, no data that
have met the QC criteria (discussed in
previous sections) should be discarded.
Because relatively few nonnormal data
points or outliers are expected, the
effect on the mean values of any
constituent should be small
IMonparametric hypothesis testing
provides no way to check for outliers (A
point can be identified as isolated from
its parent population only if a form for
the parent population is assumed.)
Fortunately, nonparametric tests are
comparatively insensitive to outliers so
the outliers may be left in the data base.
Parametric tests for detecting outliers
are in Reference 4, Appendix F; these
are designed to eliminate only one or
two outliers from a data set, but it is
absolutely invalid to drop detected
outliers and then to pass through the
data again to look for other outliers
8.7.3 Missing Data
If a small percentage of samples are
lost (e.g., due to downtime of equip-
ment), the effect of neglecting their
contributions to the precipitation
weighted average, the pH or the
concentration of a constituent is ex-
pected to be small. However, the effect
of the lost samples on the total amount
of material deposited is more important;
the deposition contributions of lost
samples should be calculated by using
the precipitation weighted mean sea-
sonal concentration values for that site
and the measured or reported (National
Weather Service) precipitation amounts
for the missed events.
8.8 Data Forms
Blank data forms on the following
pages can be used for manual transcrip-
tion of data from digital meters or from
strip chart records to prepare for data
processing. Details on preparing data
forms for each laboratory analysis are in
the respective sections of the 0 & M
manual(1).
8.9 References
1. Operations and Maintenance
Manual for Precipitation Mea-
surement Systems, United States
Environmental Protection Agency,
Research Triangle Park, N.C., (in
press).
2. An Introduction to Mathematical
Statistics. H.D. Brunk, Blaisdell
Publishing Co., Waltham, MA,
1976, p 210.
3. Hubaux A. andG. Vos, Anal.Chem.,
42, 849 (1970)
4 Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems - Vol. I- Principles, United
States Environmental Protection
Agency, Research Triangle Park,
N.C., EPA-600/9-76-005 (1976).
5. Handbook for Analytical Quality
Control in Water and Wastewater
Laboratories, United States En-
vironmental Protection Agency,
Cincinnati, OH, EPA-600/4-79-
019(1979).
6. Handbook of Chemistry and Phy-
sics, 48th ed Edited by R.C. Weast,
The Chemical Rubber Co., Cleve-
land, OH 44128, 1967-68.
7. Mero, J.L., The Mineral Resources
of the Sea, Elsevier Publ., New
York, N.Y. p.25 (1964).
8. Lodge, Jr., J.P., J.B Pate, W.
Basbergill, G.S. Swanson, K.C. Hill,
E. Lorangeand A.L. Lazrus, "Chem-
istry of United States Precipitation,"
Final Report on the National Pre-
cipitation Sampling Network, Na-
tional Center for Atmospheric
Research, Boulder CO, August
(1968)
9. Electric Power Research Institute,
"Acid Precipitation in the North-
eastern United States," Contract
Nos. RF1376-1 and RF1630-2,
Rockwell International Environ-
mental Monitoring and Services
Center (1978-1980).
10. Nehls, G. and G. Akland, J. Air
Pollut. Control. Assoc., 23, 180
(1973).
11. "The MAP3S/RAINE Precipitation
Chemistry Network: Statistical
Overview for the Period 1976-
1980," Atmos. Environ., submitted
for publication.
12. Kendall, M.G. and A. Stuart, The
Advanced Theory of Statistics, Vol.
II, Chapter 30, Hafner Publishing,
New York, (1967).
13. Mann. H.B. and A. Wald, "On the
Choice of the Number of Class
Intervals in the Application of the
Chi-squared Test," Annals of
Math. Statistics 13, 306-7 (1942).
14. Lilliefors, H.W., "On the Kolmogorov-
Smirnov Test for Normality with
Mean and Variance Unknown," J.
Amer. Statistical Assoc., 62, 399-
402(1967).
1 5. Conover, W.J., Practical Nonpara-
metric Statistics, John Wiley and
Sons, New York, 2nd Ed. (1980).
-------
Jan. 1981
Part l-Section 8.0
Gran Strong Acid Data Sheet
Sample #
Initial mv
Temp. °C
til NaOH
Injected
mv
reading
Sample #
Initial mv
Temp °C
fj/ NaOH
Injected
mv
reading
Date:
Cone NaOH N
Mict
Cc
opipette
ilibration
5 til = fjl
15 til - ,,/
Total volume
-------
Part l-Section 8.0
10
Jan. 1981
Data Sheet
N
J
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
ID
Baseline 1
Std 1
Std 2
Std 3
Std 4
Std 5
Sample 1
Sample 2
Sample 3
Sample 4
Sample 5
QC 1
Sample 6
Sample 7
Sample 8
Sample 9
Baseline 2
Sample 10
Sample 1 1
Sample 12
Std 1
Std 2
Std 3
Std 4
Std 5
Baseline 3
Peak Ht
B1
B2
B3
-------
Jan. 1981
Part l-Section 9.0
9.0 Data Quality Assessment
Precision and accuracy of precipita-
tion data provide quantitative assess-
ments of data quality. Procedures for
determining precision and accuracy
were designed for specific monitoring
and analysis methods Measurement
methods used in precipitation chemistry
monitoring systems for integrated
sampling follow a three-step process;
the first two are functions of field
operations, and the third is part of the
laboratory operations.
1. Collection of the sample in a
suitable sampler.
2. Initial field analysis for pH, specific
conductance, and weight
3. Expanded chemical analyses in the
laboratory.
At each step, QA should be assessed
Precision of sampling is determined
by using collocated samplers; field
measurement accuracy is determined
by test sample audit. Precision of
laboratory analytical methods is deter-
mined with duplicate samples; accuracy
is determined by an internal and
external blind sample audit.
9.1 Evaluation of Field Oper-
ations
9.1.1 Sampling Precision
Precision is estimated by duplicate
samplings with collocated precipitation
samplers. Each network of sites operated
by the same agency should have a
duplicate sampler of the type used for
routine monitoring. The collocated
sampler operated during routine samp-
ling should be set up consistent with
siting criteria (Section 5.0).
Data from the collocated sampler are
compared to data from the station
sampler. The measured difference in pH
(pH units), conductivity (micromho/cm),
and total mass captured between the
samplers are used to calculate preci-
sion. For each pair of measurements,
differences can be calculated using:
A, = y,-x, 9-1
where A, = difference in measure-
ments for ith precision
check (appropriate units),
y, = pH, conductivity, or
weight measured by the
duplicate collocated
sampler for ith precision
check, and
x, =pH, conductivity or
weight measured by the
the corresponding station
sampler for the ith pre-
cision check.
Quarterly, standard deviation s is
estimated for each site with collocated
samplers:
s,=
A*-1 (I A,)2
n 1=1
where s,=quarterlystandarddevia-
tion of jth instrument
during ith precision
check,
A, = difference in measure-
ments during ith pre-
cision check, and
n = number of precision
checks on the instru-
ment during calendar
quarter.
If the network contains more than one
collocated sampler, an overall precision
estimate sa for each measurement
method can be computed quarterly
Sa= r iz s2 i *
LkJ=' ' J
9-3
where
sa = overall deviations
for a specific measure-
ment method,
s, = standard deviations of
jth sampler for measure-
ment method, and
k = number of collocated
measurements within
the network.
Data from both the station sampler and
the collocated sampler are recorded on
standard Field Data forms (Section
6.8.5) with corresponding identifica-
tions. Copies of these forms should be
submitted to the QA coordinator, who
summarizes the data on the Precision
Check form (Section 9.5), which is used
in preparing QA reports (Section 9.3).
9.1.2 A ccuracy of pH and Conductiv-
ity Measurements
To assess the accuracies of field
measurements of pH and conductivity,
audits should be conducted using test
samples prepared and sent from the
central laboratory each month (Section
7.6). The samples should be measured
as soon as possible after receipt, and the
results should be returned to the
laboratory with the remainder of the
sample. The audit results should
include two laboratory analyses—one
before the sample is sent to the f itid and
one after it is returned and the field
analysis results. The QA coordinator
should record all the data on the
Monthly Field Audit Report (Section
9.4).
Accuracy for each measured variable
is estimated by computing value differ-
ences:
A, = f,-(l,,+ l(J)/2 9-4
where A,=difference in measure-
ments for jth site (appro-
priate units),
f, = field analysis of variable
for jth site,
li, = initial laboratory
analysis of variable
before shipment to the
jth site, and
lf|=final laboratory analysis
of variable after return
from jth site.
This equation should be used only if 11,, -
lf,| is less than the control limits.
Data should be summarized in the
Monthly Field Audit Report by the QA
coordinator, who also computes an
average monthly network difference Am
for the variables pH and conductivity
Am=(1/n)Z,=i
9-5
and for the monthly standard deviation
Sm =
-f? A,)2
n fi
I 1/2
9-6
where Aim ^average monthly net-
work difference for a
given variable,
A, = difference between lab \\
and field f, at the jth site,
and
n = number of sites audited
during the month.
Monthly reports should be summarized
in a quarterly report (Section 9.3).
9.1.3 Method Precision
Each network should maintain at
least one pair of collocated samplers
(Section 9.1.1). The collected samples
should be tagged so their respective
measurements in the. field and the
-------
Part l-Section 9.0
Jan. 1981
laboratory can be used to assess
quantitatively the monitoring (sampling,
handling, and measurement etc.) pre-
cision by using Equations 9-1 through
9-3 above, where y and x also represent
concentrations of species. Data should
be reported on the Precision Checks
forms (Section 9.4).
9.2 Evaluation of Laboratory
Operations
The QA coordinator must implement
routine activities to assess the precision
and accuracy of laboratory chemical
analyses of precipitation samples.
Laboratory data quality is assessed by
determining the precision, variability,
accuracy of chemical analyses.
9.2.1 Analytical Precision
To estimate the contribution of
analytical variability to total variability,
duplicate sample analysis should be
performed on about 10%of the routinely
analyzed samples. Samples randomly
selected for replicate analysis by the QC
chemist should contain a large quantity
of precipitation. The chosen samples
should be split—one-half for analysis
immediately, the other half for re-
frigeration at about 4°C to be analyzed
within a week by a different analyst. The
sample should be properly identified,
and the results of theduplicate analyses
should be recorded by the QA coordina-
tor in the Report of Duplicate Analyses
(Section'9.4).
9.2.2 Accuracy of Chemical Analysis
Accuracy of chemical analysis should
be determined monthly on blind samples
submitted to the laboratory by randomly
selected field sites. These samples,
prepared by diluting various water
standards, should be shipped in sealed
plastic bottles to the field sites. Each
sample should be identified, and ac-
companied by two postcards. On receipt
at the field site, the sample should be
refrigerated at 4°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 bucket or to the usual
container for shipment. The sample
should be weighed, and an aliquot
should be measured for pH and for
specific conductance. Afield data form
should be filled out with measured
values and the other required data; to do
this, a non-existing 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 any sample), and the
postcards mailed to the QA coordinators
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.
At regularly scheduled intervals, the
analytical laboratory should send a
printout of sample results to the QA
coordinator. The QA coordinator identi-
fies the blind samples, has the data
transferred to the QA data file, notifies
the analytical laboratory of sample
results that must be deleted from the
precipitation data file; and sends the
laboratory the expected QA sample
values. If there is a difference between
any analyzed and expected values that
is greater than the experimental error,
that analysis must be repeated. (Samples
must be stored in a laboratory refrigera-
tor until results have been approved by
the QA coordinator.) Data should be
summarized as indicated in Equations
9-4 and 9-5 (Section 9.1.2). The QA
coordinator should obtain samples from
EPA's NAP and from the USGS semi-
annually for overview purposes.
9.3 Reporting of Data
9.3.1 Laboratory Results
When analyses are complete and the
analyst has released the data for
reporting, a printout or summary of the
following must be prepared:
Analytical data
Anion-cation comparison data
Measured vs. calculated conductance
data
Quality control data
The anion-cation and the measured vs.
calculated conductance comparison
data were discussed in Sections 8.5.6
and 8.5.7. The QA data should be a
listing of the results of the QC param-
eters which pertain to the particular
data being reported and to the audit
results. Data which exceed previously
established QC limits should be flagged.
The QA coordinator should provide
the following reports to the project
manager:
Precision - Quarterly reports, sum-
marizing all data on collocated samples
and duplicate analyses performed
during the 3 mos, should include
differences in analytical results for the
split samples and the average differ-
ence for each analyte; the report should
compare average differences with the
QC ranges typical for the laboratory and
the measurement method.
Accuracy - Quarterly reports, summa-
rizing all data on three blind QA samples
in the 3 mos, should include the different
constituent concentrations and an aver-
age difference and standard deviation
computed for each. Data for the com-
parison should be obtained for each
blind sample by the QA coordinator who
summarizes the differences (mg/liter)
for each constituent.
9.3.2 Audit Data Basis for Field
Measurements
The pH and the conductivity are
generally measured in the field because
of simplicity, importance, and the
possibility of chemical change before a
sample would arrive in the laboratory.
To assess the quality of measurement
by the field operator, audit or test
samples sent from the laboratory should
be run. (To assess sampling nrecision,
collocated samplers should be run
(Section 9.1).) Results should Delude
the two laboratory analyses (before the
sample is sent to the field and after it is
returned) and the field analysis. Be-
cause the amount of data is generally
not large, computer data management
is optional. The steps involved in using
these data for checking the quality of
field measurements are in Figure 9-1.
Control limits for differences between
field and laboratory measurements
must be set during the preliminary
instrument performance documenta-
tion. Factors contributing to random
variability should include differences
between well-calibrated instruments;
storage time (several days between
measurements); storage conditions
(temperature, agitation, etc.); and
differences in field and laboratory
solution temperatures during mea-
surement (effect can be negligible if
instruments have temperature com-
pensation) After a sufficient data base
with real measurements is established,
new control limits can be calculated
from actual data. The data base should
give the standard or average deviation
for each station and for the network as a
whole. Typical control limits for 95%
confidence are approximately 0.15 unit
for pH and 10% for conductivity. Data
should also indicate when a bad pH
electrode or conductivity standard
needs replacing.
The laboratory manager should eval-
uate the data carefully if an electrode is
malfunctioning because it might be
necessary to invalidate all data taken by
that electrode since the last acceptable
audit; thus monthly test samples are
suggested.
9.4 Data Forms
Blank forms on the following pages
were taken or adapted from EPA forms
and from other references. No page-by-
page documentation is given in the top
right-hand corners on these forms. The
titles are at the top of the figures as is
customary. To relate each form to the
text, a form number is given mthe lower
right-hand corner: 1.1/9.1 indicates
form 1, version 1 (Section 9.1). A
-------
Jan. 1981
Part l-Section 9.0
Start
Prepare
Blind
Sample
(Laboratory)
Analysis
L2
(Laboratory)
Analysis
/Li
(Laboratory)
Analysis
F
(Field)
Ship
to
Field
Investigate
Field
Operations
No Problems
-End- J
Return Sample to
Laboratory
Figure 9-1. A udit procedure for field measurements of pH and conductivity.
revision of the form would be changed
to 1.2/9.1 three forms included here for
user convenience are: form 1, version 2,
and so forth.
Form Number Title
1.1/9.1 Precision Checks
2.1/9.1 Monthly Field Audit Report
3.1/9.2 Report of Duplicate Analyses
-------
Part l-Section 9.0
Jan. 1981
Site ID
Month/Year
Precision Checks*
Station Sampler #
Collocated Sampler #
Date
Weight
fgm)
S
C
D
pH
fpH units)
S
C
D
S
Conductivity C
(umho/cm) D
* S - Station Sampler
C - Collocated Sampler
D = S-C
A,= -i D,
n '=''
S,=
(n-1)
D,
]l
1/2
Weight - A,
(gm) S,
pH -A,
(pH Units) S,
Conductivity - A,
(limho/cm) 7,
QA Coordinator
Date
QA Manual for Precipitation Measurement
1.1/9.2
-------
Jan. 1981
Part l-Section 9.0
Monthly Field Audit Report
Sample #:.
Date of Preparation of Field Audit Sample:.
(Analyst Signature)
Laboratory Analysis Before Shipment*
To The Field
Laboratory Analysis After Return
From The Field
Date:
Conductivity
pH
Date:
Conductivity
pH
1.
2.
3.
±Average
Std. Dev
±A verage
. Std. Dev .
Site #
Field
Date
Laboratory Analysis of Audit Samples Vs. Field Analysis-
Conductivity (nmho/cm)
Field
Analysis
Lab"
Analysis
Diff.
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 the samples returned from the field.
b Values after return from the field.
QA Manual for Precipitation Measurement
2.1/9.2
-------
Part l-Section 9.0
Jan. 1981
Analyte
Analysis
Technique
Report of Duplicate Analysis
Routine*
(ID )
Date Result
Duplicate**
(ID ;
Date Result
Diff.
PH
Conductivity
Sulfate
Nitrate
Chloride
Phosphate
Carbonate
Bi-Carbonate
Acidity
Strong Acid
Ammonium
Sodium
Potassium
Magnesium
Calcium
* Value reported as routine sample
** Duplicate sample, might have different ID
***Diff. = Duplicate - Routine
QA Manual for Precipitation Measurement
3.1/9.2
-------
Jan. 1981
Part I - Appendix A
Appendix A
Hubaux-Vos Detection Limit Calculation
According to Hubaux and Vos(1),
DL(YD) is the limit "at which a given
analytical procedure may be relied upon
to lead to detection." The Hubaux-Vos
detection limit (DL) is significantly
greater than the smallest detectable
signal because the DL definition re-
quired a high level of certainty that a
signal represents the presence of the
analyte. An example calculation - To
establish analytical precision and
accuracy as a function of concentration,
repetitively analyze five times an
approximately 7-point calibration curve
from the lowest standard to the highest.
Then perform a linear least squares fit of
all (~35) data points. Use the slope and
the intercept values to calculate con-
centrations for all standard solutions.
Calculate the average concentration
and the standard deviation for each
concentration. Express the accuracy as
a percentage recovery found, as com-
pared to that expected. Express the
precision as percentage standard devia-
tion. This procedure was applied to Pb,
Cd, and Cr analyses by atomic absorp-
tion spectrometry. Linear least squares
fit of each calibration curve produced a
Hubaux-Vos DL for each curve. The
average detection limit was calculated.
The data in Table A-1 indicate, at the
Hubaux-Vos DL for a single calibration
curve, 10% to 20% analytical precision
and 80% to 100% analytical accuracy.
Table A-2.
Values For Student
Table A-1. Hubaux-Vos Detection Limit Data
Analyte
Pb
Pb
Cd
Cr
Highest
Cone. Std.
(ug/ml)
5.0
10.0
0.5
1.0
Lowest
Cone. Std.
(ug/ml)
0.1
0.3
0.01
0.03
Overall DL,
(ug/ml)
0.13
0.17
0.006
0.019
Avg DL,
(ug/ml)
0.224
0.325
0.011
0.036
Recovery
at Avg
DL, %
92
90
81
97
Std. Dev.
at Avg
DL, %
17
16
12
12
An Example
1. X = A + By
where X = instrument response.
A =
B =
A-1
n(Zy2 - (ZyP/n)
ZXy-(ZX)(Zy)/n.
Iy2 - (Zy)2/n
where S2 =
n = total number of data points, and
y = concentration (generally ml//g/ml or ng/ml).
S= NTS?
IX2 - (ZX)2/n- B2 (Iy2 - (Zy)2/n)
n - 2
A-2
= Students t-test for n-2|degrees of freedom, 95% confidence
(Table A-2).
Xc = A + ST V 1 /n, + 1 /N + y~2/'(Iy2 - (Iy)2/n)
YD = 2(Xc' A)(first estimate).
B
Y -.xc - A + ST V1 /n, + 1/N + '(Y'p - yF/(Iy2 - (Iy)2/n)
B
where n, = number of repeats of each concentration,
N = numoer of concentrations, and
n - total number of data points (nrN).
X' = A + BYD - ST VTTn^
delta = Xc - X .
(YD - y)2/(Iy2 ;
A-3
A-4
A-5
A-6
A-7
A-8
T-Test. 95% Confidence
Number 1'-Value
of for n-2 Degrees
Pairs of Freedom
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
>22
>32
12.706
4.303
3.182
2.776
2.571
2.447
2.365
2.306
2.262
2.228
2.201
2.179
2.160
2.145
2.131
2.120
2.110
2.101
2.093
2.086
2.06
2.04
If delta = less than some tolerance
level, stop; A-9
DL = YD; if not, replace YD by new YD
new YD = YD + delta/B. A-10
If number of iterations = less than
50, go to Equation A-7.
Reference
1. Hubaux, A., and G. Vos, Decision and
Detection Limits for Linear Calibration
Curves, Anal. Chem., 42 (8), 849 (1970).
-------
Jan. 1981
Part I - Appendix B
Appendix B
Methods Validation Study for Lead Analysis by Flame Atomic Absorption
First the flame atomic absorption
instrumental conditions were optimized,
and then a 7 point calibration curve for
lead (Pb) at concentration of 0.3 to 10 0
microgram/ml was plotted. Each con-
centration was analyzed sequentially
five times.
The strip chart recording of the raw
data is in Figure B-1. Peak heights were
measures of responses. A scale expan-
sion of 10 times was used, so instru-
mental response was approximate
0.0039 absorbance units/cm. Data
were tabulated and calculated (Section
4.2.3 of this manual), and summarized
in Table B-1. The data would be useful in
documenting instrument performance
for future reference and for communi-
cating accuracy and precision to people
outside of the laboratory. The accuracy
(% recovery) and precision (% std. dev.)
as functions of concentration are
plotted in Figure B-2. The "3s value"
would be useful as an upper control
limit in real-time quality control (Section
4.3.5).
I L
I1AIAMAIIK
t
I I I I I I I I I I I I I I I I I I I I I I I I I I I
1 1 I I I I I I I
figure B-1.
110
grooi
I 951
i 85
90\
1\ 2\ 3\ 4\ 5 6 7\ 8 9 10
Pb Concentration tig/ml
\
18
16\
14
12
10
8
6
4
2
1 2 3 4 5 6 7 8 9 10\
Pb- Concentration ng/.ml
Figure B-2. Accuracy and precision\of Pb analysis as a function of concentration.
-------
Part I - Appendix B
Jan. 1981
Table B-1
Precision-Accuracy Methods Validation Form
Cone.
ffjg/ml)
Curve0
Parameters
Slope
Intercept
Error
Con. Coef
Det. Limit
Instrumental Response (cm)
Curve 1
Curve 2
Curve 3
Curve 4
Curve 5
Spf>rif><;
Date
Srxlo*
{Analyst Signature)
Cu
Slope
Intercept
Error
Correlate
Detect/or
rve Parameters Overall
^n Cnef
I imit
Average +
Standard Deviation
3 a Value"
Regression Concentration"
Cone.
Taken
fag/ml)
Curve
1
Curve
2
Curve
3
Curve
4
Curve
5
A verage
Cone.
Found
Standard
Deviation
%
Recovery
% Std.
Dev
"For Dionex, indicate approximate umho/cm full scale (fjmho/cm scale x volts on recorder); ForAA, indicate scale expansion;
For Technicon, indicate method and flowcell length.
T/je curve parameters are calculated using the linear least squares fit of Appendix D
c From a linear least squares fit of all data points.
a'This is the average value+ 3x(standard deviation).
"This calculated from the instrumental responses given above, using the "Curve Parameters Overall".
f'Not included in linear least squares fit.
-------
March 1983 i Part II
United States Environmental Monitoring Systems
Environmental Protection Laboratory
Agency Research Triangle Park NC 27711
Research and Development EPA-600/4-82-042b Jan. 1981
svEPA Quality Assurance
Handbook for Air Pollution
Measurement Systems:
Volume V. Manual for
Precipitation Measurement
Systems
Part II. Operations and
Maintenance Manual
-------
Part II ii March 1983
Notice
This document has been reviewed in accordance with U.S. Environmental
Protection Agency policy and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
-------
March 1983 iii Part If
Foreword
Measurement and monitoring research efforts are designed to anticipate
potential environmental problems, to support regulatory actions by developing an
in-depth understanding of the nature and processes that impact health and the
ecology, to provide innovative means of monitoring compliance with regulations
and to evaluate the effectiveness of health and environmental protection efforts
through the monitoring of long-term trends. The Environmental Monitoring
Systems Laboratory, Research Triangle Park, North Carolina, is responsible for
development of: environmental monitoring technology and systems; agency-wide
quality assurance programs for air pollution measurement systems; and technical
support to EPA's Office of Air, Noise and Radiation, Office of Toxic Substances, and
Office of Enforcement.
