Most battery sizing guides fail installers because they treat backup power and solar self-consumption as the same math problem. They are not. A homeowner in Houston who wants 48 hours of refrigerator and medical equipment runtime during hurricane season needs a fundamentally different calculation than a homeowner in San Diego trying to maximize NEM 3.0 economics. Conflate the two formulas and you will sell the wrong kWh capacity, disappoint the customer, and eat the cost of a redesign.
Battery manufacturers do not make this easier. Every spec sheet leads with the nominal kilowatt-hour rating — 13.5 kWh, 16.0 kWh, 5.0 kWh — as if that number alone answers the sizing question. It does not. The usable capacity, after accounting for depth of discharge limits, round-trip efficiency losses, temperature derating, and inverter idle draw, can be 20–35% lower than the headline figure. A “13.5 kWh” Powerwall 3 delivers roughly 10.5–11.0 kWh of real, available energy to your loads on a cold morning after two years of cycling. That gap matters when you are sizing for a family of four during a winter storm.
This guide gives you two distinct sizing formulas with worked examples, a product matrix that maps real load profiles to actual battery outputs, and the electrical installation realities that determine whether your calculated kWh target is even physically possible in the customer’s panel. If you are designing storage with solar design software, the numbers in this article will help you set the right assumptions before you generate the first line item.
TL;DR: The average U.S. home consumes 30 kWh per day (EIA). If your customer wants backup power, use Formula A: (Critical Load kWh × Days of Autonomy × 1.2 Safety Factor) ÷ (DoD × Round-Trip Efficiency). If your customer wants to maximize solar self-consumption, use Formula B: (Daily Evening Peak kWh × 1.15 Efficiency Buffer) ÷ DoD. Do not mix the two. Backup sizing almost always demands more kWh than self-consumption sizing, and the wrong choice costs $5,000–$15,000 in oversized or undersized hardware.
In this guide, you will learn:
- How to separate backup power goals from self-consumption goals before you pick up a calculator
- Formula A: the step-by-step backup sizing method with a Texas family worked example
- Formula B: the self-consumption sizing method with a California array worked example
- Why the spec-sheet kWh number is a fiction — and how to calculate true usable capacity
- A side-by-side product matrix of the five most common residential batteries matched to real load requirements
- Electrical reality checks: subpanels, NEC 705.12, breaker space, and installation cost drivers
- Seasonal validation: whether December solar production can actually recharge the battery you specified
The Two Goals Every Installer Must Separate First
Before you calculate a single kilowatt-hour, determine which of two frames the customer is using: insurance or investment. The insurance frame treats the battery as protection against grid failure. The investment frame treats the battery as a tool to increase solar ROI. These two frames produce different kWh targets, different autonomy requirements, and different budget expectations. An installer who skips this distinction will recommend a 20 kWh self-consumption battery to a customer who actually needs 40 kWh for multi-day hurricane backup, or worse, sell a 40 kWh backup system to a customer who only needed 10 kWh to dodge NEM 3.0 export penalties.
The insurance-frame customer asks questions like: “How long can I run my fridge if the grid goes down for three days?” or “Will my CPAP machine work during a Nor’easter?” They are buying peace of mind, not payback period. They tolerate higher cost per kWh because the alternative — spoiled food, frozen pipes, or medical device failure — is unacceptable. For these customers, autonomy days matter more than cycle economics. You size for the worst-case scenario: a multi-day outage with no solar recharge because the roof is covered in snow or ash.
The investment-frame customer asks questions like: “How fast does this battery pay for itself?” or “Can I zero out my evening grid draw?” They are arbitraging time-of-use rates, avoiding non-bypassable charges, or maximizing self-consumption under a net metering regime that values exported energy at wholesale. For these customers, cycle depth and round-trip efficiency dominate the calculation. You size for the daily evening peak, not the once-a-decade storm. The battery becomes a financial instrument that happens to store electrons.
