How Many Lumens Does a Wall Sconce Actually Put Out?

Fewer than the box promises, and nobody publishes the figure, us included. The lumens wall sconce listings quote belong to the bulb, not to the fitting that bulb sits inside, and those are two different measurements taken by two different procedures. The fluted up down sconce has two G9 sockets with a 5W LED recommended in each, bulbs not included, and a light coverage line of 5-10 square meters. No lumen output appears on it, because we have not had the fitting measured and will not print a number nobody produced.

Two separate things happen to a bulb's light once it is screwed into a shade. Some is absorbed inside the fitting and never gets out. Some gets out perfectly well but leaves in a direction that was never going to land on your book. DOE put a figure on the first for a domestic case: an 860 lumen lamp inside a residential six-inch downlight measured 514 lumens leaving the fixture, a fixture efficiency of 60 percent.

Are the lumens wall sconce listings quote the bulb's or the fixture's?

The bulb's, and in the United States that is written into the rule that governs the label. 16 CFR 305.23 requires a lamp package to disclose "the light output of each lamp included in the package, expressed as 'Brightness' in average initial lumens rounded to the nearest five", and again at (b)(7)(i), "the lamp's average initial lumens". Every Brightness figure on a Lighting Facts panel describes a bare lamp in a laboratory, never a lamp inside a shade.

DOE draws the same line in plain words. Its LED Measurement Series fact sheet defines fixture efficiency as "the proportion of rated lamp lumens actually emitted by the fixture", and states that "for all light sources, there is a difference between rated luminous flux of the lamp and actual performance in a luminaire". A luminaire, in that document, is the complete unit: lamp, housing and optics.

So two honest numbers exist for any wall light, and a retailer can only hand you the second if somebody paid a photometric laboratory for it. Decorative brands rarely do. We have not.

How much of the bulb's light does the fitting keep for itself?

Between a third and roughly two thirds of it, on the only fixtures anyone has published numbers for. DOE's CALiPER programme tested three recessed wallwashers running the same 32 W triple-tube CFL and measured both the lamp on its own and the lumens that actually escaped the fitting.

The report those first three rows come from puts the general case bluntly: "a significant portion (50% or more) of the lumens emitted by the lamp can be trapped in the luminaire". When the same programme needed in-fixture figures for a troffer study and had only bare-lamp data, it applied a flat conversion: "values listed by LED Lighting Facts are for bare lamps. They have been multiplied by 0.70 to account for luminaire efficiency." A 0.70 planning factor is DOE's own rough answer, and losing 30 percent of a bulb's lumens is enough to change which bulb you buy.

Now the part that gets faked everywhere else. No published measurement exists for what a decorative sconce shade blocks. DOE has no report on wall sconces at all, and no standard converts "opaque shade" into a percentage. A page quoting you a sconce-specific number invented it. The rows above are the honest neighbours: real fittings, independent laboratory, wrong shape.

Does an up and down sconce fire half its light at the ceiling?

Roughly, yes, and that is the design working rather than failing. The Houseland fluted sconce is open at the top and the bottom, the spec line word for word, so a large share of the light leaves upward and reaches your book only after bouncing. CIE's international vocabulary sorts every luminaire by exactly that share.

The IES files this shape under direct-indirect lighting, "a variant of general diffuse lighting in which the luminaires emit little or no light at angles near the horizontal" — a precise way of saying that a fluted half-cylinder throws up and down and almost nothing sideways. Whatever heads for the ceiling comes back only in proportion to what the ceiling reflects, so the same fitting behaves like two different fittings in a white room and a charcoal one.

That deserves a figure too. WELL's surface design feature asks for ceiling reflectance of 80% or more across at least 80% of the area in working spaces, allows 40% or more in bedrooms, and caps residential living-room walls at 60% or less. A matt white ceiling gives most of the upward half back; a dark one keeps it, which is why a sconce that looked generous in the showroom disappoints in a moody hallway.

From the shop

Fluted Resin Hallway Sconce, Up Down Wall Light in 4 Colors, Two G9 Bulbs

The fluted sconce stands 6.5 in (16.5 cm) tall and projects 3.1 in (8 cm) from the wall, with two G9 sockets, no bulbs in the box and a 5W LED recommended per socket. Light coverage is listed as 5-10 square meters and no lumen output is published.

View the fluted up-down sconce — 4 colours & current price

How do you choose the bulb when no fixture figure exists?

Start at the bulb label, cut it down twice, then verify with a meter. Only the last of these five steps describes your actual room.

1. Read the Brightness line on the bulb's Lighting Facts panel. By regulation that is the lamp's average initial lumens, and it is the one figure in this whole transaction that somebody genuinely measured.
2. Add the sockets rather than averaging them. The Houseland fluted sconce carries two G9 sockets and recommends a 5W LED capsule in each, so two rated lumen figures add together before the fitting takes its cut.
3. Take 30 percent off as a planning allowance. That is DOE's 0.70, borrowed from troffers, so treat it as a class estimate and never as a specification for this fitting.
4. Split what is left between up and down. An open-top-and-bottom shade is direct-indirect by the IES definition, so a large share is going ceilingward and returns reduced by whatever your paint reflects.
5. Meter the surface, not the fixture. A lux meter, or a phone app you have sanity-checked against one, laid flat on the open page or the hall floor, produces the only number that is about your room.

Do that once and the strategy picks itself. A capsule costs a few dollars; the fitting behind it is wired into the wall. Buy the bulb you think is right, live with it for one evening, and swap it if the page is grey.

What brightness should a hallway or a bedside actually land on?

Aim the surface at roughly 300 to 500 lux and stop worrying about lumens, because lux is measurable and fixture lumens are not. Steve Fotios redrew the old Kruithof graph in LEUKOS in 2017 from the studies that survive scrutiny: of nine credible replications, five do not support Kruithof, one gives partial support and three are unclear. What survives is illuminance on its own. Below about 300 lux an interior tends to be rated unpleasant, 500 lux is enough, and pushing past 500 lux adds little.

Colour temperature between about 2500 K and 6500 K has a negligible effect on how bright or how pleasant people rate a room, so a warm bulb does not have to be a dim one. Boyce and Cuttle, quoted in that review, found that once a person is fully adapted, "the CCT of good colour rendering lamps in the range 2700K to 6300K has little effect on people's impressions of the lighting of the room". Pick 2700 K because you like it, then buy the lumens separately.

The light coverage line, 5-10 square meters, is a room-size suggestion, not a photometric measurement. It says the fitting is scaled for a corridor, a landing or a bedside wall rather than an open-plan living room, and says nothing about lux on any surface. Read it the way you read a rug size. If the job is reading in bed specifically, position matters more than output, and I worked through the height and angle for a bedside wall light separately. The rest of the wall lamps carry the same coverage convention.

Who is allowed to wire the thing in?

Depends entirely on your country, and two English-speaking markets answer it differently. The Houseland sconce is hardwired and the shade clips over a mounting bracket, so there is no plug in this job.

• New South Wales treats it as licensed work. The state says you need an electrical licence or certificate before you do any electrical wiring work, and states fines of $22,000 for an individual and $110,000 for a company for doing it unlicensed.
• England splits it on location. Approved Document P puts additions and alterations to existing installations outside special locations in the not-notifiable column, so a wall light added to an existing lighting circuit in a hallway does not have to be notified. A room containing a bath or shower is a special location, where all of it is notifiable.
• Not notifiable is not the same as unregulated. Paragraph 1.1 of the same document still requires the installation to be designed and installed in accordance with BS 7671, and non-notifiable work still has to be inspected and tested to those procedures.

If you have never taken down a ceiling rose, this is a job for an electrician wherever you live. I wrote out what a hardwired install actually involves, bracket by bracket, in the step-by-step for a kids' room sconce. And if the fixture you are weighing up runs on a cord or a battery instead, the battery versus plug-in comparison covers what you give up in output.

What I'd do

Fit the brightest G9 capsule that stays inside the 5W LED recommended per socket, put both in, and judge the room after a full evening rather than in the first ten seconds. Ten seconds is your eyes adapting, not the light.

Then measure. Lay a meter on the page or the floor, look for something in that 300 to 500 lux band, and if you are short, change the bulb rather than the fixture. A second capsule costs less than an hour of an electrician's time, and a hardwired fitting that comes off the wall again is a bad afternoon. The one number I will not give you is a lumen output for this sconce, because nobody has measured it and a figure invented to fill the gap is worse than the gap.