Large-area BIPV curved solar tile roof array for commercial buildings by GreenMore
Home

home energy storage system

home energy storage system

  • Five Layers of Battery Safety, and Where Each One Stops
    Five Layers of Battery Safety, and Where Each One Stops Dec 31, 2025
    A lithium battery is a box of stored chemical energy, and no amount of engineering makes that box empty. What engineering can do is put five separate barriers between a fault and a fire, so that a failure in one has to defeat the other four before anything visible happens. That is what a protection system is. This article goes through the five layers in the order a fault would meet them, says what each one actually does, and is explicit about where each one stops. If you want the general question of whether home storage is safe at all, the overview of battery safety covers it. This piece is about what is inside the product. Illustration, not a photograph: the cell stack, the BMS board and the flame-retardant case inside a household module. The Five Layers at a Glance Layer What it does Where it stops The cell LFP chemistry with a high thermal runaway threshold Survives abuse far better than NCM, but it is still a lithium cell The BMS Cell-level monitoring, fault isolation, balancing Acts on electrical and temperature signals, not on a cell already failing internally The pack and enclosure Flame-retardant case, pressure relief, module isolation, aerosol canister Contains a fault inside the pack; cannot cool a pack from outside Thermal management Keeps cells inside their working window all year Works on ambient heat, not on an internal short circuit Monitoring and response Remote alerts, fault codes, emergency stop, cabinet fire system Needs a working data link, and a person who acts on the alert Layer One: The Cell Every storage product we ship uses LiFePO4 cells. The reason is not cycle life, although that is the usual selling point. It is the temperature at which the chemistry starts to run away on its own. LFP cells sit above 800 °C before that happens, which is several hundred degrees higher than the NCM and NCA chemistries used in most electric cars. A cell that needs to reach 800 °C before it self-heats is a cell that has time to be caught. The cells are screened and aged in batches before assembly, so a weak cell is rejected at the factory rather than found in year four on a customer’s wall. The pack is rated at 6,000 cycles to 80% depth of discharge with a design life beyond ten years, and the cell design itself is flame retardant, which matters because it is the only layer that is present in every failure path. What this layer cannot do is make the cell non-flammable. A lithium cell holds energy by design; no chemistry changes that. LFP raises the bar for what it takes to trigger a runaway, and it lowers the energy released if one starts. That is a meaningful difference and it is not the same as safety. Layer Two: The BMS The battery management system watches cell voltage, pack current, cell temperature, state of charge and state of health. When a reading leaves its window, it acts in milliseconds: it cuts the contactor, isolates the affected module and reports a fault code. It also balances cells against each other, which is what keeps a pack from drifting out of step over thousands of cycles. Our packs report faults in three levels rather than one. A level-one warning is a drift that maintenance can plan for. A level-two warning is a condition that needs attention before the next season. A level-three warning is an immediate shutdown. The value of the split is that a dealer can tell the difference between “watch this” and “stop using it” without guessing from a single alarm. The limit is worth stating plainly. The BMS sees what its sensors report, at the points where the sensors sit. It stops overcharge, over-discharge, over-current, short circuit and over-temperature. It does not see a mechanical defect inside a single cell, and it cannot cool one. That is what the next two layers are for. Layer Three: The Pack, the Enclosure and the Aerosol Canister This is the layer buyers ask about least and should ask about most, because it is the one that differs between products at the same price. Our household packs carry a flame-retardant metal case, a pressure-relief vent and independent module isolation, so a fault in one module does not travel through the busbar to its neighbours. Inside the pack there is an aerosol canister, built in at pack level rather than offered as an external accessory. If a cell does reach its runaway threshold, the canister discharges into the enclosure and interrupts the reaction before it has enough heat to spread. The honest part: this is not uniform across the market, and it is not uniform across every size we build. Below 5 kWh, the fire configuration is deliberately simple, because the energy in the pack is small and the enclosure volume is tight. And in some cost-driven markets with light regulation, packs are sold with a BMS board and no extinguishing device at all, relying on the flame-retardant cell and case to pass. That is a real product category, and a buyer comparing two quotes should establish which one they are being offered. Product class Ingress protection Fire suppression Where it fits Household wall-mount and stacked, 5 kWh and above IP54 Aerosol canister built into the pack Plant room, garage, sheltered outdoor wall All-in-one and off-grid kits IP54 Aerosol canister inside the battery pack where the model includes one Remote sites, small homes, light commercial Small packs below 5 kWh IP54 Simplified; some models carry BMS protection only Cost-driven markets, light regulation Commercial and industrial cabinets IP54 to IP65, by model Built-in aerosol or NOVEC1230 system Outdoor compounds, microgrids, telecom sites The IP number deserves a note of its own. IP54 means dust-protected and resistant to water sprayed from any direction. It does not mean the unit can stand in driving rain. A cabinet rated IP54 belongs under a shelter or against a wall with an overhang; if it has to face the weather, the model needs IP65 and a mounting position that drains. We default to IP54 across the household range, and a buyer who intends to bolt a battery to an exposed wall should say so before the order rather than after. Layer Four: Thermal Management Heat is the slow version of the same problem. Every cycle produces some, and a cell held at 40 °C for years ages faster than one held at 25 °C. The fourth layer is about keeping the pack inside its window rather than rescuing it from a fault. Residential systems Household packs use natural convection and a passive heatsink, with no fan and no coolant. That keeps noise and standby consumption at zero and removes a wearing part, but it also means the enclosure is the cooling system: the fins have to be clear and the wall behind the unit has to be shaded. Our household packs are rated for a discharge range of −20 to +60 °C. That range is the envelope the unit operates in, not a flat performance band. Above roughly 45 °C the BMS limits charge current, and usable capacity falls slightly as internal resistance rises. Commercial and industrial systems C&I cabinets use forced air or liquid cooling, with the coolant loop running through cold plates between module rows. Liquid cooling holds cell temperatures closer together across a large cabinet than air does, which is the point of it: a wide spread between the hottest and coolest cell is what makes a pack age unevenly. We design to a tight spread but we do not publish a measured cell-to-cell figure, because we have not measured one on every model. If a tender requires that number, ask us for the test report for the specific cabinet rather than accepting a catalogue value. Layer Five: Monitoring and Emergency Response The fifth layer is the one that involves people. Every system reports to a cloud platform: state of charge, cell voltages, temperature, fault codes and event history. That is what turns a warranty claim into a diagnosis, and it is how a dealer notices a pack drifting before the customer notices a problem. On the unit itself there is an emergency stop that physically breaks the output, and on C&I cabinets a built-in fire system using aerosol or NOVEC1230. Household packs rely on the pack-level canister described above rather than a cabinet system, because there is no cabinet. The weakness of this layer is the same weakness as any monitoring: it needs a link and a person. In a weak-grid region the data connection fails as often as the grid, so local logging with store-and-forward matters more than the dashboard. A fault code that only exists in the cloud is not available to the technician standing in front of the unit. Our packs also expose RS485 and CAN locally, so a laptop can read the same values the platform sees. What the Certificates Cover, and What They Do Not Buyers often treat a certificate list as a safety score. It is not. Each certificate answers a specific question, and the useful ones are narrow. Standard The question it answers IEC 62619:2017 Whether the cell and battery system meet the safety requirements for industrial use, including thermal, mechanical and electrical abuse tests UN38.3 Whether the pack survives transport, including altitude, vibration, impact and short circuit. It is a shipping document, not an installation one CE, RoHS, IEC 62321 Whether the product may be placed on the EU market and whether restricted substances are controlled IEC 61000-6-1 and 6-3 Whether the electronics tolerate interference and do not emit too much of it None of these proves the installation is safe, because the installation is not part of the certificate. Clearance to combustibles, ventilation, cable sizing, the fire rating of the wall behind the unit and the local code all sit outside the test report. NFPA 855 is the clearest published example of how much of the safety case is installation rather than product, and the IATA dangerous goods rules show the same for transport: the UN38.3 report gets the pack onto a truck, and nothing more. Where the Layers Stop A protection system is not a guarantee, and the useful version of that statement is specific: The layers are designed for faults that start inside the pack. A fire that starts outside it, in a plant room or a neighbouring inverter, is a different event and needs a different answer. Water ingress beyond the IP rating voids the argument entirely. A flooded enclosure is not protected by a BMS. Mechanical damage during transport or installation can create a defect that no monitoring layer will see until it grows. A pack sold with BMS protection only, in a market where that is accepted, has four layers rather than five. The fifth is missing, not optional. LFP is more stable than NCM. It is not inert, and any supplier who describes a lithium battery as fireproof is telling you something about their sales material. The right conclusion is not that storage