Quick answer: Under Australia's R&D Tax Incentive, reusing plant water is usually ordinary engineering: where an applicable, validated design basis already predicts how the train will perform on the actual feed and operating conditions, the outcome is determinable in advance. A core R&D activity may exist where no such basis predicts what a specific loop accumulates at steady state, whether the reclaimed stream still meets the process requirement at the point of use, and whether the concentrated purge stays inside a trade waste limit — questions resolvable only by a systematic progression of work conducted to generate new knowledge. Eligibility is self-assessed activity by activity against the statutory tests.
24 August 2026 — this article describes the current rules. The 2026-27 Federal Budget announced R&DTI changes; the ATO states the measure is not yet law, and industry.gov.au states the changes will apply to income years starting on or after 1 July 2028.
A factory water loop is not a small catchment. A municipal treatment plant receives an influent shaped by weather, population and a diurnal curve; a plant that recycles its own process water receives an influent shaped by the production schedule, and then feeds part of its own output back to the head of the process. That second property is what makes the engineering interesting and the eligibility question non-trivial. Every stream a loop cannot remove concentrates until the purge carries it away, and the concentrated stream is simultaneously the water the process has to drink and the effluent the sewerage authority has to accept.
The Statutory Test, Stated Once
Under s 355-25(1) of the Income Tax Assessment Act 1997, core R&D activities are experimental activities whose outcome cannot be known or determined in advance on the basis of current knowledge, information or experience, but can only be determined by applying a systematic progression of work that is based on principles of established science and that proceeds from hypothesis to experiment, observation and evaluation, and leads to logical conclusions; and that are conducted for the purpose of generating new knowledge, including new knowledge in the form of new or improved materials, products, devices, processes or services. A supporting R&D activity must be directly related to core R&D activities (s 355-30(1)), and s 355-30(2) adds a dominant purpose bar where the activity is one referred to in s 355-25(2), produces goods or services, or is directly related to producing goods or services (business.gov.au; ITAA 1997).
Water reuse work rarely struggles with the established-science limb — mass conservation, membrane transport, solubility products and disinfection chemistry supply a settled body of principle. The more difficult questions are usually whether the outcome was determinable in advance, and whether particular activities connected with implementing or operating the reuse system require consideration under the supporting-activity provisions, including the dominant-purpose test where s 355-30(2) applies. Novelty to the company, capital cost and difficulty are not the test. What decides the question in practice is the record — the activity schedule, the hypothesis and the unresolved uncertainty as they stood before the work, the controls held constant, the pre-recorded stopping rules, and the evidence tying each alleged supporting activity to the trial rather than to production.
What Accumulates, and Where the Arithmetic Runs Out
Conservative species are governed by the purge, not by the treatment: For a species the train does not remove — sodium, chloride, sulfate, nitrate, silica, much of the hardness — the steady-state concentration in the ring is the mass entering per unit time divided by the volume leaving per unit time. Halving the purge fraction doubles the steady-state concentration of everything with a per-pass removal near zero. Adding a finer filter changes nothing. Teams spend capital on separation and are defeated by a conservative mass balance.
Steady state arrives slowly, so short trials read as passes: The loop approaches its asymptote with a time constant of roughly the ring inventory divided by the flow that removes the species. A loop holding a day and a half of inventory and purging a tenth of throughput has a time constant near a fortnight, and needs three of them to sit within about five per cent of its final value. A two-week trial on such a loop measures a transient, reports compliance, and is contradicted in month two. Any credible programme either runs long enough or deliberately accelerates the loop, and both choices have to be justified rather than assumed.
The loop changes the chemistry it feeds back into: As soluble organics and organic nitrogen accumulate, the chlorine demand of the ring rises and the breakpoint shifts, so a dose that once produced a free chlorine residual increasingly produces combined chlorine — a total-chlorine reading stops indicating disinfectant capacity at precisely the moment it matters — and disinfection by-product yield tends to rise with the accumulated organic precursor load. Recycled heat matters too: a loop returning blanching or cooling water runs warm, raising regrowth rates, shortening residual half-life and, with accumulating chloride, moving stainless steel towards conditions where pitting becomes a live risk. Where the reclaim train includes biological treatment, rising conductivity feeds back onto the biology itself — elevated and variable salinity is a well-recognised stressor on activated sludge settleability and on nitrification — so the treatment step can degrade as the loop it serves tightens.
