C3-A3

Vol. 17 No. 18
Replaces C3-A2
October 1997 Vol. 11 No. 13

Preparation and Testing of Reagent Water in the Clinical

Laboratory; Approved Guideline—Third Edition

This document provides guidelines on water purified for clinical laboratory use; methods for monitoring
water quality and testing for specific contaminants; and water system design considerations.

ABC
NCCLS...
Serving the World's Medical Science Community Through Voluntary Consensus
NCCLS is an international, interdisciplinary, nonprofit,
standards-developing and educational organization that
promotes the development and use of voluntary consensus
standards and guidelines within the healthcare community. It
is recognized wordwide for the application of its unique
consensus process in the development of standards and
guidelines for patient testing and related healthcare issues.
NCCLS is based on the principle that consensus is an
effective and cost-effective way to improve patient testing
and healthcare services.
In addition to developing and promoting the use of voluntary
consensus standards and guidelines, NCCLS provides an
open and unbiased forum to address critical issues affecting
the quality of patient testing and health care.
PUBLICATIONS
An NCCLS document is published as a standard, guideline,
or committee report.
Standard A document developed through the consensus
process that clearly identifies specific, essential requirements
for materials, methods, or practices for use in an unmodified
form. A standard may, in addition, contain discretionary
elements, which are clearly identified.
Guideline A document developed through the consensus
process describing criteria for a general operating practice,
procedure, or material for voluntary use. A guideline may be
used as written or modified by the user to fit specific needs.
Report A document that has not been subjected to con-
sensus review and is released by the Board of Directors.
CONSENSUS PROCESS
The NCCLS voluntary consensus process is a protocol
establishing formal criteria for:
! The authorization of a project
! The development and open review of documents
! The revision of documents in response to comments by
users
! The acceptance of a document as a consensus standard
or guideline.
Most NCCLS documents are subject to two levels of
consensus–"proposed" and "approved." Depending on the
need for field evaluation or data collection, documents may
also be made available for review at an intermediate (i.e.,
"tentative") consensus level.
Proposed An NCCLS consensus document undergoes the
first stage of review by the healthcare community as a
proposed standard or guideline. The document should receive
a wide and thorough technical review, including an overall
review of its scope, approach, and utility, and a line-by-line
review of its technical and editorial content.
Tentative A tentative standard or guideline is made
available for review and comment only when a recommended
method has a well-defined need for a field evaluation or
when a recommended protocol requires that specific data be
collected. It should be reviewed to ensure its utility.
Approved An approved standard or guideline has achieved
consensus within the healthcare community. It should be
reviewed to assess the utility of the final document, to
ensure attainment of consensus (i.e., that comments on
earlier versions have been satisfactorily addressed), and to
identify the need for additional consensus documents.
NCCLS standards and guidelines represent a consensus
opinion on good practices and reflect the substantial
agreement by materially affected, competent, and interested
parties obtained by following NCCLS’s established
consensus procedures. Provisions in NCCLS standards and
guidelines may be more or less stringent than applicable
regulations. Consequently, conformance to this voluntary
consensus document does not relieve the user of
responsibility for compliance with applicable regulations.
COMMENTS
The comments of users are essential to the consensus
process. Anyone may submit a comment, and all comments
are addressed, according to the consensus process, by the
NCCLS committee that wrote the document. All comments,
including those that result in a change to the document
when published at the next consensus level and those that
do not result in a change, are responded to by the committee
in an appendix to the document. Readers are strongly
encouraged to comment in any form and at any time on any
NCCLS document. Address comments to the NCCLS
Executive Offices, 940 West Valley Road, Suite 1400,
Wayne, PA 19087, USA.
VOLUNTEER PARTICIPATION
Healthcare professionals in all specialities are urged to
volunteer for participation in NCCLS projects. Please contact
the NCCLS Executive Offices for additional information on
committee participation.
October 1997 C3-A3

Preparation and Testing of Reagent Water in the Clinical

Laboratory; Approved Guideline — Third Edition

Abstract

Preparation and Testing of Reagent Water in the Clinical Laboratory; Approved Guideline — Third
Edition (NCCLS document C3-A3) provides basic information about different methods of water
purification so that laboratorians can determine which purification system(s) produces water that
best suits their specific needs. The guideline addresses issues of system design, process
specifications, storage and handling considerations, and appropriate testing methods.

[NCCLS. Preparation and Testing of Reagent Water in the Clinical Laboratory; Approved Guideline
— Third Edition. NCCLS document C3-A3 (ISBN 1-56238-336-1). NCCLS, 940 West Valley Road,
Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 1997.]

THE NCCLS consensus process, which is the mechanism for moving a document through two
or more levels of review by the healthcare community, is an ongoing process. Users should
expect revised editions of any given document. Because rapid changes in technology may
affect the procedures, methods, and protocols in a standard or guideline, users should replace
outdated editions with the current editions of NCCLS documents. Current editions are listed in
the NCCLS Catalog, which is distributed to member organizations, and to nonmembers on
request. If your organization is not a member and would like to become one, and to request a
copy of the NCCLS Catalog, contact the NCCLS Executive Offices. Telephone:
610.688.0100; Fax: 610.688.0700; E-Mail: exoffice@nccls.org.

NCCLS VOL. 17 NO. 18 i

October 1997 C3-A3

NCCLS VOL. 17 NO. 18 ii

 C3-A3
ISBN 1-56238-336-1
October 1997 ISSN 0273-3099

Preparation and Testing of Reagent Water in the Clinical

Laboratory; Approved Guideline — Third Edition

Volume 17 Number 18
David M. Jeffers
James H. Carter, Ph.D.
Gary A. Graham, Ph.D.
Anita K. Highsmith
Mona D. Jensen, Ph.D.
Richard R. Miller, Jr.
Edward A. Sasse, Ph.D.
Bette Seamonds, Ph.D.

ABC
October 1997 C3-A3

This publication is protected by copyright. No part of it may be reproduced, stored in a retrieval

system, or transmitted in any form or by any means (electronic, mechanical, photocopying,
recording, or otherwise) without written permission from NCCLS, except as stated below.

NCCLS hereby grants permission to reproduce limited portions of this publication for use in
laboratory procedure manuals at a single site, for interlibrary loan, or for use in educational
programs provided that multiple copies of such reproduction shall include the following notice, be
distributed without charge, and, in no event, contain more than 20% of the document's text.

Reproduced with permission, from NCCLS publication C3-A3 — Preparation and Testing of
Reagent Water in the Clinical Laboratory; Approved Guideline — Third Edition. Copies of
the current edition may be obtained from NCCLS, 940 West Valley Road, Suite 1400,
Wayne, Pennsylvania 19087-1898 USA.

Permission to reproduce or otherwise use the text of this document to an extent that exceeds the
exemptions granted here or under the Copyright Law must be obtained from NCCLS by written
request. To request such permission, address inquiries to the Executive Director, NCCLS, 940
West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA.

Copyright ©1997. The National Committee for Clinical Laboratory Standards.

Suggested Citation

[NCCLS. Preparation and Testing of Reagent Water in the Clinical Laboratory; Approved Guideline
— Third Edition. NCCLS document C3-A3 (ISBN 1-56238-336-1). NCCLS, 940 West Valley Road,
Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 1997.]

Proposed Standard—First Edition

January 1976

Tentative Standard—First Edition

January 1978

Approved Standard—First Edition

February 1980

Proposed Guideline—Second Edition

June 1985

Tentative Guideline—Second Edition

December 1988

Approved Guideline—Second Edition

August 1991

Approved Guideline—Third Edition

October 1997

ISBN: 1-56238-336-1
ISSN: 0273-3099

NCCLS VOL. 17 NO. 18 iv

October 1997 C3-A3

Committee Membership

Area Committee on Clinical Chemistry

Daniel A. Nealon, Ph.D. Johnson and Johnson Clinical Diagnostics

Chairholder Rochester, New York

Basil T. Doumas, Ph.D. Medical College of Wisconsin

Vice Chairholder Milwaukee, Wisconsin

Kevin D. Fallon, Ph.D. Instrumentation Laboratory

Lexington, Massachusetts

Jean C. Joseph, Ph.D. St. Mary Medical Center

Long Beach, California

Richard R. Miller, Jr. Dade International

Miami, Florida

Thomas P. Moyer, Ph.D. Mayo Clinic

Rochester, Minnesota

Gary L. Myers, Ph.D. Centers for Disease Control and Prevention

Atlanta, Georgia

Edward A. Sasse, Ph.D. Medical College of Wisconsin

Milwaukee, Wisconsin

Working Group on Reagent Water

David M. Jeffers York Hospital

Chairholder York, Pennsylvania

James H. Carter, Ph.D. Coulter Corporation

Miami, Florida

Gary A. Graham, Ph.D. Johnson and Johnson Clinical Diagnostics

Rochester, New York

Anita K. Highsmith Centers for Disease Control and Prevention

Atlanta, Georgia

Mona D. Jensen, Ph.D. Instrumentation Laboratory

Lexington, Massachusetts

Richard R. Miller, Jr. Dade International

Miami, Florida

Edward A. Sasse, Ph.D. Medical College of Wisconsin

Milwaukee, Wisconsin

Bette Seamonds, Ph.D. Swarthmore, Pennsylvania

NCCLS VOL. 17 NO. 18 v

October 1997 C3-A3

Sharon S. Ehrmeyer University of Wisconsin Hospitals

Board Liaison Madison, Wisconsin

Denise M. Lynch, M.T.(ASCP), M.S. NCCLS

Staff Liaison Wayne, Pennsylvania

Patrice E. Polgar NCCLS

Editor Wayne, Pennsylvania

Acknowledgments

The Working Group on Reagent Water acknowledges the participation of the following persons in
the review of Preparation and Testing of Reagent Water in the Clinical LaboratoryApproved
Guideline — Third Edition:

David J. Brigati, M.D. Harris Methodist Hospital

Fort Worth, Texas

Erich L. Gibbs, Ph.D. High-Q, Inc.

Wilmette, Illinois

William F. Koch, Ph.D. National Institute of Standards

and Technology
Gaithersburg, Maryland

Donald McGlory, Jr. New England Reagent Laboratory

East Providence, Rhode Island

Stephen Norton Erie Scientific

Portsmouth, New Hampshire

Gary A. O’Neill, Ph.D. Millipore Corp.

Bedford, Massachusetts

Brian Wulf George Fischer, Inc.

