United States
Environmental Protection
Agency
Environmental Monitoring and
Support Laboratory
Cincinnati OH 45268"""
Research and Development, ,.f
Drinking Water
Laboratory
Certification for
Chemistry
Course Manual
-------
June 1986
DRINKING WATER CHEMICAL LABORATORY CERTIFICATION
This manual was developed by the Environmental Protection
Agency, Environmental Monitoring and Support Laboratory-
Cincinnati with the Technical Support Division in response
to a request from the Office of Drinking Water
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Environmental Monitoring and Support Laboratory-Cincinnati
-------
DISCLAIMER
The information in this document has been funded by the United States
Environmental Protection Agency. It has been subject to the Agency's peer
and administrative review, and it has been approved for publication as an
EPA document. Reference to commercial products, trade names, or
manufacturers does not constitute endorsement.
-------
DRINKING WATER CHEMICAL LABORATORY CERTIFICATION
INTRODUCTION
Course Instructional Objectives:
This course is designed to meet the needs for training those persons
whose responsibilities include the evaluation of laboratories which analyze
potable water for chemical parameters. The course can be used by either
State or Federal personnel.
Persons desiring to attend this course should be appointed Certification
Officers by their supervisors or be responsible for certification and be
experienced professionals and hold at least a bachelor's degree in their
respective discipline.
Persons attending this course will receive sufficient instruction to be
qualified to carry out a chemical laboratory certification inspection
according to the current Criteria and Procedures Manuals. The person will
be able to carry out the inspection which will include:
1. Laboratory facilities
2. Laboratory equipment and instrument specifications
3. Sample collecting, handling and preservation
4. Methodology, including free chlorine residual," and turbidity
5. Quality control
6. Data handling
7. General laboratory practices
8. Personnel
In addition, the following topics will be presented to prepare the
Certifying Officer to deal with the personnel operating the laboratory.
1. Implementation of the laboratory certification program
2. The pre-evaluation work-up
3. The post-evaluation conference
4. Report writing
Method of Instruction:
Instruction in the course will consist of material presented in the
lecture format combined with applied laboratory evaluation in a laboratory
set up to perform chemical analysis for potable water contaminants as
specified in the Primary Drinking Water Regulations.
The present edition of this handbook has been written from the 1982
edition of the Manual for the Certification of Laboratories Analyzing
Drinking Water: Criteria and Procedures, Quality Assurance, EPA
570/4-82-002 (hereafter called "Criteria and Procedures" manual). As that
document changes and eventually becomes finalized, this manual will need
up-dating.
When this manual is used by a State to train its evaluation team(s), the
manual must be looked at to make it fit the State's certification program.
iii
-------
TABLE OF CONTENTS
Page
Introduction iii
Table of Contents iv
Implementing the Certification Program 1
Certification of Regional Laboratories and Programs 1-1
Certification of State Laboratories 1-3
Certification of Local Laboratories 1-6
Other Considerations for Laboratory Certification 1-7
Requirements for Maintaining Certification Status 1-10
Criteria and Procedures for Downgrading/Revoking
Certification Status 1-12
Reciprocity 1-15
Training 1-15
Technical Services 1-16
Alternate Analytical Techniques 1-17
Planning A Laboratory Certification 2
Purpose 2-1
The Evaluation 2-2
Analytical Methodology 3
Introduction 3-1
Nonmetals . 3-2
Metals 3—21
Organics 3-61
Operator Tests 3-68
Summary 3-71
Sampling 4
Introduction 4-1
Monitoring Requirements 4-4
Sample Containers 4-11
Sample Collecting 4-15
Summary 4-18
Personnel 5
Introduction 5-1
Guidelines for Positions and Experience 5-2
Laboratory Staffing Needs 5-6
Suiranary 5-10
General Laboratory Practices 6
Introduction 6-1
Laboratory Facilities 6-1
General Laboratory Practices 6-5
Reagents, Solvents and Gases 6-9
Distilled and/or Deionized Water 6-10
Analytical Balances 6-13
Prepacked Kits - Calibration Intervals 6-16
iv
-------
Pre- and Post Conferences 7
Introduction 7-1
Pre-Evaluation Conference 7-2
Post-Conference 7-4
Summary 7-5
Records and Reports 8
Introduction 8-1
Analytical Quality Control Data 8-1
Sampling Records 8-4
Data Retention and Reports 8-6
Summary 8-10
Basic Statistics 9
Introduction 9-1
Frequency 9-1
Measures of Central Tendency 9-4
Measures of Dispersion 9-6
Introduction to Normal Distribution Curve 9-12
Control Charts 9-21
Interpretation of Control Charts 9-32
Quality Control 10
Introduction 10-1
Critical Elements 10-2
Recommendations 10-16
Summary 10-23
Preparing a Report of the Laboratory Certification Survey 11
Introduction 11-1
The Report 11-2
Levels of Certification .' 11-6
Summary 11-7
Instrument and Equipment Needs and Specifications 12
Introduction 12-1
Inorganic 12-2
Organics 12-17
General Lab Equipment 12-22
Laboratory Safety Practices 13
Introduction 13-1
Laboratory Design and Equipment 13-2
Handling Glassware 13-8
Gases and Flanmable Solvents 13-9
Chemical Hazards 13-11
Precautionary Measures 13-14
v
-------
Unit 1
IMPLEMENTATION OF CERTIFICATION PROGRAM
I. CERTIFICATION OF REGIONAL LABORATORIES AND PROGRAMS
The Environmental Monitoring and Support Laboratory - Cincinnati
(EMSL-CI) is responsible for certifying the regional laboratory, if one
exists, for microbiological and chemical analyses, and for approving the
regional program for certifying other laboratories for these parameters.
EMSL-Las Vegas (EMSL-LV) has similar responsibilities for a region having
radiochemistry capability. The regional certification program must be
approved before a region can exercise its authority to certify other
laboratories.
A. Certification of Laboratories
Regional laboratories analyzing potable water samples under the Safe
Drinking Water Act must meet the minimum criteria specified in the
"Criteria and Procedures" manual, pass an on-site inspection at
least once every three years, and satisfactorily analyze an annual
set of PE samples. For those regions certified for radiochemistry,
satisfactory performance on two cross check samples per year is also
necessary.
1-1
-------
B. Individual(s) Responsible for Certification Program
The USEPA Region must designate the individual(s) to coordinate
drinking water certification activities. This individual(s) must be
experienced in quality assurance; hold an advanced degree or have
equivalent experience in microbiology, chemistry, or radiochemistry;
and have sufficient administrative and technical stature to be
considered a peer of the director of the principal State laboratory.
C. On-Site Evaluation Team
One or more teams must be established by the region to evaluate a
laboratory in microbiology and chemistry. Team members must be
experienced professionals, holding at least a bachelor's degree (or
equivalent education and experience) in the specific discipline
being evaluated. Team members must participate in training
activities as specified by EMSL-CI.
D. Development of Regional Plans for Certifying Local Laboratories in
Non-Primacy States
Regions are required to develop plans for certifying local drinking
water laboratories in non-primacy states. Written plans should
include the following:
1-2
-------
1. Certification official;
2. Types and numbers of laboratories;
3. Analyses to be examined;
4. Schedule for on-site evaluations; and
5. Plans for providing technical assistance to laboratories in need
of upgrading.
II. CERTIFICATION OF STATE LABORATORIES
The principal State laboratory system must have the capability to
analyze every parameter included in the drinking water regulations (40 CFR
142.10(b)(4); however, an individual laboratory which is part of a principal
State laboratory system may be certified for only one, several, or all the
cited analyses. If a principal State laboratory contracts with another
laboratory, including a laboratory outside the State, to assume the lead
role in analyzing a regulated parameter (e.g., radiochemical contaminants),
that contract laboratory will, for the purposes of this manual, be
considered part of the principal State laboratory system. In this case, the
contract laboratory must be certified either by USEPA or by the State in
which the laboratory is located for the contaminants of interest. In the
latter case, the State must have primacy.
1-3
-------
The certification process for a principal State laboratory or a local
laboratory in a non-primacy State will begin when the laboratory director
makes a formal request to the region. This application may result from the
following:
A request for first-time certification for microbiology,
chemistry and/or radiochemistry;
A request for certification to analyze additional or newly
regulated parameters;
A request to renew a laboratory's certification status after
three years; and
A request to reapply for certification after correction of
deficiencies which resulted in the downgrading/revocation of
certification status.
The Region should respond to a formal application for any of the
requests within 30 days, and a mutually agreeable date and time should be
set for the on-site laboratory evaluation. The recommended protocol for
conducting these evaluations is given in Appendix B of the "Criteria and
Procedures" manual. For certification, a laboratory must pass an on-site
inspection and satisfactorily analyze performance evaluation samples for
those parameters for which it requests certification.
1-4
-------
After the on-site visit and the review of PE sample results, the Region
can classify the laboratory for each type of analysis according to the
following ranking scheme:
o Certified - a laboratory that meets the minimum requirements of the
"Criteria and Procedures" manual. The certification shall be valid for
up to three years.
o Provisionally Certified - a laboratory which has deficiencies but can
still produce valid data.
o Not Certified - a laboratory possessing major deficiencies which in the
opinion of the Regional Administrator, cannot consistently produce
valid data.
In the case of laboratories classified as Provisionally Certified, up
to one year will be permitted for correction of the deficiencies. A
one-time extension of no more than six months may be considered by the
Region as long as the laboratory is making "good faith" progress in the
resolution of its deficiencies and the continued provisional status does not
impact the generation of valid data. A Provisionally Certified laboratory
may analyze drinking water samples for compliance purposes. In no case
should provisional certification be given if the evaluation team believes
that the laboratory lacks the capability of performing the analysis within
1-5
-------
specified limits. Once deficiencies have been corrected to the Regional
Administrator's satisfaction, the latter should upgrade the laboratory to
Certified status.
For laboratories requesting first-time certification or certification
to analyze additional or newly regulated parameters, the Region, at its
discretion, may administratively grant a laboratory Provisionally Certified
status, pending an on-site evaluation. It is granted only when the Region
judges that the laboratory has both the appropriate instrumentation and
trained personnel to perform the analyses, and the laboratory has
satisfactorily analyzed PE samples for the contaminants in question.
For those regions lacking the expertise required to certify
laboratories in radiochemistry, EMSL-LV will conduct on-site inspections.
III. CERTIFICATION OF LOCAL LABORATORIES
For the purpose Of this document, local laboratories include any State,
county, municipal, utility, Federal, or commercial laboratory, but excludes
principal State laboratories and USEPA Regional laboratories. In
non-primacy States, the Regions will certify local laboratories, using the
criteria and policies in the "Criteria and Procedures" manual.
Only primacy States where not all drinking water analyses are conducted
at State-operated laboratories are required to establish a certification
program for local laboratories (see 40 CFR 142.10(b)). All states, however,
1-6
-------
are encouraged to develop such programs. Certification must be based upon
criteria contained 1n the "Criteria and Procedures" manual or State-
developed equivalents which are at least as stringent. Those States
required by regulation to develop a certification program must appoint a
laboratory certification officer(s) who is certified by the Region as the
official(s) responsible for the State program.
The principal State laboratory system must have the technical
capability to analyze for all regulated contaminants. If.the principal
State laboratory has the resources to perform 100% of the analyses for one
contaminant (e.g., lead), but does not have adequate resources to perform
100% of the analyses for another contaminant (e.g., can only analyze 20% of
all total coliform samples), then the State certification program need only
include certification criteria for contaminants for which the State will not
be analyzing 100% of the samples.
Federal facilities must comply with all Federal, State and local
requirements with respect to the Safe Drinking Water Act. For the purpose
of certification, Federal laboratories in which routine monitoring of public
drinking water supplies is conducted are to be considered local
laboratories. The agency with primary enforcement authority, either the
State or the Region, will be responsible for carrying out certification
activities. If requested by the State, the Region may conduct on-site
evaluations of Federal laboratories in that State. USEPA will have primary
enforcement authority over any facilities on Federal Indian lands.
1-7
-------
OTHER CONSIDERATIONS FOR LABORATORY CERTIFICATION
A. Quality Assurance Plan
It is essential that all laboratories analyzing drinking water
conpliance samples adhere to defined quality assurance procedures.
This is to insure that routinely generated analytical data are
scientifically valid and defensible and are of known and acceptable
precision and accuracy. To accomplish these goals, each laboratory
should prepare a written description of its quality assurance
activities (a QA Plan). The following items should be addressed in
each QA plan:
1. Sampling procedures.
2. Sample handling procedures.
- specify procedures used to maintain integrity of all samples,
i.e., tracking samples from receipt by laboratory through
analysis to disposal.
- samples likely to be the basis for an enforcement action may
require special safeguards (see Chain-of-Custody procedures).
1-8
-------
3. Instrument or equipment calibration procedures and frequency of
their use.
4. Analytical procedures.
5. Data reduction, validation, and reporting.
- data reduction: conversion of raw data to vg/L, picocuries/L,
coliforms/100 mL, etc.
- validation: includes insuring accuracy of data transcription
and calculations.
- reporting: includes procedures and format for reporting data
to utilities, State officials, and USEPA.
6. Types of internal quality control (QC) checks and frequency of
their use.
- may include preparation of calibration curves, instrument
calibrations, replicate analyses, use of EMSL-CI provided QC
samples or calibration standards and use of QC charts.*
* QC chart for chemistry is explained in Handbook for Analytical Quality
Control in Water and Wastewater Laboratories, EPA-600/4-79-019, March 1979.
QC chart for radiochemistry is explained in Handbook for Analytical Quality
Control in Radioanalytical Laboratories, EPA-600/7-77-088, August 1977.
1-9
-------
7. Preventive maintenance procedures and schedules.
8. Specific routine procedures used to determine data precision and
accuracy for each contaminant measured.
- precision is based on the results of replicate analyses.
- accuracy is normally determined by comparison of results with
"known" concentrations in reagent water standards and by
analyses of water matrix samples before and after adding a
known contaminant "spike."
9. Corrective action contingencies.
- response to obtaining unacceptable results from analysis of PE
samples and from internal QC checks.
The QA plan may consist of already available standard operating
procedures (SOP's) which are approved by the laboratory director and
which address the listed items, or may be a separately prepared QA
document. Documentation for many of the listed QA plan items can be
by reference to appropriate sections of the "Criteria and
Procedures" manual, the laboratory's SOPs or to other literature
(e.g., "Standard Methods for the Examination of Mater and
Wastewater").
1-10
-------
If a particular listed item is not relevant, the QA plan should
state this and provide a brief explanation (e.g., some laboratories
never collect samples and thus have no need to describe sampling
procedures). A laboratory QA plan should be concise but responsive
to the above-listed items (a maximum of five pages is suggested).
Minimizing paperwork while improving dependability and quality of
data are the intended goals.
B. Chain-of-Custody Procedures
Certified laboratories which may be requested to process a sample
for possible legal action against a supplier must have a
chain-of-custody procedure available.
V. REQUIREMENTS FOR MAINTAINING CERTIFICATION STATUS
A. Periodic Performance Evaluation (PE) Samples
Certified drinking water laboratories must satisfactorily analyze PE
samples on an annual basis for each chemical, radiochemical, or
microbiological parameter (when available) for which certification
has been granted. Results must be within the acceptance limits
established by USEPA for each analysis. To main certification in
radiochemistry, the laboratory must satisfactorily analyze two
cross-check samples per year in addition to the annual set of PE
samples.
1-11
-------
B. Methodology
Laboratories must use methodologies sanctioned by the National
Interim Primary Drinking Water Regulations (40 CFR 141.21 - 141.25)
or otherwise approved by USEPA for compliance with the Safe Drinking
Water Act.
C. Notification of Certifying Authority (CA) for Major Changes
Laboratories must notify the appropriate CA (Regional Administrator
or the appropriate EMSL), in writing, within 30 days of major
changes in personnel, equipment, or laboratory location which might
impair analytical capability. A major change in personnel is
defined as the loss or replacement of the laboratory supervisor or a
situation in which a trained and experienced analyst is no longer
available to analyze a particular parameter for which certification
has been granted. The CA will discuss the situation with the
laboratory supervisor and establish a schedule for the laboratory to
rectify deficiencies.
D. On-Site Evaluation
The CA must be satisfied that a laboratory is maintaining the
required standard of quality for certification. Normally, this will
be based upon a recommendation resulting from a USEPA on-site
evaluation conducted at least every three years.
1-12
-------
CRITERIA AND PROCEDURES FOR DOWNGRADING/REVOKING CERTIFICATION STATUS
A. Criteria for Downgrading Certification Status
A laboratory may be downgraded to a Provisionally Certified status
for a particular contaminant analysis for any of the following
reasons:
1. Failure to analyze a PE sample (or an EMSL-LV cross-check sample)
within the acceptance limits established by USEPA. If more than
one concentration of a particular contaminant is provided, the
laboratory must satisfactorily analyze all concentrations, except
where otherwise stated. After downgrading to a provisionally
certified status, a laboratory may request that USEPA provide QC
samples (standard solution samples for radiochemical
contaminants) and technical assistance to help identify and
resolve the problem. Provisonally Certified status will continue
until the laboratory's analysis of a follow-up PE sample (or
EMSL-LV cross-check sample) produces data within the acceptance
limits established by USEPA.
2. Failure of a certified laboratory to notify the CA within 30 days
of major changes in personnel, equipment, or laboratory location
which might impair analytical capability.
1-13
-------
3. Failure to satisfy the CA that the laboratory is maintaining the
required standard of quality, based upon an USEPA on-site
evaluation.
During the provisional status period, which may last for up to one
year with a possible six month extension, the laboratory may
continue to analyze samples for compliance purposes until it
resolves its difficulties. It must, however, notify its clients of
its downgraded status.
B. Criteria for Revoking Certification Status
A laboratory may be downgraded from Certified or Provisionally
Certified status to a Not Certified classification for a particular
contaminant analysis for the following reasons:
1. Failure to analyze an initial and follow-up PE sample (or EMSL-LV
cross-check sample) for a particular contaminant within the
acceptance limits established by USEPA.
2. Failure to correct identified deviations (including continued use
of unapproved methods and equipment) by the time specified by the
CA.
3. Submission of a PE sample to another laboratory for analysis and
reporting data as its own.
1-14
-------
4. Falsification of data or other deceptive practices.
C. Procedures for Revocation
The CA will notify the laboratory in writing (registered or
certified mail) of the intent to revoke certification. If the
laboratory wishes to challenge this decision, a notice of appeal
must be submitted in writing to the CA within 30 days of receipt of
the notice of intent to revoke certification. If no notice of
appeal is so filed, certification will be revoked.
The notice of appeal must be supported with an explanation of the
reasons for the challenge and must be signed by a responsible
official from the laboratory such as the president/owner for a
commercial laboratory, or the laboratory supervisor in the case of a
municipal laboratory.
Within 60 days of receipt of the appeal, the CA will make a decision
and notify the laboratory in writing. Denial of the appeal results
in immediate revocation of the laboratory's certification. The CA
will request the laboratory to notify its clients of its status in
writing, and to submit verification that this has been
accomplished. Once certification is revoked, a laboratory may not
analyze drinking water samples for compliance purposes until its
certification has been reinstated.
1-15
-------
If the appeal is determined to be valid, the CA will take
appropriate measures to reevaluate the facility and to issue to the
laboratory within 60 days a written decision on its certification
status.
D. Reinstatement of Certification
Certification will be reinstated when and if the laboratory can
demonstrate to the CA's satisfaction that the deficiencies which
produced Provisionally Certified status or revocation have been
corrected. This may include an on-site evaluation, a successful
analysis of samples on the next regularly scheduled EMSL water
supply performance evaluation study, or any other measure the CA
deems appropriate.
VII. RECIPROCITY
Reciprocity, which is defined as mutually acceptable certification among
Regions and States, is endorsed by USEPA as a highly desirable element in
the certification program for drinking water laboratories. States are
encouraged to adopt provisions in their laws and regulations to permit it.
States may request USEPA to arbitrate disputes involving reciprocity. Data
from USEPA certified laboratories will be acceptable under the Safe Drinking
Water Act in jurisdictions where USEPA has primary enforcement
responsibility.
1-16
-------
VIII. TRAINING
Training is an interal part of the laboratory certification process
for: a) personnel responsible for certifying laboratories either on behalf
of the Regional office or a primacy State; b) the laboratory analysts
responsible for microbiological, chemical, and radiochemical measurements.
Mechanisms for providing adequate training should be examined by primacy
agencies or other groups.
IX. TECHNICAL SERVICES
A. Reference Samples
There are four types of EMSL reference samples: calibration
standards, quality control (QC), performance evaluation (PE), and
intercomparison cross-check samples. EMSL-CI provides QC and PE
samples for all regulated chemical and microbiological contaminants
and residual chlorine, and in addition, provides calibration
standards for trace organic chemicals. As part of the QC/PE sample
packages, contaminant concentrations are furnished along with
detailed user instructions. EMSL-LV provides calibration standards,
PE samples, and intercomparison cross-check samples for all
regulated radiochemical contaminants.
1-17
-------
QC samples and standards are provided on request as part of a
laboratory's own quality assurance activities. They serve as
independent checks on reagents, instruments, and analytical
techniques; as an aid for testing or training analysts; or for
determining intralaboratory precision and accuracy. Although no
certification or other formal USEPA evaluation functions result from
using these samples, their routine use is considered fundamental to
a proper laboratory QA plan.
EMSL-CI and EMSL-LV conduct periodic water supply performance
evaluation studies using PE samples as a requirement for
certification. In contrast to QC samples and calibration standards,
contaminant concentrations are not furnished before analysis.
Laboratories should request PE samples through the appropriate
Regional office for chemistry and microbiology, or EMSL-LV for
radiochemistry.*
At the conclusion of each study, the EMSLs prepare individual
reports for each laboratory and provide them to the participants.
The certifying authority reviews unacceptable data with the
laboratory to identify and resolve problems. QC samples and
calibration standards are useful for this purpose. Once problems
are corrected, the laboratory must analyze a second series of PE
samples for problem parameters during a follow-up EMSL study.
*EMSL-LV address is USEPA/EMSL, P.O. Box 15027, Las Vegas, Nevada 89114
1-18
-------
In addition to the annual PE sample requirement, EMSL-LV also
requires satisfactory performance in two intercomparison cross-check
studies per year. Intercomparison cross-check samples differ from
PE sanqples in that the former contain only one or two radionuclides
(e.g., radium-226 and radium-228), while PE samples for
radiochemistry are complex mixtures of alpha, beta, and
photon-emitting radionuclides. In neither case are contaminant
concentrations furnished to the laboratory until after completion of
the study.
X. ALTERNATE ANALYTICAL TECHNIQUES
Section 141.27 of the National Interim Primary Drinking Water
Regulations permits approval of alternate analytical techniques. Such a
technique, also known as an alternate test procedure
(ATP), shall be accepted "...only if it is substantially equivalent to
the prescribed test in both precision and accuracy as it relates to the
determination of compliance with any maximum contaminant level."
EMSL-CI, through its Equivalency Staff, is responsible for and provides
coordination of a program to determine the acceptability of proposed
techniques, and makes recommendations for approval or denial to the
appropriate authority. Applications for approval of an ATP may be made
on a limited or a nationwide basis. Requests for limited use approvals
1-19
-------
are submitted to the appropriate Regional Administrator or designee who,
after receiving recoirmendations from the Director of EMSL-CI and the
Director of Office of Drinking Water, has the final authority to approve
requests. Requests for nationwide use approvals should be forwarded
directly to the Director of EMSL-CI, who, after review by the
Equivalency Staff, will provide recommendations to the Director of
Office of Drinking Water, with whom final authority for approval rests.
Applicants who propose an ATP need to provide a step-by-step procedure,
applicable literature citations or other references, and any available
conparability data between the proposed ATP and the USEPA-approved test
procedure for the same contaminant.
In the case of new techniques where precision and accuracy data may not
be available to support an application, the applicant for a limited use
ATP will be asked to provide comparability data from the analysis of
samples collected from one to five water supply systems most
representative of those routinely analyzed. From each of these systems,
three samples, in which the concentrations range from the limit of
detection (LD) to the maximum contaminant level (MCL), should be
collected and each sample analyzed eight times, four times each, by the
proposed technique and the USEPA-approved test procedure. Samples can
be spiked as necessary to cover the concentration range between the LD
and MCL. Applicants for a nationwide ATP may also be asked to provide
comparability data. For each of a minimum of five sources, six samples
should be collected and analyzed eight times, four times by each
1-20
-------
technique. The water supply sources selected must be dispersed
geographically throughout the United States. Samples may also be spiked
as necessary to cover the concentration range between the LD and the MCLi
The EMSL-CI Equivalency Staff will apply a series of proven statistical
techniques to the data submitted with all applications to determine
equivalency between the proposed and the approved techniques, and submit
recommendations to approval authorities.
1-21
-------
Unit 2
PLANNING A LABORATORY CERTIFICATION
Before actually mentioning the steps involved in preparing for a
certification visit, a few words of introduction are needed. It is
important that the Certification Officer have the proper attitude when
carrying out a survey. He is attempting to help the laboratory, by offering
a form of technical assistance and training combined. He is to aid the
laboratory in correcting problems that exist and offer suggestions for
improvement.
I. PURPOSE
The object of reviewing a laboratory's analyses of drinking water
samples is to increase the quality of those analyses so that the water
consumer or recreational user's health is given the greatest possible
protection. It is better for the evaluator to consider the survey as a
conference on methods and procedures with emphasis on the acceptance of an
approved methods approach as verification of data reliability. This
attitude yields much better results with the majority of the laboratories
than does having the Certification Officer emphasize the regulatory aspects
of his visit. Certainly endorsement of the laboratory as a certified
laboratory by the State or Federal government does bring significant
prestige, while the discussion with recognized experts in drinking water
analysis does afford the opportunity for increased technical knowledge.
2-1
-------
II. THE EVALUATION
The frequency of evaluation and certification of a laboratory has been
set at once every three years. Past experience has shown that more
frequent evaluations yield little additional value to either the staff
or the program. However, where there are major difficulties or where
there are large turn-overs of laboratory personnel, evaluations should
be performed at more frequent intervals.
A. Requesting an Evaluation
Certification should be initiated by a formal request from the
laboratory to the appropriate certification authority. This request
should include a statement stating which type of analysis or tests
for which the laboratory is seeking certification and the name,
address, and telephone number of the laboratory's contact person. A
laboratory may be certified for one, several, or all analyses
included in the National Interim Primary Drinking Water Regulations
revised as published in the Federal Register* of August 27, 1980
(chemical and bacteriological). A timely response to the request
for the laboratory evaluation should be given (preferably within 30
days).
2-2
-------
B. Scheduling an Evaluation
In order to set a mutually agreeable date and time for the
evaluation, the person designated to schedule laboratory evaluations
should contact the laboratory director. The date selected should be
agreed upon by all members of the certification team and should
allow sufficient time to conduct the evaluation without being
rushed. This pre-evaluation conference with the laboratory director
should establish an evaluation schedule that would have a minimum
impact on the laboratory activities.
As a guide to help establish scheduling of laboratory
certifications, a minimum of one week should be allowed for each
laboratory to be certified. This allows approximately one and a
half to two days for the evaluation and about three days for the
report to be written. The on-site evaluation time assumes that the
chemical and bacteriological evaluations are carried out
simultaneously by different individuals during the same visit.
The guidelines do not consider travel time. Travel time can be
minimized if laboratories to be evaluated within a given year are
grouped into a defined geographical area so that the team can visit
several laboratories before returning to their office.
2-3
-------
C. Data Reviews
If the information is available, the chemical Certification Officer
should review the results of the annual unknown performance sample
analyzed by the laboratory. If the laboratory has participated in
any performance studies or has analyzed any known performance
samples, these results should also be evaluated. This data
evaluation could indicate to the Certification Officer which
analysis he might wish to schedule for operation during his visit.
If the laboratory has previously been evaluated by an on-site visit,
the Certification Officer should review this pertinent data,
particularly if the previous on-site visit was conducted by someone
other than himself. He should note any unacceptable areas to assure
himself they have been corrected properly. This should also incude
any equipment needs which were suggested during the previous visit.
A few words on the use of the survey forms might be appropriate
here. Remember the intent of the survey form is to serve as a
guideline for complete coverage of the laboratory's activities, not
as a grading sheet for answers supplied by the laboratory staff. It
is from this basic information that the Certification Officer
formulates his oral report in a wrap-up conference held at the
conclusion of the visit and from which a formal report is prepared
with specific comments and recommendations.
2-4
-------
Each deviation observed during the laboratory evaluation should be
discussed at the time it is observed. The discussion should include
the deviation, its effect on validity of results, remedial action,
and reasons justifying the change in procedures. The final hour of
the laboratory evaluation visit has traditionally been devoted to an
informal presentation of the material to be covered in the report.
Generally, the wrap-up conference is made to the laboratory
director, chemist in charge of the water program, and a
representative of the water supply engineering staff. If the
laboratory is a principal State laboratory, the presence of Regional
engineering staff members from the federal water program should be
encouraged whenever the evaluation involves public water supplies.
2-5
-------
Unit 3
ANALYTICAL METHODOLOGY
I. INTRODUCTION
The analytical methods section of the Criteria and Procedures manual is
one section that is to be considered mandatory. This is due to the
inclusion of the methods in the Federal Register issuance of the National
Interim Primary Drinking Water Regulations of December 24, 1975, as well' as
additions made in the National Interim Primary Drinking Water Amendments
published on August 27, 1980, and the Trihalomethane regulations published
November 29, 1979. These promulgations determine the methods that all
laboratories will use for analyzing potable water supplies.
Until the National Interim Drinking Water Regulations are revised.or
additional alternate methods are approved, the methods and the references
shown in Table I and III are to be used.
This unit will -separate the methods into inorganic and organic types of
analysis in order to facilitate the coverage. The inorganic category will
be further broken down into metal and nonmetal techniques. However, the
methods will not be covered in detail in this unit because it is assumed
that the trainee in this course has knowledge of all the methods.
When the Certifying Officer visits a laboratory there is very little
possibility of evaluating the technique of all the methods. Rather, the
3-1
-------
individual should consult the results of the unknown performance sample the
laboratory has analyzed. The records of the laboratory should contain
results of the analysis of internal laboratory check samples run as part of
the laboratory's quality control program. These data should indicate to the
Certifying Officer any trouble spots which exist in the methodology. It is
at this point where this outline will be of value, to point out spots where
trouble could most likely occur.
II. THE NONMETAL METHODS
The contaminants covered in this section are nitrate, fluoride, sodium,
and the corrosivity parameters. Nitrate and fluoride have established
MCL's. Sodium has a set monitoring frequency and a reporting requirement
but no MCL. Corrosivity must be determined and calculated, and in order to
do this the alkalinity, temperature, total dissolved residue, and calcium
must be determined. All methodology for inorganics that must be carried out
in a approved laboratory are listed in Table I.
