Nobody Sees Lux
Illuminance is what arrives at a surface. Luminance is what leaves it toward an eye, and the eye responds only to the second. Five hundred lux landing on a sheet of white paper and five hundred lux landing on an oiled cast-iron machine bed satisfy the same specification and describe two entirely different jobs, because the paper returns most of what it receives and the casting returns almost none of it. Nearly every complaint that a compliant scheme feels dark lives in that gap.
Three variables decide whether a task is comfortable. How small the detail is that has to be resolved, how much contrast that detail carries against its background, and how old the person resolving it happens to be. Only the third is fixed, and it moves in one direction: a retina at fifty-five admits roughly a third of the light a twenty-year-old's does under identical conditions. In a workshop where the average tenure runs to two decades, that is a design input and not a courtesy.
So a lighting specification worth signing names four things, never one. The maintained illuminance. The plane and the area it applies to. A uniformity ratio across that area. A ceiling on glare. Omit any of them and you can produce a scheme that defends itself on paper and gets rejected by the people standing under it.
Add a fifth where the work is judged by eye. Paint booths, timber shops, print rooms, food preparation, electronics assembly with colour-coded conductors: all of them will accept a scheme that meets the lux figure and then quietly relocate a bench toward the nearest window.
Maintained Means at the End of the Cycle
Every illuminance figure in a standard is a maintained figure — the value the installation must still deliver at the end of its maintenance cycle, not the day it was energised. The maintenance factor is the product of four separate depreciations: lamp lumen maintenance, luminaire dirt depreciation, room surface dirt depreciation, and a survival factor covering the fittings that will have failed and not been replaced.
Dirt does most of the damage and it is entirely site-specific. A sealed IP65 optic in a joinery shop holds its output far longer than an open reflector in the same room, because the sawdust settles on glass it can be wiped off instead of on a lamp cavity nobody opens. Welding fume, cement dust, cooking grease and rubber particulate each attack a different part of the fitting, and the same catalogue product carries different maintenance factors in each of those rooms.
The consequence is that a correct design is deliberately over-lit on day one, by a factor of one over the maintenance factor. Owners notice. Where the luminaires dim, that is a commissioning setting and the complaint disappears. Where they do not, you are choosing between a room that is uncomfortable in year one and a room that is short in year five, and the second choice is the one that generates a retrofit.
The maintenance factor also assumed a cleaning interval, and that assumption has to leave the design office with the drawings. Write the interval and the assumed reflectances into the operation and maintenance file. A factor of 0.9 in a sawmill is a fiction that turns into an argument three years later when somebody brings a meter.
Task, Surround, Background
EN 12464-1 divides the working plane into three zones and gives each its own number: the task area itself, an immediate surround band around it, and the background beyond that. Lighting an entire floor to task level lights the floor. A forty by twenty metre assembly hall with six benches in it does not need seven hundred and fifty lux across eight hundred square metres, and the schemes that do are the ones whose energy figures never survive review.
Where the tasks are fixed — a CNC control station, an inspection table, a goods-in desk — local lighting beats general lighting on cost, on quality and on the glare rating simultaneously. Where benches move quarterly, the task band has to be generous enough to absorb that, and the honest answer is a higher general level with the surround provision doing the work.
Task areas are defined by the work and not by the room. A bench is a task area; the gangway behind it is a background. The mistake that shows up in tender documents is a single illuminance figure applied to a whole floor plate, which either over-lights three quarters of it or under-lights the quarter that matters.