is dangerous. It is that the risk sits in specific places, and those places are inspectable. Ask for the pack drawing, the certificate numbers, the IP rating of the exact model and whether the aerosol canister is inside the pack or on the quotation as an extra. What to Ask Before You Buy Which cell chemistry, and what is the thermal runaway threshold? Is the extinguishing device inside the pack, or is it an accessory on the quotation? What is the IP rating of this exact model, and where will it be mounted? Which IEC 62619 report covers this pack, and what is its number? Does the BMS report graded warnings, or a single alarm? Can the local port be read without the cloud platform? A supplier who can answer all six in writing is selling a system. A supplier who answers three and changes the subject is selling a battery. The difference matters most in the markets where the grid is unreliable and the nearest service technician is a long drive away, which is the same set of conditions covered in the weak-grid hardware guide. For the product side, the stacked household battery and the liquid-cooled outdoor cabinet are the two formats the five layers are built into, and the wall-mounted 5 kWh unit is the smallest household format that carries the full pack-level configuration. Send the mounting location and the load list through the contact form and we will come back with a model and its actual ratings rather than a range. Illustration of an open pack. The aerosol canisters sit inside the enclosure above the cell stack, not on the accessory list. Frequently Asked Questions Can a home battery catch fire? It can, in the sense that any lithium pack stores energy and can fail. LFP chemistry makes that failure much harder to start than NCM, and the pack, enclosure, cooling and extinguishing layers are there to stop a single cell failure from becoming a room fire. What we will not claim is that the risk is zero. What happens if a cell overheats? The BMS sees the temperature rise, cuts the contactor and isolates the module, then reports a level-three fault. If the cell still reaches its runaway threshold, the aerosol canister inside the pack discharges and interrupts the reaction before it spreads to neighbouring modules. Is the fire suppression standard or an extra? On household packs of 5 kWh and above it is built into the pack. On packs below 5 kWh the configuration is simpler, and in cost-driven markets some units ship with BMS protection only. On C&I cabinets the aerosol or NOVEC1230 system is built in. Ask which configuration your quotation is for; it is the single line item most often left unspecified. Can the system be installed outdoors? Household packs are IP54 as standard, which is dust-protected and resistant to sprayed water, not to driving rain. They belong under a shelter or against a wall with an overhang. If the unit has to face the weather, specify an IP65 model and a mounting position that drains. C&I cabinets run from IP54 to IP65 by model. How is the system monitored after installation? Through the cloud platform for status, fault codes and event history, with RS485 and CAN available locally so a technician can read the same values on site without a data link. There is also a physical emergency stop on the unit that breaks the output independently of any software. Do the certificates mean the installation is safe? No. IEC 62619 covers the battery system, UN38.3 covers transport, and CE covers market access. None of them covers clearance to combustibles, ventilation, the wall behind the unit or the local code. Those are installation decisions, and they are where most of the real risk sits. P Peter Lu Energy Storage Product Manager Specifies storage systems around real installation conditions, and spends most of his time on the questions buyers ask after the price: which enclosure rating, which extinguishing configuration, which certificate covers what. Updated on November 16, 2026
  • Energy Storage Inverter vs Off-Grid Inverter: How to Choose
    Energy Storage Inverter vs Off-Grid Inverter: How to Choose Sep 02, 2025
    The two names get used as if they were interchangeable. They are not. One works with a utility and holds a battery ready for the moment the utility fails. The other is built for sites where there is no utility to fail, because there never was one. The practical difference shows up in five places: peak efficiency, backup switching time, overload capability, battery bus voltage, and whether the unit holds a grid certificate at all. Everything else follows from those. Below is how each works, the numbers that separate them, how to size an off-grid system, and which to specify. What Each One Does The storage inverter A storage inverter, often sold as a hybrid inverter, is a bidirectional converter between the battery, the solar array and the grid. It rectifies grid AC to charge the battery when power is cheap, and inverts battery DC back to AC when the site needs it. Because it runs in parallel with the utility, it follows grid voltage and frequency rather than creating them. Its second job is the one buyers care about. When the grid disappears it opens the grid-side contactor and forms a local grid for a set of backup circuits. That transfer takes milliseconds, and the backup side never sees the outage. The off-grid inverter An off-grid inverter has no grid to follow, so it creates the reference itself. It sets the voltage and frequency for the whole site and every load hangs off that. There is no transfer sequence, because there is nothing to transfer from. That changes what the design has to cover. With a grid present, an undersized inverter borrows from the utility during a load step. Off grid it trips, and every motor start becomes a system event. This is why off-grid units are specified with overload headroom, not continuous rating alone. The Numbers That Separate Them The table uses ranges, because the gap between a 4 kW wall unit and a 110 kW commercial unit is wider than the gap between the categories. Aspect Storage (hybrid) inverter Off-grid inverter Grid connection Grid-tied, islands on outage, reconnects None; forms its own reference Power flow Bidirectional, charges and discharges Mostly discharge; MPPT charges the battery Peak efficiency 97 to 98% 85 to 90% Backup switching Under 10 ms 10 ms typical, 20 ms maximum Instantaneous overload Sized for load steps, utility absorbs the rest 120% for motor start, on the better models Battery bus 48 to 51.2 V, or 100 to 1000 V high-voltage 48 V, or 60 to 145 V on transformer models Enclosure IP65, outdoor mounting Indoor wall mount, IP20 to IP21 Grid certification IEEE 1547, UL 1741 SB, VDE-AR-N 4105 Not applicable, and not obtainable Parallel operation Yes, with grid-support functions Yes, for capacity Typical unit size 3 to 15 kW residential, 50 to 110 kW commercial 1 to 10 kW per unit Two rows deserve a closer look. Peak efficiency is where the transformer costs most: the same toroidal unit that gives 120% overload gives away 10 points of loss, every hour the inverter runs. Switching time is where the hybrid wins outright, and it is the figure to ask for in writing. Grid-Forming and Grid-Following For twenty years the split was simple: grid-tied inverters followed, off-grid inverters formed. That line has moved, and the vocabulary moved with it. Why the line moved Grid-following inverters inject current and need an existing voltage reference. Grid-forming inverters hold voltage and frequency themselves, including while connected to a live grid. As grids take on more inverter-based generation and less rotating mass, operators need units that hold a network up rather than just push power into it, which is why grid-forming requirements now appear in interconnection rules that said nothing about them before. What to ask a supplier Four questions separate a grid-forming claim from a grid-forming product: Which mode does the unit run in while the grid is present, not only when it is absent? Can it black-start a dead bus with no other source online? Can it run in parallel with other grid-forming units without a communication link? Is there a test report for the transfer and the parallel behavior, or only a datasheet? A unit that only islands into a pre-charged, pre-configured state is a different product from one that can restart a network. Efficiency and Switching Time These two numbers decide most of the buying decision, and they pull in opposite directions. Peak efficiency and backup switching, the two figures that separate the categories. Where the loss goes A transformer-based off-grid inverter loses energy in the iron and copper of the transformer and in its charger magnetics. At 87% efficiency a 5 kW load burns about 750 W as heat, continuously. Over 4,000 running hours that is 3,000 kWh, which on a remote site is diesel bought and not used. A transformerless hybrid at 97.9% loses about 107 W on the same load. The off-grid unit still earns its place. The transformer is what allows 120% instantaneous overload and the surge tolerance irrigation pumps and compressors need. You trade efficiency for robustness, and that is a fair trade only when the load needs it. What 10 milliseconds buys A 10 ms transfer is invisible to almost everything: lights, routers and most controls ride through it. A 20 ms transfer will drop a desktop computer and may reset a PLC. If the backup circuits include servers or process control, specify the transfer time and ask for a test report. Sizing an Off-Grid System Off-grid sizing is not a bigger version of grid-tied sizing. It is a different calculation, and wrong in either direction costs money. A worked example Take a farmhouse drawing 30 kWh a day with an 8 kW peak, in a location with 3.5 peak sun hours in the worst month and two days of autonomy required. Step Input Result Inverter rating 8 kW peak, 25% headroom for motor starts 10 kW continuous Array size 30 kWh a day, 3.5 peak sun hours, 0.75 system efficiency 12 kWp Battery energy 30 kWh a day, 2 days autonomy, 90% depth of discharge 67 kWh Battery current 8 kW at 48 V bus 167 A Recheck Worst-month generation against worst-month load Array up 20% if the gap is negative The array divides daily energy by sun hours and system efficiency: 30 over 3.5 times 0.75 gives 11.4 kWp, rounded to 12. The battery divides two days of energy by usable depth of discharge: 60 kWh over 0.9 gives 67 kWh. Bus voltage decides the cable At 48 V, an 8 kW load draws 167 A, which means 35 mm² cable and a breaker to match. At a 120 V bus the same load draws 67 A and the cable drops two sizes. This is why off-grid systems above about 5 kW increasingly run on 96 V or 120 V batteries, and why the battery voltage range on the datasheet matters as much as the power rating. Standards and Certification A grid-tied inverter cannot be sold into most markets without certification to that market's interconnection rule. An off-grid inverter is held to safety standards only, which is a lower bar and