preparation and testing of reagent water in the clinical ... · • the acceptance of a document as...

C3-P4

Vol. 25 No. 13

Preparation and Testing of Reagent Water in the Clinical Laboratory; Proposed Guideline—Fourth Edition

PLEASE

This proposed document is published for wide and thorough review in the new, accelerated Clinical and Laboratory Standards Institute (CLSI) consensus-review process. The document will undergo concurrent consensus review, Board review, and delegate voting (i.e., candidate for advancement) for 90 days. Please send your comments on scope, approach, and technical and editorial content to CLSI.

Comment period ends

6 September 2005

The subcommittee responsible for this document will assess all comments received by the end of the comment period. Based on this assessment, a new version of the document will be issued. Readers are encouraged to send their comments to the Clinical and Laboratory Standards Institute Executive Offices, 940 West Valley Road, Suite 1400, Wayne, PA 19087-1898 USA; Fax: +610.688.0700, or to the following e-mail address: [email protected]

COMMENT

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. A guideline for global application developed through the Clinical and Laboratory Standards Institute consensus process.

Clinical and Laboratory Standards Institute Providing NCCLS standards and guidelines, ISO/TC 212 standards, and ISO/TC 76 standards The Clinical and Laboratory Standards Institute (CLSI, formerly 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 worldwide for the application of its unique consensus process in the development of standards and guidelines for patient testing and related healthcare issues. Our process 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, we provide an open and unbiased forum to address critical issues affecting the quality of patient testing and health care.

PUBLICATIONS

A 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 consensus review and is released by the Board of Directors.

CONSENSUS PROCESS

The CLSI 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 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 consensus level.

Proposed A 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.

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.

Our 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 CLSI’s established consensus procedures. Provisions in CLSI 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 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 document. Address comments to the Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, PA 19087, USA.

VOLUNTEER PARTICIPATION

Healthcare professionals in all specialties are urged to volunteer for participation in CLSI projects. Please contact us at [email protected] or +610.688.0100 for additional information on committee participation.

C3-P4 ISBN 1-56238-570-4

Volume 25 Number 13 ISSN 0273-3099

Preparation and Testing of Reagent Water in the Clinical Laboratory; Proposed Guideline—Fourth Edition W. Gregory Miller, PhD, DABCC, FACB Erich L. Gibbs, PhD Dennis W. Jay, PhD, DABCC, FACB Kenneth W. Pratt, PhD Bruno Rossi, MS Christine M. Vojt, MT(ASCP), MS Paul Whitehead, PhD, CChem, FRSC Abstract CLSI document C3-P4, Preparation and Testing of Reagent Water in the Clinical Laboratory; Proposed Guideline—Fourth Edition provides information on water purity requirements for clinical laboratory testing, validation of specifications, technology available for water purification, and test procedures to monitor and trend water purity parameters. Clinical and Laboratory Standards Institute (CLSI). Preparation and Testing of Reagent Water in the Clinical Laboratory; Proposed Guideline—Fourth Edition. CLSI document C3-P4 (ISBN 1-56238-570-4). Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 2005.

The Clinical and Laboratory Standards Institute 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 CLSI/NCCLS documents. Current editions are listed in the CLSI 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 catalog, contact us at: Telephone: 610.688.0100; Fax: 610.688.0700; E-Mail: [email protected]; Website: www.clsi.org

Number 13 C3-P4

ii

This publication is protected by copyright. No part of it may be reproduced, stored in a retrieval system, transmitted, or made available in any form or by any means (electronic, mechanical, photocopying, recording, or otherwise) without prior written permission from Clinical and Laboratory Standards Institute, except as stated below. Clinical and Laboratory Standards Institute 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 CLSI publication C3-P4—Preparation and Testing of Reagent Water in the Clinical Laboratory; Proposed Guideline—Fourth Edition (ISBN 1-56238-570-4). Copies of the current edition may be obtained from Clinical and Laboratory Standards Institute, 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 Clinical and Laboratory Standards Institute by written request. To request such permission, address inquiries to the Executive Vice President, Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898, USA. Copyright ©2005. Clinical and Laboratory Standards Institute. Suggested Citation (Clinical and Laboratory Standards Institute. Preparation and Testing of Reagent Water in the Clinical Laboratory; Proposed Guideline—Fourth Edition. CLSI document C3-P4 [ISBN 1-56238-570-4]. Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 2005.) Proposed Standard—First Edition Tentative Guideline—Second Edition January 1976 December 1988 Tentative Standard—First Edition Approved Guideline—Second Edition January 1978 August 1991 Approved Standard—First Edition Approved Guideline—Third Edition February 1980 October 1997 Proposed Guideline—Second Edition Proposed Guideline—Fourth Edition June 1985 June 2005 ISBN 1-56238-570-4 ISSN 0273-3099

Volume 25 C3-P4

iii

Committee Membership Area Committee on Clinical Chemistry and Toxicology W. Gregory Miller, PhD Chairholder Virginia Commonwealth University Richmond, Virginia David A. Armbruster, PhD, DABCC, FACB Vice-Chairholder Abbott Laboratories Abbott Park, Illinois John Rex Astles, PhD, FACB Centers for Disease Control and Prevention Atlanta, Georgia David M. Bunk, PhD National Institute of Standards and Technology Gaithersburg, Maryland Neil Greenberg, PhD Ortho-Clinical Diagnostics, Inc. Rochester, New York Christopher M. Lehman, MD Univ. of Utah Health Sciences Center Salt Lake City, Utah Richard R. Miller, Jr. Dade Behring Inc. Newark, Delaware Michael E. Ottlinger, PhD, DABT U.S. Environmental Protection Agency Cincinnati, Ohio

Linda Thienpont University of Ghent Gent, Belgium Thomas L. Williams, MD Methodist Hospital Omaha, Nebraska Advisors Larry D. Bowers, PhD, DABCC U.S. Anti-Doping Agency Colorado Springs, Colorado Robert W. Burnett, PhD Hartford Hospital Farmington, Connecticut Mary F. Burritt, PhD Mayo Clinic Rochester, Minnesota Paul D’Orazio, PhD Instrumentation Laboratory Lexington, Massachusetts Carl C. Garber, PhD, FACB Quest Diagnostics, Incorporated Teterboro, New Jersey Uttam Garg, PhD, DABCC Children’s Mercy Hospital Kansas City, Missouri

David E. Goldstein, MD Univ. of Missouri School of Medicine Columbia, Missouri Harvey W. Kaufman, MD Quest Diagnostics, Incorporated Lyndhurst, New Jersey Gary L. Myers, PhD Centers for Disease Control and Prevention Atlanta, Georgia David Sacks, MD Brigham and Women’s Hospital and Harvard Medical School Boston, Massachusetts Bette Seamonds, PhD Mercy Health Laboratory Swarthmore, Pennsylvania Dietmar Stöckl, PhD University of Ghent Gent, Belgium Hubert Vesper, PhD Centers for Disease Control and Prevention Atlanta, Georgia Jack Zakowski, PhD, FACB Beckman Coulter, Inc. Brea, California

Working Group on Reagent Water W. Gregory Miller, PhD, Chairholder Virginia Commonwealth University Richmond, Virginia Erich L. Gibbs, PhD High-Q, Inc. Wilmette, Illinois Dennis W. Jay, PhD, DABCC, FACB St. Jude Children’s Research Hospital Memphis, Tennessee

Kenneth W. Pratt, PhD National Institute of Standards and Technology Gaithersburg, Maryland Bruno Rossi, MS Millipore SAS Guyancourt, France Christine M. Vojt, MT(ASCP), MS Ortho-Clinical Diagnostics, Inc. Rochester, New York Paul Whitehead, PhD, CChem, FRSC Elga LabWater, Lane End, Bucks, United Kingdom

Advisors Kelli Buckingham-Meyer Montana State University Bozeman, Montana Darla M. Goeres, MS Montana State University Bozeman, Montana Marilyn J. Gould, PhD Falmouth, Massachusetts Zenaida Maicas, PharmD Cape Neddick, Maine

Number 13 C3-P4

iv

Advisors (Continued) Stephane Mabic Millipore SAS Guyancourt, France Alan Mortimer, CChem, FRSC Elga LabWater, Lane End, Bucks, United Kingdom Keith W. Richardson Associates of Cape Cod, Inc. Woods Hole, Massachusetts

Staff Clinical and Laboratory Standards Institute Wayne, Pennsylvania Tracy A. Dooley, BS, MLT(ASCP) Staff Liaison

Donna M. Wilhelm Editor Melissa A. Lewis Assistant Editor

Volume 25 C3-P4

v

Contents

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

Committee Membership........................................................................................................................ iii

Foreword.............................................................................................................................................. vii

1 Scope..........................................................................................................................................1

2 Introduction................................................................................................................................1

3 Definitions .................................................................................................................................2

4 Specifications.............................................................................................................................5 4.1 Organization of Water Purity Specifications ................................................................6 4.2 Clinical Laboratory Reagent Water (CLRW) ...............................................................7 4.3 Special Reagent Water (SRW)......................................................................................8 4.4 Instrument Feed Water..................................................................................................8 4.5 Water Supplied by a Method Manufacturer for Use as a Diluent or Reagent ..............8 4.6 Prepackaged Bottled Water...........................................................................................8 4.7 Autoclave and Wash Water Applications .....................................................................9

5 Validation and Trend Monitoring ..............................................................................................9 5.1 Validation of Purified Water as Fit for Its Intended Purpose in Laboratory

Procedures.....................................................................................................................9 5.2 Trend Monitoring of Water Purity Specifications ......................................................10 5.3 Water Purification System Validation ........................................................................11

6 Design Considerations .............................................................................................................12 6.1 Filters ..........................................................................................................................13 6.2 Reverse Osmosis (RO) Membranes............................................................................14 6.3 Contactor Membranes.................................................................................................15 6.4 Ion-Exchange Resins ..................................................................................................15 6.5 Activated Carbon ........................................................................................................17 6.6 Distillation ..................................................................................................................18 6.7 Ultraviolet Light .........................................................................................................20 6.8 Storage and Distribution .............................................................................................21

7 Testing .....................................................................................................................................23 7.1 Resistivity ...................................................................................................................23 7.2 Microbial Content by Colony Count...........................................................................28 7.3 Microbial Content by Epifluorescence Microscopy ...................................................29 7.4 Endotoxins ..................................................................................................................32 7.5 Determination of Oxidizable Organic Substances, Expressed as Total Organic Carbon (TOC) .............................................................................................................34

References.............................................................................................................................................40

Additional References...........................................................................................................................42

Appendix A. Permeability of Plastic Piping to Atmospheric CO2.......................................................43

Number 13 C3-P4

vi

Contents (Continued)

Appendix B. Resistivity Measurement in a Sparged Water Sample....................................................45

Appendix C. Methods for Correction or Compensation of Resistivity Measurements .......................47

Summary of Comments and Subcommittee Responses........................................................................49

The Quality System Approach..............................................................................................................50

Related CLSI/NCCLS Publications ......................................................................................................51

Volume 25 C3-P4

vii

Foreword This edition of the guideline includes updated information regarding the preparation and testing of reagent water in clinical laboratories. Specifications are based on measuring parameters that serve as indicators for the total ionic, organic, and microbial contamination. Emphasis is placed on the value of trending these parameters as an effective way to control the quality and consistency of purified water, as well as the importance of validating that a grade of reagent water is fit for its intended purpose in clinical laboratory testing. A new section provides guidelines for water purification system validation, ongoing maintenance, and revalidation on a recurring schedule. The Type I, II, III designations for grades of purified water, used in the previous edition, have been replaced with purity grades that provide more meaningful specifications for clinical laboratory testing. Clinical laboratory reagent water (CLRW) can be used in place of Type I and Type II water for most applications. Autoclave and wash water will generally be satisfactory replacement for Type III water. The definitions of the new grades of water include a number of parameters that were not used in previous editions and do not use some of the parameters that were used in previous editions. Resistivity measurement has been retained to monitor inorganic impurities. The previous edition recommended that water purification systems include a stage to reduce organic contamination, but required no control. This edition recognizes that organic contamination can be difficult to remove from feed water, can be introduced by components of water purification systems or biofilms, and must be controlled. Therefore, a total organic carbon (TOC) parameter has been added. Note that on-line or in-house measurements of TOC are not required. It is acceptable to send CLRW samples to a referral laboratory for TOC measurement. (See Section 7.5 for additional information on contamination issues when TOC is at low levels.) Plate counting of colonies is a widely used method for monitoring the level of microorganisms in purified laboratory water, and this edition continues to specify this approach. However, epifluorescence and endotoxin testing have been added as optional tests, because they provide additional information and results can be determined quickly. Specifications and methods for measuring pH, SiO2, and sterility have not been carried forward from the previous edition. Resistivity is more sensitive than pH to H+ and OH- contamination. Resistivity is not a sensitive indicator of weakly ionized contaminants, which may elute as concentrated pulses from ion- exchange beds when they approach depletion. However, the release of weakly ionized contaminants (silica being but one example) can be avoided by appropriate design and regular maintenance of ion-exchange components. A parameter for sterility has not been included in this edition of the guideline, because most laboratory applications do not require sterile water. When necessary, water can be sterilized; however, the method of sterilization must not degrade the purity of the water. Key Words Clinical laboratory water, high-purity water, purified water, reagent water, water purification Invitation for Participation in the Consensus Process An important aspect of the development of this and all CLSI documents should be emphasized, and that is the consensus process. Within the context and operation of CLSI, the term “consensus” means more than agreement. In the context of document development, “consensus” is a process by which CLSI, its members, and interested parties (1) have the opportunity to review and to comment on any CLSI publication; and (2) are assured that their comments will be given serious, competent consideration. Any

Number 13 C3-P4

viii

CLSI document will evolve as will technology affecting laboratory or healthcare procedures, methods, and protocols; and therefore, is expected to undergo cycles of evaluation and modification. The Area Committee on Clinical Chemistry and Toxicology has attempted to engage the broadest possible worldwide representation in committee deliberations. Consequently, it is reasonable to expect that issues remain unresolved at the time of publication at the proposed level. The review and comment process is the mechanism for resolving such issues. The CLSI voluntary consensus process is dependent upon the expertise of worldwide reviewers whose comments add value to the effort. At the end of a 90-day comment period, each subcommittee is obligated to review all comments and to respond in writing to all which are substantive. Where appropriate, modifications will be made to the document, and all comments along with the subcommittee’s responses will be included as an appendix to the document when it is published at the next consensus level.

Volume 25 C3-P4

©Clinical and Laboratory Standards Institute. All rights reserved. 1

Preparation and Testing of Reagent Water in the Clinical Laboratory; Proposed Guideline—Fourth Edition

1 Scope Purified water requirements are specified for clinical laboratory testing procedures. The following purity grades are described: • clinical laboratory reagent water (CLRW); • special reagent water (SRW); • instrument feed water; • autoclave and wash water; and • commercially available bottled water. Procedures are provided for measuring and trending parameters to control ionic, organic, and microbial contamination in purified laboratory water. Recommendations are provided to control particulate and colloidal contamination. The guideline requires the laboratory to validate a selected grade of water as fit for its intended purpose in laboratory tests. Suggested approaches for validation of water purification systems are included. It is beyond the scope of this guideline to recommend specific types of water purification systems. Instead, the guideline provides information about discrete purification technologies and monitoring strategies that can be applied in various combinations to achieve and maintain the required water purity. 2 Introduction A key element of success in the clinical laboratory is the constancy of test result quality. The physician or caregiver counts on results that represent only patient analyte measurement, not the measurement of microbial or chemical contaminants that may be extraneously introduced during a laboratory procedure. Purified water constitutes the major component of many reagents, buffers, and diluents used in clinical laboratory testing. It can also become an indirect component of tests when it is used for washing and sanitizing instruments and laboratory ware, generating autoclave steam, etc. Purified water is a potential cause of laboratory error. This guideline recommends measuring certain parameters of purified water used in clinical laboratory applications as a means of quality control for purification systems. The parameters are: resistivity, an indicator of ionic contamination; viable plate counts, an indicator of microorganism contamination; and total organic carbon, an indicator of organic contamination. Epifluorescence and endotoxin testing are included as optional approaches for measuring contamination from microbial sources. Particulate contamination is controlled by including appropriate filtration, or distillation, in the purification system. The guideline is not intended to assure the adequacy of purified water for a given laboratory application; rather, water of a specified purity must be validated as fit for use in a particular laboratory application. Any parameters used to specify a grade of purified water, or to monitor the performance of a purification system, should be measured frequently enough to detect potential changes in the system, and the measurements should be trended to detect drift and anticipate maintenance. Other organizations have published guidelines and specifications for purified water intended for various applications. Water conforming to the guidelines and specifications of these organizations may or may not be equivalent to the grades of purified water described in this CLSI guideline. Any type of purified water should be validated as fit for purpose before being used in clinical laboratory testing.

Number 13 C3-P4


3 Definitions absorption – a process by which a substance is taken up chemically or physically in bulk by a material (absorbent) and held in pores or interstices in the interior; NOTE: See also adsorption, sorption. accuracy – closeness of agreement between a test result and the accepted reference value (ISO 3534-1)1; NOTE: accuracy of a measurement is defined as the closeness of the agreement between the result of a measurement and a true value of the measurand (VIM93).2 activated carbon – porous carbon material used for removal of impurities; NOTE: See Section 6.5 for details. adsorption – adherence of molecules, atoms, and ionized species of gas or liquid to the surface of another substance (solid or liquid) as the result of a variety of weak attractions that involve ionic, Van der Waals, and surface-active (hydrophobic/hydrophilic) forces; NOTE: See also absorption, sorption. anion exchange resin – an ion-exchange resin with immobilized positively charged exchange sites, which can bind negatively charged ionized species. azeotrope – a blend of two or more components with equilibrium vapor phase and liquid phase compositions that are the same at a given temperature and pressure. bactericide – a chemical or physical agent that kills bacteria. biocide – a chemical or physical agent that kills microorganisms (as used in this document). biofilm – microorganisms, enclosed in a glycoprotein/polysaccharide matrix, that adhere to each other and/or to surfaces and may form macroscopic layers.3 CA membrane – a reverse osmosis membrane constructed of cellulose diacetate/triacetate. calibration – the set of operations that establish, under specified conditions, the relationship between values of quantities indicated by a measuring instrument or measuring system, or values represented by a material measure or a reference material, and the corresponding values realized by standards (VIM93).2 catalyst – a substance that increases the kinetics of a chemical reaction without being consumed in the reaction. cation exchange resin – an ion-exchange resin with immobilized negatively charged exchange sites, which can bind positively charged ionized species. colloid – small, solid particles that will not settle out of a solution. concentrate – the liquid containing dissolved and suspended matter that concentrates on one side of a membrane. condenser – the stage of a distillation system that removes sufficient heat from a vaporized liquid to cause the vapor to change to a liquid phase. conductivity – conductivity is the reciprocal of resistivity; NOTE: For water purification systems, conductivity is usually reported in microsiemens per centimeter (µS/cm). contactor membrane – a hydrophobic membrane used in removing dissolved gases from water.