The bands below recur throughout EN 12464-1 for indoor work places. The standard adopted on your project sets the governing value, a client specification can and frequently does sit above it, and the figure written into the contract is the one you design to.
| Work | Typical maintained illuminance | What drives the figure |
|---|---|---|
| Circulation, corridors, plant rooms | 100 lx | Safe movement and recognising a hazard, no task resolution |
| Storage racks, occupied continuously | 200 lx | Reading labels on a vertical face, not a horizontal plane |
| Filing, copying, reception | 300 lx | Intermittent short-duration reading of good-contrast print |
| Rough machining, general assembly | 300 lx | Large components, coarse tolerances, high contrast |
| Offices, writing, screen-based work | 500 lx | Sustained reading with screen luminance competing |
| Medium machining, fine assembly | 500 lx | Smaller detail, moving parts, metal-on-metal contrast |
| Technical drawing, precision bench work | 750 lx | Fine detail on low-contrast media, sustained |
| Inspection, colour matching, quality control | 1000 lx | Detecting defects and judging colour, both contrast-limited |
The Room Does More Work Than the Fitting
Utilisation factor is the fraction of bare lamp lumens that actually reaches the working plane, and it is decided by the room, not the catalogue. Two properties drive it: the proportions of the space and the reflectance of its surfaces. A luminaire that delivers a utilisation factor of 0.6 in a wide open hall may manage 0.35 in a narrow bay with the same ceiling height.
Room index captures the proportions. K equals length times width, divided by the mounting height above the working plane multiplied by the sum of length and width. A long thin corridor produces a low index and a poor utilisation factor because most of the flux meets a wall on its way down. A square hall produces a high index and keeps its light.
Reflectance is where workshops fall over. Photometric software defaults to something like 0.7 ceiling, 0.5 wall, 0.2 floor. An engineering shop with an exposed dark steel deck, machinery in grey enamel and an oil-stained slab is nowhere near those figures, and the same fittings will deliver visibly less. Paint on the deck and the upper walls buys lux more cheaply per square metre than additional luminaires do, and it improves uniformity at the same time.
The lumen method assembles all of it: number of luminaires equals target illuminance times area, divided by the product of luminaire output, utilisation factor and maintenance factor. Round the answer up, then go back and check what rounding did to the spacing, because a count that satisfies the average can land the fittings in a rhythm the room cannot accept.
The method this section assembles, in one pass: the room, the luminaires' height above the working plane, the target, the luminaire's output and the two factors give the count; the spacing-to-height ratio then sets the grid, and the page reports the average illuminance that grid actually gives — the forward check on any count out of a manufacturer's selector.
The longer dimension of the room.
The shorter dimension.
The height the luminaires hang or sit at, measured from the floor.
The surface the light is designed for — a desk, a bench, or the floor itself.
The maintained average the task needs, from BS EN 12464-1 or the IES recommendation.
The light output of one luminaire, from its data sheet — the luminaire's lumens, not the bare lamp's.
The share of the lumens that reaches the working plane, read from the luminaire's table at this room index and these surface reflectances.
The share of the initial output left at the end of the maintenance cycle, for dirt and lumen depreciation.
From the luminaire's photometric data: the widest pitch, as a multiple of the height above the working plane, that keeps the light even.
Luminaires needed
52 luminaires
The lumen method's count is rounded up; the grid is the fewest luminaires in rows that keep every pitch within the spacing-to-height limit, and the outside rows sit half a pitch from the walls.
- Height of the luminaires above the working plane
- 16.5 ft
- Room index K
- 2.39
- Luminaires by the lumen method, before rounding up
- 51.65
- Rows along the length
- 13
- Luminaires in each row, across the width
- 4
- Luminaires in that grid
- 52
- Pitch along the length
- 7.54 ft
- Pitch across the width
- 16.5 ft
- Largest pitch the SHR allows
- 20.62 ft
- Average illuminance the grid gives
- 46.31 fc
They open the calculator with your figures already in it
Lumen Method Lighting Calculator (Room Index, Luminaire Count and Spacing): 52 luminaires — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Estimated cost — your price
This site holds no price list for this material — local prices vary too much to publish honestly. Enter your supplier's price and the result is costed with it.
What this calculation does not cover
- The utilisation factor comes from the luminaire's own table at this room index and the room's reflectances; the default is a placeholder, and a dark room can halve it.
- The lumen method gives an average. Uniformity, glare and the light on walls and ceilings need a point-by-point calculation in lighting software, which is also where obstructions and the real ceiling grid come in.
- The grid is the arithmetic ideal; ceiling tiles, joists, sprinkler heads and ducts move luminaires on site, and a luminaire moved has left the layout the calculation described.