A. Nitrate
1. Manual Cadium Reduction Method
The Cadmium Reduction method reduces all nitrate to nitrite by
passing the sample through a column containing copperized
cadmium. The colorimetric step diazotizes the nitrite with
sulfanilamide and couples with N(1 naphthyl)-ethylenediamine to
form a highly colored azo dye which is measured. In order to get
3-2
-------
Table I
Approved Methodology for Inorganic Contaminants
to be done in Approved Laboratories
Contaminant
Methodology
EPA1
ASTM2
SM3
USGC4
AlkalInity
Methylorange titrimetric or
potentiometric
310.1
D1067-70B
403
Arsenic
Atomic Abs; Furnace
206.2
Atomic Abs; gaseous hydride
206.3
D2972-78B
301A-VII
1-1062-78*
Spectrophote; silver diethyldi-
thiocarbamate
206.4
D2972-78A
404A after
8(4)
Barium
Atomic Abs; direct aspiration
208.1
301A-IV
Atomic Abs; furnace
208 i 2
Cadmi um
Atomic Abs; direct aspiration
213.1
D3557-78A or B
301A-11 or
111
Atomic Abs; furnace
213.2
Calcium
Atomic Abs; direct aspiration
215.1
D2576-780
301A-11
EDTA; titrimetric
215.2
D1126-67B
306C
Chromium
Atomic Abs; direct aspiration
218.1
D1687-77D
301A-11 or
111
Atomic Abs; furnace
218.2
Corrosivity
Langelier Index
203
Agressive Index
C400-773
3-3
-------
Table 1
(continued)
Contaminant
Methodology
EPA1
ASTM2
SM
Other
Fluoride
Colorimetric; SPADNS with
distillation
340.1
D1179-72A
414A and C
Potentionmetric; ion selective
electrode
340.2
D1179-72B
414B
Automated Alizarin blue; with
distillation
340.3
603
129-71W6
Zirconium eriochrome cyanine R
with distillation
1-3325-784
Lead
Automated ion selective electrode
Atomic Abs; direct aspiration 239.1
D3559-78A or B
301A-11 or 111
380-75WF?
Atomic Abs; furnace
239. Z
Mercury
Manual cold vapor
245.1
D3223-79
301A-VI
Automated cold vapor
245.2
Nitrate
Colorimetric; brucine
352.1
D992-71
419D
Spectrophotometry; cadmium
reduction
353.3
D3867-79B
419C
Automated hydrozine reduction
353.1
Automated cadmium reduction
353.2
D3867-79A
605
Ion selective electrode
93MM-798
Ion Chromatography
300.0
Selenium
Atomic Abs; furnace
270.2
Atomic Abs; gaseous hydride
270.3
D3859-79
301
Silver
Atomic Abs; direct aspiration
272.1
301A-11
Atomic Abs; furnace 272.2
Sodium Atomic Abs; direct aspiration 273.1
Atomic Abs; furnace 273.2
Flame photometric D1428-64A 320A
Total filterable residue Gravimetric 160.1 208B
3-4
-------
1. "Methods of Chemical Analysis of Water and Wastes" EPA Environmental Monitoring and Support Laboratory,
Cincinnati, Ohio 45268 (EPA-600/4-79-020) March 1979. Available from 0R0 Publications, CERI, EPA,
Cincinnati, Ohio 45268. For approved analytical procedures for metals; the technique applicable to total
metals must be used.
2. Annual Book of ASTM STandards, Part 31 Water, American Society for Testing Materials, 1916 Race Street,
Philadelphia, Pennsylvania 19108.
3. "Standard Methods for the Examination of Water and Wastewater", 14^ Edition, American Public Health
Association, American Water Works Association, Water Pollution Control Federation, 1975.
4. Techniques of Water Resources Investigation of the United States Geological Survey, Chapter A-l "Methods
for Determination of Inorganic Substances in Water and Fluvial Sediments," Book 5 (1979, Stock
#024-001-03177-9). Available from Superintendent of Documents, U.S. Government Printing Office,
Washington, D.C. 20402.
5. "AWWA Standard for Asbestos-Cement Pipe, 4 in. through 24 in. for Water and Other Liquids" AWWA C400-77,
Revision of C400-75, AWWA, Denver, Colorado.
6. "Fluoride in Water and Wastewater, Industrial Method #129-71W". Technicon Industrial Systems, Tarrytown,
New York 10591, December 1972.
7. "Fluoride in Water and Wastewater", Technicon Industrial Systems, Tarrytown, New York, February 1976.
8. "Methods Manual-93 Series Electrodes" Form 93 MM/9790, pp 3-6, 1970. Onion Research Incorporated,
Cambridge, Massachusetts.
3-5
-------
a single answer for nitrate alone the sample would have to be run
with and without passing the sample through the column, obtaining
two values for nitrite only and for nitrate plus nitrite present
in the sample. The nitrate concentration is found -by difference
between the reduced and nonreduced values.
The sample should be filtered through a glass fiber filter or a
0.45u membrane filter before passage through the column. Build-up
of suspended matter in the reduction column will restrict sample
flow. Highly turbid samples may be pretreated with zinc sulfate
before filtration to remove the bulk of particulate matter present
in the sample. Other possibilities for interferences are the
presences of high concentrations of iron, copper or other metals.
EDTA is added to the sample to eliminate this interference.
Care must be taken when preparing the reduction column. If the
column is not properly prepared, the column efficiency for
converting nitrate will be low. This column efficiency should be
calculated for all newly prepared columns and each time the column
is used. For the primary determination of the column's
efficiency, a series of nitrite standards should be compared to
reduced nitrate standards of the same concentrations to check the
column efficiency. Then, each time the column is used, at least
one nitrite standard should be compared to a reduced nitrate
standard of the same concentration. Efficiency should range from
96 to 104%.
3-6
-------
The Certifying Officer should also be aware of the toxicity of
cadmiun so he can caution the analyst about proper techniques
for handling the cadmium during, column preparation and in the
proper disposal of column washings which might contain some of
the metal.
The cadmium should not be allowed to dry out; it should be
stored when not in use by addition of the ammonium
chloride-EDTA solution to the column assuring that it covers
the cadmium. Also, after the sample has been reduced the
colorimetric procedure should be carried out as soon as
possible and in no case should the reduced sample be allowed to
stand longer than 15 minutes before color development is
begun. Finally, the flow rate from the column is critical and
should be kept between the specified rate of 7-10 mL per minute.
2. Automated Cadmium Reduction Method
The procedure involved is the same as that of the manual
method. Care must be taken to be assured that air is not
trapped in the reduction column. In order to minimize
entrapment of air the straight column is inclined at a 20°
angle and the U shaped column should be packed wet. The column
flow characteristics must not be impaired; by sample turbidity
or formation of colloidal copper. The operator should not run
a distilled water wash, rather, he should use ammonium chloride
while the column is in the system.
3-7
-------
3. Automated Hydrazine Reduction
This method utilizes the Auto Analyzer, which can run about 20
samples per hour. The apparatus included in the system must
contain a heating bath as well as a continuous filter to remove
the precipitate produced in the reaction. Time should be
allowed between peaks for complete washout of the sample, i.e.
the recorder should return to near baseline before the next
peak.
The Hydrazine sulfate reagent is toxic and the analyst should
be aware of this. The color developing reagent is stable for
about one month and should be kept in a dark bottle in the
refrigerator.
Standards should be run about every 20 samples to assure that
the standard curve is still correct. The curve is produced by
plotting known concentration of standards against peak heights.
4. Brucine Method
The Brucine method analyzes for nitrate itself so two runs are
not necessary as in the Cadmium Reduction method. The reaction
between nitrate and Brucine produces a yellow color which can
be used for the colorimetric determination of nitrate. The
main drawback of this method is the need for reproducibility of
technique and reaction conditions. Consequently, it is
recommended that at least two standards for each batch of
samples are analyzed as a check.
3-8
-------
All strong oxidizing or reducing agents interfere. The
presence of oxidizing agents may be determined by adding
orthotolidine reagent, as in the measurement of residual
chlorine. The interference by residual chlorine may be
eliminated by adding sodium arsenite, provided that the
residual chlorine does not exceed 5 mg/liter. A slight excess
of sodium arsenite will not affect the determination. Ferrous,
ferric iron, and quadrivalent manganese give slight positive
interferences, but in concentrations less than 1 mg/liter these
are negligible. An interference due to nitrite of up to 0.5 mg
NOg-N/liter is eliminated by the use of sulfanilic acid.
Chloride interference is masked by addition of excess NaCl.
As mentioned previously, the duplication of reaction conditions
is a prime concern in the Brucine procedure. Consequently,
cautions are given about spacing the reaction tubes in racks
with each tube being surrounded by empty spaces. Mixing the
sample thoroughly and not using "vortex" type mixers are also
items where caution is prescribed.
The method calls for the light path of the spectrophotometer to
be at least 2.5 cm (1 inch) long and recommends using reaction
tubes approximately 2.5 x 15 cm in conjunction with the
spectrophotometer to determine transmittance while avoiding the
necessity of transfer following the reaction. In addition,
3-9
-------
using a stirred water bath with a gabled top is recommended in
order to maintain a temperature of at least 95°C when cooled
samples are introduced.
Again, the reagents Brucine sulfate and sodium arsenite are
toxic and precautions should be taken in the laboratory so that
these reagents are not ingested.
Of the two manual methods, the Brucine method is probably best
used in larger laboratories where large numbers of samples are
analyzed in batches. The procedure of the Brucine method would
be at its best under these circumstances. Having the same
analyst using the method as often as possible would also
enhance its performance, as good technique is often generated
by constant repetition. Smaller laboratories would not need
the stirred water bath and special tubes if the Cadmium
Reduction method were used. Since the column can be stored,
provided it is not allowed to dry out, small numbers of samples
are best carried out via this procedure.
5. Electrode Method
This procedure is based on the use of a nitrate electrode and
the measurement of a potential developed across the membrane
when the latter is immersed in a sample or standard. A pH
3-10
-------
meter with an expanded millivolt scale or a selective ion meter
must be used with the electrode. In addition, a double
junction reference electrode is needed.
The electrode must be given time to stabilize after being
introduced into the sample. One to two minutes, with constant
stirring, is usually sufficient.
There are a number of materials and circumstances that can
cause difficulties in obtaining the true potential reading.
However, the procedure calls for the use of a buffer which was
developed for use with the nitrate electrode. When added in
equal proportions with the sample, this buffer alleviates most
difficulties.
The certifier should be aware that the calibration curve is not
linear; therefore, the analyst must have enough standards to
cover the working range. Standards should be run every ten
samples to ensure that the calibration curve is still valid.
The curve is produced by plotting known concentration values of
standards against the obtained millivolt readings.
B. Fluoride
The NIPDWR's maximum contaminant level for fluoride varies with the
annual average of the maximum daily air temperature. The basis for
3-11
-------
this is the increased consumption of water in higher temperature
locations. The values are listed in Table II.
1. SPADNS Method
The SPADNS method is a colorimetric method based on the loss of
color resulting from the reaction of fluoride with the
zirconyl-SPADNS dye. Several of the interfering substances in
the SPADNS procedure are materials found in use in water
treatment plants. For example, aluminum and hexametaphosphates
are two materials which cause considerable interference. Both
these materials are added to some types of waters as part of
the treatment. Consequently, distillation is specified in the
Federal Register as a preliminary step to the use of the SPADNS
procedure.
The preliminary distillation step separates the fluoride from
other constituents in water by distilling the fluoride as
fluorosilic (or hydrofluoric) acid from a solution of the
sanple, in an acid with a higher boiling point. The procedure
uses a diluted sulfuric acid mixture which is conditioned by
heating it to the same end temperature (180°C) as the sample
will be heated. Precautions in the distillation procedure
include complete mixing of the heavy sulfuric acid and the
aqueous sample. Failure to do this could result in the entire
mixture splattering out of the distilling flask and injuring
the analyst and/or the equipment.
3-12
-------
TABLE II
Annual Average of Maximum
Daily Air Temperatures
Maximum Fluoride
Level-mg/liter
Degrees F
Degrees C
53.7 and below
12.0 and below
2.4
2.2
2.0
1.8
1.6
1.4
53.8 and 58.3
58.4 to 63.8
63.9 to 70.6
70.7 to 79.2
79.3 to 90.5
12.1 to 14.6
14.7 to 17.6
t7.7 to 21.4
21.5 to 26.2
26.3 to 32.5
Heating the distillation flask should be stopped inmediately
upon reaching the end temperature of 180°C. Proceeding beyond
this point will allow sulfates to be carried over. Sulfates
are another positive interference. In this same case, the
flame under the distillation flask must never touch the sides
of the flask above the liquid level. Superheating of the vapor
may result ip high sulfate carryover.
When high-fluoride samples are distilled, repeat the
distillation using 30 mL of distilled water. An analysis of
the distillate will indicate completeness of the fluoride
recovery. If substantial amounts of fluoride appear in the
second distillate, add this quantity to that obtained initially
and flush again. Quantities of less than 0.03 mg/300 mL may be
disregarded.
Because of the simplicity of apparatus and procedure, the
distillation procedure can be readily automated. However, the
final heating temperature and sample volumes should be adhered
to. Some manufacturers utilize different temperatures and
3-13
-------
volumes for their automated stills for fluoride; however,
unless the automated still uses the approved temperatures and
volumes of both sample and reagents, it cannot be used as an
accepted procedure. Automatic stills are generally designed
for samples whose fluoride content is within a narrow range
from day to day. Should this range be surpassed the results of
the distilltion could be in error.
The actual colorimetric procedure is sinple and requires only
minor precautions. The most important of these is that the
amount of reagent (10 mL) be very carefully added, since the
fluoride concentration is measured as a difference of the
absorbance in the blank and the sample, and a small error in
the reagent addition is the most prominent source of error.
After the reagent has been added, mix throughly to assure a
homogeneous mixture.
Since almost all methods covered in this unit prior to this
point have been colorimetric in nature, some points on good
techniques should be pointed out for the certification officers
benefit. When standard curves are utilized it is wise to check
two points of the curve with each batch of samples to assure
that the standard curve is still valid. Also check for use of
matched cells. Proper maintenance of the equipment is another
point that will indicate to the Certification Officer that
proper technique is being followed.
3-14
-------
2. Electrode Method
The procedure is based on the use of a lanthanum fluoride
crystal and the measurement of a potential developed across
this crystal by fluoride ions. There are a number of materials
and circumstances that can interfere with the measurement of
this potential. However, the procedure calls for a special
buffer developed for use with the electrode. When added in
equal proportions with the sample, this buffer deletes most of
these interferences and eliminates the need for distilling the
potable water sample.
The electrode must be given time to stabilize after being
introduced into the sample. Usually a three minute time with
constant stirring is sufficient. However, at concentrations
under 0.5 mg/liter F~, up to five minutes may be needed to
achieve a stable meter reading; higher concentrations stabilize
more quickly.
A pH meter with an expanded mv scale or a selective ion meter
must be used with a selective ion electrode. In addition, a
single junction sleeve type reference electrode, is needed. A
magnetic stirrer and stirring bar is a good piece.of equipment
to have since the sample must have constant agitation while the
electrode is immersed.
3-15
-------
3. Zirconium Eriochrome Cyanine R
This procedure, like the SPADNS method, requires an initial
manual distillation procedure. See pages 3-9 to 3-12 for the
procedure and comments.
Since the quality of Eriochrome Cyanine R. varies with the
source, it is necessary to test the reagent each time it is
prepared. The individual absorbance curves show corresponding
differences, and the sensitivity of fluoride between reagents
may differ by 20 percent.
As many interferences are the same as with the SPADNS reagent,
up to lOmg/L of Aluminum can be tolerated by allowing the
solution to stand for two hours before reading. Sulfate
interference is more drastic and special care must be taken
with the distillation procedure so as not to have sulfate
carry-over. If particularly high concentrations of sulfate are
present, precipation as barium sulfate may be necessary.
The procedure shows a salt effect when dissolved solids are
present at concentrations of 10,000 mg/L or greater.
Sensitivity may be depressed by as much as 5-10 percent.
Distillation should prevent this.
3-16
-------
Automated Alizarin Blue (Complexone)
These procedures require the use of automated equipment such as
the Technicon AutoAnalyzer or equivalent. This equipment will
usually only be found in the larger laboratories because of
cost. A fair degree of operator skill and knowledge together
with adequate instructions, are required for successful
automated analysis.
When produced during the reaction, sample color and turbidity
can escape detection of the analyst consequently causing
interference. Therefore, some sample knowledge and regular use
of standards should be recommended.
Automated Electrode Method
When run at forty samples per hour the equilibration time for
the electrode may be surpassed. Consequently an error of 1.5%
of full scale may be expected when a low sairple succeeds a high
sample and an error of 3% of scale may be expected when a high
sample succeeds a low sample. A slower rate of analysis will
produce less error. This procedure utilizes the electrode and
an "AutoAnalyzer" or its equivalent.
3-17
-------
C. Corrosivity
On August 27, 1980, the National Interim Primary Drinking Water
Regulations were revised and paragraph 141.42 was added to require
special monitoring for corrosivity characteristics. The monitoring
and reporting requirement is limited to one year. The purpose was
to identify those public water systems which were distributing
corrosive waters.
The Langelier Index and the Aggressive Index are used to indicate
the corrosive tendencies of a water. In order to calculate these
values, the temperature, calcium, alkalinity, and total filterable
residue values must be known.
The methods for measuring alkalinity and total filterable residue
are covered in this section. Only the EDTA titration procedure for
calcium will be covered here; the atomic absorption procedure will
be covered in the metals section.
1. Total Alkalinity: Methyl Orange Titrimetric or Potentiometric
Method.
This procedure involves titration of the sample with 0.02 N
sulfuric acid to an end pH of 4.5 as determined by a color
change of methyl orange indicator or by use of a pH meter.
3-18
-------
The sample should not be filtered or altered in any way,
including by dilution or preservation. The sample may be
cooled to 4°C for preservation.
Color and turbidity in the sample can cause error by making it
difficult to detect the proper end point color. It is also
important that the same volume of methyl orange indicator be
added to each sample.
Soaps, oily matter, suspended solids, and precipitates may
interfere. These materials may have an effect on the color
change or cause a sluggish response of the electrode.
The pH meter should be standardized at pH 4.0 to 4.6 with
buffers. Frequent standardization checks should be made and
care taken to insure that the electrodes are in good condition.
For both procedures, laboratory control samples should be
analyzed before each run to assure proper operation. The
sulfuric acid should be standardized frequently, preferably
before each use.
2. Total Filterable Residue: (Total dissolved solids)
This method is utilized to determine the corrosivity value for
drinking waters. When evaluating this method it should be
3-19
-------
determined that the laboratory has a copy of an approved method
reference and the proper equipment. Ideally one person should
be assigned to carry out this method, since the best values are
due to technique, and repeated use of the method would lead to
the best values.
Equipment needed for this method would include a good four-
place analytical balance. It should be determined if the
laboratory has a service contract on the balance and a set of
class S weights to check calibration of the balance. Some type
of documentation, such as a notebook, should be available to
show routine checks of the balance calibration. In addition, a
good drying oven and desiccator should be available.
When reviewing the data from this method there should be an
indication that the analyst is weighing until a constant weight
is achieved. This will indicate that no water is left in the
sample.
Laboratory quality control samples should be analyzed each time
the method is used to indicate the method is in control.
3. Calcium - EDTA Titrimetric
This procedure involves the titration of calcium with EDTA at a
pH of between 11 or 12. The main indicator for this procedure
3-20
-------
is Murexide, however, for some the end point is very difficult
to detect. Therefore, other indicators specifically designed
for calcium are used.
Magnesium is the major interference, but at a pH above 11
magnesium is precipitated as the hydroxide. For samples with
high hardness, a small sample (less than 50 mL) should be
analyzed to prevent the pH from changing during the titration.
For samples that are preserved with acid, more NaOH may be
needed to ensure that pH is 11 to 12.
Laboratory control samples containing magnesium and calcium
should be used to determine the efficiency of the indicator.
III. METHODS FOR METAL ANALYSIS
This grouping will cover the following 14 contaminants.
MCL, mg/L
1. Arsenic (As)
2. Barium (Ba)
3. Cadmium (Cd)
4. Chromium (Cr)
5. Lead (Pb)
0.05
0.05
0.010
0.05
1
6. Mercury (Hg)
0.002
3-21
-------
MCL, mg/L
7.
Selenium (Se)
0.01
8.
Silver (Ag)
0.05
9.
Sodium (Na)
—
10.
Calcium (Ca)
—
11.
Copper (Cu)
1
12.
Iron (Fe)
0.3
13.
Manganese (Mn}
0.05
14.
Zinc (Zn)
5
The first eight elements are primary contaminants. For safeguarding
public health, they are required to be monitored in community water
supplies. Each has an MCL which is expressed in concentration and should
not be exceeded. Sodium must also be monitored to protect public health,
but it does not have an established MCL. Calcium is monitored as a
parameter in the measurement of corrosivity while the remaining four
elements are secondary contaminants and are monitored for aesthetic
qualities. The secondary contaminants also have MCL's, and although not
enforceable, it is recommended that they be analyzed at the same interval as
the primary contaminants.
A total element measurement is required for the analysis of these
contaminants. Except for the arsenic colorimetric method and the calcium
EDTA titrimetric method, all other methods are atomic spectrometric
procedures. To better understand the analytical approach and use of these
spectrometric methods, they will be discussed according to technique.
3-22
-------
The different techniques can be grouped into two categories: 1)
techniques (like colorimetric methods) that always require a chemical
reaction or pretreatment sample digestion step prior to analysis and, 2)
those considered to be total element techniques that are of sufficient
energy to vaporize the analyte into an atomic state for either atomic
absorption or atomic emission. Included in the first category are atomic
absorption techniques like flameless cold vapor (CV), chelation-extraction
(CE), and gaseous hydride (GH) methods. The second category consists of
direct flame aspiration atomic absorption (FLAA), graphite furnace atomic
absorption (GFAA) and inductively coupled plasma atomic emission
spectroscopy (ICP). The following sections provide a general discussion of
the techniques, sample preparation, and specific requirements particular to
the individual contaminants.
A. Labware Cleaning, Reagents and Standards
The analysis of trace elements in drinking water necessitates the
labware be kept free from contamination. Care should be taken to
ensure that labware, whether glass, plastic or Teflon has been
thoroughly washed with detergent and tap water; rinsed with 1:1
nitric acid, tap water 1:1 hydrochloric acid, tap water and finally
deionized distilled water in that order.
3-23
-------
Deionized distilled water equivalent to or of better quality than
ASTM Type II reagent water should be used in the preparation of all
reagents, calibration standards, and as dilution water. All acids
should be analyzed to ensure no contaminants are present. The use
of redistilled acid or acids of ultra-high purity grade are
preferred.
The directions for preparing stock standards and calibration
standards are given in each method. When different starting
materials are available, the preferred order of selection is high
quality pure metals, oxides and stable salts. Compounds which are
hydroscopic or can vary in waters of hydration should be avoided
when possible. Purchased standard solution may also be used, but
in all cases, whether purchased or prepared, the concentration of
the calibration standard should be verified with a standard
obtained from an independent outside source (Quality Assurance
Branch, Environmental Monitoring and Support Laboratory -
Cincinnati).
B. Sample Preparation for Total Element Analysis Techniques
Irrespective of the valence state or chemical species, the term
"total" refers to the sum of the elemental concentration in the
dissolved and suspended fractions of a sample. The sample is not
filtered, but immediately preserved at the time of collection with
nitric acid to a pH of less than 2. When suspended material is not
3-24
-------
present, as in most finished drinking water, the contaminant
concentration in solution or the dissolved fraction of an acid
preserved sample will equal a "total" element determination.
Under these circumstances, a direct analysis without acid digestion
using a total element technique will provide a "total" element
analysis. The technique most frequently used for this analysis is
GFAA. Although pretreatment may be omitted, compliance with the
other analytical steps, as noted in each GFAA method, is still
required.
The same situation would not be true for gaseous hydride methods,
atomic absorption chelation-extraction, colorimetric methods, or
the cold vapor technique because the "total" element concentration
must be chemically converted to a particular species before the
analysis can be completed. In each of these cases, the sample
preparation procedure is either given or referenced in the method.
While GFAA is recognized as being the most sensitive technique, an
improved sensitivity can be achieved for FLAA and ICP if the sample
is concentrated by evaporation prior to analysis. This is usually
accomplished during sample preparation using an approved procedure
with concern given to the amount of dissolved solids in the
analysis solution. For FLAA, dissolved solids should be limited to
5000 mg/L and the analysis completed using a high solids burner.
In ICP, the calcium concentration of the analysis solution should
be limited to 400 mg/L, or analysis by method of standard additions
may be required.
3-25
-------
If a sample contains suspended or particulate matter, a
pretreatment acid dissolution step is required prior to analysis.
The purpose of the acid dissolution step is to solublize that
portion of the contaminant that may be occluded or adhering to the
suspended material. Included in the FLAA methods are two acceptable
sample preparation procedures. One is a vigorous nitric acid
11
digestion/ * ' while the other is a total recoverable acid
solubilization procedure using a mixture of nitric' and hydrochloric
{3 4}
acids.' ' The total recoverable procedure is preferred for
drinking water analysis because there is less chance of losses from
volatilization, the formation of insoluble oxides or occlusion in
precipitated silicates. The EPA atomic absorption procedures also
describe sample preparation procedures for 6FAA where only nitric
(1 3)
acid is used and the hydrochloric acid is omitted/ ' ' Certain
elements (arsenic & selenium) have specific preparation procedures
for GFAA and these are described under the individual element
method.^
To determine if an acid dissolution step is needed for GFAA, an
aliquot of a well mixed acidified sample should be transferred to a
Griffin beaker and visually examined. If the solution is free of
suspended or particulate matter, the acid pretreatment step for
GFAA analyses is not needed.
3-26
-------
If drinking water is to be analyzed by ICP using the Appendix to
EPA Method 200.7,the following modified "total"
recoverable acid solubilization procedure must be used for the
preparation of aj_l_ samples.
Sample Preparation Procedure for ICP Analysis of Drinking Water
From a well mixed acid preserved sample, transfer of 200 mL
aliquot to a Griffin beaker. Add 1.0 mL of (1+1) HN03 and
5.0 mL (1+1) HCL to the sample and heat on a steam bath or hot
plate until the volume has been reduced to near 20 mL making
certain the sample does not boil. Allow the sample to cool,
transfer to a 50 mL volumetric flask, dilute to the mark with
deionized-distilied water and mix. The sample is now ready for
analysis. (Note: If after preparation the sample contains
particulate material, an aliquot should be centrifuged or the
sample allowed to settle by gravity before aspiration.)
It is important to note, irrespective of the method or technique, that
once the sample is ready for analysis the calibration standards should
be prepared using the same type of acid or acids, and at the same
concentrations as will result in the sample to be analyzed either
directly or after processing.
3-27
-------
C. Atomic Absorption Techniques
The following techniques can differ in sensitivity and length of
analysis time, but each offers a particular advantage depending on
the chemistry of the contaminant and the concentration required to
be measured. Some contaminants can be analyzed by more than one
technique. Those circumstances along with a preferential ranking
of acceptable methods are addressed in the discussions of each
individual contaminant.
1. Description of Instrumental Needs
Although the design of the atomic absorption spectrophotometer
(whether single or double beam) is not specified, the
instrument should be properly optimized prior to the analysis
of any samples. If the instrument is converted to different
set-ups, for example graphite furnace determinations, the
optimization can be especially important. The burner or
atomizer should be properly aligned, the lamp properly
positioned, the wavelength peaked and the proper slit width
used.
There are two pieces of accessory equipment important for use
with the atomic absorption. First, a recorder is
recommended. This will provide a graphic record of each
3-28
-------
sample analyzed. Should any questions arise on the results of
the runs, the QC samples will give proof of proper technique
when their peaks on the recorder chart are examined.
Secondly, the atomic absorption should have a vent installed
about 15 to 30 cm (6 to 12 inches) above the burner
compartment to remove fumes and vapors from the flame or
graphite atomizer. This protects the laboratory personnel
from toxic vapors and protects the instrument from corrosive
vapors.
Except for the cold vapor analysis of mercury, gases are used
in all other atomic absorption techniques as either a flame or
an inert atmosphere for GFAA. For FLAA two gases are needed
in addition to a supply of dry filtered air. They are
acetylene and nitrous oxide (which is used exclusively for
barium analysis). For GFAA analyses, argon is recommended
because it provides greater sensitivity than nitrogen and can
be used for the analysis of all elements. However, nitrogen
is slightly less expensive, but should not be used in the
analysis of chromium because of possible cyanogen band
interference. For gaseous hydride analyses, an argon-hydrogen
or nitrogen-hydrogen flame is used.
3-29
-------
The two light sources used in atomic absorption are hollow
cathode and electrodeless discharge lamps. The hollow cathode
lamps may be purchased as a single or multi-element lamp. The
single element lamp is preferred because of its greater
emission intensity and improved signal to noise ratio. As
hollow cathode lamps age they may become noisy or weak. A
noisy lamp will give imprecise data, while a weak lamp will
require more gain on the photomultiplier tube and have a
smaller linear concentration range. Electrodeless discharge
lamps provide greater output energy and are longer lasting
than hollow cathode lamps. Their main disadvantages are a
separate power supply is needed and they have a longer warm-up
time. Electrodeless discharge lamps are highly recommended
for arsenic and selenium analyses.
When GFAA techniques are employed the instrument should be
equipped with a background correction device. Its purpose is
to correct for additive non-specific absorption inter-
ferences. If a deuterium source or a tungsten halide lamp is
employed it should be properly aligned and its output energy
should be set equal to the output energy of the hollow cathode
lamp. The use of the Zeeman and Smith-Hieftje background
correction techniques is also acceptable and does not require
application for a method variance.
3-30
-------
The use of an automatic sampler for GFAA is strongly
encouraged. They are more reliable in providing an accurate
furnace aliquot which improves both precision and accuracy.
Either direct injection sanqjlers or aersol spray units are
acceptable. However, if an automatic sampler is not available
all pipet tips should be washed to remove any possible
contamination. This is usually done by soaking in 1:5 HNO^.
For GFAA analyses either pyrolytic or both pyrolytic and
nonpyrolytic graphite tubes must be available. Only pyrolytic
graphite can be used for the analysis of barium while all
other contaminants can be analyzed using either type of
graphite. Although the pyrolytic tube is longer lasting and
usually produces a signal with greater sensitivity, the
optimum analytical concentration range is less than with
nonpyrolytic graphite. Because of the increased sensitivity
samples may require dilution before analysis and a
multi-standard second order regression plot may be necessary
to provide an adequate calibration curve.
