Type 1 type 2 and feed water in plain language

A method line that says "use type 1 water" has not yet chosen a document, a test, or a sample tap. Type 1, type 2, and feed water answer different questions, and laboratories get into trouble when one nickname is asked to do all three jobs. This guide is the decision procedure for writing the line so a second person can follow it. The cluster reference for contaminants a meter cannot see is laboratory water types and where they fail. Once the water is in a buffer, the pH workflow is preparing a buffer and checking pH.

Salts that will be dissolved in that water are specified from the reagents and chemicals catalogue. The vessel that receives a fresh dispense is a beaker or flask you can rinse and account for. Put the grade, the sample point, and the tests on a quotation request when the system or the certificate is what you are buying.

The decision this guide supports

You are deciding what to write beside "water" in a method, a batch record, or an enquiry. The useful decision has three parts. Which published specification, if any, applies. Whether you mean water entering the system or water leaving it. Which contaminant would actually spoil the assay. If you cannot answer those three, the bottle can still be wet, and the record is still incomplete.

Bench speech collapses this. People say type 1 for the polished dispense, type 2 for the general laboratory loop, and type 3 for rinse water or an autoclave feed. Those habits are local. They are not a licence to treat the Arabic numerals as an international scale in which 1 is always purer than 2, and 2 is always purer than 3, on every test.

Feed water is the water entering the system

Feed water is the stream that arrives at the purification equipment before carbon, softening, reverse osmosis, electrodeionisation, or a final polisher. It might be municipal water, a building purified loop, or the outlet of an earlier skid. It is defined by position, not by purity. Calling the final dispense "feed" because a cell-culture lab uses the word feed for media will mix two vocabularies. In this cluster, feed water enters the water system.

Product water is what you collect after the steps the skid actually runs. A resistivity cell on the product line reports the ionic quality of that product at that moment. It does not describe the feed, and a hardness measurement on the feed does not describe the product. Pretreatment is chosen for the feed. The assay specification is chosen for the product, at the point of use.

Write the sample point in plain language: "municipal feed at the skid inlet" or "dispense at the polisher, taken into a rinsed flask and used at once." A carboy filled yesterday is a third sample. Storage is a process step, whether or not anyone meant it to be.

ASTM types and ISO grades use different tests

Two documents dominate laboratory shopping language, and they do not number water the same way. ASTM D1193 specifies reagent water as types. In the tables laboratories usually cite, Type I has a resistivity minimum of 18 megohm-centimetres at 25 °C, which is a conductivity maximum of 0.056 microsiemens per centimetre. Type II has a resistivity minimum of 1 megohm-centimetre. Type III has a resistivity minimum of 4 megohm-centimetres. The ionic line is therefore not a staircase that gets worse as the Roman numeral increases. Type III is ionically tighter than Type II, while the preparation notes, filtration, and chemical limits still differ. A memory aid of "I better than II better than III" will mis-order the resistivity row.

ISO 3696 specifies grades for analytical laboratory water, not ASTM types. Grade 1, Grade 2, and Grade 3 are the names. The tests include conductivity and, depending on the grade, oxidisable matter, ultraviolet absorbance, residue on evaporation, and silica. Grade 1 is commonly cited with a conductivity maximum of 0.1 microsiemens per centimetre at 25 °C. That ionic ceiling is looser than ASTM Type I. Water that meets an ISO Grade 1 conductivity line has not automatically met an ASTM Type I resistivity line. The other columns are not interchangeable either. ISO's oxidisable-matter and absorbance tests are not an ASTM total-organic-carbon line written under another name.

There is also a Type IV in the ASTM document, and the ASTM specification can attach further microbial and endotoxin expectations as additional requirements. Those lines are extra tests. They are not implied by a type number spoken in a corridor.

What a resistivity near 18.2 actually says

Very pure water, limited by the ions from water's own dissociation, sits near 18.2 megohm-centimetres at 25 °C. That figure is an ionic statement. Temperature belongs in the sentence because resistivity moves with temperature, and a comparison to 18 or 18.2 without 25 °C is a different measurement. A cell that reads near that ceiling is telling you that dissolved ions are very scarce at the sensor.

It is not telling you the total organic carbon. Many organic molecules contribute little conductivity. It is not an endotoxin result. Lipopolysaccharide is not an ionic purity assay. It is not a nuclease assay. An enzyme that nicks RNA can ride in water the resistivity cell still likes. It is not a sterility test. A plate count, a membrane filtration, or whatever microbial method your quality system names is a different operation. Particle filtration specified for a type is likewise a process requirement, not something the resistivity number proves after the fact.

pH is a poor release test for this water. Ultrapure water has almost no buffer capacity, dissolved carbon dioxide moves the reading, and the electrode is being asked to work in a solution with almost no ionic strength. High-purity tables often leave pH unspecified for that reason. Use the pH meter on buffers, where it has something to measure. The water page and the buffer page split that work on purpose.