Mounting Height Is Measured to the Bench
Mounting height in every one of these formulas means the height of the luminaire above the working plane, and the working plane is the bench top. In a six metre shop with nine hundred millimetre benches the figure is 5.1 metres. Using 6.0 inflates the room index, flatters the utilisation factor and under-lights the room by a margin nobody can find afterwards without redoing the calculation.
Spacing comes out of the photometric file as a spacing-to-height ratio, generally somewhere around 1.25 to 1.5 for a direct distribution and higher for a wide-beam batten. Multiply it by the mounting height and you have the maximum pitch. Exceed it and the floor scallops: bright pools under the fittings, measurable dips between them, and a uniformity result that fails while the average passes comfortably.
At the perimeter the pitch halves. A row of luminaires set a full spacing off the wall starves the last two metres of the room, which is exactly where benches, racking and workstations end up. Where the wall itself is a working surface — a parts store, a paint mixing bench — pull the row in tighter still and accept the extra fitting.
Then reconcile the arithmetic with the ceiling you actually have. In a modular grid the computed pitch has to resolve onto whole tiles. In plasterboard it has to miss joists, sprinkler drops, smoke detectors, duct branches and the access panel somebody added at coordination stage. Settle that on the model, because a luminaire moved four hundred millimetres on site to clear a hanger has left the layout the uniformity calculation described.
A spacing in metres has to become a whole number of ceiling modules, and the tile grid is the thing that converts one into the other — so the tile layout and the luminaire layout are a single drawing with two legends.
SettingsSettings for this calculation
Waste is set to 5% by hand. Pick a tier above to replace it, or keep your own figure.
The length of the room.
The width of the room.
The standard suspended ceiling tile sizes.
For a 2x4 panel, which of the two room dimensions the 4 ft side runs along.
Spare panels for the border row cut to fit and for panels damaged in handling.
Estimated drop ceiling tile needed
40 tiles
- Ceiling area
- 149.5 sq ft
- Whole panels along the length
- 6
- Border at each end of the length
- 0.5 ft
- Whole panels across the width
- 5
- Border at each side of the width
- 0.75 ft
- Lengthways border with one panel taken out
- 1.5 ft
- Widthways border with one panel taken out
- 1.75 ft
They open the calculator with your figures already in it
Drop Ceiling Tile Calculator: 40 tiles — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Estimated cost — your price
This site holds no price list for this material — local prices vary too much to publish honestly. Enter your supplier's price and the result is costed with it.
What this calculation does not cover
- The answer is a panel count and nothing else: main runners, cross tees, perimeter wall angle and hanger wire never enter the arithmetic, which divides ceiling area by one tile's nominal coverage and so never sets out a grid or fixes where the border row falls.
- The allowance is multiplied onto the area rather than onto the tiles that actually get cut, so at the same setting it runs generous on a room whose sides land near a whole number of modules and thin on one that needs a part tile along all four walls.
- Both options are costed at their imperial nominal — 4 sq ft for the 2x2 and 8 sq ft for the 2x4 — so a true 600 mm (24 in) metric panel, covering 0.36 m² against the 0.372 m² assumed, is credited with roughly 3 per cent more ceiling than it lays and eats a large part of the waste margin before a single tile is cut.
- The two borders are a CENTRED grid on a true rectangle, and both of those are assumptions. A room out of square gives a border that tapers along its own wall, which no single figure describes; and the grid is routinely pulled off centre anyway to line the tees up with a window mullion, a partition head or a run of light fittings, which moves the whole border from one side to the other.
- The wider border is the one a fixer usually lays, and it costs something the narrow one does not. Two borders under half a panel come out of a single panel — cut it once and the offcut finishes the opposite wall — whereas two borders over half a panel need a panel each. Choosing the wider set-out therefore doubles the border panels, and the allowance above is a flat percentage of area that knows nothing about which one you picked.
- Length times width treats the ceiling as one unbroken rectangle, with nothing added or deducted for the items that occupy tile positions: a recessed troffer or an air diffuser can take out a whole module, while sprinkler drops, speakers and small access hatches are cut into a tile that still has to be bought.
