Distilled/Deionized Water for District Operations

In Reply Refer To: Mail Stop 412
December 20, 1991


Subject: Distilled/Deionized Water for District Operations


This memorandum discusses minimum requirements for the quality of distilled and/or deionized water for use by Districts in normal day-to-day operations. The term "deionized water" will be used throughout the rest of this memorandum with the understanding that the water referred to is produced either by distillation or deionization or a combination of both, or by some equivalent process such as reverse osmosis.


1. Final rinsing of laboratory glassware and other appropriate equipment that has been previously washed with tap water and detergents. (Note: when working with organic compounds, the deionized water rinse will usually be followed by a final rinse with an organic solvent and/or by baking the equipment [refer to OWQ Technical Memorandum 91.09] ).

2. Rinsing electrodes used to measure pH, specific conductance, dissolved oxygen, and other specific ions to assure that all traces of previous sample have been removed before placing the electrodes into a new sample.

3. Rinsing sample bottles immediately prior to filling with sample in those instances where the investigator determines that use of sample water for rinsing would be inappropriate (such as samples containing high sediment concentrations [refer to OWQ Technical Memorandum 92.01]). Data on maximum potential carryover (contamination) of the sample with deionized water are provided in Appendix II of this memorandum.

4. Washing filtration apparatus and the membrane filter to remove contaminants and to flush any wetting agents from the membrane filter. This wash with deionized water must be followed by rinsing with an adequate amount of sample water to assure removal of the deionized water from the membrane filter and filtration apparatus to prevent dilution of the sample by the deionized water during sample collection.

5. Preparation of a nitric acid rinse solution to clean equipment to be used for sampling trace elements, both in the laboratory and at field sites. (Details of this use of deionized water will made available to Districts soon).


District-produced deionized water should not be used for any of the following uses unless additional analytical steps have been taken to assure that the water is adequate for the specific purpose and is so documented.

1. As feedstock for preparation of sterile water for subsequent preparation of bacterial sampling media and other sterile solutions.

2. To prepare chemical reagents of any kind for use in chemical analysis, except as noted above to prepare acid rinses for trace- element sampling equipment.

3. To prepare specific-conductance standards.

4. To prepare blank water standards for quality-assurance studies. In addition to this memorandum, the Office of Water Quality soon will release two Technical Memoranda on the production/procurement of quality-assured special waters for District operations. The subjects of these memoranda are: (a) Blank Water for Conventional Inorganic Analyses, and (b) Blank Water for Organic Analyses. We ask that the Water Quality Specialist in each District compile a set of these three memoranda and share them with every employee who needs the information.


Most Districts produce water for their Field Service Units by running tap water through commercial ion-exchange cartridges and monitoring conductivity for cartridge replacement. In a few instances, however, Districts have resorted to using water purchased from supermarkets advertised as being "distilled water." The reasons given for using supermarket water are: (a) the cost of providing separate field-office installations is too high and deionized water can not be conveniently supplied from the District unit, or (b) the high dissolved-solids content of the available tap water consumes ion-exchange cartridges too rapidly. The use of supermarket water in water-quality operations is risky because of the unknown quality assurance employed in its production (see next paragraph). Districts should not use supermarket distilled water except in an emergency, and then only if each separate bottle is checked for conductivity immediately before using.

The American Society for Testing and Materials (1990 Annual Book of ASTM Standards, Section 11, Water and Environmental Technology, Designation: D 1193 - 77, p. 45-47) specifies four grades of reagent water ranging from Type I to Type IV. The maximum electrical conductivity of Type I through Type IV waters are 0.06, 1.0, 1.0, and 5.0 microsiemens per centimeter (µS/cm). For comparison, a recent sampling of supermarket distilled waters from around the country found conductivities ranging from about 1 to 65 µS/cm. Laboratory supply companies and water service companies more commonly refer to the resistivity of deionized water. The electrical resistivity of Type I through Type IV waters is the reciprocal of the conductivity values, namely: 16.67, 1.0, 1.0, and 0.2 megohms per centimeter.

The minimum standard for District deionized water is set by the most restricted usage; in this case, its use to prepare a nitric acid rinse for sampling equipment used in trace-element work (usage 5 cited in the list above). For this usage, and for all other normal District water-quality operations, as listed above, ASTM Type I, II, or III water with a maximum conductivity of 1.0 µS/cm is adequate. Therefore, we recommend that Districts assure that their deionized water supplies (optional sources described in Appendix I) have a maximum conductance of 1.0 µS/cm. In addition, we recommend that District deionized water be filtered through a 0.45 mm membrane filter (an in-line filter cartridge at the deionized water supply outlet is a simple way to achieve this) to equal ASTM standards for final polishing of the water. The purpose of this filter is to remove any ion-exchange resin beads that may migrate out of the cartridge and to remove many bacteria, thus enhancing shelf life of the water.