This manual has been developed to assist agencies which plan to make
precipitation measurements. The standard operating procedures in this manual
currently are EPA's recommended methods for precipitation monitoring. This
manual, when used with the Quality Assurance Manual for Precipitation
Measurement Systems, should be the basis for planning a precipitation monitoring
effort.
Thomas R. Hauser, Director
Environmental Monitoring Systems Laboratory
Office of Research and Development
-------
Part II iv March 1983
Abstract
This manual presents techniques and procedures for field and laboratory
operations associated with precipitation monitoring and analysis. Where
applicable, the analyses given are those approved by the U.S. Environmental
Protection Agency for water and wastes. The methodology described will provide
guidance for personnel and will maximize the quality as well as the quantity of data
collected.
-------
March 1983 v Part II
Contents
Section Pages Date
1.0 Introduction 2 1/1/81
2.0 Field Operations 13 1/1/81
3.0 Laboratory Support Operations for the Field 10 1/1/81
4.0 Laboratory Procedures 21 1/1/81
-------
Part II vi March 1983
Acknowledgement
This report was prepared for the Environmental Monitoring Systems Laboratory,
Research Triangle Park, North Carolina, under the direction of the Project Officer,
John C. Puzak. Material in this manual has been based on current EPA quality
assurance procedures, air and water monitoring methodology, and procedures
employed in the Electric Power Research Institute (EPRI) Acid Precipitation Study,
the National Atmospheric Deposition Program (NADP), and the Multi-State
Atmospheric Power Production Pollution Study (MAP3S). The authors wish to
extend grateful appreciation to the contributions of the staff at Rockwell
International's Environmental Monitoring & Services Center, Newbury Park,
California, and the many reviewers of the drafts of this document.
-------
Jan. 1981
Part Il-Section 1.0
1.0 Introduction
Pollutants are removed from the
atmosphere by wet and dry deposition;
these pollutants include substances
beneficial (nutrients) to the environ-
ment as well as those potentially
harmful. The harmful pollutants such as
formed from sulfur dioxide (SOa) acid
sulfates and nitrates and nitrogen oxide
(NOX) emissions, give rise to the
phenomenon commonly called acid rain
or precipitation. Precipitation is acid if
its pH, where pH = -log H+, is less than
5.6—the value due to dissolved at-
mospheric carbon dioxide (COa) alone.
Acid precipitation can affect the
ecology in sensitive unbuffered areas.
The losses of fish populations from
lakes in Scandinavia and intheAdiron-
dacksof upper New York are ascribed to
acid precipitation. Other harmful effects
on plant life and soil are being studied
but are not well understood at present.
Acid precipitation on materials of
contruction (both manmade and natural)
causes economic and esthetic losses.
Because of increasing awareness of
potentially harmful effects of acid
precipitation on the ecology, the numbers
of monitoring networks and effects
studies are increasing. Monitoring
networks require uniform, systematic,
and approved procedures for the ac-
quired data to be accurate and directly
comparable among networks.
1.1 Purpose
This operations and maintenance (O
& M) manual is written to guide
personnel involved in precipitation
chemistry monitoring while they carry
out various tasks and procedures. It is
highly recommended that all field and
laboratory operators take a course given
by a field manager or other knowledge-
able personnel in the pertinent proce-
dures discussed in this manual. The
operations covered herein include:
1. Field operations,
2. Laboratory field support opera-
tions, and
3. Laboratory operations.
Quality assurance aspects of the opera-
tions are presented in the Quality As-
surance Manual for Precipitation Mea-
surement Systems (1). However, to pro-
vide a better understanding of the meth-
odology, some duplication occurs be-
tween the quality assurance and the 0
& M manuals.
The basic goals of the procedures are
to collect representative samples with-
out 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 the basic goals, and must be
reliable. The Aerochem Metrics auto-
matic sampler has been tested and
accepted by most U.S. monitoring
networks; thus sampler discussion is
limited to this model.
For a rain gauge, pH and conductivity
meters, a balance, graduated cylinders,
or other equipment generally required
at a precipitation monitoring station,
various makes and models can meet the
goals; thus no particular models are
referred to, and the discussion is
general.
The material in this manual is based
primarily on the procedures in the EPRI
(Electric Power Research Institute) Acid
Precipitation Study in the Northeastern
United States, in NADP (National
Atmospheric Deposition Program), and
in MAP3S (Multi-State Atmospheric
Power Production Pollution Study).
The EPA handbooks for air pollution
measurements (2,3) and for water
measurements (4) were used as guides
for format and content. The analytical
procedures are based on those in the
water and wastes handbook(4). To have
the 0 & M manual stand alone without
requiring referrals to the other EPA
handbooks, some duplication of material
was required; this material is refer-
enced.
1.2 Collection Sites
Collection sites must be located to
meet the objectives of the monitoring
program—for example, baseline, re-
gional or urban, and siting criteria in
Section 5.6 of the quality assurance
manual(1). The site must yield repre-
sentative samples so must not have any
obstructions which will effect the
results. The general rule for placement
of samplers is: the distance between an
obstruction and the collector should be
at least twice the height of the larger of
the two objects. For objects of the same
or similar heights (e.g., a collector and a
rain gauge), the distance between the
two should be at least the height but
preferably twice the height of the object.
1.3 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 measured in the
field.
1. Sulfates (S04==) - Concentrations
above the baseline values are
caused mainly by human activities,
principally by the release of SOa
during the burning of fossil fuels
and during refining processes; the
SO2 is oxidized to sulfate in the
atmosphere.
2. Nitrogen Compounds (NOs", NH/)
- NOX (essentially NO + NOa)
concentrations above the baseline
values are caused primarily by the
burning of fossil fuels, such as for
transportation purposes; NH3 oc-
curs chiefly from biochemical
reactions.
3. Chloride Ion (CT) - Originates
chiefly from sea salt aerosols.
4. Phosphate (orthotribasic PO4~3) -
Source is soil, rock, and fertilizers;
an important nutrient.
5. Metal Ions (Na+, K+, Ca++, Mg++) -
Na+ originates mainlyfrom sea salt
aerosols, but all of these ions can
originate from soil dust in desert,
semiarid and intensively cultivated
areas.
6. Acidity - Both SO2 and NOz form
the acids found in precipitation.
7. Alkalinity - Calcareous material
(e.g., soil carbonate (CDs')), can
make precipitation alkaline, and
can neutralize the effects of acids.
8. pH - A quantitative measure of
precipitation alkalinity or acidity. In
a clean atmosphere, a water
sample in equilibrium with at-
mospheric CO2 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. Electrolytic Conductance - The
reciprocal of the resistance of a
solution and a measure of the
concentrations of dissolved salts.
10. Precipitation Amount - Value
required both to calculate the
weighted mean values of the
constituents and to derive the total
amount of materials deposited
over a time period.
1.4 Sampling Periods, Defini-
tion 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
-------
Part Il-Section 1.0 2 Jan. 1981
interval, commonly at least 6 h in the
winter or at least 3 h in the summer. The
sampling schedule depends on the
objectives of the program and on the
available funds. If rain data need to be
correlated with aerometric and/or
meteorological data, a subevent, event,
or at most a daily schedule must be
used. If the objective is to measure the
amount of deposition and/or its effects,
a weekly sample is generally sufficient.
Sampling periods longer than 1 week
are not advisable because important
changes can occur while the sample
remains in the collector.
1.5 References
1. Quality Assurance Manual for
Precipitation Measurement Sys-
tems, U S. Environmental Protec-
tion Agency, Research Triangle
Park, N.C., EPA-Draft.
2. Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems - Vol. I - Principles, U.S.
Environmental Protection Agency,
Research Triangle Park, N.C., EPA-
600/9-76-005 (March 1976).
3. Quality Assurance Handbook for
Air Pollution Measurement Sys-
tems — Vol. II — Ambient Air
Specific Methods, U.S. Environ-
mental Protection Agency, Re-
search Triangle Park, N.C., EPA-
600/4-77-027a (May 1977).
4. Methods for Chemical Analysis of
Water and Wastes, U.S. Environ-
mental Protection Agency, Cincin-
nati, OH, EPA-600/4-79-020
(March 1979).
-------
Jan. 1981
Part Il-Section 2.0
2.0 Field Operations
Procedures for field sampling, mea-
surement, and tests cover equipment
operation and maintenance; sample
collection, handling, preservation,
storage, and shipment; documentation;
and quality control procedures Precipi-
tation samples are very dilute so large
errors can occur due to contamination
or degradation. Procedures should
describe sample collection and handling
in a way that preserves integrity and
identity.
2.1 Equipment and Supplies
2.1.1 Station Supplies
The equipment and supplies required
depend on station classification and on
sampling objectives. Equipment and
supplies for a weekly precipitation (the
most common schedule) sampling sta-
tion are listed in Table 2-1. Equipment
required for event or daily precipitation
sampling is similar. Similar storage and
shipping procedures are also recom-
mended for event or daily sampling. For
sequential sampling, the list in Table 2-
1 should include a different type of
collector, a lower capacity (2.6 kg) more
sensitive balance, polyethylene bottles
with caps, means for storing and
shipping the sample in a cold state
(polyfoam-insulated containers and
freeze-gel packs), and few, if any,
buckets. If meteorological and/or
aerometric measurements are made,
the appropriate instruments must be
included in the list; however, these
instruments are not discussed in this 0
& M manual.
2.1.2 Spare Parts (Off Site)
Except for sampler fuses, which
should be kept at each station, spare
parts and supplies should be available
from the field manager or the central
analytical laboratory when needed.
Supplies for a network of 10 to 12 sta-
tions are listed in Table 2-2. The
polyethylene bottles are for special
sampling studies and/or for sample
storage in the laboratory.
No station should keep a spare pH
electrode. Electrodes in contact with
solution have a limited life because the
wet glass membrane ages. Only elec-
trodes that can be stored in a dry state
have a long shelf life. However, elec-
trodes should not be emptied, cleaned
and filled with electrode solution by the
station operators. When an electrode
breaks or becomes suspect (Section
2.6.3.4), it should be replaced with a
Table 2-1. Field Equipment List for Each Station
Equipment/Material
No./Site
Automatic precipitation sampler (Aerochem Metrics 301)
Collection containers (3.5 gal) for sampler and lids
Fuses for sampler
Recording rain gauge with event marker
Rain gauge mount
pH meter, electrode
Buffer. pHS.O. 4.O. 6.0, 7.0. and 8.0 (1 liter)
Conductivity meter and cell
Standard KCI solution, 74 umho/cm (50O mil
Temperature probe
Balance (20 kg capacity) or graduated cylinder (2 liter)
Set attachment weights for balance (1,2,2,5,10 kg)
Mailing cartons
Wash bottle
Test tubes, plastic (17x100 mm) disposable, or vials (35 ml)
Test tube rack
Rain gauge charts (package of 100)
Gummed labels
Envelopes
Notebook
Data forms
Kimwipes or other tissues (boxes)
Shipping tape (rolls)
Mallet, rubber
Equipment/Material
Sampler fuses (12)
Sampler sensor and motor box
Rain gauge clock
Rain gauge chart clip
Rain gauge chart paper (package of 100)
Rain gauge pens and ink (set)
pH meter
pH electrode
Buffer, pH 3.0, 4.0, 6.0, and 7.0 (1 gal)
Conductivity meter
Conductivity cell (cell constant ~1)
Standard KCI solution, 74 umho/cm
Shipping cartons and collection containers
Polyethylene sample bottles
16 oz
8 oz
4oz
2oz
Temperature probe
Test tubes, plastic (17x100 mm) disposable, 35 ml vials
Spare sampler buckets and lids
Gummed labels
Envelopes
Notebooks
Data forms
Kimwipes or other tissues (boxes)
Shipping tape (rolls)
1
5
2
1
1
1
1
1
1
1
1
1
3
1
375
1
1
300
300
1
300
15
3
1
Table 2-2. Supplies List for a Network of 10 to 12 Stations.
No.
1
1
1
3
3
1
1
3
2
1
2
(a)
36
600
600
600
200
3
7000
48
1000
WOO
12
300
36
12
"Make up as needed
-------
Part Il-Section 2.0
Jan. 1981
new tested electrode from the central
laboratory.
2.2 Sampler
2.2.1 Description
The Aerochem Metrics sampler
(Figure 2-1) for precipitation collection
has two containers and a common lid.
The lid seals the wet sample bucket
when precipitation is not occurring, and
thus minimizes evaporation and con-
tamination by dry deposition or dustfall.
When precipitation occurs, the lid
moves off the wet bucket and covers the
dry deposition bucket. The two buckets
collect wet and dry deposition, respec-
tively; they are fabricated of polyethy-
lene, which is inert to inorganic
substances (1,2). For organic constitu-
ents, glass or stainless steel containers
should be used. The common 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, the motor is actuated to
lift the lid from the collection bucket.
The sensor contains two heating
circuits: one goes on when the temper-
ature falls below approximately 2°C to
melt snow or ice on the sensor, 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 I id and by a spring load;
however, with strong winds the lid may
wobble, and some contamination may
enter the wet bucket.
2.2.2 Sampler Assembly, Operation.
Installation, and Servicing
For sampler assembly, operation,
installation, and servicing, see the
manufacturer's instructions.
2.2.3 Sampler Installation (A dditional
Instructions)
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; it may be shielded from the wind,
but it should not be put in an area where
excessive turbulence will be caused by
the shield or where there are obstruc-
tions such as trees and buildings
(Reference 3, Section 5). The rule for
obstructions is that the distance be-
tween large obstacles and the collector
(or rain gauge) should not be closer than
two times the difference between the
collector and obstruction heights, or the
angle subtended by the obstacle at the
site should be less than 30°. For the
placement of any neighboring collectors
and ram 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
Lid moves from one bucket to another.
Thermistor sensor-
plate activates
moveable lid when
wet precipitation
occurs
Plastic Buckets
Aluminum
Table
Motor Box
(under table top)
Figure 2-1. Aerochem Metrics, wet/dry precipitation collector.
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 perpendicular-
ly to the prevailing winds or with the dry
bucket downwind of the wet bucket. The
ground surface around the collector and
rain gauge should have natural vegeta-
tion 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.
2.2.4 Acceptance Tests
Sampler 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 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 freez-
ing and (5) lid cycling and sealing
observation.
1. 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 manufacturer's instructions
and Section 2.2.5.1.
2. Affix a temperature probe (ther-
mistor, thermometer, or ther-
mocouple) 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
clirnb 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, marked TH,
inside the sensor box. Directions
are given in the manufacturer's
instructions.
3. Remove the shorting object. The lid
should close immediately, and the
temperature should fall to ambient.
-------
Jan. 1981
Part Il-Section 2.0
4. During steps 2 and 3, check that
the lid does not cycle. Also check
the lid seals.
5. If the lid does not seal the wet
bucket, check to see if the plastic
foam gasket is secured in the
correct postion. To remove the
seal, see the manufacturer's in-
structions and Section 2.2.5. If this
is not the problem, contact Aero-
chem Metrics, 6832 SW 81 St.,
Miami, FL 33143, telephone (305)
661-5213.
6. 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.
7. Check the sensor heating circuital
freezing temperatures for the
Aerochem Metrics collector by
using its standard heater/am-
meter 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 unscrew-
ing 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, eitherthe problem must 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.
2.2.5 Sampler Servicing (Additional
Instructions)
2.2.5.1 Collector Motor Box Removal—
1. With the collector power discon-
nected, unplug the electrical con-
nector from the motor box.
2. Use a 7/16 in. hex wrench to
remove the bolt which connects
the connecting rod to the crank
arm.
3. Place the collector lid in the middle
position, and remove the wet
bucket.
4. RemdVe the middle screw under
the bucket.
5. Hold the motor box, and remove
the remaining three (3) screws on
the top outside surface of the
collector.
2.2.5.2 Collector Lid Pad Removal—
1. With the collector power discon-
nected, place the collector lid in the
middle position.
2. Remove the two (2) screws on the
edge of the lid.
3. Remove the two (2) L-brackets into
which the screws were threaded.
4. Remove the lid pad by prying it
gently along its edge with a coin or
a screwdriver.
2.2.6 Calibration
The precipitation collector does not
require calibration, but its proper func-
tioning should be checked frequently.
2.2.7 Winterizing the Sampler
Precipitation collection in the winter
encounters more problems than in
other periods. Snow may fall off the
sensor before melting, and it may blow
out of the collector bucket during heavy
winds. In addition, the collector lid arms
can freeze to the table, and the lid can
freeze to the collector bucket. The last
two problems are addressed here.
2.2.7.1 Prevention of Lid Freezing—
To prevent the lid from freezing to the
bucket, the following is recommended
(4):
1. 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.
2. Cut a small notch in one corner of
the roof to insert a power cord.
3. Attach the power cord inside the
roof to an air thermostat (Honey-
well or WRAP-ON) set for about
2°C (36°F); and tape the cord to the
roof arms.
4. Usea60-Wor75-Wlightbulbasa
heater; set the bulb on a piece of 9
mm (3/8 in.) Styrofoam on the lid
top to prevent a hot spot.
5. Install a piece of 18 mm (3/4 in.)
Styrofoam under the slope of the
roof to minimize heat loss.
6. To compensate for the additional
weight on the lid, add two large U-
bolts to the counterweight shaft.
2.2.7.2 Prevention of Lid Arms Freez-
ing to Table—To prevent freezing of the
lid arms to the table, insulate one from
the other.
1. Wrap and tape a plastic sheet
around each lid arm to make a
boot.
2. Tape one end of the boot to the
table and the other end to the arm.
3. 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.
2.3 Rain Gauge
2.3.1 Description
To reference all the precipitation
amounts against a standard, use a
recording rain gauge 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)
but 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, 4% for rates of 75 mm/h (3
in./h), and 6% 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 Aerochem
Metrics recommends the Belfort 5-780
series, 0 to 30 cm (0 to 12 in.) dual
traverse weighing gauge.
The rain gauge measures the amount
of precipitation. The precipitation col-
lector collects the sample for chemical
analysis. The two devices are not
interchangeable.
The recording rain gauge should have
an event marker pen to indicate when
the collector is open or closed. The times
can be read off the 8-day chart.
Aerochem Metrics will interface the
Belfort rain gauge with its collector. 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 time. Care must be taken to use
the correct event beginning or ending
time. Since the operator is seldom
present to observe the collector be-
havior during an event, the event
marker pen is invaluable for indicating a
collector malfunction.
2.3.2 Rain Gauge Assembly, Opera-
tion, Installation, and Servicing
The rain gauge assembly, operation,
installation, and servicing should* be
-------
Part Il-Section 2.0
Jan. 1981
performed according to the manufac-
turer's instructions.
2.3.3 Rain Gauge Installation (Addi-
tional Instructions)
The rules for obstructions (Section
2.2.3) pertain here also. 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 two.
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, welded to a 5.1 cm (2 in.)
diameter 1.0 m (3.5 ft.) pipe. The pipe is
sunken 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 sampler. 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 areas, a wind shield (e.g.,
swingleaf wind shields such as the one
used by the U.S. Weather Service)
should be used for 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. The
ground surface around the rain gauge
be natural vegetation or gravel; paving
may cause splashing of dirtorwater into
the gauge. Never oil any part of the
gauge except for the chart drive, and oil
this only when necessary with a light
machine oil.
2.3.4 Acceptance Tests
Rain gauge acceptance tests should
include checks on (1) sensitivity and
accuracy, (2) clock function, (3) pen and
recorder functions and (4) event pen
function.
1. With the weighing rain gauge level
and zeroed, add water equivalent
to several inches. For the Belfort
rain gauge 5-780 series, 1 in. =
820 g.
2. If the rain gauge does not read
correctly, adjust it according to the
manufacturer's instructions (Sec-
tion VI, Belfort manual).
3. With the pens inked and a chart in
place, turn the drum to produce a
zero-level trace; add water equiva-
lent 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.
For tipping bucket gauges, add
water in 0.25 mm (0.01 in.) incre-
ments, and note when the bucket
empties.
4. Wind the chart drive (or clock) until
it is fully wound, and set it for the
correct time. 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.
5 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 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.
6. To check the event pen function,
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 Aerochem Metrics.
2.3.5 Calibration
The rain gauge should be calibrated at
each 25 mm (1 in.(level according to the
manufacturer's instruction manual
after installation and at least at semi-
annual intervals thereafter. In addition,
checks should be made at half and full
scale (on the 0 to 150 mm or 0 to 6 in.
range) at monthly intervals. For the
weighing gauge, a set of weights should
be available from the manufacturer for
the checks; if the weights are not
available, weighed quantities of tap
water can be used. For the Belfort
gauge, 25.4 mm = 1 in. = 820 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; how-
ever, this range is generally not needed
because 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 (Section
2.3.6). 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.
Theiipping bucket gauge is calibrated
by adding a controlled volume of water,
using a slow drip technique. It may be
necessary to adjust the set screws
which limit the travel range of the tilting
bucket
2.3.6 Winterizing the Rain Gauge
In the winter, rain gauge problems
can be due to (1) snow filling or drifting
out of the gauge, (2) freezing of the
collected precipitation which can dam-
age the gauge bucket, and (3) the cold
affecting the clock and/or ink.
For the weighing gauge, thefollowing
precautions should be taken:
1. Remove the funnel, which is
generally in the inlet mouth.
2. Add approximately 1600g (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
10W motor oil. Do not adjust the
gauge reading after adding the
antifreeze. The gauge will 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 ap-
proximately 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.
3. In extremely cold periods, the clock
(if not new) may run slowly, and/or
the ink may not flow. Low-tempera-
ture ink is available from the rain
gauge manufacturer. If the clock is
a problem, replace it.
For the tipping bucket rain gauge,
winterizing consists of using a heater to
melt the snow in the funnel and the
bucket.
-------
Jan. 1981
Part Il-Section 2.0
2.4 Routine Sampler and
Weighing Rain Gauge Checks
and Maintenance
The following tests or procedures
should be carried out on a routine basis
on the precipitation collector and rain
gauge.
2.4.1 Sampler Tests
1. Collector Sensor Test - At daily or
weekly intervals, 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 immediately,
and the sensor should cool. 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 Aerochem Metrics model has),
check the bucket after an event or a
time period in which an event
depositing more than 0.25 mm
(0.01 in.) of precipitation has
occurred to ascertain if the bucket
contains or did contain any precipi-
tation. 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 Dry Sample Bucket - At
weekly intervals, if no event has
occurred, return the sample bucket
to the laboratory and/or test it for
cleanliness. To test it in the field,
add 250 ml of deionized distilled
water, swirl the bucket so that its
interior is washed, and measure
the specific conductance of the
solution. If the conductance is over
2 umho/cm, rinse the bucket until
the rinse water conductance is
less than 2 umho/cm. Conductiv-
ities greater than 2 umho/cm
indicate that the bucket is con-
taminated 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 hand-
ling problem are probably present
and must be corrected.
4. Minimizing Lid Lifting by Strong
Winds - Where strong winds are
common, check the lid to be sure it
does 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.
5. Replacement of Collector Lid Seal -
Replace the plastic foam undersea!
on the lid annually or as needed
(Section 2.2.5) because it will
deteriorate with time.
6. Examination of the Event Pen
Marker Trace - At weekly intervals,
inspect the event marker trace to
see if the lidcycled.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 collec-
tor problem. No lid movement
traces when the sample weight
trace shows that a n event occu rred
also indicates a collector malfunc-
tion.
7. Cleaning Techniques and Sched-
ule - Wash the collector rain
sensor monthly with deionized
water to remove dirt, salt, and film
buildup. If a film persists, clean the
sensor grid and plate with deter-
gent and a toothbrush. Wipe the
rim of the dry bucket weekly with
clean tissues (e.g., Kimwipes) to
prevent carry over of dustfall to the
sealing gasket and then to the wet
bucket.
