Guide
How do you prevent dezincification in brass water fittings, and how should a drawing call out DZR brass?
Dezincification is a form of selective corrosion in which water leaches the zinc out of a brass part and leaves behind a weak, porous layer of copper that keeps the original shape. On a water fitting it shows up as pink or coppery patches, white crusts, weeping at threads, and finally a body or stem that leaks or snaps. It is prevented by choosing a dezincification-resistant (DZR) alloy, processing it so it stays resistant, and writing that requirement on the drawing in a form a supplier cannot misread.
By Saola Manufacturing engineering team · Published · Updated
This guide is for design, quality and sourcing engineers who own a brass valve, fitting, manifold or insert in potable or process water and either have a field return that looks like dezincification or want to make sure the next lot does not produce one. It covers how to recognize the problem, what drives it, how to confirm it on a failed part and on incoming material, which alloys resist it, how the lead-free rules interact with DZR, and the exact notes to put on the drawing and the RFQ.
On this page
What does dezincification look like on a brass fitting?
The surface turns from yellow to a dull pink or copper red, often under white or gray-white crusts of zinc corrosion product. The affected metal is soft and porous, so a fitting can weep through its wall or crack at a thread long before it looks badly corroded.
Corrosion researchers describe two shapes of attack. In layer dezincification the zinc is removed fairly evenly across the wetted surface, so the part loses strength over a broad front. In plug dezincification the attack is local and grows deep, narrow plugs of porous copper through the wall, which is how a fitting can leak at one spot while the rest looks sound. Either form can be accompanied by meringue: a bulky, tenacious white build-up, usually a basic zinc carbonate, that grows inside the fitting and can restrict or stop flow. Both forms can appear on the same fitting (Zhang, Virginia Tech, 2009).
Typical field signs, roughly in the order they appear:
- White or off-white crusts on the outside of a joint, nut or valve body, often at a thread or a crevice where water seeps.
- Pink or coppery color on a scraped or cut surface, instead of bright yellow brass.
- Reduced flow through a small bore, from meringue buildup inside the fitting.
- Weeping through the body wall or at the threads with no visible crack.
- A valve stem, spindle or compression nut that shears or crumbles during service work, because the remaining copper skeleton has little strength.
None of these signs alone proves dezincification. Pink staining can also be redeposited copper, and white crusts can be scale from hard water. The only conclusive check is a metallographic section, covered below.
What causes dezincification in brass water fittings?
The main cause is an alloy with too much zinc in a susceptible phase, usually a duplex alpha-beta free-cutting brass with no inhibitor. Water chemistry, temperature and stagnation then decide how fast the attack runs.
Brass is a copper-zinc alloy, and its zinc content sets its structure. Below about 30 percent zinc a brass is generally single-phase alpha; between about 30 and 40 percent it is a duplex alpha-beta brass, the family most free-cutting and forging brasses belong to (Zhang, 2009). Resistance rises as the structure moves from all-beta to duplex to all-alpha, and alloys with less than 15 percent zinc rarely dezincify. Small additions of arsenic, antimony or phosphorus inhibit attack on all-alpha brass but are ineffective on duplex and all-beta brass, which is why an inhibitor alone does not make a duplex brass resistant (Davies, Copper Development Association, 1993). The same boundary appears in the standard test: per ISO 6509-1, the test is intended for copper alloys with a zinc mass fraction of more than 15 %.
On the water side, laboratory work published in the Journal of the American Water Works Association found a critical pH range of 8 to 9 for meringue dezincification, found that chloride increases dezincification propensity, and attributed the protective effect to alkalinity rather than hardness (Zhang and Edwards, via EPA HERO). The same research group points out that the classic Turner diagram, which predicts risk from chloride and temporary hardness, came from British waters that blocked hot-water pipes with meringue, covers only that one form of attack, and leaves out modern inhibitors and disinfectants (Zhang, 2009). Leaks are also reported more often in hot systems than in cold ones, and coupling brass to a more noble metal such as copper raises the probability of attack (Hendrix Group archive). Treat water chemistry as a reason to specify DZR, never as a reason to skip it.
