Professional Documents
Culture Documents
Vol. 17 No. 18
Replaces C3-A2
October 1997 Vol. 11 No. 13
This document provides guidelines on water purified for clinical laboratory use; methods for monitoring
water quality and testing for specific contaminants; and water system design considerations.
ABC
NCCLS...
Serving the World's Medical Science Community Through Voluntary Consensus
NCCLS is an international, interdisciplinary, nonprofit, Proposed An NCCLS consensus document undergoes the
standards-developing and educational organization that first stage of review by the healthcare community as a
promotes the development and use of voluntary consensus proposed standard or guideline. The document should receive
standards and guidelines within the healthcare community. It a wide and thorough technical review, including an overall
is recognized wordwide for the application of its unique review of its scope, approach, and utility, and a line-by-line
consensus process in the development of standards and review of its technical and editorial content.
guidelines for patient testing and related healthcare issues.
NCCLS is based on the principle that consensus is an Tentative A tentative standard or guideline is made
effective and cost-effective way to improve patient testing available for review and comment only when a recommended
and healthcare services. method has a well-defined need for a field evaluation or
when a recommended protocol requires that specific data be
In addition to developing and promoting the use of voluntary collected. It should be reviewed to ensure its utility.
consensus standards and guidelines, NCCLS provides an
open and unbiased forum to address critical issues affecting Approved An approved standard or guideline has achieved
the quality of patient testing and health care. 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
PUBLICATIONS earlier versions have been satisfactorily addressed), and to
identify the need for additional consensus documents.
An NCCLS document is published as a standard, guideline,
or committee report. NCCLS standards and guidelines represent a consensus
opinion on good practices and reflect the substantial
Standard A document developed through the consensus agreement by materially affected, competent, and interested
process that clearly identifies specific, essential requirements parties obtained by following NCCLS’s established
for materials, methods, or practices for use in an unmodified consensus procedures. Provisions in NCCLS standards and
form. A standard may, in addition, contain discretionary guidelines may be more or less stringent than applicable
elements, which are clearly identified. regulations. Consequently, conformance to this voluntary
consensus document does not relieve the user of
Guideline A document developed through the consensus responsibility for compliance with applicable regulations.
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. COMMENTS
Report A document that has not been subjected to con- The comments of users are essential to the consensus
sensus review and is released by the Board of Directors. process. Anyone may submit a comment, and all comments
are addressed, according to the consensus process, by the
NCCLS committee that wrote the document. All comments,
CONSENSUS PROCESS including those that result in a change to the document
when published at the next consensus level and those that
The NCCLS voluntary consensus process is a protocol do not result in a change, are responded to by the committee
establishing formal criteria for: in an appendix to the document. Readers are strongly
encouraged to comment in any form and at any time on any
! The authorization of a project NCCLS document. Address comments to the NCCLS
Executive Offices, 940 West Valley Road, Suite 1400,
! The development and open review of documents Wayne, PA 19087, USA.
Abstract
Preparation and Testing of Reagent Water in the Clinical Laboratory; Approved Guideline — Third
Edition (NCCLS document C3-A3) provides basic information about different methods of water
purification so that laboratorians can determine which purification system(s) produces water that
best suits their specific needs. The guideline addresses issues of system design, process
specifications, storage and handling considerations, and appropriate testing methods.
[NCCLS. Preparation and Testing of Reagent Water in the Clinical Laboratory; Approved Guideline
— Third Edition. NCCLS document C3-A3 (ISBN 1-56238-336-1). NCCLS, 940 West Valley Road,
Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 1997.]
THE NCCLS consensus process, which is the mechanism for moving a document through two
or more levels of review by the healthcare community, is an ongoing process. Users should
expect revised editions of any given document. Because rapid changes in technology may
affect the procedures, methods, and protocols in a standard or guideline, users should replace
outdated editions with the current editions of NCCLS documents. Current editions are listed in
the NCCLS Catalog, which is distributed to member organizations, and to nonmembers on
request. If your organization is not a member and would like to become one, and to request a
copy of the NCCLS Catalog, contact the NCCLS Executive Offices. Telephone:
610.688.0100; Fax: 610.688.0700; E-Mail: exoffice@nccls.org.
Volume 17 Number 18
David M. Jeffers
James H. Carter, Ph.D.
Gary A. Graham, Ph.D.
Anita K. Highsmith
Mona D. Jensen, Ph.D.
Richard R. Miller, Jr.
Edward A. Sasse, Ph.D.
Bette Seamonds, Ph.D.
ABC
October 1997 C3-A3
NCCLS hereby grants permission to reproduce limited portions of this publication for use in
laboratory procedure manuals at a single site, for interlibrary loan, or for use in educational
programs provided that multiple copies of such reproduction shall include the following notice, be
distributed without charge, and, in no event, contain more than 20% of the document's text.
Reproduced with permission, from NCCLS publication C3-A3 — Preparation and Testing of
Reagent Water in the Clinical Laboratory; Approved Guideline — Third Edition. Copies of
the current edition may be obtained from NCCLS, 940 West Valley Road, Suite 1400,
Wayne, Pennsylvania 19087-1898 USA.
Permission to reproduce or otherwise use the text of this document to an extent that exceeds the
exemptions granted here or under the Copyright Law must be obtained from NCCLS by written
request. To request such permission, address inquiries to the Executive Director, NCCLS, 940
West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA.
Suggested Citation
[NCCLS. Preparation and Testing of Reagent Water in the Clinical Laboratory; Approved Guideline
— Third Edition. NCCLS document C3-A3 (ISBN 1-56238-336-1). NCCLS, 940 West Valley Road,
Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 1997.]
ISBN: 1-56238-336-1
ISSN: 0273-3099
Committee Membership
Acknowledgments
The Working Group on Reagent Water acknowledges the participation of the following persons in
the review of Preparation and Testing of Reagent Water in the Clinical LaboratoryApproved
Guideline — Third Edition:
Associate Active Members Harris Methodist Fort Worth North Carolina Laboratory of
(TX) Public Health
Affinity Health System (WI) Hartford Hospital (CT) North Shore University Hospital
Allegheny University of the Health Alliance Laboratory (NY)
Health Sciences (PA) Services (OH) Olin E. Teague Medical Center
Alton Ochsner Medical Heritage Hospital (MI) (TX)
Foundation (LA) Hopital Saint Pierre (Belgium) Omni Laboratory (MI)
American Oncologic Hospital Incstar Corporation (MN) Our Lady of Lourdes Hospital
(PA) Integris Baptist Medical Center (NJ)
Anzac House (Australia) of Oklahoma Our Lady of the Resurrection
Associated Regional & International Health Managment Medical Center (IL)
University Pathologists (UT) Associates, Inc. (IL) PAPP Clinic P.A. (GA)
Baptist Memorial Healthcare Kaiser Permanente (CA) Pathology Associates
System (TX) Kangnam St. Mary’s Hospital Laboratories (CA)
Beth Israel Medical Center (NY) (Korea) Permanente Medical Group (CA)
Brazosport Memorial Hospital Kenora-Rainy River Regional PLIVA d.d. Research Institute
(TX) Laboratory Program (Dryden, (Croatia)
Bristol Regional Medical Center Ontario, Canada) Polly Ryon Memorial Hospital
(TN) Klinisches Institute für (TX)
Brooke Army Medical Center Medizinische (Austria) Providence Medical Center (WA)
(TX) Laboratoire de Santé Publique Puckett Laboratories (MS)
Brooks Air Force Base (TX) du Quebec (Canada) Quest Diagnostics (MI)
Broward General Medical Center Laboratory Corporation of Quest Diagnostics (PA)
(FL) America (NC) Reid Hosptial & Health Care
Canterbury Health Laboratories Laboratory Corporation of Services (IN)
(New Zealand) America (NJ) Sacred Heart Hospital (MD)
Central Peninsula General Lancaster General Hospital (PA) St. Boniface General Hospital
Hospital (AK) Libero Instituto Univ. Campus (Winnipeg, Canada)
Childrens Hospital Los Angeles Biomedico (Italy) St. Francis Medical Center (CA)
(CA) Louisiana State University St. John Regional Hospital (St.