| Decision Factor | Backup Power (Insurance Frame) | Self-Consumption (Investment Frame) |
|---|---|---|
| Primary goal | Survive grid outage | Reduce electricity bill |
| Sizing driver | Critical load kWh × autonomy days | Evening peak kWh × daily cycle |
| Typical kWh range | 10–50 kWh | 5–20 kWh |
| DoD tolerance | Conservative (80–90%) | Aggressive (90–100% for LFP) |
| Solar dependency | Nice to have, not required | Required for economic viability |
| Payback expectation | Often none — bought for resilience | 5–10 years under favorable TOU rates |
| Worst-case scenario | Multi-day outage, no sun | High evening rates for 20 years |
| Typical customer | Houston, Miami, Phoenix, rural areas | California, Hawaii, Germany, Australia |
Most U.S. residential storage installations fall somewhere between these poles. A customer in Austin might want two days of backup for ice storms and daily arbitrage under a time-of-use plan. In those cases, size for the larger of the two formulas — almost always backup — and accept that the self-consumption economics will be stronger than calculated because the battery is already paid for by the resilience requirement.
The link between these goals and your design workflow matters. When you model a system in solar software, the battery profile should match the dominant frame. A backup-dominant profile uses shallow cycles, standby losses, and infrequent deep discharges. A self-consumption-dominant profile uses daily 80–100% depth of discharge, temperature-dependent efficiency curves, and time-of-use charge/discharge schedules. Feed the wrong profile into your design tool and your production estimates, payback calculations, and inverter load calculations will all be wrong.
Formula A — Sizing for Backup Power
Backup sizing starts with the load, not the battery. You cannot size storage until you know what must stay on, for how long, and under what weather conditions. The formula is straightforward, but every input requires field verification, not guesswork.
Formula A — Backup Sizing:
Required Usable kWh = (Critical Load kWh/day × Days of Autonomy × 1.2 Safety Factor) ÷ (DoD × Round-Trip Efficiency)
The 1.2 safety factor covers inverter idle draw, wiring losses, battery aging over 5–10 years, and the reality that customers always add “just one more” circuit after installation. Days of autonomy is your call, but it should map to local outage history and customer risk tolerance, not optimism.
Worked Example: Texas Family, 10 kWh Critical Load
Consider a family of four in Houston. Their critical load panel includes:
- Refrigerator/freezer: 150 W × 24 h = 3.6 kWh/day
- Two CPAP machines: 90 W × 8 h = 0.72 kWh/day
- LED lighting (whole house, selective circuits): 200 W × 6 h = 1.2 kWh/day
- Internet modem and router: 50 W × 24 h = 1.2 kWh/day
- Phone and device charging: 100 W × 4 h = 0.4 kWh/day
- Gas furnace blower: 800 W × 4 h = 3.2 kWh/day
- One small window AC unit (select room): 900 W × 3 h = 2.7 kWh/day
Total critical load: 13.02 kWh/day. The family rounds to 10 kWh/day by omitting the window AC unit and accepting 78°F indoor temperatures during an outage. They want two full days of autonomy based on Hurricane Harvey memories, when some neighborhoods lost power for 48–72 hours.
They are considering a Tesla Powerwall 3 (13.5 kWh nominal, 11.5 kW continuous, 89% round-trip efficiency, 100% DoD usable per Tesla’s specification). Plugging into Formula A:
Required Usable kWh = (10 kWh/day × 2 days × 1.2) ÷ (1.0 × 0.89) = 24 ÷ 0.89 = 26.97 kWh
A single Powerwall 3 at 13.5 kWh nominal and 89% efficiency delivers roughly 12.0 kWh usable. Two units deliver 24.0 kWh usable. That is still short of the 27.0 kWh target. Three Powerwall 3 units (40.5 kWh nominal, ~36.0 kWh usable) would be required to hit the formula with margin. At $700–$780/kWh installed, three units cost approximately $28,000–$32,000 before incentives. The family must decide whether to accept one day of autonomy with two units ($19,000), pay for three units, or reduce critical loads further.
This is why backup sizing is an economic negotiation, not a pure engineering exercise. The formula gives you the physics. The customer gives you the budget. Your job is to show the trade-off table.
