Water and rinse management: conductivity and drag-out

A panel leaving a degrease tank carries a film of liquid with it. At 60 µm average film over 1 square metre that is 60 millilitres — a sixth of a cup — and over a shift of a few hundred racks it becomes the entire reason a pretreatment line consumes water and generates effluent.

Rinse water management follows from that film. The volume of it sets the flow, the concentration in it sets the number of rinse stages, and the conductivity of the last tank it touches sets the adhesion result under the powder. Drag-out is the variable worth controlling, and most lines control the tap instead.

This page sets out the litres per square metre that realistic geometry carries, the counterflow ratio that turns a small flow into a clean rinse, the conductivity limits to hold at each stage, and the record that shows the line is running to them. The chemistry of the tanks themselves is covered in degreasing mild steel parts, and the line layout is in the seven-tank pretreatment walkthrough.

Drag-out: the number rinse water management starts from

Rinse water management starts with an estimate of drag-out: 0.1–0.15 l/m² for flat sheet after a 30-second drain, 0.2–0.3 l/m² for a pressed panel with hems and stiffening beads, and above 0.5 l/m² for a boxed weldment with blind corners and lapped joints.

The film thickness is what varies, not the area. A mirror-flat panel drains to well under 50 µm within half a minute because there is nothing to hold the liquid. A hem flange that has been closed over a wet surface holds a reservoir that weeps for the rest of the line. A box section with a 6 mm drain hole passes liquid slowly because the hole is at the wrong end.

Vertical orientation is the single biggest lever. A panel hung on edge drains under gravity; the same panel laid flat on a rack holds a pool in every bead and drains almost none of it. That one decision changes the drag-out by a factor of three and changes the rinse flow that follows.

The counterflow ratio that sets consumption

The rinse ratio is the volume of fresh water entering a stage divided by the volume of drag-out leaving the previous one. A ratio of 4 means 4 litres of water for every litre of dragged liquid.

In a single rinse tank at steady state, the carry-over concentration into the next stage is 1 ÷ (1 + ratio). At a ratio of 4 the next tank receives 20% of the previous tank's concentration, at 9 it receives 10%, at 99 it receives 1%.

In a cascade of n stages with the same total flow, the carry-over is 1 ÷ (1 + ratio)ⁿ. Two stages at a ratio of 9 give 1%, which is the 99% removal a conversion coating needs. Two stages at a ratio of 4 give 4%, and four stages at 4 give 0.16%.

That is the whole argument for a second tank. Getting from 90% removal to 99% in one stage costs ten times the water. Getting there with a second tank costs nothing but the tank, because the same flow is now split across two dilutions. A line that is failing adhesion because the rinse conductivity is high usually needs plumbing, not a bigger pump.

A worked consumption figure for one rack

Take a rack of six panels at 0.9 m² each, which is 5.4 m² of processed surface. At 0.15 l/m² of drag-out the rack carries 0.81 litres out of each tank. At 200 racks a day the line drags 162 litres a day into the rinse cascade, before any spillage or tank dumping.

Set the cascade at a ratio of 9 and the fresh water demand is 1,458 litres a day, about 1.5 m³, or 437 m³ over 300 working days. Add the evaporative loss from the 55–60 °C degrease tank, the blowdown needed to keep the bath titration in band, and the final deionised rinse, and the realistic figure lands near 2 l/m² of surface processed.

Now halve the drag-out. A 45-second drain, parts hung on edge and a 4 mm drain hole at the low corner of each bead take the film from 0.15 to 0.075 l/m². The water demand halves to 729 litres a day and the effluent load halves with it. Nothing else on the line changes, and the coating result is identical or better because the rinse is now doing less work.

That is why the sequence for a water reduction programme is drain time first, orientation second, rack design third, air knife fourth, and flow control last. Flow control at the end of the list is a tap. The first four items are free.

Conductivity: the reading that predicts adhesion failure

Conductivity measures dissolved ions, and dissolved ions are what remain on the surface when the water evaporates. Tap water in most Indian industrial areas runs 300–800 µS/cm; deionised water runs below 10 µS/cm. The gap between those two numbers is the gap between a rinse that passes and one that blisters.

Take the reading in the tank, not from a sample carried to the lab. A beaker sample drifts as it equilibrates with the air. A handheld meter with automatic temperature compensation, dipped into the overflow weir of the final tank, gives the number in five seconds.

The final rinse is the only one that has to be near-deionised. Earlier stages can sit at 800–1,500 µS/cm. What it does not tolerate is oil, and oil is not conductive — which is the trap. A rinse can read a perfect 40 µS/cm and still be useless because it is floating a film of drawing lubricant.

ISO 8502-9 covers the conductometric determination of soluble salts on a prepared steel surface, and it is the method to quote when a customer asks for a salt limit rather than a rinse limit. The Bresle method extracts the salt from a defined area and measures the extract; the acceptance limit is usually written in mg/m² of chloride or of total salts, and it belongs on the drawing only with the extraction method named.

What to measure, how often, and what breaks first

Five readings cover the whole line, and each one has a frequency that matches how fast it drifts.

Conductivity in every rinse tank is a shift check, read at the start and the middle of the shift. It moves within hours when a rack carries a heavy drag-out or when a top-up valve is left open. Log the number against the tank, not a summary average; a line that reads 20 µS/cm on the final rinse every morning and 180 µS/cm every evening is being topped up all day and overflowing only at night.

Bath titration of the degrease and the conversion tank is a daily check, done on a sample pulled before the first rack. Concentration falls as parts go through, and a bath that starts at 5% and ends the week at 3.5% is cleaning the first two hundred racks properly and the rest badly.

