Technology
How an electrostatic precipitator takes grease out of the air
No media, no bag, no consumable in the gas path. An ESP filter charges the particles and then collects them on plates — which is why it holds efficiency for years and why the only thing it asks of you is a wash.
- Two zones in one pull-out cell
- 12 kV ionisation, earthed collection
- 107 – 176 Pa pressure drop
- 95% – 98% on grease and smoke
The short answer. Cooking fumes pass an equalising screen that knocks out the coarse oil mist, then a 12 kV ionisation zone that puts a negative charge on everything still airborne, then a stack of earthed plates 7.5 mm apart that pull those charged particles out of the airstream. Clean air leaves at 1–2 mg/m³. The captured grease runs down the plates to a drain, and the cell is washed rather than replaced.
An electrostatic precipitator is not a filter in the sense a bag house or a HEPA panel is a filter. Nothing is trapped in a medium that slowly blinds over. The collecting surface is bare stainless steel, and the particle is moved onto it by an electric field rather than by being squeezed out of the air. That single difference explains most of the behaviour that follows: the near-flat pressure drop, the long service life, the indifference to grease loading until the plates are actually full, and the fact that efficiency is recovered completely by washing rather than partially recovered by replacement.
The consequence for specifying one is that the numbers that matter are geometric and electrical rather than surface-area based. Plate spacing, plate depth, field length, voltage stability and airflow distribution decide the result. A cabinet that is physically larger but has a shorter field, or a supply that sags under load, will underperform a smaller unit with better geometry — which is why comparing ESP filters by kilogram or by footprint tells you very little.
The pass, stage by stage
Four stages, one cell, no bypass
Air moves through the cabinet in a single pass. There is no recirculation loop and no path around the field — the cell fills the duct cross-section, so anything that enters the cabinet is treated.
Equalising screen
The first screen does two jobs: it flattens the velocity profile so the middle of the duct does not travel three times faster than the edges, and it catches the coarse oil mist and any solid carry-over before it reaches the field. Uneven velocity is the most common reason a correctly sized ESP still smokes at one corner.
Ionisation zone
Fine tungsten wires sit between earthed rods at 12 kV. The corona they produce charges every particle that passes within a few millimetres, grease aerosols and sub-micron smoke alike. This zone is deliberately shallow and separate: charging is a fast process that does not want to be disturbed by the collecting field behind it.
Collection zone
Parallel earthed plates, 7.5 mm apart, at low voltage relative to the wires. Charged particles migrate sideways and stick. The plates are deep — 195 mm of adsorption width — because residence time in the field, not plate count, sets how much of the fine fraction is caught.
Final screen and discharge
A last screen equalises the leaving air and stops any re-entrained droplet. What passes it is 1–2 mg/m³ — below the threshold at which a plume is visible from the roof, and well inside the 2.0 mg/m³ most municipal cooking-fume limits work to.

The four stages in a KL-D-J cell. Grease leaves the airstream at the collection plates and runs down to the cabinet drain.

Cells pull out on rails as a single unit. The plates are bare stainless — washing restores them completely.
Why geometry decides everything
7.5 mm is not an arbitrary number
Field strength between two plates is voltage divided by distance. Halve the gap and you double the field for the same supply — until the gap is small enough that grease bridging causes a short, or that a slightly warped plate closes it entirely. 7.5 mm is where collection efficiency, arcing margin and manufacturability meet for cooking effluent at these voltages.
Holding that gap is the harder half. A cell built to 7.5 mm ± 0.8 mm over a 195 mm depth, on 0.8 mm stainless formed in twelve press passes, behaves predictably for a decade. One built to ±3 mm develops weak bands where particles slip through, and the symptom looks like a failed power supply when it is really a tolerance problem.
- Spacing sets field strength; depth sets residence time
- Tolerance sets whether the field is uniform across the whole face
- Plate thickness sets how many wash cycles it survives without deforming
- Rails and handles set whether anyone actually pulls it out
The power supply
A supply that tracks the cell, not one that just energises it
As grease accumulates on the plates, the effective gap narrows and the cell's behaviour changes. A fixed supply either arcs and trips, or backs off to a voltage where it still runs — and quietly collects less. Constant-voltage control measures the cell continuously and holds the working point, so the last shift before a wash performs like the first.