EIA Outage Data and Autonomy Targets
The U.S. Energy Information Administration tracks System Average Interruption Duration Index (SAIDI) data annually. In 2024, the average U.S. customer experienced roughly 11 hours of outage per year total. However, that headline masks two very different outage profiles:
| Outage Type | Average Duration (U.S.) | Implication for Sizing |
|---|---|---|
| Everyday disturbances (equipment failure, trees, animals) | ~2 hours/year | A 5–10 kWh battery covers 99% of these events |
| Major storms (hurricanes, ice storms, wildfires) | ~9 hours/year | Drives the autonomy decision; 1–3 days typical target |
| Total SAIDI (2024) | ~11 hours/year | Do not size to the average — size to the customer’s worst historical event |
A customer in Maine who lost power for six days during a 2023 ice storm does not care that their annual average is 8 hours. Their autonomy target is 144 hours, not 8. That customer needs a generator, not a battery — or they need to accept that battery backup for six days is economically irrational at current lithium-ion prices.
Whole-Home Backup: The 30–50 kWh Reality
Whole-home backup — keeping every circuit alive as if the grid never failed — is expensive. The average U.S. home uses 30 kWh per day (EIA, 899 kWh/month). With HVAC, electric water heating, and pool pumps, consumption climbs to 40–55 kWh/day in hot climates. Formula A with two days of autonomy and a 1.2 safety factor demands:
- 30 kWh/day home: (30 × 2 × 1.2) ÷ (0.9 × 0.89) = 72 ÷ 0.801 = 89.9 kWh required
- 45 kWh/day home: (45 × 2 × 1.2) ÷ 0.801 = 134.8 kWh required
At $1,037/kWh median installed cost (EnergySage), a 90 kWh whole-home backup system costs approximately $93,000. Even at Tesla’s aggressive $700/kWh pricing, that is $63,000. This is why almost no installer recommends whole-home battery backup for average homes. The standard recommendation is critical-load-panel backup: 10–15 kWh for essential circuits, sized to 12–24 hours of autonomy, with a generator port for extended outages.
Formula B — Sizing for Solar Self-Consumption
Self-consumption sizing is a daily cycle problem, not an autonomy problem. The battery charges from excess solar production during midday and discharges during evening peak hours when grid electricity is expensive or when net metering credits are minimal. The goal is to minimize energy flowing to the grid at low value and energy flowing from the grid at high value.
Formula B — Self-Consumption Sizing:
Required Nominal kWh = (Daily Evening Peak kWh × 1.15 Efficiency Buffer) ÷ DoD
The 1.15 efficiency buffer accounts for round-trip losses, inverter idle draw during standby, and the fact that some solar production that could have charged the battery is instead used by concurrent home loads. Unlike backup sizing, self-consumption does not use a days-of-autonomy multiplier because the battery recharges every sunny day.
Worked Example: 8,000 kWh/year California Array, 35% Self-Consumption
A homeowner in San Diego has an 8,000 kWh/year solar array (roughly 5.5 kW DC) and is on NEM 3.0, where exported energy is credited at avoided cost rates ($0.05–$0.08/kWh) while imported energy costs $0.32–$0.42/kWh during peak hours (4 PM–9 PM). Their pre-solar usage is 10,000 kWh/year. Without storage, their solar self-consumption ratio is 35%: 2,800 kWh consumed directly, 5,200 kWh exported cheaply, and 7,200 kWh imported expensively.
Their daily evening peak usage (6 PM–10 PM) averages 8.5 kWh. This is the target the battery must cover.
Required Nominal kWh = (8.5 kWh × 1.15) ÷ 0.95 = 9.775 ÷ 0.95 = 10.3 kWh
A single LG RESU10H Prime (9.6 kWh nominal, 94.5–97.5% DC efficiency) would be slightly undersized. Two Enphase IQ Battery 5P units (10.0 kWh total nominal, 90% AC round-trip efficiency) would deliver 9.0 kWh usable and meet the target with minor grid import on high-usage days. A single Tesla Powerwall 3 (13.5 kWh) would provide comfortable headroom and allow some midday arbitrage or partial next-morning coverage.
Under NEM 3.0, every kWh stored and used during peak hours avoids $0.32–$0.42 in imported electricity. At 300 cycles per year and $0.36/kWh average peak rate, a 10 kWh battery saves approximately $1,080/year in avoided imports. If the battery costs $10,000–$12,000 installed after the 30% federal tax credit, simple payback is 7–8 years — acceptable for a product with a 10–15 year warranty, though not exceptional.