a different one. Standard What it covers Where it applies IEC 62109-1 General safety of power converters for renewable systems Everywhere IEC 62109-2 Particular requirements for inverters Everywhere IEEE 1547 Interconnection and interoperability of distributed resources North America UL 1741 SB Certification of grid-support inverters to IEEE 1547 United States and Canada VDE-AR-N 4105 Connection of generation to the low-voltage grid Germany EN 50549-1 Connection to low-voltage distribution networks European Union AS/NZS 4777.2 Grid connection of energy systems through inverters Australia and New Zealand G98 and G99 Distribution code for embedded generation United Kingdom The IEC documents can be checked by number in the IEC webstore, the UL standard through UL Solutions. Ask for the certificate, not the claim: a unit that meets a standard and one that says it does look identical on a datasheet. How to Choose Three questions decide it: is there a grid, is it reliable, and will the site ever be connected. Grid available and reliable Specify a storage inverter with a battery. You get self-consumption of solar, lower peak purchases and backup for the circuits you choose. This is the default for homes, shops and most commercial buildings. No grid at all Specify an off-grid inverter and size array, battery and inverter together against the worst month, not the average. Keep a generator for long low-sun periods, and confirm the inverter accepts generator input and charges from it. Grid present but unreliable A hybrid inverter that can island is the most flexible answer: it earns money when the grid is up and carries the load when it is down. A pure off-grid design only makes sense if reconnection is genuinely not planned, since retrofitting grid certification is not possible. What GreenMore Builds We make both types, and the difference shows in the specifications rather than the marketing. A plant room installation: hybrid inverters on the wall, battery packs below, cabling in conduit. The GM-HW series is our storage inverter: 6 to 15 kW across five models, 97.9% peak efficiency, IP65 for outdoor mounting, 100% unbalanced load capability, and backup transfer in under 10 ms. It takes 48 V or 51.2 V lithium iron phosphate and lead-acid batteries over CAN or RS485, and carries a 3-year warranty. The GM-DP series is our off-grid inverter: 4.8 to 10 kW across four models, toroidal transformer with 120% instantaneous overload for motor starts, 10 ms typical switching, and an optional 120 A MPPT charger at 98% tracking efficiency. Wall-mount indoor unit, 512 by 320 by 136 mm, same 3-year warranty. Both are built for the export markets we serve and specified against the standards above rather than one domestic rule. Choosing between them for a site is usually a matter of sending us the load list. Frequently Asked Questions Can an off-grid inverter be connected to the grid later? Not directly. Off-grid inverters are not certified to synchronize and export to a utility network, and no retrofit changes that. If a site may be connected later, specify a hybrid or storage inverter, which supports both modes. Can a hybrid inverter work without a battery? Many can run as a standard grid-tied solar inverter before a battery is added, which lets you stage the investment. This varies by model, so confirm battery-less operation in writing if that is the plan. Which is better for a farm or remote business? Beyond grid reach, an off-grid inverter with solar, battery and backup generation is standard, sized against the worst month. Where grid power is present but unreliable, a hybrid inverter with islanding gives grid benefits plus backup and is more flexible. Why is correct sizing more important off-grid? Because there is no utility to borrow from. Undersized, loads get shed or the battery is discharged too deeply and ages early. Oversized, the capital sits idle. Real load data and a realistic autonomy figure matter most. What is the difference between grid-forming and grid-following? A grid-following inverter injects current into an existing voltage and frequency reference. A grid-forming inverter creates that reference, which lets it run an island, black-start a dead bus or support a weak grid. Most grid-tied storage inverters follow; most off-grid inverters form. Does an off-grid inverter need a generator? On most sites above a few kW, yes, at least as backup. Two days of autonomy covers normal weather, not a week of heavy cloud in winter. A generator sized to the average load, running a few hundred hours a year, is cheaper than the battery it would take to cover that week. PL Peter Lu, Energy Storage Product Manager GreenMore · Inverters and storage systems Peter works on inverter and battery configurations at GreenMore, matching unit ratings and bus voltages to the loads a site actually has rather than the loads in a quotation. Updated on November 17, 2026
  • How do I add battery storage to my solar system?
    How do I add battery storage to my solar system? Jun 11, 2025
    One installer quotes A$8,284 to add 10 kWh to a house in Adelaide that already has panels. Another, in Kraków, quotes €7,800 for the same job. A third, working three streets from the second, says the inverter has to come off the wall first, and quotes €10,900. The batteries are the same. The box that has been converting sunlight since 2019 is not. Start with the label on your inverter Find the model number before anyone quotes you. It is on a sticker on the side or underside of the inverter, beside the DC isolator, and it decides which route you are on. Hybrid inverters have a battery terminal already. Huawei SUN2000-M1, Sungrow SH-RS, GoodWe ET and EH, Solis S6, Fronius GEN24 Plus, SolaX X1-Hybrid. If the sticker says any of those, a battery bolts onto the DC bus and the inverter stays put. String inverters do not: Fronius Primo and Symo, SMA Sunny Boy, older GoodWe units. These need a second inverter on the AC side, or replacement. Microinverters have no DC bus to reach, so the battery arrives with its own inverter and lands at the switchboard. What is on the wall How the battery connects What you are paying for Hybrid inverter Straight onto the DC bus, one machine Battery and commissioning String inverter Second inverter on the AC side Battery, battery inverter, second isolator Microinverters AC battery at the switchboard Battery with built-in inverter, CT clamp The model number also decides which batteries it will accept, which is where people lose money. Compatibility lists live on the inverter maker's own site, not a distributor page. Search the model plus "approved battery list" and read the PDF. An all-in-one unit with the inverter inside skips the problem. Two ways in, and what each costs AC coupling keeps everything you own. The existing inverter carries on making power for the house, and a second one sits beside it, charging from the AC side and pushing back after dark. Two machines, two isolators, one extra conversion. DC coupling means removing a working inverter and fitting a hybrid. More hardware, more labour, one machine at the end. The gap is smaller than people expect at small capacity and widens with size, since the second inverter is a fixed cost. Market AC coupling, 10 kWh Replace with hybrid Which wins Australia A$9,000–14,000 before rebate A$2,500–3,500 on top AC, unless the inverter is near end of life United States $12,000–18,000 installed $2,000–5,000 on top AC Poland €7,800 €10,900 AC by €3,100 India ₹1.4–1.8 lakh ₹2.3–2.9 lakh for 10 kWh AC The upper path stores before the inverter. The lower path stores after it, which is why it needs a second machine and loses one more conversion. The Australian price index puts the difference at A$500 on 10 kWh, A$800 at 20 and A$600 at 30, the same story backwards. For the battery itself across six markets, the 20 kWh cost breakdown goes line by line. The rule installers use is age. Seven years or more of warranty left, couple on the AC side. Three years from the end, replace it now, or you pay for the AC route twice. The efficiency gap, in money DC coupling is more efficient, and not by a little. Round-trip lands at 94 to 96 percent because the energy crosses the inverter once. An AC-coupled battery crosses two in and two out, and lands at 89 to 92. Line DC coupled AC coupled Through the battery, per year 2,500 kWh 2,500 kWh Round-trip efficiency 96.5% 90% Lost in conversion 88 kWh 250 kWh Cost at €0.30/kWh €26 €75 Over fifteen years €396 €1,125 A house using 6,000 kWh a year, an 8 kWp array, a 10 kWh battery cycling 250 times: about 2,500 kWh passes through the battery, so the gap is €49 a year. The Polish quote charges €3,100 for the efficient route, which needs 64 years to repay. Nobody keeps a battery that long. System level it is smaller again: thirty to forty percent of output reaches the house without touching the battery, so the gap sits near 4 percent of generation. None of which makes AC coupling better. It makes efficiency the wrong argument. If a salesperson leads with round-trip percentage, ask what it is worth a year. What your inverter will refuse to talk to The largest cause of wasted money in retrofits is buying before checking the list. Half the major brands run closed ecosystems, and the pack either will not connect or connects unmanaged. Inverter Batteries it accepts Huawei SUN2000 LUNA2000 only Sungrow SH series SBR only SolarEdge Home Hub SolarEdge Home Battery only Enphase IQ8 IQ Battery only Tesla Powerwall 3 Tesla expansion units only Fronius GEN24 Plus BYD, LG RESU FLEX, Fronius Reserva Deye, Sol-Ark, EG4, MPP Pylontech-protocol packs, published lists Victron Tier 1 list, about 28 brands tested Victron publishes three tiers: Tier 1 tested with closed-loop communication, about 28 brands; Tier 2 same protocol, not listed; Tier 3 open loop, voltages set by hand. On the 48-volt side Pylontech CAN is the de facto handshake, which is why Deye, Sol-Ark, Growatt and EG4 take the same third-party packs. GoodWe ET and EH run an open high-voltage interface, but need ARM firmware v14 or later, which older units in the field lack. One detail separates a battery that works from one that misbehaves in month four. A managed pack has to lower its charge voltage limit as it fills. Packs that clamp current without dropping the ceiling trip overvoltage warnings at the top of every cycle, and the fault log blames the inverter. If your inverter is on the closed list, you buy that brand's battery or change the inverter. Both are legitimate. Finding out after delivery is not. Clipping: the one thing AC coupling cannot recover Most arrays are oversized against their inverter. 