EN

Lighting calculator: how many light fixtures (lux)

How many panels, battens or downlights does a room actually need? The lumen method answers by starting from the illuminance you want in lux, the floor area, and the luminous flux of one fitting in lumens — never its wattage, which has told you nothing about light output since LEDs arrived.

Two coefficients stand between the light emitted and the light that does the work: the utilisation factor, which depends on the shape of the room and how light the surfaces are, and the maintenance factor, which allows for dirt and ageing.

The mounting height asked for is not the ceiling height: it is the distance from the working plane (0.75 m above the floor in an office, floor level in a corridor) up to the fitting. Confusing the two throws off the room index, and the whole calculation with it.

The room and the fittings

Room dimensions

Working plane to fitting. Office 2.80 m to the ceiling, desk at 0.75 m: h = 2.05 m.

Target illuminance

Typical values: Corridor, stairs 100 lx · Living room, bedroom 150 lx · Kitchen, bathroom 300 lx · Office, classroom 500 lx · Workshop, general assembly 500 lx · Fine work, drawing office 1000 lx.

Printed on the box, in lumens. A common 600 × 600 LED panel puts out 3,000 to 4,000 lm.

Coefficients and wattage (optional)

Leave empty to estimate it from the room index. The manufacturer's photometric table always wins.

0.8 in a clean, maintained space; 0.6 to 0.7 in dusty conditions.

To get the installed load and the power density in W/m².

2.67 m2.50 mL = 8.00 mW = 5.00 m3 × 2 = 6 fixtures, plan view
Worked example, drawn to the demonstration values. Each fitting sits in the middle of its own cell, so the spacing at the wall is half the spacing between fittings — which keeps the corners out of shadow. The figure redraws with your own values once you calculate.

Lumen, lux and candela: three quantities, only one of which you can see

Three units, three jobs, and only one of them is what a desk looks like. The lumen is the flux leaving the lamp — all directions added up. The candela is that flux per steradian, so a tight spotlight can be thousands of candelas from the same 1,000 lm that a frosted globe smears around the room. The lux is lumens per square metre on a surface. That is illuminance. That is what you see.

This lighting calculation is a lux calculation. Lumens go in so we can count fittings; candelas stay on the photometric file, which the lumen method never opens. A 1,000 lm downlight on a 1 m² patch is 1,000 lx if every lumen lands there, and a few dozen lux if it does not. The utilisation factor is how we admit that they do not.

Why watts stopped telling you anything about light output

A 100 W incandescent was about 1,300 lm, so watts used to be a slang for brightness. LEDs broke that slang. A 10 W LED can be 1,000–1,500 lm; a 36 W fluorescent tube around 3,000. Efficacy is lm/W, and it now runs from a tired 80 to well over 160 on a decent panel, past 200 on the best boards.

Compare lumens when you are counting fittings. Compare watts when you are counting kilowatt-hours. This page will even print W/m² once it has N and the wattage per fitting — that is a running-cost number, not a brightness number. In fact two 3,000 lm panels at 25 W and 40 W light a room the same way and heat it differently.

Reference illuminances room by room, and what workplace regulations require

EN 12464-1 (and the national versions that cite it) gives maintained average illuminances: what you still have after dirt and ageing, not the sparkle on day one. A few workplace anchors:

  • Corridors, circulation — 50 to 100 lx
  • Offices, classrooms — 500 lx
  • Drawing, fine inspection — 750 to 1,000 lx
  • Rough assembly — 200 lx; fine assembly — 500 lx
  • Sales floors — 300 to 500 lx

Homes are not under that standard. 100–200 lx in a living room, 300–500 in a kitchen, 200–300 in a bathroom is the usual advice, not a legal floor. The target you type on this form is the maintained average on the working plane. If you type 500 and then design for a dirty factory with a 0.6 maintenance factor, you are asking for more initial flux than a clean office at the same 500.

Working plane and mounting height: the mistake that skews the whole calculation

The working plane is where the lux is specified — 0.85 m for a desk, 0.75 m for a lot of seated work, the floor for a corridor, a bench height in a shop. Mounting height h is from that plane up to the luminaire, not from the floor to the ceiling. A 3.00 m room, desks at 0.85 m, recessed panels: h = 2.15 m. Type 3.00 and the room index K collapses, the utilisation factor you pick is the wrong row, and N comes out as a different project.

h = hceiling − hworking plane

Pendant fittings make h smaller still. High-bay industrial is the opposite: a 8 m mounting height on a warehouse floor is a lighting calculation that barely resembles an office, even at the same lux target.

Room index K: what it measures and why it governs efficiency

K is a shape number: how wide and long the room is, compared with how high the lights sit. A large low room (open-plan office) has a high K; a tall small room (a shaft, a stair) has a low one. High K means more of the flux finds the working plane, including via the walls. Low K means the walls eat a bigger share.

K = (L × W) / (h × (L + W))

Typical range 0.6 to 5. You do not pick a utilisation factor from a feeling; you take K to the manufacturer's table. That is the whole job of this index. A square room and a long thin room with the same area do not share a K, and they should not share a U.

Utilisation factor: reading a manufacturer's table instead of taking an average

Utilisation factor U (sometimes UF, sometimes “coefficient of utilisation”) is useful flux on the working plane divided by lamp flux. It folds in the luminaire's polar curve, K, and the three reflectances. 0.5 is a middle-of-the-road guess, not a design. Real tables run from about 0.3 (dark cave, a lot of upward light wasted) to 0.8 and a bit (pale office, a direct batwing that was made for this K).

