Case Study: Electrostatic Coating Line Transfer Efficiency
A European metal furniture plant moved from air-assisted airless guns to a reciprocator electrostatic coating line: transfer efficiency 38 to 85 percent.
A steel cabinet plant bought 133 tonnes of paint in one year and landed 51 tonnes of it on the parts. The other 82 tonnes met the booth filters, the floor grating and the stack. At 7 euro per kilogram that overspray cost about 517,000 euro a year and returned nothing but filter changes.
The film that landed was uneven. Panel edges measured 90 to 95 µm of dry film while pockets 30 to 40 mm deep measured 20 to 30 µm, against a 65 µm target inside a 40 to 80 µm band. First-pass yield ran at 93.2 percent, so 6.8 percent of the coated surface went to a hand rework pass.
This account describes a European plant of that class, not a named customer, and every figure sits inside published industry ranges. The line replaced two reciprocating air-assisted airless guns with two direct charge electrostatic rotary bells on one reciprocator. Transfer efficiency moved from 38 to 85 percent, inside the 65 to 90 percent band published for rotary bells.
Case Snapshot
The table below sets the reference configuration for the rest of this account.
| Item | Reference configuration |
|---|---|
| Industry | Steel furniture and cabinet making, liquid primer and topcoat |
| Duty | Both faces of welded steel cabinets and flat panels, one primer pass and one topcoat pass |
| Line | Overhead conveyor at 3.0 m/min, 3,200 productive hours a year, 0.50 m² of surface per metre of conveyor |
| Part family | Box cabinets and panels 900 mm tall, 500 mm wide, with recesses 30 to 50 mm wide and 20 to 40 mm deep |
| Original setup | Two reciprocating air-assisted airless guns, hand touch-up in the booth, transfer efficiency 38 percent |
| System chosen | Two direct charge electrostatic rotary bells on one reciprocator, 45,000 rpm |
| Operating point | 70 kV, 120 mL/min per bell, 250 mm standoff, 200 L/min shaping air |
| Coating | Solvent-borne 2K acrylic, 48 percent volume solids, 420 g/L VOC, resistivity 5 to 50 megohm-cm |
| Booth condition | 20 to 25 °C, 45 to 65 percent relative humidity, 0.35 m/s downdraft |
| Film target | 65 µm dry total, held inside a 40 to 80 µm band |
| Grounding limit | 1 megohm from part to ground rail, hooks stripped every two weeks |
The line ran 288,000 m² of coated surface a year. That figure anchors every later number, because paint use, solvent load and rejects follow area rather than part count.
The Challenge
The mass balance was the first problem, and it was invisible on the shop floor. Two air-assisted airless guns at 250 mm standoff put about 38 percent of the atomised paint on the parts. The rest flew past the panels into the filter bank. Nothing measured that split, so it was handled as a paint price problem.
The film distribution was the second. Air throws drops where the air stream carries them, so a box cabinet takes more paint on the faces and edges square to the guns and less in the pockets. Gauges showed 90 to 95 µm on edges against 20 to 30 µm in the recesses. Operators raised delivery to cover the thin places, which pushed the edges further out of band and brought sag onto the vertical faces.
Rework was the third. Recesses reading under 40 µm after the oven were scuffed and hand sprayed at a station beside the line. That station handled about 19,600 m² a year, or 6.8 percent of output.
Booth load was the fourth. About 82.7 tonnes of wet paint a year met the filters, the grating and the ducting. The plant changed 52 sets of media a year and spent 46 crew hours a month on booth cleaning, sludge removal and gun washing.
Solvent was the fifth. The coating carries 420 g/L of solvent, about 323 g per kilogram of paint. At 133 tonnes of purchased paint a year the line sent 43.1 tonnes of solvent vapour to a recuperative oxidiser, so overspray was paid for twice.
The paint kitchen was the sixth. The plant had no meter for resistivity, no record of the solvent blend and no viscosity target. When the supplier changed a reducer, the same settings ran on a paint with different charge behaviour, and nothing on the line could show it.
Grounding was the seventh. The conveyor and hangers had never been checked with a resistance meter. Hooks carried cured paint from months of service, and a cured 60 µm film does not pass charge. Two identical panels could come off one pass with films 20 percent apart.
The Solution
The plant wrote a duty for the booth instead of treating it as a paint supply. The duty starts with the surface rate in square metres per minute, converts it to grams per minute, and only then decides what applies it.