6.6 kW of panels on a 5 kW inverter is normal, and for two hours on a clear day the extra 1.6 kW is not harvested. A DC-coupled battery sits before that bottleneck and absorbs the excess. An AC-coupled one sits after it and never sees energy already thrown away. On a 1.3 DC-to-AC ratio in a sunny climate that is worth real money; on 1.1 in northern Europe it is close to nothing. Paperwork that comes before the drill In most markets a battery is not an appliance but a change to a registered generator. The rules are dull and the penalties are not. The United Kingdom is the clearest case. A battery inverter counts as a generation unit, and its capacity is added to the solar inverter on site. Above 16 amps per phase, 3.68 kW single phase, G98 notification becomes a full G99 application. A 3.6 kW array with a 5 kW battery is 8.6 kW combined, so it is a G99: around 45 working days, and no energising until granted. Market The threshold that catches people Time United Kingdom Combined inverter output above 16 A per phase triggers G99 45 working days Australia, Queensland Any change to a registered system needs approval first 10 or 65 working days California Rule 21 applies even to systems set not to export Varies by utility New Jersey An existing PTO does not cover added equipment New application New Jersey says it plainly: a battery is a modification, the original permission to operate does not extend to it, and a fresh review is required even if it never exports. California does the same under Rule 21. What the application asks for is the same everywhere, and none of it is hard to produce: A single-line diagram showing the new inverter or battery alongside the existing array The inverter model number and its G98, G99 or equivalent certificate The battery model and its test certificate to IEC 62619 or equivalent The installer's accreditation number for the scheme in force Ignoring this rarely produces a fine. It produces an inspector who will not sign, and an insurer asking why the array no longer matches the certificate on file. In Australia the Clean Energy Council battery list is the one that matters. When the answer is no Five situations where the arithmetic fails. Net metering is the first. Where exports earn the retail rate, storing a kilowatt-hour and using it later is worth the same as exporting it now, minus the round-trip loss. You pay thousands to lose 5 percent. Wait for the scheme to change, as the Netherlands does on 1 January 2027. the payback analysis. A high export tariff is the second. Australian benchmark feed-in tariffs have run at 3 to 6.5 cents a kilowatt-hour, and shifting load off a 30-cent import rate earns well; 20 cents for exports leaves little spread. An inverter near the end of warranty is the third. The fourth is a legacy fixed feed-in contract. Several schemes require notification when storage is added, and a few treat it as material change. Read the contract first. The fifth is the United States, where the arithmetic changed in 2026. The residential clean energy credit under section 25D no longer applies to property placed in service after 31 December 2025. A homeowner buying a battery this year has no federal credit to subtract, and quotes showing a 30 percent line are working from last year's rules. British buyers keep the VAT zero rate until 31 March 2027, if the same contractor supplies and installs. Backup is a second purchase A battery does not come with backup. It comes with the ability, if the inverter has an islanding mode, the switchboard has a backed-up circuit, and the transfer is fast enough for what is plugged in. Transfer time is the number nobody quotes. A desktop power supply holds up for 16 to 20 milliseconds when the grid drops. A domestic battery in backup mode switches in under 20, the good ones under 10. At 30 the desktop reboots. The second number is the backed-up load. A fridge at 150 W, lights and sockets at 200 W, the router at 30 W and a boiler at 120 W draw about 500 W: roughly 16 hours on 10 kWh, 32 on 20. Whole-home backup is a different product, with a bigger inverter and soft-start. In the UK position matters as much as capacity. BS 7671 and PAS 63100 put batteries out of lofts, bedrooms and escape routes, so a pack in the garage changes the cable run. What to put in the quote request Eight questions, sent before anyone visits. The answers separate an installer from a salesperson, and you can send them to us first. Ask for Why it changes the price Your inverter model and firmware version Decides AC or DC before anything else The approved battery list for that model Stops you buying a pack it refuses Whether a G99 or equivalent application is included That is 45 working days of waiting Measured round-trip efficiency, not nameplate Quoted figures flatter Backup circuits and transfer time in milliseconds "Backup ready" means nothing alone Whether the existing inverter warranty survives Some void on third-party batteries A single-line diagram The one drawing every inspector wants Who signs the commissioning certificate A name, not a company A wall-mounted 5.12 kWh unit paralleled eight deep reaches 40.96 kWh on the same inverter, often the cheapest way to grow later. The day itself A battery on an existing hybrid is three to five hours. An AC-coupled system with a new circuit and a CT clamp runs six to eight. Replacing a string inverter is a full day, most of it in the switchboard. Power is off for two to six hours, and the board decides which. A 1980s consumer unit with no spare ways adds a day. Approval is the long part: 10 to 65 working days in Australia, around 45 in the UK. Order the paperwork first, or the battery sits in its crate. Photograph the inverter label and the inside of the switchboard first. Every installer asks for both, and having them ready takes a visit off the quote. Frequently Asked Questions Can I add a battery to my existing solar system without replacing the inverter? Yes, if the inverter is a hybrid with an unused battery terminal, or if you accept a second inverter on the AC side. A string inverter cannot take a DC battery but need not be removed. How much does it cost to add a battery to an existing solar system? A$9,000 to 14,000 in Australia before the rebate for 10 kWh AC-coupled, $12,000 to 18,000 in the United States, €7,800 in Poland for the AC route against €10,900 to replace the inverter. The battery is 50 to 60 percent of the invoice. Do I need a new inverter to add a battery? Only for DC coupling. With a string inverter, AC coupling adds a battery inverter and leaves the original in place. Replacing becomes cheaper within three years of the warranty ending. Can I add a battery to a system with microinverters? Yes. With no DC bus, the battery arrives as an AC unit with its own inverter and connects at the switchboard, plus a current transformer on the incoming supply to see whether the house is importing. Will adding a battery affect my solar feed-in tariff? It can. Legacy fixed tariffs sometimes require notification, and a few treat storage as a material change. Under net metering a battery adds little anyway, because exports already earn the retail rate. How long does it take to add a battery to an existing system? Three to five hours on an existing hybrid, six to eight with a new AC circuit, a full day if the inverter is replaced. The grid application is the real timeline: 45 working days in the UK, 10 to 65 in Australia. PL Peter Lu Energy Storage Product Manager Works on retrofit projects where the inverter is already on the wall and the budget is fixed. Writes about what the label says, what the grid operator will accept, and which numbers actually move a decision. Updated on November 17, 2026
  • What size solar system do I need for my home?
    What size solar system do I need for my home? May 06, 2025
    A house using 10 kWh a day needs 2.1 kW of panels in Dubai and 4.3 kW in London. Same bill, same roof, twice the array. Nothing about the house changed. Peak sun hours did. The number that decides system size is not how much sun a place gets in a year, it is how many hours a day the sun arrives at the 1,000 W/m² the panels are rated at, and that figure runs from under 3 in northern Europe to nearly 6 in the Gulf. What follows is the whole chain: your daily kWh, your peak sun hours, the performance ratio that turns one into the other, the roof that has to hold it, the inverter that has to clip it, and what the result costs in 2026. Every figure is traceable, and every table is one you can check against your own bill. What your daily kWh actually is Start with kilowatt-hours, not panels. A 400 W module is a unit of hardware; a kilowatt-hour is a unit of work. Sizing backwards from panel count is how people end up with an array that covers half their bill and a roof they cannot extend. Read it off the bill Most utilities print twelve months of kWh on the bill or in the online account. Add them and divide by 365. If you only have one month, divide by 30 and accept the error, because a January bill in Chicago and a July bill in Phoenix are not describing the same house. A worked example: 3,650 kWh a year is 10 kWh a day. That is a small household with gas heating, or a well-insulated one with a heat pump and no EV. Build it up from appliances With no bill to read, add the loads. The figures below use the US Energy Information Administration end-use shares for an average household of 10,791 kWh a year, with the large single loads called out separately because they dominate the total. Load kWh per year Notes Space heating and cooling 4,500 EIA puts HVAC at 42 percent of US household electricity Electric resistance water heater 3,500–4,000 A heat pump water heater cuts this to roughly 1,000 EV home charging, 15,000 miles 4,500 At 0.30 kWh per mile, the US average across vehicle classes Heat pump, mixed climate 5,500 About 4,000 in mild climates, over 8,000 in cold ones Lighting 970 9 percent of the average household Refrigerator 430 4 percent; a 1990s unit can use twice this Clothes dryer 430–610 4 percent of the average; heat pump dryers run lower Cooking, electric oven and range 430 4 percent Television and electronics 430 4 percent Standby and everything else 2,050 About 19 percent, and the easiest place to find savings Two lines deserve attention before anything else. An EV driven 15,000 miles adds 4,500 kWh a year, which is 43 percent of an average household total, and a heat pump adds more than that. A home planning either one should size for the load it will have in three years rather than the one on last year bill. The seasonal trap A 30 kWh a day average can hide a 55 kWh August and a 14 kWh April. A grid-tied system does not care, because the grid absorbs the swing across the year. Off-grid systems, and anyone targeting a fixed self-consumption percentage, do care, and for them the sizing input is the worst month rather than the annual average. Pull both from the same source before choosing. Peak sun hours, by place A peak sun hour is one hour of 1,000 W/m² irradiance. Four peak sun hours a day means the array receives the energy equivalent of four hours at full rated output. It is neither daylight hours nor sunshine hours, and the confusion between the three is behind most wildly optimistic system estimates. The spread is the whole story. Annual