Arizona fabs keep pushing water recovery higher. TSMC Arizona says the existing in-house water centre at its north Phoenix site reaches 65% recycling today, and that its new Industrial Reclamation Water Plant, due in 2028, will start at 85% and aim for 90% or more (Arizona Technology Council, September 2025). Each step up in recovery shrinks the concentrate stream. None of them removes the salt. In the Phoenix area, deciding where that salt ends up is the hard part of semiconductor wastewater ZLD, and the answer depends on Arizona-specific rules, geology and water supply.
Brine Consulting is an independent industrial water advisory firm in Amsterdam and sells no treatment equipment, so we have no stake in whether your fab ends up with a crystallizer. We work on Arizona projects remotely or through a defined engagement: reclaim data, concentrate analyses and supplier offers come to us, and a recovery target and outlet comparison go back. Permits and the relationship with ADEQ and your city stay with your Arizona engineers.
For the reuse side of a fab (stream segregation, reclaim grades and recovery back to ultrapure water) see our guide to water reuse in the semiconductor industry. This page starts where that one stops: the concentrate.
Why recovery keeps rising in the Phoenix area
Three local pressures push Phoenix-area fabs toward higher recovery: a smaller Colorado River supply, very large fab water demand, and a sewer outlet that is already permitted and watched.
Colorado River cuts are now a fixed planning input. On 21 August 2026 the Interior Department issued the 2027-2028 Operating Guidelines and Record of Decision. Lower Basin deliveries will be cut by 1.25 million acre-feet in 2027. If the Lower Basin states implement their proposed sharing agreement, Arizona’s share is 760,000 acre-feet (Reclamation, 21 August 2026).
Fab demand is large and visible. AZFamily reported in August 2026 that TSMC uses 4.75 million gallons of water a day in Phoenix (AZFamily, 24 August 2026).
The wastewater side is already permitted. Phoenix issues Class A industrial wastewater discharge permits to companies discharging an average of 25,000 gallons a day of process wastewater, or whose effluent could affect the city’s treatment plants; City Code Chapter 28 sets out the requirements (City of Phoenix). Fab process wastewater also falls under the federal semiconductor effluent guideline, 40 CFR Part 469 Subpart A. That 1983 rule was not revised after EPA’s November 2022 study, but EPA says its PFAS data are limited and it will keep monitoring PFAS from the category (EPA, page updated 17 March 2026).
Where fab concentrate can go in central Arizona
Central Arizona has no ocean outfall, so fab concentrate has four possible destinations, and each one is limited by something different: the city plant’s salt budget, land, geology, or the energy and landfill that full ZLD needs.
| Outlet | Regulator or decision-maker | What limits it in central Arizona |
|---|---|---|
| City sewer | The city’s industrial pretreatment program, applying 40 CFR Part 469 and local limits | The salt passes through the city’s treatment plant and into its effluent. The Bureau of Reclamation’s Central Arizona Salinity Study (Phase II, 2006) found that sewer discharge raises effluent salinity and does not work well for large concentrate volumes (CASS Phase II) |
| Evaporation ponds | ADEQ. Ponds, lagoons and other surface impoundments are discharging facilities that need an Aquifer Protection Permit unless exempt (A.R.S. 49-241) | Land. The same study called ponds land-intensive and workable mainly for small rural facilities |
| Class I deep well | ADEQ, which took over Underground Injection Control for well Classes I to VI on 15 October 2025, except on Indian lands (EPA final rule) | Geology. In 2006 the study said deep well injection “has just about been ruled out” in central Arizona because of unsuitable geology |
| Thermal ZLD to dry solids | Off site, the receiving landfill. Disposing of the solids on site would itself be a discharging facility under A.R.S. 49-241 | Energy, capital, and a salt cake that still needs a landfill |
The salinity study numbers explain why salt is a regional issue and not only a plant issue. In its 2003 Phase I report, Reclamation estimated that about 1.5 million tons of salt a year enter the Phoenix metropolitan area and about 400,000 tons leave, so almost three-quarters stays (CASS Phase I). The study is old, and its land and well findings should be re-checked for any specific site. The basic geography has not changed: there is no ocean outfall, and the one regional answer studied, a pipeline to carry concentrate from Tucson and Phoenix to Yuma, was rejected in the late 1990s on cost and water loss.
What the Chandler brine reduction plant shows
The closest local precedent is in Chandler. Intel funded an advanced reverse osmosis facility, built in partnership with the City of Chandler, that treats process wastewater from its Ocotillo campus fabs to drinking water standards and returns it to the aquifer. By the time of an EPA Region 9 case study, more than 3.5 billion gallons had been recharged (EPA Region 9, archived).
That RO plant produces its own brine. When the campus expanded, the brine side was upgraded as the Ocotillo Brine Reduction Facility. Its contractor, Sundt, describes a 1.5 million gallon a day plant completed in 2014 that uses lime softening, high-pH reverse osmosis and a thermal brine concentrator, with 90% recovery on the RO and 97% for the whole system, to reduce the flow going to existing evaporation ponds (Sundt, contractor’s project page). The Associated General Contractors of America reported that brine leaving the facility fell to less than 60 gallons a minute (AGC, 2015). That is consistent with 3% of 1.5 million gallons a day, about 31 gallons a minute. The City of Chandler lists the facility among the plants it operates (City of Chandler job posting, closing January 2025).
Three lessons carry over to any Arizona fab:
Even 97% recovery leaves a liquid stream. The plant reduced the brine sent to ponds. It did not eliminate the ponds.
The outlet came first. The ponds already existed, so the design stopped at a brine concentrator and did not need a crystallizer.
Ownership is a design choice. Here a city owns and runs the brine plant. A fab that builds its own takes on the permits, the operators and the residual.
AWP rules: when the city’s reuse plans reach back into your discharge
Arizona’s Advanced Water Purification rules, 18 A.A.C. 9, Article 8, took effect on 4 March 2025. They let cities and water providers apply for permits to purify treated wastewater for potable use (ADEQ). For a fab that sends concentrate to the sewer, the important part is source control. The rule defines an Enhanced Source Control program to prevent constituents of concern at their source, and it separates “potentially impactful” from “impactful” non-domestic dischargers (R18-9-A801). Its Tier 2 chemical control starts from a chemical inventory (ADEQ Tier 2 flowchart).
In June 2026 ADEQ circulated an informal draft of a limited update, not yet a formal proposed rule, that would narrow Tier 2 to potentially impactful non-domestic dischargers and allow class-representative as well as individual discharger analysis. It would also define what counts as a chemical used, stored or discharged. The informal comment period closed on 20 July 2026 (ADEQ, June 2026); check ADEQ for the final text.
The practical point: if your sewershed feeds a future AWP plant, the sewer outlet is no longer only a question of salt. The city’s source control program can look at your chemical inventory and at trace contaminants such as PFAS, not only at your salt. Our page on direct potable reuse versus industrial reuse compares the two uses of the same effluent.
A worked example: what higher recovery changes, and what it does not
This is a hypothetical case with stated assumptions. A fab sends 3.0 million gallons a day (MGD) of combined process wastewater to its reclaim plant at 1,200 mg/L total dissolved solids, and the membranes reject essentially all of the dissolved salt. The salt load is 3.0 × 1,200 × 8.34 = about 30,000 lb a day, or 15 short tons.
| Recovery | Concentrate flow | Concentrate TDS | Salt in the concentrate |
|---|---|---|---|
| 65% | 1.05 MGD | about 3,400 mg/L | 15 tons a day |
| 85% | 0.45 MGD | 8,000 mg/L | 15 tons a day |
| 90% | 0.30 MGD | 12,000 mg/L | 15 tons a day |
| 97% | 0.09 MGD (about 63 gpm) | 40,000 mg/L | 15 tons a day |
Moving from 85% to 90% saves 0.15 MGD, about 168 acre-feet a year. The salt column never changes. That matters because the four outlets are limited by different things:
Ponds are limited by volume. Every step up in recovery cuts the pond area you need.
The sewer is limited by salt mass. The city plant receives the same 15 tons a day whether it arrives in 1.05 MGD or 0.09 MGD.
A landfill is limited by mass. Full ZLD in this example leaves roughly 15 tons a day of dry salt, about 5,500 tons a year, before any water bound in the solids.
Scaling limits how far membranes can go. At 97% the concentrate is close to seawater salinity, and Phoenix supply water carries more calcium and magnesium than Taiwan’s; TSMC says its Arizona design takes that into account. Silica, calcium sulfate and fluoride decide where membranes stop and evaporators start.
So the right recovery target depends on the outlet you can secure. Test your own figures with our brine concentration factor calculator and MLD vs ZLD screener.
Near-zero or full ZLD?
TSMC describes its plant as “near zero liquid discharge”, and in our view that is often the sensible target. Full ZLD adds a crystallizer, a solids handling line and a landfill contract to remove the last few percent of liquid. Settle these points before a supplier sizes anything:
which outlet is available for the residual, in writing, and for how long
which streams stay out of the salt train altogether (fluoride, ammonia, CMP slurry and solvent-bearing wastes are usually treated separately)
what PFAS and other trace data exist for each stream, given the AWP source-control rules
who owns the residual and its permits: the fab, the city or a contractor
Our guides on minimal liquid discharge versus ZLD, ZLD evaporators and crystallizers and zero liquid discharge cost cover each step in more depth. For the wider choice of endpoints, see brine disposal options compared.
Where independent review fits in a fab water project
You will need Arizona engineers for permits and a supplier to design, build and guarantee the plant. An independent adviser works upstream of both:
checking per-stream analyses (silica, calcium, sulfate, fluoride and any PFAS data) before they become the design basis
setting the recovery target from the outlet you can secure, not from a supplier’s best case
costing near-ZLD against full ZLD with the salt cake and its landfill contract included
reading reclaim and brine plant offers for exclusions, recovery guarantees and who carries the residual
independent technical due diligence when a lender, buyer or board asks for it
If a supplier has already proposed a ZLD plant, start with our ZLD vendor proposal checklist. Our services: water reuse and ZLD systems, water treatment and ZLD consulting and brine resource recovery. For other US states, see our USA overview and the Texas page.
Does Arizona require zero liquid discharge for semiconductor plants?
Not as a blanket rule. The limit comes from whichever outlet you use: the city pretreatment permit for the sewer, an Aquifer Protection Permit for ponds, a UIC permit for a well. Full ZLD is how a fab meets them when none of those is open, and it still leaves a salt cake that needs a landfill.
What should a fab have ready before asking suppliers to quote?
Per-stream flows and analyses that include silica, calcium, sulfate and fluoride; any PFAS data you hold; the outlet you can secure for the residual, and for how long; and whether your sewershed is planned to feed an AWP plant. In our view, without these a supplier has to choose the design assumptions, and its quote reflects that choice rather than your site.
Send us the concentrate analysis
If a reclaim expansion, a new fab or a brine plant is on your schedule, send us the concentrate analysis and the outlet you are counting on. A scoping call is usually enough to see whether the recovery target and the ZLD scope hold up. Book a scoping call, send us your project details or write to christos@brineconsulting.com.
Written by the Brine Consulting advisory team. The Brine Consulting advisory practice includes Paolo Mattiuzza, Executive Advisor for Brine Management and ZLD Systems, with more than 20 years of international experience in environmental technologies, industrial water treatment and resource recovery, and Wendy Garcia, PhD, Part Time Consultant, with more than 15 years of experience in electrodialysis, bipolar membrane systems, brine treatment and resource recovery. Meet our experts.