8. Winter Maintenance (Section 2.2.7)
- Check the sensor temperature
when the ambient temperature
falls below freezing to ensure that
the heater is working. This may
best be done by adding snow to the
sensor. If freezing occurs, encase
the lid arms in plastic, and tape one
end of the boot to the table. If
necessary to prevent the lid freez-
ing to the bucket, attach a heater
(e.g., a 60-W light bulb)to the top of
the lid.
9. Lid Cycling - As a common occur-
rence, 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 oc-
curence, lower the temperature by
turning the screw marked TH (see
manufacturer's instructions). Sec-
ond, the lid arm can be loose or too
far out from the magnetic switch in
the motor box. Third, the switch
may be bad. (For the last two, see
Section 2.2.4, step 5).
10. Lid Malfunctioning - Another
common source of collector prob-
lems is a faulty sensor. The lid may
remain open, not open or open only
periodically. The lid staying open
indicates a shorted rain sensor. A
short can be verified by unscrew-
ing the sensor cannon connector
at the motor 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.
11. 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.
12. Other Tests - Any other mainte-
nance advised by the equipment
manufacturer should be carried
out at the recommended time
periods.
2.4.2 Weighing Rain Gauge Tests
1. Adjusting the Zero Setting of the
Rain Gauge - 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 Rain Gauge 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 Rain
Gauge Clock - Weekly, 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. 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. = 820g.
5. Inspection of Rain Gauge Pens and
Ink - Weekly, inspect the pens to
see if they have ink a nd are writing.
If they are not writing, clean the
pens, refill them, and be sure they
-------
Part Il-Section 2.0
Jan. 1981
are working. To help start the pens
writing, use a flat toothpick to
make the ink from the pen reser-
voir form a small pool at the point
of contact between the pen and the
chart.
6. Rain Gauge Chart Replacement -
At the prescribed interval, generally
weekly, remove the old chart and
replace it with a new one. Close
the access door to the chart.
7. Rain Gauge Level Check - At
bimonthly intervals, measure the
gauge level to ensure that it is still
horizontal.
8. Rain Gauge Calibration - At 6-mo
intervals (unless test 4 shows it is
necessary sooner), calibrate and
adjust the rain gauge at each 25
mm (1 in.) level.
9. Winter Maintenance - Remove the
funnel that is generally in the inlet,
and add ethylene glycol antifreeze
to the pail (Section 2.3.6).
10. Other Tests - Any other mainte-
nance recommended by the manu-
facturer should be carried out.
2.4.3 Corrective Action
Record any indication of a malfunc-
tion in the logbook. Attempt to diagnose
and correct the problem as soon as
possible. If the problem cannot be
corrected, ask the field manager or the
manufacturer for advice and direction.
Record the diagnosis and corrective ac-
tion taken in the logbook.
Collection and
2.5 Sample
Handling
2.5.1 A voiding Contamination
Careful handling of equipment and
samples to prevent contamination is
extremely important. The dissolved
substances have very low concentra-
tions, 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 laboratory. Only the
materials (e.g., sample buckets, elec-
trodes, 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; either can
be purchased at markets or drug stores.
The water should have a conductivity
less than 1.5 umho/cm; if not, try
another source. Never use tap water for
cleaning.
2.5.2 Sampling Schedules
Sampling schedules generally used
include weekly, daily, event, and
subevent. The sample should be re-
moved at the same time 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 weekly
schedules, remove the sample on the
same day each week, and never extend
the period more than 24 h. For event
schedules, remove the sample immedi-
ately 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
made.
The samples are identified and
measured for amount, pH, and con-
ductivity. They are then sealed and
stored in a refrigerator until shipment
(Sections 2.6 and 2.7).
2.5.3 Collection and Handling Pro-
cedures
2.5.3.1 Wet Bucket Labeling—Im-
mediately before use, label the sealed
sampler wet bucket (or for sequential
sampling, the capture bottles). The label
should contain the station identifica-
tion, the date used, 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.
2.5.3.2 Container and Sample Hand-
ling—The containers for the wet samples
have been cleaned and do not require
rinsing in the field before use. Never
substitute a dry bucket for a wet bucket.
At all times, take care not to contact the
inside walls of a container, a lid or a cap
with any object-especially one's finger
which can leave a deposit of salt and oil.
The container should be capped until
immediately before use, and must be
resealed immediately after use. Since
human breath contains ammonia, do
not exhale into a container.
1. Do not remove a clean bucket from
the plastic bag in which it is
shipped until it is to be placed in
the sampler.
2. Check the wet sample 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
FROM THE WET BUCKET WITHOUT
SHORTING THE SENSOR.
3. If the wet sample bucket contains
precipitation, remove it from the
collector, and replace it with a
clean, weighed, labeled bucket.
4. Remove the lid from the new
bucket after it has been placed in
the sampler, and cover the re-
moved sample bucket with the
new lid to minimize the chance of
contamination from the lid Fasten
lid on with masking tape.
5. Removeandreplacetheraingauge
chart. Record readings (times of
start and end of precipitation) on
data form. For the final amount of
precipitation reading, us6 the
maximum value on the rain chart
at end of event because loss of
water by evaporation will occur on
standing.
6. Slwirl the sample to help assure
that it is homogeneous
7. H' 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 add new antifreeze.
8. If no sample is present, seal the
empty bucket and return it to the
laboratory, or rinse it in the field for
reuse. In the first case, proceed to
Section 2.7.3; m the latter case,
rinse the bucket with 250 ml of
distilled or deionized water, and
measure the conductivity of the
last rinse. This measurement
provides a blank which reflects the
previous cleaning, operator hand-
ling, collector sealing, and so forth.
Check the conductance before
using the distilled deionized water.
After the container is cleaned,
shake it dry and reuse it.
9. For event or daily sampling, mini-
mize the number of buckets re-
quired as well as storage and ship-
ment space, by transferring the
sample from the bucket after it is
weighed to a 500-ml labeled, wide
mouth polyethylene bottle. If suf-
ficient sample is present (e.g.,
more than 300 ml), use some
(about 50 ml) to prerinse the
shipping bottle. One 500-ml bottle
per event is sufficient sample for
all measurements; the rest of the
sample may be discarded. Make
the sample transfer directly from
the bucket to the bottle, and rinse
the bucket with distilled deionized
water before reuse
10. For sequential samples, which are
collected through a funnel directly
into prenumbered, prelabeled
polyethylene bottles, seal the
bottles immediately after the
samples are collected.
The samples are now ready for field
measurement; check that the con-
tainers are sealed and correctly labeled.
-------
Jan. 1981
Part Il-Section 2.0
2.6 Field Measurements
The field measurement procedures
for weighing, pH, specific conductance,
and temperature 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 measured both in the
field and at the laboratory. These
measurements are used as a check to
detect sample changes. If less than 70 g
of sample are collected, forward the
sample to the laboratory without
measuring its conductivity and pH.
2.6.1 Weighing Sample Containers
2.6,1.1 Balance Specifications—The
balance must have a capacity of 20 kg
and a precision of at least +10 g.
2.6.1.2 Procedure—
1. With the balance level, adjust to
zero (see manufacturer's instruc-
tions).
2. Before sampling, place a new
bucket without its lid (and/or
bottle with its lid) on the balance,
and weight it 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
2.8.3).
3. After sampling, tap the lid to knock
any water drops off its inside
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.
4. After the balance has been zeroed,
place the bucket without its lid on
the balance pan, and weigh to the
nearest gram.
5. Record the weight on the bucket
label and on the field data form.
6. Subtract the initial weight of the
empty container from the final
weight of container plus sample to
obtain the sample weight.
7. Avoid breathing ontothe sample to
prevent ammonia contamination.
8. If sample is more than 70 g,
remove an aliquot of about 20gfor
conductance and pH measure-
ments. 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.
9. Seal container with lid; obtain and
record total weight to be shipped to
the central laboratory. If sample is
in bucket, secure the lid with a
rubber mallet.
2.6.2 Specific Conductance
2.6.2.1 Apparatus Requirements—
The conductivity meter and cell must
have a measurement range of 0 to 1000
umho/cm, a precision of ±0.5% of
range and an accuracy of ±1.0% of
range. The range most frequently used
is 10 to 100 umho/cm. A temperature-
compensated cell with a cell constant of
1 is preferred.
2.6.2.2 Background—The conductiv-
ity of a solution is the reciprocal of its
resistance, and is related to the total
concentration and the species of free
ions present. Since the conductivity (or
resistance) also depends on the elec-
trode area and separation as well as on
the temperature, the measuring ap-
paratus must be calibrated to obtain the
cell constant or to adjust the meter. For
calibration, use a KCI solution of known
conductivity, and be sure the tempera-
ture of the KCI standard and the sample
are the same. For rain samples, use a
0.0005M KCI solution with a specific
conductance of 74 umho/cm at 25°C.
The conductivity of the sample can be
measured on the same aliquot as used
for pH. If this is to be done, measure the
conductivity before measuring the pH to
avoid any possible error due to salt
contamination from the pH electrode.
2.6.2.3 Procedure—Measure the
specific conductance for all samples
over 70 g, using the procedure in
Method 120.2 (Specific conductance)
(5). This procedure is based on the
method in standard texts (6,7).
2.6.2.3.1 Scope and application—Th i s
method is applicable to rain, drinking,
surface, and saline waters and to
domestic and industrial wastes.
2.6.2.3.2 Summary of Method—
1. Measure the specific conductance
of a sample by using a self-
contained conductivity meter,
Wheatstone bridge-type or equiva-
lent.
2. 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.
2.6.2.3.3 Comments—
1. Standardize the instrument with
74 umho KCI solution before daily
use.
2. Keep the conductivity cell clean.
3. Minimize the temperature varia-
tions and corrections—a large
source of potential error.
2.6.2.3.4 Sample Handling and Pre-
servation—
1. Perform analyses either in thefield
or the laboratory.
2. If analysis is not completed within
24 h of sample collection, store
sample at 4°C. Wash the apparatus
with high quality distilled deionized
water, and prerinse with sample
before use.
3. Remove sample aliquot for mea-
surement, and seal the bulk sample.
2.6.2.3.5 Apparatus—
1. Conductivity bridge: range 1 to
1000 umho/cm.
2. Conductivity cell: cell constant 1.0,
or dipping-type microcell with
constant, 1.0 (e.g., YSI 3403 or
equivalent).
3. Thermometer.
2.6.2.3.6 Reagents—
1. Standard 0.0005M KCI: supplied
by the central laboratory.
2.6.2.3.7 Cell Calibration—The ana-
lyst should use the standard KCI solu-
tion and the table below to check or to
calibrate the cell, if the meter does not
have automatic temperature compen-
sation.
Conductivity of 0.0005M KCI
°C Micromhos/cm
20
21
22
23
24
25
26
27
28
66.8
68.2
69.5
71.0
72.4
73.9
75.4
76.9
78.4
If the apparatus cannot be adjusted,
record the measured value of the
standard on the data form. NOTE: Do not
insert the cell into the bulk solution.
2.6.2.3.8 Procedure—
1. Followthe manufacturer's instruc-
tions for the operation of the
instrument.
2. Allow sample aliquot to come to
room temperature (23°-27°C), if
possible.
3. Determine the temperature of
sample, ±0.5°C. If the tempera-
ture of the sample is not 25°C,
make the correction (Section
2.6.2.3.9) 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.
2.6.2.3.9 Calculation-The following
temperature corrections are based on
the standard KCI solution, and are used
with instruments with no automatic
temperature compensation.
1. If the temperature of the sample is
below 25°C, arid 2% of the reading
per degree.
-------
Part Il-Section 2.0
Jan. 1981
2 If the temperature is above 25°C,
subtract 2% of the reading per
degree. Report results as specific
conductance, (umho/cm) at 25°C
on the data form.
2.6.2.3.10 Precision and Accuracy—
1 For 41 analysts in 17 laboratories
analyzing six synthetic water
samples containing increments of
inorganic salts with conductivity of
about 100 umho/cm at 25°C the
standard deviation was about 7%,
and the accuracy was within 1 to
2%.
2. In a single laboratory (Environ-
mental Monitoring and Support
Laboratory, Cincinnati, OH) using
surface water samples with an
average conductivity of 536 umho/
cm at 25°C, the standard deviation
was +6 umho/cm
2.6.2.4 Conductance Problems and
Tests—The conductivity cell generally
has few problems; store the cells as
recommended by the manufacturer.
However, the working conductivity
standard is a very dilute 0.0005M KCI,
which may degrade slowly or become
contaminated easily. To minimize errors
due to changes in the calibration
standard, replace the 74 umho/cm
working solution at approximately
monthly intervals.
When a new working standard is
received, check it against the old
working standard. Report the measured
value of the old working standard to the
laboratory, and always return enough of
the old standard to the central labora-
tory so that it can be remeasured. Never-
return the old working standard before
checking it against the new solution
Seal and store the conductivity
standards in a refrigerator to minimize
changes. Generally, changes of less
than 5% will occur in a month and may
be ignored. If the change is more than
5%, notify the central laboratory at once
for a new standard.
If the conductivity meter has its own
builtin standardization circuit, use it to
check the KCI standard byfoflowmg the
manufacturer's instructions. If the KCI
standard has changed from its original
value by more-than 5%, inform the
central laboratory immediately. Since
the internal meter calibration is not a
traceable standard, it must not be
substituted for the KCI solution.
Another means of evaluating the
working conductivity standard is to use
the same procedure as for rain samples
to measure unknown test samples
received monthly from the central
laboratory. Return the test samples to
the central laboratory with the next
sample shipment for remeasurement. If
the laboratory finds that the station
conductivity differs from the laboratory
value by more than 10%, the laboratory
will inform the proper personnel and
replace the old conductivity standard.
Store the cells as recommended by the
manufacturer.
2.6.3 pH
2.6.3.1 Apparatus Requirements—
The pH meter and electrode must be
capable of measuring to ±0.02 unit
precision 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 less
washing than two separate electrodes.
Electrodes with shielded bulbs are more
difficult to clean, and thus are more
prone to yield errors. When a new
electrode is obtained, equilibrate it
overnight in a solution recommended by
the manufacturer before it is used.
Store the electrode in pure water, in a
74 umho/cm KCI conductivity standard,
or in a 10~4N acid (H2S04> solution.
These storage media may cause the
electrode to require longer times for
stabilization than do the concentrated
buffer solutions, but they increase the
life of the electrode.
2.6.3.2 Background—The pH of a
solution is related to the free acid
activity by the equation
pH = -log (H+)
2-1
where (H+) is the H+ activity or free H +
concentration. Thus pH does not mea-
sure the total acid concentration.
Use a pH meter which measures the
difference in electrical potential be-
tween a reference electrode and an H+
(glass) electrode. The glass electrode
potential varies as the activity (or
effective concentration)of H+in solution
varies. Although the meter actually
measures electrical potential (volts), it is
generally calibrated to give data as pH.
Each station will receive the required
calibration buffer solution from the
central laboratory. The stations should
notify the laboratory when the buffers
need to be replaced.
2.6.3.3 Procedure— The pH is mea-
sured for all samples weighing over 70
g. If the measurement is made on the
same aliquot as that used for conductiv-
ity, the pH must be measured after the
conductivity by following Method 150.2
(Electrometric) taken from Methods for
Chemical Analysis of Water and Wastes
(5). This procedure is based on the
method in standard texts (6,7).
2.6.3.3. / Scope and Application—
This method is applicable to ram,
drinking, surface, and saline waters.
2.6.3.3.2 Summary of Method—The
pH of a sample is determined electro-
metrically by using either a glass
electrode with a reference electrode or a
combination electrode
2.6.33.3 Sample Hand/ing and Pre-
servation—
1 Perform the analyses on site
immediately after sample collec-
tion.
2. After removal of sample aliquot,
seal the bulk sample container; if
the container is a bucket use a
rubber mallet to secure the lid.
2.5.3.3.4 Interferences—
1. The glass electrode, in general, is
not subject to solution interferences
from color, turbidity, colloidal
matter, oxidants, reductants, or
high salinity.
2. Sodium error at pH levels greater
than 10 is generally not a problem
since precipitation rarely has such
a high pH, but the error can be
minimized or eliminated by using a
"low sodium" error electrode.
3. Temperature effects on the electro-
metric measurement of pH arise
from two sources. The first is the
change in electrode output at
various temperatures; this inter-
ference can be controlled with
instruments having temperature
compensation or by calibrating the
electrode-instrument system at
the temperature of the samples.
The second is the pH change
inherent in the sample at various
temperatures; this error is sample
dependent, and cannot be con-
trolled; so it should be noted by
reporting both the pH and the
temperature at the time of analy-
sis.
2.6.3.3.5 Apparatus—
1. pH meter (Field Model): A wide
variety of instruments are com-
mercially available with various
specifications and optional equip-
ment. The meter should have both
an intercept and a slope adjust-
ment. Older pH meters were
calibrated by using a single buffer
and by electronically setting the
intercept; this single point calibra-
tion assumed that the correct
theoretical slope was achieved by
the electrode being used, but this
was a poor assumption because
commercial electrodes have slopes
which vary from 95% to 102% of
the theoretical value (8). Modern
pH meters therefore allow calibra-
tion at two points: one to establish
-------
Jan. 1981
Part Il-Section 2.0
the intercept and the other to
establish the slope. The instru-
ment should be left in the "stand-
by" mode, or it should be allowed
to warm up at least 1/2 h before
measurements ard made.
2. Glass electrode.
3. Reference electrode a calomel,
silver-silver chloride, or other
reference electrode of constant
potential. NOTE: combination elec-
trodes with both measuring and
reference functions are conven-
ient. The pH electrodes should be
stored in deionized water, in a 74
umho/cm KCI standard, or 10~4N
acid (H2SO<) solution. Prior to use,
the pH electrode should be thor-
oughly rinsed with deionized water.
4. Thermometer or temperature sen-
sor for automatic compensation.
2.6.3.3.6 Reagents—Standard buffer
solutions are available from the central
laboratory.
2.6.3.3.7 Calibration—
1. Because of the wide variety of pH
meters and accessories, detailed
operating procedures cannot be
incorporated into this method.
Each analyst must be acquainted
with the operation of the site
system, and must be familiar with
all instrument functions. Special
attention to care of the electrodes
is recommended.
2. Each instrument/electrode sys-
tem must be calibrated at a mini-
mum of two points that bracket the
expected pH of the samples and
that are approximately three pH
units apart. Most rainwater sam-
ples have pH values of 3.0to6.0or
4.0 to 7.0, depending on the area
sampled; therefore, the pH 3.0 and
6.0 or the pH 4.0 and 7.0 buffer
solutions are suggested.
3. Before calibration, "top off" the
electrode (if it is not sealed) with
filling solution available from the
manufacturer, and rinse carefully.
Measure both the standard buffers
and the samples at the same
temperature without allowing the
electrode to touch the bottom of
the solution vessel.
4. Various instrument designs may
use a "balance" or "standardize"
dial and a slope or temperature
adjustment, as outlined in the
manufacturer's instructions. The
pH 6.0 or 7.0 buffer is used to set
the intercept of the pH response
with the standardization knob; the
3.0 or 4.0 buffer is used to adjust
the slope of the pH response with
the slope or temperature function
control. After each calibration.
rinse the electrode thoroughly
with deionized water. Repeat the
calibration until both readings are
within +0.02 unit of the buffer
value. After the last buffer mea-
surement, measure the pH of
deionized water in several test
tubes until a stable value is
obtained to minimize the chances
of buffer contamination of any
sample. Calibrate the pH meter
before and after each set (at most
20) of samples measured at one
setting.
2.6.3.3.8 Procedure—
1. Standardize the meter and the
electrode system (Section 2.6.3.3.7).
2. Place the sample in a clean vessel
that has been rinsed with the
sample from the conductance
measurement. (Use the rinse for
the electrodes below.) Use a
sufficient volume of sample to'
cover the sensing elements of the
electrodes. Never insert the elec-
trode into the bulk sample con-
tainer.
3. If the sample temperature differs
by more than 2°C from that of the
buffer solution, correct the mea-
sured pH values. Some instru-
ments are equipped with auto-
matic or manual compensators
that electronically adjust for tem-
perature differences. Refer to
manufacturer's instructions.
4. After rinsing and gently shaking
the electrodes (do not tap on a hard
surface), immerse them into the
sample test vessel. To ensure even
contact between the sample and
the electrode(s) sensing elements,
swirl the sample gently for a few
seconds after immersion.
5. Allow the electrode(s) to equili-
brate. Do not disturb the solution
and the air-water interface while
measuring. If sample is not in
equilibrium with the atmosphere,
pH values will change as the
dissolved gases are either ab-
sorbed or desorbed. Record sample
pH and temperature inthe logbook.
2.6.3.3.9 Calculation—Read pH meters
directly in pH units. Report the pH to the
nearest 0.01 unit and the temperature
to the nearest degree (°C) on the data
form.
2.6.3.3.10 Precision and Accuracy—
1. For 44 analysts in 20 laboratories,
six synthetic water samples contain-
ing exact increments of H* - OH"
ions in the pH range 3.5 to 8.0, the
standard deviation was 0.14 unit.
2. In a single laboratory (Environ-
mental Monitoring and Support
Laboratory, Cincinnati, OH) using
surface water samples with an
average pH of 7.7, the standard
deviation was ±0.1.
2.6.3.4 Electrode Problems and Tests-
One of the chief problems that occurs
with pH measurements is the aging of
the electrode. Diagnostic tests for this
are presented here.
The first means of diagnosing elec-
trode or procedural problems is the use
of test samples sent out to the field sta-
tions by the central laboratory on a
monthly basis. The samples have rain-
type pH and conductance values, and
should be measured for both variables;
these will be measures of the station's
accuracy if the laboratory value is
assumed to be correct and if no solution
change occurs in shipment. The samples
are returned to the laboratory with the
results for recheckand evaluation. If the
field pH value differs from the laboratory
by more than ±0.15 unit, the pH
electrode probably needs replacing.
Consultation with the field operator on
the technique should confirm the
source of the problem.
A second test is the use of a reference
solution, which has a known pH and a
conductivity similar to those of rain
samples, to check the pH electrode at
biweekly intervals. The measurement
procedure should be identical to that
used for a rain sample. Store the solu-
tion in a refrigerator, and replace it
when needed or when the solution pH
or conductivity appears to have changed.
From the average value and deviation
of five or more test samples measured
consecutively 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 = [I(x, - x~)2/(n - 1)]1/2 2-2
where xi and xare the measured and the
average pH readings of the series and
where n is the number of sample tubes
measured.
For acceptable electrode behavior, s
should not be greater than 0.05 pH unit
for the test sample set, and the average
pH value should agree with the previous
month's value to ±0.10 unit. Frequently,
the first measurement will differ more
than ±0.1 unit from the others. If so, this
value should be excluded from the
average, and 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 trie data
form. The time to attain a stable reading,
(i.e., when pH is constant to ±0.02 unit
-------
Part Il-Section 2.0
10
Jan. 1981
for 1 min) should be less than 5 minutes
for a well-behaved electrode Results of
these tests serve as guides for both the
measurement technique and the equili-
bration time to be used for precipitation
sample measurements If an electrode
test at any time shows behavior poorer
than that given by the above criteria, the
electrode should be replaced. If the
average pH has changed from the
previous month's by more than 0.10
unit, check the solution conductivity. If
the conductivity has changed by more
than 10% from its original value, the
solution has degraded and should be
replaced. Always return enough of the
solution so that it can be checked by the
laboratory.
2.6.4 Temperature
2.6.4.1 Requirements—The temper-
ature probe must have an accuracy of at
least 1°C and a precision of ±0.5°C. A
thermistor, thermocouple, or ther-
mometer can be used The probe should
be calibrated yearly by the central
laboratory.
2.6.4.2 Procedure—
1. Before measuring a solution, wash
the temperature probe, and shake
it dry
2. To minimize contamination, do not
insert the probe into any solution
until after the other measure-
ments (i.e., conductivity and pH)
have been made, or use a second
sample treated in a similar manner
as the first only for temperature
measurement
3. Read and record the temperature
to the nearest 0.5°C.
2.7 Sample Indentification,
Preservation, Storage, and
Shipment
2.7.1 Background
Sample degradation can be due to
chemical interactions—for example,
with particulates or gases, or bio-
chemical reactions Preservation of
sample intergrity after removal from the
sampler can be maximized by filtration,
sealing the sample, and storage in the
dark at low temperatures (e.g , about
4°C). Do not freeze. To minimize
contamination, perform filtrations only
in the central laboratory Although
chemical preservatives are effective for
various constituents, they can interfere
in some of the measurements or
analyses, and so are used only in special.
cases (e.g., for SOa measurement).