average peak sun hours at the best fixed tilt, from PVGIS 5.3, with the array needed to cover 10 kWh a day at a performance ratio of 0.8: City Peak sun hours kWh per kW a year 10 kWh a day needs Roof at 5–7 m²/kW Dubai, UAE 5.95 1,737 2.10 kW 11–15 m² Phoenix, AZ 5.88 1,717 2.13 kW 11–15 m² Los Angeles, CA 5.59 1,632 2.24 kW 11–16 m² Sydney, AU 4.72 1,378 2.65 kW 13–19 m² Houston, TX 4.59 1,340 2.72 kW 14–19 m² Chicago, IL 4.22 1,232 2.96 kW 15–21 m² Boston, MA 3.91 1,142 3.20 kW 16–22 m² Seattle, WA 3.40 993 3.68 kW 18–26 m² Berlin, DE 3.06 894 4.08 kW 20–29 m² London, UK 2.93 856 4.27 kW 21–30 m² Dubai needs 2.10 kW and London needs 4.27 kW for the identical bill. That is 2.03 times the array, and the roof goes from 11 m² to 30 m². A guide that says "3 to 6 peak sun hours" and stops has handed you a range wide enough to double your equipment order in either direction. Identical bill, four cities. Peak sun hours set the array, not the kWh. Pull your own figure from PVGIS, the European Commission JRC tool, which is free and covers the whole world, or from the World Bank Global Solar Atlas. Both report monthly values as well as annual, and the monthly view is what tells you whether a December will still cover your load. The IEA PVPS programme publishes the country-level yield statistics behind those national averages, which is the honest way to sanity-check a single site against its peers. From kWh to kW, and where the 20 percent goes Divide daily kWh by peak sun hours and you get a number no real array delivers. Energy is lost between the cell and the meter, and the planning figure for that loss is the performance ratio. Required array (kW) = daily kWh ÷ peak sun hours ÷ performance ratio A home using 10 kWh a day at 4.5 peak sun hours: 10 ÷ 4.5 ÷ 0.8 = 2.78 kW. Round up to 3 kW and the array has a little headroom without paying for a second string. The 0.8 is not a fudge factor. It is the product of four losses that behave differently in different places, and knowing which one dominates is what separates a conservative design from a guess. Loss Typical What drives it Module temperature 5–10% Cell temperature above 25 °C. Coefficient −0.34%/°C for PERC, −0.32 for TOPCon, −0.24 for HJT Inverter conversion 2–5% European efficiency 97–98% on good units, 94–96% on budget ones Wiring and connections 1–3% DC cable runs, connectors, string mismatch Soiling and shading 2–5% Dust, pollen, snow, chimneys, trees. Worse where it rarely rains Temperature is the loss that moves most, and the one almost nobody quotes. Cell temperature is ambient plus irradiance times a mounting factor: 0.030 K·m²/W for a module on rails with air behind it, 0.050 for a module with no air gap. At 1,000 W/m² and 25 °C ambient, that is 55 °C ventilated and 75 °C integrated. A solar tile is the second case. It is the weather surface, so there is no air gap and no rear ventilation, and it runs about 20 °C hotter than a panel on rails. On a −0.34%/°C module that costs roughly 7 points of performance ratio, which is why an integrated roof needs more installed kW for the same annual kWh, and why the fair comparison between a tile roof and a panel roof is kWh per square metre of roof rather than watts per tile. The Fraunhofer ISE Photovoltaics Report carries the measured efficiency and temperature data behind those coefficients. Cold works the other way. A UK array runs a mean weighted cell temperature of 35 to 42 °C, so its temperature loss is 5 to 6 percent rather than the 10 percent a Gulf site sees. The same 0.8 planning figure is conservative in London and optimistic in Dubai, which is another reason to build the number from local data rather than borrowing one. What the array has to fit A 400 W module measures 1,722 × 1,134 mm, about 1.95 m², and weighs 21 kg. Power density is roughly 205 W/m² at the module, and that is the number that decides how much fits on your roof. Array Panels at 400 W Module area Roof area needed 2 kW 5 9.8 m² 13–15 m² 3 kW 8 15.6 m² 20–25 m² 4 kW 10 19.5 m² 25–30 m² 5 kW 13 25.4 m² 32–38 m² 6 kW 15 29.3 m² 37–44 m² 8 kW 20 39.0 m² 49–59 m² 10 kW 25 48.8 m² 61–73 m² Module area is not roof area. Edge margins, row spacing, chimneys, skylights, vent stacks and the access path an installer needs all come out of the usable figure, and fire-code setbacks take more. A 2025 Berkeley Lab sample found residential arrays covering 16 to 39 percent of total roof area, which is another way of saying most roofs have far less usable plane than the plan drawing suggests. South is best in the northern hemisphere and north in the southern. East or west costs roughly 10 to 15 percent of annual yield for the same array. A north-facing plane at 40° latitude is not worth the mounting hardware. Splitting an array east and west flattens the production curve, which helps self-consumption and costs nothing except peak output. Roofs covered in tile or slate need one more decision. A photovoltaic tile replaces the covering rather than mounting on it, which changes both the area calculation and the cost comparison. The solar tile range shows what that looks like on a pitched roof, and what BIPV actually means covers when the integrated route beats a panel array on cost as well as on looks. Inverter sizing, and why the array is bigger than the inverter An array rated 6 kW rarely has a 6 kW inverter behind it. The ratio of DC nameplate to AC output is the inverter loading ratio, and residential designs run 1.2 to 1.3 almost everywhere. Above 1.0 the inverter clips at midday, and the extra modules earn their keep in the morning, the evening and the cloudy hours. DC/AC ratio Clipping loss Where it fits 1.15 Under 1% Low-irradiance markets: northern Europe, UK, Pacific Northwest 1.20 About 1.5% Standard commercial design, moderate climates 1.25 About 2% The residential default in most markets 1.30 About 3% High irradiance with time-of-use export: California, Arizona, the Gulf 1.35 About 4% High latitude, or generous net metering 1.50 About 8% Only where modules are very cheap against inverters Moving from 1.0 to 1.3 adds 6 to 12 percent annual yield for the price of the extra modules, because inverter cost per watt exceeds module cost per watt. NREL describes 1.2 to 1.5 as the typical range for new systems and assumes 1.34 for utility-scale in its 2024 ATB. What caps the ratio in practice is not economics but the datasheet: every inverter publishes a maximum DC input, and exceeding it voids the warranty before it hurts the yield. For a buyer this matters at the quoting stage. A 6 kW array on a 4.8 kW inverter is a 1.25 ratio and a sensible design. A 6 kW array on a 6 kW inverter is a 1.0 ratio, and you paid for AC capacity you will use a few hours a year. The IEC 62548 standard covers the array design and protection side of that decision. What it costs US residential solar installed at $3.36 per watt in the third quarter of 2026, the SEIA and Wood Mackenzie benchmark. That is gross, before any incentive, and it covers panels, inverter, racking, wiring, labour, permits and interconnection. One thing changed for 2026 and it changes the arithmetic. The 30 percent federal residential credit under Section 25D expired on 31 December 2025. A homeowner buying a system outright in 2026 gets no federal credit. Third-party owned systems, prepaid leases and power purchase agreements still carry a business-claimed credit through 2027, which is why lease quotes look better this year than they did last year. The US Department of Energy homeowner guide tracks the current position. Array Gross cost at $3.36/W Roof needed Panels at 400 W 3 kW $10,080 20–25 m² 8 5 kW $16,800 32–38 m² 13 6 kW $20,160 37–44 m² 15 8 kW $26,880 49–59 m² 20 10 kW $33,600 61–73 m² 25 15 kW $50,400 92–110 m² 38 Where the $3.36 goes, per watt: modules $0.25–0.35, inverter $0.15–0.35, racking $0.10–0.20, balance of system $0.10–0.15, labour $0.50–0.80, and soft costs — permitting, design, sales, overhead and margin — $0.60–1.00. Soft costs are where two quotes on identical hardware diverge, and where negotiation actually moves anything. Over 25 years the benchmark price works out to 9 to 13 US cents per kWh, against an average US residential rate of 18.34 cents in June 2026. That gap is the whole investment case, and it is widest in California at 34.74 cents and thinnest in the states near 12 cents. Outside the US the number moves a lot. India and Pakistan install at $0.80–1.50 per watt. Australia runs AUD 1.50–2.50. The UK and Germany run EUR 1.50–2.50, which is more than the US in dollars but still pays back faster on European retail rates. Above 15 kW the ground shifts again: three-phase connection, export limits and, in many jurisdictions, a different permitting track. A 15 kW residential system with storage is the usual next step for a house that has added both an EV and a heat pump, and it is the size at which battery economics start working on their own. Battery sizing A grid-tied array does not need a battery. It sells the surplus or banks it as credit. A battery earns its place when evening use is high, when export is paid badly, or when backup matters, and it should be sized from the evening or backup load rather than from the day generation. Nameplate capacity is not usable capacity. Three things separate them: LFP chemistry: 90–100% depth of discharge, 92–96% round-trip, 6,000–8,000 cycles Lead-acid: 50% depth of discharge, 80–85% round-trip, 500–1,200 cycles 10 kWh usable from LFP: about 10.5–11 kWh nominal 10 kWh usable from lead-acid: about 20–23 kWh nominal Installed battery cost in the US ran $700–1,300 per kWh in 2026, so a 13.5 kWh unit is $9,450 to $17,550 before any state incentive. The federal credit that used to apply to homeowner purchases is gone after 31 December 2025. The number that decides whether a battery pays is the export rate, not the battery price. California net billing pays roughly $0.06–0.10 per kWh for exports against a $0.3474 retail rate, so every kWh shifted from midday export to evening use is worth three to five times as much. New York and Florida still pay full retail for exports, and there a battery is mostly a backup purchase. Runtime under real load is worth reading before choosing a capacity, because the same kWh lasts very differently across households. For off-grid, size from daily load times autonomous days, divided by round-trip efficiency and depth of discharge. A 30 kWh a day home wanting three days of cover at 0.92 round-trip and 0.9 depth of discharge: 30 × 3 ÷ 0.92 ÷ 0.9 = 109 kWh. That is a serious battery, and it is why off-grid builds almost always start with load reduction. Domestic packs in the home energy storage range are almost all LFP now, which is why the depth-of-discharge figures above are the realistic ones rather than the marketing ones. If the array is going onto an integrated tile roof, how solar roof tiles work with storage covers the DC and AC coupling differences that come with that. Where sizing goes wrong Six lines account for