The table is for that fitting, not for “LED panels” as a species. A wide-distribution panel and a narrow high-bay at the same K are different rows. If you do not have the table, U = 0.5 is an order of magnitude so you can see whether you need eight fittings or forty. It is not a tender.

Ceiling, wall and floor reflectances: the room lights itself too

A slice of every lumen hits a surface and comes back. Pale rooms are free fittings. Typical ρ:

  • Ceiling — white 0.80, light 0.70, dark 0.50, very dark 0.30
  • Walls — white 0.70, light 0.50, dark 0.30, very dark 0.10
  • Floor — light 0.30, medium 0.20, dark 0.10

A white ceiling and light walls can lift U by 20–30 % against a dark brick box. The utilisation table is entered with a triplet, often written 70/50/20. Paint the walls and you may drop a whole fitting. Type 70/50/20 on a charcoal studio and N will be optimistic.

Maintenance factor: dirt, lumen depreciation and how often to clean

Maintenance factor M (the form's M; some texts say MF or “light loss factor”) is why a new installation is brighter than the standard asked for. Dust on the diffuser, lumen depreciation of the LED package, walls that yellow. Clean office, yearly wipe: around 0.8. Ordinary: 0.7. Dusty shop, rare cleaning: 0.6 or below.

Emaintained = Einitial × M · Φt = E × S / (U × M)

The lumen method sizes for the maintained value, so a lower M means more initial flux, hence more fittings (or a bigger lumen package). It is cheaper, sometimes, to clean than to hang extra panels. It is never cheaper to type 0.9 on a mill.

Uniformity and maximum spacing: the spacing-to-height rule

Average lux can hide a dark desk. Uniformity U0 = Emin / Eaverage. Offices typically want ≥ 0.6; circulation can live with 0.4. The lumen method does not compute U0. Spacing does, roughly: keep the gap between fittings around 1.5 × h for a diffusing general luminaire, closer to 1.0 × h for a narrow optic or for precision work.

N from this page can be “enough lumens” and still be “too few rows”. If the suggested layout wants 1.8 × h, add fittings or pick a wider distribution. The extra lux is a side-effect; the point is the holes between the cones of light.

General, localised and task lighting: layering beats brightening

You do not have to hang 500 lx on every square metre because one bench needs 500. General lighting can sit at 300, a desk lamp or a local pendant takes the workstation to 500, and a bench light covers the occasional job. Less installed load, less glare in the circulation, a room that does not look like a theatre set.

  • General — uniform base on the working plane.
  • Localised — extra on a known desk or machine.
  • Task — extra for a job that comes and goes (island, press, inspection).

This calculator is a general-lighting count. If you layer, type the general target (300, say), not the task target, or you will light the corridor like a drawing office.

Glare (UGR) and colour temperature: comfort beyond the lux figure

500 lx can be a pleasure or a headache. UGR (unified glare rating) is the discomfort number: EN 12464-1 wants UGR ≤ 19 in offices, ≤ 22 in many workshops. A bare lamp in the field of view will blow that, whatever the lux. Photometric files and a layout, not this form, give UGR.

Colour temperature is the other comfort knob. 2,700–3,000 K is warm domestic; 4,000 K is the office default; 6,000 K is a cold “daylight” that some people like in a warehouse and almost nobody likes at a screen all day. CRI (Ra ≥ 80 typical, ≥ 90 for colour work) is yet another axis. None of them change N. All of them change whether anyone will thank you for N.

Annual running cost, dimming and presence detection

Installed watts × hours is the bill. A 12 × 30 W office, 10 h/day, 250 days, is 900 kWh/year before anyone talks about dimming.

kWh/year = PkW × hours/day × days/year

Daylight-linked dimming can take 30–60 % off a perimeter zone. Presence detection in corridors and toilets can take half. A time schedule stops the “we left the panels on all weekend” story. The fittings you counted on this page are the full-on load. Controls are how that load is not the annual load. They do not let you hang fewer panels unless you are willing to miss the maintained lux on a dark January morning.

Why N is rounded up, then E is worked backwards

Φt / Φ is rarely an integer. You round up, because a fraction of a panel is not a product. The actual maintained illuminance is then N × Φ × U × M / S, which is why the result can sit a few percent above the target. That is not a bug. It is the nearest catalogue count. If it sits 20 % high, you can drop a row or pick a smaller lumen package rather than live with a permanently over-lit room.

Lumen method vs a point-by-point layout

The lumen method is an average. It assumes a reasonably regular grid and a room that is a rectangular box. Atria, coffered ceilings, a single row of wall-washers, emergency-only lighting — those want a photometric layout (point-by-point, Dialux or similar). Use this page to size the order of magnitude and to check a quote. Do not use it to sign off a museum.

Panels, battens and downlights for the same flux

3,000 lm is 3,000 lm on the lamp sticker. A 600×600 panel, a T8 LED batten and a cluster of GU10 downlights with the same total flux do not share a utilisation factor or a spacing rule. Downlights are usually closer together (the 1.0 × h end) and waste more on the ceiling void if they are recessed and poorly reflector-ed. Battens on a workshop ceiling can be a high U. Swap the fitting type, retake U from the right table — do not recycle 0.5 because last week's office used 0.5.

h is mounting height above the working plane, U comes from a table for that luminaire at that K and those reflectances, and M is a maintained-lux number, not an initial-lux number. Guess all three and the count of fittings is a guess with extra decimal places.

Related calculators