Beverage plants are tapping their RO “waste” — and the payback is under a year

In beverage manufacturing, water is both ingredient and lifeblood — and the numbers are finally moving. Global bottlers track a water use ratio (WUR: liters of water per liter of product). Coca‑Cola’s WUR dropped from ~2.1 in 2013 to ~1.85 by 2020 (www.yumpu.com), yet millions of liters still vanish as wastewater, clean‑in‑place rinses (CIP: cleaning of process equipment without disassembly), and cooling blowdown.

New playbooks focus on squeezing more from reverse osmosis (RO: membrane process that separates water from dissolved salts), then reusing the RO concentrate (reject) in non‑product roles like cooling towers or initial cleaning. One Indonesian dairy/beverage facility recycled RO effluent via ultrafiltration (UF: membrane that removes suspended solids) and saved ~5.5 m³/day of freshwater — roughly 30% of its cooling tower demand (jurnal.polinema.ac.id) (jurnal.polinema.ac.id).

A larger beverage plant in Tanzania pushed RO recovery to 91% and saved 606 m³/day — trimming concentrate waste by 26% and translating to USD $635,000 per year in reduced water purchases (greenminds.world). The same audit‑to‑action approach is increasingly standard in ingredient water preparation, where a single RO train can dictate the plant’s overall WUR.

Technology providers are leaning into this shift with integrated membrane suites; facilities commonly mix and match RO, NF, and UF systems to meet fluctuating source water and reclaim targets.

Water use ratio and audit baselines

WUR provides a simple scorecard, but the savings come from mapping losses. In the Tanzania case, a comprehensive audit uncovered excessive RO reject and intermittent supply issues; by optimizing borewell conditioning and RO staging, the plant cut reject by 26% (greenminds.world). Tracking RO recovery percentage (share of feed converted to permeate) alongside WUR sets a real‑time target — >90% recovery in that case (greenminds.world).

High‑recovery RO engineering

Plants are re‑plumbing for higher RO recovery, well above the typical 60–75% range. One beverage facility installed three high‑recovery RO units totaling 110 m³/h and lifted recovery from <65% to 91%, saving 606 m³/day of fresh water — nearly 70% of the initial 932 m³/day waste — while cutting brine volume (greenminds.world) (greenminds.world) (greenminds.world).

Where salinity fits the envelope, brackish‑water designs help dial in recovery; many plants standardize on brackish water RO to handle maximum total dissolved solids (TDS: sum of salts) in the 10,000 mg/L class while holding energy and fouling in check. Membrane scaling control upstream of the high‑recovery stage often hinges on antiscalant dosing; process teams formalize this with membrane antiscalants tuned to their carbonate/sulfate risk profile.

Pretreatment barriers and fouling control

Fouling caps recovery and drives cleanings. Robust pretreatment — multimedia filtration, activated carbon, softening, or low‑pressure membranes — keeps RO flux stable and extends interval between cleans. An Indonesian study confirmed that a pretreatment train (multimedia + activated carbon + softener) ahead of RO yielded higher‑quality recycle water and better throughput (jurnalkesehatan.unisla.ac.id).

The Nutricia case used hollow‑fiber UF ahead of RO to keep a 1,225 mg/L TDS feed from clogging membranes and enabled recycling to save ~5.46 m³/day (jurnal.polinema.ac.id) (jurnal.polinema.ac.id). Plants commonly specify ultrafiltration for this barrier step. To reduce organics and chlorine that can damage membranes, teams also standardize on upstream activated carbon. For grit and fines ahead of UF, dual‑media filters such as sand/silica remain a workhorse.

When hardness is the limiter, ion exchange softening in pretreatment is straightforward. Beverage facilities implement this with a dedicated softener to pull calcium and magnesium, curbing calcium sulfate or carbonate scale at higher RO recovery.

Cooling tower makeup from RO concentrate

RO still leaves a concentrated brine (often 25–35% of feed). Instead of sending it to sewer, plants are routing it to non‑potable loops, with cooling towers as the prime target. Cooling towers can consume >30% of a facility’s total water (www.prochemtech.com), and RO brine is typically soft but salty — manageable with the right inhibitors and controls.