Membrane rejection is a variable, not a specification: Rejection of monovalent ions generally falls as feed concentration and recovery rise, and scaling is set by the concentration at the membrane wall rather than in the bulk, so calcium phosphate, calcium sulfate and silica saturation limits are reached earlier than a bulk calculation suggests. A vendor's rejection figure is quoted on a standard test solution, not on a food factory's reclaim water.
And the source term moves with the production plan: A clean-in-place sequence delivers a caustic pulse, then an acid pulse, then rinse, within an hour; a product changeover changes the soil, the organic load and sometimes the allergen status of the whole loop. A mass balance can take a schedule-derived source term as an input — that is ordinary practice. The open question is whether the actual loads, the per-pass removals and the loop's own feedback effects stay predictable across that schedule once the water starts coming back around.
Two Limits Pulled in Opposite Directions
Reclaimed water inside a food factory has to satisfy the process, and the concentrated remainder has to satisfy the sewerage authority. Closing the loop moves both in the wrong direction at once.
Process side — Food Standards Code
Clause 4(2) of Standard 3.2.3 of the Australia New Zealand Food Standards Code requires a food business to use potable water for all activities that use water conducted on the food premises, and clause 4(3) permits non-potable water for a purpose where the business demonstrates that its use for that purpose will not adversely affect the safety of the food handled (legislation.gov.au; FSANZ). Fitness for a specific purpose is therefore something the business has to establish by evidence, purpose by purpose, and the national water recycling guidelines set out the risk-based method for doing it (waterquality.gov.au).
Discharge side — Trade Waste Limits & Dilution Prohibition
The national trade waste acceptance guidelines express most criteria as concentrations, with mass-load criteria available where an authority elects to apply them, and they prohibit diluting wastes with potable, bore, ground, stormwater or any non-process water (waterquality.gov.au). A site that recovers most of its water has removed the carrier that used to keep a concentration-based limit satisfied; the daily mass of salt leaving the site is unchanged; and the compliant remedy of adding water is unavailable. Reuse converts a volume problem into a concentration problem.
Where an outcome could have been determined in advance from current knowledge, the activity is generally unlikely to be a core R&D activity on those facts, subject to the activity's own facts and the statutory tests. See what does not qualify.
A Worked Example (Hypothetical and Illustrative Only)
Invented to show where the technical boundary falls. It is not a real project and says nothing about whether any actual claim would be accepted.
A vegetable processing factory in Adelaide's northern suburbs washes, lye-peels, dices and blanches potato and carrot for foodservice. Mains intake averages 1,060 kL/d: 420 raw fluming and destoning, 260 post-peel flume, 110 peeler wash, 90 blanching and hydro-cooling, 180 clean-in-place and floor wash. Mains water runs about 300 mg/L total dissolved solids, 40 mg/L sodium and 60 mg/L chloride. Discharge to sewer is about 900 kL/d — the remaining 160 kL/d leaves as evaporation, product moisture and solids — at about 1,150 mg/L TDS, 290 mg/L sodium and 240 mg/L chloride, so the site sends roughly 1,035 kg/d of dissolved solids and 260 kg/d of sodium to the sewer. The trade waste authorisation carries illustrative limits of BOD 600 mg/L, suspended solids 600 mg/L, TDS 4,000 mg/L and pH 6–10, measured at the connection point. Baseline reclaim-ring water quality does not exist, because there is no ring.
The target, recorded 3 March 2026: Meet ≥ 55% of process water demand from reclaimed water at steady state; at all product-contact points of use hold E. coli not detected/100 mL, turbidity ≤ 1 NTU, free chlorine ≥ 0.5 mg/L at last flume, total THMs ≤ 0.10 mg/L, Na ≤ 180 mg/L, Cl ≤ 250 mg/L; keep discharge within trade waste limits; and maintain peeled-carrot surface Na/dice firmness in control limits, with no difference in a 30-panellist triangle test (≥ 15 correct needed for a 5% difference).