most of the gap between a quote that fits and one that gets revised in month three. Sizing from the annual average in a house with a summer peak. A 30 kWh average with a 55 kWh August needs a different answer if self-consumption is the goal. Forgetting the EV. 15,000 miles is 4,500 kWh a year, which is 3.85 kW of extra array at the US median yield. Reading peak sun hours as daylight hours. London gets 16 hours of daylight in June and 2.93 peak sun hours across the year. Counting the whole roof. Usable plane is 25 to 40 percent less than total after setbacks, obstructions and access. Choosing the inverter before the ratio. A 1.0 design costs more and produces less than a 1.25 design on the same array. Treating nameplate battery capacity as usable. A 10 kWh lead-acid block delivers about 5 kWh. The fix for all six is the same: build the number from your own bill, your own location data and your own roof measurements, then check the result against a second source before signing anything. If you are sizing a roof and an array at the same time, the solar system archive covers the mounting, wiring and protection side, and our team will run the numbers against your own bill and your own roof measurements. Frequently Asked Questions How many solar panels do I need for a typical home? A home using 10 kWh a day at 4.5 peak sun hours needs 2.78 kW, which is seven 400 W panels on 14 m² of module area and about 20 m² of roof. At 3 peak sun hours the same home needs 4.17 kW, or eleven panels. The panel count follows from the peak sun hours, not from a national average. Is it better to oversize a solar system? A little headroom pays. Ten to twenty percent above the calculated size covers module degradation, a new appliance and a cloudy week. Beyond that, a grid-tied system exports the surplus at whatever the local rate is, and in California that is $0.06–0.10 against a $0.3474 retail rate. Extra panels make sense with a battery or under full-retail net metering, and much less sense without either. What size battery do I need to go off-grid? Size from daily load times autonomous days, divided by round-trip efficiency and depth of discharge. A 30 kWh a day home wanting three days of cover at 0.92 round-trip and 0.9 depth of discharge needs 109 kWh of nominal capacity. Most off-grid builds cut the load first, because the battery is the most expensive component per kWh delivered. Can I add a battery later? Usually yes, and it is cheaper if the original inverter was a hybrid. Adding storage to a string inverter means either a second inverter or an AC-coupled battery, and both cost more than choosing a hybrid unit at the start. Check that the battery voltage window matches the inverter before ordering. How do I know if my roof is big enough? Measure each unobstructed plane and allow 25 to 40 percent off for setbacks, chimneys, vents and access. A 5 kW array needs 32 to 38 m² of usable roof. If the plane is short, higher-wattage modules or an east-west split recover some of it, and both cost less than a ground mount. Does a solar tile roof need more area than panels? Yes, considerably. Flat photovoltaic tiles run at about 200 W per square metre of roof, and the curved three-arch profile at about 103 W/m² because of its shape. A 400 W framed panel is about 205 W/m² of module and roughly 160 W/m² once spacing is counted. A tile roof wins on appearance and weather sealing, not on watts per square metre, so compare on annual kWh per square metre of finished roof rather than on nameplate. The tiled BIPV roofing range lists the profiles with their per-square-metre output. L Luke Product Manager, GreenMore Manages the solar product line and works with distributors and installers on array sizing, roof integration and storage pairing for residential and light-commercial projects. Has worked on photovoltaic systems since 2017. Updated on November 17, 2026
  • What is a high frequency solar inverter?
    What is a high frequency solar inverter? Jul 30, 2024
    P Peter Lu Energy Storage Product Manager, GreenMore A high frequency solar inverter — sometimes called a high frequency link inverter or HF inverter — converts DC from solar panels or batteries into AC using high frequency switching (typically 20 kHz to 100 kHz) instead of a bulky 50 Hz / 60 Hz power transformer. The high frequency allows the transformer to be much smaller and lighter than a line frequency transformer of the same power rating, which is why most modern residential and small commercial inverters use this topology. How a high frequency inverter works The power stage has three main steps. 1. DC boost via PWM. The low voltage DC from the solar array or battery (for example 12 V, 24 V or 48 V) is chopped at high frequency using PWM (pulse width modulation). This produces a high frequency AC waveform at low voltage. 2. High frequency transformer. The high frequency AC is stepped up through a small, lightweight high frequency transformer to a high voltage DC bus, typically around 300 V to 400 V. Because the transformer operates at tens of kHz rather than 50 Hz, its core and windings are a fraction of the size of a line frequency transformer of the same power. 3. Inversion to grid frequency. The high voltage DC is then inverted by an IGBT or MOSFET bridge into a 50 Hz or 60 Hz pure sine wave AC output, filtered to meet grid or load requirements. Some designs skip the high frequency transformer entirely and use a transformerless topology, which pushes efficiency even higher but gives up galvanic isolation. Transformerless inverters are common in residential rooftop solar in markets where the grid code allows them. Where high frequency topology fits The compact size, light weight and high light-load efficiency of high frequency inverters make them the default choice for a few scenarios. Scenario Why high frequency fits Residential rooftop solar Space on the wall or in the garage is limited; weight matters; loads are mostly resistive or light electronic Small commercial distributed PV Higher efficiency at partial load improves annual yield per installed kW; compact units fit in electrical rooms Off-grid or hybrid systems with light loads Good efficiency at the low average loads typical of cabins, telecom shelters and small off-grid sites Retrofit projects with tight space Lighter, smaller units reduce structural and mounting constraints compared with heavy transformer-based replacements High frequency vs low frequency: the honest trade-offs Neither topology is universally better. The right choice depends on the load profile, the installation environment and the budget. High-frequency inverter power supply architecture and resonant rectifier topology Attribute High frequency (switching) Low frequency (transformer-based) Weight and size Lighter and more compact; the transformer operates at tens of kHz Heavier and bulkier due to the line frequency transformer Surge / inrush handling Typically limited to short, ms-scale peaks Strong; can sustain multi-cycle overloads Isolation Often transformerless; isolation depends on topology and design Galvanic isolation built into the main power path Peak efficiency Higher peak efficiency at light and medium loads Good but typically a few percentage points lower at light load Cost per kW Lower bill of materials for the same power rating Copper and steel make larger units relatively expensive Best fit Residential rooftop solar, space-constrained installs, light loads Inductive loads, off-grid, harsh environment, critical backup The efficiency numbers often quoted on brochures (for example "above 95%") are peak values measured under specific test conditions. Real-world efficiency depends on the load profile, the ambient temperature and the quality of the MPPT. Always compare efficiency curves across the load range your system will actually see, not just the headline peak number. What to check before choosing Load type. If your largest loads are resistive (lights, heaters, electronics), high frequency is usually fine. If you have motors, compressors or pumps that draw high starting current, check whether the inverter's surge rating covers them — a high frequency unit may not sustain the start. Output waveform. For any inductive load, insist on a pure sine wave output. Modified sine wave or square wave inverters can overheat motors and damage sensitive electronics. MPPT range. Verify that the inverter's MPPT voltage window covers your solar array's Vmp across the expected temperature range, not just at STC. Certification. For grid-tied or backup applications, verify that the inverter carries the certifications required in your market (CE, IEC, UL, local grid code approval). Communication. If you plan to integrate the inverter with a battery or EMS, check which protocols are supported (RS485, CAN, Modbus) and whether the battery BMS is on the compatible list. If you are sizing a residential or small commercial system, see our residential range on the home energy storage category page or the commercial range on the commercial energy storage category page. For project-specific questions, use the contact page. FAQ Is a high frequency inverter always more efficient than a low frequency one? At light and medium loads, usually yes. At heavy continuous loads close to the rated power, the gap narrows and a well-designed low frequency unit can be competitive. The fair comparison is the efficiency curve across your actual load profile, not the single peak number on the datasheet. Can a high frequency inverter start a motor or compressor? Some can, if they are sized with enough surge margin. But most high frequency inverters are not designed to sustain multi-cycle overloads the way a low frequency inverter can. If motor starting is a regular requirement, a low frequency topology is usually the safer choice. What is the difference between transformerless and high frequency with HF transformer? Both use high frequency switching. A transformerless design skips the isolation transformer entirely, which saves weight and cost and pushes efficiency higher, but the DC side is not galvanically isolated from the AC side. A high frequency inverter with an HF transformer keeps isolation while still being much smaller than a line frequency unit. The right choice depends on the grid code and the safety requirements of the installation. Do high frequency inverters work with solar batteries? Yes. In a solar plus storage system, the inverter sits between the battery and the AC loads or the grid. The key is to match the inverter's communication protocol and voltage window to the battery BMS, and to size the inverter power against the actual peak load, not just the battery capacity. P Written by Peter Lu — Energy Storage Product Manager, GreenMore Peter works with distributors, installers and system integrators to size residential and commercial storage around real load profiles. He focuses on LiFePO4 systems that match the inverter, grid rules and backup needs of each destination market.