In one ProChemTech case, a chemical plant used RO concentrate (23,760 gpd) as 5‑COC cooling makeup and saved ~$91,800/yr in water/sewer costs (www.prochemtech.com). Capital was ~US$64,000 for a 671‑ton capacity system, for a simple ROI of <1 year (www.prochemtech.com), with dosing of a corrosion/deposition inhibitor and biocide enabling stable 5–6× cycles of concentration (COC: ratio of dissolved solids in recirculating water to makeup water) (www.prochemtech.com). Many beverage plants also reuse RO filtrate in towers after pH adjustment (www.chemtreat.com).

For these high‑salinity loops, treatment programs lean on tower chemistry. Plants typically budget for scale inhibitors to manage deposition, corrosion inhibitors to counter chloride stress, and biocides to control biofilm at elevated cycles. ProChemTech also notes the need for flow equalization (a brine storage tank with level controls), automatic makeup blending, and dedicated dosing to maintain stability (www.prochemtech.com).

Process rinse and first‑stage cleaning reuse

Beyond towers, facilities are reclaiming RO rinse water for equipment rinsing and floor washdowns. One U.S. bottling line reused sterile RO post‑pasteurization rinse in cooling and recirculation loops. By stabilizing pH with small‑dose chemical addition, it reused ~20–30% of RO rinse water and saved up to 1.76 million gallons/year (www.chemtreat.com).

RO concentrate can also be routed to initial CIP or non‑critical flushing, provided the stream is conditioned to avoid residue, including partial blending or softening. Many sites address hardness and scale risk by incorporating a softening step before using the stream in cleaning. Studies also show membrane trains can polish high‑load streams: one pilot treated soft‑drink wastewater to 100% remove TSS and turbidity, with COD reduction of 65% via UF and 29% via nanofiltration (NF: membrane that rejects multivalent salts/organics) — though further polishing was needed for stringent COD limits (www.researchgate.net). Plants exploring NF for polishing often evaluate packaged nano‑filtration skids for this role.

Treatment steps and control systems

pH and corrosion control come first. RO reject often has lower pH (from dissolved CO₂) and higher chloride/sulfate; buffering up to neutral range protects metallic systems. In ChemTreat’s case, a proprietary buffering agent kept pH≈6.5–8.2 before routing water to cooling towers (www.chemtreat.com). Automated chemical feeds with a dosing pump help maintain setpoints under changing loads.

Additional filtration/softening may be added to lower fouling potential. If hardness is elevated, removing Ca/Mg prevents calcium sulfate scale at higher COC; in one referenced case, pretreatment softening of the RO feed left the brine largely free of calcium carbonate, easing reuse as coolant (www.prochemtech.com). Plants also add fine filtration at the point of use; skid builders specify replaceable elements in stainless cartridge housings for hygienic service.

Monitoring closes the loop. Typical reuse skids include level sensors on a brine storage tank, flow meters to control blend ratios, and conductivity or titration‑based controllers to maintain target COC. Controls like the MegaTron CDI system (as in [13]) automatically dose based on pH or conductivity, ensuring blended reuse water meets setpoints. Remote alarms and a soft‑water backup valve that opens if the brine tank runs low are common safeguards (www.prochemtech.com). For hardware and instrumentation, plants typically roll these into water‑treatment ancillaries on a single panel.

Cost–benefit outcomes and ROI

Utility savings add up quickly. In Tucson, AZ, water + sewer rates of ~$10.73 per 1000 gal allowed a 100 gpm RO (12 hr/d) to reclaim 23,760 gpd as cooling makeup, cutting ~$91,800/yr from water/sewer bills (www.prochemtech.com). Installed capital for a reuse skid was about $64,000 — ProChemTech cites both a 671‑ton capacity system and, elsewhere, a ~500‑ton cooling system reuse skid (handling ~23,760 gpd brine) — with a simple ROI under one year in each description (www.prochemtech.com) (www.prochemtech.com).

