Headlamp Runtime, and Why You Carry Three Light Sources

Total darkness underground doesn't dim gradually the way dusk does; it's either lit or it's total, with nothing in between and no ambient light to fall back on while your eyes adjust. That single fact is why caving lighting culture is stricter than almost any other outdoor activity's, and why the standard practice among experienced cavers is to carry three independent light sources on every trip, not one good one. This article works through the actual arithmetic behind a headlamp's real runtime — using the site's own Headlamp Runtime Calculator — for a representative primary light, a backup, and a minimal third light, so "three lights" stops being a rule of thumb and becomes a set of numbers you can check against your own kit.

The formula behind the calculator

Runtime comes down to a straightforward energy budget. A battery's usable energy, in watt-hours, is its capacity in amp-hours multiplied by its voltage. That energy doesn't all reach the LED, though: a driver-efficiency factor accounts for real-world losses in the electronics that regulate the light's output, and, for caving specifically, a cold-derate factor accounts for the usable capacity a battery loses simply from running colder than the room-temperature conditions most manufacturer specs are measured in. Multiply the two loss factors together, apply that to the total energy, and divide by the lamp's power draw at your chosen brightness, and the result is a genuinely real-world runtime figure rather than a best-case spec-sheet number.

Worked example: a primary light

Take a representative dedicated caving headlamp: a 3,400mAh lithium-ion pack at 3.7 volts, drawing 1.6 watts at a general-use brightness setting, with an 85% driver-efficiency factor and a 10% cold derate for typical cave temperatures. Run those figures through headlampRuntime(): the pack holds 3.4 × 3.7 = 12.58 watt-hours of energy. Applying the 85% driver efficiency and the 10% cold derate together (0.85 × 0.90 = 0.765 effective factor) gives 9.62 effective watt-hours. Divide that by the 1.6-watt draw and the result is a runtime of just over 6 hours — 6.01 hours, to the tool's own precision — on a single charge, at that brightness, in cold conditions. That's the number worth planning around, not whatever larger figure came printed on the box.

Worked example: a backup light

A sensible backup light, built around disposable AA-type cells specifically so it doesn't depend on remembering to charge anything, might run at roughly 2,500mAh and 4.5 volts (three cells in series), drawing a more modest 1.0 watt, with a slightly lower 80% driver efficiency and the same 10% cold derate. That works out to 11.25 watt-hours of energy, 8.1 effective watt-hours after losses, and a runtime of 8.1 hours — genuinely longer than the primary, because it's drawing less power for a dimmer, but entirely adequate, output. A backup doesn't need to match the primary's brightness; it needs to reliably get you out.

Worked example: the third, minimal light

The third light in the kit is deliberately not a scaled-down headlamp; it's a genuinely minimal device — a small keychain LED, say, on a coin-cell-class battery around 220mAh at 3 volts, drawing a tiny 0.06 watts, with a 90% driver efficiency and a light 5% cold derate given how little heat-sensitive electronics it has. Run that through the same formula: 0.66 watt-hours of energy, 0.56 effective watt-hours, and a runtime of 9.41 hours — longer, in this case, than either of the other two lights, precisely because it draws so little power to begin with. It won't light a large chamber, but its entire job is different: enough light to move safely and calmly to easier ground if both other lights are somehow gone, not to finish the planned route.

Lay the three side by side and the logic of "three lights, not just one very good one" becomes concrete rather than aspirational: three independent failure points, three independent batteries, and — worked honestly through the numbers — three genuinely different runtime and brightness trade-offs, each suited to a different job in the same emergency.

Why cold costs you real capacity, not just a vague "less"

The cold-derate factor in the formula isn't a rule-of-thumb fudge; it reflects real battery chemistry. Most rechargeable cells rely on ions moving through a liquid or gel electrolyte to deliver current, and that movement slows measurably as temperature drops, which raises the cell's effective internal resistance and reduces how much of its rated capacity is actually usable at a given discharge rate. Caves run cooler than most surface trips, often close to a chilly, unchanging single-digit or low double-digit temperature year-round, which is exactly the condition manufacturer runtime claims — usually measured at a comfortable room temperature — don't reflect. A 10% derate is a reasonable planning figure for many lithium-ion packs in typical cave conditions; colder systems, or older, more degraded cells, can lose meaningfully more than that, which is one more reason to treat a spec-sheet runtime as an upper bound, not a promise.