Around that sits the protection layer. Cooking effluent is a dirty, humid, occasionally sparking load, and the electronics are expected to survive it for years without nuisance tripping a kitchen into shutdown.

The high-voltage section sits outside the gas path. Everything that touches grease is stainless and removable.
Cost in pressure
What the cell takes out of the fan
An ESP is a low-resistance device compared with a media filter, but it is not free. Pressure drop across the 13 KL-D-J models runs from 107 Pa at the small end to 176 Pa on the widest cabinet — roughly 0.43 to 0.71 in. wg. That figure belongs in the fan selection, next to duct friction, not on top of it as an afterthought.
| Model | Airflow (m³/h) | Airflow (CFM) | Pressure drop (Pa) | Pressure drop (in. wg) | Power (W) |
|---|---|---|---|---|---|
| KL-D-J40 | 4,000 | 2,354 | 107 | 0.43 | 300 |
| KL-D-J60 | 6,000 | 3,531 | 113 | 0.45 | 300 |
| KL-D-J80 | 8,000 | 4,709 | 133 | 0.53 | 300 |
| KL-D-J100 | 10,000 | 5,886 | 124 | 0.50 | 600 |
| KL-D-J120 | 12,000 | 7,063 | 131 | 0.53 | 600 |
| KL-D-J160 | 16,000 | 9,417 | 125 | 0.50 | 600 |
| KL-D-J200 | 20,000 | 11,772 | 129 | 0.52 | 1,200 |
| KL-D-J240 | 24,000 | 14,126 | 125 | 0.50 | 1,200 |
| KL-D-J320 | 32,000 | 18,834 | 121 | 0.49 | 1,200 |
| KL-D-J400 | 40,000 | 23,543 | 153 | 0.61 | 2,400 |
| KL-D-J480 (double) | 48,000 | 28,251 | 173 | 0.69 | 2,400 |
| KL-D-J480 (triple) | 48,000 | 28,251 | 152 | 0.61 | 1,800 |
| KL-D-J640 | 64,000 | 37,669 | 176 | 0.71 | 2,400 |
Figures are at nominal airflow on a clean cell and rise as grease builds, which is why the wash interval is also a fan-load decision. The airflow calculator adds this drop to duct friction and elbow allowances to give a total system pressure.
Dry and wet
The same field, two different problems
A dry ESP collects grease that is already an aerosol. A wet ESP — the KL-D-SPL arrangement — puts water in the gas path first, which changes what the cell is asked to do. The distinction matters most on solid-fuel cooking.
| Dry ESP (KL-D-J) | Wet ESP (KL-D-SPL ahead of the cell) | |
|---|---|---|
| What it is for | Grease and smoke already in aerosol form — the normal case for gas, induction and electric cooking. | Cooking that produces sparks, ash, visible smoke and radiant heat — charcoal, tandoor, solid-fuel grills, some wok ranges. |
| Fire risk | Low, provided the cell is washed on schedule. Grease load is the variable. | Lower still: the water stage quenches sparks and cools the gas before it reaches the field. |
| What the cell sees | Oily aerosol at duct temperature. | Cooled, humidified gas with the coarse fraction already washed out. |
| Water supply | None. Wash water is used off-line. | Continuous. Recirculation via the KL-D-QT tank where several units share a plant deck. |
| Wash interval | 7 – 30 days depending on cooking load; the KL-D-Q automates it. | Cell washes less often; the water stage needs its own sump management. |
| Typical chain | Hood → J → fan → discharge | Hood → SPL → J → fan → W → discharge |
Compared with the alternatives
Why not just use a media filter, a HEPA or UV?