Volume 25 C3-P4


copolymer – a polymer formed from two or more different monomers. deadleg/dead volume – a region or volume of stagnation in an apparatus or distribution system. distillation – a purification process that takes advantage of changing the phase of a substance from liquid to vapor and back to liquid, usually at the boiling temperature of the substance, in order to separate it from other substances with higher or lower boiling points. electrodeionization (EDI) – technology combining ion-exchange resins and ion-selective membranes with direct current to remove impurity ionized species from water and maintain the resin in regenerated condition. endotoxin – a thermostable lipopolysaccharide component from the cell wall of viable or nonviable gram-negative microorganisms. epifluorescence – method of fluorescence microscopy in which the excitation light is transmitted through the objective lens onto the specimen, and the fluorescence light is transmitted back through the objective lens to the eyepiece; NOTE: Fluorescence is the immediate emission of electromagnetic radiation, typically visible light, from molecules following absorption of light with a shorter wavelength. feed water – the water that is introduced into a purification process. filtration – a purification process in which the passage of fluid through a porous material results in the removal of impurities based on the physical interaction of the impurities with that porous material. fines – see particulates. gram-negative – refers to bacteria that do not retain the primary violet stain in the decolorization step in the procedure originally described by Gram. gram-positive – refers to bacteria that absorb and retain the primary violet stain in the decolorization step in the procedure originally described by Gram. ion exchange – a reversible chemical reaction between a solid containing immobilized ionic sites (ion exchanger) and a fluid (often water) by means of which ionized species may be exchanged from one substance to another. measurand – particular quantity subject to measurement (VIM93).2 microorganism – any organism that is too small to be viewed by the unaided eye, such as bacteria, viruses, molds, yeast, protozoa, and some fungi and algae. nonpurgeable organic carbon (NPOC) – the concentration of organic carbon remaining after sparging a sample to remove inorganic carbon. off-line – in water monitoring systems, referring to measurement devices that are not directly coupled to the water stream. on-line – in water monitoring systems, referring to measurement devices directly coupled to the water stream. particulates – discrete quantities of solid matter dispersed in water.

Number 13 C3-P4


permeate – substances passing through a semipermeable membrane. planktonic – a term used to describe aquatic microorganisms that float. plasticizer – a chemical component of plastics to make them softer and more flexible. polishing – the final treatment stage(s) of a water purification system. potable water – water that meets regulations as suitable for ingestion by humans. precision – closeness of agreement between independent test results obtained under stipulated conditions (ISO 3534-11); NOTE: Precision depends only on the distribution of random errors and does not relate to the true value or the specified value (ISO 3534-11). purgeable organic carbon (POC) – the concentration of carbon that escapes the sample in the gas phase during the process of sparging the sample to remove inorganic carbon prior to measuring the organic carbon. qualification – the act of establishing with documented evidence that the process, equipment, and/or materials are designed, installed, operated and perform according to the predetermined specifications. reservoir – in water purification systems, a container holding quantities of purified water. resistivity – the electrical resistance between opposite faces of a one-centimeter cube of a given material at a specified temperature; NOTE 1: Resistivity is the reciprocal of conductivity; NOTE 2: For water analysis, resistivity is usually reported in megohm-centimeters (MΩ•cm). reverse osmosis (RO) – a process in which water is forced under pressure through a semipermeable membrane, leaving behind dissolved organic, dissolved ionic, and suspended impurities. sanitization – chemical and/or physical processes used to kill microorganisms and reduce contamination from microorganisms. softening – a water treatment process whereby cations are exchanged for sodium using cation-exchange resins in the sodium form. sorption – either or both of the processes of absorption and adsorption. sparging – injection of gas below the water surface to remove other dissolved gases and volatile organic compounds. stagnation – state of a liquid without current or circulation. sterilization – validated process used to render a product free from microorganisms (ISO 151904). total carbon (TC) – total concentration of carbon (organic and inorganic) in a sample. total inorganic carbon (TIC) – total concentration of carbon as carbonates, bicarbonates, or dissolved carbon dioxide. total organic carbon (TOC) – total concentration of carbon in the form of organic compounds.

Volume 25 C3-P4


validation – confirmation, through the provision of objective evidence, that requirements for a specific intended use or application have been fulfilled; NOTE 1: The term “validated” is used to designate the corresponding status (ISO 90005); NOTE 2: The use conditions for validation can be real or simulated (ISO 90005); NOTE 3: A term used by the FDA for a study used to determine whether a test system meets user needs (12 CFR Parts 808, 812, and 820)6; NOTE 4: WHO defines validation as “the action of proving that a procedure, process, system, equipment, or method used works as expected and achieves the intended result” (WHO-BS/95.1793). verification – confirmation, through the provision of objective evidence, that specified requirements have been fulfilled; NOTE 1: The term “verified” is used to designate the corresponding status (ISO 90005); NOTE 2: Confirmation can comprise activities such as: performing alternative calculations; comparing a new design specification with similar proven design specifications; undertaking tests and demonstrations; and reviewing the document prior to issue (ISO 90005); NOTE 3: ISO 8402 defines verification as “confirmation, through the provision of objective evidence, that specified requirements have been fulfilled”); NOTE 4: The FDA defines verification as “a study used to determine whether a test system meets specifications” (21 CFR Parts 808, 812, and 820).6 4 Specifications Specifications are provided for five principal grades of purified water intended for different needs in clinical laboratory testing. At some stage in the preparation of every grade of purified water, the water must meet or exceed regulations for potable drinking water comparable to those of the European Union, Japan, or the United States. All of the parameters associated with a water specification must be measured while water purification systems are operating routinely. Insofar as it is reasonable to do so, samples must be obtained for measurement, or on-line measurements made, after the last purification component and as close to the purification system output as possible. Where any purification or storage components exist after an on-line measurement, the user should validate that the water remains fit for purpose. When systems include a recirculating loop to distribute water to remote points of use, samples must be obtained for measurement, or on-line measurements made, at or after the last port of the loop to ensure that contamination is not introduced by the loop or as the result of backflow at one or more ports on the loop. Using water that meets the specified limits for all of the parameters will reduce the probability of contaminants existing at levels that could interfere with clinical laboratory tests. However, the purpose of water specifications based on the parameters used by this guideline is to monitor for continued control of water purification systems, not to ensure that the water they produce is necessarily fit for specific applications. Purified water must be validated separately as fit for a particular laboratory application (see Section 5.1). To confirm the stability of a purification system, any parameters used to specify a grade of purified water, or to monitor the performance of the system, should be measured frequently enough and the results trended to detect changes and anticipate maintenance (see Section 5.2). Each laboratory must determine how often to measure the parameters of its purified water and purification system and how often to revalidate the water as fit for purpose, based on a balance of risk and practicality. Measurements made at intervals have the risk that an out-of-specification condition could have existed, and adversely affected clinical testing during the interval between measurements. Risk will increase as the time between measurements, and between revalidations, increases with respect to the stability of the purification system. Setting an alert threshold for a measured parameter at a more stringent level than the validated water purity can reduce the risk from gradual drift; however, this strategy does not protect against abrupt changes. A risk assessment should be carried out to establish an appropriate monitoring program.

Number 13 C3-P4


4.1 Organization of Water Purity Specifications This subsection is provided only for the purpose of assisting with the location of information relating to each of the measured parameters within this guideline. It is not intended for use outside the context of the complete guideline. 4.1.1 General Concepts Applicable to All Grades of Water The product water meeting a set specification must be validated as fit for purpose for each laboratory procedure in which it is to be used (see Section 5.1). The system producing purified water must be validated to meet the user requirements specification (see Section 5.3). Regular monitoring and trending of appropriate measured parameters must be carried out and documented to verify that water purification technologies and systems are working effectively (see Sections 5.2, 6, and 7). Procedures must be established for system maintenance to keep the system in conformance with water purity specifications (see Sections 5.2 and 6). 4.1.2 Clinical Laboratory Reagent Water (CLRW) 4.1.2.1 Ionic Impurities • Resistivity ≥10 MΩ • cm referenced to 25 °C (see Sections 4.2.1 and 7.1). NOTE: Pretreatment to remove CO2 may be needed for purified water that contains dissolved CO2

(see Section 7.1.2.3). 4.1.2.2 Microbiological Impurities • Total heterotrophic plate count <10 CFU/mL (see Sections 4.2.2 and 7.2).

NOTE: Cell counts by epifluorescence microscopy are an acceptable alternative to plate counts, provided the laboratory has established criteria for interpreting the results (see Sections 4.2.2 and 7.3).

4.1.2.3 Organic Impurities • TOC <500 ng/g (ppb) (see Sections 4.2.3 and 7.5). 4.1.2.4 Particulate Content • Purification systems must include a stage that blocks the passage of particles ≥0.22 µm at, or near, the

output stage (see Sections 4.2.4 and 6.1). 4.1.3 Other Grades of Purified Water for Use in the Clinical Laboratory • Special reagent water (see Section 4.3). • Instrument feed water (see Section 4.4). • Water supplied by a method manufacturer for use as a diluent or reagent (see Section 4.5). • Prepackaged bottled water (see Section 4.6). • Autoclave and wash water applications (see Section 4.7).

Volume 25 C3-P4


4.2 Clinical Laboratory Reagent Water (CLRW) Water that meets the specification for clinical laboratory reagent water (CLRW) should be pure enough to satisfy the requirements of most routine clinical laboratory testing. The specified parameters and limits are described in this section, and the methods for measuring the parameters are described in Section 7. Some laboratory testing applications may require special reagent water (see Section 4.2). 4.2.1 Ionic Impurities Ionic impurities are monitored by resistivity measurement. The resistivity specification for CLRW is a minimum 10 MΩ • cm, referenced to 25 °C. The resistivity of water is a function of several factors, including the charge, concentration, and mobility of the dissolved ions. Resistivity is not an absolute measure of ionic impurity concentration, but is an effective means for monitoring and trending ionic impurity levels. As a point of reference, a solution of NaCl in pure water with a concentration of 0.34 µmol/L will have a resistivity of 10 MΩ • cm, referenced to 25 °C. At the same molar concentration, other inorganic, monovalent ions will have approximately the same resistivity. Inorganic polyvalent ions carry incrementally more current than monovalent ions, so fewer ions (i.e., lower concentrations) are required to produce a solution with resistivity of 10 MΩ • cm. Ions that move relatively slowly in water (i.e., low mobility ions) do not carry current as efficiently as Na+ or Cl-.Therefore, higher concentrations of low mobility ions will produce a solution with resistivity of 10 MΩ·cm, referenced to 25 °C. If a substance is weakly ionized, only a fraction of the molecules form ions, and the concentration of the substance required to produce a solution with resistivity of 10 MΩ • cm, referenced to 25 °C, will be higher. See Sections 6.4 and 7.1 for more details. 4.2.1.1 Distillation-Based Purification Systems Dissolved CO2 may need to be removed for the purpose of making a resistivity measurement when water is purified by distillation or other means that do not exclude CO2. See Section 7.1.2.3. 4.2.2 Microbiological Impurities The specification for microbial content of CLRW at the point where the water exits a purification system for use in the laboratory is a maximum of 10 CFU/mL, measured by the plate count method described in Section 7.2. Microbial content can also be measured by epifluorescence; however, there are no consensus specifications, so laboratories must establish appropriate specifications based on validation of the water as fit for purpose. See Section 7.3 for more information. 4.2.3 Organic Impurities Organic impurities are measured as total organic carbon (TOC). The specification for TOC is a maximum of 500 ng/g (ppb). The TOC specification is a new requirement in this edition of the guideline. The 500 ng/g limit is a practical limit for organic contamination of CLRW that is consistent with the limit used by the pharmaceutical industry for purified water.7 It is included, because organic contamination of purified water can be an important source of interference in test procedures, and, as is true for all parameters, is dependent on both feed water composition and operating characteristics of water purification systems. A TOC specification is used, because the measurement provides a practical way to monitor organic contamination levels and to identify changes in levels that may require corrective action in the water purification system. Because the mole fraction of carbon is not the same for all molecules, a TOC value

Number 13 C3-P4


cannot be equated with the total mass of organic material present in an unknown water sample. See Section 7.5 for further information on TOC. 4.2.4 Particulate and Colloid Content Particulate and colloid specifications are based on the process of filtration. A particle-rejecting stage (e.g., reverse osmosis, ultrafiltration, microfiltration, distillation) that will remove all particles greater than or equal to 0.22 µm must be included at or near the final output stage of the purification system. A 0.22-µm, or finer, filter will remove most microorganisms and particulates, including those released from purification components. Reverse osmosis and ultrafiltration will also remove colloids (see Sections 6.1.2 and 6.2). 4.3 Special Reagent Water (SRW) When applications require water of different purity than CLRW, clinical laboratories should specify a special reagent water. The parameters used to specify CLRW should be included in any SRW specification, but the limits may be different and additional parameters may be added if needed. Some examples of applications that may require SRW include: • Trace organic analysis, which may require a lower TOC or an ultraviolet (UV) spectrophotometric

absorbance specification; • DNA and RNA testing, which may require specification for levels of nucleic acids, proteases,

nuclease activity; • Trace metals analysis, which typically requires an acceptable blank response for each metal to be

measured; • Cell/tissue/organ culture and fluorescent antibody detection of microorganisms, which may require an

endotoxin specification; and • Low CO2 water to prepare standard buffers for pH calibration. 4.4 Instrument Feed Water Instrument feed water is intended for the internal rinsing, dilution, and water bath functions of automated instruments. Use of CLRW for this application must be confirmed with the manufacturer of a specific instrument. Water meeting the manufacturer’s specifications must be used. 4.5 Water Supplied by a Method Manufacturer for Use as a Diluent or Reagent Water that is provided as a diluent or reagent by the manufacturer of a particular analytic system is intended for use only as described in the product labeling for that system. Such water has been qualified by the manufacturer specifically for the uses stated in the product labeling, and is not an acceptable substitute for CLRW or SRW, unless it meets the appropriate specification. Such water is also not acceptable for other testing procedures, unless the laboratory validates it for those procedures. 4.6 Prepackaged Bottled Water Commercially available, prepackaged, bottled CLRW or SRW must meet the required specifications. The water should be packaged in a manner that protects it from environmental contamination or degradation during transportation and storage, and from the effects of the container itself. Containers must be cleaned

Volume 25 C3-P4


to remove surface contamination and made of materials that do not leach significant inorganic or organic contamination. Containers must also be sufficiently impermeable to prevent significant external contamination from entering. Note that some types of plastic containers are sufficiently permeable to air to allow growth of autotrophic microorganisms, if they have a source of energy such as light. 4.6.1 Proper Labeling Commercial, bottled CLRW or SRW must include a lot number and expiration date as well as the values for resistivity (assuming CO2 has been removed), microbial count, TOC, and information on particulate control, plus other applicable parameters, for the lot of water as it exists in the container. 4.6.2 Validation in the Clinical Laboratory The laboratory must validate that the bottled water is fit for its intended purpose in clinical laboratory testing. If a bottle is opened and reused over a period of time, the laboratory must validate the water remains fit for purpose throughout the entire period of use. Each new lot of bottled water must be validated for acceptable performance in the test procedures for which it will be used. Typical validation procedures are described in Section 5.1. Bottled water should be purchased in a package size appropriate to the rate at which it is used to avoid prolonged storage of an opened container. Note that microbial and other airborne contaminants can be introduced during the period of use. An expiration date for an opened container must be established by the laboratory based on storage conditions and validation of fitness for purpose. Water stored in an opened, prepackaged bottle should be used in a timely manner to prevent accumulation of microbial or other contaminants. Note that pharmaceutical grades of purified water (e.g., purified water, sterile water for injection, sterile water for irrigation) are not acceptable alternatives for CLRW or SRW. A pharmaceutical grade of purified water must be validated for the specific laboratory testing procedures for which it will be used. 4.7 Autoclave and Wash Water Applications Autoclave and wash water is intended for use as feed water for autoclaves and for automatic laboratory dishwashers with heat drying cycles. This type of water is purified to low levels of inorganic, organic, and particulate impurities that otherwise could contaminate solutions and media in an autoclave or remain on washed laboratory ware. Various purification technologies can be used to produce this type of water. A consensus specification for this water is not available. 5 Validation and Trend Monitoring 5.1 Validation of Purified Water as Fit for Its Intended Purpose in Laboratory Procedures The laboratory must establish the chemical, microbial, and particulate quality requirements of purified water to be used for each of their applications and define one or more grades of water that will meet those requirements. Once water purity specifications have been defined, the purified water must be validated as fit for its intended purpose in each laboratory procedure. Validation is confirmation, through the provision of objective evidence, that requirements for a specific intended use or application have been fulfilled. The choice of a procedure for validation that a grade of purified water is fit for its intended purpose must consider the technical details and potential sources of interference in a laboratory procedure. Approaches are listed below that either individually or in combination may be useful for validation.

Number 13 C3-P4


• The candidate water can be used as a “blank” sample in the procedure to determine that the appropriate response is obtained for the test method. In many cases, the appropriate response will be no signal, a zero result, or absence of a negative effect on the method’s response.

• The candidate water can be used to prepare reagents or media used in the test procedure. Recovery of

expected results for quality control samples that have preassigned expected values (such as reference materials or calibrators provided by the method manufacturer) can be used to validate the suitability of the candidate water used in reagents or media. Recovery of expected results for patient samples and/or quality control samples assayed using reagents or media previously verified to give acceptable results (e.g., from another laboratory or from previous results) can be used to validate the candidate water. The latter approach is also applicable to revalidation of a new lot of prepackaged, bottled water, or revalidation of a water purification system following maintenance or sanitization.

• The candidate water can be used to prepare quality control samples (when reconstitution with water is

appropriate), and results for those quality control samples can be compared to results for the quality control samples prepared with an authenticated source of purified water. Authenticated sources of purified water may include: 1) a different lot that was previously validated; 2) a sample of validated water that was saved prior to a purification system sanitization, and was stored under conditions that would prevent contamination; or 3) a different supplier or an alternate purification system with specifications that are likely to provide water that is fit-for-purpose. If clinical laboratory results using either water source agree, it is likely the candidate water is acceptable.

• Water from an existing purification system can be validated based on review of existing quality

control records for the laboratory testing procedures, and confirmation that the patient results from the procedures meet the clinical requirements of the healthcare system.

5.2 Trend Monitoring of Water Purity Specifications After validation that a grade of water (e.g., CLRW) is fit for purpose, it is critical to ensure that the water continues to meet its specifications. Verification of purity specifications is accomplished by measuring the appropriate parameters at established regular intervals. The data should be monitored using trend graphs analogous to Levey-Jennings7 or Shewhart 8 control charts commonly used in clinical laboratories. The purpose of monitoring water quality is twofold: 1) to document the water is of the specified quality at a point in time; and 2) to detect deterioration of purification components before the deterioration impacts the acceptability of the water. Results from system monitoring must be evaluated at regular intervals to provide evidence that the preventative maintenance schedule is adequate to maintain the purified water specifications. The laboratory should also establish a procedure for responses to excursions from the established specifications that includes a review of past water testing results, an evaluation of the impact of the excursion, including the possible effect on released test results, and documentation of corrective action. The laboratory should also re-evaluate the adequacy of the testing frequency when such an excursion from specifications occurs. It is important to recognize that deterioration in a measured parameter can indicate the need for system maintenance before the value for the parameter no longer meets the specification. For example, a purification system that uses mixed bed ion-exchange technology may produce water with a resistivity close to 18 MΩ • cm, referenced to 25 °C. For such a system, even though the resistivity is above the 10 MΩ • cm specification, a 1 to 2 MΩ • cm decrease in output resistivity may represent a significant change that may require prompt maintenance to replace deteriorating components. As an ion-exchange bed becomes exhausted, strongly ionized species may displace weakly ionized species that conduct little current and can reach unacceptable concentrations, even though the resistivity of the product water is greater than the minimum specification. Refer to Section 6.4 for more details on ion-exchange technology.