When pyroltyic graphite first became available, contamination
and erratic data presented real problems. The coating process
has since improved and although nonpyrolytic graphite is
recorranended for analysis of many elements, the choice of
graphite remains at the discretion of the analyst.
3-31
-------
In GFAA the only restriction on the type of atomizer used is
that it must be graphite. At least one instrument
manufacturer offers a platform insert for the furnace. The
sample is volatilized from the surface of the platform into an
isothermal inert atmosphere that has already reached
atomization temperature. The slight momentary delay in the
vaporization of the analyte results in reduced matix effects.
A method variance is not required for the use of the platform.
For the cold vapor analysis of mercury special accessories for
the atomic absorption spectrophotometer are required. The
accessories for a closed system consist of a absorption cell,
air pump, pump tubing, bubbler, detector, and scrubber. If an
open system is used or the vapor is not collected in a
scrubber, it must be vented to an exhaust hood. The
absorption cell, which is transparent to UV radiation, is
carefully strapped to a burner head and placed in the hollow
cathode light path. The bubbler is inserted into a BOD bottle
to aerate the reduced mercury from solution into the atomic
vapor state for absorption in the cell. Dedicated instruments
designed specifically for the measurement of mercury may be
substituted for the atomic absorption spectrophotometer.
In the gaseous hydride technique, the hydride of arsenic or
selenium is formed in a special glass reaction vessel. After
formation, the hydride, using argon, is swept directly from
3-32
-------
the vessel into either an argon-hydrogen flame or a heated
quartz cell for disassociation and absorption. Each method
specifies the appropriate equipment needed. Commercially
avaiiable hydride systems can be used if they can be operated
using the same procedures and chemistry as given in the
approved methodoiogy.
Finally, when using chelat ion-extraction methods, it is
important for precise and accurate data that the extraction
process be well controlled. When possible, the extraction
should be done using the same volume of liquid in separatory
funnels of the same size. The use of a platform or wrist
shaker is highly recommended for controlling the critical
parameter of shaking rate and length of extraction time.
Although separatory funnels are common to most laboratories, a
mechanical shaker may be considered special equipment.
However, if many extractions are performed its use becomes, a
necessity.
2. Direct Flame Atomic Absorption (FLAA)
Although this technique is the easiest and most convenient to
use, it usually lacks the required sensitivity for drinking
water analysis. Of the primary and secondary contaminants
three elements, arsenic, mercury and selenium, are not
3-33
-------
analyzed by this technique. Of the remaining nine, only
barium, copper, iron and zinc have MCL's at a concentration
that do not require concentrating the sample prior to analysis.
For the analysis of cadmium, chromium, lead, silver and
majaganese it is recommended that the sample be concentrated (<_
10X) by evaporation using a combination acid solubilization
(3 4}
procedure.v ' ' In concentrating by evaporation, attention
must be given to the amount of dissolved solids (<5000 mg/L)
in the analysis solution. Concentrated samples should be
analyzed using scale expansion (10X) and a high solids
(boiling) burner. Background correction should be used unless
it can be verified that nonatomic absorption is not occurring
from the matrix.
The opposite condition is true for the analysis of sodium and
calcium. For these two elements dilution is usually
required. The linear range of FLAA analysis of these elements
is limited and the burner head is normally rotated 45° from
the hollow cathode line of light.
When the use of FLAA is possible, the air-acetylene flame is
used except for the analysis of barium. For barium, the
nitrous oxide-acetylene flame is used which requires a special
burner head.
3-34
-------
The EPA method for chromium specifies the nitrous oxide flame
with the optional use of the more sensitive air-acetylene
flame for concentrations below 200 ug/L. Even with
corvcentrating the sample, the added sensitivity of the
air-acetylene flame is needed for the analysis of chromium in
drinking water. However, concentrating the sample will
increase the calcium and magnesium concentration, which can
cause a possible matrix interference which does not occur in
the nitrous oxide-acetylene. To eliminate that possibility
and the different absorption response of the two chromium
valence states, samples and standards should be prepared to
contain 1% ammonium bifluoride (NH4HF2) and 0.2% sodium
sulfate (I^SO^).
Unlike chromiunm, concentrating the sample does not cause a
noticeable chemical matrix effect in the analysis of silver,
cadmium, manganese and lead. The sensitivity of the lead
analysis can be improved while maintaining accuracy if an
electrodeless discharge lamp is available and the more
sensitive 217.0 nm wavelength is used.
To control ionization in both barium and sodium analyses,
potassium chloride (1000 ug/L) is added to each sample
aliquot, calibration standard and reagent blank. To control
anion interferences in the calcium analysis, lanthanum
3-35
-------
chloride is added to each sample aliquot and calibration
standard alike. If the concentrations of the acids in the
samples and standards are well controlled, the necessity for
the addition of lanthanum is minimized.
Silver in tap water is most probably present as silver
chloride. Acidification with nitric acid to pH 2 will not
solubilize nor prevent the precipitation of silver chloride.
It also will not desorb silver from the walls of the sample
container. However, low concentrations of silver chloride
near the MCL will remain soluble in tap water after the sample
has been acid preserved. In samples where a measurable
quantity of silver is detected, the sample should be made
basic with the addition of ammonium hydroxide prior to
treatment with cyanogen iodide to solubilize the precipitated
silver chloride. Of course this is not done until after all
other analyses have been completed, and the remaining volume
of the sample is measured. The sample container should also
be rinsed and treated in a similar manner to desorb silver
from the walls of the container. The sample concentration
reported from the combined determinations must be based on the
original sample volume collected.
Of the 11 contaminants which can be analyzed using FLAA
methods, FLAA is the preferred atomic absorption technique for
the analysis of calcium, copper, iron, manganese, sodium, and
3-36
-------
zinc. Although the analysis of manganese will probably
require scale expansion or concentration by evaporation, it.is
included with the FLAA group because of the limited analytical
concentration range of the GFAA method and the poor stability
of the manganese chelate. Barium also can be analyzed by
FLAA, but the GFAA method is preferred because of normally
occurring low concentrations. The preferred atomic absorption
technique for the other four contaminants (silver, cadmium,
chromium and lead) is GFAA methodology. If a graphite
atomizer is not available, it is more convenient and accurate
to concentrate the sample by evaporation and compensate for
possible matrix interferences and nonspecific absorption than
complete- the analysis using the chelaton-extraction technique.
3. Graphite Furnace Atomic Absorption (GFAA)
The GFAA technique is two to three orders of magnitude more
sensitive than FLAA. This added sensitivity plus being a
total element technique make GFAA the preferred technique for
the analysis of most primary drinking water contaminants.
As with FLAA methods, when visible particulate or suspended
material are present in the sample, an acceptable digestion or
solubilization procedure must be used. In GFAA analyses, the
addition of hydrochloric acid should be avoided in the
3-37
-------
solubilization procedures. (See Section B-Sample Preparation
for Total Element Analysis Techniques.)
When doing GFAA analyses, method.of standard addition (MSAl
must be followed unless it can be verified it is not
required. Verification is usually accomplished by determining
the percent recovery of a single spike at the MCL or the
midpoint of the calibration curve. MSA is not required if
recovery is within either ± 10 percent or three standard
deviations of the mean value of the spike, whichever is the
smaller. Verification is required on every new and unusual
sample matrix. Further, when the sample analyte concentration
as determined from the calibration curve is within 10 percent
of the MCL or above, the concentration should be confirmed by
reanalyzing the sample by MSA.
In GFAA analyses it is important that the standards and
samples be matrix-matched.. This not only includes any matrix
modifier that is added, but also the type of acid and its
concentration. For most analyses, only nitric acid is used at
a concentration between 0.5 percent and 1 percent (v/v). When
MSA is used, the matching requirement is satisfied because the
standards are diluted in the sample matrix. When it is
verified with spiking that MSA is not required and the sample
concentration is read from the calibration curve, standard
operating procedures, along with records of reagent additions
and sample dilutions, will serve as evidence of matrix
matching.
3-38
-------
Once standards and samples have been prepared and matrix
matched, the same size volume of aliquot should be used for
all furnace injections. If a sample requires dilutions or
concentrating, it should not be done by changing the volume of
the aliquot injected. The reconmended aliquot volume is 20
microliters, with the maximum not to exceed 50 microliters.
f
The composition of the sample matrix and its effect is the
main reason why samples must be analyzed by MSA. Matrix
effects can be reduced and sometimes eliminated with the use
of matrix modifiers. An example is the addition of ammonium
nitrate to a sodium chloride matrix to form both ammonium
chloride and sodium nitrate during the drying cycle. These
compounds are then volatilized during the relatively low
temperature char cycle and thereby eliminating the sodium
chloride interference before atomization.
The individual graphite furnace methods describe procedures
for matrix modification for some of the known interferences.
The recommended matrix modifier used in a given procedure may
not necessarily be applicable to all presently known
interferences. Since the publication of the USEPA methods,
more information on interferences and the use of other
acceptable matrix modifiers has been gained. Although the
approved method should be followed when possible, it is not
intended to restrict the analyst to the use of only certain
3-39
-------
modifiers. The most appropriate modifier should be used for
the type of interference encountered.
The use of a different matrix modifier does not require a
variance, but the analyst should document the purpose of the
modifier and the full range of its effects. An example is the
addition nickel nitrate to arsenic and selenium analyses to
prevent loss during charring, to enhance the signal with the
use of higher atomization temperatures, and in the case of
selenium, lessen sulfate interference.
The use of son® matrix modifiers prevent losses during the
char cycle, but this may not be true in all cases. The
analyst should verify that the instrumental settings or
temperature used are below the critical point where the
analyte could be lost. Also, when relatively low atomization
temperatures are used (< 2000°C), it is advisable to sometimes
use a high temperature burn (2700°C) with continuous purge gas
flow to clean the furnace following atomization. The need for
such a step will depend on the sample matrix and modifier
used, and must be experimentally determined by the analyst.
Two types of graphite are available, pyrolytic and non-
pyrolytic. The user should be aware of the advantages and
disadvantages of each type before using them. (See Section 1
- Description of Instrumental Needs).
3-40
-------
A new graphite tube usually needs conditioning before use.
This is especially true for pyrolytic tubes. This can be
accomplished by injecting an acid blank, drying the liquid,
charring at 500°C for five seconds and atomizing at 1000°C for
three to five seconds. This procedure is successively
repeated five times while advancing the atomization
temperature upward in increments of 300°C. A properly
conditioned tube will result in more precise and accurate data.
Tube life will depend on the atomization temperature required
and the acid concentration of the solution injected. As the
tube ages, a second calibration may be required, but in any
event, either type of tube should last well over 100 burns.
The absorbance response in GFAA is a transient signal and is
usually read and recorded as a peak height response. If the
sample matrix or volatilization characteristics of the analyte
broaden the signal response time, a peak area measurement
would be more appropriate. Either means of reading the signal
is acceptable and selection is left to the discretion of the
analyst.
On some graphite atomizers, the operating conditions can be
changed to increase sensitivity and lower the limit of
detection. Two of these operating parameters are internal gas
flow interrupt and maximum power heating. With gas interrupt,
the internal flow is stopped immediately prior to
3-41
-------
atomization. This still allows the matrix volatilized during
the char cycle to be swept from the absorption path but the
atomized analyte will remain in the chamber for a longer
period of time, giving a higher absorbance reading. The
second option of maximum power heating brings the furnace to
atomization temperature in a'shorter time period than the
normal atomization mode. The rapid heating rate is controlled
by a sensing circuit to prevent atomizer from exceeding the
preselected temperature. When these two operating parameters
are used in conjunction with the platform furnace technique,
matrix effects are reduced resulting in more accurate and
precise data. The instrumental settings of these parameters
are at the discretion of the analysts and a variance for their
use is not required.
For GFAA analysis, the use of a background correcting device
for nonatomic absorption is highly recommended. If the
instrument is not equipped with a background corrector, a
nonabsorbing wavelength must be used. Normally the deuterium
source is only usable up to 350 nm. Above that wavelength, a
tungsten halide lamp, a nonabsorbing wavelength, the Zeeman or
Smith-Hieftje background correction technique is used. Of the
contaminants listed, only the wavelength for barium and sodium
are distinctively above the 350 nm deuterium source cutoff
point.
3-42
-------
Of the contaminants listed, only calcium and mercury do not
have GFAA methods. Although GFAA methods are provided for
sodium and zinc, the ubiquitous nature of these elements and
their sensitivity make these methods very difficult to use.
Therefore, for the analysis of sodium and zinc along with the
analyses of calcium, copper, iron, and manganese FLAA methods
should be used. For the remaining seven elements, GFAA is the
preferred atomic absorption method of analysis with barium
most probably requiring sample dilution prior to analysis.
The following comments particular to individual elements will
be of help in understanding the GFAA methods and evaluating
laboratory practice.
Arsenic and Selenium
Hydrogen peroxide (30 percent) is added to acid digestion to
ensure complete oxidation. A side benefit is that it enhances the
selenium signal and for this reason, it is added to samples not
requiring digestion. The peroxide used should not be stabilized
with tin.
Nickel nitrate is added for the formation nickel arsenide or
selenide during the dry cycle and prevents loss during the high
temperature char. The concentration of the nickel nitrate affects
3-43
-------
the absorbance signal and the slope of the calibration curve. A
0.1 percent nickel nitrate gives adequate enhancement while
serving as a matrix modifier for drinking water. If the sulfate
concentration exceeds 200 mg/U the nickel nitrate in the selenium
analysis should be increased, or the sample should be analyzed by
MSA.
Argon gas and internal gas flow interrupt should be used for both
arsenic and selenium analyses. Past research has shown that
nonpyrolytic graphite gives a higher absorbance than pyrolytic
graphite. It is now reported this phenomenon will vary with the
quality of the pyrolytic coating. In any event, the use of
nonpyrolytic graphite is recommended.
Electrode less discharge lamps should be used for arsenic and
selenium analyses. These lamps provide a more stable signal,
higher intensity, and improved sensitivity over hollow cathode
lamps.
When a 20 microliter injection is used for the selenium analyses,
the MCL is near the low end of the calibration curve. The
absorbance signal can be increased for improved concentration
resolution if a 50 microliter injection is used instead.
Barium
Pyrolytic graphite must be used to reduce the formation of barium
carbide. Maximum purge gas flow should be used for minimum memory
3-44
-------
effects. Argon is preferred to nitrogen to avoid chemical
interferences in the formation of nitrides. Even with the use of
a 20 microliter injection, dilution will probably be required to
keep the sample within the optimum concentration range.
Cadmium
Ammonium phosphate (NH^HPO^ is added to form cadmium
phosphate upon drying and to prevent the loss of cadmium during
the 500*C char cycle. Nonpyrolytic graphite should be used with
continuous internal gas flow to achieve the maximum optimum
concentration range. Even using these conditions, cadmium is only
linear to near 7 microgram/liter. In the past, pyrolytic graphite
gave erratic results and high blank values. Pipet tips that are
yellow in color contain cadmium sulfide and should not be used.
Chromium
Trivalent and hexavalent chromium have slightly different
absorbances for the same concentration. Hydrogen peroxide is
added to the acidified samples and standards to reduce all
chromium to the trivalent state.
Calcium has a suppressive effect on chromium, which increases with
increasing Ca concentration from a 10% suppression at 10 mg/L Ca
to a constant suppression of approximately 20% between 50 mg/L and
3-45
-------
100 mg/L. The effect remains constant from 200 up to 1000 mg/L
Ca. Therefore, calcium is added to both samples and standards to
a level above 200 mg/L.
Nonpyrolytic graphite is preferred for Cr analysis. Using a
nonpyrolytic tube the Cr MCL concentration is near the midpoint of
the calibration curve for the optimum concentration range.
Nitrogen should not be used as a purge gas because of possible
cyanogen band interference.
Lead
Lanthanum nitrate is added to suppress the sulfate interference.
50 mg La added to a 10 mL aliquot of sample will counteract up to
1500 mg/L sulfate interference. It is suggested by other
researchers that magnesium nitrate and ammonium phosphate be used
instead of lanthanum nitrate for both sulfate and
alkali/alkaline-earth chloride interferences. (Atomic Spectroscopy
Vol. 4, page 69, May-June 1983.)
Silver
Silver chloride is formed in tap water that contains both silver
and chloride ions. Preserving the sample by acidifying to pH 2
with nitric acid does not completely solubilize silver chloride,
nor does it prevent its precipitation, nor will it desorb silver
3-46
-------
from the walls of the sample container. If a preliminary
analyses indicates silver to be present, for an accurate
silver analysis, the sample and container must be made basic
using ammonium hydroxide and treated with cyanogen iodide
solution to solubilize the silver. The cyanogen iodide does
not interfere with the GFAA analysis of silver.
4. Chelation/Extraction, Atomic Absorption (CE)
Two CE procedures are approved for the analysis of silver,
cadmium, copper, iron, manganese, lead, zinc, and chromium*6
in drinking water. Both procedures use the diethyldi-
thiocarbamate chelating reagent but each is a different
chemical form. To use either of the CE procedures the sample
must first be taken through a vigbrous sample digestion
procedure to ensure the contaminant is available for
chelation. This must be done even for samples that do not
contain particulate or suspended material. Since trivalent
chromium does not react with the chelating reagent for a total
chromium analysis, all chromium must be converted to the
hexavalent state. This is accomplished using an acid
permanganate digestion.
In the "Standard Methods" procedure, the salt form (ammonium
1-pyrrolidine carbodithioate, CAS Registry No. 5108-96-3)
3-47
-------
commonly called ammonium pyrrolidine dithiocarbamate (APDC) is
added to an aqueous sample at pH 3 and mixed for chelation of
the trace elements. Methyl isobutyl ketone (MIBK) is then
added, followed by 30 seconds of vigorous shaking. The
organic aqueous mixture is allowed to separate and the lower
aqueous layer is drained and discarded. The. remaining MIBK
layer is centrifuged to remove entrained water and then
aspirated into an air-^acetylene flame for analysis. The same
instrument conditions are used as given in the FLAA methods,
except the nebulizer 0-ring is replaced with a cork gasket and
the flame fuel-oxidant ratio is adjusted to compensate for the
added fuel from the organic ketone solvent. MIBK must be used
as the wash solvent between samples.
This procedure not only offers the advantage of concentrating
the analyte for the analysis of lower concentrations, but also
separates the analyte from the major alkali alkaline-earth
ions. Also, the lower surface tension of the organic solvent
results in a higher rate of aspiration which increases
absorption sensitivity.
Although the procedure is useful, it has certain
disadvantages. Of the two procedures, the shaking rate and
time in APDC system is most critical. If the extraction is
prolonged beyond one minute, the extraction efficiency of
3-48
-------
cadmium is reduced, whereas three minutes of vigorous shaking
is required for the complete extraction of hexavalent
chromium. Also, the chelates of silver and manganese are not
stable and should be analyzed soon after extraction. The
small volume of organic layer limits the number of analyses
that can be completed using a single extraction and standards
must be extracted at the time of analysis for the preparation
of the calibration curve.
The second chelation extraction procedure is an EPA method
which also is included in AS7M D—19 methods. This procedure
uses the acid form of the chelate (1-pyrrolidine carbodithioic
acid, CAS Registry No. 25769-03-3), commonly called
pyrrolidine, dithiocarbamic acid (PDCA). In this procedure,
the chelating reagent is formed directly in chloroform by
adding small aliquots of carbon disulfide to the already
dissolved pyrrolidine. This reagent is very stable and can be
used for months if stored in a brown bottle in a refrigerator.
To extract the analytes, the chelating reagent in the
chloroform is added to the aqueous sample which has been
adjusted to pH 2.3. The aqueous organic mixture is vigorously
shaken for two minutes after which the denser chloroform layer
is drained into a Griffin beaker. (Although probably not
required multiple extractions the same procedure is repeated
and the chloroform extracts are combined.) The chloroform is
evaporated and the chelates are destroyed using a nitric acid
3-49
-------
digestion. The digested residue in nitric acid is diluted to
a predetermined volume and analyzed using FLAA methods. The
concentration of the diluted extract is read from the
calibration curve. A check standard which is extracted along
with the samples is used to determine the percent extraction
efficiency and used in calculating the final sample
concentration.
The PDCA extraction has some distinct advantages over the APDC
extraction. First, there is better separation between the
organic and aqueous layers and lower water entrainment in the
organic solvent. Second, chloroform being more dense is the
bottom layer in the separatory funnel and more convenient to
collect. Third, samples do not have to be analyzed the same
day they are extracted, but can be held in dilute nitric acid
solution. Finally, since a FLAA aqueous calibration curve is
used, the volume of the standards is not limited and instru-
ment changes to accommodate the organic solvent are not
required.
For the analysis of chromium, an acid permanganate digestion
step also must be included.
3-50
-------
5. Gaseous Hydride Atomic Absorption (GH)
There are two acceptable methods that use the GH technique for
the analysis of arsenic and selenium in drinking water.
Organic forms of arsenic and selenium must be converted to
inorganic compounds and organic matter must be oxidized before
beginning the analysis. Certain metals may cause
interferences with these methods but their concentrations in
potable water are not normally high enough to cause
interferences. The working range of these methods is 2 to 20
yg/L.
The first method which is described in all four reference
texts is commonly referred to as the zinc hydride method. In
this method for the analysis of arsenic the sample is
subjected to vigorous nitric-sulfuric acid digestion which is
taken to fumes of sulfur trioxide. After cooling, the sample
+3
is reduced to arsenic in hydrochloric with the additions
of potassium iodide and stannous chloride and diluted to a
predetermined volume. Zinc metal is added to the sample to
generate arsine which is swept into a heated quartz cell or
low temperature argon-hydrogen flame for dissociation and
absorption.
It should be noted that the digestion procedure given in
Standard Methods differs from the EPA, ASTM, and USGS
3-51
-------
digestion procedures in that a perchloric acid digestion step
is included. It is recommended that unless the laboratory is
equipped with a perchloric acid hood the perchloric step
should be eliminated for reasons of safety.
The second GH method for arsenic is similar to the zinc
hydride method except sodium borohydride is used in place of
zinc to generate the arsine. Of the approved references, only
the ASTM text describes this optional procedure. Sodium
borohydride methods are listed in other sources, but only the
ASTM method is approved. This method is preferred to the zinc
hydride method because the sodium borohydride is chemically a
cleaner reagent, easier to handle as an additive and results
in more precise and accurate data.
The zinc hydride method also can be used for selenium
analysis, but methods are given in only two of the texts,
Standard Methods and the EPA manual. The described methods
are the same as that for arsenic except the addition of
potassium iodide is omitted.
The USGS hydride method for selenium differs from the zinc
hydride method in the type of digestion, reduction and hydride
generation procedure that is used. In the USGS method the
sample is made acid with the addition of hydrochloric acid.
3-52
-------
Calcium chloride and potassium permanganate are then added and
the sample is heated to boiling while maintaining a purple
tint. After the selenium has been oxidized to the hexavalent
state with boiling permanganate, the sample is cooled, made
basic with sodium hydroxide and evaporate to dryness. The
formation of calcium selenate prevents the loss of selenium
during evaporation. The residue is redissolved in
hydrochloric acid and airaronium chloride, the selenium is
+4
reduced to selenium and the solution diluted to a
predetermined volume. Stannous chloride is added for final
reduction and the hydride is swept from the solution with
nitrogen into the hydrogen-nitrogen flame or heated quartz
absorption cell for analysis.
The ASTM selenium hydride method is a combination of the USGS
digestion procedure and sodium borohydride reduction.
Although the addition of sodium borohydride requires a well
controlled and reproducible technique, it provides a faster
response time and higher absorbances than the stannous
chloride reduction. The result is data with improved
precision and accuracy which makes the ASTM selenium method,
the hydride method choice.
3-53
-------
6. COLD VAPOR ATOMIC ABSORPTION (CV)
The determination of mercury is carried out by reducing the
various forms of mercury in the sample to the elemental state
through a chemical reaction. The mercury vapor formed is then
aerated from solution through a cell positioned in the light
path of an instrument which will measure the absorption. The
method does not utilize the flame of the atomic absorption
instrument and, consequently, is referred to as the cold vapor
technique. (See Section 1 - Description of Instrumental Needs)
In order to assure complete conversion of all forms of
mercury, including methylmercuric chloride care should be
taken to follow the procedure in the Methods for Chemical
Analysis of Water and Waste EPA-600/4-79-020. This procedure
has included extra steps, including a two-hour permanganate/
persulfate digestion which is carried out in a water bath at
95° C.
It is recommended that whatever type instrument is used, a
conventional atomic absorption or an instrument specifically
designed for mercury, that a recorder be attached to it. A
good quality 10 mV recorder with high sensitivity and fast
response time is needed to record the peaks resulting from the
determination of mercury.
3-54
-------
The flow system which precedes the instrument may be used in
either an open or closed mode. The closed mode allows the
mercury vapor to continuously pass through the flow system
until the operator traps or vents the vapor. The open system
allows one pass through the cell and then the vapor is trapped
or vented. Two points should be made here: First, that the
calibration of the instrument is dependent on the volume of
the flow system and should be done only after the system has
been configured and assembled in the same manner that will be
used in the analysis of samples. Second, mercury-vapor is
extremely toxic and safety precautions should be practiced to
protect the operator.
Certain volatile organic materials may interfere in the
determination of mercury. If this is expected to occur, the
sample should be analyzed by the regular procedure and again
under nonreducing conditions, that is, without the addition of
stannous chloride. The true mercury concentration can then be
obtained by subtracting the two values.
D. Inductively Coupled Plasma Atomic Emission Spectroscopy (TCP)
Like FLAA and GFAA, ICP is also a total element technique. It
operates on the principle of atomic emission which is the exact
opposite of atomic absorption. In atomic emission, excited atoms
3-55
-------
returning to lower energy levels or the ground atomic state emit
characteristic energy which can be measured at discrete
wavelengths particular to each element. The amount of energy
emitted is directly proportional to the concentration of the
element.
To produce the emitted energy the atoms must first be excited. In
ICP this is accomplished by generating a high temperature plasma
of metastable argon atoms that collide with the analyte atoms,
transfer energy, and raise the analyte atoms to a higher level of
energy or to the ionic state. Power to generate and sustain the
plasma comes from a radio frequency (R.F.) generator which is
usually operated at a frequency of 27.1 megahertz and at a power
output between 0.9 to 1.2 kilowatts (equivalent to a small radio
station). The power is transferred to the plasma through an
induction or work coil which serves as the antenna of R.F.
generator. Inside the coil is a quartz tube assembly called the
torch. The plasma is formed when the argon gas flowing through
the torch is seeded with electrons from a Tesla coil and becomes
conductive in the magnetic field that surrounds the coil.
The sample reaches the plasma as small aerosol droplets from the
nebulizer. As the aerosol enters the bottom of the plasma it
passes through a central channel where it is desolvated,
dissociated and made available for excitation.
3-56
-------
To record the energy emitted, the charactertic radiation from the
plasma is focused on the entrance slit of the spectrometer. As
the light passes through the slit it assumes the image of the slit
and strikes the diffraction grating. The imaged radiation is
diffracted into line spectra of characteristic wavelengths and
dispersed onto either a single photomultiplier tube (PMT) or
series of tubes placed on the focal curve of the spectrometer. An
exit slit is placed in front of the PMT to define the bandpass of
energy that strikes the PMT. Photocurrents from the PMT are
processed and controlled by a computer system and translated into
concentration.
In addition to the major components of the system (R.F. generator,
spectrometer, and dedicated computer) there are other hardware and
environmental requirements. Since acid and toxic fumes exhaust
with the plasma, the plasma compartment must be vented similar to
FLAA. This is especially important to those instruments where
front surface optics are exposed to these vapors.
The optical train and electronics are sensitive to both
temperature and humidity. Therefore, the environment in which the
instrument is placed should be well controlled, with constant
temperature and humidity, to ensure accurate determinations.
3-57
-------
Argon gas for the plasma may be supplied from a high pressure
cylinder containing dry argon or as a vapor from a tank of liquid
argon. The liquid argon supply is preferred to dry argon because
a tank will last up to three weeks, while a cylinder of dry argon
will last less than eight hours.
The nebulizer used to generate the analyte aerosol should be
rugged, not subject to clogging from dissolved solids, easy to
clean, and if adjustable, easy to align and give reproducible
as pi r ati on.
A peristaltic pump should be used to deliver the sample solution
to the nebulizer. The pump rate should be.set equal to or
slightly less than the aspiration rate of the nebulizer.
Spare parts which should be kept on hand are: a power tube for the
RF generator, a quartz torch, pump tubing, PMT's that cover the
full wavelength range, and computer disks and paper.
The phenomenon of atomic emission provides certain advantages over
atomic absorption.
1. Instruments can be designed and configured as either a
sequential or simultaneous system for multielement analyses.
2. ICP has similar or better detection limits than FLAA while
providing a linear analytical range equal to four orders of
3-58
-------
magnitude. This is attributed to the excellent signal to noise
ratio of the ICP, and the signal transparency of the sheathing
argon gas that surrounds the central analytical channel of the
plasma and cools the quartz torch. Self absorption is greatly
reduced, increasing linearity, because the analyte atoms for the
most part are confined to the central channel. Although the
extended linearity is not needed for the analysis of the primary
and secondary contaminants in drinking water, it is useful in the
analysis of the major constitutents (calcium, magnesium, sodium,
and potassium) for determining water quality.
3. The high temperature of plasma 6000°C to 8000°C makes it almost
totally free of chemical interferences. Although chemical
interferences are not severe in drinking water analysis, as noted
in earlier discussions, they are present in some FLAA and QFAA
methods.
4. If the total dissolved solids do not exceed 1500 mg/L, physical
interferences are not usually a problem in ICP analysis of
drinking water. The use of a peristaltic pump in the sample
introduction system can eliminate fluctuations in aspiration, can
be used to regulate the solution flow uptake rate, and provide a
uniform sample delivery to the nebulizer. If more sample is
pumped to the nebulizer than is needed, excess sample is drained
3-59
-------
to waste. The pump rate is usually set to slightly starve the
nebulizer to achieve the optimum conditions of maximum precision
with adequate intensities.
An operational disadvantage of ICP can be spectral interference.
Many elements are spectrally rich and emit energy at many
different wavelengths. Many of these wavelengths are far less
intense and not used for analytical purposes, but may overlap
and additively interfere with the more sensitive wavelengths used
for analysis. The separation between wavelengths and how well
they are resolved is a function of the grating and the design of
the optical system. The intensity of an interfering wavelength is
a function of the interfering element concentration. Fortunately,
in drinking water analysis spectral overlap interference is not
usually a problenu The analytes are of low concentration and the
major constituents (calcium, magnesium, sodium and potassium) have
relatively few lines and most are of low intensity.