Each of those is the right answer to a different question. On commercial cooking effluent, the question is a large volume of hot, oily aerosol that would blind a media filter in days.
| ESP filter | Media / baffle filter | HEPA | UV + ozone | |
|---|---|---|---|---|
| Mechanism | Electrostatic charge and collection on plates | Impingement and interception on mesh or baffles | Physical sieving through dense fibre | Photolytic breakdown of odour molecules |
| Grease removal | 95% – 98% | 40% – 70%, dropping as it loads | High, but only on pre-cleaned air | None |
| Sub-micron smoke | High — the fraction baffles miss is what the field is for | Poor | Very high | Partial, on odour only |
| Pressure drop | 107 – 176 Pa, essentially flat | Rises from ~100 Pa to several hundred as it loads | 250 – 500 Pa, rises to failure | Negligible |
| Consumable | None in the gas path | Replacement media, ongoing | Replacement panels, ongoing | Lamps every 8,000 – 12,000 h |
| Maintenance | Wash the cell | Swap the media | Swap the panel | Clean the quartz, replace lamps |
| Where it belongs | Ducted kitchen exhaust, upstream of the fan | First stage, and inside hoods | Not on cooking exhaust; on make-up or recirculation | Behind an ESP, for odour only |
In practice the stages stack. A baffle filter in the hood protects the ESP, the ESP removes the grease and smoke, and activated carbon or UV handles what is left of the smell. Expecting one stage to do all three is how systems end up failing inspection.

Voltage and current on the panel. A cell that is due a wash shows it here before it shows at the roof.
Efficiency over time
An ESP does not wear out. It loads up.
There is no medium to blind and no element to burn out. What happens instead is that grease accumulates on the plates and on the insulators, the effective gap closes, and the supply has to work harder to hold voltage. Efficiency falls slowly and then quickly, and the fall is fully reversible.
This is the single most useful thing to understand about the technology, because it turns maintenance from a cost into the mechanism that delivers the rated performance. A cell washed on schedule performs to specification for its whole service life. A cell left alone for a year will read 60% where it once read 95%, and the owner will conclude the technology does not work.
Questions we get before an order
ESP technology, answered plainly
What is the difference between an ESP and a HEPA filter?
A HEPA filter is a physical barrier: it traps particles in a mat of fibres, and it has to be replaced when it loads. An ESP filter charges particles and collects them on plates, and the plates are washed and reused. In a kitchen exhaust, a HEPA would be blinded by grease within weeks; an ESP is designed for exactly that grease load. HEPA belongs after the ESP, as a final polish on already-clean air, not instead of it.
What is the difference between a dry ESP and a wet ESP?
A dry ESP collects grease on dry plates and is cleaned by pulling the cell out and washing it offline. A wet ESP — or a water spray stage ahead of a dry cell — washes continuously or intermittently with water, which handles heat, sparks and very heavy grease loads but adds water supply, drainage and corrosion to the design. Kitchens with solid fuel or extreme wok loads use the wet arrangement; most others use dry.
Why 7.5 mm plate spacing?
Spacing sets the trade-off between collection area and the risk of short-circuiting across collected grease. 7.5 mm with tight tolerance gives a large collecting surface in a compact cell while leaving enough gap for the field to stay stable as grease builds. Holding the tolerance is what separates a cell that works for years from one that arcs.
What does the power supply actually do?
It holds the voltage across the cell at the optimum for collection as the load changes. Grease build-up changes the electrical load; a constant-voltage supply tracks it, so efficiency does not fall away between washes. A fixed supply either trips or runs below optimum, and both look the same from outside: no alarm, less collection.
Does an ESP filter remove odour?
No. It removes particulate — grease, smoke, oil mist. Odour compounds are gases and pass through. Odour needs an activated carbon stage, a UV/ozone stage or both, placed after the cell. If the complaint is smell rather than visible smoke, the specification needs an odour stage, not a bigger ESP.
How long does an ESP cell last?
A stainless cell in normal service lasts many years because it is washed rather than replaced, and because the washing is what keeps it from being damaged: cells fail when they are dropped, distorted or refitted at the wrong spacing. That is the practical argument for a self-cleaning cabinet on any site where manual washing would slip.
Next step
Send the layout. We will confirm the model.
Appliance list, hood length and the duct route are enough to come back with a model, an airflow figure and a drawing. OEM and ODM enquiries welcome.
