
Step 1: Calculate Your Daily Energy Consumption
Before sizing anything, audit your loads.
Formula: kWh/day = Watts × Hours ÷ 1000
| Appliance | Watts | Hours/day | kWh/day |
|---|---|---|---|
| Refrigerator | 150W | 24h | 3.6 |
| AC (1.5 ton) | 1,500W | 6h | 9.0 |
| Lighting | 200W | 5h | 1.0 |
| EV Charger (L1) | 1,440W | 4h | 5.8 |
| Misc. | 500W | 4h | 2.0 |
| Total | 21.4 kWh/day |
Add 20–25% buffer for inefficiencies and future load growth → Adjusted target: 21.4 × 1.25 ≈ 27 kWh/day
Step 2: Choose Your System Type
On-Grid (并网)
Solar feeds the grid; you draw from grid when needed. No battery required (but can add one).
Best for: Areas with reliable grid + net metering policy.
Pros: Lowest upfront cost, highest ROI.
Cons: Goes down when grid goes down; no true energy independence.
Off-Grid (离网)
Fully self-sufficient — solar + battery + generator backup, no grid connection.
Best for: Remote locations with no grid access.
Pros: Complete independence, no utility bills.
Cons: Highest cost; battery bank must cover 2–5 days of autonomy.
Hybrid / Grid-Hybrid (离并网) — Recommended
Connected to grid AND has battery storage. Can island during outages.
Best for: Areas with unreliable grid, high TOU rates, or backup power needs.
Pros: Best of both worlds — sell excess to grid, use battery during outages.
Cons: Higher cost than pure grid-tied; requires hybrid inverter.
Recommendation for most homeowners: Hybrid system. It provides resilience without the oversizing penalty of full off-grid.
Step 3: How to Your Battery Bank/Pack
Formula:
Usable Battery Capacity (kWh) = Daily Load (kWh) × Days of Autonomy ÷ DoD
- DoD (Depth of Discharge): LiFePO4 = 80–90% usable; Lead-acid = 50%
- Days of Autonomy: Grid-tied backup = 0.5–1 day; Hybrid = 1–2 days; Off-grid = 3–5 days
Example (Hybrid, LiFePO4, 1 day autonomy):
27 kWh × 1 day ÷ 0.9 = 30 kWh usable 30 kWh ÷ 0.9 (inverter losses) = ~33 kWh nameplate capacity needed
Voltage & configuration: 48V system is standard for residential. 100Ah at 48V = 4.8 kWh nameplate. For 33 kWh: ~7 × 100Ah 48V batteries in parallel.
LiFePO4 is strongly recommended over NMC for stationary storage: safer thermal profile, 3,000–6,000 cycle life vs. 500–1,000 for lead-acid.
Step 4: How to choose Your Inverter Power
The inverter must handle your peak simultaneous load, not just average load.
Inverter Size (W) = Peak simultaneous loads × 1.25 safety factor
| Load | Watts |
|---|---|
| AC startup surge | 4,500W (3× running watts) |
| Refrigerator | 150W |
| Lighting | 200W |
| EV Charger | 1,440W |
| Peak simultaneous | ~6,290W |
6,290W × 1.25 = ~7,900W → Select 8kW inverter
Inverter types:
- String inverter (grid-tied only): cheapest, no battery support
- Hybrid inverter (recommended): handles solar, battery, grid, and backup in one unit. Brands: Growatt, Deye, SolarEdge, SMA
- Off-grid inverter/charger: for pure off-grid (e.g., Victron MultiPlus, Schneider XW+)
Key spec: check both continuous output rating AND surge/peak rating — critical for motor loads like AC compressors.
Step 5: How to choose Your Solar Panel
5.1 Understand Peak Sun Hours (PSH)
Peak Sun Hours (PSH) is the number of hours per day when solar irradiance averages 1,000 W/m². This is NOT the same as daylight hours — a cloudy day may have 8 hours of light but only 2–3 PSH.
| Location | Avg PSH (Annual) | Worst Month PSH |
|---|---|---|
| Phoenix, AZ (USA) | 5.5h | 4.2h (Dec) |
| Los Angeles, CA (USA) | 5.0h | 3.8h (Dec) |
| Dallas, TX (USA) | 4.8h | 3.5h (Dec) |
| New York, NY (USA) | 4.2h | 2.8h (Dec) |
| London, UK | 2.8h | 1.2h (Dec) |
| Munich, Germany | 3.2h | 1.5h (Dec) |
| Madrid, Spain | 4.8h | 3.2h (Dec) |
| Coastal China (Shanghai) | 3.8h | 2.5h (Dec) |
Always size for your worst month PSH, not the annual average — otherwise your system will be undersized in winter.