Ion-exchange water-treatment units, capable of producing 0.06 to 1.0 µS/cm water, can be obtained on the commercial market for approximately $250 per unit (Appendix I). This low price makes it easy to equip each field office with its own unit to assure that the more costly analyses of water samples reflect actual stream or aquifer conditions and not contamination. In regions where the high dissolved-solids content of the tap water causes rapid consumption of the exchange capacity of the ion-exchange cartridges, supermarket distilled water can be used to provide a suitable feed water for the ion-exchange unit. This should assure a long lifetime for the ion-exchange cartridge at minimal additional expense. Advice on specific sources and brand names of deionizing units, as well as on optimal operation of units, can be found in Appendix I.

When rigorously quality-controlled water is needed for inorganic or for organic blanks, such water can be obtained in limited quantities from the Ocala Laboratory or the National Water Quality Laboratory (NWQL) respectively (OWQ Technical Memoranda in preparation as described above).

The use of 1.0 µS/cm water for the normal District purposes listed above is unlikely to cause contamination of water-quality samples because little deionized water should be carried over to the sample bottle if washing and rinsing is carried out properly (Appendix II). It is important, however, to monitor the quality of District deionized water for conductivity to assure that this small carryover does not produce contamination of low-level samples.


We recommend that a blank from each deionized water unit be submitted to the NWQL shortly after installation of the unit. An analysis for schedule 1043, ICP trace-metal scan, and schedule 1106, low-ionic-strength nutrients, should be run. Provided that the same brand of resin cartridge is reinstalled at each cartridge changing, and no other changes are made to the deionized water unit, the results of these analyses should provide data on the maximum likely contamination with trace metals or trace nutrients. In addition, no less frequently than monthly, depending on feedstock quality and usage (more frequently with concentrated feedstock and heavy usage), blanks of deionized water should be tested for specific conductance using a conductivity meter or similar device that can measure accurately in the range of 0.1 to 1.0 µS/cm. If a conductivity meter of adequate sensitivity is not available, the blank should be sent to the NWQL or to Ocala for a low-ionic-strength specific conductance test (see Appendix I for sample requirements). This monthly test should be independent of any conductivity detection device supplied with the deionizing unit (meter, light, buzzer, or color change). The results of these quality-assurance checks should be kept in a log book near the deionizing unit and be made available on request to technical review personnel.

Deionized water produced in accordance with these guidelines should be usable for the cited District needs. As previously noted, if a District requires deionized water for preparation of QA blanks or spikes of organic or inorganic constituents (other than QA of the deionized-water unit itself), such water will be provided through the NWQL or Ocala as will be described in the noted forthcoming memoranda.

David A. Rickert Chief,
Office of Water Quality



Key words: deionized water, distilled water, demineralized water, water purification

This memorandum does not supersede any previous Office of Water Quality Technical Memorandum.

Distribution: A, B, S, FO, PO



Each District should have someone who is familar with the details of deionized water services. That person should first determine the quality of the feedstock (finished tap water in most cases). Because of the highly variable quality of drinking water supplied throughout the United States, knowledge of the quality of the feedstock, as well as its consistency with time, is vital to making an intelligent decision about a deionized water source.

By: (a) examining several laboratory catalogs for their offerings, (b) talking to local water service companies that may provide either bottled deionized water or that lease and service deionizing units, and (c) conferring with knowledgeable personnel in other Districts or at the National Water Quality Laboratory or Ocala Laboratory, the District "expert" should be able to select a suitable and economical deionized water source for the District's field service unit and for each subdistrict or field office.

Suitable mixed bed (includes both cation and anion resin beads) cartridges from most laboratory supply houses are approximately 4 inches in diameter and 18 inches long. Most will have a maximum exchange capacity listed as 1,000 grains of calcium carbonate or 1,150 grains of sodium chloride. A grain of hardness as CaCO3 is equivalent to 17.1 parts per million (ppm), so a 1,000 grain cartridge should deionize the equivalent of 17,000 ppm hardness. To calculate a conservative estimate of the quantity of water that can be treated by such a cartridge, divide the total capacity as CaCO3 in ppm (17,000) by the sum of cations and anions (not just the hardness) of the feedstock water in ppm (or mg/L). This provides an estimate of the exchange capacity of the cartridge for all ions, not just calcium and carbonate-bicarbonate. The volume calculated in liters of feedstock water will give a reasonable estimate of the quantity of deionized water the cartridge will produce. Please note that only about 85 percent of the maximum capacity of a cartridge is normally usable.

