Home Energy Storage Winter Performance and Derating
I install home energy storage in cold-climate homes and every November my service inbox fills with the same three questions: “Why does my battery die at thirty percent state of charge in January?”, “Why does it refuse to charge when the sun is shining at minus ten?”, and “Is the system undersized or just derating?” Most of the time the answer is the second one. A ten-kilowatt-hour home battery in the data sheet often delivers six to seven kilowatt-hours on a February morning, and the inverter throttles long before the data sheet says it should. If you understand where those losses come from, you can size for the worst month instead of the best month.
Why Winter Exposes Every Weak Assumption in a Home Battery Design
A home battery is specified at twenty-five degrees Celsius, the temperature of a comfortable living room and almost never the temperature of the garage, utility closet, or unconditioned basement where the cabinet actually lives. Three things change below ten degrees: usable capacity shrinks, available discharge power shrinks, and the BMS begins to refuse charge. All three happen at once, which is why a system that performs perfectly in October feels broken in January.
At zero degrees an LFP pack delivers eighty to eighty-five percent of nameplate on a slow discharge, sixty to seventy at one-c-rate, and fifty to sixty at a 1.5C pulse. Internal DC resistance roughly doubles between twenty-five and zero, and triples between twenty-five and minus twenty.
What Actually Derates in the Cold: Capacity, Power, and Charge Acceptance
Three derating mechanisms act on a home battery at the same time, and a good winter specification treats each as a separate line item rather than a single “cold weather factor”.
- Capacity derate. Slow-discharge energy at zero is eighty to eighty-five percent of rated for LFP, seventy to eighty for NMC811, eighty-five to ninety-two for sodium-ion.
- Power derate. Peak discharge current is set from a cell-temperature look-up. Most packs derate linearly from one hundred percent at ten to forty to fifty at minus twenty. Surge headroom for an induction stove or heat-pump compressor drops in the same proportion.
- Charge derate. Below zero, lithium plating risk forces the BMS to block charge. Between zero and ten, charge is throttled to 0.05C to 0.1C for LFP with a thirty- to sixty-minute preheat. Sodium-ion tolerates a slow charge at zero to minus ten but the limit is well below the summer rate.
Stack the three losses and a 10 kWh pack at one-c-rate at twenty-five may only give 5.5 to 6.5 kWh of useful energy in real winter. AC-coupled round-trip efficiency drops from ninety-three to ninety-six percent at twenty-five to eighty-five to eighty-eight at zero and eighty to eighty-five at minus ten.
The 0 Degree Celsius Charging Rule and Why Lithium Plating Is Not Recoverable
The most common warranty fight I see is over a system that quietly charged itself to destruction on a sunny morning after a cold night. At zero degrees the lithium intercalation reaction in a graphite anode slows faster than the deposition reaction, so any charge current above a small threshold deposits metallic lithium on the surface of the anode instead of inserting into the graphite. Each plating event is two to five percent of usable capacity lost permanently, plus a real safety penalty.
Most LFP residential packs implement a hard charge block below zero at the cell level, not the cabinet level. The thermocouple that matters is the coldest one in the pack, not the average, not the cabinet wall, not the room thermostat. Good BMS firmware runs a preheat sequence until the coldest cell reads five degrees, then releases the charge limit. The preheat costs thirty to one hundred watts for a 10 kWh cabinet, roughly three to eight percent of the next day’s energy harvest.
Sodium-ion tolerates a slow 0.1C to 0.2C charge at zero to minus ten without the same plating mechanism, so a sodium-ion home battery can accept PV power on a frosty morning when an LFP pack refuses. The trade-off is calendar life: sodium-ion self-discharge is slightly higher and the high-temperature window is narrower.
Winter Sizing Math: Autonomy Days, Not Nameplate Kilowatt-Hours
Most home storage sizing is done in kilowatt-hours, which is the wrong unit for cold-climate homes. A better unit is “autonomy days at the design temperature”, which combines worst-day solar forecast, worst-night load forecast, and worst-temperature derate into one number. For a Canadian prairie home, a New England farmhouse, or a Hokkaido residence, I usually target two days of autonomy at the one-percent coldest temperature, not the annual mean.