Net Metering Impact by Regime
The economic case for self-consumption sizing varies dramatically by net metering rules. Your battery recommendation should change based on the local regulatory environment:
| Net Metering Regime | Export Credit Rate | Battery Economics | Recommended Sizing Strategy |
|---|---|---|---|
| Full NEM 1:1 (legacy) | Retail rate (~$0.12–$0.30) | Poor — battery rarely pays back | Skip battery or size minimal (5 kWh) for backup only |
| NEM 2.0 (time-of-use) | Retail minus non-bypassable charges | Moderate — size to evening peak | 10–15 kWh for TOU arbitrage |
| NEM 3.0 / no export | Avoided cost ($0.05–$0.08) | Strong — every exported kWh is lost value | 10–20 kWh, maximize self-consumption |
| No net metering (off-grid hybrid) | Zero | Critical — battery is mandatory | Size to 1.5–2.0× daily load for cloudy days |
If you are modeling this for a customer, use a generation and financial tool that accepts time-of-use rate structures and NEM 3.0 export matrices. Flat-rate modeling will understate the value of storage by 30–50% in NEM 3.0 territories. The solar designing workflow should also incorporate shading and orientation losses that reduce the midday excess available for battery charging — a heavily shaded east-west array produces less midday surplus than an unshaded south-facing system, which directly reduces the battery’s daily charge opportunity.
Want to model battery payback for your specific project? SurgePV’s solar proposal software includes NEM 3.0, time-of-use, and battery degradation modeling so you can show customers exact 25-year savings by kWh configuration.
Technical Derating: What the Spec Sheet Hides
Battery manufacturers publish nominal kilowatt-hour ratings at 25°C (77°F) with brand-new cells discharged at the optimal C-rate over 5–20 hours. None of those conditions describe a real installation. A battery in an unconditioned garage in Minneapolis in January operates at a very different performance level than the same battery in a San Diego closet. You must derate the headline number before it reaches your load calculation.
Temperature Derating
Lithium iron phosphate (LFP) batteries — now the dominant chemistry for residential storage — suffer significant capacity loss at low temperatures. Unlike NMC chemistries, LFP cannot charge below freezing without risking lithium plating and permanent damage. Most LFP systems include internal heaters or prohibit charging under 32°F (0°C). Even with heaters, available capacity drops:
| Temperature | LFP Usable Capacity | Charge Limitation | Real-World Impact |
|---|---|---|---|
| 14°F (-10°C) | 60–70% of nominal | Prohibited or heater-limited | A 13.5 kWh battery acts like an 8–9 kWh battery |
| 32°F (0°C) | 75–85% of nominal | Slow charge, reduced discharge | Winter storm backup is 15–25% weaker than spec |
| 77°F (25°C) | 100% (baseline) | Full capability | The number on the brochure |
| 104°F (40°C) | 95–98% of nominal | Accelerated aging | Slight capacity loss, but cycle life degrades faster |
For backup sizing in cold climates, apply an additional 0.80 multiplier to the usable capacity if the battery will be installed in an unconditioned space. A Powerwall 3 in a Minnesota garage in February is not a 13.5 kWh battery. It is a 10.8 kWh battery on paper and closer to 9.5 kWh in practice after efficiency losses.
Inverter Idle Draw
The battery stores DC energy. Your home uses AC energy. The inverter that bridges the two consumes power whenever the battery is active — and sometimes when it is not. The Enphase IQ Battery 5P draws approximately 1.5 kWh per day in idle/standby mode. That is 547 kWh per year of parasitic load, paid for by the customer whether the battery cycles or not. Over 10 years, that is 5,470 kWh — equivalent to $600–$1,900 in electricity cost depending on local rates.
Tesla does not publish idle draw for Powerwall 3, but field measurements suggest 0.8–1.2 kWh/day in standby. LG RESU systems paired with SolarEdge or SMA inverters show 0.5–1.0 kWh/day depending on inverter model. These numbers seem small against a 13.5 kWh battery, but they are significant against a 5.0 kWh battery. A 5 kWh Enphase unit with 1.5 kWh/day idle draw loses 30% of its nominal capacity to standby before a single load is served.