  • What is a low frequency solar inverter ?
    What is a low frequency solar inverter ? Aug 24, 2023
    P Peter Lu Energy Storage Product Manager, GreenMore A low frequency solar inverter — also called a line frequency or transformer-based inverter — converts DC from solar panels or batteries into AC using a low frequency (50 Hz or 60 Hz) power transformer as the core voltage-conversion and isolation element. The transformer sits between the inverter bridge and the AC output, doing two jobs at once: stepping the voltage up or down, and providing galvanic isolation between the DC side and the AC side. This topology has been used in off-grid and industrial power systems for decades, and it still makes sense in a specific set of applications. How a low frequency inverter works The power stage is straightforward. A DC input is switched by a full-bridge (or push-pull) inverter circuit into low frequency AC, which then passes through a power frequency transformer for voltage transformation and isolation. A filter stage smooths the output into a clean sine wave, and a control module manages voltage regulation, overload protection and, in solar systems, MPPT on the DC input side. The transformer is the defining component. Its core is built from laminated silicon steel sheets, which keep hysteresis and eddy current losses manageable at 50 Hz / 60 Hz. Because the transformer operates at line frequency, it must be physically larger than the high frequency transformer used in a switching inverter of the same power rating. That size is the source of both the main advantage and the main trade-off of this topology. Where low frequency topology fits The heavy transformer gives low frequency inverters two properties that high frequency switching inverters struggle to match at the same price point: high surge current capability and galvanic isolation by design. That makes them a natural fit for a few specific scenarios. Scenario Why low frequency fits Inductive loads with high starting current (motors, compressors, pumps, refrigerators) Transformer-based topology handles 3× to 5× rated current surges for several cycles without tripping, which most high frequency inverters cannot sustain Off-grid or weak-grid installations (rural sites, islands, telecom towers) Galvanic isolation protects the DC side from AC-side faults and lightning-induced transients; the design tolerates dirty, unstable grids better than sensitive high frequency electronics Commercial and industrial backup with mixed loads Continuous overload capacity and simple, well-understood failure modes reduce unplanned downtime in critical loads Harsh environments (high temperature, high humidity, dust) Fewer high frequency switching stages means fewer failure points; thermal design is dominated by the transformer and large heatsinks rather than dense power electronics Low frequency vs high frequency: the honest trade-offs Neither topology is universally better. The right choice depends on the load profile, the installation environment and the budget. High-frequency inverter power stage circuit topology principle Attribute Low frequency (transformer-based) High frequency (switching) Weight and size Heavier and bulkier due to the line frequency transformer Lighter and more compact; the transformer operates at tens of kHz Surge / inrush handling Strong; can sustain multi-cycle overloads Typically limited to short, ms-scale peaks Isolation Galvanic isolation built into the main power path Often transformerless; isolation depends on topology and design Peak efficiency Good but typically a few percentage points lower than a well-designed high frequency unit at light load Higher peak efficiency at light and medium loads Cost per kW Copper and steel make larger units relatively expensive Lower bill of materials for the same power rating Best fit Inductive loads, off-grid, harsh environment, critical backup Residential rooftop solar, space-constrained installs, light loads What to check before choosing A few practical items matter more than the topology label on the brochure. Starting current of your largest load. A motor or compressor can draw 3× to 7× its rated current for a few cycles at start. Size the inverter's surge rating against that number, not against the steady-state load. AC input quality. If the grid or generator feeding the inverter is unstable, a low frequency unit with galvanic isolation will usually tolerate it better than a transformerless design. Weight and mounting. Low frequency inverters are heavy. Confirm that the wall or rack can support the unit and that the installation location is accessible for maintenance. Certification. For grid-tied or backup applications, verify that the inverter carries the certifications required in your market (CE, IEC, UL, local grid code approval). Efficiency at your actual load. Look at the efficiency curve, not just the peak number. A unit rated at 95% peak may only be 88% at the 20% load where it actually runs most of the time. If you are sizing a system around motor loads or off-grid backup, see our residential range on the home energy storage category page or the commercial range on the commercial energy storage category page. For project-specific questions, use the contact page. FAQ Is a low frequency inverter always better than a high frequency one? No. Low frequency topology wins on surge handling and isolation; high frequency wins on weight, size and light-load efficiency. For a typical rooftop solar system with light loads, high frequency is usually the better choice. For motor-heavy or off-grid systems, low frequency is often the safer choice. Why are low frequency inverters heavier? The line frequency transformer must be large enough to handle the full power at 50 Hz or 60 Hz. Transformer size scales inversely with frequency, so a 50 Hz transformer is much larger and heavier than a 20 kHz switching transformer of the same power rating. Can a low frequency inverter run a well pump or air conditioner? Yes, this is one of the scenarios where the topology is a strong fit. The key is to size the inverter for the starting current of the motor, not just the running current. A 3 kW motor may need a 6 kW to 9 kW inverter for a clean start. Do low frequency inverters work with solar batteries? Yes. In a solar plus storage system, the low frequency inverter sits between the battery bank and the AC loads. It converts battery DC into clean AC and, in off-grid mode, also manages the charging path from the solar array through the charge controller. P Written by Peter Lu — Energy Storage Product Manager, GreenMore Peter works with distributors, installers and system integrators to size residential and commercial storage around real load profiles. He focuses on LiFePO4 systems that match the inverter, grid rules and backup needs of each destination market.
  • How many kWh is 51.2V 200Ah?