Held and varied: Held: production schedule, peeler setpoints, blanch time/temp, sanitiser, sampling protocol (1-min online data + daily composites). Ring volume ~2,600 kL. Each configuration ran ≥ 3 time constants before a 10-day evaluation window. Stopping rules: flux decline > 20% or any trade waste parameter exceeded.
Configuration A (Days 1–41, UF with 30% purge — Failed): Suspended solids fell to 6 mg/L, BOD to 95 mg/L. But conductivity climbed to 4,800 µS/cm, ring Na reached 870 mg/L (535 mg/L at points of use vs 180 mg/L limit), Cl reached 415 mg/L, and THMs hit 0.19 mg/L. Ruled out: particulate/organic separation alone without managing conservative dissolved ions.
Configuration B (Days 48–66, RO at 85% recovery, concentrate as purge — Stopped): Point-of-use Na fell to 15 mg/L and Cl to 20 mg/L. But connection-point TDS crossed the 4,000 mg/L limit on day 17 and reached 4,170 mg/L on day 19 (headed to 6,000 mg/L asymptote). Flux fell 22% due to calcium phosphate/silica scaling at membrane walls. Stopping rule triggered on both grounds; eliminated high-recovery RO on combined stream.
Configuration C (Days 73–128, met predefined criteria): Identified that lye peeler rinse (44 kL/d, 4.2% of water) carried 71% of Na and 60% of Cl. Segregated that stream directly to sewer. Loop ran drum screen, DAF, MBR, UF with a 150 kL/d RO polishing slipstream at 70% recovery (below scaling limit) and automated free-chlorine trim. Ring Na dropped to 230 mg/L (155 mg/L at points of use), Cl to 172 mg/L, turbidity 0.4 NTU, THMs 0.06 mg/L, E. coli 0/100 mL, and mains intake fell to 445 kL/d (58% reuse). Discharge ran 285 kL/d at 2,980 mg/L TDS (within 4,000 limit). Triangle test returned 13/30 (no difference detected).
Result & candidate boundary: Established that segregating 4% of volume controlled the steady state of the remaining 96%. Candidate core activity spans recorded hypothesis through to evaluation in the day 118–128 window. Skid procurement, trade waste variations, operator training, routine compliance sampling, and production operation sit outside, needing supporting-activity analysis under s 355-30.
What Is Generally Unlikely to Be Core on Those Facts
Activity
Why
Installing a packaged reclamation plant to a vendor's guaranteed permeate specification
A validated design basis already predicts the result
Sizing an equalisation tank from the production schedule and a mass balance
Established method; outcome determinable in advance
Routine trade waste compliance sampling and the annual return
Activity associated with complying with statutory requirements — s 355-25(2)(f)
Daily food-safety verification sampling of product-contact water
Routine testing and analysis in connection with complying with statutory requirements — s 355-25(2)(f)(iii)
Adjusting membrane clean-in-place frequency inside the supplier's operating envelope
Applying a documented product as designed
Substituting an equivalent sanitiser to reduce chemical cost
No unresolved question requiring experiment
Where an RSP Fits
An RSP is a scientific or technical service provider registered in specific research fields that a company can engage to conduct R&D activities on its behalf (business.gov.au). Ignition Research is registered in manufacturing engineering (ANZSRC 4014) and environmental engineering (ANZSRC 4011). On a reuse loop the work sits upstream of the trial: writing the fit-for-purpose specification against the actual points of use rather than the treatment outlet, sizing the trial to the loop's time constant instead of the shutdown calendar, recording the stopping rules before the first run, choosing instrumentation that resolves the species the mass balance says will govern, and building the observation and evaluation record while the line is running (record keeping).
