ElectrifyBoating: methodology

01 · WhyWhy this database has to exist

Everyone selling — or buying — an electric boat hears the same question first: how far does it really go? A spec sheet can't answer it. And no one can measure every possible boat, motor and battery combination one by one. Range depends on speed, load, hull form and propeller in ways no single brochure number captures, and the numbers that do get printed tend to be best-case. So buyers hesitate, and the people advising them have little independent data to point at.

We think the fix is unglamorous: measure real boats, on real water, under real loads — and publish the results openly, with every number labeled by how it was obtained. That database doesn't exist anywhere yet. We're building it, starting with boats from people who work with them every day.

02 · MeasuredThe measured run

The heart of the site is a test any owner can do in about twenty minutes, with nothing but the drive display and a GPS:

1. 1

   Set the throttle until the drive display reads 10% of rated motor power, and note the GPS speed.

2. 2

   Repeat at 25%, 50%, 75% and 100% — five points in all.

3. 3

   Run each leg twice, in opposite directions, and average the two speeds.

03 · ReportedThe reported path — real numbers without a test day

Not every boat will get a full measured run, and a database that insists on perfection stays small. So a second submission path is open: reported. An owner types in two numbers they already know — the boat's top speed, and its speed at 50% power, both read straight off the drive display.

Those two real anchor points pin a performance curve whose shape comes from measured boats of the same hull class in the fleet. The cruise speed and range we derive from that curve are labeled inferred — never presented as measured — and a reported boat carries a permanently lower data-quality ceiling than a measured one. Real numbers, honestly ranked below a real test: measured tells you what a boat does across its whole speed band; reported anchors the fleet's physics with two true points from that boat.

04 · OutputsWhat we calculate from a run

Five speed-and-power points become a complete performance picture:

05 · MethodHow the numbers are made

Not the full engineering story — but enough to judge whether the method is sound:

• From five points to one curve. We fit a single smooth power-versus-speed curve to the measured points. Its steepness is itself diagnostic: a displacement hull's power demand climbs gently at low speed, then rises sharply as the boat approaches its theoretical hull speed — the wave-making wall every displacement sailor knows. That wall sits at a fixed speed-to-length ratio, what naval architects call the Froude number — the yardstick that makes an 8-metre and a 12-metre hull comparable, and the reason a longer waterline cruises faster. Cruise, range and the most efficient speed are read off this fitted curve, so they aren't limited to the five test throttles themselves.
• Energy and range. Watt-hours per nautical mile is the power draw divided by the speed at that point — the boat's fuel economy, in electric units. Range is the battery's usable capacity divided by that economy, and "usable" respects chemistry: we count 95% of a LiFePO4 bank, 80% of NMC, and only 50% of lead-acid.
• Hull resistance and true efficiency. From length, beam, weight and hull form we estimate the drag the hull presents at each measured speed. The methods behind it are the industry's standard ones — Holtrop's resistance prediction for displacement hulls, Savitsky's for planing, blended for semi-displacement — in deliberately simplified form: the full methods demand inputs like a hull lines plan that no boat owner can supply, so we use reduced versions built on what owners actually know, and let the measured fleet calibrate them (the empirical estimate from the rules below). Comparing the power that drag requires with the power the motor actually drew yields overall system efficiency: motor, drivetrain and propeller losses in one figure. This is the number behind the health check — a boat that needs far more input power than its hull explains usually has a propeller or sizing problem.
• Every hull is analysed in its real regime. Displacement, semi-displacement and planing hulls each get their own physics path, and the boundary between "pushing water" and "riding on top of it" is judged on the same Froude yardstick — speed scaled against the boat's size, one definition of planing used consistently everywhere in the system. Honesty comes first: a planing hull that never actually gets on plane at its working load — a heavily loaded tender, say — is analysed as the displacement boat it is in practice. Catamarans are computed per demihull, not as one implausibly wide monohull. And cruise itself follows the regime: for planing tenders it is the minimum planing speed; for everything else, the speed at 65% of installed power — the throttle people actually use.

06 · EstimatesThe buyer estimate — computed, not looked up

Buyers can try any boat, motor and battery combination and get an estimated cruise speed and range. Methodologically, the point is this: the estimate is computed physics for those exact inputs — hull resistance from length, weight and hull form, run through the same engine that processes real submissions — not a lookup of the nearest similar boat. And it is labeled as the estimate it is.

At a hundred measured boats, the same calculation gets sharper, by design: the model's constants can be calibrated per hull class and cross-validated against runs the model never saw, so an estimate can state its evidence and an honest error band. That is the goal — a computed answer for a combination nobody has ever measured, anchored by a fleet of boats that have been.

07 · RulesThe rules we hold ourselves to

The value of this database is honesty, so every figure carries its provenance:

• Estimates say so, out loud. Our hull-resistance model is an empirical estimate in this first phase, and its constants are provisional — labeled so, openly. Every measured boat added becomes a calibration anchor: the database is the instrument that turns today's estimate into tomorrow's precision. Until then, we'd rather show you a labeled estimate than an unlabeled guess.
• Ranges, not false precision. An estimate is shown as a single honest figure, never dressed in a fabricated confidence band. Where several comparable submissions exist, we show min–max and a count.
• Quality is visible. Every submission carries a data-quality score — test conditions, wind, hull state, whether anything was estimated — shown openly, so rigorous data is findable.
• No rankings. Boats are too different for a leaderboard to mean anything. You can browse, filter, sort and compare — we won't pretend a ranking is honest.

08 · NextWhere this is going

The site opens deliberately small: a measured fleet you can trust, a five-minute reported path alongside it, a full report for every boat, and the live estimate tool for buyers. Here is what it grows into as the fleet grows: