Calculating battery capacity correctly

What you are actually calculating

The question is never "how big should the battery be," but rather "how much energy do I draw between two charges, and how much of that can the battery provide?" These two numbers separate the calculation into a consumption part and a technology part. Those who mix them often end up with a battery bank that is too small.

The consumption part is calculated in amp-hours per day. One amp-hour is one ampere over one hour. A consumer that draws 3.8 A and runs for 12 hours consumes 45.6 Ah. As long as everything is on the same voltage level – i.e., 12 V or 24 V – you can simply add amp-hours. Only when you are calculating across converters or different voltage levels do you need to convert to watt-hours.

The five steps

1. Create a consumer list

Each consumer should be listed with its current draw in amperes and daily runtime in hours. For cyclical consumers—refrigerators, pressurized water pumps, heater fans—the actual runtime counts, not the operating time. A compressor refrigerator that is in operation for 24 hours but only runs for half of that time is listed with 12 hours.

If only the power in watts is stated on the type plate: Amperes = Watts divided by Voltage. 45 W at 12 V is 3.75 A.

2. Add quiescent current

Gas detectors, alarm systems, onboard computers in standby, displays, bilge pump automatics, and charging electronics run for 24 hours. In practice, this often amounts to 0.3 to 0.8 A, or 7 to 19 Ah per day—often more than all the lighting combined. This item is most reliably forgotten.

Measuring is better than estimating: Attach a clamp ammeter to the battery's negative terminal and switch off all intentionally used consumers. What still flows then is the quiescent current.

3. Determine autonomy days

How long should the vessel be able to operate without shore power and without the engine running? Two to three days is a common approach for cruising. If you plan for fixed charging sources—solar with a reliable daily yield, daily engine hours—you can offset their yield, but you should estimate it conservatively: a solar array in September does not yield what it does in June.

4. Convert to usable capacity

FormulaNominal capacity = (Daily consumption × Autonomy days) ÷ Permissible depth of discharge

The permissible depth of discharge is a characteristic of the technology: 0.5 for lead-acid, 0.8 to 0.9 for LiFePO4. For lithium, calculate with 0.8 — this is the operating point to which typical cycle specifications refer, and it leaves you some headroom. For lead-acid, a factor of 0.5 means you need double the nominal capacity to draw the same energy.

5. Add reserve

10 to 20 percent for aging, cold, and consumers added later. This is not a random safety margin: a battery is considered to be at the end of its life when it still delivers 80 percent of its nominal capacity—meaning for the last few years, the system will operate with a measurably smaller battery bank.

Example: Cruiser, 12 V, three autonomy days

Typical values for a 38-foot vessel in summer operation. The refrigerator accounts for about half—this is the norm, not the exception.

90 Ah times three days results in a requirement of 270 Ah. With a 10 percent reserve, approximately 300 Ah should be used. From this follows:

• LiFePO4 at 80 percent depth of discharge: 300 ÷ 0.8 = 375 Ah nominal capacity, approx. 45 kg
• AGM at 50 percent depth of discharge: 300 ÷ 0.5 = 600 Ah nominal capacity, approx. 180 kg

Exceptions and limitations of the method

The battery bank must also be recharged. A theoretically suitable capacity is useless if the charging sources cannot fill it. For lead-acid, the charging current should not fall below 0.2 to 0.3 C continuously, otherwise the battery bank will remain partially charged and sulfated. As a rough cross-check: does the daily available charging energy match the daily consumption?

Temperature. The datasheet capacity applies at 25 °C. For lead-acid, the usable capacity significantly decreases with temperature, less so for LiFePO4. Calculate winter operation as a separate load case.

Peukert effect. For lead-acid, the usable capacity decreases as the discharge current increases. If you operate a lead-acid battery bank with large inverter loads, you will get less than the calculation promises. For LiFePO4, this effect is practically negligible.

Starting currents are not included in this calculation. They determine cable cross-section and fusing, not capacity.

Typical errors

• Forgetting quiescent current. The most common individual error. It runs around the clock and adds up.
• Estimating refrigerator consumption by operating time instead of actual run time. This quickly turns 45 Ah into 90 Ah—or vice versa, too little.
• Calculating with nominal capacity instead of usable capacity. Halves the actual autonomy for lead-acid.
• Mixing watts and amperes. For multiple voltage levels, consistently calculate in watt-hours.
• Omitting reserve. A battery bank at the end of its life only delivers 80 percent.
• Only calculating for summer conditions. Heating and lighting in winter significantly change the result.

Check the result

The calculation is a planning figure, not a measurement. It only becomes reliable with an amp-hour meter including a shunt in the common negative line: after two or three trips, you will see the actual daily consumption and can refine the design. For LiFePO4, this is mandatory anyway—the flat discharge curve makes a voltage display practically useless as a state-of-charge indicator.