Samples must be adequately identi-
fied 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
In the case of duplicate (collocated) or
sequential samplers, treat each sample
container as a separate sample For
duplicate samplers, distinguish the
samples by adding a -1 and -2tothesta-
tion 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) and by air
(UPS Blue or air parcel post) so that
samples are received in the laboratory
before the weekend All samples must
be well identified, and should be
accompanied by appropriate data forms.
A duplicate copy of the data form is
mailed separately to the central labora-
tory to help trace a shipment.
The laboratory, upon receipt of the
shipment, will replace the used sample
buckets or containers with clean ones
by return mail or other delivery mode.
2.7.2 Procedure
Label each sample with station
identification, date of sampling period,
and sample weight (Section 2.B.3.1).
2.7.2.1 Weekly Cumulative Samples—
1. Be sure the sample is sealed,
identified, and accompanied by its
data form
2. Pack the cumulative weekly sam-
ple collection bucket into a card-
board carton or other protective
box.
3. Seal the carton, and ship it to the
central laboratory by the method
and carrier prescribed for the
program.
4. Mail asecondcopyofthedataform
separately to laboratory.
2.7.2.2 Daily, Event or Sequential
Samples—Refrigerate event and se-
quential samples until they are shipped,
and keep them cold during shipment.
1. Be sure the samples are sealed,
identified, and accompanied by
their data forms.
2. Pack the samples incardboard-
enclosed Styrofoam boxes (Poly-
foam Packers Corp., Chicago, IL.)
with gel freeze-packs. Keep the
freeze-packs in the freezer com-
partment 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 unfrozen. Generally, four
packs per box is sufficient to keep
the samples cold for 4 or 5 days.
3 Seal the cartons, and ship by the
prescribed method and carrier.
4 Mail second copies of data forms
to the laboratory.
2.8 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 made out in triplicate (carbon
paper may be used) with one copy kept
in the siation records, one shipped with
the sample, and the last mailed sepa-
rately to the central laboratory on a
weekly basis. The logbook entries are
made out in duplicate. One copy is kept
at the station and the other is mailed
with the data forms and the ramchartto
the central laboratory
2.8.1 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 copyto
the central laboratory and keep the
other at the station.
2.8.2 Rain Gauge Charts
Mark the rain gauge chart with sta-
tion identification, dates and notations
for tests, problems, and so forth, and
submit weekly to the central laboratory.
2.8.3 Field Data Forms
The field data form must contain all of
the following information, station,
operator, date, sample weight, sampling
times, field values of pH and conductiv-
ity, problems, and so forth
Start a new form when a clean bucket
is installed in the sampler. (For daily,
event or sequential sampling use a new
form for each sample collected.) Com-
plete the form when the sample is
removed Irom the collector. An example
of a data form used by NADP for event
sampling is in Figure 2-2 The numbered
items below refer to the numbers in
Figure 2-2.
1. STATION NAME and ID supplied by
the project coordinator.
2. OBSERVER'S signature and printed
initials; person completing the
form even if substituting for regular
observer.
3. If both WET and DRY buckets to be
analyzed, indicate sample type on
the form at the beginning of the
collection period; check the wet
box for a WET sample even if no
precipitation occurred.
-------
Jan. 1981
11
Part Il-Section 2.0
4. Enter precise DATE ON and OFF
(mo/day/yr) and the local TIME
when sample buckets are installed
and removed Specify 24-h time
and indicate the TIME ZONE by
circling the appropriate three-
letter code.
5. Check appropriate boxes for the
three SITE OPERATIONS (Section
2.4). Diagnose items 1 and 3 from
the event pen trace on the rain
chart. Add evidence for item 1 —
the lack or presence of moisture in
the dry bucket and the reason-
ableness of agreement between
the collector and rain gauge a-
mounts in item 8 below. For opera-
tion 2, be sure the weight trace is
complete for the sampling time
period.
6. SAMPLE CONDITION is a qualita-
tive observation of precipitation
quality. Note any comment on
obvious causes of the condition
under item 11 REMARKS.
7. Complete the form for SAMPLE
WEIGHT by entering weight of
SAMPLE BUCKET with LID. In-
clude total weight of sealed bucket,
lid, and sample shipped, beneath
SAMPLE WEIGHT designation.
Start a new form for newly installed
bucket by entering BUCKET
WEIGHT of clean bucket (Section
2.5) with its lid. Lid used is
weighed separately and added to
BUCKET weight. Obtain the weight
of precipitation in exposed bucket
by subtracting BUCKET WEIGHT
from BUCKET + SAMPLE WEIGHT,
and entering it as SAMPLE
WEIGHT.
8. 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
samples, add all daily rain gauge
amounts, and record TOTAL SAMP-
LING PERIOD PRECIPITATION
(in.). Do not merely subtract initial
reading for week from final reading
because errors occur due to evap-
oration. Con vert TOTAL SAMPLER
PRECIPITATION amount collected
from grams to inches by multiply-
ing SAMPLE WEIGHT (Section 7)
by 0.00058 in./g, and record in
appropriate boxes.
CAL/NREL USE ONLY
BULK
DA
QA
NS/Exclude
±3-
LD
ND
1 Station
Name.
ID [
2 Observer
Signature
3 Sample Bucket
Check f—|
One Dry Side
5 Site Operations Check Yes or No for each item for wet-side samples only
if No. explain in remarks
1 Collector jftprar*, to have operated properly and sampled atl preciptl.ttion
events during entire sampling period
2 Rain gauge appears to have operated properly during the week
3 Collector opened and closed at least once during the week
4 Bucket
On
Bucket
Off
Circle Time Zone
EDJI-II MST/PDT121
EST/CDTIOI PSTI3I
CST/MDT/tl AKDT 141
AKST MST 151
American 161
Samoa
Condition Complete for all samples containing precipitation
1 Bird Droppings
2 Cloudy or Discolored
3 Lots of Soot or Dirt
Describe all else in remarks
7 Sample Weight - Grams
Only lor Buckets with water ice or snow
Bucket &
Sample & Lid
I Sample
8 Precipitation Record
For wet-side samples
only
Type
Circle one for each
day of precipitation
Amount In inches
or circle
one
R-Rain only s-Snow only m-Mixture u Unknown
z-zero t Trace mm-Missing
Bucket On To Bucket Off
Z T MM
R S M U
Z T MM
Thur
T MM
Fri
R S M U
Z T MM
R S M U
Z T MM
Sun
R S M U
Z T MM
Mon
R S M U
Total sampling period precipitation from rain gauge
Total precipitation from sampler ^ sample weight x 000058
inches/gram
R S M U
inches
tncnes
9 Sample Chemistry .____.___.
Only for wet side buckets with precipitation Conductance | I |*| I A'S/Um
Distilled water
Mo
Dav
_
rm
=
_
Aliquot removed standard certified standard measured correction factor
Correction Factor Sample Measured Sample Corrected
pH
pH 4 Observed
Sample
10 Supplies
Circle if needed
pH 4
pH 7
75 juS/cm
Field Forms
11 Remarks
For example. Contamination by operator, equipment malfunction, harvesting in
Figure 2-2. Field data form.
-------
Part Il-Section 2.0
12
Jan. 1981
9. Space is provided for ONE mea-
surement 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 mea-
surement, and record onlythefinal
value. Mention only problems in
the remarks section. (Instructions
in Sections 2.6.2 and 2.6.3.) Enter
DATE of determination as well as
total weight of sample ALIQUOT
REMOVED. Record CONDUCT-
ANCE of DISTILLED (or DEIONIZED)
WATER used for rinses and SAM-
PLE MEASURED conductance cor-
rected to 25°C. If resistance bridge
cannot be adjusted, insert mea-
sured value of 74 umho/cm stan-
dard in STANDARD MEASURED to
calculate CORRECTION FACTOR;
then calculate and record the
SAMPLE CORRECTED value. For
conductivity meters adjusted to 74
umho/cm value using KCI stan-
dard, the correction factor is unity.
The sample aliquot used for the
conductance measurement can
also be used for pH measurement.
Never return any aliquot to the bulk
sample. Avoid contaminating bulk
sample or aliquot. Measure pH of
SAMPLE aliquot (Section 2.6.3).
After the measurement is com-
pleted, recheck the pH 4 buffer
value, and enter it in OBSERVED 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.
10. Obtain SUPPLIES by circling the
appropriate material. If pH 3, 6, or 8
is needed, write it in this section.
To avoid running out, request new
material when about one-fourth of
original supply remains.
11. 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 oper-
ator, moisture in the dry bucket,
plowing, harvesting, burning, in-
creased atmospheric pollution or
dust, or power outage. The im-
portance of the information re-
quested in the remarks section
cannot be overemphasized. Care-
ful observation of the sample and
occurrences in the surrounding
environment can aid greatly in
evaluating the validity of the
sample and in the interpretation of
the data collected. For mailing
instructions see Section 2.7.
2.9 Quality Control
Quality control procedures are used
when possible to help assure the collec-
tion of high quality data Complete docu-
mentation 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 examples of quality
controls. In addition, two quality control
functions involve the testing of field
operators by outside personnel and one
function uses field personnel as middle
men in testing the central laboratory.
The first two functions use (1) unknown
test samples made up by the quality
assurance or central laboratory for pH
and conductance measurements and (2)
field audits by an experienced observer
The third merely requires the field
personnel to forward a sample received
from quality assurance personnel to the
central laboratory disguised as a regular
precipitation sample.
2.9.1 Unknown or Quality Assurance
Test Samples for the Field
To evaluate the quality of each sta-
tion's pH and conductivity measure-
ments as well as to detect problems
with these measurements, test samples
of rain-type composition should be
received from the quality assurance
laboratory on a regular (e.g., monthly)
basis.
1. Measure these samples of pH and
conductance as soon as possible
after receipt. Use the same pro-
cedure as for precipitation samples
(Sections 2.6.2 and 2.6.3).
2. Fill out items 1 and 2 of the data
form; record the data and the
results in item 9, SAMPLE CHEM-
ISTRY, and identify the sample
under item 11 REMARKS.
3. Return the results on a data form,
and the remainder of the sample by
air mail to the laboratory.
The central laboratory remeasures
the sample to be sure it has not changed
during shipmenttoand from the station.
The results indicate problems in these
measurements due to technique, equip-
ment and/or standards. If a problem is
found, it should be corrected as soon as
possible.
2.9.2 Field Audits
To review the quality assurance
system and to evaluate each station's
performance firsthand, an audit should
be conducted at least once annually by
experienced personnel from the central
laboratory. The audit should cover all
aspects of site operation.
1. About 4 to 6 weeks before the
audit, a questionnaire should be
sent from the central laboratory to
the field personnel. They fill in the
questionnaire (Section 6.7.3, qual-
ity assurance handbook (1)) 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 add a test sample
to a clean bucket at the station, and
the operator will weigh the sam-
ple, measure its pH and conductiv-
ity, and record the data on a Data
form.
3. The auditor will inspect all equip-
ment, calibrate 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
field manager.
5. The site personnel will be informed
of the results at the end of the
audit.
2.9.3 Blind Samples for the Labora-
tory
Blind (reference) 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 follow-
ing are instructions and guidelines tobe
followed:
1. A reference sample will be shipped
in a 500-ml polyethylene bottle
with two 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
(not been used in the field) sample
bucket when you are ready to
submit the reference sample;
weigh, and record as usual on a
dala form.
4. Remove your normal aliquot and
measure pH and 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. Complete the information re-
quested on the two postcards to
identify the sample, and mail the
cards.
7. Place a clean bucket in the col-
lector, and proceed as usual.
-------
Jan. 1981
13
Part Il-Section 2.0
2.10 Field Procedure Sum-
mary
To serve as an outline, an operating
procedure summary is given below. It
includes the basic .steps, but it is not
complete. The conductance and pH pro-
cedures 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 change the summary
wording if necessary.
2.10.1 Site Visits
1. Daily: check raingauge for event. If
event occurred, record date and
time, number of lid openings, and
amount of precipitation from gauge.
Note the weather. Weekly: change
chart, fill pens, and wind clock.
Monthly: check rain gauge calibra-
tion, and clean collector sensor.
2. Check dry bucketsfor moisture and
other unusual occurrences. If not
interested in dry bucket, wipe off
rim.
3. If event occurred, replace wet
bucket with new weighed one. Put
new weighed lid firmly on sample
bucket. Record observations on
data form and in logbook.
4. Check collector, sensor, and rain
gauge for problems.
2.10.2 At Site Laboratory
1. Sample handling - wipe outside of
bucket dry; tap lid to knock off
drops; remove, weigh bucket to
calculate precipitation amount (to
nearest 1.0 g). Record on sample
data form. If sample is sent to
laboratory in bottle rather than
bucket, pour sample into sample
bottle. Use smallest bottle that will
accommodate sample. Discard
sample over 16 02. bottle quantity.
Record total sample weight on bot-
tle label. Rinse sample bucketwith
deionized or distilled water, shake,
and drain. For sample in sealed
bucket or bottle, allow at least 1 h
for sample to reach room tempera-
ture.
2. Conductivity meter standardizing
for adjustable meter - Use 74
umho/cm standard. For dip tube
cell, rinse and shake test tube or
vial three times with deionized or
distilled water. Add 1.3 cm (0.5 in.)
of 74 umho/cm solution; swirl to
coat walls; drain. Add 20 ml of
solution or enough to cover elec-
trodes; insert rinsed conductivity
cell. Remove and shake; repeat
two times. 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.
Discard solution; shake cell and
tube dry. Put a second aliquot of 74
umho/cm solution in same tube;
check reading. Readjust if neces-
sary. Discard solution For closed
bottom type cell, use above instruc-
tions omitting the test tube, and
add sufficient water or solution to
cover electrodes.
2.10.3 Conductivity Test
1. 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.
2. Pour deionized or distilled water
into test tube or vial. Dip and shake
cell three times before reading. If
conductivity exceeds 1 5 umho/
cm, repeat rinses until below 1.5.
Record latter of two readings in
item 9 of field data form.
3. Drain and shake tube; shake cell
dry
4. 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 measure conductivity
and record.
5. Save this sample for pH test.
6. Rinse cell; shake, blot and store.
For closed bottom type cell, use
similar procedure, and add suffi-
cient sample to cover electrodes.
2.10.4 pH Meter Calibration
1. Rinse electrode in deionized or
distilled water and shake dry.
2. Put pH 7 buffer (or pH 6 if sample
pH range 3 to 6 is more appropri-
ate) in new tube; insert electrode,
don't touch bottom. Wait 3 min.,
adjust standardize control to 7 (or
6). Remove electrode; let electrode
dram and discard solution; repeat
with new solution in same tube.
3. Put pH 4 buffer (or pH 3 if 3 to 6
range is used) in a newtube. Rinse
and dry electrode; insert in tube,
wait 3 min. Set slope control to 4
(or 3). Rinse and let electrode dry.
4. Rinse electrode, dry, and check pH
7 (or 6) buffer. If reading is off by
more than ±0.02, repeat calibra-
tion.
5. Measure temperature of pH 7 (or 6)
buffer solution, and rinse probe.
2.10.5 pHTest
1. Rinse electrode well and let it
drain; insert in sample tube (from
2.10.3, step 5 above). Do not let
electrode touch bottom of tube.
Remove electrode; shake lightly to
remove solution drops, then rein-
sert. Remove; shake lightly; and
insert in tube.
2. WaitS to 5 min (do not stir sample)
until pH is constant to ±0.02 for 1
min, then read and record pH.
3. Remove electrode; rinse and drain;
reinsert it in pH 4 buffer for check.
If reading is off by more than 0.03
unit, recalibrate and remeasure.
4. Remove electrode; rinse and store
in test tube with 0.0001 N acid
solution (e.g., HzSCu), 74 umho/
cm KCI or deionized water. Insert
temperature probe in sample tube;
wait 1 min; read and record
temperature.
5. Remove temperature probe from
sample and rinse. Discard solu-
tion.
6. Weigh sealed sample container
after tests, and record weight on
container and on data form.
2.10.6 Storage and Shipping
1. Store sealed, labeled sample
containers in a refrigerator until
they are packed for shipment.
2. Pack and ship to the central
laboratory on scheduled day each
week by approved carrier and
method.
2.11 References
1. Galloway, J.N., and G.E. Likens,
Water, Air and Soil Pollut. 6, 241
(1976).
2. Galloway, J.N., and G.E. Likens,
Tellus30, 71 (1978).
3. Quality Assurance Handbook for
Precipitation Measurement Sys-
tems, U.S. Environmental Protec-
tion Agency, Research Triangle
Park, NC, EPA-Draft.
4. Martin, C.W. NADP Winter Opera-
tion of Sampler, Hubbard Brook
Experimental Forest, West Thorn-
ton, NH; letter to V.C. Bowersox,
March 25, 1980.
5. Methods for Chemical Analysis of
Water and Wastes, U.S. Environ-
mental Protection Agency, Cincin-
nati, OH, EPA-600/4-79-020 (March
1979).
6. Standard Methods for theExamina-
tion of Water and Wastewater,
American Public Health Associa-
tion, Washington, D.C., 14th edi-
tion, (1975).
7. Annual Book of ASTM Standards,
Part 31, Water, Standards D1125-
64 and 01293-65, p. 120, 178
(1976).
8. Galloway, J.N., B.J. Cosby, and
G.E. Likens, J. Limnol. Oceanogr.
24, 1161 (1979).
-------
Jan. 1981
Part 11-Sect ion 3.0
3.0 Laboratory Support Operations for the Field
The laboratory must supply clean
containers to the field sites, and it must
prepare standards to be used to cali-
brate field instruments as well as
quality control samples for use in the
field. This section discusses the care of
glass and plasticware, the preparation
of reference solutions, and the evalua-
tion of field equipment by the central
laboratory. The referred data forms are
in Section 3.6.
3.1 Cleaning and Supplying
of Glassware and Plasticware
3.1.1 Cleaning of New or Used Plastic-
ware
1. Rinse with deionized water 6 to 10
times. NOTE: If the plastic needs to
be rubbed to remove a film, use a
natural sponge.
2. Let stand, filled with deionized or
distilled water for 48 h. Empty and
dry in a 70°C oven.
3. 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
500 ml of deionized water to the
cleaned container, seal the con-
tainer with a cap or with Parafilm,
and slowly rotate it so that the
water touches all inner surfaces.
DO NOT SHAKE. Check the con-
ductivity of the water (Section 4.5);
it should be less than 2.0 umho/
cm. If any of the containers fail the
check, rerinse all of the containers
cleaned with the checked samples
thoroughly and retest 5%.
4. After the plasticware 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 Glassware Used for Metal
Analyses—
1. Rinse with deionized water twice
and with 10% HN03 once.
2. Rinse 6 to 10 times with deionized
water.
3.1.2.2 Glassware Used for Anions
and NH/—
1. Rinse with deionized water twice
and with 10% KOH solution once.
2. Rinse 6 to 10 times with deionized
water.
3. If water beads on the inner surface,
the glassware needs to be cleaned
more thoroughly. Wash with soap,
and then wash using the routine
10% KOH cleaning procedure. If
water still beads, soak the glass-
ware overnight in 10% KOH, and
rinse 6 to 10 times with 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
containers with clean ones. The clean,
sealed containers are shipped in plastic
bags and reusable shipping cartons to
the field sites on a routine basis (e.g.,
weekly) 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, or replace with a new one.
The shipment can be made by ground
transport since each site should have a
3-week supply of these materials on
hand.
3.2 Preparation and Mea-
surement of Conductivity Stan-
dards for the Field
1. Weigh out 7.456 g of predried (2 h
at 105°C) KCI and dissolve it in 1
liter of deionized water (0.1 OM
KCI).
2. Dilute 50 ml of the 0.1M KCI to 10
liters with deionized water (0.0005M
KCI)
3. Fill washed 0.5-liter bottles with
the 0.0005M KCI solution to be
sent to the field. Keep the solution
refrigerated.
4. Measure the actual conductivity of
the solution in each bottle (Section
4.5.2).
.5. Fill out the Field Conductivity
Standard form and label the bottle
with the measured conductivity.
6. Send new standards to the field
monthly. When old standards are
returned to the laboratory, remea-
sure the conductivity. Complete
the Field Conductivity Standard
form.
3.3 Preparation and Mea-
surement of pH Electrode
Reference Solution for Field
and Laboratory
1. Weigh out ~2.6 g (NH4)2S04 and
—2.8 g NH4HSO4, and dissolve in 2
liters of deionized water.
2. Prepare the pH electrode reference
solution by diluting 5 ml of the
stock solution (Section 3.3, step 1)
to 1 liter with deionized water.
Protect this solution from air, in a
sealed container.
3. Alternately, a 10"" to 10~5N H2S04
solution can be used for the
electrode reference solution.
4. Fill 500-ml bottles with the pH
electrode reference solution. Keep
the solutions refrigerated.
5. Measure the pH and conductivity
of the solution in each bottle (Sec-
tions 4.5.2 and 4.3).
6. Fill out the Field pH Electrode Test
Solution form, and label the bottle
with the measured pH and con-
ductivity.
7. Remeasure the pH and conductiv-
ity of these solutions after they are
returned from the field. Complete
the Field pH Electrode Test Solu-
tion form.
3.4 Preparation of Quality
Control Samples for the Field
1. Weigh out 0.70 g NH4HS04, and
dilute to 1 liter with deionized
water.
2. Weigh out 0.70 g (NH4)2S04, and
dilute to 1 liter with deionized
water.
3. Monthly, prepare an audit sample
by diluting approximately 10 ml of
each of the two stock solutions
(Section 3.4, Steps 1 and 2) to 1
liter with deionized water. Monthly,
change the volume ratio of NH4HS04
to (NH4)2SO4, or add 10~3 to 10~6 N
H2S04, to vary the samples, pH and
conductivity.
4. Alternately, 10~3 to 10~6N acid
solutions (e.g., H2SO4)canbeused.
5. Fill 60-ml bottles with the mixed
audit sample, and send each site
one sample. Three bottles should
be retained by the laboratory.
6. Immediately measure the three
samples kept by the laboratory for
pH (Section 4.3) and conductivity
(Section 4.5). Check the laboratory
electrode against another backup
electrode for one sample. Fill out
the appropriate section of the Field
Quality Control Audit Sample
Report. Refrigerate the laboratory
samples.
7. Whenthefieldqualitycontrolaudit
samples from all sites have been
returned to the laboratory, reana-
lyze the samples along with the
laboratory's three aliquots. Check
the laboratory electrode against
another backup electrode for one
-------
Part Il-Section 3.0
Jan. 1981
sample. Complete the Summary
Field Quality Control Audit Sample
Report
3.5 Evaluation of Field Equip-
ment
All meters and electrodes should be
tested before they are shipped to the
field The meters usually have a serial
number affixed, but the electrodes do
not. An identification number should
therefore be affixed to each electrode.
3.5.1 Evaluation of Conductivity Me-
ters and Cells
3.5.1.1 Prepare a 0.0003M KCI Test
Solution— Dilute 3 ml of the stock
0.10M KCI solution (Section 3.2.1 ) to 1
liter with deionized water.
3.5.1.2 Calibrate the Field Conductiv-
ity Meter— Calibrate as indicated by the
manufacturer or as described in Section
4.5.
3.5.1.3 Fill 11 Vials or Plastic
(17x1 00 mm) Tubes — Fill to a depth of 3
cm (or to cover the electrodes) with the
0.0003M KCI. The first tube is to be used
as a rinse tube.
3.5.1.4 Measure the Conductivity of
the 10 Solutions — Between each mea-
surement, rinse the conductivity cell
thoroughly with distilled water, careful-
ly shake it dry, and dip it in the rinse
solution three times.
3.5.1.5 Calculate an Average Value
and the Standard Deviation — Use the
following relationships.
and
where
3-1
3-2
(in
x, = the measured value
umho/cm or pH units),
x = the average value,
s = standard deviation, and
n = the number of values.