| Cause | Why it drives attack | What to check first |
|---|---|---|
| Wrong alloy: uninhibited duplex free-cutting brass | Beta phase loses zinc readily and nothing inhibits it | Alloy on the drawing and on the mill certificate |
| Right alloy, wrong condition after hot working | Forging or hot stamping leaves beta phase that the anneal was meant to remove | Heat treatment record and an ISO 6509-1 result on the finished part |
| Substituted bar or a lead-free alloy assumed to be DZR | Lead-free and DZR are separate properties; some lead-free brasses are beta-rich | Alloy designation and chemistry against the drawing |
| Aggressive water | Chloride and a pH in the meringue range raise attack rate | Utility water report for chloride, pH and alkalinity |
| Hot service or a galvanic couple | Elevated temperature and contact with a more noble metal such as copper raise the probability and rate of attack | Service temperature, mating materials, crevices and low-flow sections |
How do you confirm dezincification on a failed part or an incoming lot?
Section the part through the suspect area and look at it under an optical microscope: dezincified metal shows as a distinct copper-colored porous zone, and its depth can be measured. For incoming material, ask for an ISO 6509-1 test result and check the alloy chemistry against the drawing.
On a field return. Cut through the leak point or the white crust, mount and polish the section, and examine it under a microscope. The attacked zone reads as a red, spongy band that stands apart from the sound brass under it, and the section shows whether the attack is a layer or deep local plugs, and whether it followed one phase. If the laboratory has a scanning electron microscope with EDS, an EDS spot on the spongy band that shows copper with most of the zinc gone confirms dezincification rather than copper redeposited on the surface. Record the alloy, the service water and temperature, and the age of the installation, because a metallurgist needs all three to separate an alloy problem from a water problem.
On incoming material. ISO 6509-1 is the accelerated laboratory method: a prepared test surface cut from the material or finished part is exposed for 24 h to a freshly prepared 1 % copper(II) chloride solution held at 75 °C, then sectioned at right angles to that surface, and the maximum and average depth of dezincification are measured under an optical microscope (ISO 6509-1:2014 preview). The method gives a depth, not a pass or fail. The acceptance criteria are in a separate document, ISO 6509-2:2017, which applies to semi-finished products and finished parts such as fittings and valves but not to complex products such as flow meters or pump parts. ISO 6509-2 also states that it is not intended to validate dezincification after a failure in service, so judge a field return by its own section, not by a test pass or fail. In Australia and New Zealand, dezincification resistance of copper alloys is covered by AS 2345, and resistant fittings are commonly identified with a “DR” mark; check the current edition for the exact test and marking rule.
On chemistry. A handheld XRF gun confirms the alloy family and the lead level quickly, as the PMI testing guide explains. It is a poor tool for proving a small arsenic inhibitor addition, partly because arsenic and lead lines overlap in the spectrum, so rely on the certificate chemistry from the bar producer or a laboratory analysis for that element. Chemistry alone also cannot show whether a forged part received its anneal; only the ISO 6509-1 section shows that.
A quick diagnostic checklist for a return:
- Does the drawing name a DZR alloy, or just “brass”?
- Does the mill certificate match that alloy, including the inhibitor element?
- Was the part machined from bar, or forged or cast and then machined?
- If forged, is there a heat treatment record for that lot?
- Does a section show dezincification, and is it confined to one phase?
- Is the water report in the higher-risk range for chloride and pH?
- Was the fitting brazed or otherwise reheated after its anneal, which can undo the resistance?
- Is it coupled to copper or another more noble metal, or in hot-water service?
If a lot is suspect, quarantine the remaining stock, pull samples from each heat or forging lot, and send them for ISO 6509-1 testing against the ISO 6509-2 criteria before releasing them. Check fittings of the same part number installed in the same water, since an alloy or anneal problem rarely affects only one piece.
Which brass alloys prevent dezincification, and how do you choose one?
Use an alloy that is resistant by design and verified by ISO 6509-1: an arsenic-inhibited DZR brass such as CW602N, a silicon brass such as CW724R or C69300, or a low-zinc alloy such as silicon bronze. For US potable water the alloy must also meet the lead-free limit, which rules out the leaded DZR grades.