Volume 25 C3-P4


5.2.1 Maintenance of the Water Purification System Procedures must be established for maintenance and/or replacement of water purification system components to ensure that the product water continues to meet specifications. Components can fail for a variety of reasons: UV lamps deteriorate with use, ion-exchange or sorption beds become exhausted as they take up contaminants and filters can become blocked, perforated, or contaminated. Trend monitoring of parameters that measure product water specifications makes it possible to anticipate some maintenance. The measurement equipment itself requires calibration, and system controls require inspection and testing on a regular schedule. The frequency of maintenance activities should follow, as a minimum, the manufacturer’s recommendations. It is essential that the product water be tested to verify that it meets specifications following maintenance or component replacement. 5.2.2 Water Purification System Sanitization Sanitization of the water purification and distribution system is critical to ensure microbial contamination is controlled within specifications. Sanitization frequency must be adequate to maintain the purity specifications and is established based on system usage, regular quality control trend data, and the system manufacturer’s recommendation. Validation of the sanitization process should include verification that the system has been flushed sufficiently to remove or reduce traces of the sanitizing agent to levels that are consistent with the water being fit for purpose (see Section 5.1). 5.3 Water Purification System Validation Once a grade of purified water has been validated as fit for purpose in the laboratory testing procedures, the purity specifications for that water should be incorporated into a validation procedure for the water purification system itself. A validation procedure for the water purification system is used to document the system’s ability to deliver adequate volumes of purified water of the stated specifications. 5.3.1 Validation of New and Major Upgrades to Water Purification Systems: A Master Validation Plan The master validation plan encompasses all major steps to be taken in order to accomplish successfully the project design, implementation, verification of fitness for use, and use of the purified water system according to the validation plan. The following formal documentation is recommended: • a URS for the purified water grade required by the laboratory; • documents relating to consultation with water system manufacturers/suppliers regarding design

features, and system support by the manufacturer, as well as documentation (to be provided by the manufacturer) of the prospective system’s capability to meet or exceed the laboratory’s purified water requirements;

• after the system has been installed, verification and documentation of installation qualification,

operational qualification, and performance qualification over time; • written and, where appropriate, validated procedures and schedule for calibration, maintenance,

sanitization, sampling and testing; • schedule and criteria for requalification of the water system, for example, following replacement of

critical components, following unexpected failures, when introducing new laboratory procedures, or as a periodic verification; and

• training procedures and log of trained personnel on all pertinent procedures and practices.

Number 13 C3-P4


Information gathered during the installation, operational, and performance qualifications can be used to determine initial operating parameters, monitoring procedures, and preventive maintenance schedules. 5.3.1.1 Installation Qualification

The installation qualification document establishes that the installed equipment meets all criteria and specifications detailed in the purchase order; that utilities (electricity, city water, or other feed water type) with the proper specifications are available; and that each component of the system (monitoring devices, alarms, gauges, motors, valves, switches) is correctly installed. An approved drawing of the system should be provided and validated for conformity to design. Evidence of calibration of the measuring and/or controlling instruments should be provided (if pertinent).

5.3.1.2 Operational Qualification

The operational qualification document establishes through testing and/or verification that each module or component of the system operates in full conformance with the agreed URS (mechanical, instrument, alarm, ON/OFF switches, applications software, etc.). Training of personnel involved in the testing and maintenance of the system should be planned at this stage.

5.3.1.3 Performance Qualification

The performance qualification document establishes the capability of the system to deliver water of the specified purity under normal operation and under adverse operating conditions. Adverse conditions might need to be simulated and could include potential fluctuations in feed water, changes in ambient temperature, etc., that could challenge the purification system.

5.3.2 Retrospective Qualification of Existing Water Purification Systems During routine operation, an existing water purification system may be qualified based on review of existing parameter measurement data gathered since the beginning of the equipment’s installation. Because the data have not been generated under the preplanned conditions usually specified in a validation protocol, the operational and performance limits of the system have not necessarily been checked, and consequently may not be known to the user. Review of existing trend monitoring data brings awareness of reliability issues, adequacy of maintenance and sanitization practices, and other useful facts about the system. The data review should include test results from water samples collected on a regular basis over a significant time period (typically two years), if available. The purpose of this data review is to assist with creation of a prospective validation plan. Generation of statistical trend graphs or use of other statistical monitoring tools are recommended to review the retrospective data, as it visually brings attention to outliers. Follow-up of outlier data will identify conditions that may warrant modification in the parameter test frequency or changes in the system maintenance protocol. 6 Design Considerations The purpose of this section is to provide laboratories with sufficient knowledge about water purification technologies to communicate effectively with water purification specialists and make informed decisions regarding the purchase and use of water purification systems. The section is not intended to be an in-depth review of water purification technologies; detailed information is available in texts and review articles.9,10 Complete systems are not discussed and recommendations for combining discrete technologies into systems have not been provided, because there are numerous effective configurations to produce water that meets the specifications for CLRW or SRW. However, there are a number of important generalizations that apply to any system.

Volume 25 C3-P4


The contaminant levels of a municipal water supply can vary over the course of a year, and a contamination profile over a year or two can be helpful in the selection of a water purification system. The selection of the initial stages of a purification system, frequently referred to as “pretreatment stages,” will depend on the characteristics of the feed water. The primary purpose of initial treatment stages is to reduce damage to subsequent components to ensure reliable operation of a water purification system, and to decrease the cost of operation by preventing excessively frequent replacement of more expensive components. For example, a bench model purification system without pretreatment may be capable of producing CLRW when fed directly with water from the municipal supply; however, costly expendable components may exhaust so frequently that the system will be unnecessarily expensive to operate, and the purity of the product water may be inconsistent. Microorganisms and their by-products represent a particular challenge. In the absence of continuous biocidal conditions, microorganisms will grow as biofilms on all the wetted surfaces of water purification components, including storage tanks and the plumbing of a distribution system. A biofilm is a layer of microorganisms embedded in an organic matrix composed mostly of glycoproteins and heteropolysaccharides. Organisms in biofilms can multiply even when the concentration of nutrients in the main volume of water is very low, and the layer protects the organisms from periodic treatment with biocides that are primarily effective in killing planktonic (free-floating) microorganisms. The large surface areas of filters and beds of carbon or ion-exchange resins are likely to support the growth of microorganisms. Filters and beds also concentrate water contaminants that microorganisms can use as nutrients. Sloughing biofilm and by-products of microorganism growth and metabolism (e.g., endotoxins) are always potential contaminants of water. The potential for microorganism contamination and its control are discussed for each type of water purification technology. 6.1 Filters 6.1.1 Microporous Filters There are two major types of microporous filtration media: depth filters, characterized by nominal particle size ratings, and screen filters, characterized by absolute particle size ratings. Depth filters are matted fiber or material compressed to form a matrix that retains particles by random adsorption or entrapment. Screen filters have uniform molecular structures that, like sieves, retain all particles larger than the precisely controlled pore size on their surfaces. Depth filters (typically 1 to 50 µm) are commonly used as an economical way to remove the bulk of suspended solids and to protect downstream purification technologies from fouling and clogging. Screen filters (0.05 to 0.22 µm) are typically used as close as possible to the point of use to trap microorganisms and fine particulates. Trapped particulates, including microorganisms or their metabolic products, and soluble matter, can be leached from filters, and suitable maintenance (regular sanitization and periodic replacement) is necessary to maintain desired levels of performance. Newly installed filters usually require rinsing before use to remove extractable contaminants to levels that meet testing parameter specifications. 6.1.2 Ultrafilters “Ultrafilter” is a designation for a membrane filter that removes particles as small as protein macromolecules. Ultrafilters are characterized by molecular weight cutoffs (usually described in dalton units), and the efficiency with which they reduce the concentration of relevant contaminants to acceptable levels. Ultrafilters are usually installed near the outlet of a water purification system to reduce the concentration of microorganisms and large organic molecules, including nucleases and endotoxins. Ultrafilters must be

Number 13 C3-P4


regularly sanitized and/or replaced to maintain their effectiveness. Ultrafilters can be installed in traditional fashion, in which all the water flow is straight through the membrane, or in “cross-flow” (tangential flow) fashion, in which a portion of the input water flows across the membrane surface to reduce fouling by continuously rinsing away contaminants. 6.1.3 Vent Filters Hydrophobic microporous filters or ultrafilters are often fitted to water storage containers as vent filters in order to prevent particulates, including bacteria, from entering the stored water. By combining absorptive media with filter media, multicomponent vent filters can also minimize CO2 and organic contamination of stored water. Regular replacement is essential to maintain effectiveness; the manufacturer’s recommendations should be followed. 6.2 Reverse Osmosis (RO) Membranes RO membranes are used to remove contaminants that are less than 1 nm (10 Å) nominal diameter. The RO process typically removes over 90% of ionic contamination, most organic contamination, and nearly all particulate contamination from water. RO removal of nonionic contaminants with molecular weights <100 daltons is dependent on many variables and can be low. Rejection rates increase toward complete removal as molecular weights increase above 100 daltons. In theory, the removal of particulate material, including colloids and microorganisms, and molecules with molecular weights of >300 daltons will be complete; however, RO membranes may be imperfect and some particulates may pass through them. RO is frequently used as an early stage in water purification systems. During the RO process, feed water is pumped past the input side of an RO membrane under pressure (typically 4 to 15 bar, 60 to 220 psi) in cross-flow fashion. Pretreatment of the feed water with microporous depth filters and activated carbon is usually required (see Section 6.5) to protect the membrane from large particulates, transition metals, and free chlorine. Typically, 15 to 25% of the feed water passes through the membrane as permeate, and the rest exits the membrane as concentrate that contains most of the salts, most of the organics, and essentially all of the particulates. The ratio of the volume of permeate to the volume of feed water is referred to as the “recovery.” Operating an RO system with a low recovery will reduce membrane fouling, especially fouling that results from precipitation of low solubility salts. However, recoveries of 75% are possible, depending on the quality of the feed water and the use of filtration and softening pretreatment. The performance of the RO component of a water purification system is typically monitored by measuring the percent ionic rejection, which is defined as follows for either conductivity (Equation 1, more common) or resistivity (Equation 2) measurements. Both methods yield equivalent results; provided the conductivity (κinput and κpermeate) or resistivity (ρinput and ρpermeate) are expressed to a corresponding number of significant figures.

Percent ionic rejection = %100•κκ-κ

=ROinput

permeateinput (1)

Percent ionic rejection = %100•ρ

ρρ=OR

permeate

inputpermeate - (2)

The percent ionic rejection is used as a measure of membrane performance and not as a gauge of water quality. A typical percent rejection is ≥90%. The percent ionic rejection will vary with the feed water and condition of the RO membrane. The percent rejection should be recorded when a new membrane is installed and trended thereafter. Sudden changes require investigation and possible maintenance. After

Volume 25 C3-P4


extended periods of use, RO membranes deteriorate and must be replaced. Manufacturer guidelines should be followed. RO membranes are available in two common types: CA membranes made of cellulose diacetate/triacetate materials; and TF (thin film) membranes made of polyamide materials. TF membranes are stable over a greater pH range, reject monovalent ions somewhat more effectively, and provide approximately twice the throughput of CA membranes. However, oxidizing agents such as chlorine damage TF membranes, whereas CA membranes are relatively chlorine tolerant. The chlorine tolerance of CA membranes can be an advantage in some systems. Chlorine passes through RO membranes, so the growth of microorganisms on both sides of the membrane and in any components of a distribution system will be limited. The residual chlorine or chloramine in the RO water can be removed at the periphery of the distribution system during further steps of purification. 6.3 Contactor Membranes A contactor device uses a hydrophobic membrane filter to remove volatiles (e.g., CO2, O2) from water. The water stream passes on one side of the membrane, and a flush gas or vacuum removes volatiles from the other side of the membrane. The removal rate of a species is dependent on the permeability of the membrane, the contact area, contact time, and the difference in partial pressure across the membrane.

6.4 Ion-Exchange Resins Beds of ion-exchange resins can efficiently remove ionized species from water by exchanging them for H+ and OH- ions. Ion-exchange resins are porous beads made of highly cross-linked, insoluble polymers with large numbers of strongly ionic exchange sites. Ions in solution migrate into the gel matrix (internal) and crevice (external) spaces within the beads where, as a function of their relative charge densities (charge per hydrated volume), they compete for the exchange sites. Strong cation resins are polysulfonic acid derivatives of polystyrene cross-linked with divinylbenzene. Strong anion resins are benzyltrimethyl quaternary ammonium hydroxide (Type 1) or benzyldimethylethyl quaternary ammonium hydroxide (Type 2) derivatives of polystyrene cross-linked with divinylbenzene. The degree of cross-linking in the resins can be controlled, with increased cross-linking producing a more rigid, less porous gel matrix. Resins can be specialized for certain applications by altering the nature of the charged groups, the degree of cross-linking, and the size of crevices. However, there are tradeoffs. Resins optimized for the removal of well-ionized inorganic salts are extremely efficient at this task, comparatively rugged, and regenerate well. On the other hand, macro reticular resins, which are loosely cross-linked and contain large crevices to improve the uptake of organic ionized species, are more fragile and difficult to regenerate. Ion-exchange resins are available in a variety of types and grades, which are selected based on the source water and application requirements. Beds of ion-exchange resins are available as cartridges or tanks and are typically used for a period of time and then replaced with new, or regenerated, beds (batch operation). Keeping the anion and cation resins separated in their own tanks makes regeneration simpler; however, mixing the resins results in more efficient ion removal. When a bed of exchange resin is regenerated off premise, it may be blended with beds from other locations during bulk regeneration, and the possibility of unexpected contamination should be considered and discussed with the supplier. Use of virgin resins avoids this potential problem. The limitations of ion-exchange resins arise primarily from their large surface area and the batch mode in which they are commonly operated. Ion-exchange materials must be highly porous and have extremely large surface areas in order to function effectively; however, this makes them a potential breeding place for microorganisms. The release of fines and soluble components depends on the initial quality of a resin. Regeneration can damage the molecular structures of resins, resulting in greater release of fines and soluble components. For all these reasons, bed volumes should be kept as small as reasonably possible. Filters are typically installed after the beds to trap fines and other particulate matter. Note that lower grade

Number 13 C3-P4


ion-exchange resins can contribute soluble organic contaminants that may not be removed by subsequent stages of purification. Beds should be designed and operated to minimize the separation of resin components and entrapment of air, which will substantially reduce the efficiency of performance. Ion-exchange materials shed more contamination when they are first installed than they do under steady-state conditions, so it is important to rinse them after installation. The amount of rinsing required varies with the type of resin used, but a minimum of several bed volumes is typically required. The manufacturer’s instructions for the rinse procedure should be followed when beds are exchanged.

Release of Organics and Silica from Resin Pack on Exhaustion

15

15.5

16

16.5

17

17.5

18

18.5

Volume0

10

20

30

40

50

60

70

80

90

Con

cent

ratio

n (p

pb)

Resistivity

Silica

TOC

Figure 1. Release of Organics and Silica From an Ion-Exchange Resin on Exhaustion. (From Whitehead P. High purity water—Are users getting the quality they expect? Ultrapure Water. 1996;13(6):37-40. Reprinted with permission.) Because most ion-exchange beds used in batch mode operate in quasichromatographic fashion, they do not simply become less efficient as they exhaust; they release pulses of contaminants that have accumulated from the water stream since the time the beds were first placed in service. Strongly bound contaminants will displace weakly bound contaminants, so the first pulses of contaminants to be displaced from exhausting ion-exchange beds may be weakly ionized substances that have little effect on the resistivity of the product water. Thus, resistivity monitoring is unlikely to detect the initial release of these weakly ionized species, including charged organics, silicates, borates, etc. This situation is illustrated in Figure 1, which shows the relationship between resistivity, silica, and TOC measurements as an ion-exchange bed begins to exhaust. It is impractical to monitor continually for the potential release of all types of weakly ionized substances; instead, purification systems should be designed, operated, and maintained to prevent their release in concentrations that can affect applications. A validated schedule for bed replacement, or an appropriate alert level, is essential to prevent release of impurities (see Section 5.1). Examples of approaches that have been reported to control the release of weakly ionized substances include: • Use of ion-exchange beds in a passive, redundant fashion in which two identical beds are connected

in series. When the first bed exhausts, as indicated by monitoring resistivity after the first bed, any weakly bound ionic substances that have been released from the first bed will have been bound by the

Res

istiv

ity (M

Ω •

cm)

Volume 25 C3-P4


second bed and not released into the product water. At this point, the first bed can be replaced with the second bed and a new bed installed in the second position, or both beds can be replaced at the same time.

• Use of bed components that include specialized resins that are optimized to bind certain weakly

ionized species.11 6.4.1 Electrodeionization (EDI) Electrodeionization is a technology combining ion-exchange resins and ion-selective membranes with direct current to remove ionized species from water and regenerate the resins.12 Ions that become bound to the ion-exchange resins migrate to a separate chamber under the influence of an externally applied electric field, which also produces the H+ and OH- necessary to maintain the resins in their regenerated state. Ions in the separate chamber are flushed to waste. The ion-exchange beds in EDI systems are regenerated continuously, so they do not exhaust in the manner of ion-exchange beds that are operated in batch mode. EDI beds are typically smaller and remain in service for longer periods than batch mode resins. EDI is typically combined with reverse osmosis or other pretreatment to ensure it is not overloaded with high levels of salts, and to minimize growth of microorganisms. The small volume of resins results in less bleed of organic molecules. EDI components are typically designed to produce product water with resistivity above 5 MΩ • cm at 25 °C.13 6.5 Activated Carbon Activated carbon is used to remove oxidizing biocides and organic compounds from water. Most activated carbon is produced from charcoal that is created by destructive distillation of coconut, pecan shells, wood, or coal at 450 °C. The charcoal is then ground and “activated” in a process involving roasting at 800 to 1000 °C in the presence of water vapor and CO2. Acid washing removes much of the residual alkaline oxide ash (e.g., MgO, CaO, Na2O, and K2O), iron, and other soluble material. Activated carbon can also be made from polymer beads (referred to as synthetic activated carbon) which contain fewer noncarbon contaminants than natural sources of carbon and are less likely to produce fines (see below). Activated carbon used in water treatment usually has pore sizes ranging from 5 000 to 10 000 Å and a surface area on the order of 1000 m2/g. Activated carbon takes up water contaminants through adsorption (to the surface) and absorption (into the spaces) by virtue of moderate attraction (ionic forces), weak attraction (polar and Van der Waals forces), and by surface-active attraction. The combination of adsorption and absorption is referred to as sorption. A major use of carbon is the removal of oxidizing biocides from water, so they do not damage membrane filters and ion-exchange resins. One gram of activated carbon will react chemically with 2 to 4 g of chlorine, oxidizing the carbon and producing chlorides. This reaction is very rapid, and small carbon filters can effectively remove chlorine from water. Organic fouling can interfere with the removal of chlorine by activated carbon. Such fouling is dependent on the local water supply, and this should be considered when sizing units. An increasing number of municipalities have begun using chloramines as the primary disinfectant, because they are safer to use and less likely to halogenate naturally occurring organic impurities to produce toxic substances in water supplies. Chloramines also cause much less damage to membranes. The breakdown of chloramine by carbon is a catalytic reaction producing ammonia, nitrogen, and chlorides, and the reaction is slow compared to the breakdown of chlorine. It is necessary to ensure adequate carbon volume for the feed water characteristics and application. The supplier should be consulted for sizing recommendations. The second major application of activated carbon is in the removal of organic compounds from purified water. Activated carbon does not remove all dissolved organic contaminants found in water, but its use can produce a significant reduction in TOC. Contact times necessary to remove organics vary with the structure of the molecules. Some organic molecules have large contact time requirements and will not be

Number 13 C3-P4


effectively removed by carbon bed sorption. The trace capacity number, the uptake of acetoxime, is a predictor of how well a particular type, or lot, of activated carbon can remove organics from water.14-16 Beds should be designed and operated to minimize the entrapment of air that will substantially reduce their efficiency and performance. Activated carbons have extremely large surface areas in order to function effectively, and they are a source of endogenous and accumulated inorganic and organic substances, which makes them a breeding place for microorganisms.17-19 Activated carbon is available with added insoluble biocides that are intended to retard the growth of microorganisms. Activated carbon beds need to be changed frequently enough, based on monitoring microbial contamination, to limit bacterial build-up to levels that are acceptable for the purity of the final product water. Activated carbon beds are prone to releasing fines and soluble components into the water stream. Fine particles will be present in new beds of activated carbon, but they also develop as the result of mechanical and chemical effects during use. Molded and encapsulated activated carbon cartridges produce fewer fines than granular beds. Pulses of weakly sorbed organic contaminants will be displaced from beds of activated carbon by more strongly sorbed contaminants as the beds exhaust. However, activated carbon beds may also release pulses of previously bound contaminants as the result of abrupt changes in the input water (e.g., temperature; flow rate; concentration of ions, especially divalent ions; surface active contaminants). 6.6 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. Distillation can remove all classes of water contaminants, with the exception of those that have vapor pressures close to water (including azeotropes), and should not contribute organic or microbial contamination. Stills can be used as a final stage of purification to minimize microorganism and organic contamination. Not all laboratory stills are designed to prevent CO2 in the air from dissolving in the distilled water, which means that the resistivity of the water they produce cannot be more than about 1 MΩ • cm, referenced to 25 °C. In order to verify the CLRW resistivity specification, measurements must be performed after CO2 has been removed (see Section 7.1.2.3). 6.6.1 Central Stills Single-stage, central stills are generally energy inefficient. Their boilers are commonly heated by boiler plant steam; and if a heat exchanger has developed pinholes, boiler steam containing azeotropic additives, such as morpholine, will mix with the boiler water. If the water feeding the boiler is not pretreated and boiler overflow is used to reduce scaling, overflow will increase the concentration of volatile contaminants in the distillate. Condensers of central stills must be well maintained or they may contribute metals, and possibly other contaminants, to the distilled water. Modern, centralized stills use multiple effects (ME) or vapor compression (VC) designs to achieve energy efficiency and improved purity. Such stills can reduce the total dissolved solids by more than 100 times, though they may not be especially effective at reducing the concentration of volatile water contaminants. The water feeding their boilers is usually pretreated to further improve purity and to minimize maintenance. 6.6.2 Laboratory (Small-Scale) Stills Single-stage laboratory stills are most frequently used to improve the purity of already purified water (e.g., RO, deionized, or centrally distilled).