    How many kWh is 51.2V 200Ah? Mar 26, 2025
    A 51.2V 200Ah battery stores 10.24 kWh of energy. You get there by multiplying voltage by ampere-hours and dividing by 1,000: 51.2 × 200 ÷ 1,000 = 10.24 kWh. That is one of the most common capacities in 48V-class home storage, and it is usually built from 16 LiFePO4 cells in series, since 16 × 3.2V = 51.2V nominal. The arithmetic is simple enough to do in your head. The parts worth slowing down for are the ones that decide whether 10.24 kWh is really the number you should be comparing: where the voltage and the ampere-hours each come from, how much of that figure you can actually draw, and what it will run in a real house. The calculation, step by step Why ampere-hours alone tell you nothing Ampere-hours describe how much charge a battery can move. Volts describe the pressure it moves that charge at. Neither number on its own says how much work the battery can do, which is why a 100Ah rating is meaningless until you know the voltage it belongs to. The gap is not academic. A 12.8V 100Ah battery holds 1.28 kWh. A 25.6V 100Ah battery holds 2.56 kWh. A 51.2V 100Ah battery holds 5.12 kWh. Same 100Ah label on all three, four times the energy between the smallest and the largest. Buyers who compare packs by Ah across different voltages are comparing the wrong number. The formula, with a worked example Multiply volts by ampere-hours to get watt-hours, then divide by 1,000 to get kilowatt-hours. Energy (kWh) = Voltage (V) × Capacity (Ah) ÷ 1,000 = 51.2 × 200 ÷ 1,000 = 10,240 Wh = 10.24 kWh Two things trip people up here. The first is the division by 1,000, which converts watt-hours to kilowatt-hours — skip it and you will be out by a factor of a thousand. The second is that 51.2V is a nominal figure, not a constant. The pack sits somewhere between about 40V and 58.4V depending on state of charge, so the energy it delivers is calculated from the nominal voltage, while the actual current at any moment depends on where in that range it happens to be. The same rule at other sizes Once the formula is fixed, every 51.2V pack converts to a predictable ladder. The rows below are the sizes that actually show up in residential quotes. Capacity at 51.2V Nominal energy Where it usually appears 50 Ah 2.56 kWh Small balcony or backup-only systems 100 Ah 5.12 kWh Entry module in a modular stack 150 Ah 7.68 kWh Evening self-consumption for a small home 200 Ah 10.24 kWh The mainstream residential size 280 Ah 14.34 kWh Larger homes, heat pump loads 300 Ah 15.36 kWh Three-module stacks and mid-size ESS 400 Ah 20.48 kWh Four-module stacks, small commercial Where 51.2V and 200Ah come from Voltage is a cell count A lithium iron phosphate cell has a nominal voltage of about 3.2V. Stack 16 of them in series and the voltages add: 16 × 3.2V = 51.2V. That is why 51.2V packs are so common — they are the natural result of a 16-cell series string, and 16 happens to land neatly on the 48V-class inverter input window that the industry standardised on years ago. Change the cell count and you change the class. Fifteen cells give 48V nominal and are sometimes used for legacy 48V equipment; seventeen give 54.4V. Both exist, but neither is as common as 16S in home storage, and mixing them with a 51.2V inverter usually means a firmware workaround rather than a clean fit. Capacity is a parallel count The 200Ah figure comes from the cell capacity and how many strings run in parallel. A pack built from 200Ah cells needs one parallel string: 16 cells, 16S1P, 51.2V and 200Ah. A pack built from 100Ah cells — the format most Chinese cell makers ship in volume, and the one GreenMore uses — needs two strings to reach 200Ah: 16S2P, 32 cells in total, still 51.2V and 200Ah. The energy is identical either way. What changes is the internal layout, the cell count, and often the weight and the price. It is worth knowing which one you are being quoted, because a 16S2P pack has twice as many connections to get right and a different failure profile from a 16S1P pack built from large-format cells. Nominal versus usable capacity What gets subtracted The 10.24 kWh figure is nominal energy. In real use you cannot draw all of it. Most systems hold back a reserve at the bottom of the range to keep the cells out of deep discharge, and some energy is lost in the battery and again in the inverter during conversion, so what reaches your loads is lower than what the label says. How much lower depends on the depth-of-discharge limit the manufacturer sets. At 90% depth of discharge, a 10.24 kWh pack offers about 9.2 kWh before conversion losses. At 80%, the same pack offers about 8.2 kWh. That spread of roughly one whole kilowatt-hour is bigger than the difference between many competing quotes, which is why a nominal figure alone cannot settle a comparison. Why two "10 kWh" batteries differ Two units both advertised at 10 kWh can deliver noticeably different energy to the wall. Depth of discharge is the largest factor, round-trip efficiency is the second, and the inverter's own standby draw is the third — a hybrid inverter left on overnight can consume a few tens of watts continuously, which adds up over a long backup event. When you compare, ask for usable kWh, round-trip efficiency and the depth-of-discharge limit as three separate numbers. A supplier who can only give you the nominal figure has not answered the question. Capacity on the label translates into hours of home use — the larger the pack, the longer the backup. How far 10.24 kWh goes in a home How far 10.24 kWh goes depends entirely on the household. As a rough guide, it can carry efficient essentials — lighting, a refrigerator, communications and some electronics — through an evening or an outage, while heavy heating, cooling and electric cooking will draw it down quickly. Average daily consumption varies widely by country and by home, so the most reliable estimate comes from your own electricity bills rather than a generic figure. A worked backup estimate Take a two- to three-person household in a temperate climate and list only the loads worth backing up. The figures below are typical daily energy, not instantaneous draw. Backed-up load Typical daily energy Basis Refrigerator 1.5 kWh Modern 300 L unit, compressor cycling included Lighting, 10 LED lamps 0.5 kWh About 5 hours per evening Router and ONT 0.2 kWh Continuous, 24 hours Television 0.6 kWh About 4 hours Total 2.8 kWh Per day Runtime at 8.2 kWh usable About 2.9 days No solar recharge, 80% depth of discharge Treat those numbers as a starting point rather than a promise. A hot climate moves the refrigerator figure up, a cold one moves it down, and adding a heat pump or an EV charger changes the calculation completely. The point of the exercise is that 10.24 kWh covers days of essentials or hours of heavy loads — the same battery, two very different answers, depending on what you put behind it. When one unit is not enough If the load list above looks like your house, one 200Ah unit is a sensible starting point. If it does not, the table below shows what parallel units buy you. Configuration Nominal energy Typical role 1 × 51.2V 200Ah 10.24 kWh Evening self-consumption and essential backup for a small to medium home 2 units in parallel 20.48 kWh Larger evening load and longer backup runtime 3 units in parallel 30.72 kWh Higher-consuming home or extended off-grid use with solar recharge Why capacity is given in kWh It matches the bill. Electricity is charged in kWh, so a stored kWh figure tells you directly how much evening or peak-tariff energy the battery can replace. It matches the inverter. Inverter and array sizing is done in kW and kWh, so the battery's energy figure is the one that shows whether the system is balanced. It is what the nameplate carries. Rated voltage and rated capacity have to be declared for certification, and the resulting energy figure is what appears on the label — a 51.2V pack is listed by its kWh rating, not only by its Ah. That is the number an installer or an inspector reads first. Expanding with parallel modules Modular 48V-class batteries can often be added in parallel to raise total energy while keeping the same voltage, which lets you start with one unit and add more later. The number of modules supported and the cabling rules depend on the model and its BMS, so check the specification rather than assuming any number can be connected. Adding modules in stages is usually more practical than buying the full capacity up front, provided the inverter and wiring are rated for the final size from the start. What parallel expansion does to current Energy adds up in parallel; current does too. A 10.24 kWh pack running a 5 kW inverter draws roughly 100 A at 51.2V. Put three of them in parallel and the pack can deliver three times that, which is why the busbar, the DC breaker and the cable gauge have to be sized for the final configuration rather than for the first module. Sizing the DC side for 10 kWh and then adding two more units is the most common installation mistake in this class, and it is an expensive one to fix after the wall is closed. The other half of the equation is the discharge rate. A 10.24 kWh pack on a 5 kW inverter works at about 0.5C, which is a comfortable continuous rate for LiFePO4. The same inverter on a 20.48 kWh pack runs at about 0.25C, which is gentler on the cells and usually extends service life. If you expect to expand, a larger pack with the same inverter is not wasted capacity — it is a lower-stress duty cycle. Work with GreenMore GreenMore supplies home energy storage batteries built around 51.2V LiFePO4 modules with a BMS for monitoring, balancing and protection, together with compatible inverters. The stacked series runs from a single 5.12 kWh module to 20.48 kWh across four modules, and the wall-mounted 5 kWh unit covers the fixed-requirement case. We size the number of modules against your actual usage and backup needs rather than quoting a fixed package. Browse the home storage range, or send us your recent electricity bill through the contact page and we will come back with a module count and a runtime estimate for your loads. Frequently asked questions Is 51.2V the same as a 48V battery? Essentially, yes. A 16-cell LiFePO4 battery has a nominal voltage of 51.2V but is commonly called a 48V system because it works with 48V inverters. Its actual voltage moves with state of charge, from roughly 40V when empty to about 58.4V fully charged. The 48V label is a system class, not a measurement. Is the full 10.24 kWh available to use? Not always. The figure is nominal. Systems typically reserve some energy at the bottom to protect the cells, and conversion through the inverter causes small losses. At 90% depth of discharge you get about 9.2 kWh before losses; at 80%, about 8.2 kWh. Check the stated usable capacity when comparing models, and check it against the same depth-of-discharge basis. Is 100Ah always the same amount of energy? No, and this is the most expensive misunderstanding in the category. At 12.8V, 100Ah is 1.28 kWh. At 25.6V it is 2.56 kWh. At 51.2V it is 5.12 kWh. The Ah figure only becomes energy once you know the voltage, so always convert to kWh before comparing two packs. Can I add more batteries later? Usually, if the model supports parallel expansion and the inverter and cabling are rated for the larger size. Confirm the maximum number of modules and the installation requirements before you buy, so the system can grow without being replaced. If the DC side is sized for the final capacity at the start, adding a module later is a short job. How many 10.24 kWh units does a home need? That depends on daily usage and backup goals. Many homes start with one unit for evening self-consumption and essential backup, adding a second or third if they use more power or want longer runtime. Your electricity bills give the most accurate starting point — a daily figure in kWh, divided by about 8 usable kWh per unit, is a reasonable first approximation. Does a 200Ah battery need a special charger? It needs a charger, or an inverter, whose charge profile matches LiFePO4 and whose current rating suits the pack size. A 10.24 kWh pack is usually charged at somewhere between 0.2C and 0.5C, which is 40 A to 100 A at 51.2V. Charging well below that is safe but slow; charging above it shortens cell life. Check the datasheet for the maximum continuous charge current rather than relying on the inverter's default. Updated on November 12, 2026 P Peter Lu Energy Storage Product Manager, GreenMore Peter helps homeowners and installers convert battery voltage and Ah figures into usable kWh, size parallel modules against real load profiles, and understand the difference between nominal and usable capacity.

Subscribe To Our Newsletter

Stay up to date with GreenMore. We share new BIPV solar tile releases, product and certification updates, and field notes from our projects — no fluff, just the things that matter to installers, distributors, and system integrators. If there's something you'd like to see covered, drop us a line — we read every reply.

Copyright @ 2026 GreenMore All Rights Reserved. Network Supported

Sitemap / Blog / Xml / Privacy Policy

leave a message

leave a message
If you are interested in our products and want to know more details,please leave a message here,we will reply you as soon as we can.
submit

Home

Products

WhatsApp

contact