3.5.1.6 Record the Results— Record
on the Conductivity Meter/Cell Ac-
ceptance Test form and the Conductivity
Acceptance Test Summary form. The
conductivity meter and cell are accept-
able if the average value is within 2% of
the theoretical value of 44.6 umho/cm
(25°C) and if the standard deviation is
less than 2%.
3.5.2 Evaluation of pH Meters
3.5.2. 1 Calibrate the Field pH Meter-
Calibrate as indicated by the manufac-
turer or as described in Section 4.3. A
laboratory pH electrode of documented
performance should be used.
5.5.2.2 Fill 11 Vials or Plastic
(17x100mm) Tubes—Fill to a depth of 3
cm with the pH electrode reference
solution (Section 3.3). The first tube is to
be used as a rinse tube.
3.5.2.3 Measure the pH of the 10
Solutions—Between each measure-
ment, 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.
3.5.2.4 Calculate an Average Value
and the Standard Deviation—see Sec-
tion 3.5.1.5.
3.5.2.5 Record the Results—Record
on the pH Meter/Electrode Acceptance
Test form and the pH Acceptance Test
Summary form. The pH meter is accept-
able if the average pH and the standard
deviation are within 0.03 unit of the
documented values for the electrode
used.
3.5.3 Evaluation of pH Electrodes
3.5.3.1 Assign Each New pH Elec-
trode an Identification Number—Allow
it to equilibrate overnight in the storage
solution recommended by the manu-
facturer.
3.5.3.2 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.
3.5.3.3 Calibrate the Laboratory pH
Meter—Calibrate as indicated by the
manufacturer or as described in Section
4.2.
3.5.3.4 Measure the pH of 10 Tubes-
Measure pH reference solutions as
described in Sections 3.5.2.2 and
3.5.2.3.
3.5.3.5 Calculate an Average Value
and the Standard Deviation— (Section
3.5.1.5).
3.5.3.6 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 average historical
value and if the standard deviation is
less than 0.03 pH unit.
3.5.4 Evaluation of Field Balance and
Thermometers
Reference weights traceable to NBS
are necessary for daily balance calibra-
tion. Each laboratory should purchase a
set of NBS-certified weights to be used
to certify a set of working weights used
daily in the field and laboratory. The
procedure used to certify weights is as
follows:
1. zero the balance .according to
manufacturer's recommendations,
2. weigh the working certified 1.0
and 5.0 Kg weights,
3. weigh reference 1 0 and 5 0 Kg
weights,
4. repeat this procedure five times,
and
5. complete the Certification of Work-
ing Weights to NBS form.
Working reference weights should be
certified by this procedure once every 6
mo; then the reference weights are
used for daily balance calibration, and
the NBS-certified weights are kept as
primary standards.
Each laboratory should have an NBS-
calibraied thermometer. One ther-
mometer in the laboratory should be
certified against the NBS-calibrated
standard. Keep the NBS-calibrated
thermometer as a primary standard.
Assign all laboratory and field ther-
mometers (or temperature probes)
identification numbers, and then cali-
brate 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. Generally,
calibration at two temperatures, nearO0
and 25°C, is sufficient and linear
temperature behavior may be assumed.
3.6 Report Forms
Blank data forms are included in this
section for the convenience of the
manual user. Uses of these are dis-
cussed throughout Section 3. To relate
the forms to the test being documented,
a number is given in the lower righthand
corner of each form (e.g., 1.1/3.4
indicates form 1, version 1, as discussed
in Section 3.4). The forms included are
listed below:
Form
Title
1.1/3.2 Field Conductivity Standard
Form
2.1/3.3 Field pH Electrode Test Solu-
tion
3.1/3.4 Field Quality Control Audit
4.1/3.5.1 Conductivity Meter/Cell
Acceptance Test Form
5.1/3.5.1 Conductivity Acceptance Test
Summary Form
6.1/3.5.2 pH Meter/Electrode Accept-
ance Test Form
7.1/3.5.2 pH Acceptance Test Sum-
mary Form
8.1/3.5.4 Certification of Working
Weights to NBS Form
-------
Jan. 1981 3 Part Il-Section 3.0
Field Conductivity Standard Form
Date of Preparation of
0.1M KCI Stock Solution.
Date of Preparation of
Dilute Field Standard:
(Analyst Signature)
Laboratory Analysis Before Shipment to the Field
1.
2
3.
Average Cond.
- Std. Dev.
Laboratory Values After Use In The Field:
Date of Lab Lab Value
Field Site = Analysis (umhos/cm)
OM Manual for Precipitation Measurement 1.1/3.2
-------
Part Il-Section 3.0 4 Jan. 1981
Field pH Electrode Test Solution
Date of Preparation of
Stock Solution
(Analyst Signature)
Volume of Stock Taken
Final Dilution Volume of
Field pH Electrode Test Solution
Laboratory Analysis Before Shipment to the Field
Conductivity pH Aye ± Sfd Dev
1.
2.
5
Laboratory Analysis of Aliquots Returned from the Field
Field pH Values Lab Values After Return
Site # Ave. + Std. Dev. (date) Date of Analysis pH Cond. Anal. Init.
OM Manual for Precipitation Measurement 2.1/3.3
-------
Jan. 1981
Part Il-Section 3.0
Field Quality Control Audit Sample Report
Sample #
Date of Preparation of Field Audit Sample:.
Vol. of (NHJzSO* Stock: ml:
Vol. of NHtHSO* Stock; ml;
Final Dilution Volume of
Field Audit Samples: _
.ml;
Date of Stock Preparation.
Date of Stock Preparation.
(Analyst Signature)
Laboratory Analysis Before Shipment3
To the Field
Laboratory Analysis After Return*
From the Field
1
2.
3.
Conductivity
pH
A verage
+ Std. Dev.
1.
2
3.
Conductivity
A verage
± Std. Dev
pH
Site #
Laboratory Analysis of Aliquots Returned From the Field
Date
Of Lab Conductivity pH
Analysis Value Value
Analyst
Initials
a 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.
OM Manual for Precipitation Measurement
3.1/3.4
-------
Part Il-Section 3.0
Jan. 1981
Conductivity Meter/Cell Acceptance Test Form
Date of Test:
Preparation Date of
A 0 0003 M KCI Reference Solution:
Meter Type/Serial No.
Conductivity Cell Type/Serial No
Indicate Whether Test of Meter or .
Cell
(Analyst Signature}
Conductivity Values Obtained:
Aliquot 1.
Aliquot 2.
Aliquot 3:
Aliquot 4.
Aliquot 5.
Aliquot 6:
Aliquot 7.
Aliquot 9.
10:
Test Solution
A verage conductivity
± standard deviation:
Accepted.
Rejected _
OM Manual for Precipitation Measurement
4.1/3.5.1
-------
Jan. 1981
Part Il-Section 3.0
Conductivity Acceptance Test Summary Form
Meter Type/
Serial #
Cell Type/Serial #
Date of Ftef
Soln. Prep
Date of
Check
Conductivity Value
Average ± Standard
Deviation (umho/cmj
Number
of
Values
Analyst
Initials
OM Manual for Precipitation Measurement
5.1/3.5.1
-------
Part Il-Section 3.0
Jan. 1981
PH Meter/Electrode Acceptance Test Form
Date of Test.
(Analyst Signature]
Preparation Date of
pH Electrode Reference Solution:
Meter Type/Serial No
pH Electrode Type/Serial No. /_
Indicate whether test of meter or
pH Values Obtained.
. electrode
4 0 (3.0) Buffer before:
7.0 (6.0) Buffer before:
Aliquot 1'
Aliquot 2
A liquot 3.
Aliquot 4
Aliquot 5:
Aliquot 6'
Aliquot 7:
Aliquot 8
Aliquot 9:
Aliquot 10
40 (3 O) Buffer after:
7 0 (6.01 Buffer after
pH Electrode Reference Solution:
Average pH ± standard deviation:
Accepted
Rejected.
OM Manual for Precipitation Measurement
6.1/3.5.2
-------
Jan. 1981
Part Il-Section 3.0
pH Acceptance Test Summary Form
Meter Type/
Serial tt
Electrode Type/
Serial #
Date of Ref
Soln. Prep.
Date of
Check
pH Value
Average + Standard
Deviation
Number
of
Values
Analyst
Initials
OM Manual for Precipitation Measurement
7.1/3.5.2
-------
Part 11-Section 3.0
10
Jan. 1981
Certification of Working Weights to NBS Form
Date of Certification.
Weight Set Serial ft:
Summary:
Balance 0
NBS 1kg
NBS 5kg
Test 1 kg
Test 5kg
A verage ± Standard Deviation
(Analyst Signature)
Balance 0
MRS 1kg
NBS 5kg
fpcf 1kg
Tp.cf 5kg
Balance ft
/VPS 1kg
NBS 5kg
Tfs( 1 kg
Balance 0
NBS 1kg
MRS 5kg
Tptt 1kg
7>>.«f f>kg
Balance n
NBS 5kg
Test 1kg
Tf?St ^kg
Balance ft
NBS 5kg
Test 1 kg
Test 5kg
Balance 0
NBS 1kg
NRS 5kg
Test 1kg
Test 5kg
OM Manual for Precipitation Measurement
8.1/3.5.4
-------
Jan. 1981
Part Il-Section 4.0
4.0 Laboratory Procedures
The laboratory procedures herein are
for chemical measurements and anal-
yses of precipitation samples. All forms
referred to are in Section 4.10.
4.1 Gravimetric Measure-
ments
In both the field and the laboratory the
volume of rainwater is determined by
measuring the mass of the ram and
multiplying the mass by 1 g/cm3 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 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; in either, the sample is
reweighed in the laboratory.
4.1.1 Apparatus
The balance should have a capacity of
20 kg and a precision of at least ±10 g.
4.1.2 Calibration
Calibrate the balance monthly, using
weights traceable to NBS-certified
weights. Store the NBS-certified
weights (primary references) in the
laboratory, certify the working calibra-
tion weights against these, and
complete the Certification of Working
Weights to NBS form. Recertify all
working calibration weights against the
NBS-certified weights every 6 mo. The
procedure for weight certification is the
same as for field balances (Section
3.5.4). 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 NHs contamination.
1. Be sure that the balance is level,
and then -adjust its zero knob so
that the balance zeroes (see
manufacturer's instructions).
2. If a bucket containing a sample is
being weighed, tap the lid to knock
off any water drops from its inside
surface into the bucket, and then
remove the tape and lid from the
bucket.
3. Place the bucket without its lid or
the plastic bottle on the balance
pan, and weigh it to the nearest
gram.
4. Record the weight on the bucket or
bottle label on the Field Data form.
5. 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 (2,3)
4.2.1 Scope and Application
This procedure is applicable to
rainwater and to drinking, surface, and
saline waters. The pH of a sample is
determined electrometrically, using a
glass electrode in combination with a
reference electrode. The glass and
reference electrodes may be separate or
combined in one unit. The measure-
ment should be performed on site
immediately after sample collection.
Upon arrival in the laboratory, the
sample should be measured for pH to
check the field measurement and to
indicate if the sample degraded. The pH
of rainwater is usually in the 6 to 4 pH
range.
Temperature effects on electrometric
pH measurements can be caused by two
changes. The first is the change in
electrode output at various tempera-
tures; this interference can be
controlled by using instruments that
have temperature compensation or by
calibrating the electrode instrument
system at the sample temperature. The
second is the change of pH inherent in
the sample at various temperatures;
this error is sample dependent, and
cannot be controlled so it should be
noted by reporting both the pH and the
temperature at the time of analysis.
4.2.2 Apparatus
A wide variety of pH meters are
commercially available with different
specifications and optional equipment.
A meter should have both an intercept
and a slope adjustment and should be
capable of measuring to ±0.01 pH unit.
Older pH meters used to becalibrated by
using a single buffer to electronically set
the intercept; this single point calibra-
tion assumed that the correct theoret-
ical slope was achieved by the elec-
trode, but this is a poor assumption
because commercial electrodes have
slopes which vary from 95% to 102% of
the theoretical slope(4). Modern pH
meters allow calibration at two points:
one to establish the intercept and the
other to establish the slope. When the
pH meter is not in use, leave it in the
standby mode. Always allow the meter
to warm up at least 30 min. before
making measurements if it has been
disconnected.
A glass electrode and a reference
electrode must be used with the pH
meter. The reference electrode can be a
calomel, silver-silver chloride or other
reference electrode of constant poten-
tial. Combination electrodes with both
measuring and reference functions can
be used. All electrodes should be
checked for acceptable performance
(Section 3.5.3). Store the pH electrodes
in deionized water, 74 umho/cm KCI
standard or 10~4N acid(H2SC>4)solution.
Prior to use, rinse the pH electrode
thoroughly with deionized water.
A thermometer or a temperature
sensor must be used to measure the
sample and buffer solution tempera-
ture.
4.2.3 Reagents
Use primary standard buffer salts,
available from the Natural Bureau of
Standards (NBS), when extreme ac-
curacy is necessary. Prepare reference
solutions from these salts, using special
precautions and handling (5) —that is,
low-conductivity dilution water, drying
ovens, and COa-free gas for purging.
Replace these solutions at least once
each month.
Prepare secondary standard buffers
from NBS salts or use pH buffers
purchased from commercial vendors.
Use of commercially available solutions
already validated by comparisons with
NBS standards is recommended for
routine use.
4.2.4 Calibration
Because of the variety of pH meters
and accessories, detailed operating
procedures cannot be incorporated into
this manual; thus each analyst must
thoroughly understand the operation of
his instrument and all functions.
Calibrate each instrument/electrode
system at a minimum of two points that
bracket the expected pH of the samples
and that are 3 or more pH units apart.
Most rainwater samples have pH's in
the range of 3.0 to 6.0 or 4.0 to 7.0,
depending on the area sampled; there-
fore, either pH 3.0 and 6.0 or pH 4.0 and
7.0 buffer solutions are suggested.
Calibrate the pH meter before and after
each set of (at most 20) samples.
Before calibration, top off the elec-
trode, if it is not sealed, with filling solu-
-------
Part Il-Section 4.0
Jan. 1981
tion (available from the manufacturer),
and then rinse it carefully with deion-
ized water. Measure both the standard
buffers and the samples at the same
temperature.
Various pH meters have a "balance"
or "standardize" dial and a slope or
temperature adjustment (outlined in the
manufacturer's instructions) Use the
pH 6.0 or 7.0 buffer to set the intercept
of the pH response with the standardiza-
tion knob, and then use the 3.0 to 4.0
buffer to adjust the slope control of the
pH response or the temperature func-
tion control. Then recheck the mea-
sured pH of the other buffer until both
calibration standards are within ±0.02
unit of the buffer value.
After each calibration, rinse the
electrode thoroughly with deionized
water, and then measure the pH of the
deionized water in several test tubes
until a stable value is obtained. This
procedure minimizes the chances of
buffer contamination of any sample.
4.2.5 Procedure
1. Calibrate the meter and the elec-
trode system (Section 4.2.4).
2. Place a sufficient volume of the
sample in a clean vessel to cover
the sensing elements of the elec-
trodes.
3. If the sample and the buffer solu-
tion temperatures differ by more
than 2°C, correct the measured pH
values. Some pH meters have
automatic or manual compen-
sators that electronically adjustfor
differences (refer to manufac-
turer's instructions).
4. Rinse and gently shake drops off
the electrodes, immerse them in
the sample vessel and swirl the
sample gently for a few seconds
after immersion to ensure even
contact between the sample and
the electrode sensing elements.
5. Allow the electrodes to equilibrate.
Do not disturb the air-water inter-
face or the solution while the
measurement is being made. (If
the sample is not in equilibrium
with the atmosphere, pH values
can change as the dissolved gases
are either absorbed or desorbed.)
Record the pH and temperature of
the sample.
4.2.6 Calculations
Read the pH meters directly in pH
units, and record on the data form pH to
the nearest 0.01 unit, and the temper-
ature to the nearest degree (°C).
4.2.7 Quality Control
Immediately after setup, measure the
pH of the electrode reference solution
(Section 3.3). Plot and evaluate the
measured pH on the real-time plots (QA
manual. Section 7.6.5.1) (6).
Check the calibration (Section 4.2.4)
after 20 samples; if the calibration has
drifted more than ±0.02, stop the
analysis, readjust and check the meter
and the electrodes.
Forty-four analysts in 20 laboratories
analyzed 6 synthetic water samples
containing exact increments of H+ and
OH" ions with the following results for
precision and accuracy.
In a single laboratory (EMSL Cincin-
nati, OH) using surface water samples
with an average pH of 7.7, the standard
deviation was ±0 1.
4.3 Conductance Measure-
ment
4.3.1 Conductance Using One Stan-
dard (2)
4.3.1.1 Scope and Application—This
procedure is applicable to rainwater,
drinking, surface, and saline waters,
and to domestic and industrial wastes.
The specific conductance of a sample is
measured by a self-contained conduc-
tivity meter, Wheatstone bridge type, or
equivalent. Samples are analyzed
preferably at 25°C; if not, temperature
corrections are made, and results
reported at 25°C The instrument must
be standardized with KCI solution
before daily use; the conductivity cell
must be kept clean; and temperature
variations and corrections represent a
large source of potential error.
Analyses can be performed either in
the field or the laboratory. If the analysis
is not completed within 24 h of sample
collection, filter the sample through a
0.45-u filter and store it at 4°C. Wash
the filter and apparatus with high
quality deionized water, and prerinse
with sample before use.
4.3.1.2 Apparatus—The following
are needed for this procedure:
1. Conductivity bridge, range 1 to
1000 umho/cm.
2. Conductivity cell, cell constant 1.0
or dipping-type microcell with 1.0
constant, YSI #3403 or equivalent.
3. Thermometer.
Precision as
Standard Accuracy
Deviation as Bias
pH Units pH Units % pH Units
3.5
3.5
7.1
7.2
8.0
8.0
0.10
0.11
0.20
0.18
0.13
0.12
-0.29
-0.00
+1.01
-0.03
-0.12
+0.16
-O.01
+0.07
-0.002
0.01
+0.01
4.3.1.3 Reagents—
1. Stock solution, 0.1 OM KCI: Dis-
solve 7.456 g of predried (2 h al
105°C) KCI in deionized water, anc
dilute to 1 liter at 25°C.
2. Standard solution 00005M KCI:
Dilute 5 ml of the 0.10M KCI to
1000 ml with deionized water.
3. Spike solution, 0.0005M KCI Pre-
pare as in Step 2, but use an inde-
pendently prepared stock KCI solu-
tion.
4.3.1.4 Calibrations—The analyst
should use the standard KCI solution
and the table below to check the
accuracy of the cell constant.
4.3.1.5 Procedure—Follow the manu-
facturer's directions for operation of the
instrument.
1 Allow samples to come to room
temperature (23° to 27°C) if
possible.
2. Measure the temperature of sam-
ples. If the temperature is not
25°C, correct the reading to 25°C
(Section 4.3.1.6).
4.3.1 6 Calculations—Most of the
instruments are equipped to give the
conductivity of a solution as a direct
readout.
These temperature corrections are
based on standard KCI solution, and are
used with instruments having no
automatic temperature compensation
1. Add 2% of the reading per degree
below 25°C or
Conductivity of 0.0005M KCI
°C umhos/cm
20
21
22
23
24
25
26
27
28
66.8
68.2
69.5
71.0
72.4
73.9
75.4
76.9
78.4
(FWPCA Method Study 1. Mineral and
Physical Analyses.)
2. Subtract 2% of the reading per
degree above 25°C.
3. Report specific conductance, in
umho/cm at 25°C.
4.3.1.7 Quality Control—Measure
analyst spikes daily. Plot and evaluate
the values on real-time plots (QA
manual, Section 7.6.5.1). Forty-one
analysts in 17 laboratories analyzed six
synthetic water samples containing
increments of inorganic salts, with the
following results for precision and
accuracy.
In a single laboratory (EMSL, Cincinnati,
OH) using surface water samples with
-------
Jan. 1981
Part Il-Section 4.0
an average conductivity of 536 umhos/
cm at 25°C, the standard deviation was
found to be ±6.
4.3.2 Conductance Using Multiple
Standards (Alternate Procedure)
4.3.2.1 Scope and Application—This
procedure is applicable to rainwater and
to drinking and surface waters. The
conductivities of calibration standards
and samples are measured by a self-
contained conductivity meter, Wheat-
stone bridge type or equivalent. The
specific conductivities of the samples
are obtained by comparing the mea-
sured values with a plot of the measured
conductivities versus the specific
conductances of the standards. The
instrument must be standardized with
KCI solutions before daily use and the
conductivity cell must be kept clean.
Temperature variations and corrections
represent a large source of potential
error; therefore, the calibration stand-
ards and the samples should be mea-
sured at the same temperature to
minimize error For sample handling
and preservation, refer to Section
4.3.1.1, and for apparatus refer to Sec-
tion 4.3.1.2.
Specific
Conductance.
umho/cm
TOO
106
808
848
1640
1710
Standard
Deviation
in umho/cm
755
8 14
66 1
796
106
119
Accuracy as
Bias.
% umhos/cm
-2 02 -2 0
-0 76 -0 8
-3 63 -29 3
-4 54 -38 5
-5 36 -87 9
-5 08 86 9
/FWPCA Method Study 1, Mineral and Physical
Analyses)
4.3.2.2 Reagents—
1. Stock solution, 0.1 OM KCI—Dis-
solve 0.7456 g of predned (2 h at
105°C) KCI in deiomzed water, and
dilute to 100 ml at 25°C with
deionized water.
2. Stock solution, 0.01 M KCI—Dilute
10mlof0.10MKCIstockto100ml
with deionized water.
3. Spike solution, 0.0005M KCI—
Dilute 5 ml of a stock KCI solution
(0.1 OM) prepared independently
from the stock used for the calibra-
tion standards to 1 liter with
deionized water.
4.3.2.3 Calibration—Laboratory cali-
bration for conductivity measurements
is multipoint. Prepare stock 0.1M and
0.01 M KCI solutions as needed. Each
day of analysis, prepare a deionized
water blank and 0.0001 M, 0.0005M,
and 0.001 M KCI solutions from the
0.01M stock by serial dilution with
deionized water.
Calibrate the conductivity meter
using KCI solutions of known con-
ductivity (Table 4-1).
Table 4-1. Specific Conductances
of KCI Solutions
Specific Conductance
Concentration. M umho/cm @ 25°C
00001
00005
0.001
1494
7390
14700
Directly before and after analysis,
analyze the calibration standards along
with the deionized water used to dilute
the standards. Subtract the response of
the deionized water from the response
of the calibration standards analyzed at
the same time. Calculate a linear least
squares fit of the specific conductances
vs. the measured conductivities, or plot
a graph of the same parameters for all
calibration points measured before and
after the samples.
4.3.2.4 Procedure—Follow the manu-
facturer's directions for the operation of
the instrument.
1. Allow samples and standards to
come to room temperature if
possible; all must be at the same
temperature.
2. Read the measured conductivities
for all standards and samples;
thoroughly rinse the cell with
deionized water between samples
and shake off the excess water.
4.3.2.5 Calculations—Calculate con-
ductivities using a graph or using a
linear least squares fit.
1. Graph - Read the sample conduc-
tivity directly from the plot of
specific conductance vs. measured
conductivity.
2. Linear least squares fit - Use equa-
tions available in most elementary
statistics books(9) to obtain a linear
least squares fit. The calculation
yields the following parameters.
slope (m), intercept (b), error of fit
(e), and correlation coefficient (r).
For standard i, the slope and
intercept define a relationship
between the specific conductance
(x,) of the standards and the instru-
ment response (y,)
Y, - mx, + b.
4-1
Equation 4-1 is the preferred fit
where major components of ran-
dom variance are assumed to be
in instrument response. Rear-
rangement of the equation yields
the specific conductance (x,) cor-
responding to an instrumental
response (Y,) of a sample j,
4.3.2.6 Quality Control—Measure
analyst spikes daily. Plot and evaluate
the value obtained on the real-time plots
described in the QA Manual, Section
7.6 5 1(6). No precision or accuracy data
are available
4.4 Sample Filtration
After measuring the pH and conduc-
tivity, but before measuring 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
recommended in a recent study(1). The
recommended filter material is a 0.45-
um membrane filter (Millipore HA); the
filter funnel should be plastic. Before
each filtration, thoroughly rinse the
apparatus, including the filter, with
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 labora-
tory air If the filtered particulates are to
be analyzed, they should be washed
with deionized water, placed in a
labeled petri dish, allowed to dry in a
desiccator, and stored until they are
extracted.