A second type of spectral interference that can occur is when the
sample matrix causes a shift in background intensity. Correction
for this type of interference is accomplished by alternately
measuring the background intensity adjacent to the wavelength peak
during analysis. For this type of correction to be valid the
location selected for reading the background intensity must change
in the same manner and degree as the background intensity under
the wavelength peak.
3-60
-------
For operating the ICP instrument, the manufacturer's instructions
should be followed. Prior to calibration allow sufficient warmup
time (30 minutes) for the instrument to become thermally stable.
If the system is a simultaneous instrument, the entrance slit
should be optically profiled before calibration. On simultaneous
instruments the exist slits also must be profiled, but this is
only done at the time of installation and then usually verified
every three months. If an exit slit is found to be out of
profile, it is reprofiled and then checked daily until verified
with certainty that the exit slit profile is correct and stable.
It has been determined that a well controlled and steady plasma
will greatly reduce calibration drift and improve precision. The
use of mass flow controllers on the argon gas supplied to the
plasma has been found to be an excellent means to regulate and
maintain an even, steady gas flow. Although not required, the
purchase of mass flow controllers is greatly encouraged. A
successful routine for reproducing plasma conditions on a
day-to-day basis has been to use a mass flow controllers to adjust
the argon flow rate of the central channel so as to match the
copper/manganese emission intensity ratio to an experimentally
predetermined value. If interelement correction factors are used
to correct spectral interferences, it becomes of particular
importance that the plasma condition be reproducible for the
factors to remain valid.
3-61
-------
Other important instrumental parameters that affect plasma
reproducibility and the intensity of the analyte signal are the
R.F. incident power, proper adjustment of the torch in the work
coil, the height above the work coil where the plasma is observed
and the sample flow rate to the nebulizer. It is interesting to
note most instruments are operated using similar instrumental
conditions. Power is set at or near 1.1 kilowatts with a plasma
observation height of 15 to 16 mm above the work coil. Alignment
of the torch in the coil is a straightforward procedure and the
sample peristaltic pump rate is usually set at 1.2 to 1.4 ml/min.
or just below the normal draw of the nebulizer. Once the
operating conditions have been experimentally determined, they
should be maintained and fine adjustment of the plasma be
accomplished using the mass flow controller.
The most critical parameter of the plasma is the sample argon flow
rate which carries the analyte through the central channel of the
plasma. The concentration of the excited atoms is increased as
the analyte resident time in the analytical zone of the plasma is
increased. The lower the flow rate, the more intense the signal
of the ion wavelengths. The sample argon flow also affects the
penetration and size of the central channel in the plasma. The
final flow rate selected is always a compromise between the
desired intensities of atom and ion lines, but once selected must
be maintained.
3-62
-------
If it becomes necessary to replace the torch and the replacement
is not a close replica of the previous torch the emission
intensities will be different. The change in intensities will not
be uniform across the spectrum and will demand the determination
and calculation of new interelement correction factors.
Standard calibration curves for each element are stored in the
computer and used in determining concentration. These curves are
either verified or a new calibration is performed at the start of
each analysis run. During the analysis, sample data is usually
stored in a computer file for printout of a hardcopy report at the
end of the analysis run. Also, included in most instrumental
software programs are routines for performing some statistical
analysis of the data and other routines such as wavelength
scanning used for determining the presence or absence of spectral
interference.
The recognized ICP method for the analysis of water and wastes is
USEPA Method 200.7.^ This method has been amended with an
Appendix^ which is specific to the analysis of drinking
water. The Appendix is applicable to all listed elemental
contaminants except mercury and selenium. Mercury could not be
included in the Appendix because it is not a listed analyte in
Method 200.7. Also, the ICP detection limit for both elements is
above its respective established MCL's.
3-63
-------
Included in the Appendix is a required sample preparation
procedure for drinking water analyses. This procedure is
described in Section B - Sample Preparation for Total Element
Analysis Techniques. Incorporated in the procedure is a provision
for concentrating the sample 4X before analysis. The purpose of
concentrating the sample is to raise the concentration of the
analyte so that the precision about the measurement will allow the
analyst to determine if the contaminant is in-or-out of compliance
with a 95 percent level of certainty. This provision allows the
contaminants to be determined by ICP analysis with a precision and
accuracy similar to GFAA.
It has been reported that large concentrations of the major
cations (Ca, Mg, and Na) can affect the distribution of the
analyte in the different size aerosol droplets that are formed
during nebulization. This matrix effect is called "aerosol ionic
redistribution" (AIR). Only small droplets enter the plasma
(while large droplets drain to waste), any change in their analyte
concentration so that they no longer reflect the concentration of
the bulk solution being aspirated will result in a biassed
analysis. A 5 percent reduction in the cadmium intensity for a
concentration 4X the MCL has been observed when the concentration
of calcium exceeds 400 mg/L. This suppression increases with
increasing Ca concentration. The matrix effect can be corrected
using method of standard addition (MSA). It is recommended when
the concentration of a primary
3-64
-------
contaminant is determined to be within 10 percent of its MCL or
above, and the calcium concentration in the analysis solution
exceeds 400 mg/L, that the absence of AIR be verified by spiking
or the sample be analyzed by MSA.
Included in Method 200.7 are recommended wavelengths, preparation
procedures for calibration standards, expanded definitions of the
type of interferences, a procedure for MSA, and instrumental
quality control requirements. Included in the Appendix is a
discussion on the requirements of drinking water analysis, tests
and criteria for determining the presence of spectral and matrix
interferences, typical instrumental operating conditions,
mandatory quality control and data reporting requirements.
Although the initial purchase price of an ICP system is 3 to 6
times that of atomic absorption instrumentation, it can provide
cost savings because less time is required to complete the
analysis. Using the Appendix to Method 200.7 the ICP cost for the
analysis of 12 analytes in 20 samples is 1/6 that of GFAA
analysis. For this reason and the other advantages listed
earlier, ICP analysis is the most preferred method of analysis for
all elemental contaminants except mercury and selenium.
3-65
-------
IV. ORGANIC METHODS
Techniques for the determination of organics are relatively new to
most water supplies. Until the passage of the Safe Drinking Water Act,
there were no regulations concerning specific compounds unless the supply
had decided to analyze for these parameters. Analysis for organic compounds
which utilize gas chromatography requires a good deal of experience to
obtain reproductible qualitation and quantitation. The methods for the
organochlorine pesticides, the chlorinated pheno^y acid herbicides and the
trihalomethanes are recommended for use only by experienced analysts or
under close supervision of such qualified persons.
The methods to be used, including extraction and clean-up procedures,
are not included in either of the usual references. However, they are
available from the U. S. Environmental Protection Agency, Environmental
Monitoring and Support Laboratory, Cincinnati, Ohio 45268.
Additional analysis for organics has been required with the
publication of the final trihalomethane regulations on November 29, 1979.
These are well as the pesticide and herbicide information is listed in
Table III.
3-66
-------
A. Organochlorine Pesticides
The compounds as listed in the National Interim Primary Drinking
Water Regulations may be determined by the methods referenced in
the accompaning table. However, special precautions must be taken
to measure the toxaphene peaks. Interferences are experienced
when contamination or other compounds such as polychlorinated
biphenyls, phthalate esters and some, organophosphorus pesticides
are present. The referenced methods covers clean-up and
concentration techniques to be used when interferences can not be
tolerated. The method is recommended for use only by experienced
pesticide analysts or under close supervision of such qualified
persons.
Care should be taken to check such things as the state of
activation of the florisil and the amount of f1 orisi1 being used.
Trial runs are recommended using standard mixtures of the
pesticides to be analyzed. The gas chromatographic operating
conditions should also be checked, via an included procedure, to
determine if acceptable conditions exist. Standards should be
3-67
-------
Table III
Approved Methodology for Organic Contaminants
Contaminant
Methodology
EPA
1
ASTM
SMk
USGC
Chlorinated hydrocarbons
endrin
Solvent extraction, gas
chromatograpy
pp 1-19
D3086-79 509A pp 24-39
lindane
methoxychor
toxaphene
Chlorophenoxys
27*30—
2,4,5-TP
Solvent extraction, derivatization
gas chromatography .
pp 20-35 D3478-79 509B pp 24-39
Total Trihalomethanes Purge and trap-gas chromatography
JiL
Solvent extraction, gas
chromatography
Gas chromatography - mass
spectrometry
(6)
(7) and (8)
1.
2.
3.
4.
"Methods for Organochlorine Pesticides and Chlorophenoxy Acid Herbicides in Drinking Water and Raw
Source Water"*
Annual Book of ASTM Standards, Part 31 Water, American Society for Testing and Materials, 1916 Race
Street, Philadelphia, Pennsylvania 19103
"Standard Methods for the Examination of Water and Wastewater", 14^^ Ed. American Public Health
Association, American Water Works Association, Water Pollution Control Federation, 1975.
Techniques of Water-Resources Investigation of the United States Geological Survey, Chapter A-3,
"Methods for Analysis of Organic Substances in Water", Book 5, 1972. Available from Superintendent
of Documents, U.S. Government Printing Office, Washington, D.C. 20402
3-68
-------
5. "The Analysis of Trihalomethanes in Finished Waters by the Purge and Trap Method", Method 501.1 EMSL,
EPA, Cincinnati, Ohio 45268.*
6. "The Analysis of Trihalomethanes in Drinking Water by Liquid/Liquid Extraction", Method 501.2, EMSL,
EPA, Cincinnati, Ohio 45268.*
7. "Measurement of Trihalomethanes in Drinking Water by Gas Chromatography/Mass Spectrometry and
Selected Ion Monitoring", Method 501.3, EMSL, EPA, Cincinnati, Ohio 45268.*
8. "Measurement of Purgable Organic Compounds in Drinking Water by Gas Chromatography/Mass
Spectrometry", Method 524, EMSL, EPA, Cincinnati, Ohio 45268
~Available from ORD Publications, CERI, EPA, Cincinnati, Ohio 45268
3-69
-------
Table IV
Approved Methodology for Measurements riot Required to be Done in Approved Laboratories
Contaminant Methodology EPA^- ASTM^ . SM^ Other
Free Chlorine Residual
Colorimetric or titrimetric DPD
409E or F
Colorimetric syringaldazine
408G**
PH
Potentionmetric
150.1
D1293-78A or B 424
Temperature
Thermometric
212
Turbidity
Nephelometric
180.1
214A
1. "Methods of Chemical Analysis of Water and Wastes," EPA Environmental Monitoring and Support
Laboratory, Cincinnati, Ohio 45268 (EPA 600/4-79-020), March 1979. Available from ORD Publications,
CERI, EPA, Cincinnati, Ohio 45268
2. Annual Book of ASTM Standards, Part 31 Water, American Society for Testing Materials, 1916 Race
Street, Philadelphia, Pennsylvania 19103
3. "Standard Methods for the Examination of Water and Wastewater," 14*h Ed, American Public Health
Association, American Water Works Association, Water Pollution Control Federation, 1975.
4. "Standard Methods for the Examination of Water and Wastewater", 15^h Edition, American Public
Health Association, American Water Works Association, Water Pollution Control Federation, 1980.
3-70
-------
injected frequently as a check on the stability of the operating
conditions. Care should also be taken to assure good technique both
in the cleaning of glassware and use of the chromatograph are
observed.
B. Herbicides
Phenoxy acid herbicides are used extensively for weed control.
These materials could enter the source supply by run offs and be
taken.into the water supply system. These chlorinated phenoxy acids
and their esters are extracted from acidified water samples with
ethyl ether. The esters are hydrolyzed to acids and extraneous
organic material is removed by a solvent wash. The acids are
converted to methyl esters which are extracted from the aqueous
phase. The extract is cleaned up by passing it through a
micro-absorption column. Identification of the esters is made by
selective gas chromatography. Again, the identification should be
corroborated through the use of two or more unlike columns.
Organic acids, especially chlorinated acids, cause the most direct
interferences with the determination. Phenols, including
chlorophenols, will also interfere. Solvents, reagents, glassware
and other sample processing material may cause problems. All these
should be checked to prove they are free from interferences.
3-71
-------
Some of the reagents are toxic while others can present hazardous
circumstances if improperly handled. Consequently, care must be
exercised when using these materials. This method is recommended
for use only by experienced pesticide analysts or under the close
supervision of such qualified persons.
Quality control is another area that should be carried out with
regularity. This topic will be covered in another instructional
unit.
C. Trihalomethanes
Of the two procedures the liquid - liquid extraction method is the
simplest to carry out and the least expensive. However the sample
must be relatively free from interferences and is applicable only
to the determination of the four trihalomethanes. For compounds
other than these four, confirmation by GC/MS must be provided.
If analysis of raw source water is carried out, the purge and trap
technique must be performed to characterize each raw source water
if peaks appear as interferences.
Confirmatory evidence is obtained using dissimilar columns and
temperature programming. If concentrations are high enough (> 50
yg/L) halogen specific detectors may be employed for improved
specificity.
3-72
-------
Although the purge and trap procedure is applicable to many more
compounds it is more difficult to carry out and is much more
expensive. The halogen specific detector is used to ensure
specificity. This technique should be carried out only by persons
experienced in gas chromatography or under the supervision of an
experienced person.
In both procedures frequent analysis of standards should be
carried out. Quality control is written into the methodology
available from EPA. Traps should be placed in the line before
instrumentation to assure gases do not contribute peaks and the
traps should be changed regularly. These changes should be
recorded as part of the instrument maintenance procedure.
V. Operator Tests
It was the intent of the National Interim Primary Drinking Water
Regulations that all analysis be performed in approved laboratories. There
were two exceptions in the original regulations, the revisions in 1980 added
several more. These tests may be carried out by "any person acceptable to
the state". The list of these methods now includes turbidity, free chlorine
residual, temperature and pH. It is felt that Certification Officers should
be aware of these methods. The approved methodology is listed in Table IV.
3-73
-------
A. Residual Chlorine
1. DPD. This method was originally specified on the basis of the
improved accuracy and sensitivity, particularly when compared
with the o-tolidine procedure in common use. The o-tolidine
was popular because of its simplicity and availability of
field test kits. There are now available test kits designed
around the DPD procedure. There should be no sample
preservation; analyses must be made as soon as possible, or
within one hour. The sample may be taken in either plastic or
glass. Test kits are available from many sources, e.g., Hach
Chemical Co. and LaMotte Chemical Company.
2. Syringaldazine. According to Standard Methods, this method
has been shown to be the most specific colorimetric test for
measuring free available chlorine. The accuracy and precision
are comparable to the DPD method. One manufacturer of this
type kit is Ames Division of Miles Laboratory Inc. in Elkhart,
Indiana.
B. Turbidity
The method is based upon a comparison of the intensity of light
scattered by the sample under defined conditions compared with the
intensity of light scattered by a standard reference suspension.
3-74
-------
Readings in NTU's are made in a nephelometer designed according to
specified specifications. A standard suspension of Formazin,
prepared under closely defined conditions is used to calibrate the
instrument.
The presence of floating debris and coarse sediments which settle
out rapidly will give low readings. Finely divided air bubbles
will affect the results in a positive manner. The instruments
should detect differences of 0.02 units in waters having
turbidities of less than one unit. The instrument should read
from 0 to 40 units. Scales above 40 units are generally not used
except for dilution purposes.
The sample tubes to be used must be of clear, colorless glass.
They should be kept scrupulously clean, both inside and out, and
discarded when they become scratched or etched.
The use of a prepared turbidity standard of styrene divinyl benzene
polymer was permitted in the March 3, 1982 Federal Register Vol.
47, No. 42, Page 8997. This standard was permitted and formazin
standard retained to allow the choice by the analyst. The new
standard may be purchased from Amco Standards International.
3-75
-------
C. Temperature
Temperature measurements may be made with any good grade of
mercury filled Celsius thermometer. As a minimum, the thermometer
should have a scale marked for every 1.0°C or finer. Markings
should be etched on glass. The thermometer should have a minimal
thermal capacity to permit rapid equilibration. It should be
checked against a precision thermometer certified by the National
Bureau of Standards. For field use it is a good idea to provide a
metal case to prevent breakage.
D. pH (electrometric)
The certifier should be assured that the meter and electrodes are
in working order and a sufficient supply and range of buffers' are
on hand. It would be good to see if the instrument manufacturers
manual with operating instructions are available. The accepted
equipment specifications are listed in the outline on Instrument
and Equipment Needs.
VI. SUMMARY
Due to the number of chemical methods that a laboratory will be using,
the laboratory Certification Officer cannot evaluate each one during his
on-site inspection. The results of the required performance sample, as well
3-76
-------
as the information available in the records of the daily quality control,
will indicate problem areas in the methodology. At this point the
certification person can offer assistance in guiding the laboratory toward
correction of the problem.
The inspector will be expected to answer questions from technicians on
the methods and to point out any deviations that may be detected. The
inspector can supply information on equipment and instruct laboratories on
methods should the laboratory decide to utilize a procedure different from
that procedure currently in use.
It is suggested that the inspector schedule ahead of his visit those
tests he may wish to observe. These can be tests which indicate problems,
any test added since a previous visit, or an arbitrary analysis performed
from a specific method.
3-77
-------
REFERENCES
1. Part 4.1.3 of the Atomic Absorption Methods Section, "Methods for
Chemical Analysis of Water and Wastes," U.S. Environmental Protection
Agency, Environmental Monitoring and Support Laboratory, Cincinnati,
Ohio, 45268 (EPA-600/4-79-020), March 1979.
2. Part 301A-I.1 "Standard Methods for the Examination of Water and
Wastewater," 14th Edition, American Public Health Association,
American Water Works Association, Water Pollution Control Federation,
1975.
3. Part 4.1.4 of the Atomic Absorption Method Section "Methods of
Chemical Analysis of Water and Wastes," U.S. Environmental Protection
Agency, Environmental Monitoring and Support Laboratory, Cincinnati,
Ohio, 45268 (EPA-600/4-79-020), March 1979.
4. Methods, D3557-78A (Cd), D1687-77D (Cr), D3559-784 (Pb), and 03859-79
(Ag), "Annual Book of ASTM Standards", Part 31 Water, American Society
for Testing and Materials, 1916 Race Street, Philadelphia,
Pennsylvania, 19103.
5. Methods 206.2 (As) and 270.2 (Se), "Methods of Chemical Analysis of
Water and Wastes" EPA Environmental Monitoring and Support Laboratory,
Cincinnati, Ohio, 45268 (EPA-600/4-79-020), March 1979.
6. "Inductively Coupled Plasma Atomic Emission Spectrometric Method for-
Trace Element Analysis of Water and Wastes," Method 200.7, from
"Methods for Chemical Analysis of Water and Wastes, EPA 600/4-79-020",
U.S. Environmental Protection Agency, Environmental Monitoring and
Support Laboratory, Cincinnati, Ohio 45268, December 1982.
7. "Inductively Coupled Plasma Atomic Emission Analysis of Drinking
Water," Appendix to Method 200.7, U.S. Environmental Protection
Agency, Environmental Monitoring and Support Laboratory, Cincinnati,
Ohio 45268, January 1985.
3-78
-------
Unit 4
SAMPLING
I. INTRODUCTION
Since the intent of the Safe Drinking Water Act is to insure proper
drinking water quality, meaningful analysis of the water is imperative in
order to know if the water meets the standards. This analysis can only be
meaningful if it is performed on a sample that is representative of the
water that the consumer will drink. Consequently, the proper sampling
technique, use of proper containers, proper preservation and adherence to
the set frequency of sampling must be carefully observed.
In many instances the laboratories themselves will not be responsible
for sampling. However, it is necessary that all laboratories be aware of
what constitutes a representative, properly collected sample. It is the
responsibility of all laboratories sampling for parameters under the Safe
Drinking Water Act to call for a resample if the initial one does not meet
proper sampling procedures. To analyze a sample which has been collected
dubiously is to present data which is dubious in meaning. If the laboratory
is responsible for collecting the samples, it is doubly important that the
persons in the laboratory be aware of proper techniques.
The Critical Elements portion of Chapter IV of the "Criteria and
Procedures" Manual spells out the requirements that must be adhered to for
the drinking water sampling. It is attached here for student reference.
4-1
-------
When the laboratory has been delegated responsibility for sample
collection, handling, and preservation, there needs to be strict adherence
to correct sampling procedures, complete identification of the sample, and
prompt transfer of the sample to the laboratory.
The collector should be trained in sampling procedures and approved by
the State regulatory authority or its delegated representative. The sample
must be representative of the potable water system. The water tap must be
sampled after maintaining a steady flow for two to three minutes to clear
the service line. The tap must be free of aerator, strainer, hose
attachment, or water purification devices.
The sample report form should be completed immediately after collection
with location, date and time of collection, collector's name and any special
remarks concerning the sample.
A. Inorganic Contaminants
The type of sample container and the required preservative for each
inorganic chemical contaminant are listed in Table 4-1.
It is essential that all samples be analyzed within the maximum
holding times listed in Table 4-1. Where maximum holding times
cannot be met, the sample is to be discarded and resampling
requested.
4-2
-------
B. Organic Contaminants
The type of sample container and the required preservative for the
organic chemical contaminants are listed in Table 4-1.
When sampling chlorinated waters for TTHM analysis, sodium
thiosulfate or sodium sulfite should be added to the empty sample
bottles prior to shipping to the sampling site.
The TTHM bottles need to be filled in such a manner that no air
bubbles pass through the sample as the bottle is filled. The bottle
is to be sealed so that no air bubbles are entrapped. The hermetic
seal on the sample bottle needs to be maintained until analysis.
It is essential that all samples be analyzed within the maximum
holding times listed in Table 4-1. Where maximum holding times
cannot be met, the sample is to be discarded and resampling
requested.
A footnote to Tables IV-4 and IV—5 of the "Criteria and Procedures"
manual which deals with sampling cautions that, if a laboratory has
no control over sample collection, handling, and preservation, then
it is critical that a laboratory director rejects any samples not
meeting the criteria in the tables and so notify the authority
requesting the analyses.
4-3
-------
MONITORING REQUIREMENTS
A. Inorganic
1. . Analyses for all community water systems utilizing surface
water sources shall be completed within one year following
June 1977. These analyses shall be repeated at yearly
intervals.
2. Analyses for all community water systems utilizing only
groundwater sources shall be completed within two years of
June' 1977. These analyses shall be repeated at three-year
intervals.
3. For noncommunity water systems, whether supplied by surface or
groundwater sources, analysis for nitrate shall be completed
by December 24, 1980. These analyses shall be repeated at
intervals determined by the State.
4. When the maximum contaminant level is surpassed, the frequency
of resample shall be designated by the state and shall
continue until the maximum contaminant level has not been
exceeded in two successive samples or until a monitoring
schedule as a condition to a variance, exemption or
enforcement action shall become effective.
4-4
-------
B. Organic, Other than TTHM's
1. For all community water systems utilizing surface water sources,
analyses shall be conpleted within one year of June 1977.
Samples analyzed shall be collected during the period of the
year designated by the State as the time when contamination by
pesticides is most likely to occur. These analyses shall be
repeated at intervals specified by the State, but in no less
frequency than at three year intervals.
2. For conmunity water systems utilizing only groundwater sources,
analysis shall be completed by those systems specified by the
State.
3. If the results of an analysis indicate that the level of any
contaminant exceeds the maximum contaminant level, the supplier
of water shall report to the State within seven days and
initiate three additional analyses within one month.
4. When the average of four analyses exceeds the maximum
contaminant level the supplier of water shall report to the
State. Monitoring after public notification shall be at a
frequency designated by the State and shall continue until the
maximum contaminant level has not been exceeded in two
successive samples or until a monitoring schedule as a condition
to a variance, exemption or enforcement action shall become
effective.
4-5
-------
C. TTHM's
All community water systems serving a population of more than 10,000
that adds a disinfectant to the water shall analyze for TTHM's. The
minimum number of samples required to be taken shall be based on the
number of treatment plants used by the system.
1. For supplies using surface water as well as those using
groundwaters that have not been determined by the State to
qualify for reduced monitoring frequencies, the sampling
frequency is at least four samples per quarter per treatment
plant. At least 25 percent of the samples shall be taken at
locations within the distribution system reflecting the longest
residence time in the system. The other 75 percent shall be
taken at representative locations in the distribution system.
2. The State has the authority to reduce this monitoring frequency
to a minimum of one sanple analyzed for TTHM's per quarter taken
at a point in the distribution system reflecting the maximum
residence time of the water in the distribution system.
3. Water supplies using only groundwaters may request permission
from the State to reduce monitoring to one sample analyzed for
maximum trihalomethanes potential per year for each treatment
4-5
-------
Table 4-1
Contaminant
Arsenic, Barium, Cadmium,
Chromium, Lead, Selenium,
Silver
Fluoride
Mercury
Nitrate
Chlorinated Supplies
Nonchlorinated Supplies
Sodium-Calcium
Alkalinity
Total filterable residue
Chlorinated Hydrocarbons
Chlorophenoxys
TTHM
2
Preservative
Cone. HNO3 to pH<2
None
Cone. HNO3 to pH<2
Cool to 4°C
Cone. H2SO4 to pH<2
Cone. HNO3 to pH<2
Cool to 4°C
Cool to 4°C
Cool to 4*C as soon as
possible after collection
Cool to 4*C as soon as
possible after collection
Sodium thiosulfate
or sodium sulfite
3 4
Container Maximum Holding Time
P or G 6 months
P or G 1 month
G 38 days
P 14 days
P or G 28 days
P or G 14 days
P or G 6 months
P or G 14 days
P or G 7 days
Glass; with foil or Teflon-
lined cap 14 days5
Glass; with foil or Teflon-
lined cap 7 days5
Glass; with Teflon-lined
septum** 28 days
-------
Table 4-1 (Cont'd)
1. If a laboratory has no control over these factors it is critical that the laboratory director reject any
samples not meeting these criteria and so notify the authority requesting the analysis.
2. If HNO3 is to be used and cannot be used because of shipping restrictions the sample may be initially
preserved by icing and immediately shipping it to the laboratory. Upon receipt in the laboratory, the sample
must be acidifed with conc. HNO3 to pH<2. At the time of analysis, the sample container should be
thoroughly rinsed with 1:1 HNO3; washings should be added to the sample.
3. P = Plastic, hard or soft; G = Glass, hard or soft
4. In all cases, samples, should be analyzed as soon after collection as possible.
5. Well stoppered and refrigerated extracts can be held up to 30 days.
6. All samples are collected in duplicate.
-------
plant used by the system. This sample should be taken at a
point in the system reflecting the maximum residence time of the
water in the system.
0. Unregulated
1. Sodium
With the publication of the amendment to the National Interim
Primary Drinking Water Regulations, special requirements were
set down for monitoring of sodium and corrosivity. Sampling is
required at the rate of one sample per plant at the entry point
to the distribution system. Samples should be taken annually
for systems using surface waters, and at least every three years
for systems using only groundwaters.
No maximum level has been set for sodium. The monitoring was
required to provide data for the medical community when
requiring low sodium intake for individuals.
Allowance is made for States to increase the monitoring
frequency if needed. The supplier must notify the State or EPA
of the concentration found. The goal set was 20 mg/L. This
concentration is a goal and is not enforceable.
4-9
-------
2. Corrosivity
Suppliers of water for community public water systems are
required to collect samples from a representative entry point to
the distribution systems. The frequency is two samples per
plant, one sample during mid-winter and one during mid-summer.
Supplies using only groundwater are required to sample only once
per year. Again the State has the authority to set the
frequency of sampling.
The intent of monitoring for corrosion characteristics is to
ascertain the potential of harmful material entering the water
due to the agressive nature of the water on the pipes and or the
joint material. These materials could be lead, copper, and
asbestos.
The corrosivity is determined by calculation of the Langelier
Index or the Aggressive Index. In order to carry out this
calculation, the content in the water of Total Filterable
Residue, Calcium, Alkalinity, and Temperature are required.
Consequently the laboratories must be able to determine these
parameters.
4-10
-------
. SAMPLE CONTAINERS
Types
Generally two types of containers are acceptable: these are glass and
plastic. Plastic is the more convenient from a shipping standpoint;
however, plastic may not be used for the organic parameters. The glass
containers should preferably be made from a hard borosilicate glass
(Kimax or Pyrex; however, other forms may be used).
All these various materials have certain advantages and disadvantages.
The hard glass is inert to most materials. Conventional polyethylene
is to be used when plastic is acceptable because of reasonable cost and
less adsorption. Disposable type plastic containers, such as the
molded polyethylene "Cubitainer" are convenient to use.
Usually, a wide mouth container is preferred. This allows easy sample
removal and easier cleaning.
The sample containers for the total trihalomethanes should be a 25-mL
capacity or larger screw cap vial sealed with Teflon faced silicone
septa. These TTHM bottles should be the equivalent to the Pierce No.
13075 bottle and the septa equivalent to the Pierce No. 12722.
4-11
-------
The pesticide containers should be a wide mouth container of 1-liter or
1-quart capacity with a screw cap lined with Teflon. If the Teflon
liner is not available, aluminum foil may be used. Amber bottles are
recotranended, but if not available the sample should be protected from
light. The same type container should be used to collect the sample
for the chlorophenoxy herbicides.
Depending on the State or Regional requirements, the following is an
estimate of the number and types of containers needed for a surface
supply excluding those needed for resample.
Inorganics:
One 1 gal - plastic - with HNOg - all metals including sodium, calcium
One 1 gal - plastic - cooled 4°C - Nitrate, Alkalinity, Total
Filterable Residue, Fluoride.
One 1 qt - plastic - with H2S04 - Nitrate (nonchlorinated supplies)
One 1 qt - glass - with HNO-j - Mercury (for longer holding time)
Organics
One 1 qt - glass (Teflon or foil lined cap) - cooled 4°C - pesticides
One 1 qt - glass (Teflon or foil lined cap) - cooled 4°C - herbicides
Thirty-two - > 25 mL screw cap bottles with Teflon faced silicone
septa - sodium thiosulfate or sodium sulfite added - for TTHM's
Additional containers would be needed for resampling, blanks and
duplicates. A noncommunity supply would need one - 1-quart plastic
container per year unless resampling is to be carried out.
4-12
-------
In addition to the sample containers, some type of shipping containers
must be provided for each sample. For those samples which require
cooling a type of ice chest or other insulated container must be
provided.