- Only a single rectangle with each side up to 30 m (98 ft) can be entered, so an L-shaped room, a bulkhead or a dropped soffit has to be run as separate rectangles and the counts added, since one figure for the overall envelope misses the extra cut tiles every internal edge creates.
- Tiles come back as loose units, while they are sold in sealed cartons of a fixed panel count, so the order is rounded up a second time at the merchant and the surplus that reaches site is whatever the carton size forces rather than the allowance set here.
Uniformity Is the Number That Gets Rejected
Uniformity is minimum illuminance divided by average illuminance across the task area, and EN 12464-1 asks for 0.6 over most task areas with 0.4 across the immediate surround. It is the criterion schemes fail on, because average illuminance conceals everything about where the light went.
A scheme built around six high bays in a shed can hit its average with pools under the fittings and shadow in between, and the spreadsheet will report a pass. Only a point grid shows it. Ask for the calculation at the spacing the standard specifies, look at the minimum, and look at where the minimum sits — if it is under a bench, the scheme is wrong regardless of what the average says.
Obstruction is the second cause. A warehouse designed as an empty box and then racked out to eight metres is not the room that was calculated. Light has to run along the aisles, luminaires have to sit over the aisles instead of over the racks, and the calculation has to be rerun with the racking modelled as solid. Doing it afterwards is a rewire.
There is a perceptual reason behind the ratio as well as an arithmetic one. The eye adapts to the brightest thing in its field, so a task surrounded by darkness reads as dimmer than a meter says it is. The standard's background provision exists to stop a technically correct task area sitting in a cave.
Glare Arrives From Two Directions
Direct glare is quantified as a Unified Glare Rating, the method set out in CIE 117. Limits in common use run to 16 for technical drawing, 19 for offices and control rooms, 22 for general industrial work and 25 for rough work and circulation. The important property is that a UGR figure belongs to a room, an observer position and a viewing direction — a fitting advertised as UGR under 19 is quoting a standard reference room that is probably not yours.
Modern sources make this harder. A bare LED board behind a clear cover concentrates enormous luminance into a very small emitting area, which is precisely the condition the rating punishes. Micro-prismatic optics, diffusers and deep louvres exist to cut luminance at the high elevation angles the eye occupies, and they cost output. A scheme designed on bare-board efficacy and then fitted with proper optics is short before it is switched on.
Reflected glare is geometry and not brightness. A gloss drawing, a monitor, a polished machine table hands the luminaire straight back into the operator's eye, and adding output makes it worse. The fixes are positional: move the fitting out of the offending zone — roughly the mirror image of the eye about the task surface — turn the workstation, or make the source physically larger and correspondingly dimmer.
Retrofit lamps in old housings deserve their own warning. An LED tube dropped into a prismatic diffuser designed around a fluorescent lamp that emitted through three hundred and sixty degrees now emits through about a hundred and twenty. The top of the diffuser goes dark, the bottom becomes a bright band, and the room measures acceptably while looking wrong to everyone in it.
Colour, and the Flicker Nobody Specifies
Correlated colour temperature is a preference decision with two hard cases. Around 4000 K is the workshop default and reads as neutral under high illuminance; 3000 K suits offices and welfare areas where people consistently prefer it; 5000 K and above belongs to inspection and colour matching. Mixing temperatures within one room is visible, and mixing them along a single ceiling line is obvious to anybody who walks under it.
Colour rendering needs two numbers, not one. Ra of 80 or better covers general work, 90 or better covers anything judged by eye. The value that collapses in cheap 80 Ra product is R9, the saturated red, and that is exactly the sample that matters for timber, paint, meat, skin and wiring insulation. Ask for the full R1 to R15 set on any project where colour is part of the job.
Chromaticity binning is the defect that cannot be repaired later. Two nominally identical fittings drawn from different production batches sit in different bins and read as different colours across a ceiling; ANSI C78.377 defines the tolerance a product is allowed to sit inside. Buy the room in one batch, and buy the spares in the same batch, because a replacement ordered in year four will not match.
