About Fifty Watts: What a Person Can Pedal

The evidence range

Rock the Bike, after more than a decade of pedal-powered stages in San Francisco, puts the effort per person at "approx. 50–75 Watts … the amount that an average audience member can continuously provide" ¹. Electric Pedals' founder Colin Tonks puts it at 40–50 W sustained for an hour ². The KLAK manual rates its roller trainers at up to 250 W peak per bike, averaging 50–60 W per bike over a screening ³. Low-tech Magazine's 2022 household build, measured with volt and ampere meters at the generator output, recorded 45–75 W during a moderate, sustainable effort and roughly 100 W only in a short, heavy burst ⁴. A UK festival-planning guide uses 60 W per adult bicycle, 30 W per child bicycle and 10 W per hand-cycle ⁵.

Specifications

What do these labels mean?

Field	Value	Confidence	Sources
Rock the Bike comfortable continuous output	50–75 W electrical per rider"the amount that an average audience member can continuously provide."	Operator-stated	¹
Electric Pedals (Tonks) average output	40–50 W electrical, sustained for an hour	Operator-stated	²
KLAK roller-trainer rating	up to 250 W peak per bike; 50–60 W average per bike"bis zu 250W (im Schnitt 50–60 Watt pro Rad)."	Documented	³
Low-tech Magazine measured household build	45–75 W sustained; approximately 100 W short heavy effortAnalogue volt/ampere meters at the generator output; non-athlete adults.	Measured	⁴
UK festival planning averages	60 W adult / 30 W child / 10 W hand-cycle, per bikeIndustry planning guide, not an instrumented measurement.	Documented	⁵
Coggan untrained-male functional threshold power	2.04 W/kg, about 143 W at 70 kg body mass (maximal one-hour effort)	Documented	⁶
Appropedia ergometer figures	37 W indefinite; 71 W for 60–120 s; 110 W for 10 min at 80 rpmMaximal-effort laboratory tests, not voluntary event pace.	Documented	⁷
Published power-meter population medians	3.8–4.95 W/kg (20-min and FTP medians)Drawn from a self-selected, trained, power-meter-owning population; would overestimate public-event capability by a factor of roughly two to four.	Uncertain	^8,9
Gross metabolic efficiency range	17–25 %	Documented	¹⁰
Zoladz et al. measured gross efficiency	21.4 ± 1.1 % at a 120 W load17 untrained young males; peer-reviewed.	Measured	¹¹
Metabolic cost of 100 W mechanical output	400–500 W total (gross convention) or about 330 W additional (delta convention, ~30% efficiency)Roughly 3–4 W metabolised per watt delivered. Heat shed is 300–400 W under the gross convention, about 230 W under the delta convention.	Documented	^11,12
Resting-metabolism share at a 50 W workload	about 30 % of exercise oxygen uptakeDepresses gross efficiency at the low intensities typical of cinema pedalling.	Documented	¹³
Attention penalty, typing-while-pedalling study	38 ± 14 Wn = 10, recumbent ergometer, 2014.	Measured	¹⁴
Karlsruhe school event buffer draw	about 30 % of auxiliary battery consumedDespite live wattage feedback; attributed to bikes not staying continuously occupied late in the film.	Operator-stated	¹⁵
Logged single-rider session (WattsVIEW)	65–80 W average, 150–175 W peak, about 83 minutes	Measured	¹⁶
VéloKino Zürich per-bike average	50 W per bike (0.5 kWh/hour from 10 pedalers)	Operator-stated	¹⁷
Ludwigsburg schoolchildren event	400–500 W total from 10 bikes	Operator-stated	¹⁸
BBC Human Power Station	up to 14 kW peak from 80 experienced cyclists (about 175 W each, peak, not sustained)Thirty of those cyclists were needed just to boil a kettle.	Documented	¹⁹
Chained AC loss budget (Low-tech Magazine, 2011)	generator 10–20%, regulator ~25%, battery 10–35%, inverter 5–15%; worst-case net about 32.5 % (30–60% of pedal power lost across a typical chain)A literature-sourced cascade, not an instrumented measurement of any one system.	Documented	¹⁹
Measured generator efficiency by load point (Low-tech Magazine, 2022)	75–78% above 72 W output, 60% at 36 W, below 45 % at 24 WInstrumented on the magazine's own build; the curve behind Worked Example B.	Measured	⁴
Optoma ML750e-class projector load, for the single-rider case	45 minimum / 65 W maximum (Bright mode)Optoma's own rating; the load a single rider has to reach.	Manufacturer rating	²⁰
Worked Example A: AC-chain sizing	250 W load ÷ 0.35 efficiency ≈ 714 W at the pedals ≈ 14 riders at 50 W, or ≈ 10 riders at 75 WThis report's arithmetic; load figure is the CycleCinemaClub beamer-plus-speakers class.	Estimated	^19,21
Worked Example B: single-rider direct-DC sizing	pedals-to-DC net roughly 60–65%; at least 150 W pedal input needed per 100 W electrical outputDerived from Low-tech Magazine's measured generator-efficiency curve; a 60 W rider nets only about 40 W DC.	Estimated	⁴
Promotional claim: New Atlas Pedal-A-Watt	100 to 320 W claimed per average adult riderResearch correction:Contradicted by every operator and measured figure in this collection; the defensible constant is 50–60 W electrical.	Promotional / press-reported — not verified	²²
Promotional claim: Chris Joseph NRG	100 to 250 W claimed per average personResearch correction:Contradicted by every operator and measured figure in this collection; the defensible constant is 50–60 W electrical.	Promotional / press-reported — not verified	²³
Promotional claim: BikeCine (Brazil)	about 200 W per cyclist; roughly 2,500 W total across 12–16 stationsResearch correction:Roughly four times the 50–60 W constant documented by KLAK, Vienna's CycleCinemaClub and Electric Pedals; best read as a mechanical, short-burst figure rather than sustained electrical output.Press coverage, not independently metered.	Promotional / press-reported — not verified	²⁴