Two such cartridges and a mounting bracket for attaching to a wall above the field-service-unit sink cost about $175. An additional bracket and an adsorber cartridge to remove some organics and chlorine compounds will cost an additional $100. The adsorber pretreatment is strongly recommended to extend the life of the deionizer cartridges. If the deionizer cartridges do not incorporate a color-change indicator, a detection device (meter, light, or buzzer) should be purchased to provide a continuous visual or audible warning of deteriorating water quality. Please note however (as described under "Quality Assurance of District- Produced Deionized Water"), that no less than monthly measurements of completed water are required using a separate and sensitive conductivity meter.

The Ocala, Florida, laboratory currently stocks Cole-Parmer Ion-X- Changer filter cartridges and mounting brackets, and they can supply other brands if desired. The Cole-Parmer Ion-X-Changer "Research" cartridges stocked by Ocala are claimed to produce deionized water with a resistivity of 15 megohms-cm or better. Ocala offers a catalog of their available laboratory supplies, and they can be reached through PRIME EDOC MAIL at: OCALAMAN.

Most deionized water units consist of at least two identical resin cartridges in series. The correct location for the detection device (meter, light, or buzzer) is downstream from the first cartridge. This configuration assures that when the first cartridge becomes exhausted, the second cartridge will protect the deionized water supply until a cartridge change is made. When a cartridge change is required, the exhausted cartridge closest to the tap water intake (first cartridge) should be replaced with the next cartridge downstream, and a new, fresh cartridge added at the downstream end of the deionized water unit. This procedure will ensure using each cartridge to its maximum capacity.

Deionization systems tend to produce water of acceptable quality until the resins are nearly exhausted, at which point the water quality abruptly becomes unacceptable. Therefore, output water from the first cartridge should have a conductivity of about 0.1 µS/cm until the cartridge is nearly exhausted, at which point the detection device (meter, light, or color of the resin beads) will indicate that fact, and the first cartridge should be replaced as described above. Output water from the second (or last) cartridge should be monitored with a specific conductance meter periodically and the values recorded in a log book that is kept in the laboratory.

Some laboratory specific-conductance meters can measure accurately from 0.1 to 1.0 µS/cm and can be used by a District to monitor the output of the deionized-water unit. Otherwise, a deionized water blank should be sent to the NWQL or to Ocala for a low-ionic- strength specific-conductance test. Scrupulous cleaning of the sample bottle is required to obtain a good blank of deionized water for either chemical analysis or specific conductance.

The following cleaning steps should be followed:

  1. Fill the pre-cleaned sample container with deionized water,
  2. Rinse the cap with deionized water,
  3. Cap the container and allow to stand for about 5 minutes,
  4. Empty the container, and
  5. Repeat the process five or more times before sealing the cap on the sample.



by Julija M. Laenen, Portland, OR

Six new 250 mL and six new 500 mL polyethylene sample bottles, supplied by the Central Lab were used to determine what volume of water might be retained when the containers are rinsed with distilled water. First the empty containers were weighed with black caps on; then a 50 mL volume of distilled water was dispensed into each container using a 50 mL volumetric pipet. The container was then shaken and the water poured into a graduated cylinder to measure the volume poured out of the container. The total length of time to pour out the sample to the last drop was less than 5 seconds. I used a 100 mL cylinder for this measurement, but found that I could not read a difference in volume from the initial 50 mL used, so I capped the container and reweighed the container to determine the amount of water retained in the containers.

               Unrinsed       Rinsed
                 bottle         bottle        Difference
Container       (grams)        (grams)          (grams)

250 ml
1                23.9           24.0             0.1
2                22.7           22.8             0.1
3                24.3           24.4             0.1
4                23.8           24.0             0.2
5                24.8           24.9             0.1
6                23.8           23.9             0.1

500 ml
11               30.8           31.1             0.3
12               35.0           35.3             0.3
13               35.4           35.7             0.3
14               32.1           32.3             0.2
15               35.3           35.6             0.3
16               35.0           35.3             0.3
The Branch of Quality Assurance (BQA) recently completed a similar experiment with a 250 ml bottle. Their results are comparable to those listed above. BQA also notes that a 0.10 ml carryover (approximately 0.10 gram) into a 250 ml bottle results in a dilution of 0.04 percent and a carryover of 0.25 ml results in a dilution of 0.10 percent.