How Much Solar Does a Boat Need? Size It From the Fridge, Not the Roof

Start With the Budget, Not the Panel

A solar array is a supply. You cannot size a supply until you know the demand, and on a boat the demand is measured in amp-hours per day: the current a device draws, multiplied by the hours it runs. A 12 V fridge compressor draws 4–6 A when it is running and runs perhaps 40–50% of the time in a warm cabin, so it uses 40–60 Ah between one sunrise and the next. An LED anchor light draws 0.15 A for ten hours and uses 1.5 Ah. Write every load down that way, add them up, and you have the number the whole system is built around.

Do it with a clamp meter or a battery monitor rather than from the labels if you can. Labels give the maximum draw, not the duty cycle, and duty cycle is everything for the fridge. A well insulated box with a keel-cooled compressor in a Scottish loch might run 20 Ah/day. The same box on a black-hulled boat in a Greek August, opened forty times a day for cold drinks, can pass 70 Ah. That single line moves more than every other line on the table put together, which is why the fridge, not the roof, is where sizing begins. Fifty millimetres of extra insulation in the box is often cheaper than the 100 W of panel it saves.

The worked example for a 10–12 m (33–40 ft) cruising boat with a fridge, an autopilot, a chartplotter, LED lights and a family's worth of phones comes to 100–150 Ah/day, with the low end at anchor and the high end on a long hot passage. A smaller boat without a fridge might be 20–30 Ah. A 14 m boat with a watermaker, a freezer and an inverter for the espresso machine can pass 250 Ah, and at that point the conversation changes, which we come to at the end.

What a Panel Really Produces

The number on the panel is its output under laboratory conditions: 1,000 W/m² of light, a cell temperature of 25°C, perpendicular sun. A boat provides none of those. To get from rated watts to amp-hours in the battery, you multiply by the hours of full-strength sun and then take off four losses in turn.

Peak sun hours. A day's worth of sunlight, compressed into the equivalent number of hours at full strength. Mediterranean summer gives about 5, the Caribbean 5–6, a British summer about 3 averaged across cloudy and clear days, a northern winter under 1. These are averages published by meteorological agencies and solar irradiance databases, and they are the single biggest reason the same panel behaves like two different products in Corfu and in Falmouth.

Heat. Silicon loses roughly 0.4% of output per degree above 25°C. A panel lying flat on a bimini in July runs at 55–65°C, so it is already down 12–16% before anything else happens. A flexible panel glued to a deck with no air behind it runs hotter still.

Shading. This is the one people underestimate, and it is the other half of the thesis. Solar cells in a panel are wired in series, so a shadow on one cell chokes the current through all of them, like a kink in one section of a hose. On a sailing boat the boom, the backstay, the radar dome and the furled mainsail all cast shadows that walk across the panels as the boat swings at anchor. A shadow the width of a boom across a single row of cells can halve the panel's output while it lasts, and bypass diodes only soften that, they do not remove it. A smaller array in clean sun outperforms a larger one under the boom every day of the season. Two panels each with its own MPPT controller, so that a shaded panel does not drag down its neighbour, is the practical fix on a boat where shade cannot be avoided.

Controller and wiring losses. A few per cent in a good MPPT, more in a PWM that throws away the panel's surplus voltage, and another few per cent in undersized cable.

Put together, the rule of thumb for a well-mounted array in Mediterranean summer is 25–35 Ah/day per 100 W. Against a 100–150 Ah budget, that is 300–400 W for a fridge-and-autopilot boat with a little margin for a cloudy day. Against a British summer's 15–22 Ah per 100 W, the same boat needs 600 W or more to do the same job, which most 10–12 m boats cannot physically carry, so the honest UK answer is a 300–400 W array plus a couple of hours of engine every second or third day. The table below gives the same figures by climate with the assumptions behind them, because the assumptions are what you should check against your own boat.

