How the numbers are made.

BatteryCaliber publishes every equation, unit, default and boundary it uses so that you, your installer or a reviewer can check the arithmetic. The calculator is a first-order planning model. It does not size conductors, select fuses or certify compliance with any code or standard.

What the calculator estimates

1. Battery energy for reserve — the energy your loads draw over the no-charge reserve period (Wh / kWh).
2. Estimated nameplate battery capacity — the rated capacity needed once usable depth of discharge and capacity derating are added back (Wh / kWh, and Ah at your actual nominal voltage).
3. Estimated continuous and surge DC current — under separate load assumptions (A).
4. First-order copper cable estimates — loop resistance, continuous voltage drop and percentage, continuous cable heat loss and surge drop.
5. Explicit unresolved status for conductor ampacity and overcurrent protection.

Not included: solar array or charge-controller sizing, charging time, AC branch design, fault-current study, automatic wire or fuse sizing, complete BMS validation, equipment compatibility, earthing/bonding design, thermal simulation or installation approval.

Reserve energy vs. nameplate capacity, amps vs. amp-hours

Reserve energy is what your loads consume during the period with no charging. Nameplate capacity is larger: you normally do not plan to use 100% of a battery, and available capacity falls with age and temperature. The calculator divides reserve energy by both allowances.

Amps (A) describe the rate of current flow at a moment; amp-hours (Ah) describe stored charge. A battery’s Ah rating does not tell you whether it can deliver a given continuous or surge current—check the manufacturer’s discharge ratings.

LiFePO4 banks are often described by a familiar class (12 V, 24 V, 48 V), but their actual nominal voltage is usually different (for example 12.8, 25.6 or 51.2 V). Always enter the manufacturer’s rated nominal voltage.

Inputs, units and accepted ranges

Ranges are software bounds that keep the arithmetic meaningful. They are not safe operating ratings. Percentages are entered as percent and converted to fractions once inside the model.

The example resistance is hypothetical—not an AWG conversion or a cable recommendation. The temperature is an assumed conductor temperature, not ambient, not a predicted temperature and not an approved terminal rating. Watts are real power, not VA; motor start behaviour depends on power factor, VA limits and duration, so consult your inverter and motor specifications.

Voltage-class selector

These are example values only—replace them with your equipment specifications. They are not BMS cutoff recommendations, and choosing a class does not enforce compatibility.

What the energy budget includes—and leaves out

The daily energy at the battery adds three separately entered quantities:

• AC appliance energy is what your appliances use, measured or estimated downstream of the inverter. It is divided by the inverter efficiency you enter to estimate the energy drawn from the battery for those loads.
• Direct DC energy is for loads wired to the battery bus without the inverter (lights, pumps, DC fridges). It is added as entered.
• Inverter overhead is standby or no-load draw not already counted in the efficiency figure or appliance demand, multiplied by the hours per day it is present.

Cable and connection losses are not added automatically. The cable section reports a peak heat loss at the continuous current you entered. Multiplying that figure by 24 hours would assume the full load runs all day and could double-count losses already reflected in measured appliance energy, so the calculator does not do it. If you know your wiring losses over a typical day, include them in the DC energy figure.

The estimate also excludes battery self-discharge and any extra reserve margin beyond the days, depth of discharge and capacity derating you enter.

Supported chemistry. The capacity step is a simplified energy model intended for LiFePO4-type lithium banks: usable energy is the nameplate energy multiplied by the usable depth of discharge and available-capacity percentages you enter. It does not model rate-dependent capacity (Peukert effect), temperature–capacity curves, the deeper derating usually applied to lead-acid batteries, or other chemistries’ discharge behaviour. For other chemistries, use the manufacturer’s rated usable capacity at your discharge rate and temperature.

Equations (formula version 1.0.0)

All intermediate values are kept unrounded.

η  = eff / 100         ηs = surgeEff / 100
D  = dod / 100         F  = retention / 100

Eday       = ac / η + dc + idle × hours                 [Wh/day]
Ereserve   = Eday × days                                [Wh]
Enameplate = Ereserve / (D × F)                         [Wh]
Ah_exact   = Enameplate / nominal                       [Ah]
Ah_display = ceil(Ah_exact)                             [whole Ah]

Icontinuous = watts / (η × minv) + dcamps + idle / minv  [A]
Isurge      = surge / (ηs × minv) + dcamps + idle / minv [A]

Rloop       = 2 × length × resistance / 1000
              × [1 + 0.00393 × (temp − 20)]              [Ω]
Vdrop_cont  = Icontinuous × Rloop                        [V]
Drop_%      = 100 × Vdrop_cont / minv                    [%]
Cable_heat  = Icontinuous² × Rloop                       [W]
Vdrop_surge = Isurge × Rloop                             [V]

Model conventions

• Overhead: enter idle overhead only if it is not already included in your efficiency or demand figures. It is added to continuous and surge current as a conservative concurrent assumption, even though its daily duration is entered separately.
• Minimum voltage: the minimum inverter-terminal voltage already represents loaded voltage. The computed cable drop is not subtracted again—this is a first-order estimate, not a solved constant-power circuit.
• Cable current: make sure the direct DC current you enter actually flows through the modelled cable segment. This is not a branching-circuit solver.
• Temperature correction is a linear copper approximation referenced to about 20 °C; alloy, stranding and connectors vary. Use your supplier’s data.
• Heat loss is peak continuous I²R loss, not a predicted cable temperature, and it is not added to daily energy (duty cycle is unknown). Budget distribution losses separately.
• No charging is credited during the reserve period. No self-discharge, contact resistance, dynamic voltage sag or temperature-capacity curve is modelled.
• Depth of discharge is a planning assumption. An 80% figure does not guarantee any cycle life.