Round-Trip Efficiency Penalty
Round-trip efficiency measures how much energy you get out after putting energy in. AC-coupled systems (battery → inverter → AC bus → home) typically show 85–90% round-trip efficiency. DC-coupled systems (solar DC → battery DC → hybrid inverter → AC home) can reach 94–97% because they avoid one inversion step. The difference matters at scale:
| System Architecture | Typical RTE | Energy Lost per 10 kWh Cycle | 10-Year Loss (3,000 cycles) |
|---|---|---|---|
| AC-coupled (Enphase, Tesla with standalone inverter) | 85–90% | 1.0–1.5 kWh | 3,000–4,500 kWh |
| DC-coupled (LG with hybrid inverter, SolarEdge) | 93–97% | 0.3–0.7 kWh | 900–2,100 kWh |
For self-consumption economics, every lost kilowatt-hour is a kilowatt-hour that could have avoided peak-rate grid imports. At $0.36/kWh peak rates, 4,500 kWh of 10-year losses equals $1,620 in missed savings. DC-coupled systems command a premium upfront but often recover that premium through efficiency gains in high-rate territories.
Aging Curves and Warranty DoD Limits
All lithium-ion batteries degrade. A battery warrantied for 10 years and 6,000 cycles at 70% remaining capacity will deliver significantly less energy in year 8 than in year 1. Tesla warranties Powerwall 3 for 70% capacity retention at 10 years. Enphase warranties IQ Battery 5P for 60% capacity or 15 years. LG warranties RESU10H Prime for 10 years with no explicit capacity floor, but the warranty is voided if average DoD exceeds specified limits.
For conservative backup sizing, apply a 0.80 end-of-life multiplier to the nominal capacity. A 13.5 kWh battery becomes 10.8 kWh in year 10. If you sized backup for 10 kWh usable on day one, you may have only 8.6 kWh usable in year 10 — below the critical load requirement. Size with aging in mind, or plan a replacement schedule.
Product Matrix: Matching kWh, kW, and Surge to Real Loads
The kilowatt-hour rating tells you how long the battery lasts. The kilowatt rating tells you what it can run at the same time. A 10 kWh battery with 3 kW output cannot start a 4-ton central air conditioner, even if the total runtime math works. You must match both numbers to the load profile, plus surge capacity for motor starting.
| Specification | Tesla Powerwall 3 | Enphase IQ Battery 5P | LG RESU10H Prime | LG RESU16H Prime | sonnenBatterie Eco |
|---|---|---|---|---|---|
| Nominal Capacity | 13.5 kWh | 5.0 kWh per unit | 9.6 kWh | 16.0 kWh | 2.5–15.0 kWh (modular) |
| Continuous Power | 11.5 kW | 3.84 kVA | 5.0 kW | 7.0 kW | 2.5–3.3 kW |
| Surge Power | Unknown (high) | Limited by microinverters | 7.0 kW (10 sec) | 11.0 kW (10 sec) | 3.3–5.0 kW |
| Round-Trip Efficiency | 89% | 90% (AC) | 94.5–97.5% (DC) | 94.5–97.5% (DC) | ~90% |
| Chemistry | NMC (nickel manganese cobalt) | LFP (lithium iron phosphate) | LFP | LFP | LFP |
| Max Units per Site | 4 + 3 expansion packs (94.5 kWh) | 16 units (80 kWh) | Typically 2 | Typically 2 | Modular stacking |
| Installed Cost/kWh | ~$700–$780 | ~$1,510 | ~$900–$1,200 | ~$850–$1,100 | ~$1,200–$1,500 |
| Warranty | 10 years, 70% capacity | 15 years / 60% capacity | 10 years | 10 years | 10 years, 70% capacity |
Surge and Locked-Rotor Amps
Motor starting — air conditioners, well pumps, sump pumps, refrigerators — demands 3–7× the running wattage for 100–500 milliseconds. A standard refrigerator compressor draws 600 W running but 1,800–2,500 W to start. A 3-ton AC unit draws 3,500 W running but 12,000–18,000 W (50–75 A at 240V) locked-rotor amps to start.