At larger scale, the Tanzania plant’s 606 m³/day savings equated to ~$635,000/yr (implied water cost ≈$2.9/m³) (greenminds.world). A cost–benefit analysis at a Kenyan brewery reported a Benefit‑Cost Ratio (BCR) of ~3.1 for recycling just 10% of annual water via membranes, rising to ~3.6–4.0 at 25–50% recycling (www.researchgate.net) (www.researchgate.net).

Operating costs do rise modestly for pumping and chemicals, but case reports note the increase is small relative to savings. In the ChemTreat example, an online controller that maintained pH and documented probe data reduced downtime and labor (www.chemtreat.com).

Measured impacts and regulatory context

Facilities report 20–30% reductions in fresh water demand by recycling RO brine — for instance, 5.46 m³/day saved versus 21 m³/day cooling water use in the Indonesian study (www.prochemtech.com) (www.researchgate.net). Typical ROI is <1–2 years in moderate‑to‑high water cost settings (www.prochemtech.com) (www.researchgate.net), and utility bill cuts can range from tens to hundreds of thousands USD per year for medium plants — from $91k/yr at 100 gpm to $635k/yr at 110 m³/h (www.prochemtech.com) (greenminds.world).

There are also resilience gains: the Tanzania project linked greater water security to a 27% increase in product throughput (greenminds.world). In Indonesia, while specific reuse regulations are not stringent for industrial internal use, aligning with global best practices (e.g., complying with Wastewater standards PP 22/2021) can ease permitting. Each percent point of recycled water (relative to baseline use) returns roughly $0.5–$1+ million avoided costs per year in large factories (per [19]).

Implementation toolkit

Ingredient water teams typically deploy a few standard building blocks: high‑recovery RO with upstream pretreatment, a reuse skid with brine storage and blending, and tower chemistry tuned for higher COC. UF pretreatment is often packaged on hygienic frames (see ultrafiltration), with point‑of‑use polish in cartridge filters. Cooling‑side chemistry is metered with dosing pumps, and instrumentation plus safeties are consolidated under water‑treatment ancillaries.

Summary of measurable impacts

• Reducing fresh water demand by 20–30% is realistic when recycling RO brine (e.g., 5.46 m³/day saved vs. 21 m³/day cooling use in the Indonesian example; or ~boys recoup rate in www.researchgate.net).
• Typical ROI is <1–2 years in moderate‑to‑high water cost settings (www.prochemtech.com) (www.researchgate.net).
• Utility bill cuts can be tens to hundreds of thousands USD per year for medium plants ($91k/yr at 100 gpm and $635k/yr at 110 m³/h) (www.prochemtech.com) (greenminds.world).
• Each percent point of recycled water returns roughly $0.5–$1+ million avoided costs per year in large factories (per [19]).

References

Taufik H. Ramadhon et al., “Daur ulang air limbah menggunakan RO untuk Make‑Up Water Cooling Tower…”, Distilat J. Teknol. Separasi 9(2) (2023) (jurnal.polinema.ac.id) (jurnal.polinema.ac.id).

Timothy Keister, “Reuse of Reverse Osmosis Concentrate as Cooling Tower Makeup”, ProChemTech (11 Sep 2011) (www.prochemtech.com) (www.prochemtech.com).

Greenminds (Case Study), “Transforming RO Reject into a Resource…”, (22 Jan 2025) (greenminds.world).

ChemTreat, Inc. (Case Study), “Water Reuse Program Saves Beverage Plant…” (2021) (www.chemtreat.com).

M. Wakhungu, “Cost Benefit Analysis of Water Recycling in Manufacturing (EABL, Kenya)” (2016, thesis) (www.researchgate.net).

M. Çağlar Gürbüz & Göksen Çapar, “Water Reuse via Membrane Technology: …Cooling Water…”, Akademik Gıda (Aug 2024) (www.researchgate.net).