Validation, rounding and status

• Blank, non-numeric, partial (e.g. “12abc”), infinite and out-of-range values are rejected with a field-specific message. Results are cleared—never left stale—and the report cannot be printed until inputs are valid.
• Cross-field rules: minimum inverter voltage ≤ nominal voltage; surge power ≥ continuous power; AC + DC daily energy > 0. Advisories appear when AC energy is entered with zero continuous power, or when DC energy and DC current disagree.
• Capacity is rounded upward to whole Ah (a tiny 1×10⁻⁹ Ah allowance absorbs floating-point noise only). Other results are rounded for display; very small non-zero values show as “<0.01”.
• Voltage-drop status compares unrounded values: within, at (exact equality) or above your target. If continuous drop reaches 25% of minimum voltage, or surge drop reaches 50%, cable drop, heat loss and surge drop are not shown. They are replaced on screen and in the printed report with “Not available—outside this model’s useful range”, the triggering condition and “professional analysis required.” Battery-energy and load-current results remain available because they do not depend on the cable. These thresholds mark where a first-order resistive model stops being a useful description (current would no longer be set by the minimum voltage you entered); they are not safety limits and are pending independent electrical review.
• Battery, load-current and cable sections are validated separately. Leaving both cable length and resistance blank produces a battery and current estimate with the cable marked “Not calculated”. An error in one section clears that section’s results (and any section that depends on it) immediately, and printing is disabled until every entered value is valid.
• Cable length may be entered in metres or feet (1 ft = 0.3048 m exactly); the calculation always uses metres and the report shows both.
• The cable status is a voltage-drop criterion only—a comparison with the percentage you chose—not a verdict on overall system safety.
• A within-target result means only that the stated numerical criterion was met.

Worked example (default inputs)

Why there is no automatic wire or fuse size

Safe conductor ampacity depends on the applicable installation standard and edition, conductor construction and insulation, terminal ratings, wiring method, ambient conditions, bundling, duty cycle and more. Overcurrent protection additionally depends on maximum DC voltage, available fault current, interrupt capability at that voltage, time-current behaviour, surge magnitude and duration, and placement. A handful of calculator inputs cannot determine these, and adding a disclaimer to an arbitrary table would not make it safe. Any future sizing module would require independent qualified electrical review and versioned acceptance tests before release.

Testing

Calculator v1.1.0 (formula v1.0.0) is checked by 470 automated model tests with independently hand-calculated expected values—for example the default case (8,257.78 Wh reserve, 11,469.14 Wh nameplate, 225 Ah); 3,500 Wh/day for 2 days at 100% efficiency, 80% DoD and 48 V must display 183 Ah, not 182; a 1,000 Wh/day DC-only load at 12 V must display 84 Ah; and 10 A through 1 m leads at 1 Ω/km and 20 °C against 10 V gives 0.02 V, 0.20% and 0.2 W. Tests also cover every field’s boundaries, blank and malformed entries, cross-field conflicts, every status path including the out-of-range boundary, feet/metres conversion and the appliance worksheet. Headless-browser checks cover two calculators on one page, keyboard operation, error association, result clearing and recovery, layouts from 360 px to 1440 px and the printed report. Automated tests are not a professional engineering review; independent electrical review is pending.

Sources consulted

These references support the principles and cautionary wording on this site. They do not certify BatteryCaliber, and no source organization sponsors or has approved these formulas.

Sources consulted October 2026.

Version history

• Calculator 1.1.0 / formula 1.0.0 — 24 October 2026. No change to the equations. Battery, load and cable sections now work independently (cable optional); out-of-range cable figures are withheld on screen and in reports; energy-budget limits and supported chemistry shown beside the result; status labelled “voltage-drop criterion only”; optional feet/metres entry and appliance worksheet; standalone printed report.
• Calculator 1.0.1 / formula 1.0.0 — 24 October 2026. Layout fix only: the sticky results panel clears the site header.
• Calculator 1.0.0 / formula 1.0.0 — 24 October 2026. Initial public version. Separate continuous and surge current; minimum-voltage current basis with explicit efficiency; capacity rounded upward; strict validation with no stale results; “at target” and “outside model range” statuses; no automatic conductor or fuse sizing.

Found a possible error? Email contact@batterycaliber.com with your inputs and what you expected. Formula changes are versioned, tested and recorded here.