Tustin, California

NCCLS VOL. 17 NO. 18 vi

October 1997
ACTIVE MEMBERSHIP (as of 1 October 1997)
Sustaining Members
American Association for
Clinical Chemistry
Bayer Corporation
Beckman Instruments, Inc.
Becton Dickinson and Company
Boehringer Mannheim
Diagnostics, Inc.
College of American
Pathologists
Coulter Corporation
Dade International Inc.
Johnson & Johnson Clinical
Diagnostics
Professional Members
American Academy of Family
Physicians
American Association of
Bioanalysts
American Association of Blood
Banks
American Association for
Clinical Chemistry
American Association for
Respiratory Care
American Chemical Society
American Medical Technologists
American Public Health
Association
American Society for Clinical
Laboratory Science
American Society of
Hematology
American Society for
Microbiology
American Society of
Parasitologists, Inc.
American Type Culture
Collection, Inc.
Asociacion Espanola Primera de
Socorros
ASQC Food, Drug and Cosmetic
Division
Assoc. Micro. Clinici Italiani-
A.M.C.L.I.
Australasian Association of
Clinical Biochemists
Canadian Society of Laboratory
Technologists
Clinical Laboratory Management
Association
College of American
Pathologists
NCCLS VOL. 17 NO. 18
College of Medical Laboratory
Technologists of Ontario
College of Physicians and
Surgeons of Saskatchewan
Commission on Office
Laboratory Accreditation
Institut für Stand. und Dok. im
Med. Lab. (INSTAND)
International Council for
Standardization in
Haematology
International Federation of
Clinical Chemistry
International Society for
Analytical Cytology
Italian Society of Clinical
Biochemistry
Japan Assn. Of Medical
Technologists (Osaka)
Japanese Assn. Of Medical
Technologists (Tokyo)
Japanese Committee for Clinical
Laboratory Standards
Joint Commission on
Accreditation of Healthcare
Organizations
National Academy of Clinical
Biochemistry
National Society for
Histotechnology, Inc.
Ontario Medical Association
Laboratory Proficiency Testing
Program
Ordre professionnel des
technologistes médicaux du
Québec
Sociedade Brasileira de Analises
Clinicas
Sociedad Espanola de Quimica
Clinica
VKCN (The Netherlands)
Government Members
Armed Forces Institute of
Pathology
Association of State and
Territorial Public Health
Laboratory Directors
BC Centre for Disease Control
Centers for Disease Control and
Prevention
China National Centre for the
Clinical Laboratory
Commonwealth of Pennsylvania
Bureau of Laboratories
C3-A3
Department of Veterans Affairs
Deutsches Institut für Normung
(DIN)
FDA Center for Devices and
Radiological Health
FDA Division of Anti-Infective
Drug Products
Federacion Bioquimica de la
Provincia (Argentina)
Health Care Financing
Administration
INMETRO
Instituto de Salud Publica de
Chile
Instituto Scientifico HS.
Raffaele
(Italy)
Iowa State Hygienic Laboratory
Manitoba Health
Massachusetts Department of
Public Health Laboratories
Michigan Department of Public
Health
National Health Laboratory
(Luxembourg)
National Institute of Standards
and Technology
Ohio Department of Health
Oklahoma State Department of
Health
Ontario Ministry of Health
Saskatchewan Health-
Government of Saskatchewan
South African Institute for
Medical Research
Swedish Institute for Infectious
Disease Control
Industry Members
AB Biodisk
Abbott Laboratories
ABC Consulting Group, Ltd.
AccuMed International, Inc.
aejes
Ammirati Regulatory Consulting
Asséssor
Atlantis Laboratory Systems
Avecor Cardiovascular, Inc.
Bayer Corporation - Elkhart, IN
Bayer Corporation - Middletown,
VA
Bayer Corporation - Tarrytown,
NY
Bayer Corporation - West
Haven, CT
vii
October 1997
Bayer-Sankyo Co., Ltd.
Beckman Instruments, Inc.
Becton Dickinson and Company
Becton Dickinson Consumer
Products
Becton Dickinson
Immunocytometry Systems
Becton Dickinson Italia S.P.A.
Becton Dickinson Microbiology
Systems
Becton Dickinson VACUTAINER
Systems
Behring Diagnostics Inc.
Behring Diagnostics Inc. - San
Jose, CA
bioMérieux Vitek, Inc.
Biometrology Consultants
Bio-Rad Laboratories, Inc.
Bio-Reg Associates, Inc.
Biosite Diagnostics
Biotest AG
Boehringer Mannheim
Diagnostics, Inc.
Boehringer Mannheim GmbH
Bristol-Myers Squibb Company
Canadian Reference Laboratory
Ltd.
CASCO Standards
Checkpoint Development Inc.
ChemTrak
Chiron Diagnostics Corporation
Chiron Diagnostics Corporation -
International Operations
Chiron Diagnostics Corporation -
Reagent Systems
Clinical Lab Engineering
COBE Laboratories, Inc.
Control Lab (Brazil)
Cosmetic Ingredient Review
Coulter Corporation
Cytometrics, Inc.
CYTYC Corporation
Dade International - Deerfield, IL
Dade International - Glasgow,
DE
Dade International - Miami, FL
Dade International -
Sacramento, CA
DAKO A/S
Diagnostic Products Corporation
Diametrics Medical, Inc.
Difco Laboratories, Inc.
Eiken Chemical Company, Ltd.
Enterprise Analysis Corporation
Donna M. Falcone Consultants
Fujisawa Pharmaceutical Co.
Ltd.
Gen-Probe
NCCLS VOL. 17 NO. 18
Glaxo Inc.
Greiner Meditech, Inc.
Health Systems Concepts, Inc.
Helena Laboratories
Higman Healthcare
Hoechst Marion Roussel, Inc.
Hybritech, Incorporated
Hycor Biomedical Inc.
i-STAT Corporation
Integ, Inc.
International Biomedical
Consultants
International Remote Imaging
Systems (IRIS)
International Technidyne
Corporation
Johnson & Johnson Clinical
Diagnostics
Johnson & Johnson Health Care
Systems, Inc.
Kimble/Kontes
Labtest Sistemas Diagnosticos
Ltda.
LifeScan, Inc. (a Johnson &
Johnson Company)
Lilly Research Laboratories
Luminex Corporation
Mallinckrodt Sensor Systems
MBG Industries, Inc.
Medical Device Consultants,
Inc.
Medical Laboratory Automation
Inc.
MediSense, Inc.
Merck & Company, Inc.
Metra Biosystems
Neometrics Inc.
Nissui Pharmaceutical Co., Ltd.
Norfolk Associates, Inc.
North American Biologicals, Inc.
Olympus Corporation
Optical Sensors, Inc.
Organon Teknika Corporation
Orion Diagnostica, Inc.
Ortho Clinical Diagnostics, Inc.
Otsuka America Pharmaceutical,
Inc.
Pfizer Canada, Inc.
Pfizer Inc
Pharmacia & Upjohn (MI)
Pharmacia & Upjohn (Sweden)
Procter & Gamble
Pharmaceuticals, Inc.
The Product Development Group
Radiometer America, Inc.
Radiometer Medical A/S
Research Inc.
C3-A3
David G. Rhoads Associates,
Inc.
Rhône-Poulenc Rorer
Roche Diagnostic Systems
(Div. Hoffmann-La Roche
Inc.)
Roche Laboratories (Div.
Hoffmann-La Roche Inc.)
ROSCO Diagnostica
The R.W. Johnson
Pharmaceutical Research
Institute
Sarstedt, Inc.
Schering Corporation
Schleicher & Schuell, Inc.
Second Opinion
SenDx Medical, Inc.
Sherwood-Davis & Geck
Shionogi & Company, Ltd.
Showa Yakuhin Kako Company,
Ltd.
Sienna Biotech
SmithKline Beecham
Corporation
SmithKline Beecham (NZ) Ltd.
SmithKline Beecham, S.A.
SmithKline Diagnostics, Inc.
(Sub. Beckman Instruments,
Inc.)
SRL, Inc.
Streck Laboratories, Inc.
Sumitomo Metal Bioscience Inc.
Sysmex Corporation
TOA Medical Electronics
Unipath Co (Oxoid Division)
Vetoquinol S.A.
Vysis, Inc.
Wallac Oy
Warner-Lambert Company
Wyeth-Ayerst
Xyletech Systems, Inc.
Yeongdong Pharmaceutical
Corp.
Zeneca
Trade Associations
Association of Medical
Diagnostic Manufacturers
Health Industry Manufacturers
Association
Japan Association Clinical
Reagents Ind. (Tokyo, Japan)
Medical Industry Association
of Australia
National Association of Testing
Authorities - Australia
viii
October 1997
Associate Active Members
Affinity Health System (WI)
Allegheny University of the
Health Sciences (PA)
Alton Ochsner Medical
Foundation (LA)
American Oncologic Hospital
(PA)
Anzac House (Australia)
Associated Regional &
University Pathologists (UT)
Baptist Memorial Healthcare
System (TX)
Beth Israel Medical Center (NY)
Brazosport Memorial Hospital
(TX)
Bristol Regional Medical Center
(TN)
Brooke Army Medical Center
(TX)
Brooks Air Force Base (TX)
Broward General Medical Center
(FL)
Canterbury Health Laboratories
(New Zealand)
Central Peninsula General
Hospital (AK)
Childrens Hospital Los Angeles
(CA)
Children's Hospital Medical
Center (Akron, OH)
Clendo Lab (Puerto Rico)
Clinical Diagnostic Services (NJ)
Commonwealth of Kentucky
CompuNet Clinical Laboratories
(OH)
Consolidated Laboratory
Services (CA)
Consultants Laboratory (WI)
Detroit Health Department (MI)
Dhahran Health Center (Saudi
Arabia)
Duke University Medical Center
(NC)
Dwight David Eisenhower Army
Medical Center (Ft. Gordon,
GA)
East Side Clinical Laboratory (RI)
Easton Hospital (PA)
Federal Medical Center (MN)
Frye Regional Medical Center
(NC)
Gila Regional Medical Center
(NM)
Grady Memorial Hospital (GA)
Great Smokies Diagnostic
Laboratory (NC)
Gulhane Military Medical
Academy (Turkey)
NCCLS VOL. 17 NO. 18
Harris Methodist Fort Worth
(TX)
Hartford Hospital (CT)
Health Alliance Laboratory
Services (OH)
Heritage Hospital (MI)
Hopital Saint Pierre (Belgium)
Incstar Corporation (MN)
Integris Baptist Medical Center
of Oklahoma
International Health Managment
Associates, Inc. (IL)
Kaiser Permanente (CA)
Kangnam St. Mary’s Hospital
(Korea)
Kenora-Rainy River Regional
Laboratory Program (Dryden,
Ontario, Canada)
Klinisches Institute für
Medizinische (Austria)
Laboratoire de Santé Publique
du Quebec (Canada)
Laboratory Corporation of
America (NC)
Laboratory Corporation of
America (NJ)
Lancaster General Hospital (PA)
Libero Instituto Univ. Campus
Biomedico (Italy)
Louisiana State University
Medical Center
Maimonides Medical Center
(NY)
Maine Medical Center
Massachusetts General Hospital
MDS-Sciex (Concord, ON,
Canada)
Melbourne Pathology (Australia)
Memorial Medical Center (LA)
Methodist Hospital (TX)
Methodist Hospital Indiana
Methodist Hospitals of Memphis
(TN)
Montreal Children’s Hospital
(Canada)
Mount Sinai Hospital (NY)
Mount Sinai Hospital (Toronto,
Ontario, Canada)
National Genetics Institute (CA)
Naval Hospital Cherry Point (NC)
New Britain General Hospital
(CT)
New Hampshire Medical
Laboratories
The New York Blood Center
New York State Department of
Health
New York State Library
New York University Medical
Center
C3-A3
North Carolina Laboratory of
Public Health
North Shore University Hospital
(NY)
Olin E. Teague Medical Center
(TX)
Omni Laboratory (MI)
Our Lady of Lourdes Hospital
(NJ)
Our Lady of the Resurrection
Medical Center (IL)
PAPP Clinic P.A. (GA)
Pathology Associates
Laboratories (CA)
Permanente Medical Group (CA)
PLIVA d.d. Research Institute
(Croatia)
Polly Ryon Memorial Hospital
(TX)
Providence Medical Center (WA)
Puckett Laboratories (MS)
Quest Diagnostics (MI)
Quest Diagnostics (PA)
Reid Hosptial & Health Care
Services (IN)
Sacred Heart Hospital (MD)
St. Boniface General Hospital
(Winnipeg, Canada)
St. Francis Medical Center (CA)
St. John Regional Hospital (St.
John, NB, Canada)
St. Joseph’s Hospital -
Marshfield Clinic (WI)
St. Luke’s Hospital (PA)
St. Luke’s Regional Medical
Center (IA)
St. Luke’s-Roosevelt Hospital
Center (NY)
St. Mary of the Plains Hospital
(TX)
St. Paul Ramsey Medical Center
(MN)
St. Vincent Medical Center (CA)
San Francisco General Hospital
(CA)
Seoul Nat’l University Hospital
(Korea)
Shanghai Center for the Clinical
Laboratory (China)
Shore Memorial Hospital (NJ)
SmithKline Beecham Clinical
Laboratories (GA)
SmithKline Beecham Clinical
Laboratories (TX)
South Bend Medical Foundation
(IN)
So. California Permanente
Medical Group
Southeastern Regional Medical
Center (NC)
ix
October 1997
SUNY @ Stony Brook (NY)
Tampa General Hospital (FL)
UNC Hospitals (NC)
University of Alberta Hospitals
(Canada)
University of Florida
University Hospital (Gent)
(Belgium)
University Hospital (London,
Ontario, Canada)
University Hosptial (Linkoping,
Sweden)
University Hospital (IN)
University Hospital of
Cleveland (OH)
The University Hospitals (OK)
University of Medicine &
Dentistry, NJ University
Hospital
OFFICERS
A. Samuel Koenig, III, M.D.,
President
Family Medical Care
William F. Koch, Ph.D.,
President Elect
National Institute of Standards
and Technology
F. Alan Andersen, Ph.D.,
Secretary
Cosmetic Ingredient Review
Donna M. Meyer, Ph.D.,
Treasurer
Sisters of Charity Health Care
System
Charles F. Galanaugh, Past
President
Becton Dickinson and
Company (Retired)
John V. Bergen, Ph.D.,
Executive Director
NCCLS VOL. 17 NO. 18
University of Michigan
University of Nebraska Medical
Center
University of the Ryukyus
(Japan)
The University of Texas Medical
Branch
University of Virginia Medical
Center
U.S. Army Hospital, Heidelberg
UZ-KUL Medical Center
(Belgium)
VA (Albuquerque) Medical
Center (NM)
VA (Ann Arbor) Medical Center
(MI)
VA (Denver) Medical Center
(CO)
BOARD OF DIRECTORS
Carl A. Burtis, Ph.D.
Oak Ridge National Laboratory
Sharon S. Ehrmeyer, Ph.D.
University of Wisconsin
Elizabeth D. Jacobson, Ph.D.
FDA Center for Devices and
Radiological Health
Hartmut Jung, Ph.D.
Boehringer Mannaheim GmbH
Tadashi Kawai, M.D., Ph.D.
International Clinical Pathology
Center
Kenneth D. McClatchey, M.D.,
D.D.S.
Loyola University Medical
Center
C3-A3
VA (Jackson) Medical Center
(MS)
VA (Long Beach) Medical Center
(CA)
VA (Miami) Medical Center (FL)
Venice Hospital (FL)
Veterans General Hospital
(Republic of China)
Virginia Baptist Hospital
Warde Medical Laboratory (MI)
William Beaumont Hospital (MI)
Winn Army Community Hospital
(GA)
Wisconsin State Laboratory of
Hygiene
Yonsei University College of
Medicine (Korea)
York Hospital (PA)
Zale Lipshy University Hospital
(TX)
Robert F. Moran, Ph.D.,
FCCM, FAIC
mvi Sciences
David E. Nevalainen, Ph.D.
Abbott Laboratories
Donald M. Powers, Ph.D.
Johnson & Johnson Clinical
Diagnostics
Eric J. Sampson, Ph.D.
Centers for Disease Control
and Prevention
Marianne C. Watters,
M.T.(ASCP)
Parkland Memorial Hospital
Ann M. Willey, Ph.D.
New York State Department of
Health
x
October 1997 C3-A3

Contents

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i

Committee Membership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

Active Membership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

3 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

4 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

5 Design Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

5.1 Initial Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

5.2 General Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.3 Materials for Distribution and Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

6 Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

6.1 Requirements for Reagent Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

7 Storage and Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

7.1 Type I Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

7.2 Type II and Type III Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
7.3 Special Reagent Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
7.4 Handling Precautions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

8 Commercially Available Reagent Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

8.1 Diluent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
8.2 Sterile Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
8.3 Purchased Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

9 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.1 Microbial Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.2 pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
9.3 Resistivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
9.4 Soluble Silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
9.5 Organic Contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
9.6 Special Water Considerations: Endotoxins and Specifications . . . . . . . . . . . . . . . . 22

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Appendix A: Description of Purification Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

NCCLS VOL. 17 NO. 18 xi

October 1997 C3-A3

Appendix B: Quality Assurance Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

Appendix C: Methods for Correction or Compensation of Resistivity Measurements . . . . . . 31

Summary of Comments and Working Group Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Related NCCLS Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

NCCLS VOL. 17 NO. 18 xii

October 1997 C3-A3

Foreword

This guideline describes water of three specific levels of quality (Types I, II, and III) and the
methods for producing and testing such water. The classifications and specifications are designed
to enable laboratory scientists and supporting industries to specify the quality of water to be used
in such procedures as, for example, reagent preparation, reconstitution of lyophilized materials, and
sample dilution.

The committee believes that the criteria and measurements specified for monitoring water quality
are the minimum necessary. The parameters are as follows:

! Resistivity
! Microbial content
! pH
! Silicate content
! Particulate matter
! Organic content.

Measurements of resistivity are practical and readily available, and they provide significant
information about the water sampled. As the sensitivity of laboratory analytical processes
increases and sample size decreases, microbial content of the reagent water becomes increasingly
important. Microorganisms can inactivate reagents, contribute to total organic contamination, or
alter optical properties of the test solutions.

The monitoring of other parameters—namely pH, silicate content, particulate matter, and organic
content—depends on many variables. Each laboratory should assess the need for, and frequency
of, monitoring any of these on a routine basis. If the source water and the purification system
produce water that is typically negative for some contaminant, the frequency of testing for that
contaminant can be decreased. However, it is necessary to ensure occasionally that the end
product is free from all significant contaminants.

This guideline recommends that the laboratory examine the acceptability of the type of reagent
water to be used and record the rationale for this decision. A laboratory should also check to see if
there are requirements applicable to its specific uses.

Key Words

Laboratory water, microbial contamination, reagent water, resistivity, special-purpose water,

specifications, testing, water contamination, water purity, water-soluble silicates.

NCCLS VOL. 17 NO. 18 xiii

October 1997 C3-A3

NCCLS VOL. 17 NO. 18 xiv

October 1997
Preparation and Testing of Reagent Water
in the Clinical Laboratory; Approved Guideline — Third Edition
1 Introduction
This guideline describes water of three
specific levels of quality (Types I, II, and III)
and the methods for producing and testing
such water in the clinical laboratory (e.g.,
chemistry, hematology, and microbiology).
The classifications and specifications are
designed to enable laboratory scientists and
supporting industries to specify the quality of
water to be used in such procedures, e.g.,
reagent preparation, reconstitution of
lyophilized materials, and sample dilution.
No one specific method is recommended for
producing purified water. A single method or
combination of methods may be used
satisfactorily, provided that the end product
meets the required specifications stated in
this guideline. Understand that any changes
to a previously qualified water-purification
process or source water require a revalidation
of the reagent water system.
2 Scope
This document addresses requirements for
water purified for laboratory use as described
below (see Table 1 on page 4), irrespective of
the site of water production. Three grades of
water are specified, and special reagent water
is also addressed (see Table 2 on page 8):
! Clinical laboratory reagent water, Type I
! Clinical laboratory reagent water, Type II
! Clinical laboratory reagent water, Type III
! Special reagent water.
Water that conforms to specifications
published by the American Chemical Society
(ACS),a the American Society for Testing and
American Chemical Society. Reagent Chemicals, Eighth
a
Ed., American Chemical Society Specifications (April 1993),
pp. 69, 70, and 777–778. American Chemical Society,
1155 Sixteenth Street, N.W., Washington, DC 20036.
NCCLS VOL. 17 NO. 18
C3-A3
Materials (ASTM),b the College of American
Pathologists (CAP),c and the United States
Pharmacopeia (USP)d may or may not be
equivalent to the reagent water described in
this document. USP specifications apply to a
variety of in vitro and in vivo uses.
Classification is based on the ability of the
purified water to pass a series of designated
tests rather than on its ability to meet
definitive concentrations of contaminants.
3 Definitions
Absorption, n - Process by which a substance
is taken up in bulk by a material (absorbent)
and held in pores or interstices in the interior.
Note: Contrast with adsorption, a process by
which a substance is bound at the surface of
a material (adsorbent).
Activated carbon, n - Porous carbon material
used for adsorption of organic contaminants
and chlorine.
Adsorption, n - Process in which molecules,
atoms, and ions become attached to the
surfaces of solids and liquids. Activated
carbon will remove some organic compounds
by adsorption. Note: Contrast with absorb.
Biofilm, n - A thin layer of organisms
embedded in an organic matrix, composed
mostly of glycoproteins and
heteropolysaccharides. The organisms in this
layer can multiply even in reagent water and
the layer protects them from periodic
b
ASTM. Standard Specification for Reagent Water.
ASTM document D 1193–91 (1991). ASTM, 100 Barr
Harbor Drive, West Conshohocken, Pennsylvania, 19428-
2959.
c
College of American Pathologists Commission on
Laboratory Inspection and Accreditation. Reagent Water
Specifications (1985). College of American Pathologists,
325 Waukegan Road, Northfield, Illinois 60093-2750.
d
USP 23, Official Monographs: Water, pp.
1635–1637; High Purity Water, pp. 1782; Water for
Pharmaceutical Purposes, pp. 1984; Reagents, Indicators,
and Solutions, pp.1987. United States Pharmacopeia,
12601 Twinbrook Parkway, Rockville, Maryland 20852.
1
October 1997
treatment with many biocides that are
effective in killing free-floating organisms.
Carbon adsorption, n - A process during
which
the surface of carbon takes on, or adsorbs in
an extremely thin layer, molecules of gases,
of dissolved substances, or of liquids with
which it is in contact.
Chemical oxidation, n - The process of using
chemicals (oxidizing agents, such as ozone,
peroxide, chlorine) to cause a chemical
reaction wherein electrons are transferred
from a reactive chemical in the water to the
oxidizing agent.
Conductivity//electrolytic conductivity//
specific conductance, n - Electrolytic
conductivity is a quantitative measure of the
ability of a solution to carry an electric
current. It is the electrical conductance of an
aqueous solution measured between opposite
parallel faces of a 1-cm cube at a specified
temperature. The unit of conductance is the
siemens (S), formerly the mho (reciprocal
ohm). For these specifications, electrolytic
conductivity should be reported at 25 BC in
microsiemens per centimeter (µS/cm). The
reciprocal of electrolytic conductivity is
resistivity. To illustrate, a solution with an
electrolytic conductivity of 0.1 µS/cm will
have a resistivity of 10 MSCcm. See
Resistivity.
Deadleg, n - A specific volume or region of
stagnation in an apparatus. Note: Commonly
used in piping and tubing terminology to
describe the volume included in a length that
is six times the diameter of the unused
portion of the pipe or tube.
Deionization, n - A purification process that
uses synthetic resins to accomplish a
selective exchange. The removal of ions from
a solution by ion exchange.
Dissolved ionized gases, n - Charged
molecules that possess perfect molecular
mobility, the ability to expand indefinitely,
and are dispersed in water.
Dissolved ionized solids, n - The mass of
charged constituents in a filtered water
NCCLS VOL. 17 NO. 18
C3-A3
sample. For operational purposes, the filter
pore is usually 0.45 µm.
Dissolved organics, n - Matter, composed of
hydrocarbons or their derivatives, that is
dispersed in water to give a single
homogenous liquid phase.
Distillation, n - The volatilization, or
evaporation and subsequent condensation of
a liquid, to purify or concentrate a substance
or separate one substance from another.
Endotoxin/pyrogen, n - A thermostable
component of viable or nonviable gram-
negative microorganisms that can cause a
fever when injected or infused.
Filtration, n - A purification process in which
the passage of liquid through a porous
substance results in the removal of impurities
based on the interaction of the impurities
with that porous substance. Note: This
interaction is usually physical in nature and is
often based on particle size.
High performance liquid chromatography
(HPLC), n - An analytical technique for
performing chromatographic separations of
organic compounds in which the mobile
phase, eluent, or carrier, is a liquid under
pressure.
Microbial content, n - In reagent water
testing, the quantity of viable organisms, as
determined by total colony count after
incubation at 36 ± 1 BC for 24 hours,
followed by 24 hours at ambient temperature
(23 ± 3 BC) and reported as colony-forming
units per milliliter (CFU/mL).
Microorganism, n - Any organism that is too
small to be viewed by the unaided eye, such
as bacteria, viruses, molds, yeasts, protozoa,
and some fungi and algae.
Nanofiltration, n - The process by which
particles are separated from water by passing
through a permeable material with very small
pore sizes (10-9 m).
Particulate matter, n - Discrete quantities of
solid matter dispersed in water.
2
October 1997
Passivation, n - A process used to remove
surface iron from stainless steel piping by
acid etching and to oxidize the remaining
chromium and nickel surfaces to impervious
oxides.
pH, n - [The symbol for the power of
hydrogen defined as] the negative decadic
logarithm of the [relative molal] hydrogen ion
activity.
"Polish,” n - A term that describes the post-
treatment processing of source water to
remove all or some of the remaining
contaminants, depending on the intended
use.
Pyrogen, n - See Endotoxin.
Qualification, n— The act of establishing that
the process, equipment, and/or materials are
useable and will result in acceptable results.
Reagent water, n - Water purified and
classified for specific analytical uses.
Resistivity//specific resistance, n - The
electrical resistance in ohms measured
between opposite parallel faces of a 1-cm
cube of an aqueous solution at a specified
temperature. For these specifications, the
resistivity is corrected to 25 BC and reported
as megohmCcm (MS-cm). The reciprocal of
resistivity is electrolytic conductivity (formerly
referred to as specific conductance). See
Conductivity.
Reverse osmosis, n - A process in which
water is forced under pressure through a
semipermeable membrane leaving behind a
percentage of dissolved organic, dissolved
ionic, and suspended impurities.
Sanitization, n - Chemical and/or physical
processes used to kill microorganisms or
reduce contamination.
Source water, n - The water that is
introduced into the purification process.
Total microorganisms, n - All aerobic and
facultative anaerobic heterotrophic microor-
ganisms in water.
NCCLS VOL. 17 NO. 18
C3-A3
Total organic carbon (TOC), n - Carbon in the
form of organic compounds.
Ultrafiltration, n - A process during which
water is forced under pressure through a
semipermeable membrane leaving behind a
percentage of dissolved organic and
suspended impurities. The dissolved organic
and suspended impurities are filtered based
on molecular weight and size.
Ultraviolet (photochemical) oxidation, n - A
process by which an ultraviolet light source
(185 nm) is used to convert carbon to carbon
dioxide.
Ultraviolet sterilization, n - A process by
which an ultraviolet light source (254 nm) is
used to destroy microorganisms.
Validation, n— The process of determining
that the process, equipment, and/or materials
will result in acceptable results. Note: In the
United States, the Food and Drug
Administration requires documented evidence
that provides a high degree of assurance that
a specific process, equipment, and/or
materials will consistently produce a product
meeting its predetermined specifications and
quality characteristics.
4 Preparation
An acceptable method for water purification
produces water that meets the specifications
stated in Section 6. Each preparation process
has its own source water requirements and
can also have residual contaminants that
should be considered. It is important to
recognize that systems that are improperly
chosen, designed, or maintained can actually
add contaminants to the water.
Because there are many options in water-
purification technology, the working group
cannot recommend a particular purification
process to produce a particular grade of
water. The decision as to which system or
systems to install depends on past
experience, and on present and future needs.
A major consideration is the quality of the
source water. If there is a high concentration
of total dissolved solids, it is likely that the
water will have to be pretreated before
3
October 1997 C3-A3