Volume 25 C3-P4


Distilling water as a single batch (i.e., no feed) in a fractional still with many plates can produce very pure water. Organics can be digested in the boiler with oxidizing agents such as permanganate and volatile contaminants discarded before the pure water fraction is collected. However, batch distillation is rarely performed in laboratories, because it is labor intensive and very costly to automate. The great majority of laboratory stills are continuous; as boiler water is distilled away, it is replaced with fresh feed water. Chemically resistant glasses (i.e., borosilicates) are the construction materials of choice, because they are stable and can be fashioned into the intricate, seamless shapes required. In order for a continuous still to take maximum advantage of the distillation process, it must boil the feed water under conditions that approach ideal and condense the product water in equilibrium with the most uncontaminated steam possible. 6.6.3 Boiler and Boiler-Condenser Transition Stages From a maintenance point of view, laboratory still boilers should be designed to withstand the aggressive dissolving power of pretreated water. Laboratory stills should not be fed with water that precipitates during boiling, because precipitates coat and damage heaters and creep along still surfaces. Contaminants that have vapor pressures lower than water must be excluded from the steam before it enters the condenser stage of a still, or they will not be removed. Therefore, the still boiler and boiler-condenser transition stage must be designed to minimize contamination of the steam entering the condenser. Contaminants with negligible vapor pressures compared to water (e.g., most inorganic salts; particulates, including microorganisms; colloids) should be eliminated by distillation. However, undistilled boiler water can enter the condenser stage by splashing, surface entrainment, and steam entrainment. • Splashing – As water boils, splashing is inevitable; however, a still boiler can be designed to

minimize splashing, and splashes that do occur can be prevented from reaching the condenser by providing enough height above the surface of the boiling water and by trapping rogue splashes in a transition stage between the boiler and condenser.

• Surface Entrainment – The draft of steam in a still can propel a film of contaminated water along

wetted surfaces from the boiler into the condenser. Such surface entrainment can be minimized by designing a still that avoids excessive steam velocities, places a vertical hydrophobic barrier in a transition stage between the boiler and condenser to interrupt the film of water, and condenses sufficient steam on vertical wetted surfaces to reverse the net flow of the water film.

• Steam Entrainment – Bursting bubbles of boiler water will form mist that can be carried into a still’s

condenser by the draft of steam flowing from the boiler to the condenser. This steam entrainment can be controlled by designing stills with boilers that have sufficient surface area, so that boiling is not unnecessarily vigorous, and by placing an effective mist filter/trap in a transition stage between the boiler and condenser. Mist particles can be very fine, and simple baffles are not likely to be adequate.

The concentration of contaminants in the boiler steam, which have significant, but lower, vapor pressures relative to water, will be a function of their concentration in the boiler water. Means must be provided for regulating their concentration in the boiler water, or the purity of the product water will be inconsistent. This can be accomplished by periodically draining the still boiler or by means of overflow. Overflow is less efficient than draining and may increase the boiler steam concentration of contaminants that have vapor pressures higher than water. 6.6.4 Condenser Stage Contaminants that have vapor pressures higher than water are removed in the condenser stage of a still.

Number 13 C3-P4


Compound (multistage) condensers that equilibrate steam and boiling-hot water in multiple, specialized compartments are necessary in order to remove these contaminants efficiently. Such condensers are designed to equilibrate the product water with a compartment of steam containing low concentrations of contaminants, while concentrating the contaminants in a separate compartment of steam for efficient venting. Simple (single-stage) condensers cannot be effective, even if they operate close to the temperature of boiling water, because the distilled water product is in contact with the steam that is being vented. Increasing the concentration of contaminants in the vent steam of a simple condenser will necessarily increase their concentration in the product water. Condensers that operate at temperatures below the boiling point of water will also be ineffective, because they cannot concentrate contaminants for effective venting. Coil and jacket heat exchanges will have cold regions (potentially near freezing during winter at more polar latitudes) where the cooling water enters, unless they are fed from temperature-stabilized recirculating tanks. The purity of the product water will also depend on how effectively contamination in the ambient air (e.g., dust, volatiles) is prevented from entering the condenser. When a continuous still is not running, its condenser is not under a positive pressure of steam, and particulates in the ambient air can enter through the vents. Means must be provided for sanitizing a still’s condenser each time the still begins a cycle and for diverting any distilled water produced during the sanitizing period away from the storage/distribution system. Contaminants with vapor pressures higher than water will evaporate from boiling water relatively more quickly than water. Therefore, their concentration in the steam that enters a still’s condenser will be a function of the rate at which feed water enters the still’s boiler. Overflowing water through the boiler will increase the concentrations of these contaminants in the steam that enters the condenser above what their concentrations were in the boiler feed water. Means must be provided for regulating the rate at which water flows into the boiler of a continuous still, or the purity of the product water will be inconsistent. When water overflows from the boiler of a still designed with a simple coil or water jacket condenser, the concentration of contaminants with vapor pressures higher than water is likely to be greater in the product water than the feed water. 6.7 Ultraviolet Light Ultraviolet light is used as a bactericide and to breakdown and oxidize organic contaminants. Various UV sources may be used, but the most common are low-pressure mercury lamps that give effective emissions predominantly at wavelengths of 185 and 254 nm. Medium-pressure mercury lamps are used for larger water treatment systems and produce higher intensities of broader band emissions. The intensity of UV radiation actually penetrating a purified water stream can be reduced by films, including biofilms, developing on the UV source window, or by the source lamp and protective sleeves becoming less transparent to UV light. A UV meter, placed on the opposite side of the flow chamber from the UV source window, may be used to provide quality control information for large systems but may be impractical for smaller point-of-use systems. When it is not practical to install a UV monitor, the lamps must be replaced on a schedule that will assure adequate performance. 6.7.1 Bacterial Contamination Reduction UV light damages microbial DNA and enzymes (such as RNA polymerase) and is an effective biocide at sufficient dosages. A sufficiently high UV dose is needed to prevent the production of live, but nonreplicating, microorganisms that will not be identified by a plate count (see Section 7.2). Epifluorescence microscopy can be used to determine the actual number of living microorganisms in water that have been irradiated with UV light, and endotoxin testing will provide information regarding the concentration of cell wall material from living and fragmented microorganisms.

Volume 25 C3-P4


6.7.2 Organic Contamination Reduction Passing water over a source of combined 185/254 nm UV light can effectively reduce residual organic contamination in water. The UV light breaks large organic molecules into smaller ionized components which can then be removed by a downstream ion-exchange resin bed. This process must be engineered carefully to avoid reintroducing organic and microbial contaminants, which may bleed from the ion-exchange resin. High purity ion-exchange resins that do not bleed significant levels of organic contaminants are necessary and must be monitored carefully for replacement at appropriate intervals. 6.8 Storage and Distribution High-purity water is likely to become contaminated when it is stored or distributed through piping. Potential contaminants fall into the categories of soluble inorganic and organic compounds including dissolved gases, particles, and microorganisms. Contamination can result from contact with the materials from which storage and distribution systems are constructed, it can enter from the environment, and it can be produced by microorganisms within the system. The impact of potential contaminants should be assessed as part of the validation of a purified water as fit for purpose. Bacterial contamination is the most persistent problem in water storage and distribution systems. Bacteria have a great capacity for adjusting to nutrient-poor environments, such as pure water systems.20 Once bacteria establish, they provide the nutrients for other microorganisms to grow within the biofilm. Biofilms can exist in containers or distribution systems and their products can have an insidious effect on a wide range of applications. Good design and proper maintenance programs are needed to minimize the impact of bacterial contamination. 6.8.1 Materials The materials from which containers and distribution systems are constructed must have low permeability, must not contribute significant contamination to the water, and must withstand the agents and conditions used for sanitization without being damaged. Smooth surfaces and a minimum of crevices make controlling bacterial growth somewhat easier. Materials used to join plastic and metal pipes can be an unexpected source of chemical or microbiological contamination. Materials that are apparently inert in pure form can contaminate water with soluble substances, such as: a) polymerization catalysts, monomers, or short chains of polymers; b) plasticizers; c) substances used to facilitate release from production molds; and d) recycling contaminants. As a general rule, virgin, pure materials that have a history and data to support use with high-purity water are recommended. 6.8.1.1 Glass Chemically resistant glass (i.e., borosilicates) may be useful for small-scale storage systems. It is nonporous, impermeable, and easily sanitized. Borosilicate glass will contribute trace amounts of ions to purified water. Soft glasses are not recommended because they contribute significant amounts of ions to purified water. The fragility of glass precludes its use for large containers or distribution systems. 6.8.1.2 Metals Stainless steel distribution systems (Types 304 and 316) have been widely used. Other metals are rarely used because of either expense or contamination issues. Welded, polished joints are preferable. After assembly, stainless systems must be polished and passivated with nitric acid, a process that dissolves exposed iron from the surface and produces an impervious chromium and nickel oxide surface layer. The surface finish should be inspected at regular intervals to determine when repassivation needs to be performed. Inadequate passivation can result in visible rust, called rouge, which supports biofilms.

Number 13 C3-P4


Biofilms have been credited with pitting and rusting stainless systems. Stainless steel is widely used when high pressure and high temperature steam sterilization is used for sanitization. 6.8.1.3 Plastics Plastics have been widely used to construct high-purity water storage containers and distribution systems. Polyvinylidene fluoride (PVDF), acrylonitrile-butadiene-styrene (ABS), and clean grades of polyvinylchloride (PVC) are used for rigid piping systems. PVDF is suitable for high-temperature heat sterilization applications. Polyethylene (PE), polypropylene (PP), polyethylene terepthalate copolymers (PET), polyacetal, polyamide (PA), and high-density polyethylene (HDPE) are commonly used when flexible tubing is required. Less commonly used plastics include: polyether-ether-ketone (PEEK), perfluoroalkoxy vinyl ether (PFA), polytetrafluoroethylene (PTFE), fluorinated ethylenepropylene (FEP). New polymers or combinations of polymers are constantly being developed. Composite multilayer pipes and tubes provide a variety of desirable features. In all cases, care should be taken to ensure that materials used are of suitable purity and do not contain additives or fillers that can contaminate the water. Many of these materials can be welded. Ultrasonic welding of plastic tubing can produce very clean connections. If solvent welding is used, allowance has to be made for an extended rinsing period. Compression or push-fit connections have been used successfully. All plastics are to some extent permeable to volatile substances, but there is a wide range of permeability. The significance of gas permeability to the purity of the water depends on factors such as wall thickness and residence time of the water. Appendix A provides examples of CO2 permeability under various conditions. Other higher-molecular weight volatile compounds are not likely to significantly permeate through plastic because of their relatively low ambient concentrations and lower permeability. 6.8.2 Storage Storage containers should be constructed of low gas permeability materials that contribute minimum contamination to the water and are not damaged by the sanitizing agents that will be used. The container should be opaque, or protected from light, to minimize the growth of autotrophic organisms. Storage containers that are small enough to be used to transport water should not be continually topped off. They should be sized to ensure that they are refilled at least every few days, drained between fillings to prevent stagnation, and sanitized on a regular maintenance schedule. It is good practice to equip the vents with hydrophobic filters (typically ≤0.22 µm) to prevent particulates and microorganisms from entering. Storage tanks that are part of permanent systems should be kept as small as practical. The water in a reservoir may be recirculated continuously or intermittently through water purification processes to maintain purity. If recirculation is used, attention should be paid to potential temperature increase that could contribute to microbial growth. It is good practice to minimize contact with ambient air by equipping storage tank vents with hydrophobic filters, and if applicable, with adsorbents/absorbents to remove organic vapors and carbon dioxide. Inert gas blanketing can help to maintain a high resistivity and, when make-up water is not already saturated with oxygen and nitrogen, reduce the growth of biofilm. High-purity argon, essentially free of CO2, has an advantage over nitrogen because many organisms can fix nitrogen. UV irradiation may also be used to control microorganism growth. Tanks should be designed so they can be easily drained to facilitate cleaning. Tanks should be drained regularly to avoid static conditions. When an overflow is used, it should be protected to avoid back contamination from the drain.

Volume 25 C3-P4


6.8.3 Distribution Systems The design and choice of materials for distribution systems depends on the purity of the water being distributed, the size of the system and the sanitization method to be used. The components used must withstand the sanitizing procedure that will be used, and should be constructed of materials that contribute a minimum of contamination to the water. Valves at dispense points in distribution systems should be designed to prevent backflow and introduction of contamination. Distribution systems should be designed so that water recirculates through as much of the piping as possible, with minimal dead-legs to prevent parts of the system from becoming stagnant. Ideally, dead-legs should be eliminated; however when they exist, they should be flushed at regular intervals to minimize stagnation effects. Recirculation should include enough water purification processes to maintain a consistent level of purity in the loop. The recirculation flow rate in larger distribution systems is usually continuous, while in smaller laboratory-based systems, intermittent flow may be used to control temperature build-up that could contribute to microbial growth. Historically, a reasonable linear flow rate has been considered advantageous from the standpoint of limiting biofilm growth; however, biofilm will develop at any flow rate and sanitization is the only way to combat it.21-25 The length and complexity of a distribution system is an important consideration in deciding on approaches to control biofilm and other sources of contamination. There are various approaches to controlling biofilm including regular sanitization, distributing moderately pure water containing an easily removed biocide with final purification and quality monitoring at the periphery with point-of-use systems, and maintaining the water in the distribution system at 65 °C or above.26 6.8.4 Sanitization of Storage and Distribution Systems Sanitizing storage and distribution systems must be performed often enough to prevent a significant buildup of biofilm. A variety of sanitization procedures are available and vary with system design.9,10 Validation of the sanitization process should include verification that the system, including all distribution legs to points of use, has been flushed sufficiently to reduce the concentration of any sanitizing agents to acceptable levels. 7 Testing

7.1 Resistivity The measurement of electrolytic resistivity provides a useful, nonspecific assessment of the ionic content of purified water. 7.1.1 Summary of the Theory and Practice of Resistivity Measurement Electrolytic conductivity, κ, and its reciprocal, resistivity, ρ, are related to the ionic content of water. At low concentrations, for a strongly ionized salt, the value of κ is approximately proportional to (and the value of ρ is approximately inversely proportional to) the concentration of the salt or its corresponding ion pair (e.g., NaCl and Na+ + Cl-, respectively) in solution, although the proportionality constant varies from ion to ion. The following relationship exists between κ and the ionic mobilities, u+ and u-, of the component ions, for a salt of 1:1 stoichiometry (e.g., NaCl, MgSO4):

( )−+ +== uuFzcρ

κ 1 (3)

Number 13 C3-P4


Where F is the Faraday constant (96485.3383 C mol-1), c is the molar concentration of the salt, and z is the charge on the cation or anion. For the commonly used units for κ (S cm-1) and u (cm2 s-2 V-1), the conforming units of c are mol cm-3. For mixtures of salts and for salts with non-1:1 stoichiometry, an analogous situation applies. In such cases, the form of Equation 3 becomes more complex, but the basic principles remain the same.27 To the extent that u+ and u- are independent of c, κ is linearly proportional to c. The proportionality is exact at infinite dilution, but the negative deviation increases with increasing c. The values of u+ and u- also depend strongly on the viscosity of the solution, hence on the temperature of the water. For many ions, the relative temperature coefficient of u is around +2%/°C. In addition to this dependence of u+ and u- on the temperature, t, the value of the equilibrium constant for the dissociation of water, Kw, is also temperature dependent. As a result, the net dependence of the conductivity of pure water, κΗ2Ο, on t is greater than that for solutions (see Section 7.1.2.4).

Substances that carry electric current efficiently in water represent only a portion of the spectrum of water contaminants and high-resistivity water can be significantly contaminated with a range of substances that interfere with laboratory applications. The measurement of ρ cannot indicate the presence or concentration of nonionized chemical species, nor of ionized chemical species at the ng/g level. Resistivity monitors consist of a conductivity cell (sensor) with cable and a meter or display unit with associated electronics, frequently provided with temperature compensation. The meter measures the resistance, R, between the sensing electrodes of the conductivity cell. R, ρ, and κ are related by Equation 4, where Kcell is the cell constant for the given cell [sensor]:

cell

1K

R==

κρ (4)

Kcell has the dimensions of length-1. Its value is governed by the geometry of the cell [sensor]. Owing to production tolerances, the value of Kcell typically varies up to 3% from unit to unit within a production run, with an average deviation of ±0.5%.28 The value of Kcell for the specific sensor is normally determined by the manufacturer and is entered in the electronics at installation of the sensor. Recalibration is required if any change occurs to the sensor (e.g., stress/bending at removal or installation) and periodically (typically annually) for verification. If a new sensor is installed, the new Kcell must be taken into account if a valid resistivity is to be displayed. Some conductivity systems may combine this Kcell correction with a revised calibration factor for the meter. In most high-purity water systems, the monitor is configured to provide readout in terms of resistivity. The following discussion is structured in terms of ρ to minimize the need for interconversion. 7.1.2 Equipment and Materials 7.1.2.1 Resistivity Meters Commercially available resistivity meters range from battery-operated, threshold-indicating meters to microprocessor-based digital display meters. Meters with analog display are also acceptable. Manufacturers of reagent-grade water systems frequently have built-in resistivity meters with temperature compensation that are designed to provide accurate measurements in the range of water quality produced. The manufacturer’s instructions should be followed in using all meters.