.4.5 Acidity Measurements
Acid precipitation analyses should
distinguish between acidity due to
strong and to weak acids. Strong acid is
measured by the Gran titration (7,8);
total acidity is measured by titrating to
pH 8.3 to include other acids (except
NH«+); and weak acid is calculated by
subtracting the strong acid value from
the total acid value.
4.5.1 Strong A cid
4.5.1.1 Scope and Application—The
Gran titration is applicable to rainwater,
surface, and other waters and gives a
measure of the concentration of strong
acid. The detection limit of this method
depends on the volume of sample
titrated. (When a 3 ml sample is titrated
with 0.01 N NaOH, the detection limit is
0.005 A/g/ml.)
The Gran titration is based on the
Nernstian response of the pH electrode.
For the Gran plot (7,8), a function of the
potential
-------
Part Il-Section 4.0
Jan. 1981
K=1/CB10E°/S 4-4
where Cb = concentration of base,
E0 = standard potential, and
S = electrode slope.
The K value is determined, not calcu-
lated, and the ip function is given by
Equation 4-5.
= (Vo + V) 10 E s
4-5
where V0 = initial volume of solution in
titration vessel,
E = potential measured after
volume V of base added, and
S = electrode slope defined by
Equation 4-6.
S = 2.303RT/F=0.1984T(°K)mv
= 59.16 mvat 25°C
4-6
where R = gas constant,
T = absolute temperature (°C +
273°), and
F = Faraday constant.
Experimentally, one titrates and
calculates the (fi function for each V and
E value; calculates the linear least
squares fit of V and tp for the initial
values of the titration; obtains the
intercept Ve; and calculates the amount
of strong acid in the titration vessel by
multiplying Ve by the concentration of
the base added.
4.5.1.2 Apparatus—The following
are needed for the Gran titration.
1. pH meter which measures to ±0.1
mv.
2. Microelectrode which passes the
electrode test (Section 3.5.3).
3. Thermostatted ~15-ml titration
vessel.
4. Small magnetic stirring bar and
variable speed magnetic stirrer.
5. Micropipette capable of reproduci-
bly delivering 5 ul of solution.
6. A thermometer good to ±0.1 °C,
and a titration vessel constructed
so that the thermometer can be
submerged in each sample.
7. A source of N2 gas to purge C02
from the airspace above the sam-
ple.
4.5.1.3 Reagents—
1. Deionized Water - Prepare all
reagent dilutions with C02-free
water.
2. Conditioning Solution - Prepare
5.0 x 10"5N H2S04 fresh daily from
0.1 N H2S(X,. Dilute 10-ml of 0.1 N
H2SO< (a commercial DILUT-IT) to
100 ml with deionized water, and
dilute 10 ml of the resulting solu-
tion to 2000 ml with deionized
water.
3. Titrant - PrepareO.01 N NaOHdaily
from 1.0N NaOH. Dilute 10 ml of
1 .ON NaOH (a commercial DILUT-
IT) to 1 liter with deionized water
(0.01N NaOH). Standardize the
1 .ON NaOH once a month against
potassium acid phthalate
4. Spike Solution - Prepare 10 x
10~5N and 3.0 x 10~5N H2SO4 solu-
tions as "analyst spikes" daily
from a separate 0 1N stock solu-
tion.
4.5.1.4 Calibrations—Calibrate the
micropipette once a week and each time
a new glass tip is placed on the
micropipette. Calibrate the micropipette
gravimetrically or with the procedure
presented below.
Add about 2.5 ml of deionized water in
five 5-ml volumetric flasks. Use the
micropipette to deliver 5 ul of a 1000
ug/ml or 1 ug/ul Cu atomic absorption
standard to each flask. Dilute to volume
and analyze for Cu using flame atomic
absorption spectrophotometry. Calcu-
late the micrograms per milliliter found
from the calibration standards made by
serial dilution of the 1000 ug/ml Cu
using Class A volumetric pipettes. (The
background matrix for the calibration
standards should be the same as for the
calibration samples plus deionized
water.) Calculate the micropipette
calibration:
Micrograms Cu delivered = micrograms
per milliliter found x 5.0 ml.
Since the spiking solution has a
concentration of 1 ug/ul, the number of
micrograms is equal to the number of
microliters. Obtain an average value
and standard deviation of microliters
delivered; the standard deviation should
be less than 2%.
4.5.1.5 Procedure—
1. Condition the Titration Cell -
Pipette 7.0 ml of 5x10"5N H2SO4
into the titration cell; allow time for
thermal equilibration at 25°C;
measure the potential of the solu-
tion; discard the solution; rinse the
cell with conditioning solution; and
pipette another 7.0-ml aliquot of
conditioning solution into it. Re-
peat this procedure until the
potential is within 1 mv of the
previous measurement to condi-
tion the titration cell for microde-
termination of strong acid.
2. Measure Acid in the Conditioning
Solution - Pipette 7.0 ml of 5x10"5N
H j.804 into the conditioned titration
cell; start the stirrer; add 3 ml of
deionized water; allow time for
thermal equilibrium (25°C); titrate
the conditioning solution by adding
5-ul increments of 0.01 N NaOH
with a calibrated micropipette; and
record, after each addition, the
resulting potential. Continue the
titration until the change in the
potential from that originally
measured is 15 to 20 mv Data
points taken for the Gran plot are
the initial potential and all points
up to the 20-mv change; there
should be at least five points.
3. Measure Strong Acid in Samples -
Rinse the cell with conditioning
solution; pipette 7 0 ml of condi-
tioning solution into the condi-
tioned titration cell; monitor the
initial millivolt reading, and if it is
not within ±1.2 mv of the original
conditioning solution readings,
rerinse the cell and repeat until the
voltage is, then pipette 3.0 ml of
sample into the conditioning solu-
tion, and titrate as indicated in step
2 above.
4.5.1.6 Calculations—Calculation of
the concentration of strong acid de-
pends on the linear relationship be-
tween the volume of base added and the
Gran function tfi (Equation 4-3). A
sample data form and calculations are
in Figure 4-1. The data form gives the
potential readings as a function of the
volume of base added. The ifj function is
calculated for the potential at each
titration point using Equation 4-5. To
determine the equivalence point vol-
ume of base (Ve), a linear least squares
fit of the volume of base added (V)
versus function tfj is calculated. The
intercept is the equivalence point
volume of base. The tfj function can be
plotted versus the volume of base added
(V), and the intercept found graphically
(Figure 4-2). Blank data forms are in
Section 4.10. To obtain the acid content
of the sample, subtract the value of the
conditioning acid solution equivalents
from those obtained from the sample
titration.
4.5.1.7 Quality Control—Daily after
setup, analyze three conditioning
solutions and one spike (Section 4.5.1.5).
1. Plot the magnitude of V for each
conditioning solution; determine
the percent recovery for the spike
and add it to the real-time plots
described in the QA Manual (Sec-
tion 7.6.5.1).
2. If calculations are performed
mathematically, evaluate the cor-
relation coefficients; if less than
0.9990, troubleshooting should be
performed and the conditioning
solution and/or sample reana-
lyzed.
3. Monitor the initial conditioning
solution potential for each sample;
if not within ±1.2 mv of the
potential for the conditioning
solution at the beginning of the
run, empty and refill the titration
vessel with conditioning solution.
-------
Jan. 1981
Part Il-Section 4.0
Sample #
Initial mv
Temp. °C
CStt3
155.8mv
220°
Hi NaOH
Injected
0
5
10
15
20
35
50
65
30
mv
reading
146.6
1428
138.4
133.2
1266
79.1
-106.8
-133.4
-1460
Sample #
Initial mv
Temp °C
fjl NaOH
Injected
mv
reading
Date.
rnnf Mann u
Micropipette Calibration
fi /it - fit
Tntal \jnl(imR
Calculations
S = 0.1984 (T) = 0.1984 mv(273.15°C + 22.0°C) = 58.56 mv
Strong Acid
ffjl injected)
0
5
10
15
20
10
E/S
10.000
10.005
10.010
10.015
10.020
Linear least squares fit of V vs ifi:
1466
142.8
138.4
133.2
126.6
Slope
Intercept fVe)
Correlation Coef.
31872
274.49
230.88
188.19
145.17
3187.2
2746.3
2311.1
1884.7
14546
-0.01156
3677
1 0000
Amount of strong acid in titration vessel:
*) = 36. 77 ulx0.01N = 0.368 ^equivalent
Figure 4- 1 . Gran strong acid data form and calculations
§
5
CD
10 15 20 25
V 0.01 N NaOH added (fj/J
30
Ve
36.8
Figure 4-2. Plot of the Gran function (1)1) versus the volume of titrant added (V) in the
determination of strong acid. *
4. Analyze an additional conditioning
solution and spike at the end of a
day's analyses
b. Precision and accuracy data are
not available.
4.5.2 Acidity
4.5.2.1 Scope and Application—This
procedure is applicable to rainwater,
surface, and other waters with pH less
than 8.3; is a measure of the concentra-
tion of strong and weak acids that react
with OH~ ions (including dissolved
gases); and has a range dependent on
the volume of sample titrated and on the
precision with which the increments of
titrant can be measured
Titrate samples with 0.02N C03-free
NaOH solution (if 10 ml of sample is
available for analysis, use a 50-ul
syringe for dispensing the titrant to
achieve a precision better than 10
ueq/L), measure the endpoint with a pH
meter; and report results as ueq/L
The sample container must be filled
completely, sealed, and stored at 4°C.
Care must be taken to minimize expo-
sure of the sample to the atmosphere.
Open the sample container immediately
before analysis Analysis should be
performed as soon as possible after
collection Samples with an initial pH
between 4 3 and 8.3 are subjectto error
due to the loss or gain of dissolved gases
during sampling, storage, and analysis.
4.5.2.2 Apparatus—
1. pH meter and electrode (Section
4.3).
2 Microburette or microsyringe.
3 Teflon or glass magnetic stirring
bar.
4. Magnetic stirrer.
5 Beakers or flasks.
4.5.2.3 Reagents—
1. Standard solution, 1 ON NaOH—
Dissolve 40 g NaOH in 250 ml
deionized water. Cool and dilute to
1 liter with CO2-free distilled
water. Store in a polyolefin bottle
fitted with a soda lime tube or a
tight cap to protect from atmos-
pheric COa.
2. Standard titrant, 0.02N NaOH—
Dilute 20.0 ml of 1.0N NaOH with
CO2-free deionized water to 1 liter.
Store in rubber stoppered bottle.
Protect from atmospheric CO2 by
using a soda lime tube. Stan-
dardize against a 0.02N potassium
acid phthalate solution prepared
by dissolving 4.085g of anhydrous
KHC8H4O4 in C02-free distilled
water and diluting to 1 liter.
3. Spike Solutions, 1.0x10"5N and
3.0 x 10~5 NH2S04—Prepare
"analyst spikes" daily from an
independent stock solution (4.5.1.3).
-------
Part Il-Section 4.0
Jan. 1981
4.5.2.4 Procedure—
1. Pipette an appropriate aliquot of
sample into a beaker or flask
containing a small Teflon or glass
stirring bar; use extreme care to
minimize the sample surface
disturbance.
2 Immerse pH electrode(s) into
sample; stir at a rate which does
not cause sample surface disturb-
ance
3. Titrate with 0.02N NaOH to pH 8.3
as quickly as possible to prevent
absorption of atmospheric CO2.
4. Record the volume of the titrant.
4.5.2.5 Calculation—Acidity is cal-
culated as follows:
ueq/L=mle x NB x lOVmls 4-7
where mle = Volume NaOH titrant, ml,
mls = Volume of sample, ml,
and
NB = normality of titrant
4.5.2.6 Quality Control—Analyst
spikes should be titrated daily. Plot and
evaluate the values on the real-time
plots (QA manual, Section 7.6.5.1)
Precision and accuracy data are not
available.
4.6 Ammonium Determina-
tions Using Automated Colori-
metry (10)
4.6.1 Scope and Application
This procedure covers determinations
of NH4+ in drinking, rain, and surface
waters in the range of 0.05 to 4.0 mg/l
NH/; this range is for photometric
measurements made at 630 to 660 nm
in a 15-mm or 50-mm tubular flow cell.
Higher concentrations can be deter-
mined by sample dilution. Approxi-
mately 20 to 60 sampies/h can be
analyzed.
Alkaline phenol and hypochlonte
react with NH4+toform indophenolblue,
the absorption of which is proportional
to the NH4+ concentration. The blue
color is intensified with sodium nitro-
prusside.
Beware of interferences during the
measurements. If Ca++ and Mg++ ions
are present in sufficient concentrations,
precipitation problems can occur during
analysis. A Na-K tartrate solution
prevents precipitation of the alkaline
earth hydroxides.
Sample turbidity and color may
interfere; before analysis, turbidity must
be removed by filtration. Sample color
that absorbs in the photometric range
should be avoided.
4.6.2 Apparatus
Techmcon AutoAnalyzer Unit (AAI or
AAII) consisting of.
1. Sampler,
2 Manifold (AAI) or analytical cart-
ridge (AAII),
3 Proportioning pump,
4. Heating bath with doubledelay coil
(AAI),
5 Colorimeter equipped with 15-mm
tubular flow cell and 630 to 660-
nm filters,
6. Recorder, and
7. Digital printer for AAII (optional)
4.6.3 Reagents
1. Deionized Water - All solutions
must be made using NH4+-free
water (Section 7.1.11 QA manual
(6))
2. Sulfuric Acid, 5N (air scrubber
solution) - Carefully add 139 ml of
cone H2SO4 to approximately 500
ml of deionized water; cool to room
temperature, and dilute 1-liter
with deionized water
3. Sodium Phenolate - Using a 1 -liter
Erlenmeyer flask, dissolve 83 g
phenol in 500 ml of deionized
water In small increments, cau-
tiously add with agitation, 32 g of
NaOH. Periodically, cool flask
under water faucet. When cool,
dilute to 1 liter with deionized
water.
4. Sodium Hypochlorite Solution -
Dilute 250 ml of a bleach solution
containing 5.25% NaOCI (e.g.,
Clorox) to 500 ml with deionized
water. Available chlorine level
should approximate 2% to 3%.
(Since Clorox is a proprietary
product, its formulation is subject
to change; the analyst must re-
main alert to detect significant
change.) Due to the instability of
this product, storage over an
extended period should be avoided.
5. Sodium Potassium Tartrate Solu-
tion NaKC4H4O6 • 4H2O - To 900ml
of deionized water add 100 g of
NaKC4H4O6 • 4H20. Add two pel-
lets of NaOH and a few boiling chips;
boil gently for 45 min. Cover, cool,
and dilute to 1 liter with NH/-free
deionized water Adjust pH to 5.2
+0.05 with H2SO4. After allowing
to settle overnight in a cool place,
filter to remove precipitate. Add
0.5 ml Brij 35 solution(11) (avail-
able from Techmcon Corp.), and
store in stoppered bottle.
6. Sodium Nitroprusside (0.05%) -
Dissolve 0.5 g of sodium nitroprus-
side in 1 liter of deionized water.
7. Stock Solution - Dissolve 0.2966 g
of anhydrous NH4CI, dried at
105°C, in deionized water, and
dilute to 1000 ml (100 mg/l NH/).
8. Standard Solution A - Dilute 10.0
ml of stock solution to 200 ml with
deionized water (5.0 mg/l NH4*).
9. Standard Solution B - Dilute 40.0
ml of stock solution to 100.0 ml
with deionized water (40 0 mg/l
NH4)
10. Working Standards - Using stan-
dard solutions A and B, prepare
fresh daily the following dilutions
in 100-ml volumetric flasks
Standard
A
8
C
D
E
F
G
H
NH/
mg/l
0.05
0 10
020
040
0.8
1.6
32
4.0
Milliliters of
Standard Solution
A or B per 100 ml
working standard
1 Oof A
2 Oof A
4.0 of A
8 Oof A
2.0 of B
4.0 of B
8 Oof B
1 0 0 of B
4.6.4 Procedure
For a working range of 0.05 to 4.0 mg
NH4+ (AAI), set up the manifold as shown
in Figure 4-3. For a working range of
0.05 to 1.6 mg NH4* (AAII), set up the
manifold as shown in Figure4-4. Higher
sample concentrations may be accom-
modated by dilution.
1. Allow both the colorimeter and the
recorder to warm up for 30 min;
obtain a stable baseline with all
reagents, while feeding deionized
water through the sample line.
2. For the AAI system, sample at a
rate of 20/h with the 1:1 cam; for
the AAII use a rate of 60/h with the
6:1 cam and a common wash.
3. Arrange MM** standards in the
sampler in order of decreasing
concentration; complete the load-
ing of the sampler tray with
unknown samples.
4. Switch the sample line from
deionized distilled water to the
sampler, and begin the analysis.
5. Analyze the full calibration curve
every 30 samples and between
curves analyze one standard every
10 samples; periodically, check the
baseline by turning off the sampler
for approximately l.min.
6. When an analytical run is com-
pleted, draw a baseline between
the baseline sections of the chart;
measure the peak heights of the
standards and samples; and record
them on the data form (Figure4-5).
4.6.4.1 Notes on Operation—For the
low-level iNH/ analysis, take extreme
care to avoid NH4+ contamination since
NH4+ is ubiquitous in the laboratory.
Rinse the glassware immediately before
-------
Jan. 1981
Part Il-Section 4.0
SM = Small Mixing Coil
Proportioning
Wash Water __
to Sampler
SM
0000
\
Heating f
Bath 37°C \
1
LM
00000000
i
— *
-^
W
.,..,, ,^
LM 00000000
SM 0000
> f
—Waste
r~
(S
W
"V
Colorimeter
J5 mm Flow Cell
650-660 nm Filter
P B
G G
R R
G G
W W
W W
R R
P P
ml/mm
2 9 Wash
2.0 Sample
0.8 Tart rate
2.0 Air*
0.6 Phenolate
0
Sampler
20/hr.
1:1
0 6 Hypochlorite
0.6 Nitroprusside
2.5
J Waste
Recorder
*Scrubbed Through
5 N HzSOt
Figure 4-3. Ammonia manifold AA I.
Proportioning
Pump
Sampler
60/hr.
6:1
Heating
Bath
SO°C
Wash Water
to Sampler
0000
.
§
Wash
^
•«.
Recorder
G
O
0
R
0
G
W
0
R
O
Black
O
0
Blue
C3
mi/mm
2.0 Wash
0.23 Air" °
0.42 Sample
0.8 Tartrate
0 42 Phenolate
0.32 Hypochlorite
0.42 Nitroprusside
1-6 !!/„„,„
Digital
Printer
Co/or/meter
50 mm Flow Cell
650-660 nm Filter
Figure 4-4. Ammonia manifold AA II.
use Do not wipe the pipettes before
delivering the solution. Prepare all
reagents with NH4*-free water. Wear
gloves when handling reagents or
samples. Connect the airlines which
introduce bubbles into the automated
flow system to a bubbler filled with 5N
HaSO4 so that NHs gas is scrubbed from
the air before the air is introduced to the
system.
After setup, be sure the bubble
patterns are evenly segmented, that all
"Scrubbed Through
connectors are properly butted, that no
leaks are present, and that aspiration of
the sample is adequate.
Before analysis, the recorder re-
sponse is set. The recorder scale is set at
50-mv full-scale response (20 mv if
expanded scale NH4+ is being run). The
recorder input should be shorted, and
the recorder baseline adjust should be
set at 0% of the chart; then the recorder
input should be connected to the
colorimeter. (White wire is positive,
black is negative, green is ground, and
red is reference voltage.)
The colorimeter display rotary switch
should be set to zero and the output on
the recorder should be set at 0% of the
chart by adjusting the zero screw on the
colorimeter operating panel; then the
colorimeter display rotary switch should
be set to FULL SCALE, and the output on
the recorder should be set to 100% of
chart by adjusting the FULL SCALE
screw on the colorimeter operation
panel; and finally, the colorimeter
display rotary screw should be placed in
the NORMAL position.
The baseline should be set after all
reagents have been flowing for at least
15 min; the aperture for the sample
beam (knob A) should be opened all the
way; and then the aperture for the
reference beam should be adjusted to
set the baseline on the recorder. The
aperture adjustment should be con-
sidered a coarse adjustment of the
baseline, and the fine adjustment is
done with the BASELINE knob on the
colorimeter aperture panel.
After the baseline is set, thefull-scale
response should be set by placing the
sample probe in a standard solution
giving a full-scale response. When a
response is obtained, the colorimeter
scale expansion is adjusted by turning
the STD CAL to bring the full-scale
response to the desired level on the strip
chart recorder. The STD CAL should be
set so that baseline noise of 1% of full
scale is obtained when deionized water
is analyzed.
A second recorder may be used to
monitor the lower half of the analytical
range. Although the nominal output is
50 mv, the system can be set up so that
1% baseline noise is obtained with 20
mv. For routine analysis, the output may
be monitored with one recorder set at
20 mv full scale and with another at 50
mvfull scale. Note the STD CAL number
on the strip chart and on the data form.
Release the tension on the pump
tubes after each day of analysis.
Replace pump tubes after 3 days of use.
4.6.4.2 Troubleshooting—Detailed
troubleshooting procedures are general-
ly given by the the instrument manufac-
turer. Presented briefly below are a
static (electronic) test and a dynamic
test. The static test is performed by
pumping water through the flow cell;
turning off the pump; artd adjusting the
baseline (using the A and B aperture
controls on the colorimeter) to bring the
recorder response to 50%. The dynamic
test is performed with the normal
system set up and flowing. Deionized
water is sampled; all other delivery
tubes are placed in the appropriate
reagents; and the baseline should be set
-------
Part fl-Section 4.0
Jan. 1981
DATA SHEET
It-tl'&O
_ f
XD
to ms/
B
V.3
9.3
6.0
39-
k
7_
8.
9_
/c.
LL
Stikt,
50. /
27.1
/Jfc
/I
/60Q
/bol
/40^
20
2i
11*0 (s
/0.3
ZC
Z7
11*01
^70. /
/uog
ar.-V
/(r//
JO
Ilili
IM3
23.0
*1
jri
3V
O.J
lifts'
37
IU&
/z./
Figure 4-5. Ammonium analysis - peak heights.
-------
Jan. 1981
Part Il-Section 4.0
at 50% of full scale (using the A and B
aperture controls on the colorimeter)
AS A FIRST STEP IN TROUBLESHOOT-
ING, CHECK ALL ELECTRICAL CON-
NECTIONS TO SEE THAT GOOD CON-
TACT IS MADE AND THAT NO SHORTS
EXIST. Below are four typical problems
and procedures for solving them
1. Baseline Drift
Static Test - a drift noted with the
static test may be due to inade-
quate warmup of electronic com-
ponents, defective colorimeter
lamp, defective photocell or a
defective recorder. Replace the
suspected defective components.
Check the photocells by switching
the reference and sample posi-
tions; if the baseline drift changes
direction, one photocell is bad. If
the baseline does not drift during
the static test, the problem is not
electronic; the problem is the
dynamics of the system.
Dynamic Test - If baseline drift is
noted with the dynamic test, check
the bubble flow pattern. An erratic
bubble flow pattern may be due to
a leak which usually can be
stopped by tightening connections
and making sure glass tubing
abuts together. Other causes of
drift include poorly mixed rea-
gents, bad pump tubes, and rea-
gent degradation; replace sus-
pected tubes and reagents.
2. Baseline Noise
Static Test - If baseline noise is
observed, the problem is electri-
cal. Check to see if the gain adjust-
ment (STD CAL) is too high; if it is,
the problem is sensitivity. If STD
CAL is normal, the problem may be
poor electrical connections, a
defective lamp, a defective photo-
cell, or a defective recorder; check
each of these, and replace the
defective components as neces-
sary.
Dynamic Test - Baseline noise in
the dynamic system may be due to
leaks observed as an erratic
bubble pattern. If so, proceed as
indicated above (baseline drift-
dynamic test). Other causes of
dynamic baseline noise include
wornout pump tubes, degassing of
reagents (bubbles in the flow),
poorly mixed reagents, improperly
pumped reagents, and reagents
containing particulates or air
bubbles in the flow cell. Replace
reagents and pump tubes as
necessary. If a bubble is suspected
in the flow cell, pinch its exit tube
for 5 s to dislodge the bubble.