Two air-assisted airless guns left the reciprocator and two direct charge rotary bells went on. A cup at 45,000 rpm shears the paint into drops with a Sauter mean diameter near 25 µm, and the cup sits at 65 to 75 kV. Air still shapes the pattern, but the field steers the deposit. The electrostatic spraying guide covers the charging arrangement behind that change.
Grounding came before any tuning. Hangers moved to a bare steel contact hook, rail joints were bonded, and every hook is read with a 500 V insulation tester at shift start. The limit is 1 megohm from part to ground rail, and a hook that fails goes to stripping.
Two parameters joined the purchase specification: resistivity of 5 to 50 megohm-cm, and spray viscosity of 9 to 18 s in a Ford 4 cup at 20 °C. A bench meter reads resistivity on every batch, with the temperature noted on the batch sheet. A batch below the band gets 2 to 4 percent of polar reducer rather than a trial run.
One settings sheet per colour carries the kV, bell speed, fluid rate per bell, shaping air and stroke pattern. Voltage is not one setting: 70 kV covers the flat faces and 55 to 60 kV covers the recess pass, because a lower field penetrates a pocket better. The old guns left no such record, and the airless paint spraying guide sets out the tip and pressure chain the plant had run by feel.
Booth air came next. Downdraft is held at 0.35 m/s across the working zone, measured at the panel rather than at the plenum, and shaping air sits at 200 L/min per bell. Door openings and duct joints were sealed, because a cross draft of a few tenths of a metre per second pushes a charged cloud off a panel.
Measurement came last. The routine is a film gauge reading at five points per panel, resistivity per batch, ground resistance per hanger, a pattern check per shift and a log of bell current in microamps. Rising bell current at a stable voltage points at a conductive path.
The Engineering Detail
Charge, field, and how a charged drop behaves
A rotary bell is a direct charge device. The cup sits at 65 to 75 kV, and the paint film on the cup surface takes up charge as it flows out to the edge, so drops leave already charged. The grounded steel panel completes the circuit through the conveyor. A corona gun reaches the same result by ionising air at a raised pin, which is how hand held electrostatic guns charge their paint.
Field strength at the part is roughly bell voltage divided by standoff, about 280 kV/m at 70 kV and 250 mm. That figure sets the balance of forces on the drop. Air carries the drop for most of the flight and the field takes over in the last 100 to 200 mm, where the image charge in the grounded steel pulls it in. The object being steered is small: a 25 µm drop of paint at 1.30 g/cm³ holds about 11 nanograms of material.
Field lines bend around the part, which produces wrap. A charged cloud that passes the front face follows the field lines around the edge and lands on the back. The plant measures about 15 µm of extra film on the rear of a cabinet when one bell runs alone, which is why the two bells work opposite sides and are balanced.
The conductivity window and what the meter reads
Paint has to hold charge long enough to reach the part, then release it on landing, and both sides of that balance come from one property. Published practice puts the working window for direct charge solvent-borne paint at a resistivity of 5 to 50 megohm-cm, with the sweet spot near 20 to 30. Read as conductivity, the same band is 0.002 to 0.02 mS/m.
The two ends fail in different ways. Below about 1 megohm-cm the paint conducts enough to drain charge back up the fluid line, so the supply cannot hold voltage, bell current climbs and the coat arrives thin. Above about 100 megohm-cm the drops take too little charge and the wrap fades, which shows as poor recess coverage and a thin colour on the returns.
A waterborne paint, or one carrying conductive pigment, sits far below this window at roughly 10 to 250 ohm-cm. That material needs an isolated supply blocked from ground, which is a different machine rather than a different setting, and this plant stayed solvent-borne for that reason. The same resistance also keeps the supply from shorting.
Viscosity acts at the other end. Drop size on a bell follows cup speed, fluid rate and viscosity, so paint outside the 9 to 18 s Ford 4 band changes the drop spectrum without a machine setting moving. Thick paint runs larger drops and a coarser finish, while thin paint sags more easily on a vertical panel.
Faraday cages, recess geometry and back ionization
A recess shields itself. Charge collects on the lips of a pocket, and the field from that charge opposes the field from the bell, so the field inside can fall close to zero. Drops then pass over the opening instead of entering it. This is the Faraday cage effect, and it is a property of geometry.
The working rule is that the field reaches about as deep as the recess is wide. A pocket 40 mm wide and 25 mm deep coats acceptably from a bell at 250 mm. A pocket 30 mm wide and 40 mm deep takes a partial coat. A slot 20 mm wide and 50 mm deep takes almost nothing at any voltage, and the plant still hand coats that feature.