A workflow you can file

Start from the assay, not from the skid brochure. Write the failure you are preventing: a trace-ion blank, a chromatographic ghost peak, a nuclease-sensitive RNA step, an endotoxin-sensitive culture, or a glassware rinse that only has to be wet and low in salt. Pick the lightest specification that covers that failure. Over-specifying everything on every carboy teaches people to ignore the words.

Then attach a document name where a document applies. "ASTM D1193 Type I, resistivity at 25 °C, plus the TOC limit in our method" is a specification. "Type 1" is a hint. If your building loop was qualified against ISO 3696 Grade 2, write Grade 2. Do not translate it into Type II in the same sentence unless you have compared the tests and accepted the gaps.

Then name the sample point and the freshness rule. Polished water taken at the dispense and used for the blank is one rule. Water drawn into a rinsed flask from beakers and flasks and held capped for a defined time is a looser rule, and it has to be chosen. Feed-water tests, if you are buying or troubleshooting a skid, are recorded on the inlet: hardness as calcium and magnesium, and disinfectant as chlorine or chloramine. Those inlet facts do not belong in the product-water sentence.

If the control fails, branch. A product resistivity below the method limit stops the sensitive work and sends you to the polisher, the dispense line, and the vessel. A passing resistivity with a failing assay blank sends you to the contaminant the cell cannot see: organics, a dirty container, a nuclease, or endotoxin, whichever the blank suggests. A feed that has changed, with new chlorine taste reports from facilities or a rising hardness, is a pretreatment problem. Do not "fix" it by reprinting the product label.

When the words and the result disagree

A passing product meter and a failing blank is the common split. Change one variable at a time. Rinse or replace the vessel, take a fresh dispense, and repeat the blank. If the blank recovers, the specification may have been fine and the storage was not. If the blank stays high, move to the test that matches the signal: organic carbon for a chromatographic ghost, a nuclease control for a vanished RNA band, endotoxin for a culture response that water chemistry does not explain.

A failing product meter with a convincing certificate from last month is a different branch. Certificates age. The sensor, the lamp, the cartridge, and the idle time are all allowed to change. Believe today's point-of-use reading for today's ionic claim, and believe a dedicated assay for anything else.

A feed-water surprise, such as chlorine reaching a membrane that the manufacturer said must see dechlorinated water, is not solved by tightening the product specification. The product number is downstream of the damage. Deal with the inlet, and follow the skid manufacturer on pretreatment. That decision is developed further when hardness and disinfectant are the problem. Here, the rule is only: do not rename the feed as a type.

Safety, research use, and what a grade cannot approve

Water grades are analytical and operational categories. They are not a biosafety approval, a diagnostic claim, or a statement that a solution is safe to drink or to inject. If the assay touches infectious material, the institutional biosafety rules govern the work. Polishing does not disinfect a buffer that already contains an organism, and autoclaving a buffer is a separate decision from choosing Type I water to prepare it.

Chemical hazards sit in pretreatment, not in the polished dispense: acids, bases, and disinfectant-quenching chemicals belong to the skid procedure and the safety data sheet. This page does not authorise a cleaning recipe. It tells you to keep feed, product, and specification from being the same word.

A warm room, a power cut, and an honest sample

In a warm laboratory the temperature displayed beside resistivity matters twice. The specification is written at 25 °C, and a carboy left open or warm will take up carbon dioxide and lose the ionic number it had at the dispense. After a power cut, a polisher that sat idle should be rinsed to the quality your method names before a sensitive blank is collected. The first water out of a stagnant line is a different sample from a flowing dispense. Write that rinse as a step if the work is fussy. Do not invent a city-wide feed composition. Measure the feed you have, and put those measurements in the record.

What the enquiry should carry

Ask for the specification by document and grade or type, the product resistivity or conductivity at 25 °C, and any TOC, microbial, endotoxin, or filtration claim as its own line. State the daily volume and the assays the water must survive. For a system, add the feed you know: hardness, chlorine or chloramine, and where the inlet sample is taken. Ask how the unit is returned to quality after it has been idle. A single word such as ultrapure cannot be checked on receipt. The reagents and chemicals catalogue is where dissolved reagents are specified separately, so the water grade is not asked to imply a salt purity it never measured.