Children's Hospital Medical Medical Center John, NB, Canada)
Center (Akron, OH) Maimonides Medical Center St. Joseph’s Hospital -
Clendo Lab (Puerto Rico) (NY) Marshfield Clinic (WI)
Clinical Diagnostic Services (NJ) Maine Medical Center St. Luke’s Hospital (PA)
Commonwealth of Kentucky Massachusetts General Hospital St. Luke’s Regional Medical
CompuNet Clinical Laboratories MDS-Sciex (Concord, ON, Center (IA)
(OH) Canada) St. Luke’s-Roosevelt Hospital
Consolidated Laboratory Melbourne Pathology (Australia) Center (NY)
Services (CA) Memorial Medical Center (LA) St. Mary of the Plains Hospital
Consultants Laboratory (WI) Methodist Hospital (TX) (TX)
Detroit Health Department (MI) Methodist Hospital Indiana St. Paul Ramsey Medical Center
Dhahran Health Center (Saudi Methodist Hospitals of Memphis (MN)
Arabia) (TN) St. Vincent Medical Center (CA)
Duke University Medical Center Montreal Children’s Hospital San Francisco General Hospital
(NC) (Canada) (CA)
Dwight David Eisenhower Army Mount Sinai Hospital (NY) Seoul Nat’l University Hospital
Medical Center (Ft. Gordon, Mount Sinai Hospital (Toronto, (Korea)
GA) Ontario, Canada) Shanghai Center for the Clinical
East Side Clinical Laboratory (RI) National Genetics Institute (CA) Laboratory (China)
Easton Hospital (PA) Naval Hospital Cherry Point (NC) Shore Memorial Hospital (NJ)
Federal Medical Center (MN) New Britain General Hospital SmithKline Beecham Clinical
Frye Regional Medical Center (CT) Laboratories (GA)
(NC) New Hampshire Medical SmithKline Beecham Clinical
Gila Regional Medical Center Laboratories Laboratories (TX)
(NM) The New York Blood Center South Bend Medical Foundation
Grady Memorial Hospital (GA) New York State Department of (IN)
Great Smokies Diagnostic Health So. California Permanente
Laboratory (NC) New York State Library Medical Group
Gulhane Military Medical New York University Medical Southeastern Regional Medical
Academy (Turkey) Center Center (NC)
A. Samuel Koenig, III, M.D., Carl A. Burtis, Ph.D. Robert F. Moran, Ph.D.,
President Oak Ridge National Laboratory FCCM, FAIC
Family Medical Care mvi Sciences
Sharon S. Ehrmeyer, Ph.D.
William F. Koch, Ph.D., University of Wisconsin David E. Nevalainen, Ph.D.
President Elect Abbott Laboratories
National Institute of Standards Elizabeth D. Jacobson, Ph.D.
and Technology FDA Center for Devices and Donald M. Powers, Ph.D.
Radiological Health Johnson & Johnson Clinical
F. Alan Andersen, Ph.D., Diagnostics
Secretary Hartmut Jung, Ph.D.
Cosmetic Ingredient Review Boehringer Mannaheim GmbH Eric J. Sampson, Ph.D.
Centers for Disease Control
Donna M. Meyer, Ph.D., Tadashi Kawai, M.D., Ph.D. and Prevention
Treasurer International Clinical Pathology
Sisters of Charity Health Care Center Marianne C. Watters,
System M.T.(ASCP)
Kenneth D. McClatchey, M.D., Parkland Memorial Hospital
Charles F. Galanaugh, Past D.D.S.
President Loyola University Medical Ann M. Willey, Ph.D.
Becton Dickinson and Center New York State Department of
Company (Retired) Health
Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
Committee Membership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
3 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
4 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
5 Design Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
6 Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
8.1 Diluent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
8.2 Sterile Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
8.3 Purchased Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
9 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Foreword
This guideline describes water of three specific levels of quality (Types I, II, and III) and the
methods for producing and testing such water. The classifications and specifications are designed
to enable laboratory scientists and supporting industries to specify the quality of water to be used
in such procedures as, for example, reagent preparation, reconstitution of lyophilized materials, and
sample dilution.
The committee believes that the criteria and measurements specified for monitoring water quality
are the minimum necessary. The parameters are as follows:
! Resistivity
! Microbial content
! pH
! Silicate content
! Particulate matter
! Organic content.
Measurements of resistivity are practical and readily available, and they provide significant
information about the water sampled. As the sensitivity of laboratory analytical processes
increases and sample size decreases, microbial content of the reagent water becomes increasingly
important. Microorganisms can inactivate reagents, contribute to total organic contamination, or
alter optical properties of the test solutions.
The monitoring of other parameters—namely pH, silicate content, particulate matter, and organic
content—depends on many variables. Each laboratory should assess the need for, and frequency
of, monitoring any of these on a routine basis. If the source water and the purification system
produce water that is typically negative for some contaminant, the frequency of testing for that
contaminant can be decreased. However, it is necessary to ensure occasionally that the end
product is free from all significant contaminants.
This guideline recommends that the laboratory examine the acceptability of the type of reagent
water to be used and record the rationale for this decision. A laboratory should also check to see if
there are requirements applicable to its specific uses.
Key Words
c
College of American Pathologists Commission on
Laboratory Inspection and Accreditation. Reagent Water
Specifications (1985). College of American Pathologists,
325 Waukegan Road, Northfield, Illinois 60093-2750.
d
USP 23, Official Monographs: Water, pp.
American Chemical Society. Reagent Chemicals, Eighth
a
1635–1637; High Purity Water, pp. 1782; Water for
Ed., American Chemical Society Specifications (April 1993), Pharmaceutical Purposes, pp. 1984; Reagents, Indicators,
pp. 69, 70, and 777–778. American Chemical Society, and Solutions, pp.1987. United States Pharmacopeia,
1155 Sixteenth Street, N.W., Washington, DC 20036. 12601 Twinbrook Parkway, Rockville, Maryland 20852.
treatment with many biocides that are sample. For operational purposes, the filter
effective in killing free-floating organisms. pore is usually 0.45 µm.
Passivation, n - A process used to remove Total organic carbon (TOC), n - Carbon in the
surface iron from stainless steel piping by form of organic compounds.
acid etching and to oxidize the remaining
chromium and nickel surfaces to impervious Ultrafiltration, n - A process during which
oxides. water is forced under pressure through a
semipermeable membrane leaving behind a
pH, n - [The symbol for the power of percentage of dissolved organic and
hydrogen defined as] the negative decadic suspended impurities. The dissolved organic
logarithm of the [relative molal] hydrogen ion and suspended impurities are filtered based
activity. on molecular weight and size.
Sanitization, n - Chemical and/or physical Because there are many options in water-
processes used to kill microorganisms or purification technology, the working group
reduce contamination. cannot recommend a particular purification
process to produce a particular grade of
Source water, n - The water that is water. The decision as to which system or
introduced into the purification process. systems to install depends on past
experience, and on present and future needs.
Total microorganisms, n - All aerobic and A major consideration is the quality of the
facultative anaerobic heterotrophic microor- source water. If there is a high concentration
ganisms in water. of total dissolved solids, it is likely that the
water will have to be pretreated before
Dissolved Dissolved
Purification Process Ionized Ionized Dissolved Particulate Micro- Pyrogens/
Solids Gases Organics Matter organisms Endotoxins
Distillation E G/P G E E E
Deionization E E P P P P
Reverse osmosis G P G E E E
Carbon P P E/G P P P
adsorption/absorption
Ultrafiltration P P P E E E
Nanofiltration G/P P G E E E
Ultraviolet sterilization* P P P P G P
* = Ultraviolet light kills microorganisms but does not remove them. Another process is required to remove them.
E = Excellent (capable of complete or near total removal).
G = Good (capable of removing large percentages).
P = Poor (little or no removal).
can be an advantage in shedding biofilm. reinforced resin. Small scale tanks may be
PVDF can also withstand continuous biocidal made from any of the materials used for
concentrations of chlorine or ozone. piping. Glass may be used, depending on the
Installation costs are greater than for PVC but application. Clean, chemically resistant
might decline as the technology improves. glasses have the advantage that they
minimally leach TOC into the water and they
5.3.1.6 Perfluoro-alkoxy vinyl ether (PFA) are not permeable to volatile substances in
the ambient air. They do, however, leach
PFA is more fluorinated than PVDF; it offers some ions because all glasses are slightly
increased chemical resistance and heat soluble.
tolerance (to 260 BC). PFA is substantially
more expensive than PVDF and its field use 5.3.3 Spigots
has been limited.
Spigots should be considered an integral part
5.3.1.7 Poly-ether-ether-ketone (PEEK) of any distribution system. They should be
selected with the same care as the piping and
There are several forms of fluoropolymers, storage tanks. It is particularly important that
some of which do not mold well into piping. spigot design minimize dead spaces and use
PEEK is an essentially pure material with seals that do not leach contaminants into the
excellent chemical stability and higher water.
operating temperature than PVDF. PEEK can
be used at substantially higher pressures than NOTE: Tubing should not be left attached to
other plastics and its excellent chemical spigot outlets. If tubing is used to transfer
resistance and temperature rating (150 BC) water, it should be carefully selected for
permit a variety of options for biofilm control. chemical resistance and carefully cleaned.
There is less inorganic and organic leaching The risk of chemical and microbial
from PEEK than from any other plastic. Field- contamination when transferring water via
use of PEEK has been limited however, tubing is high.
because there is no inexpensive commercial
way to create welds and few fittings are 6 Specifications
available.