Voltage has a narrow useful range. Raising the field does not open a cage, because the shielding charge on the lips grows with the applied field. Ion current rises instead. Free ions join the cloud, land on the wet film and cannot reach the steel, so the film charges until it discharges through its own thickness. The mark is a crater ring, sometimes with a starburst centre. That is back ionization, and a trial at 85 kV produced it.
The fix is a lower voltage and a different approach angle, not more paint. The recess pass runs at 55 to 60 kV with a slower reciprocator stroke, and tall cabinets get a pass from two directions, since a pocket that is a cage from the front is an open face from the side. A small directed air jet across the opening moves ions out of the pocket before they settle on the film.
From square metres per minute to grams per minute
The sizing chain runs from line speed to fluid rate in one direction and back again in the other.
| Step | Value | Arithmetic |
|---|---|---|
| Conveyor speed | 3.0 m/min | set by the oven and the line balance |
| Coated surface per metre of conveyor | 0.50 m² | from the part mix and hanger pitch |
| Area rate | 1.50 m²/min | 3.0 x 0.50 |
| Dry film target | 65 µm | spec for primer and topcoat together |
| Volume solids | 48 percent | paint data sheet |
| Wet film | 135 µm | 65 / 0.48 |
| Wet paint per square metre | 176 g | 135 µm at 1.30 kg/L |
| Paint that must land on the part | 264 g/min | 176 x 1.50 |
| Spray delivery at 85 percent efficiency | 311 g/min | 264 / 0.85 |
| Fluid rate per bell, two bells | 155 g/min | 311 / 2 |
| Volume rate per bell | 120 mL/min | 155 / 1.30 |
| Annual paint purchase | 59.7 t | 311 g/min x 60 x 3,200 h |
| Reverse check | 207 g/m² | 59.7 t / 288,000 m² |
The reverse check is the useful part. Divide the annual purchase by the annual coated area and the answer is 207 g of paint per square metre. Subtract the 176 g that must land on the part and the overspray is 31 g per square metre, or 15 percent of the paint purchased. The same chain on the old line returns 463 g sprayed per square metre for the same 176 g landed, which is a 38 percent transfer efficiency.
Because the line runs at a steady 3.0 m/min, the average and the instantaneous rates agree.
Grounding faults arrive as coating defects
A part at full ground potential sits at 0 V against a bell at 70 kV, so the whole field is available for deposition. A part isolated by a painted hook floats toward bell potential and the field between them collapses. Paint still arrives, because air carries it, but the wrap goes with the field.
On a gauge the symptom is a film 15 to 30 percent thin on the face nearest the bell, with a band of correct thickness where the hanger made contact. An isolated part can also flash over to its hook, and that spark leaves a small round bare spot which a painter reads as contamination.
A 500 V insulation tester reads under 1 kilohm on a clean steel hook and over 10 megohm on a hook carrying a cured film. The plant set a 1 megohm limit, logged every hanger at shift start and moved stripping to a two week cycle. That routine removed a class of defects which no paint change would have touched.
Air flow across the part
Booth air does three jobs at once: it shapes the pattern, it carries the cloud to the part, and it clears overspray. The first two want low velocity and the third wants high velocity, so the balance is narrow.
Downdraft here is 0.35 m/s across the working zone, inside the 0.2 to 0.5 m/s band used for liquid coating booths. Above about 0.5 m/s the downdraft strips the charged cloud sideways before the field can place it, and efficiency falls while the pattern still looks correct. Below 0.2 m/s overspray stays in the booth and lands back on wet parts as dry spray.
Shaping air sits at 200 L/min per bell. At 350 L/min the drops arrive too fast to follow the field, so the recess film goes thin while the flat faces stay right. At 100 L/min the pattern collapses toward the cup centre and the reciprocator has to stroke faster to cover the same height.
Cross drafts are the hidden fault. An open door, a leaking duct joint or a badly aimed grille can move air across a panel at 0.3 m/s and shift the deposit along one edge. The plant reads an anemometer across the conveyor twice a shift, at working height.
The film band, the rheology and the verification routine
The spec is 65 µm dry inside a 40 to 80 µm band. The lower limit protects appearance and corrosion, and the upper limit protects cost and sag. Wet film is about twice the dry figure at 48 percent volume solids, so one pass stays under about 150 µm wet and two passes reach the total without a sag risk on a 900 mm face.