Volume 25 C3-P4


7.1.2.2 Effects of CO2 Contamination Measurement of the resistivity of high-purity water is subject to the effect of dissolved CO2. The CO2 will produce CO3

--, HCO3- and H+ ions that reduce the resistivity of otherwise pure water to approximately 1

MΩ • cm referenced to 25 °C. The resistivity of pure water at 25 °C (ρ = 18.2 MΩ • cm), exposed to the atmosphere rapidly approaches 1 MΩ • cm (the initial decrease occurs within seconds). For this reason, resistivity for CLRW is performed using an inline measurement, in which contamination by CO2 between the purification step being evaluated and the resistivity sensor is insignificant. Alternatively, the resistivity specification can be verified after removal of dissolved CO2 as described in Section 7.1.2.3. Accurate resistivity measurements of high-purity water (CLRW or SRW) are not feasible when cells are exposed to the atmosphere (e.g., in measurements with dip cells), owing to the ionic contamination outlined above. Therefore, dip cells are not recommended for resistivity measurements of high-purity water. All plastic tubing has a finite permeability to CO2.29 Where plastic tubing is used in an inline system, the residence time of the purified water between the final purification stage and the resistivity sensor should be minimized and the flow rate maintained high enough to avoid significant contamination. Permeability values of CO2 for various plastics30 can be used to estimate semiquantitatively the level of CO2 that occurs for a given flow rate, system layout (i.e., distance from purification output to resistivity sensor), and wall thickness. 7.1.2.3 Verifying the Resistivity Specification When Purified Water Contains Dissolved CO2 Purified water that has been in contact with air (e.g., distilled water, prepackaged purified water) will contain significant amounts of dissolved CO2. Verifying that the resistivity of such water meets or exceeds a specification for CO2 free water requires removing CO2 without adding ionic contamination. CO2 can be removed from purified water at room temperature by sparging with purified gas or by means of a contactor membrane. Heating a sample of purified water to just below the boiling point to drive off the CO2 is not practical because, at this temperature, the resistivity of pure water is nearly the same as that for water with significant ionic contamination. Sparging involves bubbling purified gas through the water in a closed resistivity measurement chamber to remove dissolved CO2 (see Figure 2). Sparging requires the use of very pure gas that will not introduce CO2 or other ionic contaminants to the water. Purified argon (99.9999%) contains <0.5 ppm CO2 and is sufficiently dense to blanket the surface of the sample water during sparging. Many varieties of sparging filters are available. Models commonly used for liquid chromatography have PTFE (polytetrafluoroethylene) bodies and 10-µm nominal porosity polyether-ether-ketone (PEEK) filter elements. Properly rinsed, they do not contribute ions and produce a dense stream of fine bubbles at low gas pressures. Positioning a resistivity cell inside the sparging chamber eliminates the need for connections with the potential for leaking CO2. Care must be taken to ensure that bubbles of sparge gas do not pass between the cell electrodes. If the resistivity cell is positioned external to the sparging chamber, the cell should have as small a volume as possible and be coupled to the sparging chamber as a closed loop. This arrangement will make it possible to efficiently circulate the water in the resistivity cell with the water in the sparging chamber by raising or lowering one with respect to the other. Sparging should be continued beyond the point that the resistivity reaches a maximum to confirm effective operation. A steady decline of the resistivity, after the maximum value has been attained, will indicate that either the gas or some hardware component of the measuring system is adding ion contamination to the sample. If air is leaking into the system, decreasing the flow rate of sparging gas will decrease resistivity, and increasing the flow rate of sparging gas will increase resistivity. Appendix B contains additional information on resistivity measurement in a sparging chamber.

Number 13 C3-P4


Gas

Spar

ging

Cha

mbe

r

ResistivityCell

Figure 2. Configurations for Measuring Resistivity in a Sparging Chamber 7.1.2.4 Temperature Compensation Resistivity meters in modern high-purity water systems normally incorporate a temperature sensor in the sensing unit. The temperature sensor (thermistor) measures the temperature of the water whose resistivity is simultaneously being measured. The meter corrects the measured value of ρ to a calculated value that would be obtained at 25 °C using a compensation algorithm. For the conductivity of aqueous ionic solutions at 25 °C, neglecting the contribution of the solvent water, ∆ρ/ρ is on the order of -2%/°C, owing to the temperature dependence of the mobility of the constituent ionized species. However, for pure water containing no-impurity ionized species, ∆ρ/ρ is approximately -5.3%/°C at 25 °C. This value combines two factors: the temperature dependence of the dissociation constant of pure water, and the inherent temperature coefficients of the mobilities of the resulting H+ and OH- ions themselves. As the concentration of ionic impurities approaches zero, |∆ρ/ρ| at 25 °C increases from 2%/°C to 5.3%/°C, owing to the increasing contribution of the more temperature-dependent ρ of the solvent water. (NOTE: ∆ρ/ρ is expressed as an absolute value to avoid comparing two negative numbers.) The effect for practical measurements is that the temperature coefficient is also a function of ρ. An example showing this functional relationship is given in Appendix C. It is standard practice to correct resistivity measurements to 25 °C. The uncertainty component associated with the temperature correction increases as the actual sensor temperature deviates from 25 °C. Table 1 shows the impact of temperature on ρ for pure water and for a 0.34 ng/g solution of NaCl, which has a resistivity of 10 MΩ cm at 25 °C. An experimental value for ∆ρ/ρ can be measured and used for temperature correction (see Appendix C which describes examples of temperature correction or compensation). For the most accurate measurement in compensated mode, the cell should be held relatively close to 25 °C and an effective ρ - t compensation algorithm must be used.

Volume 25 C3-P4


Table 1. Resistivity Temperature Table (Pure Water) Temperature (°C) Resistivity of pure

water (MΩ • cm31) Resistivity of 0.34 ng/g NaCl in water (MΩ •

cm32) 0 86.19 28.21 5 60.48 22.66

10 43.43 18.30 15 31.87 14.87 20 23.85 12.15 25 18.18 10.00 30 14.09 8.28 35 11.09 6.90 40 8.85 5.79 45 7.15 4.89 50 5.85 4.15

7.1.3 Calibration Calibrating resistivity instruments for measurements of high-purity water is challenging. Stable standard solutions with resistivities of >1 MΩ • cm (conductivity <1 µS • cm-1) are not feasible, because of the potential for ionic contamination, especially by CO2 from the air (see Section 7.1.2.2). Traceable, stable standards are available at lower resistivities (higher conductivities). However, if such (high-conductivity) standards are used to calibrate most commercial cells designed for high-purity water, polarization at the electrodes becomes significant. This polarization, which is a function of the electrode materials used,33 results in an error in the corresponding calibration. An approach that avoids these problems involves producing water that is free of contaminating ions (including those related to CO2) in a closed-loop purification system and using this water as a calibration standard at 18.2 MΩ • cm at 25 °C. Manufacturers of resistivity measuring systems commonly use this approach; however, the instruments and equipment required to demonstrate that the water is, in fact, free of contaminating ions are not readily available in most laboratories. Readers desiring a detailed description should consult the original literature28,34 and/or the manufacturer of the given instrument. All the components of a resistivity instrument (i.e., the meter [electronics], cable, and cell) should be calibrated as a set in accordance with the manufacturer’s specific instructions to achieve the greatest accuracy. Resistivity meters use alternating current of various frequencies and waveforms to minimize polarization of cell electrodes and reject the reactive component of the cell current. Small changes in the impedance characteristics of individual cells can result in significant errors in calibration. For the same reason, installing a standard resistor in place of a cell only verifies the ability of the meter to correctly measure a pure resistance; it is not a meaningful calibration of the complete instrument.35 As noted in Section 7.1.2.4, measured ρ values are generally corrected from the actual measured value at the ambient temperature to the temperature-corrected value at 25 °C. Thus, calibration, and subsequent verification, of the temperature-measuring device at the temperatures of interest are critical to obtaining an accurate value for the temperature-corrected ρ value. The temperature-measuring device is most often a thermistor placed adjacent to, or within, the cell. Its accuracy over the applicable range is verified by placing it into a calibrated temperature bath or by comparison with a calibrated thermometer. The algorithm used to correct ρ to the temperature-corrected value at 25 °C is also critical. The exact algorithm used in the calculation varies among manufacturers. The general approach is illustrated in Appendix C using a simplified approach (Marsh-Stokes equation). A commonly used method for field verification of a resistivity instrument involves using a separate, calibrated resistivity instrument as a reference. The inline cell of the reference instrument is placed in

Number 13 C3-P4


series with, and close to, the cell of the instrument to be verified. The test and reference cells should be close enough to avoid significant contamination of the water stream or change in its temperature as it flows between the two cells, and simultaneously far enough apart to avoid mutual electronic interference between the calibration and test instruments. The validity of this approach depends on these factors and on how accurately the reference instrument has been calibrated. 7.2 Microbial Content by Colony Count Clinical microbiology methods for growing pathogenic microorganisms emphasize handling samples to maximize the recovery of pathogens from clinical specimens. In clinical microbiology, this means providing a rich growth medium and optimal growing conditions that mimic physiologic conditions (temperature and nutrients). These same considerations are applied to recovering organisms that may be present in a purified water system. Purified water is nutrient poor, and is typically produced and distributed at ambient temperatures. Consequently, microorganisms growing in purified water become smaller and grow more slowly than organisms from clinical specimens. The procedures detailed here use conditions to maximize the recovery of microorganisms likely to be found in water systems balanced with the need for timely results. 7.2.1 Total Heterotrophic Plate Count The determination of total heterotrophic plate count (THPC) is an approximation of the viable number of microorganisms present in the system, and is expressed in colony forming units (CFU) per mL. Heterotrophic bacteria are organisms that require organic substrates to obtain carbon for growth and development. Water is tested for THPC by either the membrane filtration or spread-plate techniques. 7.2.1.1 Sample Collection • Water samples should be collected from points of use in the same manner as water is collected for

normal laboratory use. • Water samples should be collected using aseptic technique into sterile specimen containers at each

point-of-use site. Samples must be collected in a manner that prevents contamination by contact with skin or the environment.

• Samples may be refrigerated at 2 to 8 °C for up to 24 hours if testing cannot be accomplished within

one hour. Samples should be allowed to reach ambient temperature before testing. • All samples should be gently and thoroughly mixed prior to processing. 7.2.1.2 Media and Incubation Conditions The choice of media, filter size, incubation time, and incubation temperature should be validated to establish the adequacy of the methodology chosen. The recommended incubation temperature is 20 to 28 °C and the recommended incubation time is at least five days. If it is not practical to have an incubator controlled to this temperature, use a dust-free room temperature cabinet. The following low-nutrient agars are commercially available and suitable for water testing: • plate count agar • R2A agar

Volume 25 C3-P4


7.2.1.3 Membrane Filtration Technique The membrane filtration technique is recommended for water samples that typically have less than 10 CFU/mL. The choice of water volume to test should be based on consideration of typical results obtained from the samples such that test results can be counted (i.e., filter is not overgrown). The sample is filtered through a sterile filter housing using a sterile 0.22 µm filter. Bacteria that are present in the sample are trapped on the filter. The filter is then placed face up on the surface of the nutrient media and incubated. The nutrients from the agar diffuse through the filter allowing the growth of colonies. 1. Pour up to 100 mL of a mixed water sample into the sterile filter unit with the vacuum off. 2. Turn on the vacuum to pull the sample through the filter. 3. Turn off the vacuum and rinse the filter housing by adding approximately 30 mL sterile 0.1% peptone

to the unit. 4. Turn the vacuum on briefly to pull the peptone rinse through the filter. 5. Using sterile forceps, carefully lift the membrane out of the filter unit holding onto the edge of the

membrane. 6. Place the membrane (face up) onto the surface of an agar plate by gently rolling the filter membrane

onto the surface of the media to avoid the formation of air pockets between the filter and the surface of the growth medium.

7. Incubate the plates for a minimum of five days at 20 to 28 °C. Plates should be inverted to prevent

condensation on the agar surface. 8. After incubation, count the number of colonies and divide by the total volume tested to obtain the

actual results. 7.2.1.4 Spread-Plate Technique (Direct) The spread-plate technique is suitable for samples that typically contain high numbers of microorganisms, such as feed water. The sample is applied directly to the surface of the media and then spread across the surface of the media with a “hockey stick.” 1. Using a sterile pipette, transfer 1 mL of sample onto the surface of one or two agar plates, depending

on expected results, but maximum 1 mL per plate. The entire 1 mL of water must be spread on the agar surface. Two plates may be necessary to avoid water pooling on the surface.

2. Spread the sample over the entire agar surface using a sterile “hockey stick” to ensure proper sample

distribution. 3. Incubate the plates at 20 to 28 °C for a minimum of five days. Plates should be inverted to prevent

condensation on the agar surface. 7.3 Microbial Content by Epifluorescence Microscopy Epifluorescence microscopy is an effective and timely means of determining the concentrations of microorganisms in purified water.36-38 This technique can be used to detect and distinguish between living and dead microorganisms in under an hour, which makes it useful when rapid corrective action is

Number 13 C3-P4


indicated. Epifluorescence counts are likely to differ considerably from those obtained from plate counting because microorganisms growing in laboratory water purification systems do not necessarily grow quickly, well, or at all on plate media.39 Cells that may not be capable of division when plated may be alive and capable of metabolic activity. UV treatment of water may damage replication mechanisms without causing death of a microorganism. Laboratories that adopt epifluorescence monitoring will need to establish thresholds of acceptable counts based on their laboratory requirements. It may not be possible to establish a consistent relationship between plate counts and epifluorescence counts, because plate counts are not necessarily a fixed fraction of the total concentration of planktonic microorganisms in a water system. 7.3.1 Method Overview: Total Direct Count Technique for Enumerating Bulk Water Bacteria A water sample is filtered though a membrane filter, which is then appropriately stained and examined by epifluorescence microscopy. Depending on the concentration of microorganisms, the volume of water is adjusted (typically from 1 to 10 mL) to avoid significant overlap of cells. Staining with DAPI (4, 6-Diamidino-2-phenylindole) does not distinguish viable and nonviable bacteria and yields a total count. Staining with a mixture of SYTO 9 dye and PI (propidium iodide) in DMSO (dimethylsulfoxide) will make it possible to distinguish between viable and nonviable bacteria with a high level of certainty.36-41 SYTO 9 is a green-fluorescent nucleic acid stain that penetrates the walls of microorganisms, whether they are healthy or unhealthy; whereas, red-fluorescent PI nucleic acid stain only penetrates bacteria with damaged cell wall membranes. Cells with intact membranes will appear green and cells with damaged membranes will appear red. 7.3.2 Equipment and Expendables The equipment needed to perform epifluorescence microscopy is relatively expensive, but is often available at university medical centers and microbiology laboratories. • epifluorescence microscope with UV light • DAPI, FITC (fluorescein-isothiocyanate) and RITC (rhodamine-isothiocyanate), or equivalent filters • digital or video imaging system (not required, but very helpful) • membrane filtration apparatus for 25-mm diameter membrane filters consisting of a vacuum filter

funnel manifold and vacuum filter holder/supports • vacuum pump • timer • filter forceps for handling membrane filters The expendable items required to enumerate microorganisms in laboratory water by epifluorescence microscopy are readily available. • black polycarbonate membrane filters (0.22 µm, 25 mm diameter) • stains

— DAPI solution – A solution of DAPI in CLRW (0.1 mg DAPI/mL CLRW) is filtered using a sterile 0.22 µm pore size PVDF syringe filter to eliminate large crystals or other particulates. Refrigerated, the solution can be stored in an amber bottle for a year or longer.

— SYTO 9/PI solution – mix 1 mL of filter-sterilized CLRW water with 15 µL of 3.34 mmol/L

SYTO 9 dye in DMSO, and 15 µL of 20 mM PI dye in DMSO. The 1:1 ratio of the dyes gives good results in most applications; however, the ratio can be modified to optimize results for specific circumstances. The DMSO solutions of the dyes should be kept frozen and the working solution should be used within a day and kept cool and in the dark.

Volume 25 C3-P4


SAFETY NOTE: Nucleic acid stains should be handled with particular caution and the user should bear in mind that DMSO is known to facilitate the entry of organic molecules into tissues. The stains should be disposed of by pouring them through charcoal, and the charcoal must then be incinerated to destroy the stains.

• glass microscope slides • glass cover slips • nonfluorescing immersion oil • 0.22 µm pore size PVDF syringe filters to eliminate large crystals or other particulates from stain

solutions and CLRW used for rinse water • 95% v/v ethanol 7.3.3 Procedure The following procedure is not intended to cover every detail and laboratories are advised to consult with vendors of equipment and supplies regarding usage. 7.3.3.1 Sample Preparation 1. Wet the stage of the filtration apparatus with filter-sterilized, CLRW-grade water and suction the

water through with the vacuum until all of the water is gone. 2. Place a polycarbonate membrane filter on the stage with the shiny side facing up. 3. Place a filter funnel, which has been rinsed first with 95% v/v ethanol and then with filter-sterilized,

CLRW-grade water, on the stage and clamp the filter into place. Ethanol helps disinfect the filter funnel and eliminate any residual particles that may interfere with microscopy; however, it must be rinsed away or it could have adverse effects on the bacteria and/or stains.

4. Obtain a freshly collected sample. If samples cannot be tested promptly, they can be refrigerated up to

24 hours. Vortex samples for at least 30 seconds to be certain the cells are well distributed. 5. Cells should not be clumped or layered on one another, so individual cells can be easily identified for

enumeration, so different volumes (e.g., 100 mL, 10 mL, and 1 mL) of the undiluted water sample might be filtered to avoid repeating the preparation in order to determine the optimum volume.

6. Slowly add the volume of water sample to the filter, being certain to keep the entire surface of the

filter covered and turn on the vacuum to suction the sample through the filter. 7. Turn off the vacuum and release any residual suction on the filtration apparatus by gently lifting up

the filter funnel. 8. The filter is ready to be stained for total direct counts or viability counts. 7.3.3.2 Staining Total counts are performed using DAPI stain and viable/nonviable counts are performed using the SYTO 9/PI mixed stain. NOTE: The SYTO 9/PI mixed stain will not produce good results with bacteria that have thick walls or membranes, such as Mycobacterium fortuitum. Comparing live/dead counts with total DAPI counts will establish whether the live/dead counts are missing a significant number of microorganisms.

Number 13 C3-P4


1. Slowly (drop wise) add staining solution to the filter until the filter is fully covered (approximately 0.25 mL/filter).

2. Cover the filter funnel with aluminum foil and incubate at ambient temperature for 15 to 20 minutes.

The stains degrade when exposed to UV light, so covering the filtration funnels with aluminum foil during staining will help prevent photobleaching.

3. Use a vacuum pump to suction the staining solution through the filter. 7.3.3.3 Enumerating Total counts are relatively unambiguous. However, viable/nonviable counts yield much more information, some of which is open to interpretation. Cells that cannot, or do not, replicate, but which are functioning in other respects, will be stained as viable. Cells with damaged walls, but which may be capable of recovery and replication, will be stained as nonviable. 1. Place a small drop of immersion oil on a slide and place the filter on the glass microscope slide, shiny

side up. Place another small drop of immersion oil on top of the filter and cover the filter with a cover slip. Blot away any excess oil.

2. Use a 100X oil-immersion lens on an epifluorescence microscope with UV light and an appropriate filter (DAPI, FITC, RITC, or equivalent) to count the total number of cells in 20 random fields.42 Ideally, images of the fields can be stored using digital or video technologies. Record the appropriate information from the microscope (e.g., magnification, size of grid, zoom). Calculate the number of cells per volume based on the dilution factor, volume filtered, filter funnel surface area, etc. Close the shutter of the microscope whenever possible, because exposing the filter to UV light will photobleach the stains to the point that cells are no longer distinguishable from the background within ~2 minutes.