5.2 Core Sizing Formula
Solar Array Size (kWp) = Daily Load (kWh) ÷ PSH (worst month) ÷ System Efficiency
- System efficiency: 75–80% accounts for inverter loss (~5%), wiring loss (~2%), temperature derating (~5–10%), soiling (~2–3%), and mismatch losses (~2%)
Example (Dallas TX, worst month PSH = 3.5h, efficiency = 0.78):
27 kWh ÷ 3.5h ÷ 0.78 = ~9.9 kWp → Round up to 10 kWp
At 400W per panel: 10,000W ÷ 400W = 25 panels
5.3 Temperature Derating — A Critical Factor
Solar panels are rated at 25°C (STC). In real conditions, panel surface temperatures can reach 50–70°C in summer, reducing output significantly.
- Typical temperature coefficient for monocrystalline panels: -0.35% to -0.45% per °C above 25°C
- At 60°C panel temp: power loss = (60 - 25) × 0.40% = 14% power reduction
This is already factored into the 78% system efficiency above, but worth understanding when comparing panel specs.
5.4 Panel Types: Which to Choose?
| Type | Efficiency | Best For | Notes |
|---|---|---|---|
| Monocrystalline (Mono-Si) | 20–23% | Limited roof space | Best performance, higher cost |
| Polycrystalline (Poly-Si) | 15–18% | Budget installs | Lower cost, slightly lower efficiency |
| TOPCon / HJT | 22–25% | Premium installs | Best low-light performance, lowest temp coefficient |
| Bifacial | +5–15% gain | Ground mounts, flat roofs | Captures reflected light from rear side |
Recommendation: For home storage systems, monocrystalline or TOPCon panels offer the best balance of efficiency and longevity. Bifacial panels are excellent for ground-mount off-grid systems.
5.5 String Configuration & MPPT Sizing
Panels must be wired in strings that match your inverter's MPPT (Maximum Power Point Tracker) input voltage range.
- Typical residential MPPT input: 150–600V DC (check your inverter spec sheet)
- String voltage: Number of panels in series × Panel Voc (open circuit voltage)
- Example: 10 × 400W panels with Voc = 41V → String voltage = 410V ✓ (within range)
- Never exceed the inverter's maximum input voltage — this will damage the inverter
- For larger arrays, use multiple strings in parallel to stay within current limits
Rule: Always check Voc at minimum temperature (cold = higher voltage). Use the formula: Voc_cold = Voc_STC × [1 + Temp_coeff × (T_min - 25)]
5.6 Off-Grid: Size for Battery Recharge Too
For off-grid systems, the array must not only cover daily loads but also recharge the battery bank after cloudy days within 1–2 sunny days.
Recharge requirement: Battery capacity (kWh) ÷ System efficiency ÷ PSH Example: 90 kWh battery ÷ 0.78 ÷ 4.5 PSH = ~25.6 kWp minimum array
This is why off-grid systems require significantly larger solar arrays than grid-tied or hybrid systems.
5.7 Roof Space Check
A standard 400W monocrystalline panel occupies approximately 2.0 m² (21.5 sq ft).
| Array Size | Panels (400W) | Roof Area Needed |
|---|---|---|
| 5 kWp | 13 panels | ~26 m² |
| 8 kWp | 20 panels | ~40 m² |
| 10 kWp | 25 panels | ~50 m² |
| 15 kWp | 38 panels | ~76 m² |
If roof space is limited, prioritize higher-efficiency panels (TOPCon/HJT) to maximize output per m².
Quick Sizing Summary
| System Type | Battery Autonomy | Battery Size | Inverter | Solar Array |
|---|---|---|---|---|
| Grid-Tied | 0 (no battery) | — | 5–8 kW string | 7–8 kWp |
| Hybrid | 1–2 days | 30–60 kWh | 8–10 kW hybrid | 8–12 kWp |
| Off-Grid | 3–5 days | 90–150 kWh | 10–15 kW off-grid | 20–30 kWp |
Common Mistakes to Avoid
- Using annual average PSH instead of worst-month PSH — system will be undersized in winter
- Undersizing inverter surge capacity — AC compressors draw 3–5× running watts at startup
- Ignoring temperature derating — LiFePO4 capacity drops ~20% at 0°C; panels lose ~14% output at 60°C
- Assuming 100% DoD — always use manufacturer's recommended DoD
- Exceeding inverter MPPT voltage limits — can permanently damage the inverter
- Forgetting MPPT charge controller sizing — must match array Voc and Isc
- Not accounting for shading — even partial shading on one panel can reduce entire string output by 50%+
Tools & Resources
- PVWatts Calculator (NREL) — free solar irradiance and PSH data by location
- Global Solar Atlas (World Bank) — PSH maps worldwide
- Victron MPPT Calculator — battery/panel/controller sizing
- Growatt / Deye configurator — hybrid system sizing tools
- PV*SOL / PVsyst — professional simulation software for accurate yield modeling