Why laboratory figures are higher

Sports-science figures measure something else. Andrew Coggan's power-profile table classes an untrained male's maximal one-hour output at 2.04 W/kg — about 143 W for a 70 kg rider ⁶. Ergometer tests compiled by Appropedia found untrained adult men capable of 110 W for ten minutes and 71 W for one to two minutes, with 37 W sustainable indefinitely ⁷. These are maximal-effort protocols, not the pace a volunteer chooses while watching a film. Published power-meter data — a 3.8 W/kg twenty-minute median and a 4.95 W/kg median threshold power among riders under thirty — comes from a self-selected population of trained, power-meter-owning cyclists and would overestimate a general audience by a factor of roughly two to four ^8,9.

Graphic showing the rider-count sizing formula with a fourteen-rider AC example and a single-rider DC example — Sizing arithmetic graphic: riders ≈ load ÷ (efficiency × 50 W) with worked examples A and B^{1,4,2,7,6,3,18,5,19,16,14,17,11,25} Image: BicycleCinema.org, drawn from the research corpus

The metabolic cost

Gross efficiency — mechanical work out divided by total energy burned — ranges from 17% to 25%; one peer-reviewed study measured 21.4 ± 1.1% at a 120 W load in seventeen untrained young men ¹¹. Delivering 100 W of mechanical output costs the body 400–500 W of total metabolic rate under the standard convention, 300–400 W of it shed as heat — three to four watts burned for every watt that reaches the generator ^11,12. At a 50 W workload, resting metabolism alone accounts for about 30% of the oxygen the body uses ¹³.

Velocity fundraising pedal-powered screening by Pedalcents — Photo: Pedalcents (pedalcents.ie) — via pedalcents.ie

Attention and rotation

Two factors pull real audiences toward the bottom of the 40–75 W band. A 2014 study measured people typing while pedalling a recumbent ergometer at 38 ± 14 W, and an audience watching a film is the same attention-elsewhere case ¹⁴. At a school screening in Karlsruhe on 12 January 2026, the system drew down about 30% of its buffer battery despite a live on-screen wattage display, mainly because the bikes were no longer continuously occupied toward the end of the film ¹⁵. The KLAK manual's answer is procedural: ten simultaneous riders drawn from a group of at least thirty, rotating every few minutes ³.

Logged events

One belt-drive build logged 65–80 W average and 150–175 W peak over an 83-minute session ¹⁶. Zurich's VéloKino averages 50 W per bike, about 0.5 kWh an hour from ten pedallers ¹⁷. A schoolchildren's event in Ludwigsburg drew 400–500 W total from ten bikes ¹⁸. For scale, the BBC's Human Power Station reached 14 kW from eighty experienced club cyclists pedalling flat out — about 175 W each, a peak, not a sustained figure ¹⁹.

The planning figure and two worked examples

The planning figure is about 50 W of electrical output per casual adult rider, and riders needed ≈ load ÷ (chain efficiency × watts per rider). For a battery-buffered AC system, Low-tech Magazine's loss budget — generator 10–20%, regulator about 25%, battery 10–35%, inverter 5–15% — compounds to a worst-case chain efficiency of about 32.5% ¹⁹. Worked Example A: a 250 W AC load, the class Vienna's CycleCinemaClub runs, divided by 0.35 efficiency needs about 714 W at the pedals — roughly 14 riders at 50 W each, or about 10 at 75 W ^19,21. KLAK's deployed design matches: a 340 W projector plus sound from ten riders at 50–60 W each against a 500 W inverter ceiling, with a LiFePO4 buffer absorbing the shortfall ³. Worked Example B is a single rider on direct DC. Removing the battery and inverter stages and keeping the generator in its efficient band — a measured 75–78% above 72 W, falling to 60% at 36 W and below 45% at 24 W — gives a derived pedals-to-DC net of roughly 60–65%, so at least 150 W of pedal input is needed for every 100 W that reaches the load ⁴. A rider sustaining 60 W mechanical therefore delivers about 40 W DC against an Optoma ML750e-class projector rated 45–65 W, which is why the Cinebike build buffers a single rider through two batteries rather than feeding the projector live ^4,26,20. One lever is free: Low-tech Magazine's output doubled after cleaning and oiling the bicycle's drivetrain ⁴.

Promotional claims

Three promotional claims diverge from this range. A vendor cites "100 to 320 watts" per average adult rider ²²; another cites "100 to 250 watts" per average person ²³; and Brazilian operator BikeCine is reported in the press at about 200 W per cyclist ²⁴. Each contradicts every operator and measured figure above, and two category errors explain them: quoting mechanical power at the pedals rather than electrical power after 30–60% of chain losses, and quoting a short maximal burst rather than a multi-hour event pace. A proposal whose arithmetic assumes 100 W sustained per casual rider is undersized by roughly a factor of two.