The Tesla Powerwall 3’s 11.5 kW continuous output with high surge capability can start most residential central AC units. The Enphase IQ Battery 5P at 3.84 kVA cannot start a central AC unit without a soft-start kit or secondary inverter support. The LG RESU10H Prime at 7.0 kW surge can handle most refrigerators and well pumps but may struggle with larger AC units. The sonnenBatterie Eco at 3.3 kW continuous is designed for European loads and small U.S. homes; it is not a whole-home backup device.
Always check locked-rotor amps (LRA) on the motor nameplate against battery surge ratings. If LRA exceeds battery surge by more than 20%, install a soft-start device or size the battery up. A failed motor start during an outage is a warranty callback and a safety issue if the load is a sump pump or medical device.
Modularity and Future Expansion
Customer needs change. A homeowner who buys one battery for self-consumption in 2026 may add an EV in 2028 and need more evening capacity. Plan for expansion even if the initial sale is a single unit.
Tesla Powerwall 3 supports up to 4 units plus 3 expansion packs, totaling 94.5 kWh. However, Tesla controls installation through certified installers, and expansion requires additional electrical work and potentially a service panel upgrade. Enphase IQ Battery 5P supports up to 16 units (80 kWh) with incremental additions that do not require inverter replacement — each unit has its own integrated inverter. LG RESU systems are typically limited to two units per hybrid inverter, so expansion beyond 32 kWh requires a second inverter. sonnenBatterie Eco uses 2.5 kWh modules that stack vertically, making it the most granular expansion path but also the highest cost per kilowatt-hour.
When presenting options, show the expansion path and its cost. A customer who knows they can add a second battery for $7,000 later is more likely to buy one battery today than a customer who thinks they are locked into their initial choice.
Electrical Reality: Subpanels, Breakers, and Space
Your calculated kWh target is meaningless if the electrical infrastructure cannot support it. Every battery installation requires physical space, breaker capacity, and code-compliant interconnection. These constraints often force a smaller battery than the formula recommends, or they add $2,000–$5,000 in electrical upgrades that must be included in the proposal.
Critical Load Panel Checklist
Backup batteries require a critical load subpanel — also called an essential loads panel or protected loads panel — that isolates backed-up circuits from non-backed-up circuits. During a grid outage, the battery only feeds the critical load panel. The main panel and its non-critical circuits go dark.
Installing this panel requires:
- Physical space: A 12–24 circuit subpanel needs wall space near the main panel, usually within 10 feet per NEC.
- Breaker capacity: The battery backfeed breaker must fit in the main panel. If the main panel is full, you need a panel replacement or a generation panel ($1,500–$3,500).
- Conduit runs: 1.25”–2” conduit from battery to panel, often through finished basement walls or exterior walls.
- Load segregation: The installer must move selected circuits from the main panel to the critical load panel. This can take 2–4 hours of electrician time and may require rewiring if existing home run cables are too short.
- Generator interlock: If the customer wants a generator port for extended outages, the critical load panel needs a manual or automatic transfer switch ($400–$1,200).
A customer with a 200A main panel that is already full with tandem breakers cannot add a 50A battery breaker without a service upgrade or a derate. A customer with a 100A main panel in an older home may need a full service upgrade to 200A before any battery can be installed. These are not edge cases. In retrofit markets, 30–40% of battery installations require some form of panel work.
NEC 705.12 Constraints
NEC Article 705.12 governs the interconnection of distributed energy resources. The key constraint for battery installations is the 120% rule (or the sum rule in the 2020 and 2023 NEC editions). The total of the main breaker plus the battery backfeed breaker cannot exceed 120% of the busbar rating in most configurations.
For a 200A panel with a 200A main breaker:
- 120% of 200A busbar = 240A
- Existing main breaker = 200A
- Maximum battery backfeed breaker = 40A (40A × 240V = 9.6 kW)
A Tesla Powerwall 3 at 11.5 kW continuous requires a 50A or 60A breaker, which violates the 120% rule on a 200A panel with a 200A main. The solutions are: (1) derate the main breaker to 175A or 150A, (2) install a generation panel with a separate busbar, or (3) upgrade to a 225A or 400A panel. Each option adds cost and complexity.
Installers who skip this calculation during the site survey end up with change orders, delayed inspections, and angry customers. Measure the panel, photograph the breaker layout, and confirm busbar ratings before you quote a battery size.