further purification. Therefore, a combination that are commonly employed in various

of the commonly available processes for
water purification, summarized in Table 1 on
the next page, can be necessary to produce
water of the desired quality, particularly Type
I and special reagent water.
Table 1 is intended only as a guide to the
strengths of discrete commercial technologies
combinations to reduce the concentrations of
impurities in reagent water. The user is urged
to consult with suppliers and colleagues,
especially those with experience with the
same or similar source water and
applications. The user should bear in mind
that each discrete

Table 1. Water Purification Process Comparison (see Appendix A)

Major Classes of Contaminants

Dissolved Dissolved
Purification Process Ionized Ionized Dissolved Particulate Micro- Pyrogens/
Solids Gases Organics Matter organisms Endotoxins

Distillation E G/P G E E E

Deionization E E P P P P

Reverse osmosis G P G E E E

Carbon P P E/G P P P
adsorption/absorption

Filtration (0.22 Fm) P P P E E P

Ultrafiltration P P P E E E

Nanofiltration G/P P G E E E

Chemical oxidation P P P P E/G E/G

Ultraviolet oxidation* P P G P G/P P

Ultraviolet sterilization* P P P P G P

* = Ultraviolet light kills microorganisms but does not remove them. Another process is required to remove them.
E = Excellent (capable of complete or near total removal).
G = Good (capable of removing large percentages).
P = Poor (little or no removal).

technology has weaknesses. The only water by the manufacturer should be

generalization that can be made with some
degree of certainty is that some means for
monitoring the performance of a water-
purification system should be selected to
reasonably ensure that the failure of any
stage of the system will be detected.
5 Design Considerations
5.1 Initial Considerations
If a laboratory needs to install either a new
water-purification system or modify an
existing one, there are several issues to
consider. Initially, an analysis of the source
requested. This analysis should include a
minimal list of contaminants (e.g., silica,
organics, magnesium, calcium,
microorganisms, and total and free chlorine),
as well as pH and resistivity. (Both chlorine
and pH should be measured on site.) The
larger and more complex the system using
tap water, or the more diverse the intended
application(s), the more testing is required.
Also, multiple periodic testing can be
important because of seasonal variations in
contaminant concentrations. In some
instances, it can be helpful to consult local
water-testing authorities for advice on the
impurity content of the local source water.

NCCLS VOL. 17 NO. 18 4

October 1997
5.2 General Considerations
The overall design should be such that there
are no deadlegs in the system. Deadlegs
provide areas for stagnation and the potential
for microbial growth. The working group
strongly recommends a system that
recirculates with a minimal velocity of 5 feet
per second. Recirculation also helps minimize
microbial growth.
Because there are several approaches to
water purification, it is necessary to review
the capital outlay and operating costs of
various systems. It is recommended that the
cost analysis include expenditures for
maintenance of the system.
Once the selection and installation of a
system is complete, it is necessary to sanitize
the system before use and then at least semi-
annually, or more often as recommended by
the manufacturers, or as determined by
quality control criteria. Procedures for
sanitization can be performed by the
manufacturer of the system or according to a
procedure provided by the manufacturer.
With extended lack of use, there is a danger
of stagnation. If the system is shut down for
more than 72 hours, sanitization is
recommended. Reverse osmosis systems can
require the use of a disinfectant to sanitize
the membrane. Consult the manufacturer for
recommendations for appropriate
disinfectants. To ensure complete removal of
the disinfectant after the sanitization process,
the reagent water must be tested before use.
In some instances, commercial test kits are
available. Otherwise, consult the
manufacturer of the reverse osmosis system
for guidance in evaluating reagent water for
traces of residual disinfectant. Appropriate
safety and disposal precautions should be
used when handling the water after
maintenance of the membrane.
In general, the system will operate optimally
when used daily.
5.3 Materials for Distribution and
Storage
Much of the data relating to the suitability of
materials for the distribution and storage of
NCCLS VOL. 17 NO. 18
C3-A3
reagent water are proprietary. Therefore, little
objective data relating to the suitability of
these materials is available. As a general
rule, the selection of materials for the
distribution and storage of reagent water
depends on the desired water quality and
cost. A material is acceptable for a given
application as long as it functions well from a
mechanical standpoint and can be treated
adequately to prevent the growth of
microorganisms, which aid in establishing
biofilms. Additionally, the material should not
leach significant concentrations of
contaminants (a function of surface area,
flow volume, and flow rate) that are not
effectively removed during downstream
stages of purification. If a system is sanitized
on a periodic basis, the periods might have to
be as frequent as weekly, depending on local
conditions. Uncontrolled, biofilm production
produces more total organic carbon (TOC)
than TOC-leaching piping materials.
The control and elimination of microorganisms
from water-purification systems is important.
Microorganisms and chemical analytes
contribute to biofilm formation, which can
occur in or on distribution systems, surfaces,
sides of storage tanks, housing materials, as
well as membranes and ionic beds. Periodic
sanitization is a necessary step in the control
of microorganisms that make up these
biofilms (see Section 5.2). Once a biofilm
has been established, removal is difficult.
Some agents that have been shown to have a
limited effect on biofilm build-up include
ozone, sodium hydroxide, and sodium
hypochlorite.
Increasingly, plastics tend to be the materials
of choice, although stainless steel remains in
wide use. Glass, aluminum, tin, tin-lined, and
titanium piping possess various combinations
of reactivity, mechanical shortcomings, and
cost. The requirements for production
facilities are likely to be different from those
of clinical and research laboratories. In
general, it is preferable not to distribute high-
purity water but, rather, complete purification
to desired levels at the point of use.
5
October 1997
5.3.1 Piping
5.3.1.1 Stainless Steel
There are many grades of stainless steel and
the grades vary considerably with regard to
ease of machining, ease of welding,
smoothness of finish, corrosion resistance,
and leaching of contaminants. Caution is
necessary if the user performs trace-metal
analyses. After fabrication, and at intervals
thereafter, stainless systems must be
passivated with nitric acid, a procedure that
dissolves exposed iron from the stainless
steel surface and produces a resistant,
chromium oxide surface layer. Inadequate
passivation can result in visible rust.
Biofilms, which will develop in reagent water
that does not contain a biocide, are credited
with pitting stainless systems; however,
stainless steel systems can tolerate steam,
90 BC water, and biocidal concentrations of
chlorine and ozone. Leaching of TOC is low,
provided the system has been fully cleaned.
5.3.1.2 High-Density Polyethylene (HDPE)
HDPE is essentially pure organic material that
is less prone to oxidation than polypropylene.
Leaching of inorganic contaminants is low
and leaching of TOC increases with
temperature. Some of the major drawbacks
to the use of HDPE are that it is soft and has
a maximum operating temperature of 70 BC.
Its porosity, low operating temperature, and
susceptibility to oxidation make biofilm
control in HDPE systems difficult. The cost of
HDPE piping is approximately equal to PVC
piping.
5.3.1.3 Polyvinyl Chloride (PVC)
PVC piping is a complex mixture of plastics
that includes about 85% PVC by weight and
15% other stabilizer compounds, which are
usually organometallic in nature. Water
systems should utilize high-grade PVC pipe,
which has been specified to contain low
levels of problem additives. PVC can be
expected to leach a number of substances
into the water, including monomer,
plasticizers, stabilizers, coloring,
tetrahydrofuran, and methyl ethyl ketone,
among others. However, in practice, the
NCCLS VOL. 17 NO. 18
C3-A3
leaching of inorganic and organic substances
is often lower than one might expect.
Leaching increases with temperature and the
maximum operating temperature is 60 BC.
Thus the porous nature of PVC and its low
operating temperature make biofilm control
difficult. PVC is often used in stages of
water-distribution systems where the water is
not of extremely high purity and especially
where continuous biocidal levels of chlorine
can be maintained.
5.3.1.4 Polypropylene (PP)
Polypropylene belongs to the “polyolefins,” the
class of carbon chains composed only of
carbon and hydrogen. Physical properties
include strength, low weight, abrasion
resistance, impact strength, and a wide
operating temperature range (80 BC continuous,
100 BC for short periods). All polypropylene
includes additives to improve either impact
strength or resistance to temperature extremes,
UV light, or weathering; some contain
pigments, which may contribute to these
characteristics. Chemical resistance is
excellent for high-purity water. Cost is
relatively low. Material is joined by fusion
welding (socket or butt). Competing
products should be compared for design and
performance.
5.3.1.5 Polyvinylidene Fluoride (PVDF)
PVDF combines excellent chemical resistance
with good molding properties. Like most
fluoropolymers, it is essentially a pure
material with no stabilizers, pigments,
lubricants, or antioxidants. The maximum
service temperature is 140 BC. Small
amounts of fluoride can be expected to leach
into the water from a new system, but,
overall, the leaching of inorganic material is
low. TOC in PVDF systems has been
observed to decline to levels below the
detection limits of current instrumentation,
usually within hours. Exposure of PVDF to
ultraviolet light (UV), especially 185 nm, can
cause it to fail unless the system includes
light traps. Normal sunlight has not been
reported to cause a problem. PVDF piping
can be guaranteed to withstand 3.5 bar (50
psi; 3.5 x 105 pascals) steam for 25 years
and the modest thermal expansion of PVDF
6
October 1997
can be an advantage in shedding biofilm.
PVDF can also withstand continuous biocidal
concentrations of chlorine or ozone.
Installation costs are greater than for PVC but
might decline as the technology improves.
5.3.1.6 Perfluoro-alkoxy vinyl ether (PFA)
PFA is more fluorinated than PVDF; it offers
increased chemical resistance and heat
tolerance (to 260 BC). PFA is substantially
more expensive than PVDF and its field use
has been limited.
5.3.1.7 Poly-ether-ether-ketone (PEEK)
There are several forms of fluoropolymers,
some of which do not mold well into piping.
PEEK is an essentially pure material with
excellent chemical stability and higher
operating temperature than PVDF. PEEK can
be used at substantially higher pressures than
other plastics and its excellent chemical
resistance and temperature rating (150 BC)
permit a variety of options for biofilm control.
There is less inorganic and organic leaching
from PEEK than from any other plastic. Field-
use of PEEK has been limited however,
because there is no inexpensive commercial
way to create welds and few fittings are
available.
5.3.1.8 Polytetrafluoroethylene (PTFE)
The most chemically resistant of the
fluoropolymers is the fully fluorinated PTFE;
however, this material cannot be welded or
easily formed into pipes.
5.3.2 Storage
Storage in a tank not subjected to
recirculation is not recommended in large-
scale systems. Insofar as some storage is
necessary for the practical, small-scale use of
reagent water in the laboratory, the
containers should be rinsed with reagent
water at every filling, and the containers
should be of a size that ensures frequent
filling.
Large-scale tanks (hundreds or thousands of
gallons) tend to be made from stainless steel
or steam-cured, PVDF-lined, fiberglass-
NCCLS VOL. 17 NO. 18
C3-A3
reinforced resin. Small scale tanks may be
made from any of the materials used for
piping. Glass may be used, depending on the
application. Clean, chemically resistant
glasses have the advantage that they
minimally leach TOC into the water and they
are not permeable to volatile substances in
the ambient air. They do, however, leach
some ions because all glasses are slightly
soluble.
5.3.3 Spigots
Spigots should be considered an integral part
of any distribution system. They should be
selected with the same care as the piping and
storage tanks. It is particularly important that
spigot design minimize dead spaces and use
seals that do not leach contaminants into the
water.
NOTE: Tubing should not be left attached to
spigot outlets. If tubing is used to transfer
water, it should be carefully selected for
chemical resistance and carefully cleaned.
The risk of chemical and microbial
contamination when transferring water via
tubing is high.
6 Specifications
All specifications are stated for water as
measured at the time of production. The
resistivity of Type I water should be
measured inline daily; all other specifications
relate to the samples measured offline.
Additional purification can be required for
selected clinical laboratory procedures, such
as:
! Preparation of water with no endotoxin/
pyrogen levels for cell culture
! Preparation of microorganism-free water
for direct fluorescent detection of
microorganisms, such as Legionella (sp.),
for direct fluorescent antibody testing, or
for direct fluorescent stains of
mycobacteria
! Preparation of water with minimal organic
content for HPLC.
7
October 1997 C3-A3

Special requirements should be discussed this stringent requirement be necessary for an

with the manufacturer of the water-
purification system when the system is
purchased or modified.
6.1 Requirements for Reagent Water
application, then Type I water must be used
immediately after production in order to avoid
the rapid decrease in resistivity due to carbon
dioxide absorption from the air or
solubilization of ions from the container.

The specifications for reagent water Types I, The resistivity requirement for Type II water
II, and III are summarized in Table 2 on the is slightly more liberal than what will occur
next page. when theoretically pure water is allowed to
equilibrate with ambient room air.
6.1.1 Microbial Content
Theoretically, pure water that has equilibrated
Ideally, Type I water should be free of with room air (thereby absorbing carbon
microorganisms. (For a discussion of the dioxide, which hydrates to form carbonic acid
effects of microbial contamination, see and then dissociates to form conductive ions)
Section 9.1.2.) It is recognized, however, has a typical resistivity of about 0.8 megohm
that water manufactured by a continuous @ cm, or about 1.2 FS/cm expressed as
process might not be sterile at all times. The conductivity.
working group suggests that the microbial
content specification for Type I water be #10
CFU/mL at the time of production. Similarly,
the working group suggests that 1,000
CFU/mL should be considered the upper limit
for microbial content in Type II water. When
microorganisms are present in water, the
microbial content of the water changes over
time.