MPPT or PWM

A charge controller sits between the panel and the battery and stops the panel overcharging it. The two kinds differ in what they do with the panel's voltage. A 100 W panel produces its rated power at around 18 V. A PWM (pulse-width modulation) controller simply connects panel to battery and pulses the connection, so the panel is dragged down to battery voltage, 13–14 V, and the 4–5 V of surplus is lost as heat in the panel. An MPPT (maximum power point tracking) controller is a small DC-to-DC converter: it lets the panel run at its best voltage and converts the surplus into extra current at the battery.

The gain is typically 10–30%, and it is largest exactly when a boat needs it: on cold mornings when panel voltage is high, in low sun, in partial shade, and when the battery is flat and its voltage low. Above about 100 W of panel the extra cost of MPPT is recovered in the first season. For a 20–50 W winter trickle panel on a mooring, a decent PWM controller is fine and the money is better spent on a waterproof enclosure. Whichever you fit, buy a controller with a lithium profile even if you have lead-acid today, because you will probably not have lead-acid for ever.

Flexible or Rigid

Rigid panels are glass over a monocrystalline cell array in an aluminium frame: the panels on house roofs, in smaller sizes. They are the most efficient per watt, they run cooler because air passes behind them, and manufacturers routinely warrant them for 20–25 years. Their cost is that they need somewhere to live: a stern arch, a frame over the bimini, rail mounts, or the tops of the davits. On a 10–12 m sailing boat a stern arch carrying two 150–200 W rigid panels, aft of the boom's shadow, is the arrangement that works, and it is worth the expense of the arch.

Flexible panels are thin cells laminated in polymer. They weigh a fraction as much, bend to a coachroof or bimini, can be laced to canvas or glued down, and most can be walked on. Their cost is in life and heat. Against a surface with no air behind them they run hotter and lose more output, and the repeated flexing and UV that a boat provides produces micro-cracks and delamination; three to five years is a common working life, and some fail sooner. Use them where a rigid panel simply will not go, mount them on canvas rather than flat on a hot deck, and write the purchase date down so that when output drops you know why. Our bimini and dodger guide covers what canvas will carry.

Power boats and fishing boats have an easier time of it: a hardtop or T-top is a large flat surface with no boom over it, and two rigid panels there will out-produce four flexible ones on a yacht. The budget still starts at the fridge, and on a boat with a livewell pump running all day the pump joins it.

Wiring, Fuses and Where the Controller Goes

Solar wiring fails at the details, and the details are dull but short. Use tinned marine cable, sized so that the voltage drop from panel to controller and controller to battery is under 3% at full current; for a 300 W array over a 5 m run that generally means 6 mm² (AWG 10) or larger, and thicker is never wrong. ABYC E-11 and ISO 13297 set the standards for small-craft DC systems, and the principles are the same in both: every positive conductor fused or breakered at its source, connections crimped and heat-shrunk rather than twisted and taped, and nothing run through the bilge. Put a fuse on the panel side sized to the array's short-circuit current and a fuse on the battery side as close to the battery as you can get it, sized to the cable, because that cable is connected to a battery bank that can deliver hundreds of amps into a fault.

Mount the controller close to the battery, not close to the panels. Controllers regulate by reading battery voltage, and a long cable between the two makes the reading wrong by the drop in the cable, so the battery is undercharged in the afternoon. Keep it out of the engine bay, where the heat costs efficiency, and out of the anchor locker, where the salt costs the controller. Most MPPT units want a battery connected before the panels and disconnected after, so fit a switch or breaker on the panel side and use it in that order.

Lithium Changes the Maths

Everything above assumes the battery can take what the panel offers, and with lead-acid that is not quite true. A lead-acid bank accepts a high current only while it is well below full; from around 80% state of charge it enters the absorption stage, accepting less and less current for the last two or three hours. On a solar boat that stage falls in the early afternoon, when the panels are producing most, so a portion of the best sun of the day goes nowhere. Lead-acid also does not like living at 80%, which is what solar tends to leave it at, and it sulphates.