B. Container preparation and preservation:
A labortory may have the responsibility for providing sample
containers; consequently the Laboratory Certification Officer must
be able to inspect the capability of the laboratory to provide the
correct materials. This responsibility to provide, maintain, and
clean containers is dependent on how the State has elected to carry
out the certification program. The laboratory or the authority
could purchase in large lots and make available sets of containers
to each supply or the State may elect to require the supply to
provide their own containers. Generally speaking, the plastic
containers should not be reused for any trace analyses, that is the
metals. The glass containers for organics analyses samples should
follow the suggested cleaning procedure, including muffling at 400°C
for about 15 minutes. Once cleaned, these containers should be
stored and shipped in such a manner as to prevent recontamination.
Should the decision to reuse plastic containers be made, they should
be cleaned carefully before reuse. There are several cleaning
methods available. Choosing the best method involves careful
consideration of the nature of the sample and of the constituent(s)
to be determined.
4-13
-------
1. Traces of dichromate cleaning solution will interfere with metal
analyses. Use 1:1 nitric acid wash.
2. Traces of nitric acid may interfere with the nitrate analysis.
Use detergent with thorough rinses with tap and distilled water.
Shipping the containers to the sampling locations should take into
consideration the numbers to be shipped and eliminate any
contamination chances. The shipping containers to be used in
transporting the sample itself to the laboratory must be provided
either as a container for the empty sample container or in bulk form.
One item that must be given consideration is the preservative.
Postal regulations will not permit mailing of acids, particularly
nitric acid. Consequently, these materials must be purchased
locally or shipped by truck or other common carrier. If the
materials are to be purchased locally, the purity must be rigidly
controlled to assure no contaminants are present to affect results.
Even when the preservative nitric acid has been added and diluted by
the sample, postal restrictions may preclude the use of the mails.
Therefore, a special footnote has been added to the certification
procedures allowing an alternate icing followed by acidification
upon receipt of the sample in the laboratory.
4-14
-------
When shipping the sample to the laboratory for analysis, sufficient
time should be allowed to assure that the holding times are not
surpassed. Alternate forms of transportation should be checked out
beforehand to allow use if needed. The sample container must be
protected from physical damage in shipment and sufficient coolant added
to the ice chest or other form of insulated container to last through
the duration of shipment. Caps should be checked when the sample is
taken to assure that they will not leak. Upon receipt in the
laboratory, any deviation from the mandatory sampling requirements,
i.e., preservative, holding times, should be noted and, if necessary, a
resample ordered immediately.
SAMPLE COLLECTING
When the laboratory has been delegated the responsibility for sample
collection, handling, and preservation, there needs to be strict
adherence to correct sampling procedures, complete identification of
the sample and prompt transfer of the sample to the laboratory.
The collector should be trained in sampling procedures and approved by
the State regulatory authority or its delegated representative.
According to the National Interim Primary Drinking Water Regulations,
Section 141.2(c), the sampling location is the "free-flowing outlet of
4-15
-------
the ultimate consumer." Since this represents a minimal effort, one
sample can be taken at any point in the distribution system and fulfill
the regulation. Some States may require more frequent samples at
random locations, or a single composite sample taken at various
locations.
The exception to this sampling location is the turbidity sample which
must be taken at the point of entry of the water into the distribution
system.
When collecting the sample, the tap should be run to assure that the
water collected is from the distribution system and not from the
private service connection. The tap should be allowed to run at a
steady flow for two or three minutes before sampling. The sampler
should be sure that the tap is free of aerator, strainer, hose
attachment or water purification device. The sample container should
be flushed two or three times before the actual sample is taken unless
directed not to by the analytical method. The container should not be
filled completely to allow extra volume for effects of temperature
during transit. The preservative, if any, should be carefully added to
the container, the container capped and the sample shaken.
If the sample is to be cooled during shipment, the sample container
should be placed in an insulated container and sufficient coolant added
to last during shipment.
4-16
-------
The sample should be labeled to identify it during future analyses.
The information should include:
A. Date, place and time of sampling; name of person collecting the
sample.
B. Identification of the sample as to whether it is a routine
distribution system sample, check sample, raw or process water
sample, or other special purpose sample.
C. Analysis to be run on the sample as well as preservative added and
what amount has been added.
D. Any other remarks that the sampler thinks are necessary.
This information should be affixed to the sample container in such a
way as to assure that it will not become separated in later handlings.
The "Criteria and Procedures" manual states that chain of custody
procedures must be carried out on all samples taken for potential
enforcement actions only. The exact procedure and directions on this
procedure should be obtained from the appropriate certification
authority.
4-17
-------
V. FIELD MEASUREMENTS
As set down in the SDWA, there are two types of analysis which may be
carried out in other than certificed laboratories. These are the
analysis for residual chlorine and turbidity. These measurements may
be carried out in the field. In addition, should any other information
about the sample be required, such as pH, temperature, etc., these
should also be carried out in the field; It is not the scope of this
outline to discuss the procedures involved in these analyses.
State regulations may require additional procedures to be carried out
by the person taking the sample. The Interim Primary Drinking Water
Regulations do not.
VI. SUMMARY
The importance of proper sampling is the foundation of meaningful
analytical results. Consequently, a laboratory should know what
constitutes a meaningful sample in order to judge when a resample is
necessary due to improper sampling, preservation or handling techniques.
The preservative to be added, the type of container and the holding
times are spelled out in the "Criteria and Procedures" manual in a
mandatory section.
4-18
-------
The Laboratory Certification Officer must evaluate whether or not the
laboratory is conducting a proper sample receipt procedure and, if it
has the responsibility, a proper sampling of the water supplies.
4-19
-------
Unit 5
PERSONNEL
I. INTRODUCTION
The Safe Drinking Water Act, PL 93-523, has as its main intent to
provide good quality drinking water to all individuals and to insure the
continuation of this supply of high quality water. To do this a series of
MCL's and a schedule for all water supplies to follow in monitoring for
these chemical and bacteriological contaminants was established. In an
effort to continue in this intent, the National Interim Primary Drinking
Water Regulations, Section 141.28, requires that for compliance purposes
"sanples will be considered only if they have been analyzed by a laboratory
approved by the state except that measurement for turbidity and free
chlorine residual may be performed by any person acceptable to the state".
The present "Criteria and Procedures" manual has recommendations which
attempt to indicate the type of persons and experience necessary to carry
out the analysis of these contaminants.
This instructional unit will consider the personnel needs of a
laboratory doing analysis for water supplies in order to adequately carry
out the analysis of chemical contaminants. The guidelines set down in the
"Criteria and Procedures" manual are for a minimal program. This does not
in any way hinder a state from adopting guidelines or regulations more
stringent than those listed in the Federal guidelines.
5-1
-------
II. GUIDELINES FOR POSITIONS AND EXPERIENCE
Again, it should be stressed that this section is a recommendation and
not a requirement. The "Criteria and Procedures" manual has addressed
considerable thought to this section, listing the various personnel that can
be expected to be needed in a laboratory and the minimum of academic
training and experience required. The appropriate statements for each
position from the "Criteria and Procedures" manual have been included for
your reference.
A. Chemical Personnel Guidelines
1. Analyst for inorganic contaminants—that is, all chemical
measurements other than organic chemicals.
a. Academic Training: Minimum of high school diploma or its
equivalent (state certification or licensing may be
considered).
b. Experience
1) Minimum of six months of on-the-job training, under the
direct supervision of a qualified analyst, in
measurements being considered for approval.
5-2
-------
2) After six months, the analyst must demonstrate
acceptable skills through the successful participation
in the analysis of applicable performance evaluation
samples.
This individual would be responsible for all chemical measurements
other than organic chemicals. These would include all atomic
absorption techniques including the extraction portions as well as
the methods for nitrate and fluoride. Looking at the minimal
conditions a person who had graduated from high school would
receive six months, training before being considered a qualified
analyst. The section on "under the direct supervision of a
qualified analyst," could be a person who has the above mentioned
qualifications himself. The measurements carried out by this
individual during this training period would be acceptable for
compliance with National Interim Primary Drinking Water
Regulations.
The acceptability of his results would be based on the criteria
mentioned above being a guideline. The only mandatory
stipulations would be 1) that he use approved methods, i.e., those
referenced in the Federal Register; 2) that the sample have been
properly preserved; and 3) that the laboratory meets the quality
control regulations.
5-3
-------
2. Analyst for organic chemicals
a. Academic Training: Minimum of a bachelor's degree in
chemistry or its equivalent (state certification or
licensing may be considered).
b. Experience: Minimum of six months of experience in
measurements being considered for certification and two
years of experience in organic analysis. Each year of
college level training in related scientific fields or
demonstrated equivalency shall be considered equal to one
year of work experience. Such a substitution should not
exceed one-half of the required experience.
c. Supervision: Supervision by an analyst (also eligible to
analyze for organic chemicals) who has 1) a professional
degree or its equivalent with one year of course work in
organic chemistry; and 2) one year of experience in
measurement of organic chemicals; by gas chromatography.
This individual will be responsible for the operation of the gas
chromatograph and its use in determining the chlorinated
hydrocarbons, the chlorophenoxys, and the TTHM's. Some
indication has been given that if any changes in the National
Interim Primary Drinking Water Regulations are made the most
probable area would be the organics. The Regulations might be
expanded to include
5-4
-------
more extensive organic monitoring. It would be the area this
individual would be responsible for, so it might be well to know
this when this individual is being considered.
Here, too, the results generated by this individual would be
imnediately acceptable with the same provisions as listed under
the inorganic seciton.
3. GC/MS Operator: In addition to the organic analyst requirements
above, the following are recoiranended minimum standards for the
GC/MS operator, if this technique is used.
a. Training: Satisfactory completion of a minimum one week
course in GC/MS offered by an equipment manufacturer,
professional organization, university, or other qualified
operators.
b. Experience: Minimum of one year experience in the operation
of a GC/MS instrument.
4. Supervisor^
a. Academic Training: Minimum of bachelor's degree in
chemistry or its equivalent.
1 These positions should be optional in smaller laboratories.
5-5
-------
b. Experience: Minimum of two years of experience in
measurements being considered for certification.
With the equalization of a year of college-level training
to a year of experience this individual would only need his
bachelor's degree in chemistry to qualify for the position.
4. Laboratory Director*
a. Academic Training: Minimum of a bachelor of science degree
or its equivalent.
b. Experience: Minimum of five years' experience.
Larger laboratories would probably have a necessity for this
type individual to oversee the entire operation of the
laboratory. The following section on staffing will give some
consideration as to when administrative type individuals are
required.
III. LABORATORY STAFFING NEEDS
The fact that this section is not mentioned in the "Criteria and
Procedures" manual would indicate its lack of importance. However, it would
seem that an evaluator of a laboratory should have some ideas on the
subject. If individual analysts try to carry out too many analyses in a
5-6
-------
given time, their quality arid analytical precision can suffer. Laboratory
directors will have interest in this subject in order to predict future
needs should additional workloads materialize.
If the Certifying Officer has knowledge of some of the figures and
knowledge of what assumptions must be made to manipulate these numbers, he
can offer additional technical assistance to the laboratory he is inspecting
for certification.
One of the first pieces of data required in order to estimate staffing
requirements is an indication of the number of analyses a single individual
can carry out. All data of this nature are extremely variable. The
accompanying table gives some estimates of the rate of analysis for an
analyst. These data represent the number of a particular type analysis that
a single individual can run in one man-day. For example, 60 arsenic
determinations can be done in one man-day.
Several thoughts should be considered when contemplating these
figures. While it might be possible to run 150 samples/day when utilizing
an atomic absorption technique without any preliminary treatment, switching
from one technique to another would consume time. For example, to convert
the atomic absorption from its direct aspiration set-up to a gaseous hydride
configuration or to a flameless set-up would consume time.
The instrument would also need to be optimized after each conversion.
There is the possibility that more than one instrument would be available to
5-7
-------
be used, i.e., one for direct aspiration use, one for gaseous hydride and/or
one for the flameless determination of mercury.
Some consideration must also be given to how samples can be expected to
be received. Unless a laboratory can set dates for the samples to be taken,
the sampling would be at the discretion of the water treatment plant
operator. Consequently, there could presumably be times when no samples are
received in a laboratory and times when the volume of samples on hand to be
analyzed are beyond the capability of the personnel of the laboratory.
One possiblity to relieve this peaking would be position sharing. A
single analyst could be working in both water pollution and water supply and
as peaks occur could be switched from one batch of samples.to another.
Since the same parameters as determined under the NPDES program, the analyst
would have the same needs for expertise in both programs.
Of course, the individuality of analysts would dictate that all
individuals would not run the same number of samples in a given time.
Perhaps an analyst could be kept on one type of analysis, say atomic
absorption, and other on the wet analysis of nitrates of fluorides.
However, this would be possible only in a laboratory where the staffing
level would be high enough to make this feasible.
Automation of the methods will increase the number of samples that are
capable of being run in a given time, but this would increase the instrument
5-8
-------
investment considerably. Again, this would be a choice for a larger
laboratory which would have the capital. A number of possibilities exist as
to who would be doing the analysis. The size range of laboratories can vary
from small laboratories doing only selected analysis, larger commercial
laboratories doing all analysis, large state laboratories, and even combined
state laboratories doing anlysis for air, water, water supply and others.
Consequently, these extremes of possibilities make it even more difficult to
predict staffing unless more specific information is obtained.
The number of pieces of equipment available to the laboratory is
another factor. Duplication could cut the time down, but will drive the
costs up. All these items must be considered by a laboratory when
determining the manpower requirements. The figures given here are meant to
convey the simplest form, that is, one piece of equipment for each type
needed and utilization of automatic sample reading which could mean use of
equipment at night after the analyst has prepared the samples.
Once a rate of analysis has been chosen, the second number which must
be established is the workload that the laboratory can expect. Some
assistance in deciding this workload can be obtained from the sanpling
frequency set down in the National Interim Primary Drinking Water
Regulations. Then if the laboratory knows how many water supplies it will
have to be responsible for, the workload can be estimated.
5-9
-------
Here again, some additional thought will be needed. Quality control
will be recommended both from the "Criteria and Procedures" manual and from
normal operating procedures in a laboratory. This level of quality control
will place an additional workload on the laboratory by requiring check
samples to be run in addition to the normal load. Resampling will also add
to the workload. Resampling can occur from improper initial sample taking
and also to validate or disprove results not meeting the maximum contaminant
level. Thosestates having primacy could require additional sampling above
the Federal requirements; this would also raise the workload of the
laboratory. All these thoughts will need consideration before the total
workload of a laboratory can be determined.
Another demand on the laboratory personnel time other than analysis
will be a continued form of training. This could be in both providing
training to new employees, and upgrading of existing laboratory personnel
and, also being trained themselves to keep up their technical expertise and
learning new techniques.
IV. SUMMARY
The indicated recommendations in the "Criteria and Procedures" manual
for the Certification of Water Supply Laboratories lists four positions
which might be used in a laboratory. The document also lists what minimum
academic training and experience would seem to be needed to carry out the
5-10
-------
NUMBER OF ANALYSIS THAT CAN BE RUN PER MAN-DAY PER ANALYST
CHEMICAL
1 man year = 220 man days
Parameter
Model St. Prog.*l
ERCO2
EPA3
Arsenic
60
60
70
Barium
60
30
150
Cadmium
20
10
70
Chromi um
20
30
70
Lead
20
10
70
Mercury
20
20
70
Nitrate
85 (Auto-
30
45
mated)
Selenium
60
20
70
Silver
60
30
150
Fluoride (probe)
100
30
113
Chlorinated Hyc.
2
1
3
Chlorophenoxys
2
1
3
One additional piece of data, the National Sanitation Foundation in its
report estimated 4 man-days/year/system for laboratory support. However,
this also included bacteriology.
* Excluding administrative overhead, chain of custody procedures, report
writing, etc.
1 Model State Water Monitoring Program, U.S. E.P.A. Office of Water and
Hazardous Materials, Monitoring and Data Support Division
2 ERCO, Energy Resources Co., Inc. Study for EPA.
3 Handbook for Analytical Quality Control in Water and Wastewater
Laboratories EPA Analytical Quality Control Laboratory.
5-11
-------
functions necessary. Two of the four positions would seem to be needed to
carry out the functions necessary. Two of the four positions would be
considered optional depending on the size of the laboratory, i.e., the
Supervisor and laboratory director. However, in this case, it is suggested
that some type of outside consultant be retained to give the laboratory some
professional backup.
This unit tried to address the problem to how many persons a laboratory
will need. The attempt was to develop some thought on the part of the
Certifying Officer about the subject. It was felt that this type of
information could be valuable in assessing the needs of a laboratory if it
is operating above its capability or it needs to expand.
All laboratories, including those that exceed the recommendations
outlined herein, are urged to maintain and continually improve their
personnel, facilities, equipment, instrumentation, and quality control
procedures. To ensure continued production of scientifically and legally
defensible data, an ongoing training program should be an integral part of a
laboratory's program.
5-12
-------
UNIT 6
GENERAL LABORATORY PRACTICES
I. INTRODUCTION
This outline will consider general laboratory practices and services
that are fundamental in the operation and management of any analytical
laboratory. Sufficient information is offered for consideration by a
certification official to evaluate laboratory facilities, services, and
procedural details related in this.outline. Although the material
considered here serves as a guide for the certification official only and
places no mandatory requirement on the part of a participating laboratory,
deviations can be observed reflecting the need for updating and replacement
of equipment and supplies, inadequate laboratory facilities, and the lack of
attention given to procedural details for the care and maintenance of
laboratory and service equipment.
II. LABORATORY FACILITIES (GUIDELINE)
A. Space
The quality of the analytical analyses and the production of
reliable data can be adversely affected because of inadequate
laboratory facilities. Laboratory space is of concern in that
6-1
-------
space should be adequate to handle peak work loads. A laboratory
that must continually program limited work space and time for
various analyses operates under adverse conditions. As a
guideline, there should be approximately 150 to 200 square
feet/person, regardless of test, and approximately 15 linear feet
of usable bench space. Working space requirement should include
sufficient work bench space for processing samples; storage space
for chemicals, glassware and portable instrumentation; open floor
space for large equipment such as refrigeration, oven, water still,
washing and hood facilities, etc.
Considering the above space figures and taking into account the
many variables, a more practical estimate of space requirement can
be derived from observing operating conditions in any given
laboratory.
There are other variables affecting space requirements for any
individual laboratory operation. A laboratory performing routine
analyses for a small water plant may easily be housed in one room.
However, such space may be inadequate for a laboratory engaged in
more sophisticated analytical analyses or in laboratories doing
research. Multidiscipline laboratories may require separate
laboratory work area and support functions to facilitate timely
processing of water and other samples throughout the course of a
work day.
6-2
-------
B. Utilities - Water and Electric
Laboratory services and services equipment available for use in
laboratory operations can affect the quality of the laboratory
output. Every laboratory should have hot and cold running water
with a number of sinks depending on the size and workload of the
laboratory.
Electrical service first of all should conform to local, state or
national electrical codes. The adequacy of the electrical system
can best be judged by the system capacity to meet or exceed the
requirements of the laboratory operation. Consideration must be
given to a proper laboratory lighting system, operation of delicate
and sensitive laboratory instrumentation, and functioning of high
current demand equipment. Sufficient 120- and 220- volt circuits
must be located in specified areas and throughout the laboratory
allowing for better flexibility and sufficient capacity to handle
varying types of work. Independent circuits may be necessary for
operation of some instrumentation.
C. Exhaust Hoods
The handling of toxic and/or hazardous chemicals, solvents, and
gases may require the use of an exhaust ("fume") hood.
6-3
-------
Methodologies that include in the the analytical scheme such steps
as evaporation, extraction, distillation or digestion may also
require the use of an exhaust hood. The performance of a fume hood
is controlled by its designed hood face velocity, regulation of
flow over the face opening and control of outfall of fumes from the
hood face. Minimum face velocity at any point on the operating
opening should not be less than eighty percent of the average
design face velocity, which is the total air passage across the
hood face divided by the hood face area. The following table gives
average and minimum design face velocities in relation to the
nature of materials handled.
Laboratory Hood Design Face Velocities
Nature of
Materials
Handled
Gases
Vapors
Dusts, Fumes,
Mists
Mineral
Dusts
Average3
Minimum!3
High Toxic
Less
than 0.1
Less than 0.1
mg/cu meter
—
150
125
Moderately Toxic
0.1 to
100 ppm
0.1 to 15
mg/cu meter
To 5
mppcf
100
80
General
Laboratory Use
Above
100 ppm
Above 15
mg/cu meter
Above 5
mmpcf
60
50
Non-Toxic
a Total hood cfm divided by total face area.
b Lowest maintained velocity at any point across the hood face.
6-4
-------
Hood efficiency is affected by velocity gradients over the hood
face because of a hood's air distribution characteristics,
inadequate maintenance, and wear factors such as dirty vent ducts,
worn or loose fan belts, corrosion and deterioration. Exhaust
hoods should at least meet minimum safety standards as prescribed
in many states under labor codes.
Specifications should be such that the design constitutes a
balanced system taking mainly into account adequate air movement
and proper venting. Room temperature should not be affected while •
in the operating mode. Utilities should be available with water,
cup sink, lights and electrical outlets being the more important.
III. GENERAL LABORATORY PRACTICES
A. Glassware and Plasticware
Quality of laboratory data can be compromised both in test
sensitivity and reproducibility by improper choice, use, and
cleaning of laboratory glassware and plasticware. Of the many
types of glassware available, the best suited for all-round
analytical laboratory use is made from borosilicate glass
manufactured under the trade name "Pyrex" or "Kimble" brand. This
glassware is not completely inert, but is satifactory for all
6-5
-------
analyses included in EPA's methods manual. Where required other
brands are available for special purpose use possessing properties
such as resistance to heat, light, shock, and alkali.
The use of plasticware made from Teflon, polyethylene, polystyrene,
and polyproplylene has supplemented or replaced many glassware
items in the laboratory. The specific qualities of these plastics
and their relative chemical resistence must be taken into account
in regard to their use. Some points to consider in evaluating the
appropriate choice of glassware and/or plasticware are:
1. Borosilicate glassware is not completely inert, especially to
alkalies. Standard alkaline solutions and solutions of alkali
metals are best stored in plastic bottles.
2. Selection of plasticware should be predicated on the basis of its
specific properties, chemical resistence and intended use.
3. Plastic bottles and containers may be used for storage of reagents
and standard solutions where no effect from the contents and
container is encountered. Strong mineral acids and organic
solvents will attack polyethylene and should be avoided.
4. Disposable glassware is generally made of soft glass.
6-6
-------
5. Disposable glassware and plasticware should be discarded after
initial use.
6. Cracked, chipped or etched glassware should be discarded.
7. Plastic sample containers are not satisfactory for samples
collected for organic analyses.
B. Volumetric Glassware
Every laboratory should have a set of Class A volumetric glassware
for precise measurements of liquid volume. The ware includes the
usual burets, volumetric flasks and volumetric pipets. Class A
volumetric glassware so designated meet the Federal Specifications
as set by the National Bureau of Standards. Also one must be aware
of the designation of volumetric apparatus calibrated either to
delivery ("TD") or to contain ("TC") a definite volume of liquid.
Less accurate types of glassware including graduated cylinders,
serological, and measuring pipets, should never be used for precise
measurement of volume of solution. The graduated markings of such
glassware as Erlenmeyer flasks and beakers should be considered as
mere approximation of actual volume marked and probably should not
be relied upon as a volumetric measurement in any case.
6-7
-------
C. Cleaning of Glassware
The cleaning procedure adopted must be amenable to producing
chemically clean glassware and other items for reuse in the
laboratory. For most analytical needs it would be sufficient ;to
wash the items in a warm detergent solution followed by tap water
rinse and then with distilled water. However, it would be good
practice on the part of the analyst or whoever does the cleaning to
be aware of special cleaning requirements pertaining to certain
determinations and items of glassware. One may maintain separate
sets of glassware for certain procedures to avoid cross
contamination. Special cleaning requirements may be required for
volumetric glassware, especially burets and pipets. Absorption
cells used in spectrophotometers must be free of any film residues
and scratches.
Certain determinations, especially for trace metals require 1:1
nitric acid-water rinse after cleaning followed by repetitive tap
and distilled water rinses. Detergents containing phosphates
should not be used to clean glassware for phosphate
determinations. Ammonia-free water must be used to rinse glassware
for ammonia and Kjeldahl determinations.
6-8
-------
All glassware used in sample collection and organic residue
analyses should be free of organic and other contaminants.
Glassware should be thoroughly cleaned soon after use with
detergent solution, rinsed with tap water, distilled water,
redistilled acetone and a final rinse with high quality hexane.
Glassware should be allowed to dry, any mouth opening covered with
foil and ignited at 400°C for one hour. Volumetric ware, caps and
Teflon liners are not ignited. Glassware, unless cleaned for
immediate use, should be protected from environmental contamination
and stored in a closed contaminant free area. Caps and liners
should be stored in sealed containers. It has been found
advantageous to maintain a separate set of glassware (suitably
prepared) for the nitrate, mercury and lead procedures due to the
potentiality for contamination from the laboratory environment.
IV. REAGENTS, SOLVENTS AND GASES
It is intended that only the best quality chemical reagents, solvents,
and gases be enployed in lieu of any particular method requirement. The
minimum purity of reagents and chemicals should be analytical reagent
grade. Organic solvents and gases should be compatible with whatever
analytical operation is to be performed. Analytical standards should be
reference grade, when available.
6-9
-------
Factors such as the parameter being measured and sensitivity and
specificity of a detection system will determine the purity of the reagents
required. The quality of reagents and solvents may vary from lot to lot,
even within a given lot; therefore, a laboratory should check each new
container put to use. Proper storage is essential to prevent degradation.
All reagents, solvents and gases when purchased should be dated and recorded
to monitor shelf life. Subsequently, prepared reagents and standard
solutions should be monitored for changes in concentration or
deterioration. It would be expeditious to properly label these as to
compound, concentration, solvent used, name of preparer and date.
In the interest of safety, personnel should be aware that many chemical
reagents, both in their original state and prepared solutions, are hazardous
and should be handled with discretion to prevent their inhalation,
ingestion, absorption or contact through the lungs, mouth or skin.
V. DISTILLED AND/OR DEIONIZED WATER
The purity of distilled water and/or deionized water a laboratory
normally uses should be evaluated rather closely whether it is produced
in-house or purchased from a vendor.
Laboratory water should be free of substances that interfere with
sensitive chemical measurement with the degree of water purity being
6-10
-------
consistent with the needs for any particular analysis in question. Ordinary
distilled water is usually not pure and may contain minute traces of
impurities which could interfere with many colorimetric procedures.
However, for many analyses the distilled water is used without further
purification. Certain analysis may require a specific type of treatment,
namely ammonia free, carbon dioxide free or ion free water. When
determining trace organics the distilled water must be of sufficiently low
organic background. This may require further treatment of the distilled
water to eliminate this problem. To assure minimum contamination one must
make careful selection of the distillation apparatus including the still,
storage facility and any associated piping. Installation and maintenance
should be of prime concern.
Demineralized water made by passing tap water through a mixed bed ion
exchanger is applicable for many procedures. Very high purity water, less
than 0.1 micromho/cm conductance, can be produced by passing distilled water
instead of tap water through the same type filter.
Factors affecting suitability and quality of distilled and/or deionized
water:
A. Best distilled water system utilizes stainless-steel construction.
B. Other systems utilized in order of preference are quartz, vicor, or
Pyrex glass.
6-11
-------
C. Tin-lined systems are subject to corrosion exposing base metal to
contact with distilled water.
D. In order of preference, all connecting plumbing should be stainless
steel, Pyrex or special plastic made of polytetrafluoroethylene
(PTFE) material. Polyvinyl chloride (PVC) pipe should not be used
for connecting plumbing.
E. Storage tanks, in order of preference, should be stainless steel,
fiberglass, or (PTFE) plastic. Tanks should be sealed and fitted
with vented air filtering devices.
F. Ordinary distillation of water will not remove ammonia or carbon
dioxide. Volatile organics may distill over from the feed water
and nonvolatile impurities at times may come over in the steam in
the form of spray.
G. Pretreatment of the imput source water to a still by the use of any
combination of water softening, deionizing and/or a carbon
filtration will improve the quality of the distillate.
H. Post-treatment of the distillate using a deionizing column and a
carbon filter will produce very high purity water.
6-12
-------
I. A laboratory must monitor at scheduled intervals its distilled or
deionized water supply. One essential laboratory water quality
measure is conductivity. Supplemented with periodic checks for
various physical and chemical parameters will insure a continued
supply of high-quality laboratory water.
J. Maintenance of distillation equipment, including replacement or
recharging of any treatment columns should be performed on a
scheduled basis.
K. A laboratory must judge for itself the suitability of its
laboratory water supply. If the quality of the water does not
suffice, action must be taken along the lines of equipment update,
more frequent monitoring and maintenance.
ANALYTICAL BALANCES
A. Proper Care and Use of Balances
It has been stated that the most important piece of equipment in
any analytical laboratory is the analytical balance, yet the care
of the analytical balance is frequently overlooked.
6-13
-------
Most laboratories today are equipped with the single-beam, one-pan
balance. The controls and dials vary among different
manufacturers. This balance permits very rapid determinations to
be made by a substitution type weighing on a single beam with only
two knife edges.
Some laboratories are still equipped with one of the two types of
double beam analytical balances, one known as the rider and
graduated-beam type and the other called, the chain type, both types
employing the use of analytical weights.
Deviations from the following observable precautions for the proper
care and use of an analytical balance can have a cumulative effect
on the critical weighing operation.
1. Any balance should be operated according to the manufacturer's
instructions.
2. The balance should be mounted on a shock-proof table and
located away from the laboratory traffic. Area subject to
drafts, humidity changes and temperature changes should be
avoided.
6-14
-------
3. While using the balance, the balance level should be checked
frequently and adjustments made when necessary.
4. The control knob which releases the beam support should be
handled gently, so one does not jar the balance.
5. The beam should always be supported (not free-swinging) when
adding or removing any object or weight from the pan or when
adjusting a rider.
6. Never put chemicals directly on the balance pan, avoid
spillage on the pan and inside the balance case.
7. Never weigh an object while it is hot or even warm.
8. For the double beam balances, handle analytical weights only
with forceps; never use hands.
9. The balance case should be closed when taking final readings.
When through weighing, the beam should be raised from the
knife edges, weights returned to the beams or removed from the
pan, objects removed from the pan and the door closed.
6-15
-------
10. Balance accuracy and performance should be checked
periodically with the use of standardized Class "S" weights.
This may be performed in-house, however, lacking Class "S"
weights or if adjustments or repair are required, the service
of a company service man or balance consultant would be
in-order.