6.1.2 pH

The pH requirements for reagent water Types

I and II are not specified. The other
specifications for Types I and II water are so
rigorous as to render a pH specification, a
less sensitive measure of purity, less than
useful. However, the pH specification of 5.0
to 8.0 for Type III water does provide a
measure of, and limit to, the impurity for this
grade.

6.1.3 Minimum Resistivity

The resistivity limits for the three grades of

reagent water define the allowable ionic
content of the respective water (see Section
9.4).

The 10 megohm @ cm at 25 BC resistivity cut-

off for Type I water defines an ionic
concentration of less than 10-6 gram
equivalent weight, which is close to the
concentration of the hydrogen and hydroxyl
ions contributed by the water itself. Should

NCCLS VOL. 17 NO. 18 8

October 1997 C3-A3

Table 2. Reagent Water Specifications

Type I Type II Type III
Maximum microbial 10 1000 NS
content, colony
forming units per
mL (CFU/mL)

pH NS NS 5.0–8.0

Minimum resistivity, 10 (inline) 1.0 0.1

megohm @ centimeter
(megohm @ cm 25 BC)

Maximum silicate mg/L 0.05 0.1 1.0

SiO2

Particulate matter* 0.22-µm filter NS NS

Organic contaminants* Activated NS NS

carbon
or distillation
or reverse osmosis

______________________________________________
*: This is a purification process requirement and is not measured by the end user.
NS: Not specified.

6.1.4 Maximum Silicate
Soluble or colloidal silica can be present in
the source water and it might not be
adequately removed in the purification
process. Silicates
or colloidal silica can interfere with certain
assays. Levels of 0.05 mg/L or below,
measured as SiO2, do not appear to cause
interferences. The higher level of silica as
specified for Type II water, may or may not
cause interferences; however, satisfactory
use should be documented.
6.1.5 Particulate Matter
Type I water should be free of particulate
matter, including microorganisms, larger than
0.2 µm. This can be achieved by passing the
water through a post-membrane (vinyl) filter
with a mean pore size no larger than 0.22
µm. However, note that many users elect to
add a postmembrane filter with a pore size of
0.1 µm. Purification process requirements for
particulate
matter are not specified for Types II and III
reagent water.
6.1.6 Organic Contaminants
Organic contaminants should be kept to a
minimum in Type I water. The content of
organic material is reduced when water is
distilled, subjected to reverse osmosis, or
passed through activated carbon. A
combination of these processes can be more
effective in removing organic material. If
activated granular carbon is used, periodic
replacement of the activated carbon is
necessary. If distillation or reverse osmosis is
used, resistivity measurements might not
meet the requirements for Type I water.
6.1.7 Degradation of Water Quality
Because certain characteristics, such as
resistivity and microbial content, change
quickly once the water is produced, their
influence on
usability must be evaluated (see Section 7).

NCCLS VOL. 17 NO. 18 9

October 1997
7 Storage and Handling
The working group recommends that the
laboratory examine the acceptability of the
type of reagent water to be used and record
the rationale for this decision (see Section
8.3.3).
7.1 Type I Water
When Type I water is stored, its resistivity
will decrease, metals and/or organic
contaminants will be leached from the
storage container, and microbial
contamination can occur. However, these
changes may not have an impact on the
quality of certain testing applications.
Type I water can be thought of as the "ideal"
general purpose water that can be produced
with currently available water treatment/
purification technology and used at the time
of production. Type I water should be used
in test methods that require minimal
interference or when lack of interference from
water of lesser purity cannot be documented
or inferred.
7.2 Type II and Type III Water
Type II water is intended to provide the user
with water in which the presence of
microorganisms, resistivity, and silicates can
be tolerated. Type II water should be used
for routine test methods for which
requirements leading to the choice of Type I
or special reagent waters do not apply. The
handling of Type II water should be
consistent with its specific intended use.
Type III water can be used for washing
glassware, for preliminary rinsing of
glassware (for final rinsing, the water grade
suitable for the intended use of the glassware
should be used), and as source water for
producing a higher grade water.
Storage and distribution systems for both
Type II and Type III water should be
constructed of materials that will protect the
water from chemical or microbial
contamination. If water is removed from the
storage tank into a secondary vessel (flask)
for routine laboratory use, both the secondary
NCCLS VOL. 17 NO. 18
C3-A3
vessel and the water must be replaced daily.
Storing water in large vessels (carboys) for
extended periods of time is unacceptable
because of the inevitable, unpredictable rate
of degradation of the water quality.
7.3 Special Reagent Water
Special reagent water can be necessary for
highly sensitive analytical techniques, such as
HPLC and chromosomal analyses. Water of
this specified purity might be beyond that
attainable by the laboratory's existing
purification system. Such water is discussed
in Section 9.6 of this document.
7.4 Handling Precautions
To retard the further, inevitable, unpredictable
rate of water quality degradation, good
laboratory practice should be followed in
removing water from a container. The user
must not touch the lid or inside cover, or dip
pipets into the container. Water should be
poured for use into a secondary container,
and the unused portion must not be returned
to the original container.
8 Commercially Available
Reagent Water
8.1 Diluent
Water that is provided as a diluent by the
manufacturer of a particular analytic system
is intended for use only as described in that
system; it is not an acceptable substitute for
reagent water. Such water has been qualified
by the manufacturer specifically for the uses
stated in the product labeling.
8.2 Sterile Water
Typically, sterile (pharmaceutical) water is not
manufactured to meet the specifications of
reagent water and it should not be used as its
equivalent.
8.3 Purchased Water
The working group recommends the following
guidelines for the purchase of reagent water
and its use:
10
October 1997
8.3.1 Proper Labeling
Only water that is labeled with values
determined at the time of manufacture for
resistivity, microbial count, endotoxins, and
silica content should be purchased. Just as
with water prepared within the laboratory,
these measurements are valid only at the time
of manufacture, and some parameters can
begin to change immediately. Proper labeling
includes a lot number and expiration date
provided by the manufacturer.
8.3.2 Package Size
Water should be purchased in a package size
appropriate to the usage rate. Refer to
Section 7 for storage information. Water
should be packaged so as to protect it from
environmental contaminations and from the
effects of the container itself (e.g., leaching
of electrolytes from "soft" glass).
8.3.3 Lot-to-Lot Quality Assurance
The working group recommends that the
laboratory monitor the quality of each lot of
reagent water, as with other reagents, for
successful use in the respective clinical test
application.
9 Testing
It is essential to monitor water quality
through testing that addresses those
contaminants that are found in the source
water or the end product. Monitoring is
required at regular specified time intervals.
The time intervals can be seasonally
dependent for some contaminants; however,
microbial content should be monitored at
least weekly. Although such testing can be
retrospective, it provides the laboratory with
useful data and it can be helpful in the
detection of trends and impending problems.
Additional testing is necessary when a
component of the water-purification system is
changed or inline monitoring devices indicate
a decrease in the water purity. Assays to
test for water purity described in this
guideline use ACS reagent-grade or equivalent
chemicals. All tests performed, results,
actions taken, and repeated test results
should be recorded in an appropriate log (see
NCCLS VOL. 17 NO. 18
C3-A3
Appendix B, “Quality Assurance
Procedures”).
9.1 Microbial Content
9.1.1 Areas to Consider
Three areas related to microbial
contamination merit consideration:
(1) Permissible levels of microorganisms that
might be present when different types of
reagent grade water are produced.
(2) The effect of different levels of
microorganisms on the accuracy and
precision of clinical laboratory tests.
(3) The effect of microbial load on other
significant reagent water parameters
(e.g., total organics).
9.1.2 Effects of Microbial Contamination
Ideally, reagent water should be free of
microorganisms; the production and storage
of reagent water, however, make this
difficult, if not impossible. Microorganisms
can affect the quality of reagent water in the
following ways:
! Inactivating reagents or altering
substrates or metabolites by enzyme
action.
! Contributing to the total organic content
of the water.
! Altering the optical qualities of the water
and causing high background absorbance
in spectrophotometric analyses, if
microbial clumps are present.
! Producing pyrogens/endotoxins.
Laboratory tests that use adequate controls
and standards should detect reagents that
perform suboptimally. Certainly, the quality
of the reagent water used to reconstitute
such reagents can be the source of a problem
and excessive microorganisms its cause.
Manufacturers should establish and specify
the type of reagent water necessary to
reconstitute their reagents.
11
October 1997
9.1.3 General Guidelines
The limits suggested in this document for
total microbial count and storage times of
water were established after consideration of
the preceding comments. The methods are
readily adapted to supplies present in clinical
laboratories and use techniques familiar to
clinical laboratory personnel. Clinical
laboratory personnel should evaluate past
experience in their own laboratories and they
should recognize that excessive levels of
microorganisms can cause problems in
maintaining other reagent water parameters.
9.1.3.1 Quantitation
! Microbial content should include the
evaluation of total colony count
determined by a standard method, after
incubation at 36 ± 1 BC for 24 hours,
followed by 24 hours at 23 ± 3 BC.
! Microbial content is reported as colony
forming units per milliliter (CFU/mL).
9.1.3.2 Sampling
Following is the procedure for sampling:
(1) Collect the sample in a sterile, closed
container that is large enough to hold the
entire sample. Leave ample air space to
allow mixing of the sample before
examination.
(2) Open the tap fully for a minimum of 1
minute before sampling, then restrict the flow
to fill the container without splashing. The
working group emphasizes that inadequate
flushing is one of the most common causes
of an elevated microbial count.
(3) Collect a minimum of 10 mL of water
from
each sample site.
(4) Process the sample within 1 hour of
collection, or within 6 hours when stored at 2
to 8 BC.
NCCLS VOL. 17 NO. 18
C3-A3
9.1.4 Methods of Determining Total
Microbial Counts
9.1.4.1 General Guidelines
Organisms that frequently contaminate water
are gram-negative rods, including
representatives of the genera Pseudomonas,
Alcaligenes, Flavobacterium, Klebsiella,
Enterobacter, Aeromonas, and Acinetobacter.
Total microbial count procedures provide a
standardized means for determining the
population density of aerobic and facultatively
anaerobic heterotrophic microorganisms in
water. This is an empirical measurement
because microorganisms occur singly, in
pairs, chains, clusters, or packets. Also, no
single medium or set of physical or chemical
conditions can be assumed to support the
growth of all microorganisms in a water
sample. Consequently, the number of viable
microorganisms can be higher than the
number of colony forming units.
9.1.4.2 Criteria for Selecting a Method
Various standards that relate to the quality of
water recommend a variety of procedures to
determine total microbial count, including
pour-plate, filtration, and bacteriologic
sampling methods. In choosing a method,
the decision should be based on the following
considerations:
! Sensitivity of the method.
! Use of media that will support the growth
of microorganisms most commonly
isolated from water in the time frame
designated.
! Supplies available to perform the
procedure.
! Cost of the procedure.
9.1.4.3 Limitations of Methods
The methods recommended are not inclusive
of all methods that might satisfactorily
comply with the above objectives. Different
methods can recommend sampling different
volumes of water, especially when
commercially available microbial enumeration
12
October 1997
kits are used. The sensitivity of a method is
enhanced by sampling more than 1 mL of
water, because it is possible to detect
contamination at levels less than 1 CFU/mL.
Larger volumes also help negate distributional
problems with suspended microorganisms in
fluids.
To provide a representative distribution of
microorganisms for a total microbial count for
clinical laboratory reagent water, sufficient
vortexing or mixing of a specimen is
extremely important. If a commercially
available kit enumeration system is used,
follow the instructions for sampling and
enumeration and convert the results to
CFU/mL when determining compliance with
this guideline.
The working group does not recommend the
calibrated loop method because it lacks
sensitivity in determining colony counts of
less than 100 CFU/mL.
9.1.5 Incubation Conditions for
Determining Total Microbial Counts
9.1.5.1 Incubator—Air or Heat Sink
For incubation in an air or heat sink, the steps
are as follows:
(1) Maintain a temperature of 36 ± 1 BC;
check the temperature using a thermometer
with calibrations traceable to a National
Institute of Standards and Technology (NIST)-
certified thermometer
(2) Monitor moisture content periodically
(many incubators have humidity controls; a
relative humidity of 70 to 96% at 36 BC is
desirable)
(3) To provide the necessary humidity if no
humidity control is available for the incubator,
keep a pan of water in the incubator
chamber.
9.1.5.2 Incubation Conditions
To maintain proper incubation conditions, the
steps are as follows:
NCCLS VOL. 17 NO. 18
C3-A3
(1) Incubate the samples at 36 ± 1 BC for
24 hours
(2) Then incubate the samples at the
ambient temperature (23 ± 3 BC) for an
additional (minimum of) 24 hours.
The total incubation time is at least 48 hours.
9.1.6 Standard Plate Count (SPC)
9.1.6.1 Equipment
To perform the standard plate count, the
following equipment is necessary:
! Incubator (see Section 9.1.5.1).
! Binocular (dissection) microscope with
total magnification of 10 or 15X and
adequate illumination, or a Quebec colony
counter.
! Petri dishes, sterilized (15- X 100-mm).
! Vortex mixer (optional).
9.1.6.2 Reagents
Heterotrophic plate count agar (HPC),
trypticase soy agar (TSA), brain heart infusion
agar (BHI), standard plate count agar (SPC),
R2A media, or any suitable transparent
medium capable of supporting the growth of
the organisms mentioned in Section 9.1.4.1,
should be prepared according to the
manufacturer's instructions.
9.1.6.3 Procedure
The procedural steps for SPC are as follows:
(1) Mix a 10-mL water specimen by
vortexing for 10 seconds or by multiple
inversions.
(2) Transfer 1 mL from the sample to a petri
dish using a 1-mL pipet.
(3) Using aseptic technique, pour
approximately 15 mL of the melted agar
medium, cooled to 46 to 50 BC, into each
petri dish that contains a sample. Carefully
13
October 1997
mix the agar and inoculum by rotation. Allow
the agar to solidify.
(4) Invert the dishes and incubate them as
discussed in Section 9.1.5.2.
(5) Count the colonies and report the number
of viable colonies per plate (1-mL sample) as
CFU/mL. The use of magnification is
preferred for the enumeration of CFU/mL.
9.1.7 Filtration Methods†
9.1.7.1 Membrane Filtration
The following equipment is necessary for the
performance of membrane filtration:
! Equipment
— Incubator (see Section 9.1.5.1)
— Binocular (dissection) microscope
with a total magnification of 10 or
15X, or a Quebec colony counter
— Petri dishes, sterilized and
disposable (12- X 50-mm) with
absorbent pads and a tight lid
— Screw-capped dilution bottles (100-
mL)
— Membrane filters, 0.45-Fm, 47-mm,
white-gridded
— Filtration units—filter funnel,
manifold or vacuum filter flask,
tubing, and a vacuum source
— Alcohol or gas flame
— Forceps
— One-hole rubber stopper.
! Reagents
Use the following reagents in the
performance of membrane filtration:
— Total count media or any suitable
broth medium capable of supporting
†
Two representative filtration methods are described.
NCCLS VOL. 17 NO. 18
C3-A3
growth of the organisms mentioned
in Section 9.1.4.1
— 0.1% peptone water
— Sterile dilution water
— Alcohol.
! Procedure
The procedure for membrane filtration is as
follows:
(1) Dispense 1.8 to 2.0 mL of culture broth
evenly onto absorbent pads.
(2) Assemble, using standard aseptic
technique, the sterile filter funnel or filter
holder by inserting the funnel base into the
filtering flask or manifold, using a one-hole
rubber stopper of appropriate size.
(3) Place (using forceps sterilized by dipping
in ethyl alcohol and flaming) a sterile
membrane filter on the filter base, grid-side-
up, and attach the funnel to the base of the
filter unit.
(4) Shake the water sample vigorously 25
times.
(5) For sample volumes of 1 mL, to disperse
cells evenly, add 20 mL of sterile peptone
water to the funnel before adding the sample.
(6) Filter the sample and rinse the sides of
the funnel at least twice with 20 to 30 mL of
sterile dilution water. Turn off the vacuum
and remove the funnel from the filter base.
Aseptically remove the membrane filter from
the filter base and place it grid-side-up on the
pad.
(7) Incubate as described in Section 9.1.5.2.
(8) Remove the dishes from the incubator
after incubation and count the colonies, using
magnification as required.
(9) Express the results in CFU/mL.
14
October 1997
9.1.7.2 Filtration Kit Method
The following equipment is necessary for the
performance of the filtration kit method:
! Equipment
— Incubator (see Section 9.1.5.1)
— Magnifying glass, dissecting
microscope, Quebec colony counter
or similar magnification device
— Sterile collection vessel
— Enumeration or field monitor kit
— Sterile syringe with locking hub or
vacuum pump (method dependent).
! Reagents
Use the reagents contained in the kit.
! Procedure
The manufacturer's recommendations should
be followed. The technique described below
is general and can vary with different
manufacturers. The steps are as follows:
(1) Collect the recommended volume of test
sample into the sterile container.
(2) Assemble the filtration device and
connect it to a negative pressure source
(syringe or pump as appropriate).
(3) Draw the test sample into the filtration
device.
(4) Add culture medium if it is not already
contained in the device.
(5) Incubate the device according to the
manufacturer's instructions and using the
conditions specified in Section 9.1.5.2.
Some devices can be incubated without
disassembly. These require inversion so that
the filter membrane is facing down. Other
devices require the filter membrane to be
removed, placed on a petri dish with the
appropriate medium, inverted, and incubated.
NCCLS VOL. 17 NO. 18
C3-A3
! Counting Colonies
For counting colonies, the steps are as
follows:
(1) Use a gridded filter pad to assist in
counting viable CFU.
(2) Use a magnifier—such as a lens,
dissecting microscope, or Quebec colony
counter—to count colonies (without
magnification, two colonies can appear as
one).
(3) Count systematically, using the grid-lines
to establish a consistent pattern.
(4) If there are many colonies, use the
number in 10 randomly located grid squares
to estimate the total number on the filter
square. Determine the number of grids on
the filter. The CFU/vol tested is calculated
as:
CFU/Vol Tested = No. Grids @ No. CFU
10
(5) After counting the colonies, record the
number and convert it to CFU/mL.
9.1.7.3 Bacteriologic Sampler Method
Bacteriologic samplers for enumerating
CFU/mL in water are commercially available.
Fluid is absorbed by a pad impregnated with
dehydrated SPC broth.
! Equipment is an incubator as described in
Section 9.1.5.1.
! Reagents are supplied by the
manufacturer.
! Procedure
The procedural steps are as follows:
(1) Remove the paddle from the case, while
taking care to avoid touching the membrane
with one’s hands or other objects.
(2) After adequate flushing, procure the
water specimen as recommended by the
manufacturer.
15
October 1997
(3) Reinsert the paddle and immerse for the
specified time.
(4) Decant the water and incubate the
paddle according to the manufacturer's
instructions.
(5) Determine CFU/mL.
9.2 pH
9.2.1 Neutral Salt
Pure water has so few ions present that a
neutral salt, such as potassium chloride (KCl),
should be added to the sample to obtain a
stable pH reading.
9.2.2 Reagents
9.2.2.1 Reference Buffer Solutions
It is recommended that at least two buffers
be chosen: pH 7.0 and one lower than pH
7.0. Commercially available buffers may be
used if they are verified against NIST
reference buffers.
9.2.2.2 Potassium Chloride, Saturated
To avoid contaminating the water, which
could result in erroneous conclusions about
the purity of the water in question, this
solution of KCI should be free of all acidic and
alkaline impurities.
9.2.3 Standardization
The room temperature should be recorded
and the exact pH for the buffers at that
temperature should be used to standardize
the pH meter. The manufacturer's
instructions for standardization should be
followed.
9.2.4 Procedure
The procedural steps are as follows:
(1) Standardize the pH meter.
(2) Rinse the electrodes with at least three
changes of water using a flowing stream.
NCCLS VOL. 17 NO. 18
C3-A3
(3) Add one drop (.30µL) of saturated KCl to
a 50-mL water sample.
(4) Take the pH reading as quickly as
possible following the manufacturer's
instructions.
(5) Read the pH to the nearest 0.1 unit.
9.3 Resistivity
The resistivity is inversely proportional to the
ionic content of water: The higher the
water's ion concentration, the lower its
resistivity. Therefore, resistivity
measurements are useful to assess the ionic
content of the water. The measurement is
sensitive to the point where the only ionized
species are the hydrogen and hydroxyl ions
contributed by the water itself. At a
resistivity of 10 megohm @ cm at 25 BC, the
cutoff for Type I water, the concentration of
ionic species is less than 10-6 gram equivalent
weight per liter. In water of higher resistivity,
the decrease in ionic contamination is
extremely small.
The measurement of the resistivity of water
gives a nonspecific indication of the presence
and concentration of ionized chemical species
only. The resistivity cannot indicate the
presence, nature, and concentration of
nonionized chemical species, or of ionized
chemical species at the level of parts per
billion. When accurate determinations of
chemical species at the level of parts per
billion are considered, the water used must be
tested for the trace species under
consideration. If present, such trace species
may be tolerated only at a concentration that
is not significant for the concentration level at
which the analysis is to be performed.
9.3.1 Equipment and Materials
9.3.1.1 Resistivity Meters
Commercially available resistivity meters
range from battery-operated, null indicating
meters to microprocessor-based digital
display meters. Meters with analog display
are also acceptable.
16
October 1997 C3-A3