There is also a point worth knowing for smaller manufacturers: Eligible R&D expenditure incurred to a registered research service provider may be treated differently where the entity's notional R&D deductions fall below the usual $20,000 threshold, subject to the requirements in s 355-100(2) and the underlying activities satisfying the R&D eligibility tests. And using an RSP does not guarantee eligibility — you still self-assess. In mechanism, where an R&D entity's notional deductions for the year fall below $20,000, s 355-100(2) substitutes a different base for working out the offset — the first kind of expenditure in it being expenditure incurred to a registered research service provider that is not an associate of the R&D entity, for services within a research field for which the provider is registered — so RSP-conducted eligible R&D activities can still support an offset. Offset tiers are covered in our article on the refundable and non-refundable offset; see also claiming R&D under $20,000 and R&D Tax Incentive in Adelaide.
Talk to Ignition Research before the skids are ordered, while the loop can still be designed as an experiment rather than reconstructed as one. As a Registered Research Service Provider at Lot Fourteen in Adelaide, we design and conduct the experimental programme and produce the technical record while the plant runs. We are not a registered tax agent: your business self-assesses and remains responsible for its own claim. Get in touch.
Frequently Asked Questions
Q: Is installing a water recycling system in a factory an eligible R&D activity?
A: Generally not, as a core R&D activity, where the system is specified from a mass balance and purchased against a guaranteed permeate quality that has been validated on a comparable feed — the outcome is determinable in advance on the basis of current knowledge. A core activity may exist in the narrower case where no such basis predicts what a particular loop accumulates at steady state on a particular production schedule, and whether the reclaimed stream will meet the process requirement while the purge stays inside its discharge limits, and that can only be determined by the systematic progression of work described in s 355-25(1). Eligibility is self-assessed activity by activity.
Q: Why can't the steady-state composition of a closed loop be calculated in advance?
A: The conservative part often can be — a salt balance will tell you roughly where sodium or chloride ends up for a given purge fraction, and it can take the source term straight from the production schedule. What the balance does not supply is the per-pass removal of the actual mixture at the concentration it will actually reach, since membrane rejection generally falls with feed concentration and recovery, scaling is governed by wall concentration rather than bulk, and biological treatment inside the loop can degrade as the loop's own salinity rises. Nor does it tell you whether the schedule-derived loads and those feedback effects stay predictable once water is returning around the ring. The balance frames the hypothesis; it does not answer it.
Q: Does closing a water loop make trade waste compliance harder rather than easier?
A: It can. The national trade waste acceptance guidelines express most criteria as concentrations, and they prohibit diluting wastes with potable, bore, ground, stormwater or other non-process water. Recovering water removes the carrier that previously kept a concentration-based limit satisfied even though the daily mass discharged is unchanged or lower, so a site can move from compliance to exceedance without changing what it puts into the sewer. Local authorities set their own acceptance criteria, which may be stricter than the national guideline values.
Q: Can recycled water be used for product-contact purposes in a food factory?
A: Clause 4(2) of Standard 3.2.3 of the Food Standards Code requires potable water for all activities that use water conducted on food premises, and clause 4(3) permits non-potable water for a purpose where the business demonstrates that its use for that purpose will not adversely affect the safety of the food handled. The demonstration is the business's to make, on evidence, for each purpose; the national water recycling guidelines set out the risk-based method, and state and territory food regulators administer the Code. Whether generating that evidence involves any core R&D activity is a separate question turning on whether the outcome was determinable in advance.
Sources & Further Reading
legislation.gov.au — Income Tax Assessment Act 1997 — Div 355, incl. ss 355-25, 355-30 and 355-100
legislation.gov.au — Food Standards Code Standard 3.2.3, Food Premises and Equipment — cl 4, water supply
waterquality.gov.au — Guidelines for Sewerage Systems: Acceptance of Trade Waste (Industrial Waste)
waterquality.gov.au — Australian guidelines for water recycling
waterquality.gov.au — Default guideline values for water quality
Related: manufacturing engineering research field · environmental engineering research field · what does not qualify · what an RSP is · refundable vs non-refundable offset · more Insights
This article is general information from a Registered Research Service Provider about the R&D Tax Incentive. It is not tax, legal or financial advice; eligibility depends on your circumstances and you should self-assess and seek your own advice.
Thinking about a project like this?
If you're weighing up an AI, software or technical improvement project and can't tell yet whether it's implementation or research, start with a quick read on where it sits.