Deep Cycle Battery Runtime Calculator

One cell in that table carries no figure on purpose. Trojan publishes depth-of-discharge advice for the T-105 but no cycles-at-a-depth figure for that model, so none is shown; the AGM and gel rows come from Victron's own GEL and AGM datasheet, which prints a three-row cycle design life for each range, and the LiFePO4 row from Battle Born. Read the AGM and gel rows as what they are: one maker's design life for its own cells, not a figure for every battery of that type. Prices and rates on this page are our own estimates as of the date shown; they change often. Enter your own figures for a current result. The weight-per-Ah column is our own estimate; no published source backs those figures. Check your results against your product data or a licensed professional before acting.

A lead-acid battery gives back less total energy when you pull harder, and its maker measures exactly how much less. Trojan prints four ratings for the same T-105: 250Ah over 100 hours, 225Ah over 20 hours, 207Ah over 10 hours and 185Ah over 5 hours. Against the 20-hour rating most labels quote, the 5-hour figure is 17.8% lower.

This is the Peukert effect, named after the German scientist who quantified it in 1897. High current builds lead sulfate on the plates faster than it dissolves, and the blocked area stops taking part in the reaction.

The calculator above does not ignore this. It reads the capacity each maker prints at each constant-current rate, one table per chemistry: Trojan's T-105 sheet for flooded, Victron's AGM deep-cycle table for AGM, Victron's gel column for gel. Lifeline's printed table for its GPL-31T sits 3 to 8 percent below Victron's between the 8 A and 25 A rates; for AGM and gel this calculator uses Victron's table across its full range. Between two printed rows it interpolates on a straight line in log-log space, which is Peukert's law applied one segment at a time, so every printed row comes back exactly. Lifeline separately prints a recommended Peukert constant of 1.12 for its own line. Below the range of rates a maker prints, the calculator holds the slowest printed row's capacity; above it, it extends the slope of the last two printed rows; either way the result carries a live warning rather than a silent extrapolation.

LiFePO4 takes no rate correction here, because Victron measures its own nominal capacity at a 1C discharge already. It does get the nominal voltage right: Victron prints 12.8V and 1280Wh for a 100Ah block, so feeding a flat 12V would understate the pack by 6.25%.