Several manufacturers of reagent grade water 9.3.2.2 Temperature Compensation

systems have built-in resistivity meters with
temperature compensation that are designed
to provide maximum accuracy in the range of
water quality produced. The manufacturer’s
instructions should be followed in using all
meters.
9.3.1.2 Thermometer
A thermometer with graduations to 0.1 BC
and with calibration traceable to an NIST-
certified thermometer (if using a
nontemperature-compensated meter) is
required.
The resistivity of a given sample of pure
water varies significantly with temperature
(see Table 3). As a result, at standard
reference temperature (25 BC), Type I water
measures approximately 10 megohm @ cm. It
is essential to correct all resistivity
measurements at other temperatures to 25 BC
for meaningful interpretation. By convention,
the resistivity of solution is the resistivity it
exhibits at 25 BC.
(1) Type I water must be measured with a
temperature-compensated inline cell.

9.3.1.3 Potassium Chloride (2) Type II water is more accurately

Use anhydrous, high-purity potassium
chloride (KCl) to prepare calibration solutions.
To prepare a 0.01 mol/L-solution, dissolve
0.7455 g of KCl in Type I water, and dilute to
1.000 L using calibrated volumetric,
glassware. This 0.01 mol/L KCl solution will
have a resistivity of 70.7 kS@cm (electrolytic
conductivity, 1414 µS/cm) at 25.0 BC. To
prepare a 0.001 mol/L solution, dissolve
0.0746 g of KCl in Type I water, and dilute to
1.000 L using calibrated volumetric
glassware. This 0.001 mol/L KCl solution will
have a resistivity of 6.81 kS@cm (electrolytic
conductivity, 146.9 µS/cm) at 25.0 BC.
measured with an inline cell, but it can be
measured with a dip type cell if the
measurement is made as close as possible to
the time of production. It is recommended
that the water be maintained at 25 ± 2 BC
during the measurement.
(3) Type III water can be measured using a
dip-type cell. It is recommended that the
water be maintained at 25 ± 2 BC during the
measurement.
There are three methods of temperature
correction or compensation (see Appendix C).

Alternatively, commercially available

conductivity standards, as well as Standard
Reference Materials from the National
Institute of Standards and Technology, can
be used.

9.3.2 Sampling and Measuring

9.3.2.1 Daily Measurement

The resistivity of Type I water should be

recorded daily. It is recommended that the
resistivity of Type II water also be measured
daily.

The working group recommends that the

inline cell be installed as close to the point of
use as possible. Because it has been
recommended that the system recirculate, at
the minimum, the cell should be installed
downstream from the last point of use.

NCCLS VOL. 17 NO. 18 17

October 1997 C3-A3

Table 3. Resistivity Temperature Table (10 (2) Immerse the cell and temperature sensor
Megohms @ cm Type I Water) in the solution; move it up and down using a
circular motion to remove any entrapped gas
bubbles. Take a reading immediately.
Temperature Resistivity
BC Megohms @ cm
(3) If necessary, apply the manual
0 28.25 temperature compensation, according to
Appendix C.
5 22.65
NOTE: Readings taken with dip cells are
10 18.27 somewhat less accurate because
15 14.85 solutions are open to the
atmosphere. At purity levels above
20 12.15 1 megohm @ cm, dissolving
atmospheric CO2 will cause the
25 10.00 measurement to drift and ultimately
result in artificially low resistivity
30 8.28
readings.
35 6.91
9.3.3 Calibration
40 5.80

45 4.90
50 4.16
9.3.2.3 Measurement of Type I Water
The 10 megohm @ cm resistivity of Type I
water can be measured only by using an
inline cell. Measure according to the
manufacturer’s directions. Automatic
temperature compensation is required.
9.3.2.4 Measurement of Types II and III
Water
The resistivity of Types II and III water can be
measured using an inline or a diptype cell.
For inline meter/cell combinations, the
manufacturer’s instructions should be
followed.
When measuring the resistivity with an offline
meter/cell combination, the procedure is as
follows:
(1) Rinse the cell and container at least three
times with separate aliquots of water to be
tested. It is recommended that the water be
thermostated to 25 ± 2 BC during the
measurement. Otherwise, either automatic or
manual temperature compensation should be
used.
The resistivity meter and conductivity cell
should be calibrated and checked in
accordance with manufacturer’s instructions.
Some meters have an automatic check;
others have a manual test button. Note that
this type of calibration check involves
disconnecting the meter cell (either
automatically, by pressing a button, or
manually) and inserting a resistor of known
value in place of the cell to obtain a known
meter display. This only checks the
functionality of the meter electronics and
does not verify accuracy of the cell. For
offline meters and some online meters, a KCl
solution or certified conductivity standard can
be used to calibrate and monitor the day-to-
day operation of the meter/cell system. The
0.01 mol/L KCl solution has a resistivity of
70.7 kS @ cm (electrolytic conductivity, 1414
µS/cm) at 25.0 BC. The 0.001 mol/L KCl
solution has a resistivity of 6.81 kS@cm
(electrolytic conductivity, 146.9 µS/cm) at
25.0 BC. Standard reference materials
(SRMs) with certified values of electrolytic
conductivity are available from the National
Institute of Standards and Technology,
ranging from 10 S@cm (100,000 µS/cm) to
0.2 MS@cm (5 µS/cm). Commercially
available conductivity standards may also be
used. Meter tolerances will vary dependent
on the manufacturer. Generally, newer
meters are accurate to ± 5% of the actual
reading. Older models are accurate to ± 3%

NCCLS VOL. 17 NO. 18 18

October 1997
of full scale. Consult the manufacturer to
determine the expected degree of accuracy
for the meter in use.
9.3.4 Source of Error
Reading accuracy is the sum of the errors
contributed by the environment and the
various components of the measurement set-
up. The following errors and components are
included:
! Instrument accuracy: Instrument error,
worst case, is the stated accuracy for the
range being used. Typical numbers are
±1 to ± 10% of full scale.
! Cell-constant error: Cells are
manufactured to a nominal cell-constant
value. Typical numbers are ±1 to ± 8%.
This error should be taken into account in
determining the reading accuracy.
! Solution temperature offset: This is the
product of the temperature coefficient
and the temperature offset from 25 BC,
expressed as a percentage of the reading,
which would have been obtained at 25
BC.
The error is not necessarily a linear
function of temperature. Typical numbers
are ±1 to ±2% of full scale.
! Cell contamination and air bubbles.
! Cell platinization problems.
! Electrical noise.
! Cable series resistance.
! Galvanic effects.
The first three bulleted items are expected to
contribute the most substantial error.
Meter tolerances vary depending on the
manufacturer. Consult the manufacturer to
determine the expected degree of accuracy
for the meter in use.
NCCLS VOL. 17 NO. 18
C3-A3
9.4 Soluble Silica
Soluble silica in the water supply is a major
problem in certain geographical locations.
Silica can affect enzyme determinations, as
well as trace metal and electrolyte analyses.
At high concentrations, silica can also directly
interfere with some spectrophotometric
measurements.
9.4.1 General Guidelines
The source water should be evaluated to
determine if a high concentration of silicates
is present. It is recommended that the
laboratory's requirements be discussed with
the manufacturer, and the water purification
system(s) necessary to produce water with
the desired specifications should be selected.
The appropriate system or combination of
systems should eliminate the need to test
silicates routinely. If silicate testing is
required, consider using an appropriate
reference laboratory to perform atomic
absorption analysis. Alternative methods are
described in Sections 9.4.2 and 9.4.3.
Colloidal silica may also be present in the
source water. It is removed only partially by
ion exchange procedures, but it is
successfully removed by distillation or reverse
osmosis. Colloidal silica should not be
present in purified water and it is difficult to
measure because it is unreactive when used
in the molybdate method described in Section
9.4.3.
9.4.2 Commercial Kit Method
A commercially available kit method may be
used if it can perform to the appropriate level
of sensitivity.
9.4.3 Molybdate Method
Silicates react with molybdate ion to form a
complex that can be reduced by 1-amino-2-
naphthol-4-sulfonic acid to produce a blue
color. The intensity of the blue color is
proportional to the concentration of soluble
silica.
19
October 1997
9.4.3.1 Reagents
Appropriate safety precautions should be
employed in the preparation of reagents. The
following reagents should be prepared and
stored in plastic containers:
! 1-Amino-2-naphthol-4 sulfonic acid
solution (14.45 mmol/L)
The steps for preparation and storage are
as follows:
(1) Dissolve 1 g of sodium sulfite
(Na2SO3) in 50 mL of Type I water.
(2) Add 0.5 g of 1-amino-2-naphthol-4-
sulfonic acid; mix to dissolve.
(3) Add this solution to 100 mL of a
solution containing 30 g of sodium
hydrogen sulfite (NaHSO3).
(4) Dilute to 200 mL with Type I water.
(5) Store in a dark, polyethylene bottle at
room temperature.
The solution is stable for two weeks
(discard when the color darkens or
precipitate forms).
! Ammonium molybdate solution, 75 g/L
(85.9 mmol/L)
The steps for preparation and storage are
as follows:
(1) Dissolve 7.5 g of ammonium
molybdate [(NH4)6 Mo7O24 C 4H2O] in Type
I water.
(2) Dilute to 100 mL with Type I water.
(3) Store in a polyethylene bottle at room
temperature.
The solution is stable for three months.
! Hydrochloric acid, approximately 6 mol/L
The steps for preparation and storage are
as follows:
NCCLS VOL. 17 NO. 18
C3-A3
(1) Slowly add one volume of
concentrated HCl to one volume of Type I
water.
(2) Store in a polyethylene bottle at room
temperature.
The solution is stable indefinitely.
! Oxalic acid solution, 100 g/L (1.11
mmol/L)
The steps for preparation and storage are
as follows:
(1) Dissolve 10 g of oxalic acid (H2C2O4 C
2H2O) in Type I water.
(2) Dilute to 100 mL with Type I water.
(3) Store in a polyethylene bottle at room
temperature.
The solution is stable for three months.
! Silica, stock standard solution (1 mL =
0.1 mg SiO2; 16.64 µmol)
The steps for preparation and storage are
as follows:
(1) Dissolve 0.473 g of sodium
metasilicate (Na2SiO3 C 9H2O) in Type I
water.
(2) Dilute to 1 L with Type I water.
(3) Store in a polyethylene bottle at room
temperature.
The solution is stable for three months.
9.4.3.2 Equipment and Materials
The spectrophotometer should be capable of
at least a 700-nm setting. The maximum
absorbance wavelength is 815 nm, which, if
available, is the preferable setting.
9.4.3.3 Sampling
For sampling, 100-mL water samples are
collected in chemically clean polyethylene
20
October 1997
bottles. The temperature of the water
collected should not exceed 35 BC.
9.4.3.4 Standardization
For standardization, the steps are as follows:
(1) Prepare, by properly diluting the stock
standard solution, a series of standards (100
mL each) covering a concentration range of
0.01 to 1.0 mg/L of SiO2. Treat 50-mL
aliquots of each dilution as described in the
procedure.
(2) Use a 50-mL aliquot of the best available
water (preferably Type I) carried through the
procedure as a blank. If Type I water is not
available for silica standard preparation, the
water can be assumed to be "silica-free" (less
than 0.05 mg/L) if the blank gives no visible
blue color in this assay procedure.
(3) Prepare a calibration curve, plotting
absorbance against silicate concentration as
SiO2 in milligrams per liter.
9.4.3.5 Procedure
All unknowns, standards, and the blank
should be run through the following
procedure in its entirety. The steps are as
follows:
(1) Transfer, quantitatively, 50 mL of sample
to be tested to a polyethylene or another
suitable plastic container.
(2) Add, in quick succession, 1 mL of 6
mol/L HCl and 2 mL of ammonium molybdate
solution; mix well.
(3) After exactly 5 minutes, add 1.5 mL of
oxalic acid solution and mix again.
(4) After 1 minute, add 2 mL of the amino-
naphthol-sulfonic acid solution; mix well and
allow to stand for 10 minutes.
(5) Zero the spectrophotometer with the
blank and measure the absorbance of the
sample (the wavelength of choice is 815 nm
or 640 nm for higher concentrations).
NCCLS VOL. 17 NO. 18
C3-A3
(6) Silica concentration, in milligrams per
liter, may be read directly from the calibration
curve. Record the reading in a permanent
log.
9.5 Organic Contamination
9.5.1 General Guidelines
Methods for assessing contamination with
organic compounds are plentiful, but they are
impractical for routine use in the clinical
laboratory. Therefore, the working group
chose not to recommend a specific method.
Spectrophotometric evaluation of water in the
far ultraviolet region is possible only with a
research-grade spectrophotometer. However,
if such instruments are available, they can be
of some use in evaluating the presence of
organic compounds in the manufactured
water. Methods based on the reduction of
potassium permanganate are not adequate
because they cannot detect many organic
compounds. HPLC procedures, if available,
can satisfy this need. If HPLC is unavailable
and organic contamination is suspected, the
water system manufacturer should be
consulted about the problem.
9.5.2 Procedure
The working group suggests the following
procedure:
(1) The water-production system should
include a purification process that is effective
in removing or significantly reducing dissolved
organic compounds, as described in Section 4
(2) If unusual analytical results or observable
changes in source water lead to a suspicion
of organic contamination, disposable system
components should be replaced and the
water quality should be reassessed.
(3) If the problem persists, the manufacturer
of the treatment/purification system should
be consulted.
21
October 1997 C3-A3

9.6 Special Water Considerations:

Endotoxins and Specifications

Endotoxins are heat-stable metabolic products

formed from the cell walls of viable and
nonviable gram-negative bacteria. When
injected into humans in small quantities,
pyrogenic effects are observed and death can
follow. However, when ingested little affect
is seen in humans.