Flicker on a machine shop floor is a safety matter and not a comfort one. High-percentage modulation from a low-cost driver produces stroboscopic effects that can make a chuck, a saw blade or a fan appear stationary or slowly rotating. IEEE 1789 covers the modulation limits, and driver flicker performance belongs in the specification alongside efficacy — it is the one photometric property nobody checks at handover and nobody can retrofit without changing every driver in the building.
Daylight Is a Luminaire You Did Not Buy
Rooflights at ten to fifteen percent of floor area transform a shed, and north-light glazing delivers diffuse illumination without the solar gain that comes with a south-facing dome. On a single-storey industrial building the daylight scheme is usually the largest single lighting decision available, and it is made by the architect months before the electrical package is priced.
The well between the roof plane and the ceiling plane is the second optic in the system, and it is routinely wasted. A deep well finished in dark plaster absorbs most of what enters it; the same well in matt white with splayed reveals delivers a wide pool onto the floor. Well depth against well width is the ratio to look at, and the finish specification is worth arguing for because it is the cheapest lumens on the project.
Daylight saves nothing unless the electric scheme responds to it. A photocell-dimmed perimeter row, or a separately switched daylight zone, is the difference between a rooflight that earns its capital cost and one that simply adds a summer cooling load. Commission the photocell setpoints on an overcast day and record them; a sensor set on a bright afternoon holds the row off in conditions where people need it.
Daylight also produces the widest uniformity swings in the building — the same bench at noon in June and at four o'clock in January are two different rooms. Design the electric installation to stand on its own against the maintained target, then let the controls claw the energy back. A scheme that relies on daylight to reach its number fails on the shortest day of the year, which is when the complaint arrives.
The walls of the well are the second optic in any rooflight, and a matt white finish across that surface area puts more light on the floor than a larger glazed panel does.
The interior width of the light well shaft.
The interior length of the light well shaft, perpendicular to the width.
The vertical distance from the roof opening down to the ceiling opening.
Light well wall surface area
117 ft²
This assumes a straight (non-splayed/non-tapered) light well — a splayed well (wider at the ceiling than the roof, common for maximizing daylight spread) needs a per-wall trapezoidal area calculation instead of this simplified straight-shaft method.
They open the calculator with your figures already in it
Skylight Light Well Surface Area Calculator: 117 ft² — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Estimated cost — your price
This site holds no price list for this material — local prices vary too much to publish honestly. Enter your supplier's price and the result is costed with it.
What this calculation does not cover
- Measures the finish, not the daylight. How much light reaches the room is governed by the well index, the shaft's depth set against its width: a 2 m (6.5 ft) shaft over a 600 mm (24 in) opening delivers a fraction of what the same skylight gives through a 200 mm (8 in) ceiling. No amount of reflectance rescues a shaft that is too deep and too narrow — widening the ceiling opening or splaying the walls is what does.
- Counts one side of each wall. The shaft passes through the roof void, so those walls are part of the thermal envelope and need insulation on their outer face plus a continuous air seal where they meet the ceiling plane — the same dimensions, a different material, and none of it in this area. The well that skips it is the one that grows a black line of mold around the ceiling opening every winter.
From a Fixture Count to a Circuit Schedule
The wattage that belongs on a circuit schedule is driver input power, not module or lamp power. A 36 W LED board on a driver running at ninety-two percent draws about 39 W at the terminals, and across a hundred and twenty fittings that difference is an extra way in the board. Take the figure from the driver datasheet at the drive current the luminaire is actually built with, since one board is commonly sold at three current settings.
Lighting is a continuous load by definition — it runs for three hours or more — so the branch circuit is loaded to eighty percent of the protective device rating and the device is chosen at a hundred and twenty-five percent of the load. That much is arithmetic and gets done at design stage.