All specifications are stated for water as
5.3.1.8 Polytetrafluoroethylene (PTFE) measured at the time of production. The
resistivity of Type I water should be
The most chemically resistant of the measured inline daily; all other specifications
fluoropolymers is the fully fluorinated PTFE; relate to the samples measured offline.
however, this material cannot be welded or
easily formed into pipes. Additional purification can be required for
selected clinical laboratory procedures, such
5.3.2 Storage as:
The specifications for reagent water Types I, The resistivity requirement for Type II water
II, and III are summarized in Table 2 on the is slightly more liberal than what will occur
next page. when theoretically pure water is allowed to
equilibrate with ambient room air.
6.1.1 Microbial Content
Theoretically, pure water that has equilibrated
Ideally, Type I water should be free of with room air (thereby absorbing carbon
microorganisms. (For a discussion of the dioxide, which hydrates to form carbonic acid
effects of microbial contamination, see and then dissociates to form conductive ions)
Section 9.1.2.) It is recognized, however, has a typical resistivity of about 0.8 megohm
that water manufactured by a continuous @ cm, or about 1.2 FS/cm expressed as
process might not be sterile at all times. The conductivity.
working group suggests that the microbial
content specification for Type I water be #10
CFU/mL at the time of production. Similarly,
the working group suggests that 1,000
CFU/mL should be considered the upper limit
for microbial content in Type II water. When
microorganisms are present in water, the
microbial content of the water changes over
time.
6.1.2 pH
pH NS NS 5.0–8.0
______________________________________________
*: This is a purification process requirement and is not measured by the end user.
NS: Not specified.
6.1.4 Maximum Silicate matter are not specified for Types II and III
reagent water.
Soluble or colloidal silica can be present in
the source water and it might not be 6.1.6 Organic Contaminants
adequately removed in the purification
process. Silicates Organic contaminants should be kept to a
or colloidal silica can interfere with certain minimum in Type I water. The content of
assays. Levels of 0.05 mg/L or below, organic material is reduced when water is
measured as SiO2, do not appear to cause distilled, subjected to reverse osmosis, or
interferences. The higher level of silica as passed through activated carbon. A
specified for Type II water, may or may not combination of these processes can be more
cause interferences; however, satisfactory effective in removing organic material. If
use should be documented. activated granular carbon is used, periodic
6.1.5 Particulate Matter replacement of the activated carbon is
necessary. If distillation or reverse osmosis is
Type I water should be free of particulate used, resistivity measurements might not
matter, including microorganisms, larger than meet the requirements for Type I water.
0.2 µm. This can be achieved by passing the
water through a post-membrane (vinyl) filter 6.1.7 Degradation of Water Quality
with a mean pore size no larger than 0.22
µm. However, note that many users elect to Because certain characteristics, such as
add a postmembrane filter with a pore size of resistivity and microbial content, change
0.1 µm. Purification process requirements for quickly once the water is produced, their
particulate influence on
usability must be evaluated (see Section 7).
7 Storage and Handling vessel and the water must be replaced daily.
Storing water in large vessels (carboys) for
The working group recommends that the extended periods of time is unacceptable
laboratory examine the acceptability of the because of the inevitable, unpredictable rate
type of reagent water to be used and record of degradation of the water quality.
the rationale for this decision (see Section
8.3.3). 7.3 Special Reagent Water
Type I water can be thought of as the "ideal" To retard the further, inevitable, unpredictable
general purpose water that can be produced rate of water quality degradation, good
with currently available water treatment/ laboratory practice should be followed in
purification technology and used at the time removing water from a container. The user
of production. Type I water should be used must not touch the lid or inside cover, or dip
in test methods that require minimal pipets into the container. Water should be
interference or when lack of interference from poured for use into a secondary container,
water of lesser purity cannot be documented and the unused portion must not be returned
or inferred. to the original container.
(2) Open the tap fully for a minimum of 1 ! Use of media that will support the growth
minute before sampling, then restrict the flow of microorganisms most commonly
to fill the container without splashing. The isolated from water in the time frame
working group emphasizes that inadequate designated.
flushing is one of the most common causes
of an elevated microbial count. ! Supplies available to perform the
procedure.
(3) Collect a minimum of 10 mL of water
from ! Cost of the procedure.
each sample site.
9.1.4.3 Limitations of Methods
(4) Process the sample within 1 hour of
collection, or within 6 hours when stored at 2 The methods recommended are not inclusive
to 8 BC. of all methods that might satisfactorily
comply with the above objectives. Different
methods can recommend sampling different
volumes of water, especially when
commercially available microbial enumeration
kits are used. The sensitivity of a method is (1) Incubate the samples at 36 ± 1 BC for
enhanced by sampling more than 1 mL of 24 hours
water, because it is possible to detect
contamination at levels less than 1 CFU/mL. (2) Then incubate the samples at the
Larger volumes also help negate distributional ambient temperature (23 ± 3 BC) for an
problems with suspended microorganisms in additional (minimum of) 24 hours.
fluids.
The total incubation time is at least 48 hours.
To provide a representative distribution of
microorganisms for a total microbial count for 9.1.6 Standard Plate Count (SPC)
clinical laboratory reagent water, sufficient
vortexing or mixing of a specimen is 9.1.6.1 Equipment
extremely important. If a commercially
available kit enumeration system is used, To perform the standard plate count, the
follow the instructions for sampling and following equipment is necessary:
enumeration and convert the results to
CFU/mL when determining compliance with ! Incubator (see Section 9.1.5.1).
this guideline.
! Binocular (dissection) microscope with
The working group does not recommend the total magnification of 10 or 15X and
calibrated loop method because it lacks adequate illumination, or a Quebec colony
sensitivity in determining colony counts of counter.
less than 100 CFU/mL.
! Petri dishes, sterilized (15- X 100-mm).
9.1.5 Incubation Conditions for
Determining Total Microbial Counts ! Vortex mixer (optional).
For incubation in an air or heat sink, the steps Heterotrophic plate count agar (HPC),
are as follows: trypticase soy agar (TSA), brain heart infusion
agar (BHI), standard plate count agar (SPC),
(1) Maintain a temperature of 36 ± 1 BC; R2A media, or any suitable transparent
check the temperature using a thermometer medium capable of supporting the growth of
with calibrations traceable to a National the organisms mentioned in Section 9.1.4.1,
Institute of Standards and Technology (NIST)- should be prepared according to the
certified thermometer manufacturer's instructions.
mix the agar and inoculum by rotation. Allow growth of the organisms mentioned
the agar to solidify. in Section 9.1.4.1
(4) Invert the dishes and incubate them as — 0.1% peptone water
discussed in Section 9.1.5.2.
— Sterile dilution water
(5) Count the colonies and report the number
of viable colonies per plate (1-mL sample) as — Alcohol.
CFU/mL. The use of magnification is
preferred for the enumeration of CFU/mL. ! Procedure
— Filtration units—filter funnel, (6) Filter the sample and rinse the sides of
manifold or vacuum filter flask, the funnel at least twice with 20 to 30 mL of
tubing, and a vacuum source sterile dilution water. Turn off the vacuum
and remove the funnel from the filter base.
— Alcohol or gas flame Aseptically remove the membrane filter from
the filter base and place it grid-side-up on the
— Forceps pad.
— Total count media or any suitable (9) Express the results in CFU/mL.
broth medium capable of supporting
†
Two representative filtration methods are described.
The following equipment is necessary for the For counting colonies, the steps are as
performance of the filtration kit method: follows:
— Sterile syringe with locking hub or (4) If there are many colonies, use the
vacuum pump (method dependent). number in 10 randomly located grid squares
to estimate the total number on the filter
! Reagents square. Determine the number of grids on
the filter. The CFU/vol tested is calculated
Use the reagents contained in the kit. as:
(3) Reinsert the paddle and immerse for the (3) Add one drop (.30µL) of saturated KCl to
specified time. a 50-mL water sample.
(4) Decant the water and incubate the (4) Take the pH reading as quickly as
paddle according to the manufacturer's possible following the manufacturer's
instructions. instructions.
(5) Determine CFU/mL. (5) Read the pH to the nearest 0.1 unit.
Table 3. Resistivity Temperature Table (10 (2) Immerse the cell and temperature sensor
Megohms @ cm Type I Water) in the solution; move it up and down using a
circular motion to remove any entrapped gas
bubbles. Take a reading immediately.
Temperature Resistivity
BC Megohms @ cm
(3) If necessary, apply the manual
0 28.25 temperature compensation, according to
Appendix C.
5 22.65
NOTE: Readings taken with dip cells are
10 18.27 somewhat less accurate because
15 14.85 solutions are open to the
atmosphere. At purity levels above
20 12.15 1 megohm @ cm, dissolving
atmospheric CO2 will cause the
25 10.00 measurement to drift and ultimately
result in artificially low resistivity
30 8.28
readings.
