When building high-power off-grid or backup energy storage systems, connecting a massive battery bank directly to an inverter is a recipe for destroyed contactors and tripped BMS protections. Large inverters utilize massive DC bus capacitor banks (often 10,000µF to 50,000µF) to smooth high-frequency switching. If you close the main battery disconnect while these capacitors are empty, they act as a dead short, pulling thousands of amps of inrush current. To prevent this, you must know how to charge a capacitor with a resistor to create a controlled pre-charge circuit before engaging the main power path.
This guide covers the exact RC timing math for the pre-charge phase, then scales up to the broader power and energy storage system, including battery bank sizing, Peukert derating, and inverter load matching.
System Block Architecture: Source to Load Pre-Charge
A safe high-power DC system follows a strict source-to-load sequence: Battery Bank (Source) → Pre-Charge Contactor + Resistor → Inverter DC Bus Capacitor (Load) → Main Contactor.
When the system powers on, a low-current control circuit closes the pre-charge contactor. Current flows from the battery through the pre-charge resistor and into the inverter's DC bus capacitors. The resistor limits the inrush current, while the capacitor charges according to the standard RC time constant formula: $\tau = R \times C$. After a duration of $5\tau$ (five time constants), the capacitor reaches 99.3% of the battery voltage. At this point, the voltage differential across the main contactor is near zero, and it can be safely closed without arcing or inrush spikes. The pre-charge contactor then opens, removing the resistor from the continuous high-current path.
The table below provides real-world resistor sizing for common 48V hybrid inverter DC bus capacitances. Peak inrush is calculated at a nominal 48V source.
| Bus Capacitance | Resistor Value (Wirewound) | Time Constant ($\tau$) | Total Pre-Charge Time ($5\tau$) | Peak Inrush Current |
|---|---|---|---|---|
| 10,000 µF | 50 Ω (25W min) | 0.50 seconds | 2.50 seconds | 0.96 A |
| 20,000 µF | 25 Ω (50W min) | 0.50 seconds | 2.50 seconds | 1.92 A |
| 33,000 µF | 15 Ω (75W min) | 0.49 seconds | 2.47 seconds | 3.20 A |
| 47,000 µF | 10 Ω (100W min) | 0.47 seconds | 2.35 seconds | 4.80 A |
Note: Always use chassis-mount wirewound resistors (e.g., Vishay NH series). Carbon composition resistors will fail catastrophically under repetitive pulse loads. Source: Littelfuse Pre-Charge Relay Application Guidelines.
Sizing the Battery Bank: Peukert, C-Rates, and DoD
Once the DC bus capacitor is safely charged and the main contactor closes, the battery bank must sustain the inverter's continuous load. Sizing this storage requires understanding series/parallel topology, depth-of-discharge, and non-linear discharge curves.
Series vs Parallel Consequence for V and Ah
Wiring topology dictates your system voltage and capacity. Series connections sum the voltage while Ah remains constant (e.g., four 12V 100Ah batteries in series yield 48V at 100Ah). Parallel connections sum the Ah while voltage remains constant (e.g., two 48V 100Ah strings in parallel yield 48V at 200Ah).
Never parallel mismatched lithium cells, different chemistries, or cells with significantly different cycle ages. Variations in internal resistance cause current hogging, where one cell absorbs the bulk of the charge/discharge current. This leads to localized overheating, separator meltdown, and thermal runaway. Always use a BMS with individual cell balancing and install a Class-T fuse within 7 inches of the positive terminal to interrupt fault currents before they ignite the electrolyte.
Peukert's Law and Efficiency Derating
While LiFePO4 batteries exhibit a nearly flat discharge curve, lead-acid and AGM batteries suffer from Peukert's effect, where higher discharge currents yield disproportionately lower usable capacity. The formula is $t = H \times (\frac{C}{I \times H})^k$, where $C$ is rated capacity, $I$ is actual current, $H$ is the rated hour discharge time, and $k$ is the Peukert exponent (typically 1.15 to 1.30 for lead-acid, and ~1.03 for LiFePO4).