7.3.4 Quality Control A quality control check for the epifluorescence process can be completed by filtering known concentrations of fluorescent microscopic beads onto a membrane filter and processing the filter as though it were a routine sample. Known solutions of live and killed microorganisms can be mixed quantitatively and processed as samples. 7.4 Endotoxins The majority of microorganisms present in water purification systems exist in biofilms and are not free floating in the effluent water. For this reason and because microbial cells can be killed, prevented from replicating, and filtered relatively easily, high-purity water may contain a greater mass of microbiological byproducts (e.g., toxins, metabolites, cell fragments) than of viable cells that can replicate and be detected by plate count or epifluorescence. Thus, endotoxin measurements are a useful adjunct to viable cell techniques to determine the levels of water contamination by cell wall debris from microorganisms. Endotoxins should be specifically tested when they may be an interfering substance in a measurement procedure. The Limulus amebocyte lysate, LAL reagent, is an enzyme activation cascade that can be initiated by both endotoxins (lipopolysaccharide), which react with cascade Factor C, and specific types of polysaccharide [(1,3)-beta-D-glucans], which react with cascade Factor G. Lipopolysaccharides are present in the cell walls of gram-negative bacteria. Beta-glucans are present in the cell walls of Eumycota (mushrooms, sac fungi, yeast, molds, etc.), certain bacteria, and green plants, including algae. Water samples that test positive with a glucan-sensitive endotoxin assay should be evaluated with a glucan-insensitive or glucan- specific assay. It is always important to rule out cross-reactions and interferences by performing standard

Volume 25 C3-P4


additions of known concentrations of endotoxin and/or a (1,3)-beta-D-glucan, depending on which entity is of interest. The Bacterial Endotoxins Test43 recognizes three major techniques to measure the reaction of an LAL test: 1) an increase in the turbidity of the reaction mixture with time (kinetic turbidimetric method); 2) an increase in intensity of color with time (kinetic or endpoint chromogenic method); and 3) presence or absence of gelation after a given incubation period (gel-clot method). Each method requires reagent formulated for the intended method. In purified water, the sizes and numbers of microorganisms tend to be small; therefore, the concentration of lipopolysaccharide, a function of the surface area of the cell, is relatively low. The most sensitive LAL method (kinetic turbidimetric) may detect the presence of fewer than 25 small cells (or their cell wall equivalents) per mL. Gel-clot testing is substantially less sensitive than kinetic turbidimetric testing but can be performed without specialized equipment. 7.4.1 Sample Collection The terms, “sterile” and “nonpyrogenic” or “pyrogen-free,” do not define container characteristics that ensure suitability for endotoxin sample collection or storage. Containers must be certified or tested to demonstrate that they are free of detectable endotoxin and interfering substances, consistent with the sensitivity of the endotoxin assay being used. Containers must also be certified or tested to demonstrate that they will not adsorb a significant amount of the endotoxin in a sample at the temperature(s) and over the period of time that the sample will be stored prior to analysis. In order to reasonably minimize surface effects, it is recommended that sample containers be filled and that samples be at least 25 mL, though far less sample is actually required for testing. Samples should be collected from points of use in the same manner as water is collected for normal laboratory use. Endotoxins are stable and can be transported to an off-site laboratory; however, assays should be performed promptly to avoid microbial growth in the sample. Refrigeration slows the growth of most microorganisms, and therefore, reduces the chance that the concentration of endotoxin will increase during storage. Freezing and thawing is believed to promote adsorption of endotoxins to container surfaces. 7.4.2 Equipment and Procedure Endotoxin tests are either performed as titrations to a gelation endpoint (gel-clot) or by measurement of optical density (absorbance). The tests are enzymatic assays, so instrumentation must provide a constant temperature to all reaction vessels during the period of incubation. The gel-clot method requires a water bath or dry-block incubator to maintain 37 °C for the required incubation period. The gels are physically unstable, so the incubation equipment must minimize disturbance to the reaction tubes (e.g., circulating water baths may cause too much vibration). Typically, gel-clot tests are sensitive to about 0.03 EU (endotoxin units)/mL; however, some reagents may be sensitive to as little as 0.015 EU/mL. When single-test vials (STVs) are used, water samples are added to small tubes containing premeasured, gel-clot test components, mixed, and incubated for a specified period of time. When multitest vials (MTVs) are used, water samples are added to reaction tubes and then reconstituted LAL components are added, the tube contents are mixed, and the tube is incubated for a specified period of time. A positive test is defined by presence of a clot that has formed in the bottom of a tube and that is sufficiently stable, so the tube can be inverted without causing the clot to dislodge. Any tube in which a solid gel has formed contains at least the concentration of endotoxin indicated by the reagent sensitivity. The concentration of endotoxin in a sample is measured by testing serial dilutions; as endotoxin is diluted to less than the sensitivity of the reagent, an endpoint, the last positive test in a series of dilutions, is reached. The dilution factor at the endpoint multiplied by the sensitivity of the reagent is

Number 13 C3-P4


the concentration of endotoxin in the undiluted sample. Gel-clot tests are standardized by testing a series of standard reference endotoxin concentrations that bracket the labeled sensitivity of the reagent. The endpoint observed in the standard series must be within the 50 to 200% error of the test for the test to be valid. Standard incubation time for this type of test is one hour. Tests based on reading absorbance use reagents formulated so the turbidity that develops during the gelation reaction or the intensity of color that develops on hydrolysis of an artificial substrate can be measured with a spectrophotometer. There are two methods of reading photometric tests—endpoint and kinetic. For endpoint methods, the absorbance is measured after a given period of incubation. The incubation period defines the sensitivity of the test; the shorter the incubation time, the less sensitive the test. Endpoint photometric tests are typically sensitive to 0.005 EU/mL. For kinetic methods, the rate of increase in absorbance is measured. Incubation proceeds until the data are sufficient to calculate the rate (or a value related to rate called a “time of onset”—the time it takes for onset of a constantly increasing rate after time of reagent addition). Both reading techniques use curves obtained by testing known concentrations of a standard reference endotoxin to compute endotoxin levels for unknown samples. Kinetic tests may be as sensitive as 0.001 EU/mL when read intervals are short and individual reactions precisely timed. Incubation times for endpoint and kinetic methods may range from 15 to 20 minutes for sensitivities around 0.25 EU/mL and 90 minutes for sensitivities as great as 0.001 EU/mL. The accepted error of photometric methods is typically 50 to 200%, the same as for gel-clot.43 Even though photometric methods allow calculation of a numeric value for the unknown concentration of endotoxin, the error is still broad because of inherent characteristics of LAL reactions. Therefore, it is very important to follow the instructions of the LAL manufacturer to obtain the best results from their reagents and instrumentation. Endpoint tests are usually done in 96-well, plastic plates that are incubated separately from the plate reader. Kinetic tests are usually done in 96-well, plastic plates, using an incubating plate reader, or in glass reaction tubes, using instruments that have been specialized for kinetic testing. In all cases, incubation temperatures must be uniform. Kinetic tests also require short read intervals (some instruments read all reaction vessels every ten seconds). Instruments used to make turbidity measurements must not disrupt the distribution of fine suspensions in the reaction vessel. 7.4.3 Standardization Companies that provide LAL test materials usually provide standardized endotoxin for calibration purposes. The activity of these secondary standards is determined by reference to the United States Pharmacopoeia Reference Standard, which is currently the same preparation as the Second International Standard of the World Health Organization preparation (94/580) and that of the FDA (EC-6). The recommendations for performing potency determinations have been published.44 One would expect to standardize a kinetic LAL assay with a range of dilutions; however, it is also important to standardize each lot of gel-clot reagents. There are two reasons for doing so: 1) to confirm that a lot has the expected sensitivity; and 2) to become familiar with the appearance of the reaction mixture and clot over a bracketing range of dilutions. 7.5 Determination of Oxidizable Organic Substances, Expressed as Total Organic Carbon (TOC) Organic substances are monitored by oxidizing them and detecting the resulting oxidation products. By convention, the measurement is expressed as total organic carbon (TOC). When a water sample contains an unknown mix of organic molecules, as is usually the case, TOC cannot be related to the concentration of organic molecules in the unknown water because the amount of carbon is different in different molecules. For example, 100 ng/g (ppb) of carbon is present in a solution of 131 ng/g (ppb) phenol or 990 ng/g (ppb) chloroform, because phenol contains 76% by weight of carbon and

Volume 25 C3-P4


chloroform contains 10% by weight of carbon. However, in practice, changes in the type or concentration of organic contaminants are likely to change values for TOC. It is important to initiate investigation of the water purification system in response to changes in TOC, even if the values are below the specification or alert levels. A number of different approaches can be used to determine TOC and each has merits and limitations, which are discussed in the following sections. 7.5.1 Sampling 7.5.1.1 Off-line Measurements made on water samples that are collected into containers and transported to an instrument are considered to be off-line. Off-line measurement is not recommended when the TOC of a water sample is lower than 50 ng/g (ppb), because the potential for contamination of the sample by organic and inorganic substances in ambient laboratory air, transfer systems, and containers will likely introduce significant error. Sample containers must be impermeable and free of organic substances. Narrow-mouth borosilicate glass containers are recommended. They should be carefully washed, rinsed, heated in an oven to 450 °C (below the strain point of the glass) with a clean air atmosphere for approximately one hour, and allowed to cool in the oven. This oven procedure is recommended, because it avoids a final rinse stage, which may introduce organic contaminants. Screw caps should be lined with metal foil or PTFE (polytetrafluoroethylene), and carefully cleaned. Ground glass stoppers can also be used. Alternatively, commercially available, precleaned, certified, low-TOC containers can be used. Containers should be validated as suitable for TOC determination by measuring blanks. Containers should be closed immediately after filling and their closures should be protected from contamination that could enter the sample when they are opened (e.g., by placing the containers in plastic bags). Certified, precleaned containers are available with a secondary cover for their closures. If samples will not be analyzed within 24 hours, they should be protected from light and stored in a refrigerator but not frozen. 7.5.1.2 On-line For on-line measurements, the instrument is connected directly to the purified water stream. On-line instruments permit continuous or semicontinuous determination of organic contamination. On-line measurement is recommended for TOC specification levels of <50 ng/g (ppb) and highly recommended when TOC specification levels are <20 ng/g (ppb).45

On-line measurements use a variety of sampling designs appropriate to the instrumentation, accuracy, and frequency of monitoring desired. The instrument can be connected to the main water stream by means of a connection designed to prevent contamination and to divert a sample portion of the stream for continuous or intermittent analysis. Depending on instrument design and other factors, the oxidized water sample can be directed to waste or recycled through an upstream stage of the water purification system to remove contaminants introduced by oxidation. Alternatively, a portion, or all, of the main water stream may pass through the measuring cell and continue on with the main stream, in which case there should be a subsequent repurification step to remove impurities introduced by oxidation. There must be verification that such a repurification step does not introduce significant organic contamination.

Number 13 C3-P4


7.5.2 Oxidation of Organic Molecules Purified water can contain a wide variety of organic compounds. Some organic molecules are oxidized more easily than others. Oxidation does not take place instantaneously, and the kinetics of oxidation depend on the nature of the organic material present and the oxidation conditions.46 The following techniques have been successfully used in various instruments for oxidation of organic species in purified water samples: • 185 nm and 254 nm UV light; • 185/254 nm or 300 to 400 nm UV light combined with a catalyst; • persulfate and 185/254 nm UV light at room temperature or 90 °C; • persulfate at 100 °C; • ozone; and • high temperature (680 to 1050 °C) catalytic oxidation. As a general rule, complete oxidation of organic molecules contributes to the accuracy of a TOC result. Some instruments oxidize for a fixed period of time that has been qualified to ensure that all organic molecules likely to be in the water sample are fully oxidized. Other instruments use dynamic techniques to determine when oxidation is complete by, for example, waiting until no more CO2 is produced or the resistivity of the sample reaches a constant level. Some instruments are designed to partially oxidize the organic contamination in a sample and are discussed further in Section 7.5.4. Laboratories must validate that the oxidation process used is effective for the purpose intended. The validation may be performed by the user or by the instrument manufacturer, and requires measuring the recovery of standard solutions containing molecules likely to be encountered in purified water, including molecules that are difficult to oxidize. For example, sucrose is frequently used as an easily oxidizable molecule, and p-benzoquinone as one that is more difficult to oxidize.47 7.5.3 Instruments That Determine TOC Instruments that determine TOC are designed to detect CO2 selectively; their detectors respond minimally, if at all, to the other products of organic oxidation. 7.5.3.1 Terminology • TC (total carbon) is the total concentration of carbon (organic and inorganic) in a sample. • TOC (total organic carbon) is the total concentration of carbon contained in organic molecules in a

sample. Elemental carbon is included as organic carbon. • TIC (total inorganic carbon) is the total concentration of carbon in the form of carbonate,

bicarbonate, and dissolved CO2 in a sample. • POC (purgeable organic carbon) is the concentration of carbon that escapes the sample in the gas

phase during the process of sparging the sample to remove inorganic carbon prior to measuring the organic carbon. POC depends on the instrument design and operating conditions as well as the nature of specific organic molecules present.

Volume 25 C3-P4


• NPOC (nonpurgeable organic carbon) is the concentration of organic carbon remaining after sparging a sample to remove inorganic carbon. NPOC depends on the instrument design and operating conditions as well as the nature of specific organic molecules present that can be lost as POC.

7.5.3.2 Nondispersive Infrared (NDIR) Detection NDIR detectors measure CO2 in a dry gas phase and are specific for CO2. TOC instruments that use NDIR detectors can be designed to make two measurements, TIC and TC. First, the water sample is acidified (pH ≤2) to convert TIC to CO2, which is sparged by carrier gas through the NDIR detector in a recirculating loop. Because NDIR detectors are specific for CO2 molecules, the concentration of TIC can be determined, even though other inorganic gases and POC may be present in the recirculating loop. After the TIC has been determined, the recirculating gas loop is kept closed, the sample is oxidized, and the TC is determined. There is no loss of POC, because both NPOC and the recirculated POC are oxidized. Therefore, subtracting TIC from TC will equal TOC. As the TIC concentration increases with respect to TC, the uncertainty of TOC calculations will increase, because two independent measurements, each with an associated error, are being subtracted. Traces of moisture and impurities in the gas phase, the purity of the acidifying and oxidizing reagents, and the relative volume of the gas phase can also affect the limits of detection and accuracy. Alternatively, inorganic carbon can be removed from a water sample by acidifying and sparging without determining the inorganic carbon released. However, this approach determines NPOC, rather than TOC. Sparging to remove inorganic carbon also removes POC, which can include short chain aliphatics, alcohols, ketones, esters, halomethanes, and aromatic compounds such as benzene, toluene, and cyclohexane. Measuring NPOC can fail to detect significant organic contamination from POC molecules. When water purification systems produce water with a resistivity approaching that of pure water, and measurements are made on-line to prevent CO2 absorption from air, TIC will be very low and, depending on the TOC limit required, may be negligible and the step of an analysis that is intended to calculate, or remove, TIC can be omitted. In this special case, there will be no loss of POC, and instruments that are not designed to remove and determine TIC will determine TOC. 7.5.3.3 Resistivity Detection Resistivity detectors cannot be specific for CO2 unless ions other than CO3

--, HCO3- and H+ are excluded.

Interposing a membrane that is selective for small gas molecules such as CO2 (e.g., Teflon®a AF, or similar) between a sample chamber and a resistivity cell can achieve the necessary selectivity. TOC instruments that use resistivity detectors combined with such membranes can be operated in much the same way as NDIR detectors (see Section 7.5.3.2). Some of these instruments split samples into two essentially matched paths to achieve greater stability. In one path, the sample is acidified to determine TIC, and in the other path the sample is oxidized and acidified to determine TC. The potential for error associated with increasing TIC is similar to that for instruments that use NDIR detectors. A contactor membrane may be used to reduce the impact of inorganic carbon and improve the accuracy of TOC determinations. As in the case of TOC instruments that use NDIR detectors, water with a resistivity approaching that of pure water will have a very low TIC and, depending on the TOC limit required, the step of an analysis that is intended to calculate, or remove, TIC can be omitted. 7.5.3.4 Calibration and Quality Control for TOC Instruments TOC instruments are calibrated with external standards traceable to a recognized standards organization,

a Teflon® is a registered trademark of DupontTM.

Number 13 C3-P4


preferably a national metrology institute (e.g., NIST in USA, IRMM in Belgium, and LGC in UK). Potassium hydrogen phthalate (KHP), sucrose, methanol, and p-benzoquinone are widely used. KHP is traceable for replaceable hydrogen content but not necessarily for carbon content. For example, p-benzoquinone is more difficult to oxidize than some other molecules and thus serves as a more meaningful standard to verify the completeness of oxidation. A TOC instrument should be calibrated at a sufficient number of points to establish a calibration relationship that will satisfy the requirements for accuracy. Quality control procedures may vary with the technology used to determine organic contamination. In general, external quality control samples should be interspersed with purified water samples to confirm that instruments are stable. CLSI/NCCLS document C24—Statistical Quality Control for Quantitative Measurements should be consulted for suitable techniques. If the controls indicate that an instrument’s calibration has drifted significantly, the instrument should be recalibrated. 7.5.4 Instruments That Are Not Carbon Specific The term TOCEquivalent is used in this document to refer to values determined by instruments that are not carbon specific. TOCEquivalent instruments measure the change in resistivity of the water sample before and after UV oxidation, without a CO2-selective interface. They are most accurate when used with purified water that has a high resistivity, so the resistivity change due to the oxidation products can be distinguished from the baseline resistivity. TOCEquivalent instruments can be calibrated to give values consistent with TOC values when known organic compounds containing only carbon, hydrogen, and oxygen atoms are analyzed. TOCEquivalent values can differ from TOC values because TOCEquivalent instruments respond to resistivity changes from all ions, not only ions related to CO2. Depending on the mix and concentrations of organic contaminants present in a water sample, different designs of TOCEquivalent instruments may give different values, and TOCEquivalent instruments may give values that are higher or lower than those from carbon-specific instruments.48 Values higher than a carbon-specific method are likely to be reported when organic contaminants contain atoms other than carbon, hydrogen, and oxygen that oxidize to form strong ions. A chlorinated organic molecule producing H+ and Cl- ions is an example.49 TOCEquivalent instruments will report lower values when water samples contain ionized species, for example, carboxylic acids, that are oxidized to more weakly ionized species. However, in the carboxylic acid example, the difference becomes less significant as the resistivity of the water approaches 18 MΩ • cm, referenced to 25 °C, because the total concentration of ionized substances will be very low. Partial oxidation, used in some TOCEquivalent instrument designs, can also lead to TOCEquivalent values that may be higher or lower than TOC values. The oxidation products from a particular water sample may produce changes in resistivity that are different from the corresponding changes in resistivity that were produced by the standards on which the calibration was based. When TOCEquivalent instruments are used to monitor and trend organic contamination, a laboratory should establish a TOCEquivalent alert level that is consistent with the fit-for-purpose validation of the water. The TOCEquivalent alert level should be set to accommodate the potential differences between TOC Equivalent and TOC determinations, and the TOCEquivalent values should be periodically re-evaluated by comparison with values obtained by carbon-specific TOC measurements. In the case of SRW with low organic contamination, a TOCEquivalent specification may be established based on fit-for-purpose validation for the application and the measurement system used.

Volume 25 C3-P4


7.5.4.1 Calibration and Quality Control for TOCEquivalent Instruments Calibration and quality control of TOCEquivalent instruments that are designed to permit the introduction of calibration or control solutions are analogous to those described in Section 7.5.3.4, except the organic substance(s) used as standards must not contain atoms other than carbon, hydrogen, and oxygen. In the case of instruments that do not have provision for introduction of calibration or control solutions, the manufacturer’s instructions for calibration and quality control of the instrument should be reviewed.