However, laboratory procedures can be

altered by the presence of endotoxins in
reagent-grade water. The limulus ameobcyte
lysate test (LAL) is used to measure
endotoxin levels in water and other materials.
Although there is no LAL standard limit for
reagent-grade water, as with pharmaceutical
water(s), some laboratories use a cut-off level
at 0.25 endotoxin units/mL for reagent-grade
and special-purpose water.

The specifications for special water may vary

according to the application for which it is
intended. End users are encouraged to refer
to published literature references specific for
the intended use.

NCCLS VOL. 17 NO. 18 22

October 1997
Bibliography
American Society for Testing and Materials.
Standard test method for silica in water.
ASTM designation D859–88. Conshohocken,
PA: ASTM; 1988.
American Society for Testing and Materials.
Standard test methods for electrical
conductivity and resistivity of water. ASTM
designation D1125-82. Conshohocken, PA:
ASTM;1989.
American Society for Testing and Materials.
Water: atmospheric analysis. In: 1990 Annual
Book of ASTM Standards, Conshohocken,
PA: ASTM;1990; (11.1; 11.2).
HA-AWWA-WPCF. Standard Methods for
Examination of Water and Waste Water.
17th ed. Washington, DC; 1989.
Batjer JD, et al. Effects of microbial
contamination of reagent water on selected
laboratory tests. Am J Clin Pathol. 1979; 71:
19–325.
Blais P, Cooper MT. Contaminants in clinical
reverse osmosis water purification systems.
JAMA. 1980;243:649.
Bowers GN Jr. Reagent quality: A basic
analytical consideration in clinical
enzymology. CAP/ASPEN. 1982 Enzyme
Conference.
Bristol DW. Detection of trace organic
impurities in binary solvent systems: a solvent
purity test. J Chrom. 1980;188:193.
Bukey M. Deadlegs: A widespread threat to
DI water systems. Ultrapure Water.
1987;4:66– 70.
Callaghan TJ. Practical guide for the selection
of a laboratory water purification system.
Am Lab. 1988;5:60–68.
College of American Pathologists,
Commission on Laboratory Inspection and
Accreditation. Reagent water specifications.
Chicago: CAP; 1985.
NCCLS VOL. 17 NO. 18
C3-A3
Gabler R, Hegde R, Hughes D. Degradation of
high purity water on storage. J Liquid
Chrom. 1983;6:2565–2570.
Ganzi GC. Preparation of high purity
laboratory water. In: Methods in Enzymology.
New York: Academic Press; 1984;(104).
Hamilton H. Selection of materials in testing
and purifying water. Ultrapure Water.
1985;2: 36–38.
Hanselka R, Reinzuch KJ, Bukey M. Materials
of construction for water systems part II:
Real life failure modes of plastics. Ultrapure
Water. 1987;4:50–53.
Hanselka R, Williams R, Bukey M. Materials
of construction for water systems part I:
Physical and chemical properties of plastics.
Ultrapure Water. 1987;4:46–50.
Highsmith AK. Water in health care facilities.
In: Architectural Design and Indoor Microbial
Pollution. Oxford: Oxford University Press;
1988:81–102.
Highsmith AK, Kaylor BM, Reed CJ, Ades
EW. Evaluation of Water Treatment Systems
Producing Reagent Grade Water. SAE
Technical Paper: Series 901424. Society of
Automotive Engineers; 1990.
Jorgensen JH, Smith RF. Rapid detection of
contaminated intravenous fluids using the
Limulus in vitro endotoxin assay. Appl
Microbiol. 1973:26;521–524.
Kaplan LA, Pesce AJ. Methods in Clinical
Chemistry. 2nd ed. St. Louis, MO: CV
Mosby; 1989.
Malaiyandi M, Cooper MT, Blais P. Reverse
osmosis units do not remove all water
contaminants. CMA J.1980;122:15–16.
Meltzer TM. High purity water preparation for
the semiconductor, pharmaceutical, and
power industries. Littleton, CO: Tall Oaks; p.
800.
23
October 1997
Nielsen SS, Highsmith AK, Crow SA. A
Comparison of Filamentous Fungal
Populations in Potable Water, Point-of-Use
Water, and Room Air. Atlanta, GA: Centers
for Disease Control and Prevention, Water
Quality Laboratory; 1995.
Poirier SJ, Sienkiewicz PM. Organic free
water. Am Lab. 1980;12.
Rechen HC. Piping system designed and
materials for optimal performance in ultrapure
water transmission. Ultrapure Water.
1985;2: 39–42.
Sullivan JD Jr, Valois FW, Watson SW.
Endotoxins: The Limulus amebocyte lysate
system. In: AW Bernheimer, ed. Mechanisms
in Microbial Toxicology. New York: John
Wiley and Sons; 1976:217.
NCCLS VOL. 17 NO. 18
C3-A3
Tietz NW, ed. Fundamentals of Clinical
Chemistry. 4th ed. Philadelphia, PA: WB
Saunders; 1996.
Tietz NW, ed. Textbook of Clinical Chemistry.
2nd ed. Philadelphia, PA: WB Saunders;
1994.
United States Pharmacopeial Convention, Inc.
US Pharmacopeia XXI. Rockville, MD; 1985.
Winstead M. Reagent Grade Water: How,
When and Why? Austin, TX: American
Society of Medical Technologists, Steck Co.;
1967.
Wood JH. Pure water systems for hospitals.
Plumbing Engineer Magazine. July/August,
1985.
24
October 1997 C3-A3

Appendix A. Description of Purification Processes

A.1 Distillation

Distillation separates water from contaminants by changing the state of water from a liquid phase
to a gas phase and then back to a liquid phase. Each of these transitions provides an opportunity to
move pure water away from contaminants that do not make the transition with the same efficiency
as water molecules. In theory, distillation can be the most predictable of all water-purification
technologies, provided the equipment is well designed and properly utilized.

Using a still, there are three stages in the laboratory distillation process: (1) the boiler stage, (2) the
transition stage, and (3) the condenser stage. The function of the boiler stage is to produce steam
that approaches equilibrium with the boiler water and not to overheat the steam to the point where
it carries an unnecessary burden of contaminants with low, but significant, vapor pressures. It is
also a function of the boiler to minimize the production of boiler water particles in the steam during
distillation. The third function of the boiler stage is to eliminate the concentrating residual
contaminants from the still without degrading the quality of the steam produced by the boiler. The
function of the transition stage is to smooth the inevitable imperfections in boiler operation by
trapping mist particles and by equilibrating overheated steam. The function of the condenser is to
remove sufficient heat calories from the steam to permit condensation while maintaining a
steam–water equilibrium. It is also the function of the condenser to mechanically separate the
distilled water from the steam phase without permitting the redissolution of volatile contaminants,
which must be vented from the condenser at greatly increased concentrations. For a condenser to
properly perform this function, a condenser must not enclose surfaces that are significantly cooler
than the boiling point of water, a requirement that tends to preclude the use of coils or water jacket
designs. The condenser should also be designed so that the output distilled water is mechanically
partitioned from the steam containing relatively high concentrations of contaminating volatiles, and
permitted to equilibrate with steam containing concentrations of volatiles no higher than the steam
leaving the transition stage.

Most dissolved ionized solids (i.e., common inorganic salts) have insignificantly low vapor pressures
at the boiling point of water and will not make the transition from boiler water to boiler steam.
These dissolved solids must be periodically flushed from the boiler. If the transition stage is not
effective, dissolved ionized solids can migrate into the condenser stage where they can no longer
be removed.

Most dissolved ionized solids (i.e., common inorganic salts) have insignificantly low vapor pressures
at the boiling point of water and will not make the transition from boiler water to boiler steam.
These dissolved solids must be periodically flushed from the boiler. If the transition stage is not
effective, dissolved ionized solids can migrate into the condenser stage where they can no longer
be removed.

Dissolved ionized gases include gases such as carbon dioxide, sulfur dioxide, and ammonia, which
form ions in solution. These gases are the major contributors in resistivity changes in distilled
water. If a still can produce water with a resistivity approaching 5 @ 106 S-cm, the boiler,
transition, and condenser phases are operating effectively. The resistivity of the water must be
tested by the flow cell at the point of exit from the still because ionized gases in the ambient air
will dissolve rapidly into the water and lower the resistivity.

The bulk of dissolved organics is removed from the distillate because they have a lower vapor
pressure or higher vapor pressure relative to water. Those organics that have vapor pressures close
to water, including azeotropes, might not be removed completely by distillation. One strategy to
minimize this problem is to feed the still boiler with water previously purified by another purification
system, for example, reverse osmosis or deionization.

NCCLS VOL. 17 NO. 18 25

October 1997 C3-A3

Particulate matter and microorganisms should not penetrate the transition stage of a properly
designed and maintained still. However, permitting the still to rinse down, after start-up and before
collecting water, is recommended.

Pyrogens and endotoxins are large molecules that should not penetrate the transition stage of a
properly designed and maintained still.

A.2 Deionization

Ion exchange is an effective way to produce water with a high resistivity, because strong-
acid/strong-base, mixed-bed exchange resins can effectively remove the mobile ions that elevate
the conductance of water. However, the user should bear in mind that many substances present in
water are only weakly ionized (e.g., silicates) or are essentially undissociated (e.g., many organic
substances). They will not be bound by some ion-exchange resins and are not detected by
resistivity measurements. It is inherent in the gel structure of the ion-exchange resins that they
leach a variety of substances into the water stream. Resin beds provide excellent conditions for
the growth of organisms, which results in recontamination of the water. If the resins are
regenerated by an exchange service, there is a possibility that small quantities of organic and
inorganic contaminants picked up during previous use will leach into the water stream, even under
the best of circumstances. It is important to understand the limitations of deionization and to have
the supplier provide information on the specific regeneration process. The manufacturer may be
asked to provide assurance that the regeneration process will minimize contamination from previous
use and/or regeneration itself. It is important to understand that the high specific resistance
produced by ion exchange is no assurance that deionized water is “pure.”

A.3 Reverse Osmosis

Reverse osmosis (RO) is a process in which water is forced under pressure through a
semipermeable membrane leaving behind a percentage of dissolved organic, dissolved ionic, and
suspended impurities. RO can effectively remove more than 97% of monovalent ions and an even
greater percentage of divalent ions. However, volatile substances concentrate as the result of the
RO process and many low-molecular-weight organic substances pass through RO membranes.

There are three major classes of RO membranes as described in the table on the next page.
Phenol, formaldehyde, and acetic acid are passed preferentially by cellulose acetate membranes and
their concentrations in the output water increase with respect to the source water. RO
concentrates contaminants on the feed side of the membrane, causing significantly increased
concentrations of sparingly soluble salts. This can result in scaling of the RO membrance. A
common pretreatment to address this problem is water softening.

NCCLS VOL. 17 NO. 18 26

October 1997 C3-A3

Reverse Osmosis Membrane Characteristics

CA PS PA

Cellulose Acetate Polysulfone Thin film composite

Feed pH 4–7.5 4–10 4–12

Operating 250–400 psig 50–200 psig 50–200 psig

Pressure* 6 6
(1.7 x 10 – 2.8 x 10 pascals) 5 6
(3.5 x 10 –1.4 x 10 pascals) (3.5 x 105–1.4 x 106
pascals)

Resistance to Medium to high Medium to high Low

free CI2

Remarks Hydrolyzes above Preferential Ca++ Activated carbon or

pH 7.5 and Mg++ rejection reducing agent
required for
chlorine reduction

Potential of Softener required

microbial degradation

Concentrate Yes No No
low-molecular-
weight organics

*Operating pressure is reported in pound per square inch gauge pressure (psig). For example, at sea level, the
pounds per square inch absolute pressure (psia) is 14.7, whereas the gauge pressure would read 0 psig
because nearly all pressure gauges use the surrounding air pressure as their reference point.

In theory, RO membranes prevent passage of microbial contaminants. Realistically there is always

the possibility of imperfections that will allow passage of these contaminants. Consequently,
treatment with biocidal agents is necessary to control the bioburdens that can damage or foul the
RO membrane. Many biocides cannot be used, because they will damage the membranes and/or
they are not adequately rejected by the membranes and are difficult to remove from the RO product
water. Consult the manufacturer for information about appropriate decontamination agents.

A.4 Carbon Adsorption/Absorption

Carbon adsorption/absorption is virtually always used as a pretreatment step and in combination

with other purification processes. There are special grades of activated carbon and other synthetic
adsorbents that exhibit excellent capabilities for removing organic contaminants. Their use,
however, is targeted toward specific compounds and applications. The geometry and quantity of
activated carbon used should be sufficient to maximize contact time and, therefore, the adsorption

NCCLS VOL. 17 NO. 18 27

October 1997 C3-A3

of substances in the water stream. Activated carbon beds are particularly effective in removing
chlorine from water. Although carbon beds are viable locations for microbial growth, the adsorptive
affinity for chlorine can greatly increase the growth of these organisms in downstream sections of
the purification system. The limitations of carbon use are as follows:

! Carbon is mechanically degraded to produce fines that must be captured downstream

! Leaches ash minerals into the water stream

! Only weakly adsorbs other contaminants as a function of contact time.

In practice, the primary use of carbon is to remove chlorine from water entering ion-exchange resin
systems. The working group recommends consulting with the manufacturer before using carbon
for water purification.

A.5 Filtration and Ultrafiltration

For purposes of this document, filtration is defined as a mechanical process used to remove
particulate matter (including microorganisms) that is 0.22 µm or greater in size. Ultrafiltration is the
process used for the mechanical/electrochemical removal of smaller-sized dissolved and suspended
impurities. By their nature, filtration traps particulates within the filter’s matrix, whereas
ultrafiltration retains particulates based on their size, shape, and electronic charge.

These techniques are used in combination with other purification processes. Filtration is often used
both at the beginning and end of a series of water-purification processes. Ultrafiltration is almost
always a pretreatment step. Limitations of both techniques include the following:

! A build-up of particulate matter, which blocks the flow of water

! Physical penetration of the filter media by the contaminants or particles.

A.6 Nanofiltration

Nanofiltration is an emerging membrane technology for water purification. Nanofilters possess

performance characteristics between those of ultrafilters and reverse osmosis membranes. Like
ultrafiltration and reverse osmosis, nanofilters are most often configured as spiral wound cartridges.

Performance specifications for the few nanofilters currently available vary. However, in all cases,
percent rejection of divalent ions (e.g., calcium, magnesium, sulfate) is much greater than the
rejection of monovalent ions, such as sodium and chloride. Like RO membranes, nanofilters provide
excellent removal of organics that have molecular weights greater than about 300. Because of
their much better rejection of calcium and magnesium over sodium, nanofilters are sometimes
referred to as “loose” or softening RO membranes. However, a typical nanofilter can be operated
at higher water recovery than a reverse osmosis membrane, thus reducing the amount of water
wasted. The safe operating pH of nanofilters is wide; however, like PA RO membranes, most
nanofilters have a low resistance to attack by chlorine in water.

Although a number of nanofilters are currently available, the commercial use of them at this time is
limited. However, given their unique properties, water systems for laboratories might begin
employing this emerging membrane technology.

NCCLS VOL. 17 NO. 18 28

October 1997 C3-A3

A.7 Chemical Oxidation

Though not widely used in clinical laboratory water-purification systems, ozone is gaining in
popularity. Ozone is 5 to 1000 times more effective as a biocide than chlorine on a weight basis
(depending on the organism and the form in which the chlorine is dispensed). In a sense, ozone is
the perfect biocide for reagent water; it breaks down to oxygen with a half life of about 25 minutes
at 20 BC and the breakdown can be accelerated with a UV source, which makes ozone easy to
remove at its point of use. Because ozone is such a powerful oxidant, it degrades RO membranes
and most polymeric plastics, including ion-exchange resins.

A.8 Ultraviolet Oxidation and Ultraviolet Sterilization

Ultraviolet oxidation results from the absorption of 185-nm light, which produces hydroxyl radicals
that oxidize organic material to smaller ionizable components. Recirculation of water over an
ultraviolet source and a deionizer can result in a significant reduction of organic material.
Ultraviolet oxidation does not guarantee removal of all organic substances from water.

Ultraviolet sterilization results from the absorption of 254-nm light, which damages the DNA and
RNA of microorganisms causing cell death. The efficiency of both processes depends on the
amount of light that actually penetrates the water; the design of the devices is an important factor.
While ultraviolet sterilization kills microorganisms, the dead casts are not degraded or removed,
which can potentially lead to pyrogen formation.

Limitations of ultraviolet oxidation and ultraviolet sterilization include the following:

! Lamps aging (reduced light output)

! Build-up of ultraviolet absorbing films
! Presence of ultraviolet absorbing or scattering particles in the water.

Both ultraviolet oxidation and ultraviolet sterilization are used in combination with other water-
purification processes and they are often positioned in a recirculating loop.

NCCLS VOL. 17 NO. 18 29

October 1997 C3-A3

Appendix B. Quality Assurance Procedures

To ensure sample quality throughout sample collection and analysis, a set of operating procedures
should be established by each laboratory. A quality assurance program should include both quality
control and quality assessment. Guidelines addressed in the 17th edition of Standard Methods for
the Examination of Water and Wastewater can serve as the basis for program development. The
subsequent procedure manual should be prepared in accord with NCCLS document GP2-A3, Clinical
Laboratory Procedure Manual—Third Edition; Approved Guideline.