3. Poor Sensitivity - If poor sensitivity
is observed, be sure that the gain
(STD CAL) is normal, the recorder
is set on 50-mv full-scale re-
sponse, and the color filters are
correct Replace the pump tubes if
they are old. Adjust the flow cell
alignment and the colorimeter
mirror assembly alignment (manu-
facturer's instructions), and opti-
cally peak the colorimeter. If the
problem is still not solved, it may
be due to improperly prepared
reagents or calibration standards,
prepare fresh solutions if they are
suspect.
4. Sensitivity Changes - Sensitivity
changes not accompanied by
increased baseline noise are
probably due to defective pump
tubes, degradation of reagents, or
recorder problems. Replace rea-
gents and pump tubes as neces-
sary. Check the recorder response
for stability, using a constant —40
mv source.
5. Nonlinear Calibration Curves - If
none of the problems indicated
above is noted, then nonlinear
calibration curves are probably
due to incorrectly prepared rea-
gents or standards or to nonlinear
recorder responses. Prepare fresh
reagents and standards as neces-
sary. Check recorder response
using a constant voltage source
which delivers 10, 20, 30, 40, and
50 mv.
4.6.5 Calculations
Using a Graph - Prepare a standard
curve by plotting peak heights of
processed standards against their
concentrations; compute the concen-
tration of samples by comparing sample
peak heights with those on the standard
curve.
Using Linear Least Squares Fit - Use
the equations for calculation of the
linear least squares fit thatare available
in most elementary statistics books
(e.g., see Reference 9). The linear least
squares equation gives the following
parameters: slope (m), intercept (b),
error of fit (e), and correlation coefficient
(r). The slope and intercept define a rela-
tionship between the concentration of
the standard (x,) and the predicted
instrument response (Y,).
Y, = mx, + b. 4-8
Equation 4-8 is preferred for fitting
where the major components of random
variance are assumed to be in instru-
ment response; a simple rearrangement
of Equation 4-8 yields concentration (x,)
corresponding to an instrument re-
sponse (Y,) of a sample j;
x, = (Y, - b)/m.
4-9
4.6.6 Quality Control
1. After every 30 samples analyze at
least five calibration curve stan-
dards. With every 10 samples,
repeat one standard. If responses
to these standards change more
than 5%, stop the analysis, and
determine the cause.
2. Every 30 samples, analyze an
analyst spike (Section 7.6.4.1 of
QA manual (6)), a duplicate, and an
old sample. Calculate the first
spike of the day from the first
calibration curve, and plot the
value on the real-time plot (Section
7 6.5.1, QA manual (6)) Compare
these data with previous data; if
the control limits are exceeded,
stop the analysis, and search for
the problem.
3 When preparing new stock solu-
tion (Section 4.6.3), prepare dilute
calibration standards from the old
and the new stocks. Analyze the
old and new standards; compare
the calibration curves; and com-
plete the Standard Preparation
form in Section 410.
4.7 Orthophosphate Deter-
minations Using Automated
Colorimetry (12)
4.7.1 Scope and Application
This procedure covers the determina-
tion of phosphate in drinking, rain, and
surface waters; it is based on reactions
specific for dissolved orthophosphate; it
is usable for 0.02 to 0.4 mg/l PO
-------
Part Il-Section 4.0
10
Jan. 1981
Wash all glassware with hot 11 HCI
and rinse with deionized water Never
use commercial detergents that contain
phosphate
4.7.3 Reagents
1. Sulfuric Acid Solution, 5N -Slowly
add 70 ml of cone H2S04 to
approximately 400 ml of deionized
water, cool to room temperature;
and dilute to 500 ml with deionized
water
2. Antimony Potassium Tartrate Solu-
tion - Weigh 0.3 g K(SbO)C4H4O6-l/2
H20; dissolve in 50 ml of deionized
water in a 100 ml volumetric flask;
dilute to volume with deionized
water; and store at 4°C in a dark
glass-stoppered bottle
3. Ammonium Molybdate Solution -
Dissolve 4 g (NH4)6Mo7024-4H20 in
100 ml deionized water, and store
at 4°C in a plastic bottle.
4. Ascorbic Acid, 0 1M - Dissolve 1.8
g of ascorbic acid in 100 ml of
distilled water. The solution is
stable for about a week if prepared
with water containing no more
than trace amounts of heavy
metals and if stored at 4°C.
5. Combined Reagent fAAl) - Mix the
above reagents in the following
proportions for 100 ml of combined
reagent: 50 ml of 5N H2S04,5 ml of
antimony-potassium tartrate solu-
tion, 15 ml of ammonium molyb-
date solution, and 30 ml of as-
corbic acid. Mix after adding each
reagent. All reagents must be at
room temperature before they are
mixed, and must be mixed in the
order given. If turbidity forms,
shake the combined reagent, and
let stand for a few minutes (until
the turbidity disappears) before
processing. The 100 ml is suffi-
cient for 4 h of operation. Since the
stability of the combined solution
is limited, prepare fresh solution
for each run. NOTE 1: The com-
bined reagent deteriorates less
rapidly if kept cold (~5°C) while
being pumped into the system.
NOTE 2: The combined reagent is
more stable if ascorbic acid is
excluded and if the mixed reagent
(molybdate, tartrate, and acid) is
pumped through the distilled
water line and the ascorbic acid
solution (30 ml of 7.4 diluted to
100ml with distilled water) through
the original mixed reagent line.
6. Stock Phosphate Solution -Dis-
solve 0.1434 g of predried (105°C
for 1 h) KH2PO4 in deionized water,
and dilute to 1000 ml (100 mg/l
po«~3).
7. Standard Phosphate Solution -
Dilute 2.0 ml of stock solution to
100 ml with deionized water (2
mg/l PO4~3).
8. Prepare a series of standards by
diluting suitable volumes of stan-
dard solution to 100 ml with
deionized water. The following
dilutions are suggested:
Concentration Standard Phosphorus
Standard mg/l PO4-3 Solution, ml
A
B
C
D
E
F
G
0.02
004
008
012
020
0.30
040
1.0
2.0
40
60
100
15.0
200
4.7.4 Procedure
1. Set up the manifold (Figure 4-6,
AAI; Figure 4-7, AAII).
2. Allow both the colorimeter and the
recorder to arm up for 30 min.
Obtain a stable baseline with all
reagents when pumping deionized
water through the sample line.
3. For the AAI system, sample at a
rate of 20/h, 1 min sample, 2 mm
wash. For the AAII system, use a
30/h, 2:1 cam, and a common
wash.
4. Place standards in sampler m
order of decreasing concentra-
tions. Complete filling of the
sampler tray with unknown and
control samples.
5 Switch the sample line from
deionized water to sampler, and
begin the analysis.
6. Analyze the full calibration curve
every 30 samples; repeat one
standard every 10 samples. Peri-
odically, obtain a baseline by
turning off the sampler while in
deionized rinse water for approxi-
mately 1 mm.
7. When analytical run is completed,
draw a line between the baseline
sections of the chart. Measure the
peak heights of the standards and
samples, and record them on the
data form. Figure 4-5 for NH3+.
4.7.4.1 Notes on Operation—Most of
the information in Section 4 6.4.1 is
applicable to the operation of the
autoanalyzer for orthophosphates The
color complex formed in the phosphate
analysis absorbs at a wavelength of 880
nm. It is critical that the correct
interference filters and a red-sensitive
phototube (Techmcon 199-B021 -04) be
used to obtain the greatest sensitivity
4.7.4.2 Troubleshooting—The in-
formation covered in Section 4.6.4.2 is
applicable to troubleshooting for the
operation of the autoanalyzer for
orthophosphate.
4.7.5 Calculations
Using A Graph — Prepare a standard
curve by plotting peak heights of
processed standards against their
concen!rations; compute the concen-
trations of samples by comparing
sample peak heights with the standard
curve.
Using Linear Least Squares Fit —The
equations for linear least squares fit are
available in most elementary statistics
books (e.g., see Reference 9). 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 define a rela-
tionship between the concentrations of
SM = Small
LM = Large
50°C f
Heating \_
Bath
i
Colorii
50 mn
880 m
Mixing Coil
Mixing Coil
Wash Water
to Sampler
LM
00000000
)
r /
SM
0000
netric
{
SM
OOOO
Waste
[
T
1 i
W
4.
~^
^
P Bl
P B
R R
Y Y
0 O
G G
Proportioning
'Pump
ml/mm
2.9 Wash
2.9 Sample
0.8 Air
1.2 Distilled
0
Sampler
2O/hr.
2:1
Water
0.42 Mixed
R
2.00
eagent
Waste
Recorder
^ Flow Cell
TJ Filter
Figure 4-6. Phosphorus manifold A A I.
-------
Jan. 1981
11
Part Il-Section 4.0
ml/mm
Wash Water
to Sampler
Waste
Waste
Colorimetric
50 mm Flow Cell
880 nm Filter
Figure 4-7. Phosphorus manifold AA II.
standards (x,) and the predicted instru-
ment response (Y,).
Y, = mx,+ b. 4-10
Equation 4-10 is preferred for fitting
where major components of random
variance are assumed to be in instru-
ment response. A simple rearrange-
ment of Equation 4-10 yields the
concentration (x,) corresponding to an
instrumental measurement (Y,) of a
sample j.
X, = (Y, - b)/m
4-11
4.7.6 Quality Control
1. Analyze at least five calibration
curve standards every 30samples.
With every 10 samples, repeat one
standard. If the responses change
more than 5%, stop the analysis,
and determine the cause.
2. Analyze an analyst spike (Section
7.6.4.1, QA manual (6)), a duplicate,
and an old sample every 30 sam-
ples. Calculate the first spike of the
day from the first calibration curve,
and plot the value on the real-time
plot (Section 7.6.5.1, QA manual
(6)). Compare these data with
those obtained in the past. If the
control limits are exceeded, stop
the analysis, and search for the
problem.
3. When preparing new stxjck solu-
tion (Section 4.7.3), prepare dilute
calibration standards, from the old
and the new stock. Analyze the old
and new standards and compare
the calibration curves. Complete
the Standard Preparation form in
Section 4.10.
4.8 Sulfate, Nitrate, Phos-
phate, and Chloride Determina-
tion Using ion Chromatography
4.8.1 Scope and Application
This procedure covers the determina-
tion of sulfate, nitrate, phosphate and
choride in rainwater. The anions are
separated on an ion exchange column
because of their different affinities for
the exchange material. The material
commonly used for anions is a polymer
coated with quaternary ammonium
active sites. After separation, the anions
pass through a strong acid cation
exchange column (suppressor column)
which exchanges all cations for H+ ions-
All species are detected as acids by a
conductivity meter
Any species with a retention time
similar to that of the sought anions will
interfere. With the possible exception of
NOz", rainwater does not contain any
such species. If NOZ~ is present, it will
elute just after the Cf, and the shape of
the Cr peak will be assymetric. Large
amounts of anions elutmg close to those
anions of interest will interfere (e.g., 10
mg/l N03" will mask 0.1 mg/l PO/3);
thus the automated colorimetric pro-
cedure is preferred for PCV3 in rain-
water.
4.8.2 Sample Preparation
When an aqueous sample is injected,
the water passes rapidly through both
the separator and the suppressor
columns, and gives rise to a very low
conductivity. This negative water dip
interferes with the Cl~ analysis; thus it is
necessary to remove the water dip for
low-level Cl~ analysis by adding con-
centrated eluent to all samples and
standards
When manual addition is necessary,
use a micropipette to deliver concen-
trated buffer into each sample. The
recommended procedure (13) is-
1. Prepare 4 liters of the concen-
trated eluent, 0.6M NaHC03/0.48M
Na2COa, in a collapsible 4 liter
bottle protected from air.
2. Each day of analysis, prepare 4
liters of working eluent from the
concentrated eluent, and pour an
aliquot of the concentrate to be
used for preparing samples.
3. Pour each sample or standard into
a cleaned plastic test tube to a
previously established 10 ml mark;
micropipette 50 ul of the concen-
trated buffer into the test tube; seal
the top with parafilm, and shake to
mix.
NOTE: It is important that the concen-
trated eluent form the 4 liter bottle be
taken at the same point m time for
working eluent and for sample prepara-
tion so that the prepared samples are
analyzed using working eluent prepared
from the identical concentrated eluent.
4.8.3 Apparatus
1. Ion chromatograph (Dionex manual
models 10, 14, 16 or Dionex Auto
Ion System 12) with anion separa-
tor column (P/N 30827) and anion
suppresssor column (P/N 30828).
2. Automated injection system (14)
consisting of: Technicon auto
sampler, and Controller for samp-
ling and injection. This controller
cannot be purchased; it must be
built. The automatic cycle is initi-
ated when the sample probe
moves from the rinse cup to the
sample, sets a timer and causes
injection (a pre-determined time
after this event). The length of time
the valve stays in the inject posi-
tion is set with another timer.
These timing circuits cause injec-
tion by activating electrically
controlled pneumatic valves. With
a proportioning pump, either sam-
ple or rinse water is continually
pumped through the injection loop,
which is only "on-line" with the
Dionex system for the 30 s injec-
tion.
Technicon proportioning pump
and pump tubes - This pump is
used to feed sample or rinse water
continually through the injection
loop. No automated injection system
is needed if Dionex Auto Ion
System 12 is available.
-------
Part Il-Section 4.0
12
Jan. 1981
3. Data Recording System - Integra-
tion or strip chart recorders can be
used for recording the ion chro-
matographic peaks. The nominal
output to the recorder is 1.0 v. The
Dionex plane parallel electrode
conductivity detector, however,
gives a linear response with
concentration until electronic
saturation occurs at approximately
4.0 v. Therefore, several analytical
ranges on recorders set at different
full-scale voltages can be moni-
tored simultaneously.
4. Collapsible bags, 4- and 20-liter.
5. Pipettes - an assortment of sizes.
6. Volumetric flasks - an assortment
of sizes.
7. Two recorders or one recorder with
dual pens.
8 Disposable sample cups for auto-
sampler.
4.8.4 Reagents
1. Concentrated Eluent, 0.6M
A/atfC03/04.8M Na2C03 - Dissolve
100.8407 g NaHCO3and 101.7509g
Na2C03 in 2 liters of hot deionized
water.
2. Working Eluent, 0.003M NaHCO^/
0.0024M /Va2CO3 - Dilute 20 ml of
concentrated eluent (0.6M NaHCOa/
0.48M Na2C03) to 4 liters with deion-
ized water; transfer to a 4-liter collap-
sible bag. (If a 20 liter collapsible bag
is used, dilute 100 ml concen-
trated eluent to 20 liters.)
3. Regenerant, 1 .ON H2SOt - Add
11J. 1 ml of cone H2S04 to ~3 liters
of deionized water in a 4-liter col-
lapsible bag; dilute to ~4 liters.
4. Concentrated Buffer, 14% - Dilute
~7 ml of concentrated eluent (O.6M
NaHC03/0.48M Na2C03) to 59 ml
with deionized water.
5. Mixed Stock Solution, - 1000 mg/l
SO/, NO3~, P04~3; 200 mg/l CI".
Dissolve 0.3297 g NaCI, 1.433 g of
KH2P04,1.6305 gKN03 and 1.814
2g KzSO4 in 1 liter of deionized
water. Complete the Standard
Preparation form (end of Section
4.8).
6. Standard Solution A - Dilute 10 ml
of mixed stock solution to 100 ml
with deionized water (100 mg/l
SO<=, N03~ P04~3, and 20 mg/l CI").
7. Standard Solution B - Dilute 10 ml
standard solution A to 100 ml with
deionized water (10 mg/l SO/,
NO3", P04"3, and 20 mg/l CI").
8. Working Standards - Use standard
solutions A and B to prepare the
working standards listed in Table
4-2. These standards are prepared
by diluting the indicated volumes
of standard solutions A or B to 100
ml with deionized water.
Table 4-2. Concentration of Working Standards Used for the Analysis of
Rainwater Samples by Ion Chromatography.
Working
Standard
A
B
C
D
£
Concentration
SO/, NO:T, PO4~3
JO
5
1
05
0.1
in mg/l
cr
2
1
0.2
0.1
0.02
Milliliters of Standard
Solution A or B per 100 /ml
of working standard.
10 ml of A
5 ml of A
10 ml of B
5 ml of B
Jm/ofB
4.8.5 Procedure
1. Regenerate the suppressor column
with 1 .ON H2SO4 for 5 min, rinse
with deionized water for 15 min.
2. Set up the recorders for the most
sensitive range and for any addi-
tional ranges needed (Section
4.83).
3. Recommendations of optimum
Dionex sensitivity:
A 0003M NaHCO3/0.0024M
Na2COs eluent with normal
anion separator and suppres-
sor,
B. 1 5 umho/cm full scale on
recorder,
C 0.8-ml injection loop,
D. Flow rate for column back
pressure of 650 psi, and
E. Pressure gauge used as pump
stroke noise suppressor
4. Begin to pump the eluent through
the columns. After a stable base-
line is obtained, inject the highest
standard either manually or with
the autosampler. Adjust the zero to
approximately 10% of the chart; as
the highest standard for NO3~
elutes, adjust the recorder calibra-
tion to approximately 90% of the
chart
5. Fill the autosampler with stan-
dards and samples, starting with
the highest concentration stan-
dard and decreasing the concentra-
tions, or inject the samples manual-
ly in the same order. The first
sample to be analyzed should be a
quality control sample; a standard
should be analyzed after 10 sam-
ples and (if possible) a full standard
curve should be analyzed after the
samples. List the samples on the
data form in the order correspond-
ing to their position on the sample
tray.
6. Turn on the autosampler to start
analyzing samples.
7. When the run is complete, number
the samples on the strip chart;
draw a baseline for each sample or
standard; measure the peak height
with a clear plastic ruler; and
record the peak height on the strip
chart and on the data form (Figure
4-5).
4.8.6 Calculations
Usinga Graph— Construct calibra-
tion curves by plotting concentrations of
standards against the peak height for
each analyte. Read the concentration of
the analyte directly from the calibration
curves
Using a Linear Least Squares Fit —
Equations used for calculation of the
linear least squares fit are available in
most elementary statistics books (e.g.,
see Reference 9). The linear least
squares fit yields the following param-
eters1 slope (m), intercept (b), error of fit
(e) and correlation coefficient (r). The
slope and intercept define a relationship
between concentration of standard i (x,)
and the predicted instrument response
(Y,),
Y, = mx, + b.
4-12
Equation 4-12 is preferred for the major
components of random variance as-
sumed to be in instrument response A
simple rearrangement of Equation 4-12
yields the concentration (x,) correspond-
ing to an instrumental response of (Y,),
x, = (Y, - b)/m.
4-13
4.8.7 Quality Control
Analyze at least three calibration
standards of varying concentration per
recorder range for each analytical run.
After 10 samples (following the initial
standards), run another standard. End
the analytical run after an additional
—10 samples with another calibration
curve. If the peak heights of the
standards change more than 10%,
rerun the samples.
Analyze an analyst spike (Section
7.6.4.1, QA manual (6)) during every
run. Calculate the spike concentration
from the first calibration curve, and plot
the value on the real-time plot (Section
7.6.5.1, QA manual (6)). Compare these
data with those obtained in the past. If
the control limits are exceeded, stop the
analysis, and search for the problem.
Periodically, analyze duplicates and old
samples.
Periodically, check the resolution of
the Dionex anion separator column by
analyzing a standard containing Br" and
N03~ in equal concentrations close to a
-------
Jan. 1981
13
Part Il-Section 4.0
full-scale response (~1ug/ml) on the
most sensitive analytical range. After
this standard is run, complete the
Dionex Resolution Test form (Section
410) which indicates the method for
calculating the percentage resolution of
Br~ and NOa". Check the resolution
periodically, particularly if Br~ is known
to be in the samples. If Br~ is present,
use Dionex anion separator columns
which give better than 60% resolutions
of Br~ and NC>3~; if Br~ is not present, use
Dionex anion separator columns until
the SC>4° and NOa" resolution are at
least 60%. The column resolution should
be first documented as part of the instru-
ment performance study for the ion
chromatograph (Section 7.5.3, QA
manual (6)), so that the resolution of a
specific column can be monitored as it
decreases with extended use
When preparing a new stock solution
(Section 4.8.4), prepare dilute calibra-
tion standards from the old and the new
stock. Analyze the old and new stan-
dards, and compare the calibration
curves. Complete the Standard Prepara-
tion form in Section 4.10.
4.8.8 Troubleshooting
The most common problem with an
ion chromatograph is leakage. Depend-
ing on the location of the leak, low
column back pressure, a noisy baseline,
irreproducibility, or no instrumental
response may develop. Leaks may be
stopped by tightening the fittings. The
Dionex fittings should be tightened by
hand only; more pressure can be
exerted if a cloth is placed over the
fitting. If tightening does not stop the
leak, the tubing used for the fitting
should be reflared.
Another common problem is air in the
system; this problem is also observed as
a noisy baseline. If air is in the columns,
let the system run until the air is out. If
the air is not in the columns, take both
columns offline, and let the system
pump at 100% flow; the flow should
return to normal, and both columns can
be put back on line.
Other less common probems are a
leaky valve, a plugged column, or a
plugged flow cell. A leaky valve may
cause baseline noise. The leaking may
be stopped by loosening and retighten-
ing all fittings evenly, or it may be
necessary to disassemble and clean the
valve ultrasonically in water.
A plugged column is indicated by
increased back pressure and loss of
resolution. To clean a plugged column,
use the following procedure: (1) remove
and clean the inlet column fitting and
the Teflon cloth filter, (2) reassemble
and test the column. If back pressure is
still high, reduce the percentage flow to
less than 10%; place the column
backwards in the system; (with the inlet
column fitting off) pump off the top 1 /8
in. of column resin; replace the Teflon
cloth filter; reassemble the end fitting;
put the end fitting back on; pump the
resin back down into the column; place
the column back in the system with its
original orientation. A special tool is
required to remove the column end
fitting from the suppressor column.
Once the end fitting is removed, the
suppressor column filters may be
cleaned.
A plugged flow cell leads to signifi-
cant baseline noise. Tocleanandcleara
plugged flow cell, switch both columns
offline, remove the fitting from the exit
(top) of the flow cell, and pump eluent
through at 100% flow
If a loss of resolution is not due to a
plugged column, be sure the eluent was
correctly prepared. If it was correctly
prepared, consult with Dionex before
attempting any analytical column-
cleaning procedures
Poor reproducibility may be caused by
improper rinsing of the injection loop; it
should be rinsed with a sample volume
at least three times the volume of the
injection loop
If there is no pressure showing on the
gauge and if no leak is found, then the
pump may have to be primed. Remove
the top fitting from the liquid end,
increase the flow to 100%, andpurgeall
the air from the pump to reduce the flow
rate to normal; reconnect the pump
fitting.
4.9 Metal Determinations by
Atomic Absorption Spectro-
scopy (15)
4.9.1 General Method
4.9.1.1 Scope and Application—
Metals in solution may be determined
by atomic absorption spectroscopy. The
procedure is simple, rapid, and applic-
able to the determination of Na+, K+,
Ca++ and Mg++ in rainwater. Detection
limits, sensitivity, and optimum ranges
vary with makes and models of atomic
absorption spectrophotometers. Data in
sections for the individual metals
indicate ranges measurable by direct
aspirate Most concentration ranges
may be lowered by scale expansion and
extended upwards by a less sensitive
wavelength or by rotating the burner
head. Detection limits by direct aspira-
tion may be extended through concen-
tration of the sample by addition of
conventional solvent extraction tech-
niques. Concentration ranges for the
individual metals are somewhat de-
pendent on equipment—for example,
the type of spectrophotometer, the
energy source, and the degree of
electronic expansion of the output
signal.
In direct aspiration atomic absorption
spectroscopy, a sample is aspirated and
atomized in a flame. A light beam from a
hollow cathode lamp (cathode made of
the element to be determined) is
directed through the flame, into a
monochromator, and onto a detector
that measures the amount of light
absorbed. Absorption depends on the
presence of free unexcited ground-state
atoms in the flame. Since the wave-
length of the light beam is characteristic
of only the metal being determined, the
light energy absorbed by the flame is a
measure of the concentration of that
metal in the sample.