35 6.91
9.3.3 Calibration
40 5.80
45 4.90
The resistivity meter and conductivity cell
should be calibrated and checked in
50 4.16 accordance with manufacturer’s instructions.
Some meters have an automatic check;
9.3.2.3 Measurement of Type I Water others have a manual test button. Note that
this type of calibration check involves
The 10 megohm @ cm resistivity of Type I disconnecting the meter cell (either
water can be measured only by using an automatically, by pressing a button, or
inline cell. Measure according to the manually) and inserting a resistor of known
manufacturer’s directions. Automatic value in place of the cell to obtain a known
temperature compensation is required. meter display. This only checks the
functionality of the meter electronics and
9.3.2.4 Measurement of Types II and III does not verify accuracy of the cell. For
Water offline meters and some online meters, a KCl
solution or certified conductivity standard can
The resistivity of Types II and III water can be be used to calibrate and monitor the day-to-
measured using an inline or a diptype cell. day operation of the meter/cell system. The
0.01 mol/L KCl solution has a resistivity of
For inline meter/cell combinations, the 70.7 kS @ cm (electrolytic conductivity, 1414
manufacturer’s instructions should be µS/cm) at 25.0 BC. The 0.001 mol/L KCl
followed. solution has a resistivity of 6.81 kS@cm
(electrolytic conductivity, 146.9 µS/cm) at
When measuring the resistivity with an offline 25.0 BC. Standard reference materials
meter/cell combination, the procedure is as (SRMs) with certified values of electrolytic
follows: conductivity are available from the National
Institute of Standards and Technology,
(1) Rinse the cell and container at least three ranging from 10 S@cm (100,000 µS/cm) to
times with separate aliquots of water to be 0.2 MS@cm (5 µS/cm). Commercially
tested. It is recommended that the water be available conductivity standards may also be
thermostated to 25 ± 2 BC during the used. Meter tolerances will vary dependent
measurement. Otherwise, either automatic or on the manufacturer. Generally, newer
manual temperature compensation should be meters are accurate to ± 5% of the actual
used. reading. Older models are accurate to ± 3%
Meter tolerances vary depending on the Silicates react with molybdate ion to form a
manufacturer. Consult the manufacturer to complex that can be reduced by 1-amino-2-
determine the expected degree of accuracy naphthol-4-sulfonic acid to produce a blue
for the meter in use. color. The intensity of the blue color is
proportional to the concentration of soluble
silica.
(4) Dilute to 200 mL with Type I water. The solution is stable for three months.
The solution is stable for two weeks The steps for preparation and storage are
(discard when the color darkens or as follows:
precipitate forms).
(1) Dissolve 0.473 g of sodium
! Ammonium molybdate solution, 75 g/L metasilicate (Na2SiO3 C 9H2O) in Type I
(85.9 mmol/L) water.
The steps for preparation and storage are (2) Dilute to 1 L with Type I water.
as follows:
(3) Store in a polyethylene bottle at room
(1) Dissolve 7.5 g of ammonium temperature.
molybdate [(NH4)6 Mo7O24 C 4H2O] in Type
I water. The solution is stable for three months.
(2) Dilute to 100 mL with Type I water. 9.4.3.2 Equipment and Materials
The steps for preparation and storage are For sampling, 100-mL water samples are
as follows: collected in chemically clean polyethylene
bottles. The temperature of the water (6) Silica concentration, in milligrams per
collected should not exceed 35 BC. liter, may be read directly from the calibration
curve. Record the reading in a permanent
9.4.3.4 Standardization log.
Bibliography
American Society for Testing and Materials. Gabler R, Hegde R, Hughes D. Degradation of
Standard test method for silica in water. high purity water on storage. J Liquid
ASTM designation D859–88. Conshohocken, Chrom. 1983;6:2565–2570.
PA: ASTM; 1988.
Ganzi GC. Preparation of high purity
American Society for Testing and Materials. laboratory water. In: Methods in Enzymology.
Standard test methods for electrical New York: Academic Press; 1984;(104).
conductivity and resistivity of water. ASTM
designation D1125-82. Conshohocken, PA: Hamilton H. Selection of materials in testing
ASTM;1989. and purifying water. Ultrapure Water.
1985;2: 36–38.
American Society for Testing and Materials.
Water: atmospheric analysis. In: 1990 Annual Hanselka R, Reinzuch KJ, Bukey M. Materials
Book of ASTM Standards, Conshohocken, of construction for water systems part II:
PA: ASTM;1990; (11.1; 11.2). Real life failure modes of plastics. Ultrapure
Water. 1987;4:50–53.
HA-AWWA-WPCF. Standard Methods for
Examination of Water and Waste Water. Hanselka R, Williams R, Bukey M. Materials
17th ed. Washington, DC; 1989. of construction for water systems part I:
Physical and chemical properties of plastics.
Batjer JD, et al. Effects of microbial Ultrapure Water. 1987;4:46–50.
contamination of reagent water on selected
laboratory tests. Am J Clin Pathol. 1979; 71: Highsmith AK. Water in health care facilities.
19–325. In: Architectural Design and Indoor Microbial
Pollution. Oxford: Oxford University Press;
Blais P, Cooper MT. Contaminants in clinical 1988:81–102.
reverse osmosis water purification systems.
JAMA. 1980;243:649. Highsmith AK, Kaylor BM, Reed CJ, Ades
EW. Evaluation of Water Treatment Systems
Bowers GN Jr. Reagent quality: A basic Producing Reagent Grade Water. SAE
analytical consideration in clinical Technical Paper: Series 901424. Society of
enzymology. CAP/ASPEN. 1982 Enzyme Automotive Engineers; 1990.
Conference.
Jorgensen JH, Smith RF. Rapid detection of
Bristol DW. Detection of trace organic contaminated intravenous fluids using the
impurities in binary solvent systems: a solvent Limulus in vitro endotoxin assay. Appl
purity test. J Chrom. 1980;188:193. Microbiol. 1973:26;521–524.
Bukey M. Deadlegs: A widespread threat to Kaplan LA, Pesce AJ. Methods in Clinical
DI water systems. Ultrapure Water. Chemistry. 2nd ed. St. Louis, MO: CV
1987;4:66– 70. Mosby; 1989.
Callaghan TJ. Practical guide for the selection Malaiyandi M, Cooper MT, Blais P. Reverse
of a laboratory water purification system. osmosis units do not remove all water
Am Lab. 1988;5:60–68. contaminants. CMA J.1980;122:15–16.
College of American Pathologists, Meltzer TM. High purity water preparation for
Commission on Laboratory Inspection and the semiconductor, pharmaceutical, and
Accreditation. Reagent water specifications. power industries. Littleton, CO: Tall Oaks; p.
Chicago: CAP; 1985. 800.
Nielsen SS, Highsmith AK, Crow SA. A Tietz NW, ed. Fundamentals of Clinical
Comparison of Filamentous Fungal Chemistry. 4th ed. Philadelphia, PA: WB
Populations in Potable Water, Point-of-Use Saunders; 1996.
Water, and Room Air. Atlanta, GA: Centers
for Disease Control and Prevention, Water Tietz NW, ed. Textbook of Clinical Chemistry.
Quality Laboratory; 1995. 2nd ed. Philadelphia, PA: WB Saunders;
1994.
Poirier SJ, Sienkiewicz PM. Organic free
water. Am Lab. 1980;12. United States Pharmacopeial Convention, Inc.
US Pharmacopeia XXI. Rockville, MD; 1985.
Rechen HC. Piping system designed and
materials for optimal performance in ultrapure Winstead M. Reagent Grade Water: How,
water transmission. Ultrapure Water. When and Why? Austin, TX: American
1985;2: 39–42. Society of Medical Technologists, Steck Co.;
1967.
Sullivan JD Jr, Valois FW, Watson SW.
Endotoxins: The Limulus amebocyte lysate Wood JH. Pure water systems for hospitals.
system. In: AW Bernheimer, ed. Mechanisms Plumbing Engineer Magazine. July/August,
in Microbial Toxicology. New York: John 1985.
Wiley and Sons; 1976:217.
A.1 Distillation
Distillation separates water from contaminants by changing the state of water from a liquid phase
to a gas phase and then back to a liquid phase. Each of these transitions provides an opportunity to
move pure water away from contaminants that do not make the transition with the same efficiency
as water molecules. In theory, distillation can be the most predictable of all water-purification
technologies, provided the equipment is well designed and properly utilized.
