Buffers: Seconds or Hours

Two kinds of storage

A buffer stores energy electrostatically, in an ultracapacitor, or electrochemically, in a battery; a few systems carry a little of both. The chemistry sets the bridge time — how long the picture survives without pedalling — and the safety profile of the whole system.

Specifications

What do these labels mean?

Field	Value	Confidence	Sources
Ultracapacitor coulombic efficiency	greater than 99 %Manufacturer datasheet figure for ultracapacitor cells generally, not a measurement of any specific pedal-cinema system.	Documented	¹
Ultracapacitor cycle life	500,000 to over one million duty cyclesManufacturer datasheet figure for the cell class, presented as category context.	Documented	²
Ultracapacitor usable energy	approximately 75 % of stored energy, discharging from rated to half-rated voltageExplains why ultracapacitor bridges run seconds to minutes, not hours.	Documented	³
Rock the Bike Pedal Power Utility Box bridge	2–5 minutes, depending on consumption	Manufacturer rating	⁴
Cinécyclo hybrid buffer	about 40 secondsLes Échos, 2025.	Operator-stated	⁵
Cinécyclo hybrid buffer, alternate figure	55 secondsADRC sector guide, 2025.	Operator-stated	⁶
KLAK central-unit battery chemistry	LiFePO4Capacity in watt-hours is not disclosed.	Documented	⁷
Oekotrainer buffer capacity	12.8 V / 100 Ah / 1,280 Wh (LiFePO4)	Documented	⁸
Rock the Bike Event Power Box capacity	2.5 or 5 kWhOnly the 5 kWh option is published as a capacity; the standard 2.5 kWh is implied by the page's own worked example, not printed. Battery chemistry and bus voltage are not published either.	Manufacturer rating	⁹
KLAK central-unit output	230 V AC / 50 Hz, up to 500 W	Documented	⁷
LiFePO4 thermal-runaway onset	near 270 °CVendor-published comparison, consistent with the technical literature.	Documented	¹⁰
NMC thermal-runaway onset	150–210 °C	Documented	¹⁰
LiFePO4 cycle life	4,000–10,000 cycles, against 1,500–3,000 for NMC	Documented	¹⁰
Karlsruhe school event auxiliary-battery draw	about 30 %12 January 2026; cause stated as bikes not continuously occupied late in the film.	Operator-stated	¹¹
Measured generator efficiency by load point	75–78% above 72 W, 60% at 36 W, below 45% at 24 WLoad-dependent; an oversized generator on a small load idles in its lossy region.	Measured	¹²
Schottky diode rating, Low-tech Magazine build	10 A, against an 8–9 A maximum production current	Documented	¹²

Ultracapacitors: seconds to minutes

Ultracapacitor cells as a class are manufacturer-documented at greater than 99% coulombic efficiency and 500,000 to over one million duty cycles, with charge times measured in seconds ^1,2. Those figures describe the cell technology, not any pedal-cinema system: no operator using ultracapacitors has published capacitance or in-system efficiency figures of its own. The limit is energy, not power. About 75% of stored energy is usable when discharging from rated voltage to half-rated voltage ³, which is why ultracapacitor bridges run seconds to minutes rather than hours. Rock the Bike's Pedal Power Utility Box states "a 2–5 minute buffer of power (depending on consumption levels)" ⁴. Cinécyclo's hybrid kit has carried supercapacitors since 2016 "to reduce battery use to a minimum" and reports about 40 seconds of autonomy in one 2025 account and 55 seconds in another ^5,6.

Comparison card of buffer chemistries with cycle life and thermal thresholds labelled as vendor-published — Chemistry card: ultracapacitor vs LiFePO4 vs NMC — cycle life and thermal-runaway onset (vendor-published)^{5,6,4,7,3,1,2,10} Image: BicycleCinema.org, drawn from the research corpus

LiFePO4 and NMC: hours

Battery-buffered systems trade that speed for hours of ride-through. KLAK's central unit runs on a lithium iron phosphate (LiFePO4) battery whose watt-hour capacity is not disclosed; its output is 230 V AC up to 500 W ⁷. Oekotrainer's rental set specifies its buffer exactly — 12.8 V, 100 Ah, 1,280 Wh, also LiFePO4 ⁸. Rock the Bike's Event Power Box is rated at 2.5 or 5 kWh, with chemistry and bus voltage unpublished ⁹. LiFePO4's documented thermal-runaway onset sits near 270°C, against 150–210°C for lithium nickel manganese cobalt oxide (NMC) cells, and its cycle life is 4,000–10,000 against NMC's 1,500–3,000 ¹⁰. Where the chemistry is published, the documented battery-buffered systems use LiFePO4, not NMC ⁷. Either chemistry still needs a functioning battery management system; a UK coroner's report made that point in fatal terms for an unmanaged e-bike pack ¹³.

Cyclists pedalling during an Efecto Pedal outdoor event — Photo: Efecto Pedal (pantalla-inflable.blog) — via pantalla-inflable.blog

Buffer size and swap cadence

A large buffer lets a screening run on with the bicycles empty; a small one keeps the effort visible but exposes every slow changeover. At Karlsruhe's school screening on 12 January 2026, students rotated every five to fifteen minutes under a live wattage display, and the system still drew about 30% of its auxiliary battery, mainly because the bikes were no longer continuously occupied toward the end of the film ¹¹. The buffer also has to suit the generator's efficient band: one instrumented build reached 75–78% efficiency only above 72 W of output, falling to 60% at 36 W and below 45% at 24 W, so an oversized generator on a small load idles in its lossy region whatever the buffer behind it ¹². Berlin's ufaFabrik removes the buffer altogether and raises the resistance on the remaining bikes when a rider drops out ¹⁴.

Fielded protections and recommendations

Two protections are documented as fielded practice. Reverse-current isolation: Low-tech Magazine's manual warns that "without a diode, the battery could power the generator (and turn the pedals) instead of the other way around," and specifies a Schottky diode rated above the expected current — 10 A against an 8–9 A maximum ¹². Supervision: KLAK's central unit shuts down after ten minutes without pedalling and carries state-of-charge alarms ⁷. Two further measures are engineering recommendations rather than documented pedal-cinema practice: pre-charging or current-limiting an ultracapacitor bank, because a discharged capacitor "appears as a short circuit" to whatever is charging it, and ideal-diode controllers as a lower-loss alternative to a simple diode in a multi-bike array ^15,16. No pedal-cinema hardware documented here has either measure built in; both belong on a design review with a qualified engineer rather than on a parts list.

What to ask

Ask for the bridge time in writing and compare it with the slowest realistic rider swap, not the best case. Ask which chemistry is inside and whether it carries a working battery management system. Ask what happens to the picture when the buffer runs low — a fade, an alarm or a hard cut.