When standard operating procedure manuals are compiled, consideration should be given to the
following topics:

! A quality assurance plan with approval signatures, organizational charts, and responsibilities.

! Procedures for the preventive maintenance of equipment: Frequency of these procedures

should be determined through discussions with manufacturers of the system and based on the
performance of the product water.

! Calibration procedures, corrective actions, internal quality control activities, performance

audits, and data assessments.

! Special worksheets designed for reporting the results of daily, weekly, and monthly tests, etc.

! Quality control checklists.

Resistivity checks should be recorded daily. Microbiological testing should be performed weekly.
All other parameters should be tested as necessary, depending on geographical and seasonal
considerations, or the manufacturer's recommendations.

NCCLS VOL. 17 NO. 18 30

October 1997 C3-A3

Appendix C. Methods for Correction or Compensation of Resistivity Measurements

Following is a discussion of three methods used in the correction or compensation of resistivity

measurements:

(1) In automatic temperature compensation, the meter’s probe cell contains a temperature
compensation circuit element that internally adjusts the meter to read corrected resistance
directly.

(2) In manual temperature compensation, automatic compensation is not provided. An accurate

temperature measurement should be taken simultaneously with the resistivity measurement.
When the temperature of the water measurement is outside the 25 ± 2 BC range, correct the
resistivity reading to 25 BC according to the meter instructions. In the absence of meter
instructions for manual correction, determine the temperature coefficient empirically. Measure
the subject solution at two different temperatures (preferably at 5 BC above and at 5 BC below
the expected measurement temperature), then perform the following calculation :

Dt is the resistivity reading at temperature t.

D1 is the resistivity reading at temperature t1, 5 BC above temperature t.
D2 is the resistivity reading at temperature t2, 5 BC below temperature t.
Davg is the arithmetic average of the resistivity readings at temperatures t1 and t2.
D25 is the calculated, temperature-compensated resistivity at 25 BC, temperature t25.
ªt is the absolute value of the difference between t25 and t, rounded to the nearest whole
number.
f is the average fractional change in resistivity per degree Celsius at temperature, t.

D2 % D1
Da v g'
2

D2 ! D1
f '
(t1 ! t2)D a v g

If t is above 25 BC, then use the following equation:

D25 = Dt (1 + f)ªt

If t is below 25 BC, then use the following equation:

D25 = Dt (1 ! f)ªt

In the temperature range of 0 to 50 BC, the so-calculated, temperature-compensated resistivity

at 25 BC should be within 5% of the resistivity value, had it actually been measured at 25 BC.

NCCLS VOL. 17 NO. 18 31

October 1997 C3-A3

(3) The following temperature compensation equation is known as Marsh and Stokes’ equation.
This equation converts any measured electrolytic conductivity reading to its equivalent value at
25 BC.

where:

6t is the electrolytic conductivity in µS-cm, at temperature t.

625 is the electrolytic conductivity in µS-cm, at 25 BC.
t is the temperature of measurement in BC.
D25 is the calculated, temperature-compensated resistivity at 25 BC, temperature t25.

Resistivity at 25 BC is obtained by using the following relationship:

D25 = 1/625

NCCLS VOL. 17 NO. 18 32

October 1997 C3-A3

Summary of Comments and Working Group Responses

C3-A2: Preparation and Testing of Reagent Water in the Clinical Laboratory—Second Edition;
Approved Guideline

General

1. As a provincial reference laboratory, we are considering setting up a quality assurance

program in the near future for monitoring reagent water quality used in clinical laboratories
of different hospitals. NCCLS document C3-A2 has proven helpful in the planning of the
projected program.

However, in Type I water specifications, particulate matter and organic contaminants are
covered by a process specification (i.e., not measured by the end user). How then is
someone expected to discover if the process is still efficient after a given time? If an
interlaboratory quality assurance program, such as the one we are considering setting up,
were to include those two parameters, how should they be measured and according to
what "standard" should they be interpreted?

! The C3-A3 guideline recommends a process for achieving the types of water described.
The working group does not recommend a procedure for total organic contaminants and
particulates. Refer to the following ASTM standards for specific information on these
processes:

- ASTM. Standard Test Method for Total, Organic, and Inorganic

Carbon in High Purity Water by Ultraviolet (UV) or Persulfate Oxidation,
or Both, and Infrared Detection. ASTM designation D4779-93. ASTM,
West Conshohocken, PA, 1993.

- ASTM. Standard Test Method for Total Carbon and Organic Carbon in
Water by Ultraviolet, or Persulfate Oxidation, or Both, and Infrared
Detection. ASTM designation D4839-94. ASTM, West Conshohocken,
PA, 1994.

2. Are you aware of any recent reference dealing with quality assurance programs for
monitoring high-purity water? Even after computer searches through data banks, the only
useful reference we came up with was "Quality Control for a 14 to 18 Megohm-cm
Deionized Water Supply," by EJ Dravian, I Schoen, F Abrams, P Datta, EM Custer, Arch
Pathol Lab Med, 1986, 110, 228:231.

! The bibliography in C3-A3 was updated to include references for quality assurance
programs. Note that the article cited in your comment recommends a specification of 100
CFU/mL for Type I water. This level of microbial content was determined using a “loop”
method of inoculation. This method is less sensitive and cannot be considered equivalent
to the method described in C3-A3.

3. C3-A2 is a well-informed, well-documented, comprehensive, and clearly written guideline

that addresses the needs of the modern clinical laboratory. Definitions are presented
objectively; footnotes are appropriate and relevant.

Design considerations, construction materials, and specifications for systems used to

contain and to transport reagent water and specifications for Types I, II, and III of reagent
water, their preparation, intended use, verification, and handling are practical topics and
address "real-life" situations.

NCCLS VOL. 17 NO. 18 33

October 1997 C3-A3

This is one of the best NCCLS guidelines I have reviewed.

! The working group appreciates the comment.

4. The proposed standards may or may not be equivalent to existing similar standards from
ACS, ASTM, and USP. Because it is not clear why a new set of guidelines is necessary, it
seems to complicate the choice as to what is appropriate water quality for a particular
application.

! The purpose of all NCCLS documents is to put useful references in the hands of
laboratorians. The C3-A3 guideline focuses on clinical laboratories. The documents from
ACS, ASTM, and USP are valid standards written for different purposes and have served
as references for this guideline.

5. The basic concept that the level of purity of water can be established by conductivity
measurements alone is incorrect. Water that exhibits low conductivity may well contain a
variety of organic compounds; resin particles; colloidally dispersed metals; silica in solution
or silica colloidally dispersed; and some gases. It may also have any number and variety of
microbiological organisms. It follows that the concept that variations in levels of
conductivity reflect levels of purity is similarly incorrect.

Using water that is validated only with a conductivity parameter can directly influence
laboratory results. For example, colloidal iron will not influence conductivity, but it can
completely obscure a low serum iron. Silica in water will not alter conductivity, but it can
inhibit enzyme activity. The possibilities for errors involved in immunofluorescent assays
are high. I find the use of any single purity criterion inadequate.

The United States Pharmacopeial Convention (USP)-grade (purified water) is the most
generally used in industries (pharmaceutical), and it is already available in most hospital
and other laboratories. It is the starting point for higher levels of purity, such as water for
irrigation and water for injection. It is a satisfactory product for most laboratory tests.
The highest level of purity of water is "water for injection." This product meets all of the
requirements for USP purified water (as well as ACS) and the finished sterile product is
proven to be free of pyrogens. At the present time, this water represents the highest
purity water that is generally available. The NCCLS statement that injection-grade water is
not suitable for laboratory use is incorrect. The further statement that Type I water is
superior to injection-grade water for any use, in any situation, is similarly not correct.

! This document does not establish water purity on the basis of conductivity measurements
alone. However, such measurements are practical and readily available, and they provide
significant information about the water sampled. The working group does not agree with
the claims about USP grade or “water for injection.” It does believe, however, that the
wide range of issues mentioned in the comment have been considered in the revision of
C3-A3.

6. The belief that water degrades under normal ambient conditions is incorrect. Water is
extremely stable. It does not oxidize or undergo reduction even at high temperatures in
steam systems or in ice at super-low freezer temperatures. Water does degrade only at
extremely high temperatures, well over 1000 BC, as, for example, in contact with molten
metals; degradation then occurs with explosive violence.

What the document refers to as degradation is actually the equilibrium that is established
between any material and its environment. This is a basic requirement of the second law
of thermodynamics, and it applies throughout the universe.

NCCLS VOL. 17 NO. 18 34

October 1997 C3-A3

The conductivity of water increases primarily because of the atmospheric carbon dioxide
equilibrating with the water. Whatever water is used for subsequent to its production, to
prepare a saline solution, for example, yields a prepared reagent solution that, in turn,
tends to come into equilibrium with its environment. The recommendation that water be
used before it "degrades" (an NCCLS term) is meaningless.

! C3-A3 is revised to clarify that the term “degradation” refers to the degradation of water
quality.

7. Failure to differentiate between process specifications and product specifications is evident

throughout this document. One of the more pronounced examples is in the matter of
retrospective testing. This is always a type of process specification, i.e., the quality of the
material being tested at a particular time. In a successful test, it means that the process
was able, at that particular time, to function properly; it means nothing about what
happened earlier or later than the time tested. In the event that a retrospective test is not
successful, one has missed any chance to correct the problem.

! Footnote a in Table 2 on reagent water specifications was revised to help differentiate

between product specifications and purification process requirements not measured by the
end user. Although several testing recommendations in C3 are retrospective, they provide
the laboratory with useful data and can be helpful in the detection of trends and impending
problems.

The generally accepted references for water purity are as follows: American Chemical
Society (ACS); the United States Pharmacopeial Convention (USP); and good
manufacturing practices (GMP) (required by the FDA). In each case, multiple parameters
are measured and compared to established specifications.

! See the response to Comment 4.

Foreword

8. The Foreword proclaims that resistivity measurements are “simple.” In my experience,

resistivity measurements are anything but simple. The authors of this document appear to
concur that resistivity measurements are not simple, because the body of the document
contains two-and-one-half pages of instructions for this task.

! The working group revised the Foreword to clarify its originally intended meaning.

9. I recommend modifying the first sentence of the third paragraph as follows: “Resistivity
measurements are simple and provide an excellent index of ion content when the water is
first purified/manufactured.” I also suggest changing the “and” to “but” in the second
sentence.

! The Foreword of C3-A3 is revised.

10. The first two sentences of the fourth paragraph should be modified as follows: “Monitoring
of the other parameters noted is dependent upon many variables. Each laboratory has to
assess the needed frequency of monitoring any of these on laboratory manufactured,
purchased, or stored water, depending on the specific application.

! The working group did not agree to make these changes in the Foreword. It believes the
issues are addressed in the text of C3-A3.

NCCLS VOL. 17 NO. 18 35

October 1997 C3-A3

11. An additional sentence should be added to the Foreword that states, “Additionally, specific
applications may have more stringent requirements for the parameters noted or for
monitoring of additional parameters.”

! The working group does not believe that this level of detail is appropriate for the Foreword.
The issue is addressed in the body of the document.

Section 2.0

12. The second paragraph should specify the documents published by those organizations
referenced.

! C3-A3 includes the related document citations and the mailing addresses of the
corresponding organizations.

Section 3.0

13. I do not like the definition of HPLC. "An...the eluent or carrier is a liquid under pressure.”
Should read: "...the mobile phase is a liquid under pressure.”

! The definition of HPLC is revised in C3-A3.

14. I do not think that pH is "the electromotive force between....” Perhaps it should read,
"determined by measuring the electromotive force between..."

! The working group believes that the definition of pH is correct as stated.

15. Use the Random House American Dictionary for common words; then consult with NCCLS
document NRSCL8 for the other words before publication.

! There is an ongoing effort to maintain consistency of definitions in NCCLS documents.

C3-A3 was cross-checked with NRSCL8-P2, Nomenclature and Definitions for Use in
NRSCL and Other NCCLS Documents—Second Edition; Proposed Guideline.

Section 4.0

16. Please define the term "feedwater."

! The term “feedwater” was intended to be synonymous with “source water.” To reduce
confusion, the term “source water” is used throughout C3-A3.

17. Footnote a of Table 1 indicates that there is preliminary evidence of effectiveness. Is there
any new information available?

! The working group is not aware of any new published information to the contrary.

Section 5.0

18. This guideline should give some recommendations as to the space allocations required for a
water system.

! The working group cannot recommend space allocations for water systems. Space
requirements depend on the system selected and its intended use.

NCCLS VOL. 17 NO. 18 36

October 1997 C3-A3

Section 5.1

19. A sentence should be added to the end of the second paragraph that states, “Consult local
water testing authorities for advice.”

! The following sentence is added to Section 5.1, “In some instances it can be helpful to
consult local water-testing authorities for advice on the impurity content of the source
water.”

Section 5.2

20. The document makes a strong recommendation for the recirculation of water. Because I
feel that this is a critical and often overlooked component of a laboratory's water
purification system, I would like these two sentences to be in bold, or stand out in some
way. Similarly, the frequency of cleaning is buried in paragraph 3.

! As recommended, these points are emphasized by using italic type.

21. I would recommend rewording the first sentence of the third paragraph as follows:
“...sanitize prior to use and then at least as recommended by the manufacturer,
semiannually, or as determined by quality control criteria (whichever is most frequent.)”

! The sentence is modified in C3-A3.

22. In the third sentence of the third paragraph, change “will” to “may.”

! The working group did not agree with the change; it believes the statement to be correct
as written.

23. The second sentence should be modified to state, “However, additional parameters may
need to be monitored, especially if the user performs trace element analysis.”

! The sentence has been revised in C3-A3.

Section 5.3.1.2

24. Should not the second paragraph be listed first if they are the better options? Additionally,
list the alternative materials in order of priority regarding the quality of reagent water level.
Note special advantages and disadvantages for specific applications as examples.

! Section 5.3.1 of C3-A3 has been completely revised to provide the end user with
important considerations for the selection of system materials.

Section 5.3.1.3

25. To the end of the third sentence add, “...for specific applications.”

! This section has been revised to reflect the current status of construction materials.

Section 5.3.2

26. The note advising personnel not to attach tubing to a spigot is correct, because there can
be a problem with growth if the tubing is left in place. However, this is done to direct the
water effectively, and some suggestion for an alternate procedure should be offered if

NCCLS VOL. 17 NO. 18 37

October 1997 C3-A3

compliance is expected. Our internal specifications allow tubing to be attached for a day
at a time. The tubing must be removed and allowed to drain each day, and it must be
sanitized every 5 days.

! The working group does not recommend this practice.

27. Spigots should be of a diaphragm valve type, which are of a “sanitary” design.

! The working group cannot recommend the specific type of valve; there are practical
alternatives.

Section 6.0

28. To the first paragraph add the following sentence, “Storage of water will change some of
its characteristics, which may or may not affect usability of the water.” (This would result
in the deletion of Section 6.2.)

! The working group believes that the suggested change is inappropriate as part of the
specifications discussion.

Section 6.1

29. I suggest changing the section title to “Requirements for Types of Water” followed by a
sentence stating, “The requirements for each type of reagent water are specified in Table
2.”

! C3-A3 reflects these changes.

Section 6.1.1

30. One should not use a terminal filter on Type I water because the filter collects bacteria,
which, in turn, releases endotoxins in increasing amounts with time. Purity should be
imparted “up front” and maintained throughout the system.

! This section was revised as part of the review process.

31. For parallelism of form, Section 6.1.1 and 6.1.2 should be footnotes to Table 2 or the
footnotes should be consolidated and incorporated into the numbered paragraph style.

! The working group ensured that the text and table are consistent but did not change the
format as suggested.

Section 6.1.3

32. I recommend the following additions to this section, “If all the requirements for Type I
water are essential for a specific application, Type I water must be used immediately
.....Since certain characteristics, such as resistivity, will change quickly once the water is
produced, their influence on usability must be evaluated.”

! Section 6.1.3 is revised as suggested.

33. The unit of resistivity used throughout this document is the “megohm.” However, the
building engineers of my institution as well as some testing laboratories routinely use

NCCLS VOL. 17 NO. 18 38

October 1997 C3-A3

“micromos.” I think the end user would be well served if guidelines were given in both
units.

! The definition of “specific resistance or resistivity” is revised to accommodate this change.

34. What is the meaning of footnote b? If this is necessary as a criterion, it must be tested by
the manufacturer, whether the manufacturer is a laboratory, pharmacy, or some other
source. Otherwise, leave this out or describe its appropriateness.

! Footnote b is changed to footnote a and the wording is clarified.

Section 7.1

35. The resistivity of stored water on exposure to any amount of air will decrease because
some carbon dioxide dissolves in it. This does not mean, however, that the water is
impure or unfit for laboratory use. It is, therefore, inappropriate to state that, “Type I
water should be used immediately after processing.”

! Section 7.1. is revised.

36. I suggest adding a qualifier to the first sentence, “For optimal useability.”

! Section 7.1. is revised.

Section 7.1.1

37. While high resistivity generally means the absence of some impurities, low resistivity does
not always mean the presence of contamination. It is, therefore, inappropriate to make
resistivity the only measures for impurities in Type I water and to state, as does this
section, that, “Type I water cannot be stored because its resistivity will decrease, metals
and/or organic contaminants will be leached from the storage container, and bacterial
contamination will occur.”