Using a still, there are three stages in the laboratory distillation process: (1) the boiler stage, (2) the
transition stage, and (3) the condenser stage. The function of the boiler stage is to produce steam
that approaches equilibrium with the boiler water and not to overheat the steam to the point where
it carries an unnecessary burden of contaminants with low, but significant, vapor pressures. It is
also a function of the boiler to minimize the production of boiler water particles in the steam during
distillation. The third function of the boiler stage is to eliminate the concentrating residual
contaminants from the still without degrading the quality of the steam produced by the boiler. The
function of the transition stage is to smooth the inevitable imperfections in boiler operation by
trapping mist particles and by equilibrating overheated steam. The function of the condenser is to
remove sufficient heat calories from the steam to permit condensation while maintaining a
steam–water equilibrium. It is also the function of the condenser to mechanically separate the
distilled water from the steam phase without permitting the redissolution of volatile contaminants,
which must be vented from the condenser at greatly increased concentrations. For a condenser to
properly perform this function, a condenser must not enclose surfaces that are significantly cooler
than the boiling point of water, a requirement that tends to preclude the use of coils or water jacket
designs. The condenser should also be designed so that the output distilled water is mechanically
partitioned from the steam containing relatively high concentrations of contaminating volatiles, and
permitted to equilibrate with steam containing concentrations of volatiles no higher than the steam
leaving the transition stage.
Most dissolved ionized solids (i.e., common inorganic salts) have insignificantly low vapor pressures
at the boiling point of water and will not make the transition from boiler water to boiler steam.
These dissolved solids must be periodically flushed from the boiler. If the transition stage is not
effective, dissolved ionized solids can migrate into the condenser stage where they can no longer
be removed.
Most dissolved ionized solids (i.e., common inorganic salts) have insignificantly low vapor pressures
at the boiling point of water and will not make the transition from boiler water to boiler steam.
These dissolved solids must be periodically flushed from the boiler. If the transition stage is not
effective, dissolved ionized solids can migrate into the condenser stage where they can no longer
be removed.
Dissolved ionized gases include gases such as carbon dioxide, sulfur dioxide, and ammonia, which
form ions in solution. These gases are the major contributors in resistivity changes in distilled
water. If a still can produce water with a resistivity approaching 5 @ 106 S-cm, the boiler,
transition, and condenser phases are operating effectively. The resistivity of the water must be
tested by the flow cell at the point of exit from the still because ionized gases in the ambient air
will dissolve rapidly into the water and lower the resistivity.
The bulk of dissolved organics is removed from the distillate because they have a lower vapor
pressure or higher vapor pressure relative to water. Those organics that have vapor pressures close
to water, including azeotropes, might not be removed completely by distillation. One strategy to
minimize this problem is to feed the still boiler with water previously purified by another purification
system, for example, reverse osmosis or deionization.
Particulate matter and microorganisms should not penetrate the transition stage of a properly
designed and maintained still. However, permitting the still to rinse down, after start-up and before
collecting water, is recommended.
Pyrogens and endotoxins are large molecules that should not penetrate the transition stage of a
properly designed and maintained still.
A.2 Deionization
Ion exchange is an effective way to produce water with a high resistivity, because strong-
acid/strong-base, mixed-bed exchange resins can effectively remove the mobile ions that elevate
the conductance of water. However, the user should bear in mind that many substances present in
water are only weakly ionized (e.g., silicates) or are essentially undissociated (e.g., many organic
substances). They will not be bound by some ion-exchange resins and are not detected by
resistivity measurements. It is inherent in the gel structure of the ion-exchange resins that they
leach a variety of substances into the water stream. Resin beds provide excellent conditions for
the growth of organisms, which results in recontamination of the water. If the resins are
regenerated by an exchange service, there is a possibility that small quantities of organic and
inorganic contaminants picked up during previous use will leach into the water stream, even under
the best of circumstances. It is important to understand the limitations of deionization and to have
the supplier provide information on the specific regeneration process. The manufacturer may be
asked to provide assurance that the regeneration process will minimize contamination from previous
use and/or regeneration itself. It is important to understand that the high specific resistance
produced by ion exchange is no assurance that deionized water is “pure.”
Reverse osmosis (RO) is a process in which water is forced under pressure through a
semipermeable membrane leaving behind a percentage of dissolved organic, dissolved ionic, and
suspended impurities. RO can effectively remove more than 97% of monovalent ions and an even
greater percentage of divalent ions. However, volatile substances concentrate as the result of the
RO process and many low-molecular-weight organic substances pass through RO membranes.
There are three major classes of RO membranes as described in the table on the next page.
Phenol, formaldehyde, and acetic acid are passed preferentially by cellulose acetate membranes and
their concentrations in the output water increase with respect to the source water. RO
concentrates contaminants on the feed side of the membrane, causing significantly increased
concentrations of sparingly soluble salts. This can result in scaling of the RO membrance. A
common pretreatment to address this problem is water softening.
CA PS PA
Concentrate Yes No No
low-molecular-
weight organics
*Operating pressure is reported in pound per square inch gauge pressure (psig). For example, at sea level, the
pounds per square inch absolute pressure (psia) is 14.7, whereas the gauge pressure would read 0 psig
because nearly all pressure gauges use the surrounding air pressure as their reference point.
of substances in the water stream. Activated carbon beds are particularly effective in removing
chlorine from water. Although carbon beds are viable locations for microbial growth, the adsorptive
affinity for chlorine can greatly increase the growth of these organisms in downstream sections of
the purification system. The limitations of carbon use are as follows:
In practice, the primary use of carbon is to remove chlorine from water entering ion-exchange resin
systems. The working group recommends consulting with the manufacturer before using carbon
for water purification.
For purposes of this document, filtration is defined as a mechanical process used to remove
particulate matter (including microorganisms) that is 0.22 µm or greater in size. Ultrafiltration is the
process used for the mechanical/electrochemical removal of smaller-sized dissolved and suspended
impurities. By their nature, filtration traps particulates within the filter’s matrix, whereas
ultrafiltration retains particulates based on their size, shape, and electronic charge.
These techniques are used in combination with other purification processes. Filtration is often used
both at the beginning and end of a series of water-purification processes. Ultrafiltration is almost
always a pretreatment step. Limitations of both techniques include the following:
A.6 Nanofiltration
Performance specifications for the few nanofilters currently available vary. However, in all cases,
percent rejection of divalent ions (e.g., calcium, magnesium, sulfate) is much greater than the
rejection of monovalent ions, such as sodium and chloride. Like RO membranes, nanofilters provide
excellent removal of organics that have molecular weights greater than about 300. Because of
their much better rejection of calcium and magnesium over sodium, nanofilters are sometimes
referred to as “loose” or softening RO membranes. However, a typical nanofilter can be operated
at higher water recovery than a reverse osmosis membrane, thus reducing the amount of water
wasted. The safe operating pH of nanofilters is wide; however, like PA RO membranes, most
nanofilters have a low resistance to attack by chlorine in water.
Although a number of nanofilters are currently available, the commercial use of them at this time is
limited. However, given their unique properties, water systems for laboratories might begin
employing this emerging membrane technology.
Though not widely used in clinical laboratory water-purification systems, ozone is gaining in
popularity. Ozone is 5 to 1000 times more effective as a biocide than chlorine on a weight basis
(depending on the organism and the form in which the chlorine is dispensed). In a sense, ozone is
the perfect biocide for reagent water; it breaks down to oxygen with a half life of about 25 minutes
at 20 BC and the breakdown can be accelerated with a UV source, which makes ozone easy to
remove at its point of use. Because ozone is such a powerful oxidant, it degrades RO membranes
and most polymeric plastics, including ion-exchange resins.
Ultraviolet oxidation results from the absorption of 185-nm light, which produces hydroxyl radicals
that oxidize organic material to smaller ionizable components. Recirculation of water over an
ultraviolet source and a deionizer can result in a significant reduction of organic material.
Ultraviolet oxidation does not guarantee removal of all organic substances from water.
Ultraviolet sterilization results from the absorption of 254-nm light, which damages the DNA and
RNA of microorganisms causing cell death. The efficiency of both processes depends on the
amount of light that actually penetrates the water; the design of the devices is an important factor.
While ultraviolet sterilization kills microorganisms, the dead casts are not degraded or removed,
which can potentially lead to pyrogen formation.
Both ultraviolet oxidation and ultraviolet sterilization are used in combination with other water-
purification processes and they are often positioned in a recirculating loop.
To ensure sample quality throughout sample collection and analysis, a set of operating procedures
should be established by each laboratory. A quality assurance program should include both quality
control and quality assessment. Guidelines addressed in the 17th edition of Standard Methods for
the Examination of Water and Wastewater can serve as the basis for program development. The
subsequent procedure manual should be prepared in accord with NCCLS document GP2-A3, Clinical
Laboratory Procedure Manual—Third Edition; Approved Guideline.
When standard operating procedure manuals are compiled, consideration should be given to the
following topics:
! A quality assurance plan with approval signatures, organizational charts, and responsibilities.