Take the worst-day overnight load, multiply by the autonomy days, divide by usable depth-of-discharge (0.9 for LFP, 0.85 for sodium-ion), divide by inverter and DC-cable efficiency at the design temperature (0.85 to 0.90 in deep winter), divide by end-of-life derate (0.80 by year ten to fifteen), and divide again by the cold capacity factor (0.80 to 0.85 for LFP, 0.85 to 0.92 for sodium-ion). A 2.5 kWh overnight base load for two days gives 5 kWh gross load, then 5 / 0.90 / 0.88 / 0.80 / 0.82 = 9.8 kWh of nameplate. Without the cold derate the same math gives 6.3 kWh, and the homeowner will tell you in March the system was undersized.
Heat: How Much, Where, and What It Costs You
There are three ways to keep a home battery warm in winter, and the cheapest is the one nobody talks about. Self-heating from charge and discharge can lift a pack from zero to fifteen degrees over a single sunny winter day without external energy, as long as the pack is being used. The second is waste heat from a paired inverter: a five-kilowatt residential inverter at rated load dumps one hundred fifty to two hundred watts into the room from its own losses, and even at idle the standby electronics produce fifteen to thirty watts. Place the battery near the inverter and you have solved half the winter problem for free. The third is an active resistive heating pad or silicone rubber heater bonded to the cabinet base, sized between thirty and one hundred watts for a 10 kWh cabinet.
A 60 W pad running for ten hours on a minus twenty night draws 0.6 kWh, roughly six percent of a 10 kWh pack’s winter throughput. Over the four coldest months the seasonal overhead is two to four percent of annual throughput. The pad should be controlled by the BMS, not by a wall thermostat: I have walked into too many service calls where the homeowner turned the wall switch off to “save power” and the pack charged itself into a plating event two days later.
Siting and Enclosure Choices That Decide Winter Uptime
Where you put the cabinet is often the difference between a winter-derated pack and a winter-ready pack. The best location in a cold-climate home is a finished mechanical room that shares a wall with the inverter; the worst is an uninsulated garage, an attic, or an exterior wall with a north exposure. A pack mounted on an interior wall in a conditioned space typically runs eight to twelve degrees warmer than the same pack on an exterior north wall, the difference between full power and forty percent derating on a minus twenty night.
Three siting rules I follow on every cold-climate install. First, never mount on an exterior wall: the thermal bridge through the back of the cabinet will drop the cell temperature well below the room air temperature that the thermostat reports. Second, leave at least one hundred fifty millimeters of clearance on all four sides and the top so the cabinet’s own heat can rise and circulate. Third, route the DC conduit through the conditioned side of the building envelope, not the attic or rim joist.
Enclosure ratings also matter. NEMA 3R or IP54 is fine for a temperate garage, but where the temperature crosses zero fifty times a year, look for IP65 with a Gore-Tex or ePTFE vent that lets the cabinet breathe without letting in driving rain or blowing snow. Without the vent, the cabinet will pump moist winter air in and out with every swing, and a year of that will leave forty to sixty milliliters of condensate at the base, enough to corrode a busbar.
PV Harvest in Winter and the Snow Loss Nobody Models
Most home storage sizing is done in July, exactly the month when solar harvest is least predictive of winter performance. A forty-degree north latitude site that produces six kilowatt-hours per installed kilowatt per day in July will produce one and a half to two in December, closer to one to three for a steep roof with a poor tilt. If the array is sized against the inverter at the summer nameplate, the winter harvest will be twenty-five to thirty-five percent of summer. The battery cannot store energy the array did not collect.
Snow compounds the problem. A five-centimeter fresh snow layer can take a tilted array offline for two to seven days depending on tilt, temperature, and wind. Steep forty-five-degree arrays shed snow within hours; shallow ten-degree arrays can sit under snow for a week. Heated glass is a real option for critical-load homes in heavy-snow regions; it costs roughly ten percent of nameplate energy to run but keeps the array producing through every storm.
Do not trust a single-day winter generation number. Use the worst week, not the worst day, and assume one full week of array outage due to snow. For a 5 kW array in a snowy climate that is thirty to fifty kilowatt-hours of shortfall, two to four days of typical overnight load, exactly the autonomy target I set earlier.
Chemistry Choices for Cold Climates
LFP is the default for almost every residential install I do, and it remains the right answer in cold climates for one reason: ARC onset of self-heating at two hundred fifty degrees Celsius, versus one hundred ten to one hundred forty for NMC811. In a cold garage the safety margin matters more than the capacity loss, because the pack is unattended, in a wood-framed building, often within ten meters of a smoke detector. LFP at minus twenty delivers seventy to eighty percent of nameplate and refuses to charge below zero; NMC811 delivers eighty to eighty-five and tolerates a slow charge below zero, but with a thermal-runaway window one hundred to one hundred forty degrees narrower.