Installation Cost Drivers Beyond Battery Price
The battery itself is only part of the installed cost. According to EnergySage, the median U.S. residential battery installation cost is approximately $1,037/kWh, but that includes labor, permits, electrical hardware, and margin. The battery hardware alone is a fraction of the total:
| Cost Component | Typical Range | Notes |
|---|---|---|
| Battery hardware | $400–$800/kWh | Varies by manufacturer and volume |
| Inverter / gateway | $1,000–$3,000 | Required unless battery has integrated inverter |
| Critical load panel | $800–$2,500 | Includes labor to move circuits |
| Electrical upgrades (panel, conduit, breakers) | $1,500–$5,000 | 30–40% of retrofits need this |
| Permits and inspection | $300–$800 | Varies by jurisdiction |
| Labor (installation, commissioning) | $2,000–$4,000 | 8–16 hours for typical install |
| Margin and overhead | 20–30% of subtotal | Standard contractor markup |
NREL’s 2024 Annual Technology Baseline (ATB) places residential BESS pack costs at $283/kWh at the pack level. The gap between $283/kWh and $1,037/kWh installed is soft costs: labor, permitting, customer acquisition, and installer margin. As hardware prices fall — BloombergNEF reported lithium-ion pack prices at $108/kWh and LFP packs at $81/kWh as of December 2025 — soft costs become an even larger percentage of the total. The installer who can streamline permitting, standardize electrical packages, and reduce truck rolls will capture margin as hardware commoditizes.
Seasonal Validation: Will Solar Recharge This in December?
A battery sized for summer self-consumption can become an expensive paperweight in winter if solar production drops below the daily charge requirement. Before you finalize any kWh specification, validate whether the array can actually refill the battery on the shortest day of the year.
Solar Production Ratio: Summer vs Winter
Solar production varies by 2:1 to 4:1 between summer and winter depending on latitude, tilt, and weather patterns. A battery that charges fully by 2 PM in July may not reach 50% state of charge by sunset in December.
| Region | Summer Production (kWh/kW/day) | Winter Production (kWh/kW/day) | Winter/Summer Ratio | 5 kW Array Winter Daily Output |
|---|---|---|---|---|
| Phoenix, AZ | 5.5–6.0 | 3.5–4.0 | 65–70% | 17.5–20.0 kWh |
| San Diego, CA | 5.0–5.5 | 3.0–3.5 | 60–65% | 15.0–17.5 kWh |
| Denver, CO | 5.0–5.5 | 2.5–3.0 | 50–55% | 12.5–15.0 kWh |
| Boston, MA | 4.5–5.0 | 2.0–2.5 | 45–50% | 10.0–12.5 kWh |
| Seattle, WA | 4.0–4.5 | 1.0–1.5 | 25–30% | 5.0–7.5 kWh |
| Minneapolis, MN | 4.5–5.0 | 1.5–2.0 | 30–40% | 7.5–10.0 kWh |
A 10 kWh battery in Boston needs roughly 12 kWh of solar production to charge from 20% to 100% state of charge (accounting for DC-to-AC losses). A 5 kW array produces 10–12.5 kWh on a clear December day — barely enough. On a cloudy December day, production drops to 3–5 kWh, and the battery will not charge. That is acceptable if the customer sized for backup (the battery sits at standby until an outage) but problematic if they sized for daily self-consumption.
The 24-Hour Recharge Test
For self-consumption systems, apply this seasonal validation: calculate whether the battery can recharge from its typical morning state of charge to 100% before sunset on a clear winter day.
Example: A homeowner in Denver has a 10 kWh LG RESU10H Prime. They discharge to 20% (1.92 kWh remaining) by 7 AM after overnight standby and morning loads. They need 7.68 kWh to reach 100%. Their 6 kW array produces 15 kWh on a clear December day. After serving concurrent home loads (2 kWh during daylight hours), 13 kWh is available for charging. At 95% DC-coupled efficiency, 12.35 kWh reaches the battery. Recharge completes by 3 PM. The system passes the winter test.