! Special worksheets designed for reporting the results of daily, weekly, and monthly tests, etc.
Resistivity checks should be recorded daily. Microbiological testing should be performed weekly.
All other parameters should be tested as necessary, depending on geographical and seasonal
considerations, or the manufacturer's recommendations.
(1) In automatic temperature compensation, the meter’s probe cell contains a temperature
compensation circuit element that internally adjusts the meter to read corrected resistance
directly.
D2 % D1
Da v g'
2
D2 ! D1
f '
(t1 ! t2)D a v g
D25 = Dt (1 + f)ªt
D25 = Dt (1 ! f)ªt
(3) The following temperature compensation equation is known as Marsh and Stokes’ equation.
This equation converts any measured electrolytic conductivity reading to its equivalent value at
25 BC.
where:
D25 = 1/625
C3-A2: Preparation and Testing of Reagent Water in the Clinical Laboratory—Second Edition;
Approved Guideline
General
However, in Type I water specifications, particulate matter and organic contaminants are
covered by a process specification (i.e., not measured by the end user). How then is
someone expected to discover if the process is still efficient after a given time? If an
interlaboratory quality assurance program, such as the one we are considering setting up,
were to include those two parameters, how should they be measured and according to
what "standard" should they be interpreted?
! The C3-A3 guideline recommends a process for achieving the types of water described.
The working group does not recommend a procedure for total organic contaminants and
particulates. Refer to the following ASTM standards for specific information on these
processes:
- ASTM. Standard Test Method for Total Carbon and Organic Carbon in
Water by Ultraviolet, or Persulfate Oxidation, or Both, and Infrared
Detection. ASTM designation D4839-94. ASTM, West Conshohocken,
PA, 1994.
2. Are you aware of any recent reference dealing with quality assurance programs for
monitoring high-purity water? Even after computer searches through data banks, the only
useful reference we came up with was "Quality Control for a 14 to 18 Megohm-cm
Deionized Water Supply," by EJ Dravian, I Schoen, F Abrams, P Datta, EM Custer, Arch
Pathol Lab Med, 1986, 110, 228:231.
! The bibliography in C3-A3 was updated to include references for quality assurance
programs. Note that the article cited in your comment recommends a specification of 100
CFU/mL for Type I water. This level of microbial content was determined using a “loop”
method of inoculation. This method is less sensitive and cannot be considered equivalent
to the method described in C3-A3.
4. The proposed standards may or may not be equivalent to existing similar standards from
ACS, ASTM, and USP. Because it is not clear why a new set of guidelines is necessary, it
seems to complicate the choice as to what is appropriate water quality for a particular
application.
! The purpose of all NCCLS documents is to put useful references in the hands of
laboratorians. The C3-A3 guideline focuses on clinical laboratories. The documents from
ACS, ASTM, and USP are valid standards written for different purposes and have served
as references for this guideline.
5. The basic concept that the level of purity of water can be established by conductivity
measurements alone is incorrect. Water that exhibits low conductivity may well contain a
variety of organic compounds; resin particles; colloidally dispersed metals; silica in solution
or silica colloidally dispersed; and some gases. It may also have any number and variety of
microbiological organisms. It follows that the concept that variations in levels of
conductivity reflect levels of purity is similarly incorrect.
Using water that is validated only with a conductivity parameter can directly influence
laboratory results. For example, colloidal iron will not influence conductivity, but it can
completely obscure a low serum iron. Silica in water will not alter conductivity, but it can
inhibit enzyme activity. The possibilities for errors involved in immunofluorescent assays
are high. I find the use of any single purity criterion inadequate.
The United States Pharmacopeial Convention (USP)-grade (purified water) is the most
generally used in industries (pharmaceutical), and it is already available in most hospital
and other laboratories. It is the starting point for higher levels of purity, such as water for
irrigation and water for injection. It is a satisfactory product for most laboratory tests.
The highest level of purity of water is "water for injection." This product meets all of the
requirements for USP purified water (as well as ACS) and the finished sterile product is
proven to be free of pyrogens. At the present time, this water represents the highest
purity water that is generally available. The NCCLS statement that injection-grade water is
not suitable for laboratory use is incorrect. The further statement that Type I water is
superior to injection-grade water for any use, in any situation, is similarly not correct.
! This document does not establish water purity on the basis of conductivity measurements
alone. However, such measurements are practical and readily available, and they provide
significant information about the water sampled. The working group does not agree with
the claims about USP grade or “water for injection.” It does believe, however, that the
wide range of issues mentioned in the comment have been considered in the revision of
C3-A3.
6. The belief that water degrades under normal ambient conditions is incorrect. Water is
extremely stable. It does not oxidize or undergo reduction even at high temperatures in
steam systems or in ice at super-low freezer temperatures. Water does degrade only at
extremely high temperatures, well over 1000 BC, as, for example, in contact with molten
metals; degradation then occurs with explosive violence.
What the document refers to as degradation is actually the equilibrium that is established
between any material and its environment. This is a basic requirement of the second law
of thermodynamics, and it applies throughout the universe.
The conductivity of water increases primarily because of the atmospheric carbon dioxide
equilibrating with the water. Whatever water is used for subsequent to its production, to
prepare a saline solution, for example, yields a prepared reagent solution that, in turn,
tends to come into equilibrium with its environment. The recommendation that water be
used before it "degrades" (an NCCLS term) is meaningless.
! C3-A3 is revised to clarify that the term “degradation” refers to the degradation of water
quality.
The generally accepted references for water purity are as follows: American Chemical
Society (ACS); the United States Pharmacopeial Convention (USP); and good
manufacturing practices (GMP) (required by the FDA). In each case, multiple parameters
are measured and compared to established specifications.
Foreword
! The working group revised the Foreword to clarify its originally intended meaning.
9. I recommend modifying the first sentence of the third paragraph as follows: “Resistivity
measurements are simple and provide an excellent index of ion content when the water is
first purified/manufactured.” I also suggest changing the “and” to “but” in the second
sentence.
10. The first two sentences of the fourth paragraph should be modified as follows: “Monitoring
of the other parameters noted is dependent upon many variables. Each laboratory has to
assess the needed frequency of monitoring any of these on laboratory manufactured,
purchased, or stored water, depending on the specific application.
! The working group did not agree to make these changes in the Foreword. It believes the
issues are addressed in the text of C3-A3.
11. An additional sentence should be added to the Foreword that states, “Additionally, specific
applications may have more stringent requirements for the parameters noted or for
monitoring of additional parameters.”
! The working group does not believe that this level of detail is appropriate for the Foreword.
The issue is addressed in the body of the document.
Section 2.0
12. The second paragraph should specify the documents published by those organizations
referenced.
! C3-A3 includes the related document citations and the mailing addresses of the
corresponding organizations.
Section 3.0
13. I do not like the definition of HPLC. "An...the eluent or carrier is a liquid under pressure.”
Should read: "...the mobile phase is a liquid under pressure.”
14. I do not think that pH is "the electromotive force between....” Perhaps it should read,
"determined by measuring the electromotive force between..."
15. Use the Random House American Dictionary for common words; then consult with NCCLS
document NRSCL8 for the other words before publication.
Section 4.0
! The term “feedwater” was intended to be synonymous with “source water.” To reduce
confusion, the term “source water” is used throughout C3-A3.
17. Footnote a of Table 1 indicates that there is preliminary evidence of effectiveness. Is there
any new information available?
! The working group is not aware of any new published information to the contrary.
Section 5.0
18. This guideline should give some recommendations as to the space allocations required for a
water system.
! The working group cannot recommend space allocations for water systems. Space
requirements depend on the system selected and its intended use.
Section 5.1
19. A sentence should be added to the end of the second paragraph that states, “Consult local
water testing authorities for advice.”
! The following sentence is added to Section 5.1, “In some instances it can be helpful to
consult local water-testing authorities for advice on the impurity content of the source
water.”
Section 5.2
20. The document makes a strong recommendation for the recirculation of water. Because I
feel that this is a critical and often overlooked component of a laboratory's water
purification system, I would like these two sentences to be in bold, or stand out in some
way. Similarly, the frequency of cleaning is buried in paragraph 3.
21. I would recommend rewording the first sentence of the third paragraph as follows:
“...sanitize prior to use and then at least as recommended by the manufacturer,
semiannually, or as determined by quality control criteria (whichever is most frequent.)”
22. In the third sentence of the third paragraph, change “will” to “may.”
! The working group did not agree with the change; it believes the statement to be correct
as written.
23. The second sentence should be modified to state, “However, additional parameters may
need to be monitored, especially if the user performs trace element analysis.”
Section 5.3.1.2
24. Should not the second paragraph be listed first if they are the better options? Additionally,
list the alternative materials in order of priority regarding the quality of reagent water level.