VII. PREPACKAGED KITS-CALIBRATION INTERVALS
As stated in the inclementation strategy document for certification of
water supply laboratories, all kit procedures, other than the N,
N-diethyl-p-phenylenediamine (DPD) Colorimetric Test Kit and the
3,5-dimethoxy-4-hydroxy-benzaldazine (Syringaldazine), are considered
alternative technqiues.
Visual comparison devices, whether color wheels, sealed ampules or other
visual standards utilized in these test kits, should be calibrated at least
every six months. These calibrations should be documented.
6-16
-------
Unit 7
PRE- AND POST-CONFERENCES
I. INTRODUCTION
The on-site evaluation will begin and end with conferences involving
the laboratory personnel. The Certification Officers will use the
pre-conference to gain information about the laboratory, its personnel and
the analytical methodology. After the inspection, they will meet with the
laboratory personnel and discuss the results of the evaluation. They will
indicate where improvements are necessary and, in general, assist the
laboratory in improving its analytical capabilities. These meetings should
leave the laboratory director or chief chemist with the impression of having
been assisted by the certification team.
Another important aspect of an on-site visit is the ability of the
certification team to offer technical assistance to the laboratory. The
"Criteria and Procedures" manual points out that the Regional Plans for
certifying local labs should contain plans for providing technical
assistance to laboratories in need of upgrading. Again, in discussing the
review of the on-site evaluation, the "Criteria and Procedures" manual
points out that the certification team should discuss possible assistance
the Region can provide the laboratory.
7-1
-------
II. PRE—EVALUATION CONFERENCE
In attendance at this conference should be at least the laboratory's
chief chemist, chief microbiologist and the laboratory director. This will
enable the evaluation team to be assured of having someone available to
answer questions about the various portions of the laboratory. This above
list should not imply that only these individuals may be present. Any
additional individuals whom the laboratory chiefs feel should be present
should attend.
Some information will already be in the hands of the certification
team, including which methods the laboratory has requested to be certified.
In the pre-conference, the Certification Officer can begin filling out the
evaluation forms by covering the personnel section, parts of the
methodology, sampling, quality control, and data reporting sections.
The background of the laboratory personnel with regard to their formal
education, training, experience, and responsibilities can be obtained by
questioning the chief chemist or laboratory director. The number of
personnel and the number of samples received by the laboratory per month can
allow the evaluator to assess the staffing of the laboratory.
The laboratory director can indicate which methods of analysis are
being used by the laboratory for each contaminant that is currently being
analyzed and cite the reference work used to describe the method. Knowing
the frequency of operation of each method gives the Certification Officer an
7-2
-------
idea of how much quality control will be necessary for the laboratory to be
carrying out to fulfill the certification mandate.
Information on sample handling such as the types of containers used,
what preservation techniques are used for each type analysis, delivery
times, and the policies on holding times can be obtained in the conference.
The Certification Officer can observe during the laboratory walk-through,
actual sample containers. He can then see if the sample containers are new
or if they are in good condition and are they proper, that is, do the
organic sample containers have Teflon cap liners and are they of acceptable
materials. He will want to see an example of actual data report forms to be
assured that it contains the required information. He could also check the
sampling, date to assure himself that the proper holding times are not being
ignored.
The Certification Officer can also determine during the conference the
types of instruments the laboratory has available. He can look at any
performance data the laboratory might have on known reference samples and on
unknowns, if he does not have this information already. He can also
determine during the conference if the laboratory is having any problems in
the analysis and can then spend more time during the actual evaluation on
the trouble areas.
The Certification Officer should fill out as much of the check sheet
as possible then do the actual evaluation and double check some items such
as equipment repair and specifications.
7-8
-------
III. POST-CONFERENCE
Before this conference is held, the certification team should review
their evaluations and prepare themselves for the meeting. This will help
identify forgotten items and allow the laboratory to offer explanations for
any deviations. If done with a sincere effort, this preparation will
minimize any new or undiscussed material in the actual report and keep down
any misunderstanding between the laboratory and Certification Officers.
During the post evaluation conference, the laboratory director and
chief chemist, should be in attendance. Any of the laboratory staff that
might be involved in the discussion should be in attendance also. For
example, if any deviations have been detected, the persons who are
responsible for that area should be invited to be present at the conference
and to participate in the discussions.
The actual discussions should center around any deviations in
methodology, instrumentation, sampling, preservation, holding times, quality
control, or any other subjects in which the evaluator has uncovered
potential problem areas The Certification Officer should consider this an
oral report to the laboratory covering his evaluation. The topics in the
post conference should not be limited and the certification team should
allow sufficient time so as not to appear to be pressed for time. The
laboratory officials should be encouraged to discuss all topics which they
feel should be covered. As the certification document states, the post
conference should end with a discussion of how the certification authority
can aid the laboratory.
7-4
-------
IV. SUMMARY
These conferences are extremely important. The pre-conference will
provide consderable information to the certification team and also allow
them to become acquainted with individuals in the laboratory whom they have
not previously met. The post-conference is the concluding conference and
should attempt to convey the idea of a technical assistance visit rather
than any enforcement endeavor. If the entire evaluation is handled
properly, a good working relationship can be established which can prevail
through all future contacts with the laboratory and will provide more
cooperation between the two parties.
7-5
-------
Unit 8
RECORDS AND REPORTS
I. INTRODUCTION
Production of reliable analytical data is the prime effort of a
laboratory. However, an addition to this responsibility, is the keeping of
records and report writing. Records of the chemical analysis, sample
information and quality control are all examples of information that must be
stored. The return of the analytical data to the concerned treatment plant
is an exanple of the reports a laboratory must produce. Records of chemical
analysis should be kept by the laboratory for not less than three years.
This includes all raw data, calculations and quality control data. The
laboratory Certification Officer will have to review these data in order to
properly certify the laboratory. This outline will list the records and
reports that should concern the evaluation officer for chemical parameters.
The form of data and report storage may vary from laboratory to
laboratory. Some larger laboratories may store the data in some automatic
data processing system while others will have hard copy storage. If the
Certification Officer knows exactly what is needed to satisfy his needs, he
can request the laboratory to gather such records together for his visit.
II. ANALYTICAL QUALITY CONTROL DATA
When a Certification Officer certifies a laboratory, he will need to
8-1
-------
evaluate the existence and significance of the laboratory's analytical
quality control data. This will include as a critical element the following
materials for all chemical parameters for which certification is sought.
A. Results of the Annual Performance Sample
The Certification Officer may have this data available at his own
office if the appropriate regional or state principal laboratory
has forwarded this information to him.
B. Standard Calibration Curves for Each Chemical Parameter for which
the Laboratory is to be Certified
This should also include the vertification of these curves daily or
once each time the curve is used, depending on the frequency of
analysis in the laboratory. If the frequency of use for a
particular analysis exceeds 20 samples per day, the data should
include proof of verification after every 20 samples.
C. The existence of a written Quality Assurance Plan
In order to assure that the laboratory data is of known and
acceptable there should be a written QA plan. This plan should
address the following, sampling procedures, sample handling,
equipment calibration, methodology, data reduction, internal QC,
preventative maintenance, routine for determining data accuracy and
precision and corrective action contingencies.
8-2
-------
D. That blanks, reagent water and WC standards are run as necessary to
assure good data
Reagent, field blanks and duplicates should be run to assure no
contamination in sanqples.. The proper number of laboratory quality
control standards should be run to assure proper maintenance of
precision.
Although listed as recommended the Certification Officer should ask
to see if available, and encourage compliance with, the following:
1. Current service contract on all balances; which could include
the use of class S weights to check accuracy of the balances.
2. A schedule and color standards for checking the wavelength of
the spectrophotometers.
3. Availability of a thermometer certified by the National Bureau
of Standards or one of equivalent accuracy.
4. Chemical should be dated before being placed on the shelf and
replaced as needed or as shelf life has been exceeded.
The last set of guideline information that the Certification
Officer should look for is only for those laboratories
analyzing samples other than its own. This data would include
results of the laboratory's performance on a known, reference
8-3
-------
sample. This should be carried out quarterly. These
performance evaluation samples will be made available to all
principal laboratories by EPA. The principal laboratories may
then utilize these sanples for the local laboratories being
evaluated by the State.
Additional records should be available showing verification of
precision and accuracy for each method. That is, duplicate
samples every ten samples, or with each set of samples and
calculations of standard deviations and use of quality control
charts.
III. SAMPLING RECORDS
The laboratory may or may not be responsible for the actual taking of
the sanples. However, it is the responsibility of the laboratory to assure
itself that the sample is meaningful and that the mandatory handling and
preservation techniques have been carried out.
The actual sampling records retained by the laboratory will need to be
evaluated by the entire evaluation team, provided that the laboratory is
being evaluated for all the parameters. The chemical and bacteriological
evaluators will have need to see these records. The data is required to be
kept by the laboratory for at least three years.
The Certification Officers should satisfy themselves that sufficient
data is being kept by the laboratory. They should review the records for
8-4
-------
compliance with the required sampling frequency, check sampling program,
sample transit time and resampling notices, if appropriate. The suggested
data on each sample should include at least the following information:
A. Date, place and time of sanpling; name of the persons ftho collected
the sample.
B. Identification of the sample as to whether it is a routine
distribution system sample, check sample, raw or process water
sample, or other special purpose sample.
C. Date of receipt of sample and analysis.
D. Laboratory and persons responsible for performing analysis.
E. Analytical technique/method used.
F. Results of Analysis.
The Certification Officer might wish to make laboratories aware of some
of the following points. If a laboratory has been contracted with by a
public water supply to provide analytical services that laboratory might
wish to set up a meeting with that supply. The laboratory can then assure
that the proper information and sample frequencies are to be provided and
what the laboratory will do when there is doubt of the veracity of the
sample. State principal laboratories might wish to retain data on the
turbidity and residual chlorine determinations and sanpling to be assured
8-5
-------
that these analysis are correct in analytical results, frequency and
location of sampling.
IV. DATA RETENTION AND REPORTS
If a person compares the Data Reporting section of the "Criteria and
Procedures" and Section 141.33 of the National Interim Primary Drinking
Water Regulations, it should be remembered that the documents are referring
to two different responsible parties. The National Interim Primary Drinking
Water Regulations are referring to the Public Water System and requiring
data retention for ten years. The "Criteria and Procedures" manual is
referring to a laboratory which performed the analysis for that public water
supply. There will be cases where the laboratory is part of the water
supply. In this case the data will have to be kept for the longer periods
to comply with the requirements for the public supply.
The laboratory should have the records of their personnel's training
and experience available for inspection by the Certification Officer.
However, since this area is a guideline, this information would be obtained
from the laboratory director during the pre-evaluation discussion.
The Certification Officer might wish to see the reports from the
laboratory to the water supply giving the results of the analysis. This
would indicate whether sufficient data is being transmitted by the
laboratory to fulfill the needs of the water supply.
8-6
-------
Critical Element
Recommended
I. QUALITY CONTROL
A. Lab doing own
analysis only
1. General
a. Written QA Plans
b. Annual performance sample
c. Available methods manual
d. Calibration of pH meter with
each use
2. Inorganic
a. Standard curve for each
parameter
b. Verification of standard
curve daily or every 20
sanples
3. Organic
a. .Daily method blank for
pesticides & herbicides
b. Stds. daily to show linearity
c. QC standard run quarterly
d. Filled blank for TTHM with
each set of samples
e. 10% of samples should be
duplicates for TTHM's
f. Daily TTHM lab control std.
g. Proven precision of TTHM
equipment
h. Raw water blanks for Liq/Liq
ext. of TTHM's
i. Prove performance of 6C/MS by
using BFB
1. Service contract on
balances
2. Records showing cali-
bration of instruments
8-7
-------
Critical Element
Recommended
B. Lab doing other
than its own
samples
1. All of the above
1. Quarterly analysis of
known sanple
2. One duplicate sample
every 10 samples - or
with each set of -samples
3. Calculate standard
deviation
4. Daily use of
quality control charts
or tabulation of mean
and standard deviation
SAMPLING DATA
1. Retention of data for
not less than one year
Enforcement data for
three years
2. Retained data to include
a. Date of sampling
b. Place of sampling
c. Time of sampling
d. Name of sanple
collector
e. Identification of
sample
Distribution sanple
Check sanple
Raw sanple
Processed sample
Special sample
f. Date of receipt of
sample
g. Date of analysis of
sample
h. Name of lab
i. Name of analyst
j. Method use for
analysis
k. Result
8-8
-------
DATA RETENTION AND REPORTS
1. Evaluate personnel back-
ground (education and
experience)
2. Approval of plan for
turbidity and residual
chlorine
3. Retention times of dat
8-9
-------
If the laboratory is a principal State laboratory, the records should
be available to the Certification Officer showing the provisions the State
has made to approve turbidity analysis and chlorine residual analysis if
substitution is to be allowed by the State.
V. SUMMARY
An important part of the on-site visit is the evaluation of the records
and reports of the laboratory. The Certification Officer will evaluate
critical element quality control materials as well as the recommended data
of the quality control and data reporting sections of the "Criteria and
Procedures" manual.
The effort the laboratory has put forth to document the approval of the
turbidity and residual chlorine determinations should be evaluated in the
State principal laboratories. Also the time the data is kept and in what
manner it is kept should be evaluated.
This section can indicate to the Certification Officer the reliability
of the data (from the quality control materials) as well as the accuracy of
the sampling techniques, handling and preservation, and documentation of the
sanpling program.
8-10
-------
Unit 9
STATISTICS FOR CHEMISTS
I. INTRODUCTION
Statistics may be defined, for our purpose* as a collection of
mathematical methods which have been developed for handling numerical data
pertaining to samples or portions of entire populations.
The statistical methods with which we will concern ourselves deal with
the presentation and analysis of numerical data from samples.
II. FREQUENCY
A. Definitions
1. Frequency - indicates how many times a particular, individual
datum value occurs in a collection of data.
2. Frequency table - a tabular arrangement of data, ranged in
ascending or descending order of magnitude, together with the
corresponding frequencies (See Table I).
9-1
-------
3. Frequency histogram - a set of rectangles having bases (lower 4
upper class boundaries) on a horizontal axis (abscissa) with
centers at the given data values and heights equal to the
correponding frequencies depicted on the vertical axis
(ordinate). NOTE: class boundaries should not coincide with
any values in your data set (See Figure 1).
4. Frequency polygon - a line graph of frequencies plotted against
data values. This graph can be constructed by connecting
midpoints of tops of rectangles in the frequency histogram (See
Figure 1).
B. Application
Consider the application of the above definitions to the following
set of data, obtained from 12 determinations for chloride in water.
100
101
98
99
Results (uq/L)
101
100
102
101
100
102
100
99
9-2
-------
Table I
Frequency Table
Lower Upper
Class Class
Chloride (uq/L) Frequency Boundary Boundary
98 1 97.5 98.5
99 2 98.5 99.5
100 4 99.5 100.5
101 3 100.5 101.5
102 2 101.5 102.5
Figure 1
Frequency Histogram & Frequency Polygon
]
>>
g 3 -3
u
3
O"
h z —
b.
I .
7^
Ul
T
\
\
v
t
98
i
99
I
10G
I
101
«
102
Chloride yg/L
9-3
-------
III. MEASURES OF CENTRAL TENDENCY
A. Definitions
Central tendency - the tendency of values to cluster about a
particular value in the distribution.
1. Mean - arithmetic average of all the values in the sample
where there are n numbers of values.
2. Mode - that value which occurs most frequently
3. Median - midpoint of an array (arranged in order of magnitude)
of individual data values. It is the middle value or the
arithemetic mean of the two middle values. If there are an odd
number of observations, n, the median is
n
n
* * where
x
n + 1
"2
represents the
9-4
-------
n + 1
2
value in the frequency distribution. If there
are an even number of observations the median
is:
X + X
n *r
7 *
+ 1
the average of the middle two data values.
xu
4. Geometric Mean: The n root of the product of the data
points.
Tg " Vx) (x2)...(*„)
•" JtT-7
This is what we get when we transform the data to logrithms,
then take the anti-log of the arithemetic mean of the
transformed data.
Xg = antilog
log (xx) + log (x2)
• + log (*n)
B. Application
Consider the application of the above definitions to the previously
mentioned set of data, obtained from twelve determinations for
chloride in water, shown in Table I.
1. Mean
98+2 (99) + 4 (100) + 3 (101) + 2 (102) = 100.25 Uq/L
9-5
-------
2. Mode = 100 yg/L
3. Median = £ 7 + 1
1
7
= 100 + 100 =100 Ug/L
—2
4. Geometric Mean = Xg = Jl.QZ X 10^4 = 100.24 ug/L
or Xg = antilog ,,
(98) + 2 log (99) + 4 log (100) + 3 log (101) + 2 log (102) = 100.24 ug/L
IV. MEASURES OF DISPERSION
A. Definitions
1. Dispersion - spread or variability of observations in a distribution.
2. Range - the difference between the highest value in the data set and the
lowest value in the data set.
R = max - min
3. Variance - the sum of the squares of the deviations of the individual
data values from their mean divided by the total number of data values(n)
minus 1
9-6
-------
The definition of the variance can be expressed by the following
formula:
S3 -ifc ' *'2
n - 1
4. Standard deviation - the square root of the variance.
The definition of the standard deviation can be expressed by the
following formula:
s =
However, the formula commonly used because of its adaptability
to the hand calculator is the following:
where there are n number of values.
5. Standard deviation - derived from the range statistic. For
small data sets (n < 10), the range statistic is often used to
9-7
-------
estimate standard deviation:
R as
The value for dn is obtained from Table II.
Table II
Size of sample(n)
2 1.128
3 1.693
4 2.059
5 2.326
6 2.534
7 2.704
8 2.847
9 2.970
10 3.078
12 3.258
16 3.532
6. Standard deviation of the mean (s7) - the standard deviation of
individual data items(s) divided by the square root of the number
of data items(n).
The definition of the standard deviation of the mean can be expressed
by the formula:
s
9-8
-------
7. Relative standard deviation - the standard deviation(s)
expressed as a fraction of the mean:
RSD-f
The relative standard deviation is often expressed as a
percent. It is then referred to as percent relative standard
deviation:
% RSD . | . 100
The relative standard deviation is particularly helpful when
comparing the precision of a number of determination on a given
substance at different levels of concentration.
8. Relative range - the range (R) expressed as a fraction of the
mean:
9-9
-------
The relative range can also be expressed as a percent. It is
then referred to as percent relative range:
% RR = - *100
J
B. Application
Consider the application of the above definitions to the previously
mentioned set of data, obtained from twelve determinations for
chloride in water.
1. Range = 102 - 98 = 4 ug/L
xi
n
2. Variance: s2 =
(xrD2
n-l
x^X (xi - X")2 f(xr X)2
1 98 -2.25 5.06 5.06
2 99 -1.25 1.56 3.12
4 100 - .25 .06 .24
3 101 + .75 .56 1.68
2 102 +1.75 3.06 6.12
llTZ?
9-10
-------
(Xi -
*)2 - 16.22 = lA7
irrr ~~rr
3. Standard deviation - square root of the variance
fe Xj2 - fe Xi)'
n - 1
xi
fx .
fx4
1
2
4
3
2
98
99
100
101
102
98
198
400
303
204
1201
9604
9801
10000
10201
10404
9604
19602
40000
30603
20808
imu
s =
120617 - 12g32
n
120617 - 120601
n
= 1.21 ug/L
4. Standard deviation - derived from the range statistic
s = R
^T
1.23 ug/L
5. Standard deviation of the mean:
ST = £ = 1-zl = 0.35 ug/L
n TT
6. Relative Standard Deviation:
RSO => - 1,21
J 100.25
0.0121
9-11
-------
or percent relative standard deviation
% RSD = — .100 = 1,21 .100 = 1.21%
J 100.25
7. Relative range:
RR = - = —i— = 0.0399
J 100.25
or percent relative range
%RR ss — .100 = — .100 = 3.99%
J 100.25
V. INTRODUCTION TO NORMAL DISTRIBUTION CURVE
A. Statistics deal with theoretical curves which are smoother than the
frequency polygons, that are obtained from experiments in real
life. However, frequency distributions or frequency polygons of
experimental data often approximate a mathematical function called
the "normal" distribution curve. (See Figure 2)
Figure 2
Normal Distribution Curve
x
cj
e
o
c
-41
U
U.
Quantity Measured
9-12
-------
As shown in Figure 3, the frequency polygon for the 12
determinations for chloride in water is a fairly good
approximation of the normal curve.
Comparison of Normal Curve and Frequency Polygon
Chloride yg/L
If, however, in the chloride determinations we had obtained 103
ug/L instead of 98 ug/L and 104 ug/L instead of 99 ug/L this
distribution would not have been a good approximation of the
normal curve, as is shown by the frequency polygon in Figure 4.
Figure 4
Comparison of Normal Curve and Frequency Polygon
-------
Bo If a frequency distribution is a good approximation of the normal
curve, we can use some facts about the normal curve to give us
information about the frequency distribution.
The normal curve has the following equation:
Y = J e -Ui-u)2/2a2
c yjiv
where = it = 3.1416
e = 2.7183
x^ = corresponds to point on the abscissa
Y = corresponds to point on the ordinate
(j^ => variance
a = standard deviation
u a mean
The normal curve can be completely specified by a, the standard
deviation of the population of numbers represented by the curve,
and by w, the mean of the population of numbers. A population
consists of a group of items or individuals.
9-14
-------
Figure 5 shows the normal distribution in terms of the population
mean, and the standard deviation of the population o, and gives the
percent of area under the curve between certain points.
Figure 5
Normal Distribution Curve
We may check the distribution of sample data to see if it is a
"normal" distribution in the following manner. Substitute the
value of the sample mean (X) for the value of the midline and
9-15
-------
substitute the value of the sample standard deviation (s) for the
limits of the value spans where we might expect certain percentages
of the data items to occur. Then we can check the number of data
items which actually do occur within these values spans. -
Figure 6 demonstrates this application using the chloride data
values from Table 1. The data values are marked on the horizontal
line and the frequency of the occurrence of each value is marked on
the vertical. The midline of the distribution is marked at the
value of the sample mean (X = 100.25, See III B 1). The value of
the sample standard deviation (s = 1.21, See IV B 3) is used to
mark value areas under the curve where different percentages of
data values will probably occur. Thus, for the area X" * Is, X - Is
= 99.04 and X" + Is = 101.46. Therefore, according to the normal
distribution curve shown in Figure 5, we might expect about 68% of
the data items to have values between 99 and 101. (The values are
rounded to whole numbers since the data values are thus recorded).
Consulting Table 1, we find that 75% or 9 of the 12 data items have
values in this range. This percentage is shown in Figure 6 by the
frequency polygon for the data shown earlier in Ficfure 3.
Likewise assuming a normal distribution, we would expect 95% of the
observations to lie within * 2 a's from the population mean. In
fact, 100% of the observations were within * 2 s's from the sample
mean.
9-16
-------
4 -I
3 -J
>>
-------
The procedure is as follows:
1. Find the upper class boundaries (see frequency histogram II A 3).
2. Calculate the cumulative relative frequencies as percentages
using the formula:
P = 100(i—0.5)/n
where: P = cumulative relative frequency as a percentage
i = cumulative frequency
n s total number of data points
3. Plot the upper class boundaries (abcissa) versus P (ordinate).
4. Draw the line of best fit.
5. Using the pair of dots, located next to the vertical axes,
locate the point where a horizontal line will intersect the line
of best fit.
6. Project these intersections upward to where they intersect the
proper line in the upper grid.
9-18
-------
7. Construct the distribution curve.
Application: Consider the application of the above procedure to
the previously mentioned set of data, obtained from twelve
determinations for chloride in water, shown in Table I.
Frequency
Chloride (uq/L)
Frequency
Cumulative
P
Upper Class Boundary
98
1
1
4.2%
98.5
99
2
3
20.8%
99.5
100
4
7
54.2%
100.5
101
3
10
79.2%
101.5
102
2
12
95.8%
102.5
Figure 7 depicts.this data plotted on normal propability paper.
The dashed lines show how one uses the pair of dots, located
next to the vertical axes, to construct the distribution curve
(steps 5, 6, and 7 of the procedure). From this plot the mean
(IT) and the standard deviation (S) can be estimated. The mean
is found by reading the value of the abscissa where the line
intersects the cumulative frequency percent of 50 - (In our
example IT is approximately equal to 100.5 yg/L). The standard
deviation may be estimated by extrapolating the line until it
intersects the top and bottom of the graph paper. These two
points of intersection represents approximately 99.87% of the
data, or J *¦ 3S. (In our example J ± 3s is approximately equal
to 104.6 - 96.9 = 7.7; s is approximately equal to 7.7 divided
by 6 or 1.28 yg/L).
9-19
-------
normal -f&QMZILlTY flOT
%1-f I00.7 lot-?
Fig** 7
Mr
laS-f
-------
If data are not normally distributed, then the data will have to be
transformed to make it normally distributed. Some common
transformations are:
1. replacing Xi by log x-j,
2. replacing x^ by Vx7
3. replacing x^ by 1^ or 1^
xi fir
CONTROL CHARTS
A. INTRODUCTION
As long as repeated samples/analysis exhibit only random errors the
measurement process is in statistical control. The purpose of a
quality control (QC) chart is to graphically detect lack of
statistical control in the measurement process.
The relationship that exists between a normal curve and a control
chart is probably best understood by looking at Figure 8 that
depicts the two side-by-side.
9-21
-------
Figure 8
Upper Control Limit (UCL)
Central Line
Lower Control Limit (LCL)
Order of Results
Control charts may be kept on any of various characteristics (e.g.,
average 00; percent recoveries (P), standard deviation (S),
percent relative standard deviation (%RSD), range, percent relative
range (%RR), etc.
Usually, the divisions of a control chart are not integer (whole
number) factors times the standard deviation values, e.g., 3s.
This is because s is calculated from a limited number of
observations, and because the values in Table IV already have
factored in 1 (this factor comes from the fact that we are
n
concerned about fluctuations about averages - remember that the
standard deviation of the mean is defined as
9-22
-------
B. CONSTRUCTION OF CONTROL CHARTS
The American Society for Testing and Materials (ASTM) published
Special Technical Publication 15D, entitled "ASTM Manual on
Presentation of Data and Control Chart Analysis". This document is
an excellent guide to preparing QC charts. Table III is a summary
of some commonly used types of control charts.
Setting control limits
Accumulate data by analyzing sets of replicates at a given
concentration level or if historical data already exist you
may use these values.
Calculate the average of the needed characteristics.
Look up in the appropriate table the factors for control
limits (Table IV).
Calculate central lines and the upper and lower control
limits according to the formulas given in Table III:
9-23
-------
Table III.
Central Line Control Limits
Averages using S X X *
Averages using R J X * AgE"
Standard Deviation T and B^S"
Ranges ft D^R and D-jft
9-24
-------
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Table IV
Chart for
Averages
Factors for
Control Limits
*2
Chart for Standard
Deviation
Factors for
Control Limits
Chart for
Ranges
Factors for
Control Limits
1.880
1.023
0.729
0.577
0.483
0.419
0.373
0.337
0.308
0.285
0.266
0.249
0.235
0.223
0.212
0.203
0.194
0.187
0.180
0.173
0.167
0.162
0.157
0.153
2.659
1.954
1.628
1.427
1.287
1.182
1.099
1.032
0.975
0.927
0.886
0.850
0.817
0.789
0.763
0.739
0.718
0.698
0.680
0.663
0.647
0.633
0.619
0.606
0
0
0 .
0
0.030
0.118
0.185
0.239
0.284
0.321
0.354
0.382
0.406
0.428
0.448
0.466
0.482
0.497
0.510
0.523
0.534
0.545
0.555
0.565
3.267
2.568
2.266
2.089
1.970
1.882
1.815
1.761
1.716
1.679
1.646
1.618
1.594
1.572
1.552
1.534
1.518
1.503
1.490
1.477
1.466
1.455
1.445
1.435
0
0
0
0
0
0.076
0.136
0.184
0.223
0.256
0.284
0.308
0.329
0.348
0.364
0.379
0.392
0.404
0.414
0.425
0.434
0.443
0.452
0.459
3.267
2.575
2.282
2.115
2.004
1.924
1.864
1.816
1.777
1.744
1.716
1.692
1.671
1.652
1.636
1.621
1.608
1.596
1.586
1.575
1.566
1.557
1.548
1.541
9-25
-------
Examples:
a. No Historical Data Given:
For example, suppose you wanted to construct a range chart
(duplicates).
1. List the range (R) for each set of samples
2. Calculate the average range by summing the list of R values and
dividing by the number of sets of duplicates:
Duplicates
Range
10.01 10.54
10.23 9.87
10.41 10.01
9.86 10.32
10.02 10.01
9.48 9.79
10.12 9.62
9.66 10.31
9.84 10.12
10.31 9.81
0.53
0.36
0.40
0.46
0.01
0.31
0.50
0.65
0.28
0.50
m
9-26
-------
3. Calculate the Upper Control Limit (UCL) on the average range
NOTE: The number of observations (n) in this sample are two
(2) since duplicates were used here.
UCL = 04 IT = 3.267 (0.4) = 1.31
4. Calculate the Lower Control Limit (LCL) on the range
LCL = O3 IT = 0 (0.4) = 0
5. Graph R, UCLr, and LCLr
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
UCL = 1.31
TT = 0.4
1 T
5 s 7 8 9 10 rr
Order of Results
duplicate sairqple sets
6. The above precision control chart is now conplete, and can
be used to plot R values from subsequent duplicate samples
to determine if the sample/analysis is in control, out of
control (plotted R value beyond UCLR or LCL^), and/or to
detect any trends developing within the process.
9-27
-------
b. Historical Data Given:
If a method reports or states performance characteristics, then
until you can accumulate enough data to calculate your own, you
could establish control limits using the reported historical
data. NOTE: It is always preferable to collect enough data to
calculate your own performance characteristics.
For example, suppose you wanted to construct a mean recovery (X)
using an R chart (duplicates) for the analysis of toluene using
Method 503.1. Table V presents historical performance
characteristics data given in the method.