The water meter is a weekly reading converted to litres per square metre of processed surface. This is the number that catches a leak, a stuck float valve or a hose left running overnight, and it is the only one that shows consumption drifting before the bill does.

The oil skimmer and the overflow weir are a visual check at the start of the shift. A weir blocked with sludge drops the tank level, the float valve opens to refill, and the tank then never overflows. The water going in is fresh; the water going out has stopped, so the concentration in the tank climbs all week while the inlet runs continuously.

Soluble salt extraction on finished parts to ISO 8502-9 is a monthly check, and an adhesion cross-hatch on the same frequency. Together they close the loop between the conductivity number and the result the customer sees.

The failure that hides behind an automatic top-up

A float valve is not a rinse control. It maintains a level; it does not remove contamination, and a line fitted with one and no overflow behaves like a closed bath with a dilution drip. The clean water enters at the top, the oil floats, the sludge settles, and the part leaving the tank carries both.

The symptom is a batch that passes the water-break test on the visible face and fails adhesion on the flange. The cause is a rinse tank that has not overflowed in a month. Fix the weir, set the overflow to a measured rate, and put the conductivity log next to the tank where the operator can see it move.

Contamination that conductivity cannot see

Oil, metal fines and phosphate sludge all sit in a rinse tank without changing its conductivity by more than a few microsiemens. They are the other half of the rinse problem, and they cause more coating rejections than salts do.

Oil arrives from an under-performing degrease, from lubricant trapped in a hem, and from the press shop in the form of a film nobody removed before the part reached the line. It floats, it coats the part on withdrawal, and it blocks the conversion coating exactly as a salt film would.

Fines arrive from laser cutting, from the blast chamber and from grinding. They settle overnight and re-suspend when the next rack goes in, so the first rack of the morning coats worse than the last rack of the previous shift.

Sludge arrives from the conversion tank as carry-over on the rack, and it accumulates in the rinse that follows. A rinse tank with a visible sediment layer needs cleaning on a schedule, not on a complaint.

Filter the rinse on a bypass loop, skim the surface continuously, and clean the tank on a fixed interval. None of those three actions shows up in the conductivity reading, and all three show up in the adhesion result.

Effluent: the rinse is the treatment load

Every litre of drag-out that enters a rinse eventually leaves as effluent, carrying the chemistry of the tank it came from. Halving drag-out halves the neutralisation acid, the phosphate precipitation chemical and the sludge that has to be dewatered and disposed of.

The treatment itself is a sequence. Alkaline rinse water is neutralised to pH 6.5–8.5, phosphate-bearing water is precipitated with a coagulant and settled, and the clarified water goes to the effluent treatment plant before discharge. The consent limits are written in pH, COD, phosphate and suspended solids, and the plant is monitored against them continuously.

Under ISO 14001 the effluent is a monitored aspect with a legal obligation attached, which means the record is not optional. The log needs the daily pH, the dosing quantities, the sludge removal dates and the discharge test results. An auditor who sees a conductivity log that tracks the effluent pH can see the line and the treatment working as one system.

The cheapest effluent treatment on any pretreatment line is a 45-second drain. It costs nothing, it reduces the load at the source, and it makes every downstream tank and every chemical dose smaller.

What Ganesh runs

Parts are cut on the 12 kW fibre laser with the 40 ft × 16 ft bed or blanked on presses from 20 T to 300 T, formed on the 200 T 5-axis brake, welded on three CLOOS robotic cells or the manual stations, then blasted in the 16 ft × 16 ft chamber where the surface needs it and cleaned on the seven-tank pretreatment line. Powder goes on in the 18 ft × 14 ft booth and cures in the oven.

The seven tanks are separate vessels with their own control, which is what makes a counterflow cascade possible without mixing chemistries. Rinse water is metered and logged, the effluent treatment plant handles the discharge, and the plants hold ISO 9001:2015 and ISO 14001:2015. Canopies for Mahindra Powerol and elevator components for TKE run through the same route, and the paint shop page describes the sequence end to end while the plant and quality hub covers the effluent and utility records behind it.

What to specify, and what to ask a supplier for

Most drawings say "powder coat, 60–80 µm" and nothing about the surface under it. These fields close that gap.

• Substrate and surface condition on arrival: oiled, pickled, galvanised or rusted.
• Cleaning route if the customer specifies one, or a note that the part must be water-break free before conversion.
• Soluble salt limit in mg/m² with the extraction method, where the service is aggressive.
• Whether a deionised final rinse is required, and the conductivity limit that proves it.
• Dry-off temperature, and any component that cannot take it.
• Whether the part can be dipped at all: length, weight and the position of drain holes.

Ask a supplier for the conductivity log of the final rinse and the effluent pH record for the last month. Both are one page, and both are hard to produce if the line is not being run to a number.

When a rinse cascade is the wrong answer

Very long parts do not fit a tank. A 6 m bent section goes through spray pretreatment, where the rinse is a nozzle bank and the drag-out is replaced by a film that has to be rinsed off in seconds. The conductivity principle is the same; the counterflow arithmetic is not.

Galvanised and aluminium substrates carry zinc and aluminium ions into the rinse, and the pH window is narrower than for mild steel. Sharing a rinse cascade between mild steel and galvanised parts moves zinc into the steel line and causes adhesion faults on both.

And in a plant where water is scarce or the discharge consent is tight, the answer can be an evaporator or a closed-loop rinse rather than a bigger cascade. That is a capital decision that starts from the litres per square metre figure.

Reduce the drag-out, prove the rinse with a meter, and treat the effluent as the same system. Send the drawing and coating requirement through the RFQ and the quotation states the rinse route.