Number 13 C3-P4


References 1 ISO. Statistics – Vocabulary and Symbols – Part 1: Probability and General Statistical Terms. ISO 3534-1. Geneva: International

Organization for Standardization; 1993. 2 ISO. International Vocabulary of Basic and General Terms in Metrology. Geneva: International Organization for Standardization; 1993. 3 Costerton JW, Lewandowski Z, Caldwell DE, Korber DR, Lappin-Scott HM. Microbial biofilms. Ann Rev Microbiol. 1995;49:711-745. 4 ISO. Medical laboratories—Requirements for safety. ISO 15190. Geneve: International Organizaiton for Standardization; 2003. 5 ISO. Quality Management Systems – Fundamentals and vocabulary. ISO 9000. Geneva: International Organization for Standardization;

2000. 6 Federal Register: Part VII Dept. of Health and Human Services, FDA. 12 CFR Parts 808, 812, and 820. Rules and Regulations. October 7, 1996;61(195):52222-52601. 7 Levey S, Jennings ER. The use of control charts in the clinical laboratory. Am J Clin Pathol. 1950;20:1059-1066. 8 Shewhart WA. Economic control of quality of manufactured product. New York: Van Nostrand; 1931. 9 Meltzer TH. High Purity Water for the Semiconductor, Pharmaceutical, and Power Industries. Littleton, CO: Tall Oaks Publishing, Inc.;

1993;800. 10 Meltzer TH. Pharmaceutical Water Systems. Littleton, CO: Tall Oaks Publishing, Inc.; 1996. 11 Darbouret D, Kano I. Ultrapure water blank for boron trace analysis. J Anal At Spectrom. 2000;15:1395-1399. 12 Thate S, Specogna N, Eigenberger G. Electrodeionization - A comparison of different EDI concepts used for the production of high-purity

water. Ultrapure Water. 1999;16(8):42-56. 13 Salem E. Deionization - Areas to consider when selecting an EDI system. Ultrapure Water. 2000;(6):72-76. 14 Megonnell N, McClure A. Cabon and MTBE – What’s a dealer to do? Water Conditioning & Purification. 2000;72-75. 15 ASTM D3860-98(2003). Standard Practice for Determination of Adsorptive Capacity of Activated Carbon by Aqueous Phase Isotherm

Technique. ASTM International, www.astm.org. 2003. 16 Nowicki HG, Schuliger PE. Cabon software programs: sorbent performance evaluation ASTM aqueous phase isotherm program. Water

Conditioning & Purification. 2000;102-106. 17 MacLean RG. A new look at carbon: Biological purification of drinking water. Water Conditioning & Purification. 1999;56-59. 18 Camper AK, LeChevallier MW, Broadaway SC, McFetters GA. Growth and persistence of pathogens on granular activated carbon filters.

Applied Environmental Microbiology. 1985;50:1378. 19 Rollinger Y, Dott W. Survival of selected bacterial species in sterilized activated carbon filters and biological activated carbon filters.

Applied Environmental Microbiology. 1987;53:777. 20 Adley CC, Saieb F. Biofilm formation in high-purity water: Ralstonia pickettii a special case for analysis. Ultrapure Water. 2005;14-19. 21 Costerton JW, Stewart PS. Battling biofilms. Sci Am. 2001;285(1):74-81. 22 Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a common cause of persistent infections. Science. 1999;284(1):1318-1322. 23 Purevdorj-Gage LB, Stoodley P. Biofilm structure, behavior, and hydrodynamics. In: Ghannoum M, O’Toole GA, eds. Microbial Biofilms.

Washington, D.C.: ASM Press; 2004. 24 Rickard AH, McBain AJ, Stead AT, Gilbert P. Shear rate moderates community diversity in freshwater biofilms. AEM. 2004;70(12):7426-

7435. 25 Simoes M, Pereira MO, Vieira MJ. Action of a cationic surfactant on the activity and removal of bacterial biofilms formed under different

flow regimes. Water Research. 2005;39:478-486. 26 Traeger H. Microbial control: how to protect against biofilm build up in loops and tanks. Ultrapure Water. 2005;22;24-30. 27 MacInnes DA. The Principles of Electrochemistry. New York: Reinhold Publishing Corporation; 1939:Chap. 3. 28 Bevilacqua AC. Ultrapure Water - The Standard for Resistivity Measurements of Ultrapure Water. Proceedings of the 1998 Pure Water and Chemical Conference, Santa Clara, CA, USA, pp 141-172. 29 Carr G. Ultrapure Water. 2000:17-21.

Volume 25 C3-P4


30 Physical Properties of NALGENE Labware. Available at: http://nalgenelab.nalgenunc.com/techdata/resin/index2.asp?m=&p=Permeability +CO2+%28metric%29 and http://nalgenelab.nalgenunc.com/techdata/Resin/index2.asp?m=&p=Permeability+CO2. Accessed June 7, 2005. 31 Morash KR, Thornton RD, Saunders CH, Bevilacqua AC, Light TS. Measurement of the resistivity of high purity water at elevated

temperatures. Ultrapure Water. 1994;11(9):18-26. 32 Bevilacqua AC. The Effect of Temperature, Temperature Error, and Impurities on Compensated Conductivity Measurements. 16th Annual

Semiconductor Pure Water and Chemicals Conference, Santa Clara, California, March 3-6, 1997 (original data for Table 1 were provided by the author).

33 Light TS. Ultrapure Water. 1991;1-5. 34 Thornton RD, Light TS. A new approach to accurate resistivity measurement of high purity water. Ultrapure Water. 1989;14-21. 35 ASTM D5391, Vol. 11.01, April 2003. Standard Test Method for Electrical Conductivity and Resistivity of a Flowing High Purity Water

Sample. Appendix A2. 36 APHA, AWWA AND WEF. Standard Methods for the Examination of Water and Wastewater. 19th ed. Washington, D.C.: APHA;

1995:9216. 37 Kepner RL, Pratt JR. Use of fluorochromes for direct enumeration of total bacteria in environmental samples: past and present.

Microbiological Reviews. 1994;58:603-615. 38 McFeters G, Pyle B, Lisle J, Broadaway S. Rapid direct methods for the enumeration of specific, active bacteria in water and biofilms. J

Appl Microbiol. 1999;85:193-200. 39 McAlister MB, Kulakov LA, Larkin MJ, Ogden KL. Microbials - Analysis of bacterial contamination in different sections of a high-purity

water system. Ultrapure Water. 2001;18(1):18-26. 40 Boulos L, Prevost M, Barbeau B, Coallier J, Desjardins R. LIVE/DEAD BacLight: application of a new rapid staining method for direct

enumeration of viable and total bacteria in drinking water. J Microbiol Meth. 1999;37:77-86. 41 McFeters G, Yu FP, Pyle BH, Stewart PS. Physiological assessment of bacteria using fluorochromes. J Micro Meth. 1995;21:1-13. 42 AWWA, AWWA AND WEF. Standard Methods for the Examination of Water and Wastewater 19th ed. Washington, D.C.: APHA;

1995:9216. 43 USP28-NF23 S1. General Chapters <85> Bacterial Endotoxins Test - The United States Pharmacopoeial Convention Inc. 44 Food and Drug Administration publication, Guideline on validation of the Limulus Amebocyte Lysate test as an end product endotoxin test

for human and animal parenteral drugs, biological products and medical devices. December 1987.

45 Donovan RP, Painton Swiler LA, DeGenova J, Boswell T. Evaluating on-line TOC analyzers for high purity water recycle systems. Ultrapurewater. 1998;30-36.

46 Peyton GR. Marine Chemistry. 1993;41:91-103. 47 USP28-NF23 S1. General Chapters <643> Total Organic Carbon - The United States Pharmacopoeial Convention Inc. 48 Rydzewski J. Instruments - Identification of critical contaminants by applying an understanding of different TOC measuring technologies.

Ultrapure Water. 2002;19:20-27. 49 McCurdy L. Implementing TOC testing for USP 23 – A case study. Pharmaceutical Engineering. 1997;17:2-7.

Number 13 C3-P4


Additional References Byrne W. Encyclopedia of Water Treatment – Interactive CD 17 Volumes. Littleton, CO: Tall Oaks Publishing; 2004. 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. Standard Methods for the Examination of Water and Waste Water. 20th ed. Water Environment Federation, Alexandria, VA. ISBN: 0-87553-235-7 (1998).

Volume 25 C3-P4


Appendix A. Permeability of Plastic Piping to Atmospheric CO2 The following provides an example of gas permeability effects for diffusion of atmospheric CO2 through polypropylene on the resistivity of pure water (18.2 MΩ • cm, referenced to 25 °C). Static for ten minutes in 1 mm wall, 4 mm ID tubing 3.5 MΩ • cm, referenced to 25 °C Flow at 1 m/s through 1 mm wall, 4 mm ID tubing 18.0 MΩ • cm, referenced to 25 °C Static for one hour in 25 L storage vessel with 3 mm wall 16 MΩ • cm, referenced to 25 °C Static for 24 hours in 25 L storage vessel with 3 mm wall 4 MΩ • cm, referenced to 25 °C Taking for example polypropylene (PP), 6 m OD tubing 4 mm ID Permeability = 12 400 cm3 • mm • m-2 • (24 h)-1 • bar-1 Which means that 12 400 cm3 will flow through 1 m2 of 1 mm thick PP with 1 bar difference in CO2 pressure across it over a period of 24 hours. The wall of the tubing is 1 mm thick and the pressure outside is 1 bar. Taking 1 meter of tubing the outside surface area is 188.5 cm2 Gas flow/minute for 100% CO2 is 12 400 X 188.5 / (10 000 X 24 X 60) = 0.162 cm3 Carbon dioxide is 0.033% in the atmosphere, therefore the rate of gas flow will be: 0.162 X 0.033/100 = 5.3 X 10 –5 cm3/min assuming no pressure of CO2 in the water. Therefore, the maximum rate of flow is 5.3 X 10 –5 cm3/min. As CO2 enters the water, its partial pressure will rise and the transfer rate will slow. Assuming normal laboratory temperatures and converting this volume into mass using 22.4 L of CO2 weighs 44 g: Transfer rate = 105 ng CO2/min 1) Static situation The volume of water in the 1 m length of tubing is 12.6 cm3. Therefore, the rate of CO2 increase in the water is 105/12.6 = 8.3 µg/l (ng/g (ppb))/min This corresponds to a decrease in resistivity from 18.2 MΩ • cm to 3.5 MΩ • cm over a period of ten minutes. Similarly, for a reservoir of 25 L with a wall area of 0.6 m2 and a wall thickness of 3 mm, the resistivity will fall at a maximum rate to 16 MΩ • cm over a period of one hour and 4 MΩ • cm over 24 hours. For larger diameter pipe work and larger reservoirs, the effects will be proportionately less. 2) Dynamic flow situation a) Transient

Number 13 C3-P4


Appendix A. (Continued) Water passing through this length of tubing at a flow rate of 1 L/minute (i.e., 1.3 m/s) will be in the tubing for 1.26 s, in which time it will have the possibility of absorbing 1.75 ng of CO2. This will reduce the resistivity to 18.0 MΩ • cm, referenced to 25 °C.

Volume 25 C3-P4


Appendix B. Resistivity Measurement in a Sparged Water Sample Figure B1. Model for Argon Purge of Water in a Bottle NOTES: 1. V is the volume of water in the bottle (assumed constant). 2. Co, C(t), CAr(t) are the respective concentrations of CO2 at the locations noted in the figure. 3. Perfect mixing is assumed in V and in headspace. 4. f and fAr are flow rates at given locations, each assumed constant. 5. distribution coefficient: KH = C(t)/CAr(t), i.e., CAr(t) = C(t)/KH, (Henry’s Law) is assumed to hold at

all t. 6. The purge gas (Ar) is assumed to be CO2-free. 7. Rate of transfer of CO2 from gas phase to aqueous phase or vice versa is governed by physical

absorption only [i.e., any effect of the chemical kinetics of the reaction CO2(aq) ↔ H2CO3(aq) is neglected].

For the case where the removal of CO2 from solution or the absorption of CO2 into solution depends solely on physical absorption, i.e., transfer from the gas phase to aqueous solution or vice versa, the time-dependent concentration of CO2 in solution may be derived from the initial value problem (IVP) defined by the following first-order differential equation

VfC

=)t(CVKfK+f

+)t('C 0

H

HAr (B1)

with the boundary conditions for C(t) (general case):

∞=∞→= CtCCC )()0( 0 (B2) The general solution to this IVP is given by Equation 3:

( ) tVKfKfCCCtC ⎟⎟

⎠

⎞⎜⎜⎝

⎛ +−−+= ∞∞

H

HAr0 exp)( (B3)

V, C(t)Co, f C(t), f

CAr(t), fArCin = 0,flow fAr

Input water Output water

ArgonInput Output

headspace

Stirrer

V, C(t)Co, f C(t), f

CAr(t), fArCin = 0,flow fAr

Input water Output water

ArgonInput Output

headspace

Stirrer

Number 13 C3-P4


Appendix B. (Continued) For the case with boundary conditions

0HAr

H)(0)0( CfKf

fKtCC+

=∞→= (B4)

i.e., starting with CO2-free water in the bottle and introducing CO2-containing water with CCO2 = C0, Equation (3) becomes

⎥⎦

⎤⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛ +−−

+= t

VKfKf

fKffKCtC

H

HAr

HAr

H0 exp1)( (B5)

For the case with boundary conditions

0HAr

H0 )()0( C

fKffKtCCC+

=∞→= (B6)

i.e., for purging that starts with water in the bottle having CCO2 = C0, Equation (3) becomes

⎥⎦

⎤⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛ +−⎟⎟

⎠

⎞⎜⎜⎝

⎛+

++

= tVKfKf

fKff

fKffKCtC

H

HAr

HAr

Ar

HAr

H0 exp)( (B7)

Volume 25 C3-P4


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: 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.

1. 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 °C range, correct the resistivity reading to 25 °C 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 °C above and at 5 °C below the expected measurement temperature), then perform the following calculation: ρt is the resistivity reading at temperature t. ρ1 is the resistivity reading at temperature t1, 5 °C above temperature t. ρ2 is the resistivity reading at temperature t2, 5 °C below temperature t. ρavg is the arithmetic average of the resistivity readings at temperatures t1 and t2. ρ25 is the calculated, temperature-compensated resistivity at 25 °C, temperature t25. ∆t 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.

212

avgρρ

ρ+

= (C1)

avg21

12

)( ρ

ρρ

ttf

−

−= (C2)

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

( ) tt f ∆+= 125 ρρ (C3)

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

( ) tt f ∆−= 125 ρρ (C4)

2. In the temperature range of 20 °C to 30 °C, the temperature-compensated resistivity at 25 °C

calculated in this manner should be within 3% of the resistivity value, had it actually been measured at 25 °C. For larger temperature deviations, the Marsh-Stokes equation (or a similar curve fit) should be used instead of the linear (first-order) relationship assumed here.

3. The following temperature compensation equation is known as the Marsh-Stokes equation and illustrates the general approach used in many instruments. This equation converts any measured electrolytic conductivity reading in aqueous solution to its equivalent value at 25 °C. The equation assumes that the inherent (mobility-dependent) temperature coefficient of each ion is 2%/°C. NOTE: The Marsh-Stokes equation is phrased in terms of κ, not ρ.

Number 13 C3-P4


Appendix C. (Continued)

0.0549825)0.02(1

]325)(-7104.2690225)(-5106.062925)(-3102.9326[0.05498-25 +

+

•+•+•+=

-t

-t-t-ttκκ

where: κt is the electrolytic conductivity in µS/cm, at temperature t. κ25 is the electrolytic conductivity in µS/cm, at 25 °C. t is the temperature of measurement in °C. κ25 is the calculated, temperature-compensated resistivity at 25 °C. The Marsh-Stokes equation is based on the following compensation algorithm: a. Take the total measured electrolytic conductivity, κt, at the measurement temperature t. b. Subtract the conductivity of pure water at t from κt. The result (numerator of the first term of the

Marsh-Stokes Equation) is the conductivity attributable to impurity ionized species in the water. c. Correct the result from (b) to the conductivity attributable to impurity ionized species in the water at t

= 25 °C. This operation is performed by the denominator of the first term. d. Add κ for H2O at t = 25 °C to yield the corrected total conductivity at t = 25 °C, κ 25. State-of-the-art instrumentation may substitute a proprietary model in place of the Marsh-Stokes equation with a proprietary equation to model the κ - t relationship in the given instrument. Typically, the general approach is similar to the algorithm of the Marsh-Stokes equation, but the models for calculating the temperature-κ relationships (for pure water and for the impurity ion components) are more refined. Users should consult with the manufacturer or vendor to clarify the exact equation used and its validity range.

Volume 25 C3-P4


Summary of Comments and Subcommittee Responses C3-A3: Preparation and Testing of Reagent Water in the Clinical Laboratory; Approved Guideline—Third Edition Table 1 and Table 2 1. C3-A3 ranks carbon with an “E/G” rating in Table 1, where “E” is defined as, “capable of complete or near complete

removal”, and “G” is defined as, “capable of removing a large percentage.” In Table 2, the standard ranks “activated carbon” at the top of the list of means for removing organics from Type 1 reagent water. Both of these rankings are absolutely incorrect. If reverse osmosis deserves a “G” for the 98+% removal of dissolved ionized solids and a “G” for the removal of dissolved organics, which it does, the ranking for carbon in Table 1 should be “P.”

Appendix 4 explains that carbon is most effective for removing chlorine and that carbon beds shed fines, leach salts, promote the growth of microorganisms, and do not bind most organics efficiently. And there is no shortage of literature to support this position.

• Table 1, Table 2, and Appendix A4 have been removed from this edition. Section 6 provides an updated review of

various technologies, including activated carbon, used in water purification systems. Section 6.1.4 2. The C3-A3 requirement for the determination of silicates in Type 1 reagent water appears to be arbitrary and the statement,

“Silicates or colloidal silica can interfere with certain assays” appears to be incorrect. The standard is often cited as evidence of the existence of silica interferences and has been used to promote commercial products. Can CLSI cite any references in support of this statement?

If references for silica interferences in biomedical testing and research cannot be found, would CLSI remove the statement and consider dropping the requirement for silicate testing?

• Silicate testing is not included in this edition.

Clinical and Laboratory Standards Institute consensus procedures include an appeals process that is described in detail in Section 8 of the Administrative Procedures. For further information, contact CLSI or visit our website at www.clsi.org.

Number 13 C3-P4


The Quality System Approach Clinical and Laboratory Standards Institute (CLSI) subscribes to a quality system approach in the development of standards and guidelines, which facilitates project management; defines a document structure via a template; and provides a process to identify needed documents. The approach is based on the model presented in the most current edition of CLSI/NCCLS document HS1—A Quality Management System Model for Health Care. The quality system approach applies a core set of “quality system essentials” (QSEs), basic to any organization, to all operations in any healthcare service’s path of workflow (i.e., operational aspects that define how a particular product or service is provided). The QSEs provide the framework for delivery of any type of product or service, serving as a manager’s guide. The quality system essentials (QSEs) are: Documents & Records Equipment Information Management Process Improvement Organization Purchasing & Inventory Occurrence Management Service & Satisfaction Personnel Process Control Assessment Facilities & Safety C3-P4 addresses the quality system essentials (QSEs) indicated by an “X.” For a description of the other documents listed in the grid, please refer to the Related CLSI/NCCLS Publications section on the following page.

Doc

umen

ts

& R

ecor

ds

Org

aniz

atio

n

Pers

onne

l

Equi

pmen

t

Purc

hasi

ng &

In

vent

ory

Proc

ess

Con

trol

Info

rmat

ion

Man

agem

ent

Occ

urre

nce

Man

agem

ent

Ass

essm

ent

Proc

ess

Impr

ovem

ent

Serv

ice

&

Satis

fact

ion

Faci

litie

s &

Safe

ty

X

Adapted from CLSI/NCCLS document HS1—A Quality Management System Model for Health Care.