The choice comes down to three questions. Does the part touch drinking water in a market with a lead limit? Is it machined from bar, or hot forged or cast and then machined? And how much machinability does the part need? Arsenic-inhibited DZR brass such as CW602N gets its resistance from chemistry plus a structure that is close to all alpha. Its UK predecessor, CZ132, contains enough beta phase to hot stamp or extrude; forging stock is supplied without the final heat treatment, and the fittings maker heat treats the forged part in the range 450 to 550 °C to convert it to an all-alpha structure protected by the arsenic (Hendrix Group archive). Skip that heat treatment, or reheat the finished part above it, for example during brazing, and a forged body can fail the test even though its chemistry is correct. A small amount of finely dispersed beta, around 5 percent, does little harm in an otherwise all-alpha brass (Davies, 1993).
Lead-free and dezincification resistance are separate properties, and confusing them is a common cause of field failures. Under the US Safe Drinking Water Act a wetted surface is lead free when its weighted average lead content is not more than 0.25 percent (40 CFR 143 subpart B), so a leaded DZR grade like CW602N, with about 2 percent lead, will not normally meet it. Silicon brass such as C69300 has a lead limit of 0.09 percent and lists water valve bodies and stems among its typical uses on the Copper Development Association’s C69300 alloy page. Silicon is not a guarantee on its own: early CDA work found silicon brasses not immune to dezincification, though close to heat-treated arsenical DZR brass and improving with silicon content (Davies, 1993), so a silicon brass still needs an ISO 6509-1 result for the lot. A simple lead-free yellow brass with high zinc and no inhibitor, by contrast, is beta-rich and should not be assumed to be DZR. The lead-free brass guide covers the NSF 61 and NSF 372 side of the decision.
| Alloy (EN / UNS) | How it resists dezincification | US lead-free (40 CFR 143) | Best fit |
|---|---|---|---|
| CW614N (CuZn39Pb3) / C38500, CW603N / C36000 free-cutting brass | Does not; duplex structure, no inhibitor | No (leaded) | Dry or non-water parts only |
| CW602N (CuZn36Pb2As) / C35330 | Arsenic inhibitor plus near all-alpha structure; forgings need the post-forging anneal | No (leaded) | Markets that accept leaded DZR; machined from bar or forged and annealed |
| CW511L (CuZn38As) | Arsenic inhibitor plus producer processing; at this zinc level the structure, and so the resistance, depends on heat treatment | Check the weighted-average calculation | Low-lead DZR parts with modest machining, with a lot test |
| CW724R (CuZn21Si3P) / C69300 | Silicon brass; offered as dezincification resistant, but not immune by composition alone | Yes, when the finished part passes the calculation | US and EU potable water parts, bar or forged |
| Lead-free high-zinc yellow brass without inhibitor | Not by composition; beta-rich | Yes | Do not use where DZR is required |
| Silicon bronze or gunmetal (low zinc) | Zinc at or below about 15 %, where dezincification is rare | Depends on grade (gunmetals are often leaded) | Fittings where zinc loss must be ruled out |
Two process points matter as much as the alloy. First, an alloy name on its own is not proof: the same designation can be resistant or not depending on heat treatment, so ask for an ISO 6509-1 result tied to the lot. Second, residual stress is a separate risk in brass. Ammonia-bearing environments can crack stressed brass by stress corrosion, so a stress-relief anneal after heavy cold work is often called for alongside the DZR requirement.
How should a drawing call out DZR brass so the supplier cannot misread it?
Name the alloy by its full designation and product standard, require dezincification resistance tested to ISO 6509-1 with acceptance to ISO 6509-2, and state the lead-free requirement separately. A note that says only “brass” or “DZR brass” leaves the alloy, the heat treatment and the proof to someone else.
A callout that closes the common gaps looks like this:
- Material. “CW724R (CuZn21Si3P) per the applicable EN product standard, or C69300. No substitution without written approval.” Name one alloy and, if you accept an equivalent, name that too.
- Dezincification resistance. “Material and finished part shall be dezincification resistant, tested per ISO 6509-1, accepted per ISO 6509-2. Test report required per heat or lot.” For Australia, add the AS 2345 requirement and the DR mark.
- Heat treatment. For a forged or hot-stamped DZR body: “Heat treat after hot working per alloy producer’s DZR procedure; record per lot. No brazing or other reheating of the finished part above the heat treatment temperature.”
