Where does the rest of the water go? Pot ale, spent lees, and the zero-liquid-discharge rule behind your whisky

Key takeaways

• Almost none of a distillery’s water ends up in the bottle. Producing one litre of alcohol takes roughly 10–17 litres of water (Singh et al., Bioresources and Bioprocessing, 2026); the bottle receives a few hundred millilitres of it.
• Scotch malt distilling returns it as two named streams. Per litre of alcohol: about 8 L of pot ale and 10 L of spent lees and washings, plus 2.5 kg of draff (White et al., ACS Omega, 2020).
• Pot ale is a strong effluent, not dirty water. COD 47 g/L, BOD 25 g/L, pH 3.6–4.1, and copper from 1.7 to 15.3 mg/L (White et al., 2020) — one to two orders of magnitude stronger than municipal sewage.
• Molasses stillage is stronger still. 8–15 L per litre of alcohol at COD 100,000–150,000 mg/L, BOD 35,000–50,000 mg/L, pH 4.0–4.5 (Mikucka & Zielińska, Applied Biochemistry and Biotechnology, 2020).
• India’s answer is zero liquid discharge. CPCB’s Charter for molasses-based distilleries requires the shift from end-of-pipe treatment to integrated water management (CPCB, Charter for ZLD in Molasses Based Distilleries), and it has moved real numbers: water consumption down from 15 to 5.59 KL per KL of alcohol between 2016–17 and 2021–22 (Satavan et al., Indian Forester, 2023).
• Blending water is the exception that proves the rule. It is the only significant water stream in a distillery that generates no effluent — because it is not used and discarded, it is consumed into the product.

A bottle of whisky contains a few hundred millilitres of water, and producing the alcohol in it consumed somewhere between ten and seventeen litres (Singh et al., 2026). The difference does not evaporate and it does not disappear: it leaves the distillery as effluent — hot, acidic, organically loaded, and in the Scotch case carrying dissolved copper from the stills. Understanding where that water goes is the part of whisky’s water story that marketing never tells, and it is also the clearest way to see why the water added at the end of the process is categorically different from every other water in the plant.

How much water actually leaves a distillery as waste?

Nearly all of it. The Scotch Whisky Association’s industry target is 12.5 to 25 litres of water per litre of pure alcohol by 2025, a range wide enough to reflect how differently distilleries are built and sited (Scotch Whisky Association, Responsible Water Use). Indian and international figures for fermentation-based alcohol land in a similar band at 10–17 l/lpa (Singh et al., 2026).

Set that against what reaches the consumer. A 700 ml bottle at 40% ABV contains roughly 280 ml of ethanol and 420 ml of water, and only part of that water was ever added deliberately as blending water — the rest came through with the distillate. Against 12.5 litres of process water per litre of pure alcohol, the water that survives into the bottle is a rounding error.

Everything else becomes an effluent stream with a name, a permit, and a cost.

What is pot ale, and why can’t it go down the drain?

Pot ale is what remains in the wash still after the first distillation: the spent yeast, unfermentable sugars and protein left behind once the alcohol has been driven off. A detailed characterisation of a Scottish malt distillery’s output found that “for every 1 L of alcohol produced, approximately 2.5 kg of draff, 8 L of pot ale, and 10 L of spent lees (and washings) were the resulting yield” (White et al., ACS Omega, 2020).

The composition explains the permit. That study reports pot ale at pH 3.6–4.1, 5.1% total dry matter, COD 47 g/L and BOD 25 g/L, with copper between 1.7 and 15.3 mg/L and crude protein at 33% of dry matter (White et al., 2020). A COD of 47,000 mg/L is roughly a hundred times that of typical domestic sewage. Discharged untreated into a burn, it would strip the oxygen out of the water before it stripped anything else.

At national scale the volumes are industrial: the same work puts Scottish annual output at roughly 850,000 tonnes of draff and 2,720,000 tonnes of pot ale. The routes it takes are, in practice, four — concentration into pot ale syrup sold as a ruminant feed, land spreading, consented discharge to sea, and anaerobic digestion (White et al., 2020).

That last route is the one the industry has been moving toward, because the same organic load that makes pot ale a pollutant makes it a fuel. A design study for an Irish whiskey distillery producing 2 million litres a year modelled a pre-treatment-integrated anaerobic digester yielding about 900,000 m³ of methane annually — enough to cover 24% of the plant’s thermal energy requirement and 42.5% of its electricity, at €3.6 million capital cost and a 9.6-year payback (Dalton et al., Heliyon, 2022).

What are spent lees, and why are they the copper problem?