Rheology decides whether the wet film stays put. High shear viscosity sets how the paint atomises on the cup, and low shear viscosity sets whether it runs down the panel after landing. A slightly thixotropic paint holds a heavier wet film without sagging, which is why the supplier tuned the rheology package alongside the conductivity.
The verification sheet takes about 20 minutes a shift. It records film at five points on a panel, resistivity and temperature of the batch in use, bell current per bell, ground resistance of the last hanger set and booth downdraft. A drift in bell current at stable voltage points at a conductive path, and a drift in film at stable current points at grounding or at a paint batch.
The Results
The figures below compare the twelve months before the retrofit with the twelve months after, at the same line speed and the same 288,000 m² of coated surface.
| Metric | Before | After | Unit |
|---|---|---|---|
| Transfer efficiency | 38 | 85 | percent |
| Paint purchased per year | 133.4 | 59.7 | tonnes |
| Paint used per square metre | 463 | 207 | g |
| Paint cost per square metre | 3.24 | 1.45 | euro |
| Paint that missed the part | 82.7 | 8.9 | tonnes per year |
| Dry film on edges and recesses | 25 to 95 | 52 to 68 | µm |
| First-pass yield | 93.2 | 98.4 | percent |
| Surface sent to rework | 19,600 | 4,600 | m² per year |
| Solvent sent to the oxidiser | 43.1 | 19.3 | tonnes per year |
| Booth filter changes | 52 | 19 | per year |
| Booth cleaning hours | 46 | 18 | hours per month |
| Bell and hook maintenance hours | 12 | 26 | hours per month |
| Retrofit capital, one off | none | 950,000 | euro |
Paint was the largest gain by a wide margin. Purchase fell from 133.4 to 59.7 tonnes, a 55 percent cut, while the paint landing on the parts held steady at about 51 tonnes. The ratio between the two years matches the transfer efficiency ratio, at 2.23 against 2.24, which is the check that the saving came from deposition rather than from lighter films.
The quality numbers followed the film. First-pass yield rose from 93.2 to 98.4 percent and the rework area fell from 19,600 to 4,600 m² a year. The dry film mean stayed near 65 µm while the spread tightened from 25 to 95 µm down to 52 to 68 µm. The hand station stayed for the deepest slots, because no field reaches that geometry.
Two numbers moved the wrong way, and both are real costs of the change. Bell and hook maintenance rose from 12 to 26 hours a month, because rotating components and a two week stripping cycle replace two static guns. Booth and oxidiser costs fell by about 60,000 euro a year and the paint saving is 517,000 euro a year, so the 950,000 euro retrofit pays back in just under two years on paint alone.
Why This Matters
The first thing that transfers to another line is the order of the work. Surface rate in square metres per minute comes first, grams per minute follows from the film spec and the volume solids, and the applicator is chosen against that delivery figure. A plant that buys a bell and then asks what to spray keeps the old overspray and adds a capital bill.
The second is that transfer efficiency is a measurement rather than a claim. Published bands put conventional air atomising guns at 20 to 35 percent, airless and air-assisted airless guns at about 30 to 45 percent, and electrostatic rotary bells at 65 to 90 percent. Any line can place itself on that scale from paint purchased, coated area and film thickness.
Four thresholds carry into any electrostatic coating line. Keep part to ground rail resistance under 1 megohm, checked with a meter rather than by eye. Hold paint resistivity inside 5 to 50 megohm-cm, or 0.002 to 0.02 mS/m, and read it per batch. Run the recess pass at 55 to 60 kV rather than at the flat face setting, since a higher field sharpens a Faraday cage instead of opening it. Keep the wet film per pass under about 150 µm, or a vertical face sags before the oven.
Two rules cover geometry and air. A recess coats to roughly its own width in depth, so a slot 20 mm wide and 50 mm deep belongs to a hand gun. Downdraft stays between 0.2 and 0.5 m/s across the part, and any cross draft past a couple of tenths of a metre per second moves the deposit before the pattern looks wrong.
The method runs on any conveyor where metal parts take a liquid coat, from enclosures and shelving to appliance panels. If your line is fighting overspray, edge build or thin recesses, send the part drawing, the line speed and the film spec to the BoreJet application team through the contact page. The gun and lance range used for touch-up on this line is on the spray gun page, and the caps and tips for any station that stays air atomised are on the air atomizing nozzle page.
Your Duty May Differ
Send the Vessel Drawing and the Pump Curve.
We size the head class, the count and the operating point for your tank, not a catalogue guess. A coverage test on the actual vessel beats any estimate.