1. Optimum Concentration Range - A
range is the concentration limits
below which scale expansion must
be used and above which curve
correction should be considered.
This range varies with the sensi-
tivity of the instrument and the
operating conditions.
2 Sensitivity - The concentration (mg
metal/I) that produces a 1% ab-
sorption.
3. Detection Limit - Detection limits
can be either instrumental or
procedural The limits of the
former (using acid water stan-
dards) would be the signal-to-
noise ratio and the degree of scale
expansion; the latter would be the
sample matrix and the preparation
procedure. The Scientific Ap-
paratus Makers Association (SAMA)
has approved the following defini-
tion' the concentration which
would yield an absorbance value
twice the standard deviation of a
series of measurements of a solu-
tion, the concentration of which is
distinctly detectable above, but is
close to the blank absorbance
value. Detection limit values listed
on the individual analysis forms
are to be considered minimum
working limits achievable with
procedures in this manual; these
values may differ from the opti-
mum detection limits reported by
various instrument manufactur-
ers.
4. Dissolved Metals - Constituents
(metals) which pass through a 0.45
um membrane filter.
For determinations of metals, con-
tamination and loss are prime concerns.
Dust in the laboratory environment and
impurities in reagents or on apparatus
which the sample contacts are potential
contaminants. Containers for liquid
samples can introduce either positive or
negative errors in measurement of trace
metals by contributing contaminants
-------
Part Il-Section 4.0
14
Jan. 1981
through leaching or surface desorption
or by depleting concentrations through
adsorption Thus for collection and
treatment of the sample prior to analysis
the sample bottle must be linear
polyethylene, polyproplyene or Teflon; it
should be thoroughly washed with
deionized distilled water As soon as
practical after collection, the sample
must be filtered through a 0 45 um
membrane filter before analyzing dis-
solved constituents Use plastic filtra-
tion apparatus with plain nongrid-
marked membrane filters to avoid
contamination. Use the first 50 to 100
ml to rinse the filter flask, discard this
portion, and collect the required filtrate
volume To stabilize the samples, add 2
ml 5% - 0.5% HN03-La(NO3)3 solution
(4 9.2.2) to 8 ml of each sample.
The most troublesome interference m
atomic absorption spectrophotometry is
"chemical", and it is caused by lack of
absorption of atoms bound in molecular
combination in theflame This phenom-
enon can occur when the flame is not
sufficiently hot to dissociate the mole-
cule (e g , PO4'3 interference with Mg++)
or when the dissociated atom is im-
mediately oxidized to a compound that
does not dissociate further at the flame
temperature Addition of lanthanum
overcomes the PCu~3 interference in the
Mg++ and Ca++ determinations.
lonization interferences occur where
the flame temperature is sufficiently
high to cause removal of an electron
from a neutral atom and thus to
generate a cation This interference can
generally be controlled by addition to
both standard and sample solutions of a
large excess of an easily ionized ele-
ment.
Although rare, spectral interference
can occur when an element not being
determined absorbs light within the
width of the absorption line utilized for
the determination of the element of
interest. The results would be errone-
ously high due to the interfering ele-
ment
Interference can occur when reso-
nant energy from an element in a
multielement lamp or when metal
impurity in the lamp cathode is within
the band pass of the slit setting if that
metal is in the sample; this interference
may be reduced by narrowing the slit
width.
4.9.1.2 Apparatus—
1. Atomic Absorption Spectrophoto-
meter - Single or dual channel,
single or double beam instrument
having a monochromator, a photo-
multiplier detector, adjustable
slits, a wavelength range of 190 to
800 nm, and provisions for an
interfacing strip chart recorder
2. Burner -The type recommended by
the instrument manufacturer should
be used.
3. Hollow Cathode Lamps - Single
element lamps are preferred but
multielement lamps may be used.
Electrodeless discharge lamps
may be used when available
4. Strip Chart Recorder.
5. Pipettes and Volumetric Flasks -
various sizes.
6. Pressure-Reducing Valves -Sup-
plies of fuel and oxidant are
maintained at pressures some-
what higher than controlled operat-
ing pressure by suitable valves
7. Glassware - All linear polyeth-
ylene, polypropylene and Teflon
containers, including sample bot-
tles, should be washed with deter-
gent and rinsed with tap water, 1 1
NH03, tap water, 1.1 HCI, tap
water, and deionized distilled
water m that order
4.9.1.3 Reagents —
1 Deionized Water
2 Nitric Acid (cone.) - If metal
impurities are present, use a
spectrograde acid.
3. Lanthanum Nitrate, La/NO^-GHzO
- Use a grade of chemical desig-
nated for flame enhancement in
atomic absorption
4. Stock Standard Metal Solutions -
Prepare using individual metal or
salt (Section 4.9.1.4). Commercial-
ly available stock standard solu-
tions may also be used.
5. Calibration Standards - Prepare a
series of standards of the metal by
diluting the appropriate stock
metal solution to cover the concen-
tration range.
6. Fuel and Oxidant - Commercial
grade acetylene is generally ac-
ceptable. Air may be supplied from
a compressed air line, a laboratory
compressor, or a cylinder of com-
pressed air.
4.9.1.4 Calibration—Prepare stock
standard solutions from high-purity
metals, oxides, or nonhygroscopic
reagent grade salts using deionized
distilled water and redistilled HNO3 or
HCI. (See individual analysis forms for
specific instructions.) Avoid H2SO4 and
HsPCu since they adversely affect many
elements. Prepare the stock solutions at
concentrations of 1000 mg metal/l.
Commercially available standard solu-
tions may be used.
Prepare a blank and six calibration
standards by diluting the stock metal
solutions to various concentrations in
the appropriate range; for best results,
prepare fresh for each analysis, and
discard after use. Prepare calibration
standards using the same concentra-
tion of acid and lanthanum nitrate as
will result in the samples following
processing. Since the background
contamination m the HNOs and lan-
thanum nitrate may vary, the same
stock HNOa-lanthanum nitrate reagent
must be used for both the samples and
the calibration standards. Beginning
with the blank and proceeding toward
the highest standard, aspirate the solu-
tions and record the readings. Prepare a
six-point calibration curve after 20
samples, and analyze a single standard
after 10 samples.
4.9.1.5 Atomic Absorption (General
Procedure)—Differences between makes
and models of atomic absorption spectro-
photometers prevent formulation of
detailed instructions applicable to every
instrument, so the analyst shouldfollow
the manufacturer's operating instruc-
tions for the particular instrument.
1. Set up the instrument by turning
on the light source, setting the
wavelength, aligning the light
source for maximum response,
moving the burner up into the
beam and visually aligning it.
2 Lower the burner, and ignite the
flame
3. Allow the instrument to warm up
15 min to minimize the effect of
wave-length shifts due to grating
expansion.
4 Reset the wavelength.
5. Optimize the nebulization rate by
aspirating a high standard (less
than 0 2 absorbance unit response)
and adjusting the nebulization rate
for maximum response.
6. Determine the detection limits by
the baseline noise of the atomic
absorption light source. With the
flame on and the instrumental
conditions optimized, use a scale
expansion so that the baseline
noise is ±1 % of full-scale response.
(Scale expansion may be achieved
by using the instrument scale
expansion functions or by monitor-
ing the AA output with a recorder
set on a full-scale response less
than the nominal 10 mv.)
7. Reaspirate the high standard and
adjust the burner position for
maximum response; the atomic
absorption instrument should be
ready for operation.
8. Prepare a full calibration curve
every 20 samples and analyze one
standard every 10 samples.
9. When an analytical run is com-
plete, draw a baseline for all
samples; measure the peak heights
of the standards and samples; and
record them on the data form
(Figure 4-5).
-------
Jan. 1981
15
Part Il-Section 4.0
Calculations—Prepare a calibration
curve by plotting peak heights of
processed standards against their
concentrations, and compute the con-
centrations of samples by comparing
sample peak heights with those on the
standard curve, or calculate sample
data from the linear least squares fit
parameters of the bracketing calibration
standards. (Equations for the fit are in
most elementary statistics books (e g.,
see Reference 9).)
The linear least squares fit yields the
following parameters: slope (m), inter-
cept (b), error of fit (e), and the correla-
tion coefficient (r) The slope and
intercept define a relationship between
concentration standards (x,) and the
predicted instrument response (Y,),
Y, = mx, + b.
4-14
This equation is preferred for fitting
where major components of random
variance are assumed to be in instru-
ment response. A simple rearrange-
ment of this equation yields the concen-
tration (x,) corresponding to an instru-
mental response (Y,) for a sample j,
x, = (Y, - b)/m.
4-15
Concentrations obtained must be cor-
rected for the dilution made to stabilize
the sample.
Troubleshooting - Lack of sensitivity
may be due to non-optimized instru-
mental conditions (wavelength, slit
width, lamp or burner alignment,
aspiration rate, or scale expansion).
Occasionally, loss of sensitivity may be
due to a clogging of the nebulizer with
particulates; the nebulizer should be
cleaned with a thin wire, according to
the manufacturer's instructions. A
change in sensitivity may be due to
insufficiently warming up the instru-
ment, clogged aspirator or an acetylene
cylinder containing less than 1 00 psi of
acetylene.
Excessive baseline noise may result
from non-optimized instrumental condi-
tions or from degradation of a light
source which should be replaced.
Nonlineanty of response may result
from improperly prepared standards,
incorrect instrumental conditions, or
nonlinear recorder response. Linearity
of the recorder response should be
checked with a constant voltage source
(2, 4, 6, 8, and 10 mv).
4.9.1.6 Quality Control—
Minimum Requirements
1. All quality control data should be
available for easy reference or
inspection.
2. An unknown performance sample
(when available) should be analyzed
twice a year for the metals mea-
sured.
Minimum Daily Control
1. A calibration curve constructed
from a minimum of a reagentblank
and six standards should be pre-
pared daily
2. If 20 or more samples per day are
analyzed, a working standard
curve must be run every 20 sam-
ples; a single standard should be
analyzed after the first 10samples,
and checks must be within +10%
of the original curve
3 An analyst spike (Section 7.6.4.1,
QA manual (6)), prepared by the
analyst from a stock solution
different than that used to prepare
the calibration standards, should
be analyzed every 20 samples and
plotted real-time, (Section 7.6.5.1,
QA manual (6)).
4. If calculations are performed with
the aid of the linear least squares
equation, the correlation coef-
ficient of the first curve should be
evaluated before continuing with
the analysis. Experience has shown
that it should be 0.9995 or greater
for NaT, K+, Ca++ and Mg++; how-
ever, the value obtained in practice
should be compared to that ob-
tained in the instrument perform-
ance study (Section 7.5.3, QA
manual (6)). If a value is lower than
the one observed before, it is
probably due to poorly prepared
standards.
5. Periodically, old samples, duplicates,
and reagent blanks should be run
(Section 7.6.6, QA manual (6)).
6. When preparing a new stock
solution, prepare dilute calibration
standards from the old and the
new stock; analyze the old and the
new standards; compare the cali-
bration curves; complete the Stand-
ard Preparation form (Section
4.10).
Optional Requirements
1. A service contract should be in
effect on balances and on the
atomic absorption spectrophoto-
meter
2. Class-S weights should be avail-
able for periodic checks on balances.
3. Chemicals should be dated upon
receipt, and replaced as needed or
before shelf-life has been exceeded.
4. A reference sample (if available)
should be analyzed once a quarter
for the metals measured.
5. At least one duplicate sample
should be run either with every 10
samples or with each set of
samples to verify precision.
6. Standard deviations should be
documented for all measurements.
7. Quality control charts or a tabula-
tion of mean and standard devia-
tion should be used to document
data validity daily.
4.9.2 Calcium Determination
4.9.2.1 Scope and Application—The
response and range are:
Optimum concentration range: 0.2 to 1
mg/l at 422.7 nm
Sensitivity 0.08 mg/l
Detection limit: 0.01 mg/l
Instrument's parameters (general) are:
1. Calcium hollow-cathode lamp
2. Fuel: acetylene
3. Oxidant: air
4. Type of flame: reducing
For sample preparation, handling,
and preservation, see Section 4.9.1.1
4.9.2.2 Reagents—
1. Stock Solution - Suspend 1.250 g
of CaCOs (analytical reagent grade),
dried at 180°C for 1 h before
weighing, in deionized water, and
cautiously dissolve with a mini-
mum of dilute HN03; dilute to 1000
ml with deionized distilled water (1
ml = 0.5 mg Ca or 500 mg/l). A
commercially available stock solu-
tion may be used.
2. Nitric Acid-Lanthanum Nitrate
Solution - Dissolve 77.93 g La(NO3)3-
6H2O in 25 ml cone. HNO3, and
dilute to 500 ml with deionized
distilled water (to yield 5% HN03 -
0.5% La).
3. Calibration Standards - Prepare
dilutions to be used as calibration
standards at the time of analysis.
To each 100-ml volumetric flask
used for calibration standards or
samples, add 20 ml of the 5% HNO3
- 0.5% La solution before diluting
to the mark; use a proportionally
smaller amount for smaller volumes.
4.9.2.3 Analysis Procedure—For
analysis procedure and calculations,
see Section 4.9 1.5, and follow notes 1 -
4.
1. PO4"3, S04Z, and Al+3 interfere, but
are masked by the addition of
lanthanum. Since low calcium
values result if the pH of the
sample is above 7, both standards
and samples are prepared in dilute
HNO3. Concentrations of Mg++
greater than 1000 mg/l also pa use
low calcium values. Concentra-
tions of up to 500 mg/l each of Na+,
K+, and NO3~do not cause interfer-
ence.
2. Anionic chemical interferences
can be expected if lanthanum is
not used in samples and standards.
3. The 239.9-nm line may also be
used. This line has a relative
sensitivity of 120.
-------
Part Il-Section 4.0
16
Jan. 1981
4.9.3 Magnesium Determination
4.9.3.1 Scope and Application—The
response and range are
Optimum concentration range: 0,02 to
0 6 mg/l at 285.2 nm
Sensitivity: 0.007 mg/l
Detection limit. .01 mg/l
Instrument's parameters (general) are'
1. Magnesium hollow-cathode lamp
2. Fuel: acetylene
3 Oxidant air
4. Type of flame: oxidizing
For sample preparation, handling,
and preservation, see Section 4.9 1 1.
4.9.3.2 Reagents—
1. Stock Solution - Dissolve 0.829 g
MgO (analytical reagent grade) in
10 ml of redistilled HN03, and
dilute to 1 liter with deionized
water (1 ml = 0.50mgof Mgor(500
mg/l) A commercially available
stock solution may be used.
2 Nitric Acid-Lanthanum Nitrate
Solution - Dissolve 77.93 g La(NO3)3-
6H20 in 25 ml cone. HN03, and
dilute to 500 ml with deionized
water (5% HN03 - 0.5% La).
3. Calibration Standards - To each
100 ml volumetric flask used for
calibration standards, add 20 ml of
the 5% HN03 - 0.5% La solution;
use a proportionally smaller amount
for smaller volumes. Then add
appropriate volume of stock solu-
tion and dilute to volume with
deionized water.
4.9.3.3 Analysis Procedure—For the
analysis procedure and the calculation,
see Section 4.9.1.5, and follow notes 1 -
2.
1 Interference caused by Al+++ con-
centrations greater than 2 mg/l is
masked by addition of lanthanum.
Na+, K+, and Ca++ cause no inter-
ference at concentrations less
than 400 mg/l
2. 202 5 nm with relative sensitivity
of 25 may also be used.
4.9.4 Potassium Determination
4.9.4.1 Scope and Application—
Optimum concentration range. 0.1 to 1
mg/l at 766.5 nm
Sensitivity: 0.04 mg/l
Detection limit 0.03 mg/l
The instrumental parameters (general)
are:
1. Potassium hollow-cathode lamp
2. Fuel: acetylene
3. Oxidant: air
4. Type of flame: slightly oxidizing
For sample handling and pre-
servation, see Section 4.9.1.1.
4.9.4.2 Reagents—
1 Stock Solution - Dissolve 0.1907 g
of KCI (analytical reagent grade),
dried at 110°C, mdeionizedwater,
and dilute to 1 liter (1 ml=0.10mg
K or 100 mg/l). A commercially
available stock solution may be
used.
2 Nitric Acid-Lanthanum Nitrate
Solution - Dissolve 77 93 g La(NO3)3-
6H20 in 25 ml cone. HNO3, and
dilute to 500 ml with deionized
water (5% HN03 - 0.5% La)
3. Calibration Standards - To each
100 ml volumetric flask used for
calibration standards, add 10 ml of
the 5% HNO3 - 0 5% La solution;
use a proportionally smaller amount
for smaller volumes. Then add an
appropriate volume of stock solu-
tion and dilute to volume with
deionized water.
4.9.4.3 Procedure—For the analysis
procedure and the calculations, see
Section 4 9,1 1.
4.9.5 Sodium Determination
4.9.5.1 Scope and Application—The
response and range are:
Optimum concentration range: 0.03 to 1
mg/l at 589.6 nm
Sensitivity: 0.015 mg/l
Detection limit- 0.03 mg/l
The instrumental parameter (general)
are
1 Sodium hollow-cathode lamp
2. Fuel: acetylene
3. Oxidant: air
4. Type of flame- oxidizing
For sample handling and preserva-
tion, see Section 4.9.1 1.
4.9.5.2 Reagents—
1. Stock Solution - Dissolve 2.542 g
of NaCI (analytical reagent grade),
dried at 140°C, in deionized distilled
water, and dilute to 1 liter (1 ml = 1
mg Na or 1000 mg/l). A commer-
cially available stock solution may
be used
2. Nitric Acid-Lanthanum Nitrate
Solution - Dissolve 77.93 g La(NO3)3-
6H20 in 25 ml cone. HNO3, and
dilute to 500 ml with deionized
water (5% HNO3 - 0.5% La). To
each 100-ml volumetric flask used
for calibration standards, add 20
ml of the 5% HNO3-0.5% La
solution before diluting to the
mark; use a proportionally smaller
amount for sample volumes. Then
add an appropriate volume of stock
solution and dilute to volume with
deionized water.
4.9.5.3 Procedure—For the analysis
procedure and the calculations, see
Section 4.9.1.1 and follow notes 1 and
2.
1 The 330 2-nm resonance line of
N.a+, which has a relative sensitiv-
ity of 185, provides a convenient
way to avoid the need to dilute
more concentrated Na+ solutions
2 Low-temperature flames increase
the sensitivity by reducing the
extent of lonization of this easily
ionized metal, lonization may also
be controlled by adding K (1000
mg/l) to both standards and samples.
4.10 Report Forms
Blank data forms are included in this
section for the convenience of the
manual user. Use of these are discussed
throughout Section 4. To correlate the
forms to the test being documented, a
number is in the lower right-hand
corner of each form (eg., 1 1/4.4
indicates form 1, version 1, discussed in
4.4 The forms included are listed
below
Form Number Title
1.1 /4.2 2 Certification of Working
Weights to NBS Forms
2.1/4.4 1 7.2 Gran Strong Acid Data
Form
3.1/40 Standard Preparation
Form
4.1/4.7 10.4 Dionex Resolution Test
4.11 References
1. Peden, M.E., LM. Skowron, F F.
McGurk, "Precipitation Sample
Handling, Analysis, and Storage
Procedures". US DOE, Pollutant
Characterization and Safety Re-
search Division, Contract EY-76-
S-02-1199, Research Report 4,
COO-1199-57 (1979); Peden, M.E.
and L.M Skowron, "Ionic Stability
of Precipitation Samples", Atmos.
Environ. 72, 2343(1978).
2. Standard Methods for the Examine -
tion of Water and Wastewater,
14th Ed, p. 460, 71, (1975).
3 Annual Book of ASTM Standards,
Part 31, Water Standard D1293-
65, p. 178(1976).
4. Galloway, J.N., B.J. Cosby, and
G.E. Likens, J. Limnol. Oceanogr.
24, 1161 (1979).
5 National Bureau of Standards,
Special Publication 250.
6. Quality Assurance Manual for
Precipitation Measurement Sys-
tems, US EPA, Research Triangle
Park, NC, EPA-Draft.
7. Gran, G., Analyst 77, 661 (1952).
8. Liljestrand, H.M. and J.J. Morgan,
Environ. Sci. and Technol., 12,
1271 (1978).
-------
Jan. 1981
17
Part Il-Section 4.0
9. An Introduction to Mathematical
Statistics, H.D. Brunk, Blaisdell
Publishing Co., Waltham, MA
1976, p. 210.
10. Methods for Chemical Analysis of
Water and Wastes, March 1979,
EPA-600-4-79-020, Method 350.1.
11. A wetting agent recommended
and supplied by the Technicon
Corporation for use in Auto Ana-
lyzers.
12. Methods for Chemical Analysis of
Water and Wastes, March 1979,
EPA-4-79-020, Method 365.1.
13. Lipski, Allan, Personal Communi-
cation, Barringer Magenta Ltd.,
Rexdale, Ontario, Canada
14. Shepard, L.S., R.J. Schwall and G.
Colovos, "Ion Chromatographic
Analysis,Using an Automated Ion
Chromatograph," presented at a
Dionex Symposium in Sunnyvale,
CA, 21 June 1978.
15. Methods for Chemical Analysis of
Water and Wastes, March 1979,
EPA-600-4-79-020, Methods 215.1,
242.1, 2581, Atomic Absorption
Methods.
-------
Part Il-Section 4.0
18
Jan. 1981
Certification of Working Weights to NBS Form
Date of Certification-
Weight Set Serial U:.
Rftlanrp O
MRS 1kg
NBS 5kg
Test 1kg
Test 5kg
#**
Ralanre O
NBS 1kg
NPS 5kg
Te^t 1kg
Test 5kg
*-»*
Ralf»nr.f> Cl
MRS 1krj
MRS 5kg
Test 1kg
Test 5kg
Summary •
A verage ± Standard Deviation
Balance 0
NBS 1kg
NRS Zkrj
Test 1kg
Tf><;t 5kg
/Analyst Signature)
Balance 0
A/fl.
-------
Jan. 1981
19
Part Il-Section 4.0
Gran Strong Acid Data Sheet
Sample #
Initial mv
Temp. °C
/j/NaOH
Injected
mv
reading
Sample #
Initial mv
Temp. °C
fj/ NaOH
Injected
mv
reading
Date:
Cone
Micr
Ca
5/j/
Total
NaOH N
opipette
libration
i,l
vnlume-
OM Manual for Precipitation Measurement
2.1/4.4.1.7.2
-------
Part Il-Section 4.0
20
Jan. 1981
Standard Preparation Form
Species
Date
(Analyst Signature)
Species
Species
Molecular
Weight
Salt
Formula
Salt
Molecular
Weight
Gross
Weight
(grams)
Tare
Weight
(grams)
Weight
Taken
(grams)
Final
Dilution
Vol. (ml)
Expected
Cone.
ffjg/ml)
Comparsion to Old Standard: Prepare new dilute calibration standards from the above stock solution and run the old
calibration standards against the new. This should be done by analyzing the old and the new
standards successively at each concentration level. Calculate the linear least squares fit of the
old and the new calibration standards
Slope (m)
Intercept (b)
Correl. Coef.
Error
Old Standard
New Standard Date of Preparation of Old Stock Solution
Date of Preparation of Old Stock Dilute Standards .
. Slope Differencea{%): ( M"~M° W 700 = .
V Mo
Detection Limit
Intercept Difference fug/ml): bn-b0 =
a n = new standard, o = old standard, M - linear least squares fit slope, b = linear least squares fit intercept.
Operations & Maintenance Manual for Precipitation Measurement Systems - 3.1/4.0
OM Manual for Precipitation Measurement
3.1/4.0
-------
Jan. 1981
21
Part Il-Section 4.0
Dionex Resolution Test
Date .
Concentration. Br
.fig/ml.
Column Back Pressure fat max. of stroke) :_
Flow rate: ml/min
Column Serial #:
Is precolumn in system?
(a) _ cm (b)
, Date of purchase:
Yes No
. fjg/ml
psi
cm (see Figure below)
Percentage Resolution: 100 x (1-a/bj.
Test Chromatogram:*
T
a
FIGURE
*Cut out test chromatogram and attach to this form.
OM Manual for Precipitation Measurement
4.1/4.7.10.4
*US GOVERNMENT PRINTING OFFICE 1983-659-095/0731
------- |