A lithium iron phosphate bank accepts near-full charging current until it is very nearly full, has no absorption tail worth the name, and is untroubled by sitting part-charged. The same 300 W array on the same boat delivers more usable amp-hours into lithium than into lead, because none of the afternoon is wasted, and the bank gives back more of what it stores, because lithium can be cycled to 80–90% of its capacity where lead-acid should stop at 50%. How the chemistries age is a subject of its own; our marine battery guide covers it. For sizing, the effect is that a lithium boat can get away with slightly less panel, or the same panel with more confidence.

The Winter Sitting Problem

A boat on a mooring or in a marina berth with no shore power loses charge all winter: 3–5% a month of self-discharge in a lead-acid bank, plus whatever the bilge pump float switch, the gas alarm and the stereo memory quietly take. By March the bank is flat, and a lead-acid bank left flat sulphates and never recovers its full capacity. This is the most common way marine batteries die, and it happens while nobody is aboard.

A single 20–50 W panel through a controller solves it. Even a northern winter's single peak sun hour returns 1–3 Ah a day from a 30 W panel, which is more than the parasitic draw, so the bank stays full and the bilge pump stays able to run. The controller is not optional here: an unregulated panel on a full battery in a sunny March week will overcharge it and boil off the electrolyte. Winter-trickle solar belongs on the winterisation list alongside the antifreeze.

The Honest Limits

Solar covers the fridge, the electronics, the lights, the pumps and the phones, and on a well-set-up boat in the Med it covers them entirely, which means anchoring for a week without starting the engine. It does not cover air-conditioning. A small marine unit through an inverter draws 40–80 A at 12 V, so four hours on a hot night is 200–300 Ah: more than the whole array makes in a day and more than most banks can usefully give. The same is true of induction cooking, electric water heating and a full-time freezer plus watermaker. Those are generator loads, or shore-power loads, and a boat that has them needs either a generator or a plan to be in a marina when they run.

The alternator is the other half of the honest answer. On passage days the autopilot and instruments push the budget to its top end and the engine is often running anyway; a 60–120 A alternator, with a proper external regulator if the bank is lithium, makes up in an hour what the panels lose to a cloudy afternoon. Solar is for the days the engine does not run, and those are the days that matter, because they are the days at anchor.

Keeping the Record

A solar array degrades slowly and then suddenly. Glass panels lose around half a per cent a year and are still going after two decades; flexible panels can hold output for three seasons and then lose a third of it in a month from a crack nobody can see. Controllers fail, connectors corrode, a new radar dome casts a shadow that did not exist last year. None of this announces itself. The fridge just runs the battery a little lower each night. The only way to catch it is a record: the install date, the panel and controller models and their rated outputs, the battery bank's size and chemistry, and a daily amp-hour reading from the controller or the battery monitor taken at sunset. Kept in a logbook alongside the engine hours and the service history, the same July anchorage two summers apart tells you in one glance whether the array is what it was, and whether the cause of a flat battery is the panel, the controller or the fridge. That is a system record, the same kind you keep for the engine, and it belongs in the same place; our post on what to record in a ship's log sets out the rest.

The Bottom Line

Size solar from the fridge, not from the roof. Add up the amp-hours, and on most 10–12 m cruising boats the fridge is half of them and the total is 100–150 Ah/day. Expect 25–35 Ah per 100 W in Mediterranean summer, 15–22 in Britain, and much less under the shadow of the boom, so mount the panels where the boom cannot reach them and give each its own MPPT. That makes 300–400 W for a fridge-and-autopilot boat, rigid on an arch if you can, fused at both ends with the controller next to the battery. Lithium lets the same panel do more; a 30 W panel keeps the bank alive all winter; and nothing you can bolt to a boat this size will run the air-con. Get the fridge and the shadow right and the rest of the system sizes itself.