1. Calculate the central line
Historical Data
spike level » 0.40 ug/L
mean % Recovery = 93%
y = (spike level) (% Recovery) = (40yg/L) (0.93) » 0.37 ug/L
2. Calculate the range
R
dn as R a Sdn
Historical Data
s - .022 yg/L
R = (.022) (1.128) = .025 Ug/L
9-28
-------
3. Calculate Upper Control Limit (UCL)
UCL = I + A2 ^ = 0.37 Ug/L + (1.880) (.025) = 0.4170 Wg/L
4. Calculate Lower Control Limit (LCL)
LCL = I - A2 IT = 0.37 yg/L - (1.880) (.025) = 0.3230 ng/L
9-29
-------
TABLE V
Single Laboratory Accuracy and Precision for Aged Samples Containing Aromatic Compounds
and Selected Organohalides Spiked Into Chlorinated Drinking Mater
Relative Length
Spike
Day 1
Samples
Mean
Standard
Standard
of
Compound
Level
Recovery
Analyzed3
Recovery
Deviation
Deviation
Study
(ug/L)
1%)
1%)
. 1%)
(%)
(days)
Benzene
0.40
100
7
100
0.082
2.1
15
Tr i ch1or oethy1ene
0.50
103
10
104
0.037
7.0
28
a-Trichlorotoluene
0.50
100
9
89
0.048
11.0
28
Toluene
0.40
100
7
93
0.022
5.7
15
Tetrachloroethylene
0.50
108
10
104
0.040
1.1
28
Ethylbenzene
0.40
103
7
93
0.032
8.5
15
1-Ch1orocyc1ohexene-1
0.50
96
10
91
0.029
6.4
28
p-Xylene
0.40
95
7
85
0.029
8.7
15
Chlorobenzene
0.50
96
10
96
0.029
6.1
28
m-Xylene
0.40
95
7
90
0.028
7.7
15
o-Xylene
0.40
93
7
90
0.026
7.2
15
i so-Propy1benzene
0.40
93
7
88
0.030
8,7
15
Styrene
0.40
0.00
7
0.00
0*
p-Bromofluorobenzene
No Data
m-Propylbenzene
0.40
90
7
83
0.030
9.3
15
t-Butylbenzene
0.40
95
7
88
0.030
8.7
15
o-Chlorotoluene
No data
p-Chlorotoluene
0.50
94
8
93
.022
4.9
13*
Bromobenzene
0.50
96
10
93
.030
6.4
28
sec-Butylbenzene
0.40
85
7
80
0.034
11.0
15
1,3,5-Tr i methy1benzene
0.50
96
10
92
0.040
8.7
28
p-Cymene
0.80
92
5
88
0.012
2.8
6*
1,2,4-Trimethylbenzene
0.40
83
7
75
0.029
8.7
15
p-Dichlorobenzene
0.50
106
10
100
0.029
8.7
28
m-Dichlorobenzene
0.50
96
10
92
0.040
8.7
28
Cycloprepylbenzene
No data
m-Butylbenzene
0.40
90
7
78
0.049
16
15
2,3-Dibenzofuran
0.40
14
7
0.0
—
—
0*
o-Dichlorobenzene
0.50
102
9
92
0.033
7.1
28
Hexachlorobutadi ene-1,3
0.50
88
10
74
0.062
17
28
1,2,4-Trichlorobenzene
0.50
94
10
88
0.047
11
28
Naphthalene
0.50
108
8
96
0.062
13
20*
1,2,3-Tr ichlorobenzene
0.50
100
10
85
0.046
11
28
a Samples randomly analyzed throughout the study period.
* Maximum recommended holding time.
-------
5. Graph X, UCL and LCL
0.42
UCL = 0.42 ug/L
0.37
X = 0.37 Wq/L
0.32
LCL = 0.32 uq/L
I—2—3—5—5—5—7—5—5—15—U
Order of Results
Mean of Duplicates
6. The above X control chart is now complete and can be used to
plot mean values of subsequent duplicates to determine if
the sample/analysis is in control, out of control (plotted X
values beyond UCL or LCL), and/or to detect any trends
developing within the process. Note this chart is only
applicable when the dosing leel is 0.40 yg/L.
NOTE: If the plotting of 30 subsequent means of duplicate
indicates that your analytical system is in control, you
should use your own data to construct a new chart.
9-31
-------
VII. INTERPRETATION OF CONTROL CHARTS
Examples:
I. In this example, a trend has developed; the characteristic being
monitored has changed (8 consecutive points lie to one side of the
central line). If the change is detrimental then all variables in
the procedure should be checked in an attempt to stop this
deteriation of the process.
Characteristic
UCL
being
monitored
X
X
X
X X X X XX
Central Line
X
X
LCL
T 2 3 5 5 5 7 8 9 ID TT
Order of Results
When only random errors occur the probability that a result will
lie above the central line is 0.5 (50%), and the probability that a
result will lie below the central line is also 0.5 (50%). The
probability that seven consecutive results will lie on the same
side of the central line given that the one immediately preceding
the seven also lies on the same side of the central line
(conditional probability) is given by the following expression:
9-32
-------
0.5)8 (0.5)7 = .0078
"075T
The probability that this event would happen by chance alone is
.0078, (0.78%); a very unlikely event - less than one chance out of
99.
2. In this example, a gross error has occurred, the characteristic
being monitored has gone outside of the control limits. At this
point the system can be stopped and all variables in the system
checked. Once the system has been corrected, all samples between
sample set 3 and 4 should be rerun to insure the validity of the
data.
Characteristics
being
monitored
1 5 5—7 5"
Order of Results
UCL
Central Line
LCL
"TO FT
When only random errors occur the probability that a result will
lie above the upper control line is 0.01 (1%); a very unlikely
event - one chance out of 99.
9-33
-------
Control chart data are normally generated by analyzing sets of
quality control samples - the analytes of interest may be placed in
either laboratory pure water or a standard water that matches the
type of matrix normally analyzed in your laboratory. Control charts
should be reconstructed (updated) periodically, because the
performance characteristics may change (when 30 data points have
been collected, can use newer data to construct a new chart), for
the better, with time.
9-34
-------
REFERENCES
1. Bennett, C.A. and Franklin, N.L. Statistical Analysis in
Chemistry and the Chemical Industry. John Wiley and Sons, Inc.,
New York. 1954.
2. Crow, E.L., Davis, F.A., and Maxfield, M.W. Statistics Manual.
Dover Publications, Inc., New York. 1960.
3. Dixon, W.J. and Massey, F.J. Introduction to Statistical
Analysis. McGraw-Hill Book Co., Inc., New York. 1957.
4. Ostle, B. Statistics in Research. The Iowa State University
Press, Iowa. 1963
5. Youdon, W.J. Statistical Methods for Chemists. John Wiley and
Sons, Inc. New York. 1951.
6. American Society for Testing and Materials (ASTM) Committee E—11
on Statistical Methods. ASTM Manual on Presentation of Data and
Control Chart Analysis, ASTM STP 15D. ASTM, 1916 Race Street,
Philadelphia, Pennsylvania, 19103
9-35
-------
Unit 10
QUALITY CONTROL
I. INTRODUCTION
The purpose of the Safe Drinking Water Act is to assure the public of
an adequate supply of safe water. To achieve this, maximum levels of
certain contaminants were proposed along with the prescribed methodology for
analyzing for these parameters. When a laboratory performs these analyses,
the laboratory should practice quality control to assure that the results
being reported are true values and not in error.
Data developed from these examinations must be reliable and beyond
reproach. The data can be used for making judgments on technical operations
in water treatment or in legal actions involving public health hazards. For
these reasons the "Criteria and Procedures" manual has set down some
critical and some recommended QC procedures.
The entire critical elements for certification section contained in the
"Criteria and Procedures" manual is considered as the minimum acceptable
program on quality control that a laboratory can carry out and still expect
reliable results. Most laboratories will want to go beyond these minimum
critical requirements and include more quality control.
10-1
-------
This outline will cover the minimum critical quality control
procedures, then go into the recommended practices portions and proceed
further into some ideas not in the "Criteria and Procedures" manual. The
reasons for going further are to acquaint laboratory certification personnel
with sufficient information to be able to evaluate whether the laboratory
has conplied with the minimum critical sections and allow the Certification
Officer to reconmend further procedures. The topic of quality control from
all aspects in a laboratory is well covered in the Handbook for Analytical
Quality Control in Water and Wastewater Laboratories EPA—600/4—79—019
produced by the USEPA and available from the Center for Environmental
Research Information. The Certification Officer should keep in mind that
technical assistance to the laboratory he/she is evaluating is of prime
importance because through this assistance he/she can upgrade, the laboratory
to produce better results.
Assistance to state Certification Officers can be obtained from the
regional certification authority, the Analytical Quality Control Officer in
the region, or EMSL-CI.
II. CRITICAL ELEMENTS
A. General
1. The laboratory should prepare and follow a written QA plan. It
is essential that all laboratories analyzing drinking water
10-2
-------
compliance samples adhere to defined quality assurance
procedures. This is to insure that routinely generated
analytical data are scientifically valid and defensible and are
of known and acceptable precision and accuracy. To accomplish
these goals, each laboratory should prepare a written
description of its quality assurance activities (a QA Plan).
The following items should be addressed in each QA plan.
a. Sampling procedures.
b. Sample handling procedures
- specify procedures used to maintain integrity of all
samples, i.e., tracking samples from receipt by laboratory
through analysis to disposal.
- samples likely to be the basis for an enforcement action
may require special safeguards (Chain-of-Custody
procedures).
c. Instrument or equipment calibration procedures and frequency
of their use.
d. Analytical procedures.
e. Data reduction, validation and reporting.
10-3
-------
- data reduction: conversion of raw data to ug/L,
picocuries/L, coliforms/100 mL, etc.
- validation: includes insuring accuracy of data
transcription and calculations.
- reporting: includes procedures and format for reporting
data to utilities, State officials, and USEPA.
f. Types of internal quality control checks and frequency of
their use.
- may include preparation of calibration curves, instrument
calibrations, replicate analyses, use of EMSL-provided QC
samples or calibration standards and use of QC charts.*
g. Preventive maintenance procedures and schedules.
h. Specific routine procedures used to determine data precision
and accuracy for each contaminant measured.
- precision is based on the results of replicate analyses.
*QC chart for chemistry is explained in Handbook for Analytical Quality
Control in Water and Wastewater Laboratories, EPA-600/4-79-019, March 1979.
10-4
-------
- accuracy is normally determined by comparison of results
with "known" concentrations in reagent water standards
and by analyses of water matrix samples before and after
adding a known contaminant "spike".
i. Corrective action contingencies.
- response to obtaining unacceptable results from analysis
of PE samples and from internal QC checks.
The QA plan may consist of already available standard operating
procedures (SOP's) which are approved by the laboratory director
and which address the listed items, or may be a separately prepared
QA document. Documentation for many of the listed QA plan items
can be by reference to appropriate sections of this manual, the
laboratory's SOP's or to other literature (i.e., Standard Methods
for the Examination of Water and Wastewater).
If a particular listed item is not relevant, the QA plan should
state this and provide a brief explanation (e.g., some laboratories
never collect samples and thus have no need to describe sampling
procedures). A laboratory QA plan should be concise but responsive
to the above-listed items (a maximum of five pages is suggested).
Minimizing paperwork while improving dependability and quality of
data are the intended goals.
10-5
-------
2. All quality assurance data should be available for inspection.
All quality control data must be available for inspection.
This statement assures the availability of the data. The
person certifying the laboratory might wish to make use of
these data to assure himself that the laboratory is practicing
quality control and to what extent. After an amount of data
have accumulated, it can serve as a record of a continuing type
of quality control rather than a sporadic, hit or miss type.
At any time, should there be questions on the reliability of
any data, the quality control records will be available to show
the reliability of the data produced during the time period in
question.
The guidelines for data reporting recommend that the records of
chemical analyses should be kept by the laboratory for not less
than three years. It would seem prudent that all quality
control data be kept for a like period of time.
Data required would include a record of the results of the
yearly performance sample, a standard curve for each method the
laboratory has been certified for, the records showing a check
of this curve daily or each time the analysis is carried out.
If the laboratory analyzes 20 or more samples per day, records
should include the value of a standard run after every 20
samples. Again, this is for
10-6
-------
a minimal program and it would be well for laboratories to
adopt at least the recoranended procedures listed in the
"Criteria and Procedures" manual.
3. All laboratories must analyze an unknown performance evaluation
sample once a year.
Laboratories must perform on an unknown performance sample once
per year for parameters measured.
In a minimal program this yearly check sample would be the
first external indication of a problem in a laboratory to the
certifying authority. The required daily quality control data
would not be sent to the certifying authority. If unacceptable
answers were obtained for one or more parameters, the
laboratory would be asked to analyze a follow-up performance
sample. If continued problems existed, the certifying
authority could offer some form of technical assistance to
rectify the problem. If the data are borderline or perhaps
sporadic in nature, the Certifying Officer might wish to
schedule his next visit at a time when the questionable
analytical method is being performed.
The principal state laboratory, as well as local laboratories,
will be required to analyze an unknown performance sample.
This sample will be provided by the regional authority which
will certify that
10-7
-------
laboratory in each state. The USEPA also plans to make
available to states samples which can be used as performance
samples for local laboratories which the state has
responsibility for certifying. The performing laboratory will
be given results of their analysis in terms of being within or
out of the acceptable limits.
Results must be within the control limits established by USEPA
for each analysis for which the laboratory wished to be
certified.
The laboratory will be informed if they have or have not
complied with this requirement by the authority supplying the
performance sample.
4. A manual of analytical methods should be available to the
analysts.
Usually the laboratory will provide to the certification team,
before its arrival at the laboratory, a list of the approved
methods used by the laboratory. The list should include
methods for all contaminants analyzed by the laboratory. The
laboratory should at least have copies of the approved methods
reference. That is either the USEPA's Methods for Chemical
Analysis of Water and Wastes EPA-60Q/4-79-020, the 14th ed of
Standard Methods, or others as listed in the footnotes in Table
1 of Unit 3.
10-8
-------
The laboratory could also prepare a methods manual of their own
which would gather the approved methodology of the laboratory
into one volume.
Either way, analysts should be able to have access to the
approved methods. This should hold to a minimum the deviations
that analysts might carry out if left to their own procedures.
5. Meters for pH should be calibrated each use period with fresh
standard buffers.
Good procedures for the use of a pH meter should be followed, a
good discussion is included in the Standard Methods write up of
method 424.
B. Inorganic - Critical Elements
1. A standard curve must be prepared and kept for each parameters
the laboratory analyzes for. This curve must be prepared with
a minimum of blank and three standards. The references for the
analytical methods will provide the laboratory with the range
of the test. Good procedure would dictate choosing the three
standards to cover this entire range. A high, low and
mid-range standard would be best to run at least one of the
stanards should be at or below the MCL. In order to assure
10-9
-------
good coverage the laboratory should be encouraged to run more
than the minimal requirements as listed above. A good
recommended procedure is to prepare the initial curve with a
blank and eight standards covering the entire range.
If more than one analyst will run the same test, it would be
wise to have each analyst check their procedural technique by
checking the standard curve.
2. After the initial curve has been established, it should be
verified each day on which analysis are performed by the use of
at least a reagent blank and one standard which is within the
range of the standard curve. Again, the Certification Officer
should encourage more than the required minimum daily check.
The reconmiendation for good technique recommends a blank and
two standards, one high and one low concentration.
3. These required daily checks of the standard reagent curve
should be with ± 10% of the original concentration value. For
example, if the MCL was 0.50, a standard at this level analyzed
as an unknown should fall between 0.55 and 0.45. If not, the
analyst should check in the following order:
a. Any variable instrument parameters
b. Rerun check sample.
10-10
-------
c. Prepare new standard.
d. Prepare all reagents fresh.
e. Check shelf life of chemicals.
f. Check instrument.
If the value persists at the new value through all this, then
the analyst should prepare a new standard curve.
4. If 20 or more samples per day are analyzed, the working
standard curve must be verified by running an additional
standard at or near the MCL every 20 samples. Checks must be
within * 10 percent of original curve.
C. Organic - Critical Elements
1. Laboratory method blanks must be run.
For each day on which pesticides or phenoxyacid analysis are
initiated or TTHM reagent water is preapred, it is essential
that a laboratory method blank be analyzed with the same
procedures used to analyze sanples. This will assure that no
TTHM's are present in the reagent water and further that no
10-11
-------
interfering compounds are present in the reagents used in the
pesticide or herbicide analysis.
2. A minimum of three calibration standards should be analyzed
each day to calibrate the analytical system.
It is recommended that a blank and three standards that cover
the working range be used. It would be a good idea to insert a
fourth standard that would be slightly above the working
range. Thus calibration series will indicate problems with
reagents if the blank shows contamination and will also prove
that the system is functioning properly and is linear.
If the laboratory can demonstrate that the instrument response
is linear through the origin, this practice can be reduced to
one standard per day, providing the response of the standard is
within ± 15 percent of previous calibration.
3. Certified quality control check standards should be run each
quarter for each contaminant.
If the criteria established by USEPA are not met, corrective
action needs to be taken and documented. Such standards can be
obtained from EMSL-CI.
10-12
-------
4. Field blanks for trihalomethanes should be run with each set.
Several instances of accidental sample contamination have been
noted and attributed to diffusion of volatile organics through
the septum seal and into the sanple during shipment and
storage. The field blank is used as a monitor for this problem.
Impurities contained in the purge gas and organic compounds out
gasing from the plumbing ahead of the trap usually account for
the majority of contamination problems. The presence of such
interferences are eaisly monitored as a part of the quality
control program. When a positive TTHM response is noted in the
field blank, the analyst needs to analyze a method blank in
order to ascertain if the contamination problem is indigenous
to the samples or to the instrument system. .
If reportable levels of TTHM's are demonstrated to have
contaminated the field blank resampling is essential.
5. Ten percent of all TTHM samples should be run in duplicate.
This is a normally accepted quality control technique. A
continuing record of results and subsequent actions taken needs
to be maintained.
10-13
-------
6. The laboratory needs to analyze a known TTHM laboratory control
standard each day.
This will assure the analyst that the system is in control.
However, if errors exceed 20 percent of the true value, all
TTHM results taken since the previous successful test are to be
considered suspect.
7. Each time the TTHM analytical system undergoes a major
modification or prolonged period of inactivity the precision of
the system needs to be demonstrated by the analysis of
replicate laboratory control standards.
By checking the system, such variables as the degree of trap
conditioning, the degree of desporbtion from the traping column
and the purging efficiency of each compound can be compared to
previous efficiencies. Small variations in purge time, purge
flow rate, or purge temperature can affect analytical results.
By keeping proper records of these variables the same precision
of analysis can be monitored.
8. Analyisis of TTHM by liquid-liquid extraction must include
analysis of raw source waters.
10-14
-------
Since the TTHM's are formed as part of the disinfection by
chlorination in the plant, the raw source water roust be
monitored to provide data on interfering compounds. This
liquid-liquid extraction technique efficiently extracts a wide
boiling range of nonpolar organic compounds and, in addition,
extracts the polar organic components of the sample with
varying efficiencies. The absence of peaks in the raw source
water analysis with retention times similar to the
trihalomethanes is generally adequate evidence of an
interferences - free finished drinking water analysis.
When potential interferences are noted in the raw source water
analysis, an alternate analysis or the use of alternate
chromatographic columns must be used to reanalyze the sample
set If interferences are still noted, qualitative
identifications should be performed by using dissimilar columns
or by GC/MS. If the peaks are confirmed to be other than
trihalomethanes and add significantly to the total
trihalomethane value in the finished drinking water analysis,
then the sample set must be analyzed by another method.
9. When GC/MS analysis is used it is essential that performance
tests using bromofluorobenzene be conducted once each eight-
hour work shift. This test is covered in the section on
Instrument and Equipment Needs and Specifications, Unit 12, and
in the "Criteria and Procedures" manual.
10-15
-------
III. RECOMMENDATIONS
The following items are classed in the "Criteria and Procedures" manual
as recommended. However, a certain amount of importance must be attached to
each item. The committee preparing the document felt strongly enough about
these items to keep them in the document. Common laboratory practice would
assure that these items be carried out.
1. Current Service Contract on All Balances
The analytical balance is of great importance in a laboratory. As
reagents are weighed on this piece of equipment, care must be taken
to assure that it is in good working order. The laboratory
Certification Officer should question the head chemist as to the
existance of a service contract on the balances. Should the
laboratory Certification Officer need additional information on
proper care of a balance there is a section in the Handbook for
Analytical Quality Control in Water and Wastewater Laboratories
EPA-600/4-79-019.
2. Class S weights Available to Make Periodic Checks on Balances
This could be included as part of the routine service contract or a
set purchased and shared with the bacteriological laboratory which
will also have need for them. A very conplete set of directions
for checking the performance of a balance is contained in Part 30
of ASTM Standards.
10-16
-------
3. NBS - Certified Thermometer Available to Check Thermometers in
Ovens, etc.
Again this item could be a shared item between chemical and
bacteriological laboratories. The Certification Officer could
carry this item with him and provide this service to the
smaller type laboratories. Since this item is only
recommended, the Certification Officer can only question if
this thermometer is available and used.
4. Color Standard or Their Equivalent Available to Verify
Wavelength Settings on Spectrophotometers.
The spectrophotometers should be checked for wavelength
alignment. If a particular colored solution is to be used at a
closely specified wavelength, considerable loss of sensitivity
can be encountered if the wavelength control is misaligned. In
visual instruments, an excellent reference point is the maximum
absorbance for a dilute solution of potassium permanganate,
which has a dual peak at 526 mu and 546 mu. On inexpensive
graphing instruments, which possess less resolution than the
prism instruments, the permanganate peak appears at 525 to 550
mu as a single flat-topped spike.
10-17
-------
Another point that should be mentioned is the care and use of
spectrophotometry absorption cells. If possible, the
Certification Officer should observe the techniques of the
laboratory in the use of the cells. Good techniques here could
indicate good technique in all the colorimetric procedures.
Chemical Dated Upon Receipt of Shipment and Replaced as Needed
or When Shelf Life is Exceeded.
It should not be necessary to store clean glassware or
chemicals on bench tops. Floor length cabinets or about bench
cabinets should be available for storage. Chemicals themselves
should be of analytical reagent grade to assure good quality.
Dating the chemical upon receipt will give the chief chemist an
indication of the amounts to order and if the chemical can
still be relied on to have its initial quality.
Laboratories Analyzing Water Supply Samples Other Than Its Own
Should Carry Out Additional Quality Control. This section
covers additional optional items for the larger laboratories.
a. Laboratory should perform on a known reference sample (when
available) once per quarter for the parameters measured.
10-18
-------
Since the yearly known performance sample will not indicate
to the laboratory how well it is doing, other than pass or
fail, a known sample will show how the laboratory compares
in precision and accuracy to that given for the various
methods. Analysis of the known sample will allow
comparison and show any trend of the quality control of the
laboratory. These data should be available to the
Xertification Officer for inspection.
This known quality control check sample should be available
to the laboratory from the principal state laboratory. If
not, a synthetic sample prepared by the head chemist can be
used. This control can be a large sample from a natural
source known to contain the constituents of concern or a
synthetic sanple prepared in the laboratory from chemicals
of the highest purity grade. In either case, if the
control is to be kept, it should be stabilized by addition
of a suitable preservative. See the section on sampling
for the choice of preservative.
b. The measured value should be within the control limits
established by USEPA for each analysis for which the
laboratory wishes to be certified.
10-19
-------
Precision data can be found in one or the other standard
references. That is
1) Standard Methods for the Examination of Mater and
Wastewater, 13th Edition (1971).
2) Manual of Methods for the Chemical Analysis of Water
and Wastes EPA-600/4-79-020, 1983 edition. Not
available any longer.
These data have been accumulated in Table I. If this data
does not fulfill the need of the Certification Officer, he
may write to the USEPA, EMSL, Cincinnati, Ohio 45268 and
request additional information on accuracy and precision.
c. At least one duplicate sanple should be run every 10
samples, or with each set of samples, to verify the
precision of the method. Checks should be within the
control limits established by USEPA for each analysis for
which the laboratory wishes to be certified.
In order to document that reproducible results are being
obtained (i.e. precision of the method), it is necessary to
run duplicate samples. Although the frequency of such
10-20
-------
replicate analysis is, by nature dependent on such factors
as the original precision of the method, the reliability of
the instrumentation involved and the experience of the
analyst, good laboratory technique is to run duplicate
analysis at least 10 percent of the time. The resulting
data should be within the control limits established by
USEPA. If the data do not agree, the system is not under
control, and results are subject to question.
d. Standard deviation should be calculated and documented for
all measurements being conducted.
This calculation will provide the upper and lower control
limits for the test. Analysts can then determine whether
or not the data produced is acceptable. This data can be
calculated on seven replicate determinations for initial
comparison. However, as additional determinations are
performed, they should be added to existing data and the
precision data recalculated. Twenty or more runs tend to
present better statistical data.
Standard deviation calculations should be determined for
each analyst to carry out the analysis. However, the data
should not be collected until the analyst is familiar with
the procedure. The concentration used to calculate the
10-21
-------
standard deviation should be at the level expected in the
sample for those laboratories doing only their own water.
For laboratories doing determinations other than their own
supply it would be best to have the standard deviation
calculated at several concentrations. However, for a
minimal effort, the concentration should be chosen at or
close to the maximum contaminant level for the parameter.
In order to assure these data are collected, the standard
run after each 20 samples could be at the concentration
used to determine the standard deviation This would
produce a constant flow of this data for inclusion in
future updates of the standard deviation calculation.
e. Quality control charts or a tabulation of mean and standard
deviation should be used to document the validity of data
on a daily basis.
If the upper and lower control limits of ± 2 standard
deviations are calculated, the analyst will have some idea
as to the acceptability of each determination as the
results are obtained. When outliers are found the analyst
can reschedule these for analysis to asure themselves of
the result before action is taken to call for a resample of
the supply.
10-22
-------
Production of quality control charts and subsequent
graphing of the charts of data obtained in the laboratory
will give pictorial representations of the control of the
method. Tendencies toward one or the other control limit
will indicate loss of control of the method.
How to produce quality control charts and a discussion of
these statistical tools is covered in Basic Statistics
Unit 9 and in the Handbook for Analytical Quality Control
in Water and Wastewater Laboratories EPA-600/4-79-019.
IV. SUMMARY
The quality control items in the "Criteria and Procedures" manual
identify a minimal effort for all types of laboratories. Since quality
control is for the benefit of the laboratory in assuring valid data, it
would seem wise for all laboratories to practice a good deal more quality
control than is set down in the manual.
This section has discussed the quality control steps to be taken to
assure proper analytical performance in the laboratory. However, a complete
picture of quality control would include adherence to proper sampling
techniques, including collection, preservation and handling; use of
acceptable methods, and proper reporting of data to be considered. It must
be recognized (and practiced) however, that quality control begins with
collection and does not end until resulting data are reported.
10-23
-------
UNIT 11
PREPARING A REPORT OF THE
LABORATORY CERTIFICATION SURVEY
I. INTRODUCTION
A narrative report should be jointly prepared by the entire
certification team. Each team member is responsible for the section
pertaining to his expertise. This report should contain a copy of all
information pertinent to the evaluation and its disposition, including a
copy of the on-site evaluation form used by each team member involved. The
report should also recommend analysis for which certification can be awarded
upon approval by the appropriate director of laboratory certification/QA
director and/or the director of water program.
As part of the on-site evaluation, each team member should discuss the
results of the on-site visit with the laboratory director. Some
prediscussion time should be used by the team members to assure that
complete coverage of their findings are discussed. If this is done, then
the final written report will contain as little new or undiscussed material
as possible.
When the evaluation is carried out by the team members of a regional
inspection authority, copies of the report should be sent to the Regional
Administrator for consideration. The Regional Administrator will notify the
11-1
-------
State of his concurrence or exception and forward the entire report to the
State. State evaluation of local laboratories will route the report to the
officer reponsible for State certification, who will in turn inform the
local laboratories of the results. Copies of the report should always be
retained by the Certification Officer to be used in case questions arise and
to refresh his mind before the next evaluation of the laboratory.
II. THE REPORT
Since the State accepting primacy will set up its own reporting form,
there is no set format for the report. However, each section of the report
form should be discussed clearly, stating which items are critical and which
are recommendations.
A. Personnel (Recommendation)
Under this topic the Certification Officer could summarize those
personnel available to the laboratory at the present time,
pointing out, if necessary, what is required to upgrade the
personnel to meet the guidelines as stated in the "Criteria and
Procedures" manual. If a serious understaffing exists in a
laboratory, the Certification Officer should recommend the
additional personnel needed by the laboratory in this section.
The Certification Officer could also recommend any training he
11-2
-------
feels would benefit the personnel to help them to better perform
their duties. In this section the certification officer might
designate which personnel are deemed essential and for which
notification of the certifying authority is necessary if loss or
replacement occurs.
Laboratory Facilities (Recommendation)
Discuss all topics showing where improvements or changes are
suggested. For example, one of the more important items covered
under this topic is the source of distilled/deionized water. The
inspector might suggest a larger capacity or different type or
better analysis to assure quality.
Laboratory Equipment and Instrument Specifications (Recommendation)
Although classified as a recommendation this area is important.
If a lack of equipment is evident or if the instrument is not
approved for the analysis it is being used for, this would relate
directly to the methodology which is a mandatory section. The
laboratory evaluator would also point out the items which might
need to be updated or added to.
General Laboratory Practices (Recommendation)
This section would be important from the standpoint of quality
11-3
-------
assurance. Improper glassware preparation could indicate a future
trouble area in analytical analysis.
E. Methodology (Critical)
Since the methods to be used for the analysis of the various
contaminants have been set down in the Federal Register, this
section is mandatory until changed or until an alternate method
has been approved by the proper authority. The acceptable
references at this time are listed in tables I and II in the
Methodology Outline (number 3). The Certification Officer should
be sure to obtain all the pertinent information on this section.
F. San^le Collecting, Handling and Preservation (Critical)
This section must be looked at in depth by each member of the
certification team. The proper sampling techniques may be carried
out by someone other than the laboratory personnel, however, the
sanpling techniques must be known or the possibility of improper
analytical results exists. However, it is the responsibility of
the laboratory to refuse to analyze or to require a new sample if
it is known or suspected that the sample was improperly collected
or preserved. The preservation technique and sanpling containers
can be evaluated by the Certification Officer to assure compliance.
11-4
-------
The location of sample collection is spelled out in the National
Interim Primary Drinking Water Regulations as "the free flowing
outlet of the ultimate user of a public water system." The
required sample container and preservative as well as the holding
times can be found in the "Criteria and Procedures" manual.
However, each reference method should be checked for special
precautions.
6. Quality Control (Critical and Recommended)
The intent of this section is to have the Certification Officer
review enough data to assure that the analytical data being
developed is of a reliable nature. The critical items are
sufficient to provide a minimal effort of quality control. The
Certification Officer should carefully review the data available
to be sure that at least a minimal effort is being carried out by
the laboratory. If not, or if it is deemed desireable, comments
should be made here. Strong encouragement should be given to
having more quality control than just a minimal effort.
H. Data Handling (Recommendation)
This section is used as a recommendation to be sure that enough
information about the sample is being retained for a sufficient
11-5
-------
time to cover any questions that might arise. It should be
pointed out here that in the implementation section under other
considerations it is stated that for potential enforcement actions
only the chain of custody for a sanple must be maintained and
recorded from time of sampling through analysis to the final
recording. Consequently, additional data handling would be
required in this case. The laboratory should be aware of this
provision and what additional information or procedures are needed.
III. LEVELS OF CERTIFICATION
The Operational Guidance section of the implementation of the
"Criteria and Procedures" manual lists three levels of certification under
the criteria procedures. Although these are listed for principal State
laboratories, something similar could be set up for local laboratories. The
three levels are:
A. Certified - a laboratory that meets the minimum requirements
listed in the technical manuals for specific parameters. A
laboratory may be certified to perform only those analyses for
which certification is requested. For example, a laboratory may
wish to analyze for coliform bacteria and not for other
contaminants. The certification should be for three years.
11-6
-------
B. Provisionally certified - a laboratory that has been certified
subject to correction of deficiencies that do not impact the basic
validity of the test results. A grace period up to one year could
be awarded for deficiencies, but in no case should provisional
certification be given when the laboratory does not have the
capability, in the opinion of the evaluator, of performing the
analysis.
C. Not certified - a laboratory that does not have the capability of
performing the analyses as determined by "Criteria and Procedures"
manual. A laboratory would be placed in this category for an
analysis only after the State has had an opportunity for a review
of the decision and the Regional Administrator has. allowed for
"due process." Should the Regional Administrator uphold the "not
certified" classification on a principal State laboratory after
the hearing, the State must immediately correct the major
deficiencies noted and then request reinspection, or the State can
request inspection of another laboratry to perform the assigned
analytical work.
SUMMARY
The laboratory approval and certification procedure as developed by
11-7
-------
the Administrator or his designee is part of the procedure for the
determination of primary enforcement responsibility. Since the withdrawal
of laboratory approval or certification may affect primary enforcement
responsibility, the procedure shall include notice and the opportunity for a
hearing as provided in 40 CRF 142.12 and 142.13.
11-8
-------
UNIT 12
INSTRUMENT AND EQUIPMENT NEEDS AND SPECIFICATIONS
I. INTRODUCTION
The intent of the Safe Drinking Water Act was to place the
responsibility of analysis on the treatment facility personnel (40CFR No.
248, December 24, 1975, pg. 59581). However, some states have elected to do
the analysis of chemical parameters themselves. This arrangement gives rise
to several possibilities: a) the treatment facility will do the analysis;
b) the treatment facility will contract with an outside laboratory to do the
analysis; and finally; c) that the state laboratory will do the analysis.
In the first two cases the laboratories would have to be certified by
the state as being capable of doing the analysis. A further possibility
could arise when a treatment facility elects to do some, but not all of the
analysis and would request certification only for those parameters for which
it would be doing analysis.
When a laboratory is to be certified for chemical analysis, it will
need only that equipment necessary to carry out the analysis for those
parameters which it has chosen to do. The "Criteria and Procedures" manual
lists the capital equipment needed for a laboratory doing all the analysis.
12-1
-------
A laboratory doing only one or a selected number of analyses would want to
know what was needed for the methods chosen. This outline will list the
equipment needed for each analysis. The specification for the capital
equipment will not be repeated each time but referenced to the first mention
of that piece of equipment.
In the case of a laboratory which has a large volume of analyses, the
number of pieces of equipment should be taken into account. This would
insure sufficient equipment to carry out their workload.
II. INORGANIC
A. Metals run by direct aspiration or furnace analysis
1. Atomic Absorption Spectrophotometer: An instrument having
either a single or double beam design with a grating
monochromator, photomultiplier detector, adjustable slits, and
a wavelength range of at least 190 to 800 nm should be used.
In addition, an appropriate readout system that has a response
time capable of measuring the atomic absorption signal
generated is required. This includes the capability to detect
positive interference on the signal from intense non-specific
absorption.
The system should have sufficient equipment such as regulators,
tubing and gas controls to adequately monitor and deliver the
proper rate of oxidant and fuel required by the method.
12-2
-------
Sufficient hollow cathode lamps or electrodless discharge lamps
should be available.
2. Burner: The instrument should be equipped with a burner
recommended by the manufacturer for the method. The instrument
should also be vented about 15 to 30 cm (6 to 12 inches) above
the burner to remove fumes and vapors.
3. Fuel and oxidant: nitrogen or argon, hydrogen, acetylene, air
and nitrous oxide gases are needed.
4. Recorder: Having a chart width of 10 inches or 25 cm, a full
scale response time of 0.5 sec or less, 10 or 100 mv input to
match the instrument and a variable chart speed of 5 to 50
cm/min or equivalent. In furnace work the recorder must be
used to verify adequate background correction if a CRT video
readout or hard copy printer is not available.
5. Graphite Furnace: Any furnace device capable of reaching the
specified temperatures is satisfactory. In addition a supply
of graphite platforms and/or tubes should be available in
pyrolytic and nonpyrolytic forms. Also an inert gas, argon, is
needed for furnace analysis. The analyst should also have
microliter pipets with disposable tips. Sizes should be
available from 5 to 100 uL. Yellow colored tips have
12-3
-------
been found to contain Cadmium. White colored tips have been
found to be acceptable for all analyses. If a large volume of
analyses are expected an auto sampler is recommended. This can
even provide better precision in the analysis.
A background correction system or provision for a subsequent
analysis using a nonabsorbing line is required for furnace
analysis.
B. Metals run by Inductively Coupled Plasma (ICP)
1. ICP instrument: either sequential or simultaneous.
2. Spare parts: a power tube for the RF generator, quartz torch,
some pump tubing, photomultiplyer tubes and computer disks and
paper.
3. Argon gas: either liquid or dry gas.
4. A peristaltic pump: capable of at least 2 mL/min.
5. Vent for torch in order to remove toxic fumes.
6. The room should be temperature and humidity controlled.
12-4
-------
C. Arsenic and Selenium by Gaseous Hydride
An atomic absorption instrument with argon and hydrogen gases and
some form of readout device is needed. The following are also
necessary.
1. Flow Meter: Capable of measuring 1 liter/minute such as
Gilmont No. 12 or equivalent.
2. Medicine Dropper: Capable of delivering 1.5 mL; fitted into a
size "0" rubber stopper.
3. Reaction Flask: A pear-shaped vessel with two side arms and 50
mL capacity, both arms having a ground glass T 14/20 joint
(such as Scientific Glass part OM-5835 or equivalent).
4. Special Gas Inlet-Outlet Tube: Constructed from a micro cold
finger condenser by cutting off that portion below the ground
glass T 14/20 joint (such as Scientific Glass part JM-3325 or
equivalent).
5. Hollow Cathode or Electrodeless Discharge Lamp.
6. Drying Tube: 100-mm long polyethylene tube filled with glass
wool.
12-5
-------
7. Arrangement: As in Figure I below:
-V
JM-3325
Medicine
Dropper in
Size "0"
Rubber
Stopper
Argon
Flow
Meter
Drying I -—I
Tube r~ L,
K=>
Hydrogen
(Fuel)
(Auxiliaiy Air)
Argon
— (Nelmli/rr
. Air)
AA Burner Body
JM-5835
Fig. 1. Schematic Arrangement of Equipment for
Determination of Arsenic and Selenium
12-6
-------
Arsenic by Silver Diethydithiocarbamate
1. Arsine generator: A Gutzeit generator (Fisher Scientific Co.
No. 01-405) or equivalent used in conjunction with an absorber
tube or assembly.
2. Photometer: Spectrophotometer or filter photometer, capable of
measuring from 400 - 700 nm. Should be capable of using
several sizes and shapes of absorption cells providing a path
length of approximately 1 to 5 cm. The photometer should have
a maximum spectral band width, of no more than 20 nm and a
accuracy of ± 2.5 nm.
Mercury
1. Manual cold vapor using an atomic absorption unit:
a. Absorption Cell: Cells 2.5 cm in diameter and from 10 to 15
cm in length windows may be used. Suitable cells may be
constructed from plexiglass tubing, 2.5 cm (1 inch) O.D.
The ends are ground perpendicular to the longitudinal axis
and quartz windows (2.5 cm (1 inch) diameter x 0.159 cm
(1/16 inch) thickness) are cemented in place. Gas inlet and
12-7
-------
outlet ports (also from plexiglass but .06 cm (1/4 inch]
Q.D.) are attached approximately 1.27 csn 11/2 inch) from
each end. The cell is strapped to a burner for support and
aligned to the light beam.
b. Air Pump: A peristaltic pimp capable of delivering an air
flow of 1 or 2 liter/min may be used. Any other regulated
compressed air system that is clean and dry (including
cylinder air) is satisfactory.
c. Flow Meter: Capable of measuring an air flow of 1 or 2
liters/min.
d. Aeration Tubing: A straight glass frit having a coarse
porosity. Tygon tubing can be used for passage of the
mercury vapor from the sample bottle to the absorption cell
and return.
e. Drying Tube: A tube 15.2 cm {6 inches) x 1.9 era (3/4 inch)
diameter tube containing 20 g of magnesium perchlorate my
be used, [n, place of the drying tube a small reading lamp
tfith a 60 w bulb can be used to prevent condensation of
moisture inside the cell. The lamp is positioned to shine
on the absorption cell maintaining the air temperature in
the cell about 10° C above ambient.
12-8
-------
f. Mercury Vapor Trap: Because of the toxic nature of mercury
vapor a by-pass should be included in the system to either
vent the mercury vapor to an exhaust hood or pass the vapor
through some absorbing media.
Liquid type, such as, equal volumes of 0.1 M KMnO^ and 10
percent HgSO^ or 2,25 percent iodine in a 3 percent KI
solution or
Solid type, such as, a specially treated charcoal that will
absorb mercury vapor - available from Barnebey and Cheney,
E. Eight Avenue and N Cassidy Street, Columbus, Ohio 43219,
Cat. #580-22 or from Coleman Instruments, 42 Madison Street,
Maywood, Illinois 60153, Cat. #50-160.
12-9
-------
g. Arrangement: As in Figure 2 below.
o
AIR PUMP
DESICCANT
A ABSORPTION
0 BUBBLER CELL
~
SAMPLE SOLUTION
IN BOD BOTTLE
SCRUBBER
CONTAINING
A MERCURY
ABSORBING
MEDIA
Fig. 2. Apparatus for Flameless Mercury Determination
12-10
-------
2. Manual cold vapor using an instrument designed specifically for
mercury:
a. An instrument designed specifically for the measurement of
mercury using the cold vapor technique may be substituted
for the atomic absorption spectrophotometer. An instrument
such as the Coleman MAS-50, or its equivalent may be used.
b. Drying Tube: See specifications on previous page.
c. Mercury Vapor Trap: See l.f. above.
3. Automated Cold Vapor Method:
a. Atomic absorption instrument or instrument specifically
designed for mercury. See 2.a.
b. Autoanalyzer or equivalent: Exact equipment used is
specified by the individual method, should include:
sampler, proportioning pump, manifold or anaytical
cartridge, boiling bath with two distillation coils, vapor
liquid separator, recorder.
12-11
-------
F. Fluoride
The following items are required with this method.
1. Electrode Method:'
a. Electrode: The electrode method does not require
distillation when used to measure fluoride in finished
drinking waters.
b. pH Meter: Must have expanded mV scale, an accuracy and
readability of ± lmV. Laboratories purchasing a new pH
meter are strongly advised to purchase one capable of
functioning with specific ion electrodes. Unit may be
line/bench or battery/portable operated.
or
c. Specific ion meter: Readable and accurate to ± lmV, unit
may be line or battery operated.
12-12
-------
Fig. 3. Simplified Distillation Apparatus
-------
Fig. 4. Distillation Apparatus
-------
d. Electrodes: Fluoride and reference electrode or combination
type.
e. Magnetic mixer: variable speed with Teflon-coated stirring
bar.
2. SPAONS method with manual distillation: The following items
are required.
a. Photometer - spectrophotometer or filter photometer - see
previous specifications (Unit 2, 0.2).
b. Distillation equipment: See Figures three and four.
3. Automated methods with distillation: Autoanalyzer or
equivalent with: Sampler, manifold, proportioning pump,
distillation module, recorder and electrode flowthrough cell
for the potentiometric procedure. The equipment is specified
in the individual method.
12-13
-------
G. Nitrate
1. Brucine:
a. Spectrophotometer or filter photometer: suitable for
measuring absorbance at 410 nm. With specifications as
previously covered.
b. Water bath: stirred with gabled lid, for use at 100°C.
c. Water bath: for use at 10-15°C.
2. Manual cadmium reduction:
a. Spectrophotometer: For use at 540 nm and a path length of 1
cm or longer with specifications as previously stated.
b. Reduction Column: The column in Figure 5 was constructed
from a 100-mL pipet by removing the top portion. This
column may also be constructed from two pieces of tubing
joined end to end. A 10 cm length of 3 cm I.D. tubing is
joined to a 25 cm length of 3.5 mm I.D. tubing.
12-14
-------
Fig. 5. Reduction Column
-------
3. Autoanalyzer or equivalent:
a. With sanpler, manifold, pump, colorimeter, recorder, a
column for the auto cadmium reduction method, plus a heating
bath and continuous filter for the auto hydrazine method.
Set up is specified in each method.
4. Electrode methods:
a. pH meter or specific ion meter: As specified previously.
b. Electrode and reference electrode or combination: As
specified in specific method.
c. Magnetic stirrer: Variable speed with Teflon coated
stirring bar.
H. Turbidity
1. Nephelometric
a. Nephelometer instrument: Ratio types are permissible but
should be conpared to conventional nephelometers. Measures
scattered light intensity at a 90° angle to the axis of the
transmitted light. Sealed liquid turbidity standards are
acceptable if calibrated against prepared formazin standards
at least every four months. Solid turbidity standards are
not acceptable.
12-15
-------
b. Standards: Formazine or styrene divinyl benzene polymer
standards.
I. Chlorine
1. Test kit: Using DPD or syringaldazine procedures.
2. pH meter or specific ion meter plus specified electrodes for
potentiometric method.
3. Photometer as previous specified.
J. Calcium
Titrimetric: Routine laboratory glassware.
K. Sodium
1. Flame photometric method.
a. Flame photometer: either direct reading or internal
standard type for use at 589 nm.
L. Total filterable residue
Analytical balance: capable of weighing to 0.1 mg.
Glass fiber discs: 4.7 cm or 2.1 cm without organic binder.
12-16
-------
Organic Methods
A. Chlorinated hydrocarbons and chlorophenoxys
1. Gas Chromatograph: A commercial or custom designed gas
chromatograph with a column oven capable of isothermal
temperature control to at least 220 ± 0.2 C. The system should
be equipped with accurate needle-valve, gas-flow controls,
accept 1/4 inch glass columns with the option of direct
on-column injection. The system must be demonstrated to be
suitable for chlorinated hydrocarbon pesticides with a minimum
of decomposition and loss of compounds of interest. Equipped
with a glass lined injection port and either an electron
capture, microcoulometric titration or electrolytic
conductivity detector.
2. Recorder for gas chromatrograph: Strip chart recorder, having
a chart width of 10 inches or 25 cm, a full scale response time
of 1 sec. or less, 1 mv (-0.05 to 1.05) signal to match the
instrument and variable chart speeds of 5, 10, 25, or 50 mm per
min., or equivalent.
3. Kuderna-Danish (K-D) glassware (Kontes)
Snyder Column - three-ball (macro) and two-ball (micro)
Evaporative Flasks - 500 mL
Receiver Ampules - 10-mL graduated
Anpule Stoppers
12-17
-------
4. Water bath: Electric or steam heated capable of temperature
control to within 5°C with a maximum temperature of 100#C.
Concentric ring or other cover is required to support K-D
concentrators.
B. TTHM's:
1. Purge and Trap Method
a. Gas chromatograph: As above equipped with temperature
program from 45#C to 220°C at about 8°C/min and with either
microcoulometric titration or electrolytic conductivity
detector.
b. Recorder: as specified above.
c. Purge and trap system: A commercial or custom designed
system can be used. When interfaced to a compatible gas
chromatograph, the assembly should be able to detect 0.5
12-18
-------
ug/L of each of the individual TTHM's and measure them with
a reproducibility not to exceed 8 percent relative standard
deviation at 20 ug/L. The device should be designed for a 5
ml. sanple volume. The gas inlet should disperse finely
divided gas bubbles through the sample. The trap should be
capable of heating the trapping device to 180°C in one
minute with less than 40°C overshoot.
2. Liquid/liquid Extraction:
a. Gas chromatograph and recorder as specified above. The gas
chromatograph should be equipped with a linearized
(frequency modulated) electron capture detector.
3. Gas chromatography/mass spectrometry/data system (GC/MS/DS):
a. The GC must be capable of temperature programming. Any
column (either packed or capillary) that provides data with
adequate accuracy and precision (Sect. 10) can be used. If
a packed column is used, the GC usualy is interfaced to the
MS with an all-glass enrichment device and an all-glass
transfer line, but any enrichment device or transfer line
can be used if performance specifications described in this
method can be demonstrated with it. If a capillary column
is used, an enrichment device is not needed. A recommended
packed GC column for the listed analytes is 1.8 m long by 2
12-19
-------
mm ID glass packed with 1% SP-1Q00 on 60/80 mesh Carbopack
B. Recommended operating parameters for that column are:
helium carrier gas flow rate of 30 mL/min and temperature of
48°C for 4 minute, increased to 230°C for at least 25 minute
or until all expected analytes elute. An alternative
recommended packed column is 1.8 m long by 2 mm ID glass or
stainless steel packed with 0.2% Carbowax 1500 on 80/100
mesh Carbopack C.
b. Mass spectral data are obtained with electron-impact
ionization at a nominal electron energy of 70 eV. The mass
spectrometer must be capable of scanning from 35 to 450 amu
every 7 s or less and must produce a mass spectrum that
meets all criteria in Table 1 when 50 ng or less of
jj-bromofluorobenzene (BFB) is introduced into the GC. To
ensure sufficient precision of mass spectral data, the
desirable MS scan rate allows acquisition of at least five
spectra while a sample component elutes from the GC. With
capillary columns which produce narrower peaks than packed
columns that criterion may not be feasible and adequate
precision with fewer spectra per GC peak must be
demonstrated.
c. An interfaced data system (DS) is required to acquire,
store, reduce and output mass spectral data. The computer
12-20
-------
software must allow searching any. GC/MS data file for ions
of a specific mass and plotting ion abundances versus time
or scan number. This, type of plot is defined as an
extracted ion current profile (EICP). Software must also
allow integrating the abundance in any EICP between
specified time or scan number limits. A hard copy device is
necessary for data output and archiving.
For each analyte, the mean accuracy should be in the range
of 85 to 115 percent. Adequate precision is obtained when
the relative standard deviation is ± 20 percent. For some
listed analytes, this may not be feasible for low
concentration measurements.
p-Bromofluorobenzene Key Ions and Ion Abundance Criteria
Table V
Mass
Ion Abundance Criteria
50
75
95
96
173
174
175
176
177
15 to 40% of mass 95
30 to 60% of mass 95
base peak, 100% relative abundance
5 to 9% of mass 95
less than 2% of mass 174
greater than 50% of mass 95
5 to 9% of mass 174
96 to 100% mass 174
5 to 9% of mass 176
12-21
-------
IV. General Lab Equipment
In addition to the equipment listed for each method the following
equipment should also be available.
A. Conductivity meter: Useful for checking distilled water quality.
Instrument should be readable in ohms or umhos, have a range of
2-2,500,000 ohms or umhos ± 1 percent and have a sensitivity of
0.33 percent or better. The unit may be line/bench or
battery/portable operated if the above specifications are met.
B. Drying oven: Gravity and mechanical convention units with
selectable temperature control from room to 170°C or higher are
satisfactory.
C. Desiccator: Glass or plastic models may be appropriate depending
on the particular application.
D. Hot plate: Large or small units with selectable temperature
control for safe heating of laboratory reagents.
E. Refrigerator: For aqueous reagent and sample storage a standard
kitchen type domestic refrigerator will be sufficient. For storing
organics and flammable materials an "explosion proof" type of
refrigerator should be used.
12-22
-------
F. Glassware: All glassware purchased for laboratory use should be
composed of Pyrex or Kimax type glass. This type of glass is more
resistant to damage by heatt chemicals, and abuse than is regular
soft glass. All volumetric glassware should be marked Class A,
denoting that it meets federal specifications for volumetric
glassware and need not be calibrated before use.
12-23
-------
Unit 13
LABORATORY SAFETY PRACTICES
I. INTRODUCTION
While this is not an aspect of laboratory certification, evaluators
should point out, on an informal basis, potential safety problems observed
during an on-site visit. This outline is added to point out some coiranon
laboratory safety practices.
A. Safe Use, Handling, and Storage of Chemicals
1. Chemicals in any form can be safely stored, handled, and used
if their hazardous physical and chemical properties are fully
understood and the necessary precautions, including the use of
proper safeguards and personal protective equipment are
observed.
2. The management of every unit within a laboratory must give
wholehearted support to a well integrated safety policy.
B. General Rules for Laboratory Safety
1. Supervisory personnel should think "safety." Their attitude
toward fire and safety standard practices is reflected in the
behavior of their entire saff.
13-1
-------
2. A safety program is only as strong as the worker's will to do
the correct things at the right time.
3. The fundamental weakness of most safety programs lies in too
much lip service to safety rules and not enough action in
putting them into practice.
4. Safety practices should be practical and enforceable.
5. Accident prevention is based on certain common standards of
education, training of personnel and provision of safeguards
against accidents.
LABORATORY DESIGN AND EQUIPMENT
A. Type of Construction
1. Fire-resistant or noncombustible.
2. Multiple story buildings should have adequate means of exit.
3. Stairways enclosed with brick or concrete walls.
13-2
-------
4. Laboratories should have adequate exit doors to permit quick,
safe escape in an emergency and to protect the occupants from
fires or accidents in adjoining rooms. Each room should be
checked to make sure there is no chance of a person being
trapped by fire, explosions, or release of dangerous gases.
5. Laboratory rooms in which most of the work is carried out with
flammable liquids or gases should be provided with
explosion-venting windows.
B. Arrangement of Furniture and Equipment
1. Furniture should be arranged for maximum utilization of
available space and should provide working conditions that are
efficient and safe.
2. Aisles between benches should be at least 4 feet wide to
provide adequate room for passage of personnel and equipment.
3. Desks should be isolated from benches or adequately protected.
4. Every laboratory should have an eyewash station and a safety
shower.
13-3
-------
C. Hoods and Ventilation
1. Adequate hood facilities should be installed where work with
highly toxic or highly flammable materials are used.
2. Hoods should be ventilated separately and the exhaust should be
terminated at a safe distance from the building.
3. Make-up air should be supplied to rooms or to hoods to replace
the quantity of air exhausted through the hoods.
4. Hood ventilation systems are best designed to have an air flow
of not less than 60 linear feet per minute across the face of
the hood, with all doors open and 150, if toxic materials are
involved.
5. Exhaust fans should be spark-proof if exhausting flammable
vapors and corrosive resistant if handling corrosive fumes.
6. Controls for all services should be located at the front of the
hood and should be operable when the hood door is closed.
7. All laboratory rooms should have the air changed continuously
at a rate depending on the materials being handled.
13-4
-------
D. Electrical Services
1. Electrical outlets should be placed outside of hoods to afford
easy access and thus protect them from spills and corrosion by
gases.
2. Noninterchangeable plugs should be provided for multiple
electrical services.
3. Adequate outlets should be provided and should be of the
three-pole type to provide for adequate grounding.
E. Storage
1. Laboratories should provide for adequate storage space for
mechanical equipment and glassware which will be used regularly.
2. Flammable solvents should not be stored in glass bottles over
one liter in size. Large quantities should be stored in metal
safety cans. Quantities requiring containers larger than one
gallon should be stored outside the laboratory.
3. Explosion proof refrigerators should be used for the storage of
highly volatile and flammable solvents.
13-5
-------
4. Cylinders of compressed or liquified gases should not be stored
in the laboratory.
F. Housekeeping
1. Housekeeping plays an important role in reducing the frequency
of laboratory accidents. Rooms should be kept in a neat
orderly condition. Floors, shelves, and tables should be kept
free from dirt and from all apparatus and chemicals not in use.
2. A cluttered laboratory is a dangerous place to work.
Maintenance of a clean and orderly work space is indicative of
interest, personal pride, and safety-mindedness.
3. Passageways should be kept clear to all building exits and
stairways.
4. Metal containers should be provided for the disposal of broken
glassware and should be properly labeled.
5. Separate approved waste disposal cans, should be provided for
the disposal of waste chemicals.
6. Flammable liquids not miscible with water and corrosive
materials, or compounds which are likely to give off toxic
vapors should never be poured into the sink.
13-6
-------
G. Fire Protection
1. Laboratrory personnel should be adequately trained regarding
pertinent fire hazards associated with their work.
2. Personnel should know rules of fire prevention and methods of
combating fires.
3. Fire extinguishers (CC^ type) should be, provided at
convenient locations and personnel should be instructed in
their use.
4. Automatic sprinkler systems are effective for the control of
fires in chemical laboratories.
H. Alarms
1. An approved fire alarm system should be provided.
2. Wherever a hazard of accidental release of toxic gases exists,
a gas alarm system to warn occupants to evacuate the building
should be provided.
3. Gas masks of oxygen or compressed air type should be located
near exists and selected personnel trained to use them.
13-7
-------
HANDLING GLASSWARE
A. Receiving, Inspection and Storage
1. Packages containing glassware should be opened and inspected
for cracked or nicked pieces, pieces with flaws that may become
cracked in use, and badly shaped pieces.
2. Glassware should be stored on well-lighted stockroom shelves
designed and having a coping of sufficient height around the
edges to prevent the pieces from falling off.
B. Laboratory Practices
1. Select glassware that is designed for the type of work planned.
2. To cut glass tubing or a rod, make a straight clean cut with a
cutter or file at^the point where the piece is to be severed.
Place a towel over the piece to protect the hands and fingers,
then break the tubing or rod by snapping with a motion directed
away from the body.
3. Large size tubing is cut by means of a heated nichrome wire
looped around the piece at the point of severance.
13-8
-------
4. When it is necessary to insert a piece of glass tubing or a rod
through a perforated rubber or cork stopper, select the correct
bore so that the insertion can be made without excessive strain.
5. Use electric mantels for heating distillation apparatus, etc.
6. To remove glass splinters, use a whisk broom and dustpan. Very
small pieces can be picked up with a large piece of wet cotton.
GASES AND FLAMMABLE SOLVENTS
A. Gas Cylinders
1. Large cylinders must be securely fastened so that they cannot
be dislodged or tipped in any direction.
2. Connections, gauges, regulators or fittings used with other
cylinders must not be interchanged with oxygen cylinder
fittings because of the possibility of fire or explosion from a
reaction between oxygen and residual oil in the fitting.
3. Return empty cylinders promptly with protective caps replaced.
13-9
-------
B. Flammable Solvents
1. Store in designated areas well ventilated.
2. Flash point of a liquid is the temperature at which it gives
off vapor sufficient to form an ignitible mixture with the air
near the surface of the liquid or within the vessel used.
3. Ignition temperature of a substance is the minimum temperature
required to initiate or cause self-sustained combustion
independently of the heating or heated element.
4. Explosive or flammable limits. For most flammable liquids,
gases and solids there is a minimum concentration of vapor in
air or oxygen below which propagation of flame does not occur
on contact with a source of ignition. There is also a maximum
proportion of vapor or gas in air above which propagation of
flame does not occur. These limit mixtures of vapor or gas
with air, which if ignited will just propagate flame, are known
as the "lower and higher explosive or flammable limits."
5. Explosive Range. The difference between the lower and higher
explosive or flammable limits, expressed in terms of percentage
of vapor or gas in air by volume is known as the "explosive
range."
13-10
-------
6. Vapor Density is the relative density of the vapor as compared
with air.
7. Underwriter's Laboratories Classification is a standard
classification for grading the relative hazard of the various
flammable liquids. This classification is based on the
following scale:
Ether Class 100
Gasoline Class 90 - 100
Alcohol (ethyl) Class 60 - 70
Kerosene Class 30 - 40
Paraffin Oil Class 10 - 20
8. Extinguishing agents.
V. CHEMICAL HAZARDS
A. Acids and Alkalies
1. Some of the most hazardous chemicals are the "strong" or
"mineral" acids such as hydrochloric, hydrofluoric, sulfuric
and nitric.
13-11
-------
2. Organic acids are less hazardous because of their comparatively
low ionization potentials. However, such acids as phenol
(carbolic acid), hydrocyanic and oxalic are extremely hazardous
because of their toxic properties.
3. Classification of acids.
B. Oxidizing Materials
1. Such oxidizing agents as chlorates, peroxides, perchlorates and
perchloric acid, in contact with organic matter can cause
explosions and fire.
2. They are exothermic and decompose rapidly, liberating oxygen
which reacts with organic compounds.
3. Typical hazardous oxidizing agents are:
Chlorine Dioxide
Sodium Chlorate
Potassium Chromate
Chromium Trioxide
Perchloric Acid
13-12
-------
C. Explosive Power
1. Many chemicals are explosive or form compounds that are
explosive and should be treated accordingly.
2. A few of the more common examples of this class of hazardous
materials are:
Acetyl ides
Silver Fulminate
Peroxides
Peracetic Acid
Nitroglycerine
Picric Acid
Chlorine and Ethylene
Sodium Metal
Calcium Carbide
D. Toxicity
1. Laboratory chemicals improperly stored or handled can cause
injury to personnel by virtue of their toxicity.
13-13
-------
2. Types of exposure. There are four types of exposure to
chemicals:
a. Contact with the skin and eyes
b. Inhalation
c. Swallowing
d. Injection
PRECAUTIONARY MEASURES
A. Clothing and Personal Protective Equipment
1. Chemical laboratories should have special protective clothing
and equipment readily available for emergency use and for
secondary protection of personnel working with hazardous
materials.
2. Equipment should be provided for adequate:
a. Eye protection
b. Body protection
c. Respiratory protection
13-14
-------
d. Foot protection
e. Hand protection
B. Bodily Injury
1. Burns, eye injuries, and poisoning are the injuries with which
laboratory people must be most concerned.
2. First emphasis in the laboratory should be on preventing
accidents. This means observing all recognized safe practices
using necessary personal protective equipment and exercising
proper control over poisonous substances at the source of
exposure.
3. So that a physician can be summoned promptly, every laboratory
should have posted the names, telephone numbers, and addresses
of doctors to be called in an emergency requiring medical care.
REFERENCES
Guide for Safety in the Chemical Laboratory, the General Safety Coiranittee of
the Manufacturing Chemists Association, Inc., Van Nostrand, New York (1954).
13-15
------- |