What carrying multiple sets actually buys you

The calculator also accepts a battery-count input, for exactly the "how many spare sets am I carrying" question rather than "how long does one set last." Set the primary lamp's battery count to 3 instead of 1, with everything else identical, and the tool reports a total runtime of just over 18 hours across all three sets combined — effectively three separate 6.01-hour runs stacked end to end, since each fresh set starts its own discharge curve from full. That's not a new calculation, just the same per-set figure multiplied by how many sets you're actually carrying, but seeing it as a single combined number is useful for comparing directly against a trip's planned length plus margin, the way the 18-hour target for a 9-hour trip does above.

Brightness mode is the other lever, and it's a direct trade

Power draw and runtime move in direct, inverse proportion in this formula — halve the wattage and you double the runtime, all else equal — which is why choosing a lower brightness mode for general movement is such an effective way to stretch a light's usable life on a long trip. Drop that same primary lamp's draw from 1.6 watts to 0.8 watts, for instance by stepping down from a high to a medium brightness mode, and the identical energy budget now supports roughly 12 hours instead of 6, more than doubling the day's available light from the exact same battery. The practical habit that follows: reserve full brightness for when you genuinely need to see far or wide — a large chamber, a route across open space — and default to a lower, still perfectly usable mode for the ordinary business of moving through an average passage.

From one set's runtime to how many spares a trip needs

A single figure like "6.01 hours" only becomes a packing decision once it's compared against how long the trip is actually expected to run — and, per the planning habit covered elsewhere on this site, trips run long far more often than they run short. A common, reasonably conservative planning approach is to carry enough total runtime to cover roughly double the planned trip length, on the logic that the margin needs to absorb both a longer-than-planned trip and the possibility that a light degrades or fails outright. For a planned 9-hour trip, that's an 18-hour target: divide by the primary light's 6.01-hour runtime per set and round up, and the arithmetic calls for 3 sets of batteries for that one light alone, before the backup and third light are even considered. The same doubling logic, worked across a full range of trip lengths and paired with layering and gear-list figures too, is laid out in the site's Trip Planning Reference.

Verifying a runtime estimate against reality

A calculated runtime is only a starting estimate, and it's worth actually checking it against your own lamp at least once rather than trusting the arithmetic blindly forever. A simple real-world check: charge the lamp fully, note the time, run it continuously at the brightness mode you actually plan to use, somewhere reasonably representative of cave temperature if you can manage it, and note when it drops to genuinely unusable brightness. Compare that measured figure to what the calculator predicted for the same inputs. A significant gap in either direction is useful information in itself — a battery that's aged more than expected will under-perform its rated capacity, while a well-matched driver and fresh cell can sometimes slightly outperform a conservative estimate. Either way, a single real test converts a theoretical number into a figure you actually trust on a real trip, and it's worth repeating occasionally as a battery ages, since usable capacity genuinely degrades over a cell's lifetime even with good care.

Checking your own kit against these numbers

The specific figures above describe representative lights, not necessarily yours — the entire point of running them through the calculator rather than quoting a fixed table is that your own lamp's capacity, voltage, and actual power draw at the brightness you use most will differ, sometimes considerably. Take those three numbers off your own lamp's spec sheet (or measure the draw directly, where you can), apply a cold derate appropriate to the systems you actually visit, and run it through the Headlamp Runtime Calculator yourself. The output is only as good as the inputs, but the arithmetic behind it is exactly what's been worked through here, and it takes less time than reading this paragraph to redo for a different lamp, a different brightness mode, or a colder system than the one you last checked it against.

The takeaway

Three lights isn't caution for its own sake; it's three independent, genuinely different tools, each with its own real runtime once cold and driver losses are accounted for honestly rather than trusted from a box. Know your own primary, backup, and third light's real numbers, carry enough spare capacity for a trip that runs longer than planned, and treat the manufacturer's spec sheet as a starting point for your own arithmetic, not the final answer.