Note special advantages and disadvantages for specific applications as examples.
! Section 5.3.1 of C3-A3 has been completely revised to provide the end user with
important considerations for the selection of system materials.
Section 5.3.1.3
25. To the end of the third sentence add, “...for specific applications.”
! This section has been revised to reflect the current status of construction materials.
Section 5.3.2
26. The note advising personnel not to attach tubing to a spigot is correct, because there can
be a problem with growth if the tubing is left in place. However, this is done to direct the
water effectively, and some suggestion for an alternate procedure should be offered if
compliance is expected. Our internal specifications allow tubing to be attached for a day
at a time. The tubing must be removed and allowed to drain each day, and it must be
sanitized every 5 days.
27. Spigots should be of a diaphragm valve type, which are of a “sanitary” design.
! The working group cannot recommend the specific type of valve; there are practical
alternatives.
Section 6.0
28. To the first paragraph add the following sentence, “Storage of water will change some of
its characteristics, which may or may not affect usability of the water.” (This would result
in the deletion of Section 6.2.)
! The working group believes that the suggested change is inappropriate as part of the
specifications discussion.
Section 6.1
29. I suggest changing the section title to “Requirements for Types of Water” followed by a
sentence stating, “The requirements for each type of reagent water are specified in Table
2.”
Section 6.1.1
30. One should not use a terminal filter on Type I water because the filter collects bacteria,
which, in turn, releases endotoxins in increasing amounts with time. Purity should be
imparted “up front” and maintained throughout the system.
31. For parallelism of form, Section 6.1.1 and 6.1.2 should be footnotes to Table 2 or the
footnotes should be consolidated and incorporated into the numbered paragraph style.
! The working group ensured that the text and table are consistent but did not change the
format as suggested.
Section 6.1.3
32. I recommend the following additions to this section, “If all the requirements for Type I
water are essential for a specific application, Type I water must be used immediately
.....Since certain characteristics, such as resistivity, will change quickly once the water is
produced, their influence on usability must be evaluated.”
33. The unit of resistivity used throughout this document is the “megohm.” However, the
building engineers of my institution as well as some testing laboratories routinely use
“micromos.” I think the end user would be well served if guidelines were given in both
units.
34. What is the meaning of footnote b? If this is necessary as a criterion, it must be tested by
the manufacturer, whether the manufacturer is a laboratory, pharmacy, or some other
source. Otherwise, leave this out or describe its appropriateness.
Section 7.1
35. The resistivity of stored water on exposure to any amount of air will decrease because
some carbon dioxide dissolves in it. This does not mean, however, that the water is
impure or unfit for laboratory use. It is, therefore, inappropriate to state that, “Type I
water should be used immediately after processing.”
36. I suggest adding a qualifier to the first sentence, “For optimal useability.”
Section 7.1.1
37. While high resistivity generally means the absence of some impurities, low resistivity does
not always mean the presence of contamination. It is, therefore, inappropriate to make
resistivity the only measures for impurities in Type I water and to state, as does this
section, that, “Type I water cannot be stored because its resistivity will decrease, metals
and/or organic contaminants will be leached from the storage container, and bacterial
contamination will occur.”
! As stated in C3-A3, the working group considers resistivity to be only one of several
measurements used in the determination of water quality.
38. I suggest the following rewording of the first sentence, “Storage of Type I water will result
in decreases in resistivity (due to the absorption of carbon dioxide) increases in metals
and/or organic contaminants (due to leaching from the storage container), and bacterial
growth/contamination will occur.”
! The working group does not agree that Type I water can be stored.
39. Please consider adding the following paragraph to this section. “The specific effects of
this degradation on intended use of water originally produced and tested as meeting Type I
criteria, must be evaluated by the laboratory. If Type I water produced by the laboratory is
not used immediately, criteria for acceptability and monitoring must be established by the
laboratory as with any other reagent.”
Section 7.1.2
40. “The specifications for Type I water cannot be met by bottled water.” Even water purified
on site will have resistivity changes as it comes to equilibrium with its environment and
does not necessarily mean that it has become contaminated with “bad” contaminants.
41. To the end of the paragraph, add the following sentence, “However, with laboratory-
produced Type I water, if a manufacturer can produce appropriate information regarding
the characteristics of the water produced/stored, the laboratory can determine
acceptability for the intended use.”
! Section 7 is revised.
Section 7.3
42. The last sentence in Section 7.3 is awkward. Replacing “...as well as...” with “...and...”
would be an improvement.
! The former Section 7.3 is now incorporated into Section 7.2. This change is incorporated
in the section revision.
Table 3
43. We agree that "special reagent water" used for tissue/cell culture should be pyrogen-free.
In actual practice, is this mandatory and is there a level that might have been documented
as insufficient to interfere with test results? In that event, what would that level be and
what would its scope be for application in interpreting LAL results?
44. I suggest that the title of the table be modified as follows, “Typical/Suggested Uses of
Reagent Water Produced as:” and move the “Special” column after “Type III.”
Section 7.4.1
45. The description of the minimal organic content for HPLC-grade water is more confusing
than helpful. We are not told what “minimal” means or how to obtain such additionally
purified water. Putting water through ion-X or another resin should only accomplish what
should have been done to obtain Type I water. The HPLC method for determining the level
of organic contamination is poor but important enough to strengthen. What is the
monitoring wavelength? What type of gradient is to be used? Has anyone gotten advice
from the commercial people (e.g., Waters)?
Section 8.2
Section 8.3
47. This section should state that the water used in laboratories must either: (1) meet the well-
established criteria required by organizations such as United States Pharmacopeial
Convention or the American Chemical Society; or (2) be stored in nonleaching containers
with a certificate ensuring that the water is pure pursuant to the same criteria, with an
expiration date on the label. In this way the laboratory/consumer would still have the
option of purchasing bottled reagent water.
! The working group believes Types I, II, and III waters, as defined, are appropriate for
clinical laboratory uses.
Section 8.3.3
48. This section describes changes/properties that apply to both on-site generated and bottled
water.
Section 9.1
49. I have used the bacteriologic sample method for quantitating the degree of bacterial
contamination of water for many years. This method, which was developed for both urine
and water cultures, has its limitations, but it is most useful in settings that do not routinely
perform plate counts.
50. During our recent CAP inspection, one of the inspectors raised a question about one of the
tests suggested. At this time, we also looked at the microbiological testing that we had
been doing for years for our colleagues in all clinical laboratories (obviously for free!). We
have used colony counts by the calibrated loop-spreading method for many years. The
NCCLS guideline goes into an extensive description of pour-plate counts (over 6 pages),
then says in Section 9.1.4.3, “Limitations of Methods,” that the recommended methods
are not inclusive, etc. The calibrated loop or spreading methods are not mentioned at all.
Our free job will be impossible to do if we adopt the recommended pour-plate method,
start filtering, or such. We are thinking about recommending the bacteriologic sampler
method (Section 9.1.7.3), which each laboratory could do on their own.
Aside from these problems, there seems to be a general problem with the guideline; the
recommended methodologies probably were developed by industrial quality control and
public health people and not by clinical microbiologists. If the calibrated loop-spreading
method is good enough for urine cultures, it should be satisfactory for plate counts of
reagent water. We know that there will be some error, but we do not believe that we
need the accuracy of the regulators to look at different reagent waters—Type I with max.
10 CFU/mL and Type II with max. 1000 CFU/mL.
! The recommended protocol for microbial content was adopted from “Standard Methods for
the Examination of Water and Waste Water” published by the American Public Health
Association and the American Water Works Association. The working group believes that
the technology exists to meet the specifications for reagent water described in this
guideline.
Section 9.1.3.2
51. In the second paragraph, we question the representativeness of a sample collected after 1
minute of allowing the water to flow, because we do not believe that this is the common
practice of laboratories.
! Upon consideration of the comment, the working group believes its recommendations to be
appropriate.
52. In the third paragraph, we believe that 50 to 100 mL would provide better quantitation.
! The working group believes that the protocol is consistent with the recommendations of
the American Public Health Association and the American Water Works Association;
therefore, no change was made.
Section 9.1.5
53. Our laboratory has recently added a continuous recirculating ultraviolet light system to our
deionized water system. In this system, deionized water flows continuously through all
the lines and up to the base of each faucet and then through a 3'-long cylinder that has a
germicidal light running the length of it. All the water passes through a 0.2-micron filter.
All the polyvinylchloride plumbing was replaced with new tubing, and all faucets were
replaced with new plastic stems and mechanisms.