Worked Numeric Example: You have a 100Ah AGM battery rated at the 5-hour rate ($H=5$), meaning it delivers 20A for 5 hours. If your inverter pulls 50A, a linear assumption suggests it will last 2 hours. However, using a Peukert exponent of $k=1.25$:
$t = 5 \times (\frac{100}{50 \times 5})^{1.25} = 5 \times (0.4)^{1.25} = 5 \times 0.319 = 1.59 \text{ hours}$.
You lose over 20% of your expected runtime simply due to high-current derating. For further context on battery runtime calculations, refer to Battery University's runtime guide.
Charge/Discharge Limits and Inverter Sizing
Matching your battery's charge/discharge limits to the inverter's output capacity ensures you do not trigger the BMS low-voltage disconnect (LVD) or over-current protection during heavy loads.
What Charge/Discharge Limits Apply?
For standard LiFePO4 storage systems, the manufacturer charge limit is typically 0.5C (50A for a 100Ah bank), and the continuous discharge limit is 1C (100A), with a 2C peak allowed for 30 seconds. Depth-of-Discharge (DoD) should be capped at 80% to 90%. While LiFePO4 can technically be drained to 0%, chronic deep cycling accelerates capacity fade and increases the risk of cell voltage imbalance at the bottom of the curve.
Inverter/Charger Sizing for the Stated Load
Inverters are not 100% efficient; you must size the DC input based on the AC output and the inverter's efficiency curve (typically 90% to 95% at peak load).
Worked Numeric Example: Your continuous AC load is 4,000W. The inverter efficiency at this load is 93%.
1. Required DC Power: $4,000W / 0.93 = 4,301W$.
2. Continuous DC Current: At a nominal 48V (actual resting voltage ~51.2V, but calculate at the lower nominal to ensure wire sizing headroom), $I = 4,301W / 48V = 89.6A$.
3. Sizing Decision: You must size your inverter for at least 5,000W continuous output. Your battery cables, busbars, and main Class-T fuse must be rated for at least 100A continuous (which requires 2/0 AWG copper wire with 90°C insulation in a 30°C ambient environment).
Execution and Troubleshooting Decision Tree
Wiring the RC pre-charge circuit requires precision. Use a dedicated pre-charge contactor (e.g., a Gigavac or Albright rated for low-current switching) wired in parallel with the main heavy-duty contactor. The pre-charge resistor must be mounted to a heatsink or fireproof bulkhead, as it will dissipate significant heat during the charging phase ($E = \frac{1}{2}CV^2$). For a 47,000µF cap at 48V, that's over 54 Joules of heat generated in 2.35 seconds.
| Symptom | Probable Cause | Measurement / Fix |
|---|---|---|
| Main contactor chatters or welds closed | Pre-charge time too short; capacitor not fully charged before main closes. | Measure DC bus voltage with an oscilloscope. Increase RC time constant by increasing resistor value or adding capacitance. |
| Pre-charge resistor smokes or burns out | Continuous current flowing through pre-charge path; pre-charge contactor failed to open. | Verify control logic. The pre-charge contactor MUST open the millisecond the main contactor closes. Check for welded relay contacts. |
| Battery BMS trips instantly on inrush | Resistor value too low, or pre-charge contactor bypassed. | Verify resistor ohms with a multimeter (must be within 5% of spec). Ensure main contactor is physically open during the $5\tau$ window. |
| Inverter throws 'DC Bus Under-Voltage' fault | Pre-charge resistor value too high; capacitor leaks down before main contactor closes. | Check for high ESR in aging capacitors. Decrease resistor value slightly to speed up the $5\tau$ charge time. |
By correctly sizing the resistor to manage the capacitor's charge curve, and subsequently sizing the battery bank to handle the steady-state Peukert and efficiency loads, you build a storage system that survives the first millisecond of operation and delivers reliable power for the next decade.