Inrush is the constraint that gets discovered in the field instead. Every LED driver charges an input capacitor at switch-on, and the peak runs to tens of amperes for a few hundred microseconds. Twenty drivers on one way will trip a 16 A instantaneous-trip breaker every time the lights are switched on, while the running current sits placidly at three amperes. Driver datasheets publish the peak current, its duration and a maximum number of units per breaker type; use that figure, move to a slower curve, or split the way. Contactor and relay contacts take the same abuse and weld shut over a few thousand operations.
Emergency lighting is a separate scheme carrying its own duration, its own supply arrangement and its own test regime. Maintained against non-maintained, escape route against open area classification, three hours in most adopted regimes, and a test facility somebody has to physically operate every month. Design it alongside the general scheme, because the escape route luminaires influence where the general fittings can go.
Fixture count multiplied by driver input watts is the number the protective device sees, and checking it against the continuous-load allowance decides how many ways the lighting board needs.
The total number of light fixtures on this circuit.
The rated wattage of each fixture (or lamp/driver combination).
The nominal voltage supplying the circuit.
The rating of the protective device on this circuit.
Total connected lighting load
720 W
The circuit's current draw is 6.0 A, at or below 16.0 A — 80% of the breaker rating, which NEC 210.19 and 210.20 set for a continuously-loaded circuit. Under BS 7671 the design check is instead Ib ≤ In ≤ Iz — load under device, device under cable capacity — with no 80% derate for lighting, so this comparison is on the cautious side of the British rule rather than a statement of it. Being inside the rating on this one check settles nothing about the rest of the circuit — the conductors, the overcurrent device, and the work as installed are all outside it.
- Circuit current draw
- 6 A
- Maximum continuous load (80% of breaker)
- 16 A
They open the calculator with your figures already in it
Lighting Circuit Connected Load Calculator: 720 W — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Add the equipment this sizes
This result is a specification — 720 W — not a quantity. Put the thing it sizes into your project: how many, what you call it, and your supplier’s price.
What this calculation does not cover
- Watts divided by volts is the current only at unity power factor, and lighting is not at unity. LED drivers and ballasts pull current out of phase with the voltage, so the real figure is watts divided by volts times power factor — at 0.9 that is 11% more current than shown here, and uncorrected budget drivers sit nearer 0.5, which roughly doubles it. The breaker responds to current, not to watts.
- Steady-state load says nothing about the moment of switch-on. Electronic drivers charge their input capacitors in the first milliseconds at many times running current, and enough drivers on one circuit will trip a B-curve MCB or a thermal-magnetic breaker every time the lights are turned on, on a circuit that clears this check with room to spare. That is a driver-count and breaker-curve question, and it is decided nowhere on this page.
Local Light for What the General Scheme Cannot Reach
General lighting is designed for a horizontal plane, and a great deal of skilled work does not happen on one. An operator looks into a vertical face inside a machine guard, under a shelf, into a cabinet, along the underside of a weld. No ceiling scheme reaches those places at any reasonable cost, and raising the general level to compensate is the most expensive way to fail.
Task lighting delivers three to five times the local illuminance from a fitting close to the work, and it is almost always the cheaper answer. Position it outside the veiling reflection zone, shield the source from the operator's eye, and give the operator the ability to aim it. A task light nobody can adjust becomes a task light somebody unplugs.
Low-voltage tape has one characteristic failure: voltage drop along its own length. Five metres of 24 V tape fed from one end is visibly dimmer at the far end than at the near one, and the difference reads as a fault to whoever inspects it. Inject at both ends or at the midpoint, keep runs inside the manufacturer's stated maximum, and size the supply with genuine headroom, because a driver in a closed aluminium profile has nowhere to shed its heat.
Environment finishes the specification. Coolant mist, swarf, wash-down and impact are the real service conditions on a machine shop floor, and IP and IK ratings are not a formality there. Unrated tape in an open profile behind a splash guard lasts one season, and the replacement is done at production standstill.
Tape is sold by the metre and powered by the watt, and the supply has to carry the whole run with margin still in hand once the profile has taken away the driver's ability to cool itself.
The total length of LED strip in the run, including any joined reels.
Check your specific strip's spec sheet — this varies by LED density and brightness.