Volume 25 C3-P4


Related CLSI/NCCLS Publications* C24-A2 Statistical Quality Control for Quantitative Measurements: Principles and Definitions; Approved

Guideline—Second Edition (1999). This guideline provides definitions of analytical intervals; plans for quality control procedures; and guidance for quality control applications.

EP7-A Interference Testing in Clinical Chemistry; Approved Guideline (2002). This guideline provides

background information guidance, and experimental procedures for investigating, identifying, and characterizing the effects of interfering substances on clinical chemistry test results.

GP2-A4 Clinical Laboratory Technical Procedure Manuals; Approved Guideline—Fourth Edition (2002). GP2-

A4 addresses the design, preparation, maintenance, and use of technical procedure manuals in the clinical laboratory.

* Proposed-level documents are being advanced through the Clinical and Laboratory Standards Institute consensus process; therefore, readers should refer to the most recent editions.

Active Membership (as of 1 April 2005)

Sustaining Members Abbott Laboratories American Association for Clinical Chemistry Bayer Corporation BD Beckman Coulter, Inc. bioMérieux, Inc. CLMA College of American Pathologists GlaxoSmithKline Ortho-Clinical Diagnostics, Inc. Pfizer Inc Roche Diagnostics, Inc. Professional Members American Academy of Family Physicians American Association for Clinical Chemistry American Association for Respiratory Care American Chemical Society American Medical Technologists American Society for Clinical Laboratory Science American Society for Microbiology American Society of Hematology American Type Culture Collection, Inc. Asociacion Mexicana de Bioquimica Clinica A.C. Assn. of Public Health Laboratories Assoc. Micro. Clinici Italiani- A.M.C.L.I. British Society for Antimicrobial Chemotherapy Canadian Society for Medical Laboratory Science - Société Canadienne de Science de Laboratoire Médical Canadian Standards Association Clinical Laboratory Management Association COLA College of American Pathologists College of Medical Laboratory Technologists of Ontario College of Physicians and Surgeons of Saskatchewan ESCMID International Council for Standardization in Haematology International Federation of Biomedical Laboratory Science International Federation of Clinical Chemistry Italian Society of Clinical Biochemistry and Clinical Molecular Biology Japan Society of Clinical Chemistry Japanese Committee for Clinical Laboratory Standards Joint Commission on Accreditation of Healthcare Organizations National Academy of Clinical Biochemistry National Association of Testing Authorities - Australia National Society for Histotechnology, Inc. New Zealand Association of Phlebotomy Ontario Medical Association Quality Management Program-Laboratory Service RCPA Quality Assurance Programs PTY Limited Sociedad Espanola de Bioquimica Clinica y Patologia Molecular Sociedade Brasileira de Analises Clinicas Taiwanese Committee for Clinical Laboratory Standards (TCCLS) Turkish Society of Microbiology Government Members Armed Forces Institute of Pathology Association of Public Health Laboratories BC Centre for Disease Control Caribbean Epidemiology Centre Centers for Disease Control and Prevention Centers for Medicare & Medicaid Services Centers for Medicare & Medicaid Services/CLIA Program

Chinese Committee for Clinical Laboratory Standards Commonwealth of Pennsylvania Bureau of Laboratories Department of Veterans Affairs Deutsches Institut für Normung (DIN) FDA Center for Devices and Radiological Health FDA Center for Veterinary Medicine FDA Division of Anti-Infective Drug Products Iowa State Hygienic Laboratory Massachusetts Department of Public Health Laboratories National Center of Infectious and Parasitic Diseases (Bulgaria) National Health Laboratory Service (South Africa) National Institute of Standards and Technology National Pathology Accreditation Advisory Council (Australia) New York State Department of Health Ontario Ministry of Health Pennsylvania Dept. of Health Saskatchewan Health-Provincial Laboratory Scientific Institute of Public Health; Belgium Ministry of Social Affairs, Public Health and the Environment Industry Members AB Biodisk Abbott Diabetes Care Abbott Laboratories Acrometrix Corporation Advancis Pharmaceutical Corporation Affymetrix, Inc. Ammirati Regulatory Consulting Anna Longwell, PC A/S ROSCO AstraZeneca Pharmaceuticals Aventis Axis-Shield POC AS Bayer Corporation - Elkhart, IN Bayer Corporation - Tarrytown, NY Bayer Corporation - West Haven, CT BD BD Diabetes Care BD Diagnostic Systems BD VACUTAINER Systems Beckman Coulter, Inc. Beckman Coulter K.K. (Japan) Bio-Development SRL Bio-Inova Life Sciences International Biomedia Laboratories SDN BHD bioMérieux, Inc. (MO) Biometrology Consultants Bio-Rad Laboratories, Inc. Bio-Rad Laboratories, Inc. - France Bio-Rad Laboratories, Inc. - Plano, TX Blaine Healthcare Associates, Inc. Bristol-Myers Squibb Company Canadian External Quality Assessment Laboratory Cepheid Chen & Chen, LLC Chiron Corporation ChromaVision Medical Systems, Inc. Clinical Micro Sensors The Clinical Microbiology Institute Cognigen CONOSCO Copan Diagnostics Inc. Cosmetic Ingredient Review Cubist Pharmaceuticals Dade Behring Inc. - Cupertino, CA Dade Behring Inc. - Deerfield, IL Dade Behring Inc. - Glasgow, DE Dade Behring Inc. - Marburg, Germany Dade Behring Inc. - Sacramento, CA David G. Rhoads Associates, Inc. Diagnostic Products Corporation Digene Corporation Eiken Chemical Company, Ltd. Elanco Animal Health Electa Lab s.r.l. Enterprise Analysis Corporation F. Hoffman-La Roche AG Gen-Probe GlaxoSmithKline Greiner Bio-One Inc. Immunicon Corporation

ImmunoSite, Inc. Instrumentation Laboratory International Technidyne Corporation I-STAT Corporation Johnson and Johnson Pharmaceutical Research and Development, L.L.C. K.C.J. Enterprises LabNow, Inc. LifeScan, Inc. (a Johnson & Johnson Company) Machaon Diagnostics Medical Device Consultants, Inc. Merck & Company, Inc. Micromyx, LLC Minigrip/Zip-Pak Nanosphere, Inc. National Pathology Accreditation Advisory Council (Australia) Nippon Becton Dickinson Co., Ltd. Nissui Pharmaceutical Co., Ltd. Novartis Pharmaceuticals Corporation Olympus America, Inc. Optimer Pharmaceuticals, Inc. Ortho-Clinical Diagnostics, Inc. (Rochester, NY) Ortho-McNeil Pharmaceutical (Raritan, NJ) Oxoid Inc. Paratek Pharmaceuticals Pfizer Animal Health Pfizer Inc Pfizer Italia Srl Powers Consulting Services Predicant Biosciences Procter & Gamble Pharmaceuticals, Inc. QSE Consulting Radiometer America, Inc. Radiometer Medical A/S Replidyne Roche Diagnostics GmbH Roche Diagnostics, Inc. Roche Diagnostics Shanghai Ltd. Roche Laboratories (Div. Hoffmann- La Roche Inc.) Sanofi Pasteur Sarstedt, Inc. Schering Corporation Schleicher & Schuell, Inc. SFBC Anapharm Streck Laboratories, Inc. SYN X Pharma Inc. Sysmex Corporation (Japan) Sysmex Corporation (Long Grove, IL) TheraDoc Theravance Inc. Thrombodyne, Inc. THYMED GmbH Transasia Engineers Trek Diagnostic Systems, Inc. Vetoquinol S.A. Vicuron Pharmaceuticals Inc. Vysis, Inc. Wyeth Research XDX, Inc. YD Consultant YD Diagnostics (Seoul, Korea) Trade Associations AdvaMed Japan Association of Clinical Reagents Industries (Tokyo, Japan) Associate Active Members 82 MDG/SGSCL (Sheppard AFB,TX) Academisch Ziekenhuis -VUB (Belgium) ACL Laboratories (WI) All Children’s Hospital (FL) Allegheny General Hospital (PA) Allina Health System (MN) American University of Beirut Medical Center (NY) Anne Arundel Medical Center (MD) Antwerp University Hospital (Belgium) Arkansas Department of Health ARUP at University Hospital (UT) Associated Regional & University Pathologists (UT) Atlantic Health System (NJ) AZ Sint-Jan (Belgium) Azienda Ospedale Di Lecco (Italy) Barnes-Jewish Hospital (MO) Baxter Regional Medical Center (AR)

Baystate Medical Center (MA) Bbaguas Duzen Laboratories (Turkey) BC Biomedical Laboratories (Surrey, BC, Canada) Bermuda Hospitals Board Bo Ali Hospital (Iran) Bon Secours Hospital (Ireland) Brazosport Memorial Hospital (TX) Broward General Medical Center (FL) Cadham Provincial Laboratory (Winnipeg, MB, Canada) Calgary Laboratory Services (Calgary, AB, Canada) California Pacific Medical Center Cambridge Memorial Hospital (Cambridge, ON, Canada) Canterbury Health Laboratories (New Zealand) Cape Breton Healthcare Complex (Nova Scotia, Canada) Carilion Consolidated Laboratory (VA) Carolinas Medical Center (NC) Cathay General Hospital (Taiwan) Central Laboratory for Veterinarians (BC, Canada) Central Ohio Primary Care Physicians Central Texas Veterans Health Care System Centro Diagnostico Italiano (Milano, Italy) Chang Gung Memorial Hospital (Taiwan) Changi General Hospital (Singapore) Children’s Hospital (NE) Children’s Hospital Central California Children’s Hospital & Clinics (MN) Children’s Hospital Medical Center (Akron, OH) Children’s Medical Center of Dallas (TX) Chinese Association of Advanced Blood Bankers (Beijing) CHR St. Joseph Warquignies (Belgium) City of Hope National Medical Center (CA) Clarian Health - Methodist Hospital (IN) CLSI Laboratories (PA) Community Hospital of Lancaster (PA) Community Hospital of the Monterey Peninsula (CA) CompuNet Clinical Laboratories (OH) Covance Central Laboratory Services (IN) Creighton University Medical Center (NE) Detroit Health Department (MI) DFS/CLIA Certification (NC) Diagnostic Accreditation Program (Vancouver, BC, Canada) Diagnósticos da América S/A (Brazil) Dianon Systems (OK) Dr. Everett Chalmers Hospital (New Brunswick, Canada) Duke University Medical Center (NC) Dwight David Eisenhower Army Medical Center (GA) Eastern Health Pathology (Australia) Emory University Hospital (GA) Enzo Clinical Labs (NY) Evangelical Community Hospital (PA) Fairview-University Medical Center (MN) Florida Hospital East Orlando Focus Technologies (CA) Focus Technologies (VA) Foothills Hospital (Calgary, AB, Canada) Franciscan Shared Laboratory (WI) Fresno Community Hospital and Medical Center Gamma Dynacare Medical Laboratories (Ontario, Canada) Gateway Medical Center (TN) Geisinger Medical Center (PA) Guthrie Clinic Laboratories (PA) Hagerstown Medical Laboratory (MD) Harris Methodist Fort Worth (TX) Hartford Hospital (CT) Headwaters Health Authority (Alberta, Canada)

Health Network Lab (PA) Highlands Regional Medical Center (FL) Hoag Memorial Hospital Presbyterian (CA) Holy Cross Hospital (MD) Hôpital Maisonneuve - Rosemont (Montreal, Canada) Hôpital Saint-Luc (Montreal, Quebec, Canada) Hospital Consolidated Laboratories (MI) Hospital de Sousa Martins (Portugal) Hospital for Sick Children (Toronto, ON, Canada) Hotel Dieu Grace Hospital (Windsor, ON, Canada) Huddinge University Hospital (Sweden) Humility of Mary Health Partners (OH) Hunter Area Health Service (Australia) Hunterdon Medical Center (NJ) Indiana University Innova Fairfax Hospital (VA) Institute of Medical and Veterinary Science (Australia) International Health Management Associates, Inc. (IL) Jackson Health System (FL) Jacobi Medical Center (NY) John H. Stroger, Jr. Hospital of Cook County (IL) Johns Hopkins Medical Institutions (MD) Kadlec Medical Center (WA) Kaiser Permanente (MD) Kantonsspital (Switzerland) Kimball Medical Center (NJ) King Abdulaziz Medical City - Jeddah (Saudi Arabia) King Faisal Specialist Hospital (Saudi Arabia) LabCorp (NC) Laboratoire de Santé Publique du Quebec (Canada) Laboratorio Dr. Echevarne (Spain) Laboratório Fleury S/C Ltda. (Brazil) Laboratorio Manlab (Argentina) Laboratory Corporation of America (NJ) Lakeland Regional Medical Center (FL) Landstuhl Regional Medical Center (APO AE) Lawrence General Hospital (MA) Lewis-Gale Medical Center (VA) L'Hotel-Dieu de Quebec (Canada) Libero Instituto Univ. Campus BioMedico (Italy) Lindy Boggs Medical Center (LA) Loma Linda Mercantile (CA) Long Beach Memorial Medical Center (CA) Los Angeles County Public Health Lab (CA) Lourdes Hospital (KY) Maimonides Medical Center (NY)

Marion County Health Department (IN) Martin Luther King/Drew Medical Center (CA) Massachusetts General Hospital (Microbiology Laboratory) MDS Metro Laboratory Services (Burnaby, BC, Canada) Medical College of Virginia Hospital Medical Research Laboratories International (KY) Medical University of South Carolina Memorial Medical Center (Napoleon Avenue, New Orleans, LA) Methodist Hospital (Houston, TX) Methodist Hospital (San Antonio, TX) Mid America Clinical Laboratories, LLC (IN) Middlesex Hospital (CT) Montreal Children’s Hospital (Canada) Montreal General Hospital (Canada) National Serology Reference Laboratory (Australia) NB Department of Health & Wellness (New Brunswick, Canada) The Nebraska Medical Center Nevada Cancer Institute New Britain General Hospital (CT) New England Fertility Institute (CT) New York City Department of Health & Mental Hygiene NorDx (ME) North Carolina State Laboratory of Public Health North Central Medical Center (TX) North Shore - Long Island Jewish Health System Laboratories (NY) North Shore University Hospital (NY) Northwestern Memorial Hospital (IL) Ochsner Clinic Foundation (LA) Onze Lieve Vrouw Ziekenhuis (Belgium) Orlando Regional Healthcare System (FL) Ospedali Riuniti (Italy) The Ottawa Hospital (Ottawa, ON, Canada) Our Lady of the Resurrection Medical Center (IL) Pathology and Cytology Laboratories, Inc. (KY) Pathology Associates Medical Laboratories (WA) The Permanente Medical Group (CA) Phoenix College (AZ) Piedmont Hospital (GA) Pocono Medical Center (PA) Presbyterian Hospital of Dallas (TX) Providence Health Care (Vancouver, BC, Canada) Provincial Laboratory for Public Health (Edmonton, AB, Canada)

Quest Diagnostics Incorporated (CA) Quintiles Laboratories, Ltd. (GA) Regional Health Authority Four (NB, Canada) Regions Hospital Rex Healthcare (NC) Rhode Island Department of Health Laboratories Riverside Medical Center (IL) Robert Wood Johnson University Hospital (NJ) Sahlgrenska Universitetssjukhuset (Sweden) St. Alexius Medical Center (ND) St. Anthony Hospital (CO) St. Anthony’s Hospital (FL) St. Barnabas Medical Center (NJ) St. Christopher’s Hospital for Children (PA) St-Eustache Hospital (Quebec, Canada) St. John Hospital and Medical Center (MI) St. John Regional Hospital (St. John, NB, Canada) St. John’s Hospital & Health Center (CA) St. Joseph’s Hospital - Marshfield Clinic (WI) St. Jude Children’s Research Hospital (TN) St. Mary Medical Center (CA) St. Mary of the Plains Hospital (TX) St. Michael’s Hospital (Toronto, ON, Canada) St. Vincent’s University Hospital (Ireland) Ste. Justine Hospital (Montreal, PQ, Canada) Salem Clinic (OR) San Francisco General Hospital (CA) Santa Clara Valley Medical Center (CA) Seoul Nat’l University Hospital (Korea) Shands at the University of Florida South Bend Medical Foundation (IN) South Western Area Pathology Service (Australia) Southern Maine Medical Center Spartanburg Regional Medical Center (SC) Specialty Laboratories, Inc. (CA) State of Connecticut Dept. of Public Health State of Washington Department of Health Stony Brook University Hospital (NY) Stormont-Vail Regional Medical Center (KS) Sun Health-Boswell Hospital (AZ) Sunnybrook Health Science Center (ON, Canada) Sunrise Hospital and Medical Center (NV)

Swedish Medical Center - Providence Campus (WA) Temple University Hospital (PA) Tenet Odessa Regional Hospital (TX) Touro Infirmary (LA) Tripler Army Medical Center (HI) Truman Medical Center (MO) Tuen Mun Hospital (Hong Kong) UCLA Medical Center (CA) UCSF Medical Center (CA) UNC Hospitals (NC) Unidad de Patologia Clinica (Mexico) Union Clinical Laboratory (Taiwan) United Laboratories Company (Kuwait) Universita Campus Bio-Medico (Italy) University College Hospital (Galway, Ireland) University of Chicago Hospitals (IL) University of Colorado Hospital University of Debrecen Medical Health and Science Center (Hungary) University of Illinois Medical Center University of Maryland Medical System University of Medicine & Dentistry, NJ University Hospital University of the Ryukyus (Japan) University of Wisconsin Hospital The University of the West Indies University of Virginia Medical Center University of Washington US LABS, Inc. (CA) USA MEDDAC-AK UZ-KUL Medical Center (Belgium) VA (Tuskegee) Medical Center (AL) Virginia Beach General Hospital (VA) Virginia Department of Health Washington Adventist Hospital (MD) Washoe Medical Center Laboratory (NV) Waterford Regional Hospital (Ireland) Wellstar Health Systems (GA) West China Second University Hospital, Sichuan University (P.R. China) West Jefferson Medical Center (LA) Wilford Hall Medical Center (TX) William Beaumont Army Medical Center (TX) William Beaumont Hospital (MI) Winn Army Community Hospital (GA) Winnipeg Regional Health Authority (Winnipeg, Canada) Wishard Memorial Hospital (IN) York Hospital (PA)

OFFICERS

BOARD OF DIRECTORS

Thomas L. Hearn, PhD, President Centers for Disease Control and Prevention Robert L. Habig, PhD, President Elect Abbott Laboratories Wayne Brinster, Secretary BD Gerald A. Hoeltge, MD, Treasurer The Cleveland Clinic Foundation Donna M. Meyer, PhD, Immediate Past President CHRISTUS Health Glen Fine, MS, MBA, Executive Vice President

Susan Blonshine, RRT, RPFT, FAARC TechEd Maria Carballo Health Canada Kurt H. Davis, FCSMLS, CAE Canadian Society for Medical Laboratory Science Russel K. Enns, PhD Cepheid Mary Lou Gantzer, PhD Dade Behring Inc. Lillian J. Gill, DPA FDA Center for Devices and Radiological Health J. Stephen Kroger, MD, MACP COLA

Jeannie Miller, RN, MPH Centers for Medicare & Medicaid Services Gary L. Myers, PhD Centers for Disease Control and Prevention Klaus E. Stinshoff, Dr.rer.nat. Digene (Switzerland) Sàrl James A. Thomas ASTM International Kiyoaki Watanabe, MD Keio University School of Medicine

940 West Valley Road Suite 1400 Wayne, PA 19087 USA PHONE 610.688.0100 FAX 610.688.0700 E-MAIL: [email protected] WEBSITE: www.clsi.org ISBN 1-56238-570-4