Spent lees are the residue left in the spirit still after the second distillation — a dilute, acidic mixture of organic acids and water, weak in organic load compared with pot ale but awkward for two other reasons. It is the larger volume of the two, at around 10 litres per litre of alcohol including washings (White et al., 2020), and it carries dissolved copper leached from the still.

Copper is not an accident of the process; it is the process. Copper surfaces in the still strip sulphur compounds out of the vapour and are a large part of why a pot-still spirit tastes the way it does. But copper that leaves the still in solution is an aquatic toxicant, and the acidity of spent lees — generally below pH 4.2 — keeps it dissolved and mobile. Distilleries manage this with pH correction, constructed wetlands ahead of surface-water discharge, or recovery; recent work at the University of Manchester has demonstrated stripping copper ions from spent lees with magnetite nanoparticles that are then pulled out magnetically (Chemistry World).

Note what this means for the water balance. Spent lees are mostly water, and they are the closest thing in the plant to a stream that could plausibly be recycled — but they are the wrong water in the wrong place. Nobody proofs a whisky with them.

Why is Indian distillery effluent a different scale of problem?

Because the feedstock is different. India’s alcohol industry is built largely on cane molasses, and molasses stillage — known in India as spent wash — is one of the most concentrated industrial effluents produced anywhere.

A peer-reviewed review puts the numbers plainly: “for every liter of alcohol produced, molasses-based distilleries usually generate about 8–15 L of stillage,” with COD of 100,000–150,000 mg/L, BOD of 35,000–50,000 mg/L, pH 4.0–4.5, and a discharge temperature of 71–81 °C (Mikucka & Zielińska, Applied Biochemistry and Biotechnology, 2020). That is two to three times the COD of Scotch pot ale, at comparable volume.

It is also dark brown, and it stays dark brown. The colour comes from melanoidins — high-molecular-weight polymers formed by Maillard reactions during molasses processing, present at around 2% and resistant to microbial breakdown (Singh et al., 2026). Biomethanation removes much of the oxygen demand — that review records BOD falling from 68,000–80,000 mg/L to about 34,000 mg/L — but the effluent remains strongly coloured and still unfit for discharge. Treating the organic load and treating the colour are two separate problems.

What does “zero liquid discharge” actually require?

It requires that no treated or untreated effluent leaves the site as liquid. Not “cleaned to a standard and discharged” — not discharged.

India’s Central Pollution Control Board pursued this for molasses-based distilleries through a Charter whose stated aim was to move distilleries “from an end-of-pipe treatment approach to an integrated water and waste management system” (CPCB, Charter for Zero Liquid Discharge (ZLD) in Molasses Based Distilleries; Satavan et al., Indian Forester, 2023). In practice, compliance means concentrating spent wash — typically by biomethanation followed by multiple-effect evaporation — and then either incinerating the concentrate in a boiler or composting it with press mud, while condensate from the evaporators is polished and recycled back into the process.

CPCB has continued to formalise the regime: a standard operating procedure for ZLD verification of distilleries was issued on 3 July 2024, tied to certification for units seeking financial assistance to expand ethanol capacity (TeamLease RegTech).

Has it actually worked?

On the evidence, yes — and the mechanism is instructive. Reviewing the Charter’s effect on distilleries along the main stem of the Ganga, Satavan and colleagues report that 63 molasses-based distilleries achieved ZLD, that water consumption fell from 15 KL per KL of alcohol in 2016–17 to 5.59 KL/KL in 2021–22 — a 62.7% reduction — and that spent wash generation fell from 11.1 to 6.48 KL/KL, down 41.6% (Indian Forester, 2023).

The important detail is which number moved first. Faced with a rule that made discharge impossible, distilleries did not primarily build bigger treatment plants — they stopped generating the effluent. When every litre of spent wash must be evaporated at your own cost, the cheapest litre is the one you never make. A discharge standard regulates the pipe; a zero-discharge standard regulates the process.

Where does blending water sit in all of this?

Outside it, and that is the point worth understanding.

Every other water in a distillery is used and released: mashing water becomes wort and then pot ale; cooling water takes on heat and goes back to the river or the tower; boiler water becomes steam and then blowdown; cleaning water becomes CIP effluent. Each one has a permit, a treatment cost and a disposal route.

Blending water has none, because it is not a utility. It is not used to do a job and then discarded — it is consumed into the product and shipped out in glass. There is no effluent stream associated with the water that proofs a spirit, for the same reason there is no effluent stream associated with the spirit itself.

This has two practical consequences. First, on a water-balance sheet, blending water is the smallest line and the only one with no downstream liability attached. Second, it is the one water in the plant whose entire cost is quality rather than volume — nobody is trying to use less of it, because using less of it means bottling at a higher strength. What matters is only that it is right.