The installation of this system was precipitated earlier this year by the excessive colony
forming units per milliliter (CFU/mL) of bacteria we started picking up after applying NCCLS
guidelines for testing the microbial content of our presumed Type I water. Previously, we
had incubated our petri dishes for only 24 hours at 37 BC. NCCLS guidelines call for this,
plus another 24 hours at ambient temperature (23 E ±3 BC). Since applying this
incubation regimen, we have consistently found several hundred CFU/mL at each faucet
using NCCLS guidelines for collection. In all cases, this testing has been done immediately
after the system has sanitized the lines with peroxyacetic acid for 1 hour. After incubation
at 37 BC, there is no growth. It all appears after sitting another 24 hours at ambient
temperature.
The system manufacturer has also cultured our irradiated water before and after sanitizing.
Their normal protocol is to incubate for 24 hours at 37 BC. Culturing one of our faucets
before sanitizing, they found 40 CFU/mL. If they swabbed the outside of the faucet with
an alcohol swab before sampling, counts were reduced to 10 CFU/mL. If they let both of
these plates stay for an additional 24 hours at 37 BC, the results were too numerous to
count. After sanitizing, the sample showed no growth after 24 hours at 37 BC. They did
not subject this latter sample to longer incubation.
Your manual states that Type I water is a fleeting entity. From our observations, it
appears impossible to attain, at least from a microbiological standpoint using your
standards. Even though we are apparently not using Type I water, none of us can ever
recall having any technical problems that could be traced to water quality any place in the
laboratory over the last 5 years. The only thing that changed is the standard being applied
to the water, i.e., incubation for 48 hours instead of 24 hours. The majority of our
reagents contain bacteriostatic chemicals or are maintained at refrigeration temperatures
during their lifetime.
Also, we have no procedures that even come close to using our water under the 48-hour
incubation conditions specified in your manual and I suspect that this is true for most
laboratories in the country. Could we perhaps be applying a standard for water quality
that is unrealistic compared to how water is used in a typical laboratory?
Currently, we classify all of our procedures as requiring Type II or Type III water. It is our
current feeling that the water in the plumbing lines of our DI system defining the plumbing
lines as a primary container. Traditionally, laboratories have thought of their water quality
as being Type I. Now, it appears another tradition will bite the dust.
! The recommended protocol for microbial content was adopted from “Standard Methods for
the Examination of Water and Waste Water” published by the American Public Health
Association and the American Water Works Association. The working group believes that
the technology exists to meet the specifications for reagent water described in this
guideline.
54. Thank you for the opportunity to comment on what is generally an excellent guideline. Our
principal concern is with the temperature employed for microbiology testing. Most
“standard” laboratory tests are for pathogenic bacteria, while contamination with
nonpathogenic bacteria is more likely in most high-purity water systems. Standard
temperatures used in microbiology are actually inhibitory to many nonpathogenic
organisms. A better range is 30 to 35 BC. Some water laboratories actually use 28 to 30
BC.
Section 9.1.5.1
55. The abbreviation NIST should read, “National Institute of Standards and Technology.”
Section 9.1.6.2
56. Agar should be HPC (heterotrophic plate count), R2A, “standard methods” TSA, or other
media designed for “stressed” organisms in purified water systems.
Section 9.1.6.3
57. In the third paragraph, a pour plate is not optimal for water testing because the sample
size is too low.
! The working group believes that there is an alternate technique available for sampling
larger volumes. (Refer to the response to Comment 51.)
Section 9.2
! A new Section 9.6, including a discussion on endotoxin determinations, was added to C3-
A3.
Section 9.3
59. The guideline should recommend testing pH of water routinely (e.g., weekly). This
document includes pH testing without recommending a minimum frequency.
! The working group believes that it is the responsibility of each laboratory’s quality
assurance guidelines to indicate a testing frequency. If desired, in-line instruments are
commercially available.
Section 9.4.1.1
60. This may become one of the more popular NCCLS guidelines with worldwide applicability.
With the increasing use of sensitive, modern laboratory instrumentation and the improved
immunochemical in vitro diagnostics, the error caused by low-quality reagent water (and
possibly wash water) on laboratory results could be greatly amplified in developing
countries.
Instructions for the use of, and schematics for building, a simple, battery-operated, and
inexpensive resistivity meter would greatly add to the value of this guideline.
61. I question the stipulation that a meter should be capable of measuring only higher than
specification.
Section 9.4.2
62. Water samples for quality testing should be taken from the point of use. In-line meters
should be installed near the point of use.
Section 9.4.3
63. We are told that Type I water should have a resistivity of at least 10 megohms.
Meanwhile, Section 9.4.3 recommends that the resistivity meter may be calibrated with a
0.01 mol/L KCl solution, which is expected to have a resistivity of 0.707 kohm-cm. I
cannot see how a calibration performed at 0.707 kohm-cm can be used to validate
measurements of Type I water with a resistance of ~1.5 orders of magnitude greater than
the calibration value.
! Section 9.3.3 (which was Section 9.4.3 in C3-A2) and Section 9.3.1.3 are revised.
64. If there is no other way to independently validate water resistivity, then how may this
paragraph speak of the relative accuracy of resistivity meters? Perhaps the word
“precision” ought to be substituted here.
Section 9.5
65. Our procedure (a copy of which was provided to the NCCLS Executive Offices) for the
annual check of soluble silica embodies the ideas and changes we believe necessary to
make C3-A2 useful on a routine basis.
! The working group appreciates any improvements to the procedure and encourages
publication of the procedure for future reference. Section 9.4 of C3-A3 recommends the
use of an appropriate reference laboratory or commercially available kits and it provides the
existing procedure as an alternate.
Section 9.5.3
66. The descriptions of the formulations of two reagents used for the "Soluble Silica" assay
need attention. The description for preparation of the 1-amino-2-naphthol-4-sulfonic acid
(ANSA) solution calls for preparation of a 30% solution of sodium hydrogen sulfite. This is
significantly above the solubility point of this compound. I, and others in my corporation
who prepare this ANSA solution reagent, have found that the only way to create a
solution, instead of a suspension, is to slightly heat about 180 mL of water, and, to this,
add the sodium sulfite and ANSA itself, let dissolve while stirring, and then slowly add the
sodium hydrogen sulfite directly to this solution. This procedure results in a true, clear
solution. The procedure, as written in C3-A2, results in a cloudy suspension, which settles
out quickly.
The other problem is with the description of the oxalic acid solution. The molarity of this
10% solution is listed as 1.11 millimolar. Obviously it should be in the molar range.
Additionally, oxalic acid is the common name given to oxalic acid dihydrate. The dihydrate
should be specified in the text, for clarity.
! This protocol was revised to be consistent with the ASTM D859-88 procedure.
Section 9.6.1
67. Most clinical chemistry laboratories have some access to a decent spectrophotometer.
Why not give us the procedure and let the individual users decide if they can actually use it
or not?
Bibliography
68. We could not find a single location in Canada for Mrs. Winstead's often cited article,
"Reagent Grade Water: How, When and Why? American Society of Medical Technologists,
Austin, Texas. Steck Co. 1967. " Repeated requests to U.S. libraries were unsuccessful
to date. Would you kindly provide us with the necessary information to get a copy of the
above-mentioned article? It would be most appreciated if you could.
! Interested persons should contact the American Society for Clinical Laboratory Sciences
(formerly American Society for Medical Technology) for updated information (7910
Woodmont Avenue, Suite 1301, Bethesda, Maryland 20814).
Appendix C
69. We recommend that the document states that preventive maintenance procedures must
specify a schedule for the replacement of filters or other critical components.
70. This document must be withdrawn or discontinued and appropriate corrective action taken.
The problem with a simple revision process is that it tends to indicate that the essence of
the document is correct and that some improvements have been made. In fact, the basic
document is incorrect and is fundamentally a promotion of Type I water systems.
! The working group believes that the essence of the document is correct and improved its
utility by addressing the comments in this summary. Other editorial changes were made to
clarify the intended use of Type I water and commercially available reagent water. This
guideline describes water of three specific levels of quality (Types I, II, and III) and the
methods of producing and testing such water. The classification and specifications are
designed to enable laboratory scientists and supporting industries to specify the quality of
water to be used in such procedures as reagent preparation, reconstitution of lyophilized
materials, sample dilution, etc. No one specific method is recommended for producing
purified water. A single method or combination of methods may be used satisfactorily,
provided that the end product meets the required specifications stated in this guideline.
The working group does not recommend withdrawal of the guideline.
DI1-A2 Glossary and Guidelines for Immunodiagnostic Procedures, Reagents, and Reference
Materials—Second Edition; Approved Guideline (1992). DI1-A2 addresses common
terminology and basic methodology for immunodiagnostic procedures.
EP7-P Interference Testing in Clinical Chemistry; Proposed Guideline (1986). EP7-P offers
background information and procedures for characterizing the effects of interfering
substances on test results.
‡
Proposed- and tentative-level documents are being advanced through the NCCLS consensus process; therefore, readers
should refer to the most recent editions.
NOTES
NOTES
NOTES
NOTES
NOTES
NOTES