Sodium-ion is the chemistry to watch for the next two to three years. The hard-carbon anode tolerates a 0.1C to 0.2C charge at minus ten to minus twenty without the same plating risk, the high-temperature safety window is closer to LFP than to NMC, and the cold capacity retention of eighty-five to ninety-two percent at minus twenty is meaningfully better than LFP’s seventy to eighty. The trade-offs are energy density one hundred to one hundred sixty watt-hours per kilogram at pack level versus one hundred fifty to one hundred eighty for LFP, and a voltage window of 1.5 to 4.0 V per cell versus 2.5 to 3.65 V, so a forty-eight-volt string needs twenty-eight cells instead of sixteen.
Winter Commissioning and Acceptance Tests
Most residential storage commissioning is done in spring or fall, when the system looks great. Winter commissioning tells you whether the install is actually winter-ready. The five tests I run on every cold-climate install, in order: insulation at five hundred volts direct current greater than one hundred megaohms (below ten, trace to a wet conduit or compromised gasket before energizing); rest voltage spread less than thirty millivolts after a full charge and a two-hour rest with the cabinet off, to confirm cell balance; capacity at 0.2C at the design winter temperature, typically five to ten degrees, must be at least ninety-five percent of the design cold derate; cold charge acceptance, cool the pack below one degree with a thermocouple attached, attempt a 0.1C charge, confirm the BMS blocks the charge and the preheat sequence engages; infrared scan after a one-hour full-power discharge, looking for connector deltas above fifteen kelvin that need re-torque.
Codes that apply: UL 9540 and UL 9540A for the system level, NFPA 855 for siting, IEC 62619 and IEC 62133-2 for the cells, UN 38.3 for shipping and storage state of charge. The local Authority Having Jurisdiction will usually want to see the AHJ plan-review package and the rapid-shutdown labeling per NEC 706.705. A pack that passes the five tests above in October will sail through January.
Frequently Asked Questions
How much capacity does a home battery lose in winter?
A typical LFP home battery delivers eighty to eighty-five percent of nameplate at zero degrees and seventy to eighty at minus twenty on a slow discharge. Sodium-ion retains eighty-five to ninety-two percent at minus twenty.
Why does my home battery refuse to charge when it is cold?
Below zero the lithium plating reaction on the anode becomes favorable, and any charge current above a small threshold deposits metallic lithium that does not redissolve. The BMS blocks charge and runs a short preheat cycle until the coldest cell reads above five degrees.
Can a home battery charge from solar in freezing weather?
LFP cannot charge below zero, so the BMS blocks solar input until the pack is preheated. Sodium-ion can accept a slow 0.1C to 0.2C charge at zero to minus ten, so it can harvest PV on a frosty morning when an LFP pack refuses.
Should I keep my home battery indoors in winter?
If the site allows it, mount the cabinet in a conditioned mechanical room or a basement above ten degrees to eliminate most winter derate. Garages and exterior walls are the worst locations because the cabinet sees ambient air directly.
How much extra energy does a battery heater use in winter?
A sixty-watt resistive pad running eight to ten hours on a cold night draws 0.5 to 0.6 kilowatt-hours, roughly five to six percent of a 10 kilowatt-hour pack’s winter throughput. Over four months the seasonal overhead is two to four percent of annual throughput.
Does snow on the solar panels affect battery autonomy?
A five-centimeter fresh snow layer can take a tilted array offline for two to seven days depending on tilt, temperature, and wind. Steep forty-five-degree arrays shed snow within hours; shallow ten-degree arrays can sit under snow for a week.
Is LFP still the best home battery chemistry for cold climates?
LFP remains the default because of its two-hundred-fifty-degree self-heating onset and predictable abuse behavior. Sodium-ion is becoming the better choice for the coldest sites, with eighty-five to ninety-two percent capacity at minus twenty and tolerance for slow subzero charge.
What tests should I run on a home battery in winter?
Five tests in order: insulation at five hundred volts DC greater than one hundred megaohms, rest voltage spread less than thirty millivolts after a full charge, capacity at 0.2C at the design temperature within five percent of the design derate, cold charge acceptance below one degree to confirm the BMS blocks and preheats, and infrared scan after a one-hour full-power discharge to catch loose connectors.
Further Reading
References