That distinction is why a distillery can run an aggressive water-reduction programme across cooling, cleaning and mashing while tightening, rather than relaxing, the specification on the few hundred millilitres per bottle that reach the consumer. Those are not competing objectives. They are different water.

What should a buyer ask?

If you are assessing a distillery’s water story — as a trade buyer, an auditor, or a brand owner doing diligence on a contract bottler — six questions separate a real programme from a page in a brochure:

1. What is your water intensity in l/lpa, and what was it three years ago? A trend is worth more than a number.
2. Which effluent streams leave the site, and under what consent? Pot ale and spent lees should be named separately.
3. Where does the pot ale go? Syrup, land, sea, or digester — each is defensible; “we treat it” is not an answer.
4. How is copper in spent lees managed? The acidic, copper-bearing stream is the one an environmental regulator will ask about first.
5. If you operate in India, are you ZLD-certified, and by whom? After the 2024 SOP there is a defined verification route (TeamLease RegTech).
6. What is your blending water, and is it specified separately? A plant that lists it alongside cooling water has not understood its own process.

Glossary

• Pot ale — the residue left in the wash still after the first distillation of malt whisky; high in protein and organic load, acidic, and containing dissolved copper.
• Spent lees — the residue left in the spirit still after the second distillation; weaker in organic load than pot ale, acidic, copper-bearing, and the larger volume of the two.
• Draff — the spent grain left after mashing; sold as animal feed or digested for biogas.
• Spent wash (stillage, vinasse) — the equivalent of pot ale in molasses-based distilling; darker, hotter and several times stronger in COD.
• COD / BOD — chemical and biochemical oxygen demand, the two standard measures of how much oxygen an effluent will consume as it degrades. The higher the number, the more damage untreated discharge does to a watercourse.
• Melanoidin — the brown, Maillard-derived polymer responsible for the persistent colour of molasses spent wash; resists microbial breakdown, so it survives biomethanation.
• Biomethanation — anaerobic digestion of an effluent to convert its organic load into methane, reducing oxygen demand and recovering energy.
• Multiple-effect evaporator (MEE) — a staged evaporator used to concentrate spent wash to a solids content at which it can be incinerated or composted.
• Zero liquid discharge (ZLD) — a regulatory requirement that no liquid effluent leaves the site; water is recovered and recycled, and the remaining solids are disposed of by other routes.
• l/lpa — litres of water per litre of pure alcohol, the standard water-intensity metric in distilling.

Frequently asked questions

How much water does it take to make a bottle of whisky? Published intensity figures cluster between 10 and 25 litres of water per litre of pure alcohol (Singh et al., 2026; SWA). At the low end of that range, a 700 ml bottle at 40% ABV represents roughly three to seven litres of process water, almost none of which is in the bottle.

Is pot ale dangerous? Not to handle, but it is a serious pollutant if discharged untreated. At COD 47 g/L and BOD 25 g/L (White et al., 2020) it would deoxygenate a small watercourse, and the dissolved copper it carries is separately toxic to aquatic life.

Can pot ale or spent lees be reused as blending water? No. Both are the opposite of blending water: acidic, organically loaded, and in the case of spent lees copper-bearing. They are candidates for energy recovery or animal feed, not for dilution.

Why is Indian distillery effluent treated so much more strictly than Scottish effluent? Chiefly because it is far stronger. Molasses stillage runs at COD 100,000–150,000 mg/L (Mikucka & Zielińska, 2020) against pot ale’s 47,000, it resists decolourisation, and much of the industry sits inland on river basins under specific restoration programmes — conditions under which a discharge consent is harder to justify than a discharge ban.

Does zero liquid discharge apply to every Indian distillery? The Charter and the associated enforcement were directed at molasses-based distilleries, with the Ganga basin as the initial focus, and ZLD verification is now formalised through a CPCB SOP dated 3 July 2024 (TeamLease RegTech). Specific obligations depend on feedstock, location and the conditions in a unit’s own consent — the state pollution control board’s consent document, not a general article, is the authority for any given site.

Does blending water generate any effluent at all? The blending water itself does not — it leaves in the bottle. Producing it does: purification plant creates a concentrate stream, which is a separate question covered in our piece on sustainability, sourcing and purity certifications. What blending water never becomes is distillery effluent, because it is never used as a utility.

Is a distillery that cuts its water use also cutting its blending water? It should not be. Reducing blending water means bottling at a higher strength, which is a product decision, not an efficiency one. Water-reduction programmes target cooling, cleaning and mashing — the streams that become effluent. The few hundred millilitres per bottle that reach the consumer are the wrong place to look for savings and the right place to spend on quality.

Blenders International · Water Science series