! As stated in C3-A3, the working group considers resistivity to be only one of several
measurements used in the determination of water quality.

38. I suggest the following rewording of the first sentence, “Storage of Type I water will result
in decreases in resistivity (due to the absorption of carbon dioxide) increases in metals
and/or organic contaminants (due to leaching from the storage container), and bacterial
growth/contamination will occur.”

! The working group does not agree that Type I water can be stored.

39. Please consider adding the following paragraph to this section. “The specific effects of
this degradation on intended use of water originally produced and tested as meeting Type I
criteria, must be evaluated by the laboratory. If Type I water produced by the laboratory is
not used immediately, criteria for acceptability and monitoring must be established by the
laboratory as with any other reagent.”

! Section 7.1 is revised.

NCCLS VOL. 17 NO. 18 39

October 1997 C3-A3

Section 7.1.2

40. “The specifications for Type I water cannot be met by bottled water.” Even water purified
on site will have resistivity changes as it comes to equilibrium with its environment and
does not necessarily mean that it has become contaminated with “bad” contaminants.

! Section 7.1 is revised.

41. To the end of the paragraph, add the following sentence, “However, with laboratory-
produced Type I water, if a manufacturer can produce appropriate information regarding
the characteristics of the water produced/stored, the laboratory can determine
acceptability for the intended use.”

! Section 7 is revised.

Section 7.3

42. The last sentence in Section 7.3 is awkward. Replacing “...as well as...” with “...and...”
would be an improvement.

! The former Section 7.3 is now incorporated into Section 7.2. This change is incorporated
in the section revision.

Table 3

43. We agree that "special reagent water" used for tissue/cell culture should be pyrogen-free.
In actual practice, is this mandatory and is there a level that might have been documented
as insufficient to interfere with test results? In that event, what would that level be and
what would its scope be for application in interpreting LAL results?

! The working group added a test for endotoxins in Section 9.6.

44. I suggest that the title of the table be modified as follows, “Typical/Suggested Uses of
Reagent Water Produced as:” and move the “Special” column after “Type III.”

! Table 3 is deleted from the guideline.

Section 7.4.1

45. The description of the minimal organic content for HPLC-grade water is more confusing
than helpful. We are not told what “minimal” means or how to obtain such additionally
purified water. Putting water through ion-X or another resin should only accomplish what
should have been done to obtain Type I water. The HPLC method for determining the level
of organic contamination is poor but important enough to strengthen. What is the
monitoring wavelength? What type of gradient is to be used? Has anyone gotten advice
from the commercial people (e.g., Waters)?

! HPLC grade water is commercially available. Discussion of methods for determining

organic contaminants by HPLC is beyond the scope of this guideline. Refer to Bowers in
the bibliography.

Section 8.2

46. I recommend changing “has not been” to “is not typically.”

NCCLS VOL. 17 NO. 18 40

October 1997 C3-A3

! The essence of this change was incorporated into C3-A3.

Section 8.3

47. This section should state that the water used in laboratories must either: (1) meet the well-
established criteria required by organizations such as United States Pharmacopeial
Convention or the American Chemical Society; or (2) be stored in nonleaching containers
with a certificate ensuring that the water is pure pursuant to the same criteria, with an
expiration date on the label. In this way the laboratory/consumer would still have the
option of purchasing bottled reagent water.

! The working group believes Types I, II, and III waters, as defined, are appropriate for
clinical laboratory uses.

Section 8.3.3

48. This section describes changes/properties that apply to both on-site generated and bottled
water.

! Section 8.3 was revised to address issues related to purchased water.

Section 9.1

49. I have used the bacteriologic sample method for quantitating the degree of bacterial
contamination of water for many years. This method, which was developed for both urine
and water cultures, has its limitations, but it is most useful in settings that do not routinely
perform plate counts.

I recommend that your subcommittee consider the inclusion of the quantitative

bacteriologic loop procedure for estimating bacterial counts in reagent water. In large
clinical or reference laboratories, there is usually simple bacteriologic media available.
Since many clinical laboratories use quantitative loop procedures for urine cultures, it
would be appropriate for this option to be made available. I recommend that you obtain
the consultation of a clinical microbiologist when considering this proposal.

! The use of the loop is inappropriate; sample size is inadequate.

50. During our recent CAP inspection, one of the inspectors raised a question about one of the
tests suggested. At this time, we also looked at the microbiological testing that we had
been doing for years for our colleagues in all clinical laboratories (obviously for free!). We
have used colony counts by the calibrated loop-spreading method for many years. The
NCCLS guideline goes into an extensive description of pour-plate counts (over 6 pages),
then says in Section 9.1.4.3, “Limitations of Methods,” that the recommended methods
are not inclusive, etc. The calibrated loop or spreading methods are not mentioned at all.
Our free job will be impossible to do if we adopt the recommended pour-plate method,
start filtering, or such. We are thinking about recommending the bacteriologic sampler
method (Section 9.1.7.3), which each laboratory could do on their own.

Aside from these problems, there seems to be a general problem with the guideline; the
recommended methodologies probably were developed by industrial quality control and
public health people and not by clinical microbiologists. If the calibrated loop-spreading
method is good enough for urine cultures, it should be satisfactory for plate counts of
reagent water. We know that there will be some error, but we do not believe that we

NCCLS VOL. 17 NO. 18 41

October 1997 C3-A3

need the accuracy of the regulators to look at different reagent waters—Type I with max.
10 CFU/mL and Type II with max. 1000 CFU/mL.

! The recommended protocol for microbial content was adopted from “Standard Methods for
the Examination of Water and Waste Water” published by the American Public Health
Association and the American Water Works Association. The working group believes that
the technology exists to meet the specifications for reagent water described in this
guideline.

Section 9.1.3.2

51. In the second paragraph, we question the representativeness of a sample collected after 1
minute of allowing the water to flow, because we do not believe that this is the common
practice of laboratories.

! Upon consideration of the comment, the working group believes its recommendations to be
appropriate.

52. In the third paragraph, we believe that 50 to 100 mL would provide better quantitation.

! The working group believes that the protocol is consistent with the recommendations of
the American Public Health Association and the American Water Works Association;
therefore, no change was made.

Section 9.1.5

53. Our laboratory has recently added a continuous recirculating ultraviolet light system to our
deionized water system. In this system, deionized water flows continuously through all
the lines and up to the base of each faucet and then through a 3'-long cylinder that has a
germicidal light running the length of it. All the water passes through a 0.2-micron filter.
All the polyvinylchloride plumbing was replaced with new tubing, and all faucets were
replaced with new plastic stems and mechanisms.

The installation of this system was precipitated earlier this year by the excessive colony
forming units per milliliter (CFU/mL) of bacteria we started picking up after applying NCCLS
guidelines for testing the microbial content of our presumed Type I water. Previously, we
had incubated our petri dishes for only 24 hours at 37 BC. NCCLS guidelines call for this,
plus another 24 hours at ambient temperature (23 E ±3 BC). Since applying this
incubation regimen, we have consistently found several hundred CFU/mL at each faucet
using NCCLS guidelines for collection. In all cases, this testing has been done immediately
after the system has sanitized the lines with peroxyacetic acid for 1 hour. After incubation
at 37 BC, there is no growth. It all appears after sitting another 24 hours at ambient
temperature.

The system manufacturer has also cultured our irradiated water before and after sanitizing.
Their normal protocol is to incubate for 24 hours at 37 BC. Culturing one of our faucets
before sanitizing, they found 40 CFU/mL. If they swabbed the outside of the faucet with
an alcohol swab before sampling, counts were reduced to 10 CFU/mL. If they let both of
these plates stay for an additional 24 hours at 37 BC, the results were too numerous to
count. After sanitizing, the sample showed no growth after 24 hours at 37 BC. They did
not subject this latter sample to longer incubation.

Your manual states that Type I water is a fleeting entity. From our observations, it
appears impossible to attain, at least from a microbiological standpoint using your

NCCLS VOL. 17 NO. 18 42

October 1997 C3-A3

standards. Even though we are apparently not using Type I water, none of us can ever
recall having any technical problems that could be traced to water quality any place in the
laboratory over the last 5 years. The only thing that changed is the standard being applied
to the water, i.e., incubation for 48 hours instead of 24 hours. The majority of our
reagents contain bacteriostatic chemicals or are maintained at refrigeration temperatures
during their lifetime.

Also, we have no procedures that even come close to using our water under the 48-hour
incubation conditions specified in your manual and I suspect that this is true for most
laboratories in the country. Could we perhaps be applying a standard for water quality
that is unrealistic compared to how water is used in a typical laboratory?

Currently, we classify all of our procedures as requiring Type II or Type III water. It is our
current feeling that the water in the plumbing lines of our DI system defining the plumbing
lines as a primary container. Traditionally, laboratories have thought of their water quality
as being Type I. Now, it appears another tradition will bite the dust.

! The recommended protocol for microbial content was adopted from “Standard Methods for
the Examination of Water and Waste Water” published by the American Public Health
Association and the American Water Works Association. The working group believes that
the technology exists to meet the specifications for reagent water described in this
guideline.

54. Thank you for the opportunity to comment on what is generally an excellent guideline. Our
principal concern is with the temperature employed for microbiology testing. Most
“standard” laboratory tests are for pathogenic bacteria, while contamination with
nonpathogenic bacteria is more likely in most high-purity water systems. Standard
temperatures used in microbiology are actually inhibitory to many nonpathogenic
organisms. A better range is 30 to 35 BC. Some water laboratories actually use 28 to 30
BC.

! See the response to Comment 51.

Section 9.1.5.1

55. The abbreviation NIST should read, “National Institute of Standards and Technology.”

! This correction was made.

Section 9.1.6.2

56. Agar should be HPC (heterotrophic plate count), R2A, “standard methods” TSA, or other
media designed for “stressed” organisms in purified water systems.

! HPC was added to Section 9.1.6.2.

Section 9.1.6.3

57. In the third paragraph, a pour plate is not optimal for water testing because the sample
size is too low.

! The working group believes that there is an alternate technique available for sampling
larger volumes. (Refer to the response to Comment 51.)

NCCLS VOL. 17 NO. 18 43

October 1997 C3-A3

Section 9.2

58. If pyrogen-free water is to be manufactured in the laboratory, we recommend testing

pyrogens on each day of use. (If a frequency is not specified, then it will not be done
frequently enough.)

! A new Section 9.6, including a discussion on endotoxin determinations, was added to C3-
A3.

Section 9.3

59. The guideline should recommend testing pH of water routinely (e.g., weekly). This
document includes pH testing without recommending a minimum frequency.

! The working group believes that it is the responsibility of each laboratory’s quality
assurance guidelines to indicate a testing frequency. If desired, in-line instruments are
commercially available.

Section 9.4.1.1

60. This may become one of the more popular NCCLS guidelines with worldwide applicability.
With the increasing use of sensitive, modern laboratory instrumentation and the improved
immunochemical in vitro diagnostics, the error caused by low-quality reagent water (and
possibly wash water) on laboratory results could be greatly amplified in developing
countries.
Instructions for the use of, and schematics for building, a simple, battery-operated, and
inexpensive resistivity meter would greatly add to the value of this guideline.

! This is beyond the scope of C3-A3.

61. I question the stipulation that a meter should be capable of measuring only higher than
specification.

! The working group believes that the recommendation is correct.

Section 9.4.2

62. Water samples for quality testing should be taken from the point of use. In-line meters
should be installed near the point of use.

! C3-A3 is revised to incorporate this point.

Section 9.4.3

63. We are told that Type I water should have a resistivity of at least 10 megohms.
Meanwhile, Section 9.4.3 recommends that the resistivity meter may be calibrated with a
0.01 mol/L KCl solution, which is expected to have a resistivity of 0.707 kohm-cm. I
cannot see how a calibration performed at 0.707 kohm-cm can be used to validate
measurements of Type I water with a resistance of ~1.5 orders of magnitude greater than
the calibration value.

! Section 9.3.3 (which was Section 9.4.3 in C3-A2) and Section 9.3.1.3 are revised.

NCCLS VOL. 17 NO. 18 44

October 1997 C3-A3

64. If there is no other way to independently validate water resistivity, then how may this
paragraph speak of the relative accuracy of resistivity meters? Perhaps the word
“precision” ought to be substituted here.

! Section 9.3.3 was revised to address this concern.

Section 9.5

65. Our procedure (a copy of which was provided to the NCCLS Executive Offices) for the
annual check of soluble silica embodies the ideas and changes we believe necessary to
make C3-A2 useful on a routine basis.

! The working group appreciates any improvements to the procedure and encourages
publication of the procedure for future reference. Section 9.4 of C3-A3 recommends the
use of an appropriate reference laboratory or commercially available kits and it provides the
existing procedure as an alternate.

Section 9.5.3

66. The descriptions of the formulations of two reagents used for the "Soluble Silica" assay
need attention. The description for preparation of the 1-amino-2-naphthol-4-sulfonic acid
(ANSA) solution calls for preparation of a 30% solution of sodium hydrogen sulfite. This is
significantly above the solubility point of this compound. I, and others in my corporation
who prepare this ANSA solution reagent, have found that the only way to create a
solution, instead of a suspension, is to slightly heat about 180 mL of water, and, to this,
add the sodium sulfite and ANSA itself, let dissolve while stirring, and then slowly add the
sodium hydrogen sulfite directly to this solution. This procedure results in a true, clear
solution. The procedure, as written in C3-A2, results in a cloudy suspension, which settles
out quickly.

The other problem is with the description of the oxalic acid solution. The molarity of this
10% solution is listed as 1.11 millimolar. Obviously it should be in the molar range.
Additionally, oxalic acid is the common name given to oxalic acid dihydrate. The dihydrate
should be specified in the text, for clarity.

! This protocol was revised to be consistent with the ASTM D859-88 procedure.

Section 9.6.1

67. Most clinical chemistry laboratories have some access to a decent spectrophotometer.
Why not give us the procedure and let the individual users decide if they can actually use it
or not?

! Determinations for organic contaminants require HPLC.

Bibliography

68. We could not find a single location in Canada for Mrs. Winstead's often cited article,
"Reagent Grade Water: How, When and Why? American Society of Medical Technologists,
Austin, Texas. Steck Co. 1967. " Repeated requests to U.S. libraries were unsuccessful
to date. Would you kindly provide us with the necessary information to get a copy of the
above-mentioned article? It would be most appreciated if you could.

NCCLS VOL. 17 NO. 18 45

October 1997 C3-A3

! Interested persons should contact the American Society for Clinical Laboratory Sciences
(formerly American Society for Medical Technology) for updated information (7910
Woodmont Avenue, Suite 1301, Bethesda, Maryland 20814).

Appendix C

69. We recommend that the document states that preventive maintenance procedures must
specify a schedule for the replacement of filters or other critical components.

! Appendix B of C3-A3 addresses this issue.

70. This document must be withdrawn or discontinued and appropriate corrective action taken.
The problem with a simple revision process is that it tends to indicate that the essence of
the document is correct and that some improvements have been made. In fact, the basic
document is incorrect and is fundamentally a promotion of Type I water systems.

! The working group believes that the essence of the document is correct and improved its
utility by addressing the comments in this summary. Other editorial changes were made to
clarify the intended use of Type I water and commercially available reagent water. This
guideline describes water of three specific levels of quality (Types I, II, and III) and the
methods of producing and testing such water. The classification and specifications are
designed to enable laboratory scientists and supporting industries to specify the quality of
water to be used in such procedures as reagent preparation, reconstitution of lyophilized
materials, sample dilution, etc. No one specific method is recommended for producing
purified water. A single method or combination of methods may be used satisfactorily,
provided that the end product meets the required specifications stated in this guideline.
The working group does not recommend withdrawal of the guideline.

NCCLS VOL. 17 NO. 18 46

October 1997 C3-A3

Related NCCLS Publications‡

DI1-A2 Glossary and Guidelines for Immunodiagnostic Procedures, Reagents, and Reference
Materials—Second Edition; Approved Guideline (1992). DI1-A2 addresses common
terminology and basic methodology for immunodiagnostic procedures.

EP7-P Interference Testing in Clinical Chemistry; Proposed Guideline (1986). EP7-P offers
background information and procedures for characterizing the effects of interfering
substances on test results.

GP2-A3 Clinical Laboratory Technical Procedure Manuals—Third Edition; Approved Guideline

(1996). GP2-A2 addresses the design, preparation, maintenance, and use of technical
procedure manuals in the clinical laboratory.

I2-A2 Temperature Calibration of Water Baths, Instruments, and Temperature

Sensors—Second Edition; Approved Standard (1990). I2-A2 offers information about
reference thermometers and proper usage. Background information and methodology
for the performance of practical temperature calibrations using standard reference
material (SRM) thermometers is also provided.

‡
Proposed- and tentative-level documents are being advanced through the NCCLS consensus process; therefore, readers
should refer to the most recent editions.

NCCLS VOL. 17 NO. 18 47

October 1997 C3-A3

NOTES

NCCLS VOL. 17 NO. 18 48

October 1997 C3-A3

NOTES

NCCLS VOL. 17 NO. 18 49

October 1997 C3-A3

NOTES

NCCLS VOL. 17 NO. 18 50

October 1997 C3-A3

NOTES

NCCLS VOL. 17 NO. 18 51

October 1997 C3-A3

NOTES

NCCLS VOL. 17 NO. 18 52

October 1997 C3-A3

NOTES

NCCLS VOL. 17 NO. 18 53