Extra headroom so the supply isn't run at its rated maximum continuously.
Recommended power supply size
47.5 W (recommended supply rating)
Always confirm against your specific strip's datasheet watts/ft rating — this varies significantly between products.
- Raw strip load
- 39.6 W
They open the calculator with your figures already in it
LED Strip Power Supply Calculator: 47.52 W (recommended supply rating) — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Add the equipment this sizes
This result is a specification — 47.5 W (recommended supply rating) — not a quantity. Put the thing it sizes into your project: how many, what you call it, and your supplier’s price.
What this calculation does not cover
- The result is DC output watts only. It does not check voltage drop, set feed points, or size the low-voltage wiring — a long 12 V run can dim visibly at the far end even with a correctly rated supply, and the cable and connectors have to carry the current, not the watts.
- Watts per foot is treated as one flat figure for the whole run. Nothing accounts for an RGB or RGBW strip drawing more at full white than a single-channel rating suggests, for controllers, amplifiers or receivers powered from the same supply, or for the spread between a datasheet figure and the reel actually in hand.
- This is not a branch-circuit or load calculation. The figure is the driver's output rating: it says nothing about the AC circuit feeding it, the supply's own conversion losses, overcurrent protection, or whether the total exceeds what a Class 2 listed driver is allowed to deliver.
- The safety margin is flat headroom, not a manufacturer derating curve. A supply mounted in a hot, enclosed or unventilated space often has to be derated further than the percentage entered here, and nothing checks the driver's ambient or ingress rating against where it is going.
- Driver type and control method are outside the model. Constant-current strips, mains-dimmable drivers and PWM or 0-10 V dimming each carry their own minimum and maximum load requirements, so a supply sized purely on total watts can still sit below a dimmer's minimum load or be incompatible with it.
Proving It With a Meter
Measure after dark, or with every blind closed, on a grid of points at working plane height. Use a cosine-corrected and colour-corrected cell — a cheap sensor calibrated against an incandescent source reads LED light optimistically. Record the supply voltage at the same time, since a scheme measured at 253 V and a scheme measured at 216 V are not comparable.
Let the installation settle first. LED output falls during the first half hour as the board and driver come up to operating temperature, and a reading taken at switch-on can overstate steady-state output by several percent. Where the fittings are dimmable, confirm the commissioning setting the reading was taken at and write it on the sheet.
Then compare against the right target. The design figure was maintained, and the installation is new, so a correct scheme should read roughly one over the maintenance factor above target — a 0.8 factor means a fresh installation measuring about a quarter high. An installation that measures exactly at the maintained figure on handover day is already short, and it will fall below the standard well inside the cleaning cycle.
Hand over the assumptions along with the results: maintenance factor, cleaning interval, assumed reflectances, module and driver type, drive current setting, and the batch. The next person to replace a failed luminaire will otherwise match the wattage, miss the colour bin, and leave a patch in the ceiling that stays there for the life of the building.
What the lighting schedule has to name
A luminaire count on its own orders nothing. Six figures sit beside it on the schedule, and every one of them moves the count when it changes.
- Maintained illuminance, with the plane and the task area it applies to — Task band, immediate surround and background are three different figures on one floor plate.
- Maintenance factor, with the cleaning interval that justifies it — Lamp lumen maintenance, luminaire dirt, room surface dirt and survival multiplied together; a sawmill and an office are not the same number.
- Mounting height above the working plane — Measured to the bench top and not to the slab; it sets the room index and through it the utilisation factor.
- Spacing-to-height ratio taken from the photometric file — Half pitch at the walls where benches and racking sit; scalloping across the floor is what exceeding it looks like.
- Uniformity and glare limits from the adopted standard — An average passes while a corner fails, and a UGR figure describes a room and a viewing direction, not a fitting.
- Driver input watts and inrush peak per luminaire — Input power decides the circuit; inrush decides how many luminaires share a way and which breaker curve survives switch-on.
Opens the calculators above on one screen with the dimensions from this article already filled in. Quantities only — this site publishes no price list, because local prices vary too much to publish honestly.