Counter-example: A homeowner in Seattle has the same 10 kWh battery and a 5 kW array. December production is 5–7 kWh/day. After concurrent loads, only 3–4 kWh is available for charging. The battery recharges to 50–60% maximum. The customer sized for daily cycling but achieves only 3–4 cycles per week in winter. Payback stretches from 7 years to 12+ years. This customer should either size down to 5 kWh, accept seasonal underutilization, or reconsider the battery purchase entirely.
When to Size Down for Economics
If winter production cannot reliably recharge a large battery, the economically rational choice is often a smaller battery that charges fully most days rather than a large battery that rarely charges above 60%.
A 5 kWh battery charged daily delivers more annual cycling value than a 15 kWh battery charged to 40% three days per week. The 5 kWh system costs less, cycles at higher DoD (improving utilization), and does not strand capital in unused capacity.
Use shadow analysis to model December production at the actual array tilt and azimuth, accounting for roof obstructions and terrain shading. Do not use PVWatts annual average for battery sizing. The battery cares about the worst production days, not the average.
Conclusion
Residential battery sizing in kWh is not a single formula. It is two formulas applied to two different customer goals, modified by temperature, aging, electrical constraints, and seasonal solar production. Get the goal wrong and you will overspend by thousands or underserve the customer when the grid fails.
Three actions to take on your next battery project:
-
Run both formulas. Even if the customer says they only want self-consumption, calculate backup requirements. If the backup number is larger, present the trade-off table and let the customer choose with full information.
-
Verify the electrical infrastructure before you quote. Measure the panel busbar, count breaker spaces, photograph the load center, and confirm wall space for the critical load subpanel. A kWh target that cannot be installed is a wasted sales cycle.
-
Model December production before you model payback. A battery that does not charge in winter is a battery that does not meet its financial promises. Use actual winter irradiance data and solar software with monthly resolution, not annual averages.
Battery storage is the fastest-growing segment of residential solar. Installers who size accurately, communicate clearly, and respect the difference between backup and self-consumption will win repeat business and referrals. Those who treat 13.5 kWh as the answer to every question will spend 2026 rewriting proposals and processing refunds.
Frequently Asked Questions
How many kWh do I need for whole home backup?
A whole-home backup system typically requires 30–50 kWh for 24 hours. The average U.S. home uses 30 kWh per day (EIA). Factor in depth of discharge (DoD) and round-trip efficiency losses to get usable capacity. Most installers recommend sizing to critical loads — roughly 10–15 kWh for essential circuits — rather than whole-home backup.
How long will a 13.5 kWh battery last during an outage?
A 13.5 kWh battery (e.g., Tesla Powerwall 3) runs essential loads (fridge, lights, internet, phone charging) for 10–18 hours. On whole-home loads including HVAC and water heater, runtime drops to 2–4 hours. Solar recharging extends this indefinitely on sunny days.
What is the difference between kW and kWh in battery sizing?
kWh measures total stored energy (capacity) — how long the battery lasts. kW measures power output — how many appliances can run simultaneously. A battery with 10 kWh and 5 kW output can run a 5,000-watt load for 2 hours. Both ratings must match your loads.
Is a 5 kWh battery enough for a home?
A 5 kWh battery covers essential loads for 4–6 hours — suitable for overnight self-consumption on a small home. It is not sufficient for whole-home backup. For daily self-consumption optimization, most U.S. homes need 10–20 kWh to offset their evening peak usage.
What size battery do I need for a 2000 sq ft house?
A 2,000 sq ft home typically uses 25–35 kWh per day. For essential backup (fridge, lights, internet), 10–13.5 kWh covers 12–24 hours. For self-consumption optimization, 15–20 kWh offsets most evening grid draw. Exact sizing depends on HVAC type, occupancy, and local solar production.
How do I calculate battery backup time for my appliances?
List each appliance’s wattage and daily runtime hours. Multiply watts by hours to get Wh. Sum all appliances. Divide by battery usable capacity (kWh × DoD × round-trip efficiency). The result is runtime in hours. Add 20% safety margin for inverter losses and aging.
Can I add more batteries to my system later?
Yes, most residential battery systems are modular. Tesla Powerwall supports up to 4 units plus 3 expansion packs (94.5 kWh total). Enphase IQ Battery supports up to 16 units (80 kWh). sonnen offers 2.5 kWh modules. Confirm inverter compatibility and available breaker space before expanding.