- Lead. “Wetted surfaces to meet the lead-free definition of 40 CFR 143; finished product to NSF/ANSI/CAN 61 where required.” Keep this line separate from the DZR line so neither is read as covering the other.
- Marking. If your market uses a mark such as DR, show its location and size on the drawing.
- Documents. “Mill certificate with chemistry including inhibitor elements; ISO 6509-1 report.” The mill test report guide shows how to trace a certificate to the parts you receive.
On the RFQ, add the service water and temperature, whether the part will be forged, cast or machined from bar, and the markets it ships to. Those three facts let a supplier propose the right alloy instead of the cheapest brass that machines well.
How is a DZR brass part handled on a Saola order?
Machining from bar runs at our lead CNC plant, while forging, casting, annealing and plating go to partner plants we manage. Each quote names the production site and the country of origin.
Turned and milled brass fittings are cut at our lead CNC plant, T.K Viet Nam Precision Company Limited, on the equipment described on the CNC turning and CNC milling pages. Brass is not on the published bar list on the turning page today, so name the exact DZR grade and its product standard on the RFQ, and the quote states whether that bar can be sourced with a mill certificate showing the inhibitor chemistry. Silicon bronze, a low-zinc option, is on that list.
A forged or cast body, its anneal, and any plating are steps outside the lead CNC plant. Where they are part of the job they are placed with partner plants we manage, which are not necessarily in Vietnam, and the quote says where each step is done; see surface finishing for the finishes on offer. The ISO 6509-1 report comes from the bar producer or an independent laboratory, so list it as a required document on the RFQ. ISO 9001 certification is in progress, not held. Send the drawing with the material, DZR and lead notes through the quote form.
Frequently asked
Is lead-free brass automatically dezincification resistant?
No. Lead content and dezincification resistance are controlled by different things. Silicon brasses such as C69300 are both, but a high-zinc lead-free brass with no inhibitor is not. Specify both properties separately on the drawing.
Can a dezincified fitting be repaired or left in service?
It should be replaced. The dezincified copper layer has little strength, and the damage usually extends further than the surface suggests. Replace it with a DZR part and check whether neighboring fittings in the same water are the same alloy.
What do the CR and DR marks on brass fittings mean?
Both identify dezincification-resistant fittings. In the UK, fittings made from DZR brass to BS 2872 and BS 2874 have been identified by a CR mark embossed or engraved on the side; DR is the mark commonly used in Australia and New Zealand under AS 2345. A US drawing has no equivalent mark, so the requirement must be written out.
Does plating a brass fitting stop dezincification?
Not reliably on the wetted side. Plating on a bore or thread is easily damaged in assembly, and water reaches the base metal at any break. Choose a resistant alloy for wetted surfaces and treat plating as cosmetic or for the outside only.
Is dezincification the same as stress corrosion cracking in brass?
No. Dezincification removes zinc and leaves porous copper; stress corrosion cracking, sometimes called season cracking, splits a stressed brass part, typically in the presence of ammonia. A part can suffer either, so DZR and stress relief are separate requirements.
Sources
- ISO 6509-1:2014, Determination of dezincification resistance of copper alloys with zinc, Part 1: Test method (preview), ISO, 2014
- ISO 6509-2:2017, Part 2: Assessment criteria (scope), ISO, 2017, confirmed 2022
- AS 2345-2006 (R2016), Dezincification resistance of copper alloys, Standards Australia, 2006
- Zhang and Edwards, Effects of pH, chloride, bicarbonate, and phosphate on brass dezincification, Journal AWWA, 2011, via EPA HERO
- Zhang, Dezincification and brass lead leaching in premise plumbing systems, Virginia Tech, 2009
- Davies, A note on the dezincification of brass and the inhibiting effect of elemental additions, Copper Development Association, 1993
- C69300 alloy properties, Copper Development Association, accessed October 2026
- Dezincification (brasses archive), The Hendrix Group, accessed October 2026
- 40 CFR part 143 subpart B, Use of lead free pipes, fittings, fixtures, solder, and flux for drinking water, eCFR, 2020 rule, accessed October 2026