Where Do Solar-to-Battery Losses Go? A Practical LiFePO4 Energy Ledger

When a solar-battery owner sees less energy at the load than the PV array produced, the missing kWh have not necessarily disappeared at one single component. Some energy was never available after site conditions, some may have been limited by the controller or inverter, some crossed the battery charge-and-discharge boundary, and some was used by the conversion equipment or the loads themselves.

This guide answers one practical question: where do solar-to-battery losses occur, and how can you measure them without double-counting the same loss? It is an energy-ledger guide, not a promise of one universal system-efficiency percentage. Real results depend on meter locations, operating mode, temperature, state of charge, reserve settings, load profile and the exact PV, inverter, battery and wiring configuration.

Quick Answer: Solar-to-Battery Losses Are a Chain of Boundaries

Start by naming the two points you are comparing. “Panel output” might mean a modeled DC estimate, a PV inverter reading or energy after an MPPT stage. “Battery output” might mean DC energy at the battery terminals, inverter AC output or energy measured at a household load. Those are different boundaries.

The practical map is:

The basic accounting identity is simple: energy loss within a defined boundary equals energy entering that boundary minus energy leaving it. The difficult part is defining the boundary and making sure the two readings cover the same time interval and operating state.

What Does “Solar-to-Battery Losses” Mean?

The phrase can describe several different comparisons. Before calculating anything, write down a sentence such as:

> “I am comparing PV energy measured at the inverter DC input with AC energy delivered to the household loads during the same discharge cycle, excluding grid charging.”

That sentence is more useful than a bare claim such as “the system is 90% efficient.” It tells you whether the calculation includes PV-side conditions, conversion, battery storage, inverter output, standby consumption, reserve energy and direct PV-to-load operation.

The U.S. Department of Energy notes that energy storage is not 100% efficient and that solar output changes with season, time of day, clouds, dust, haze, shadows, rain, snow and dirt. Those effects belong in the system context before a battery-specific conclusion is made. See the DOE overview of solar integration, storage and system basics for the wider operating picture.

Start with an Energy Ledger, Not One Efficiency Percentage

Use separate entries for separate boundaries. A useful ledger may contain the following values for the same interval:

Do not use these symbols as if they were automatically available from every product. Some systems report only AC-side energy; some combine PV production and load flow; some estimate state of charge; and some counters reset or use different time bases. The purpose of the ledger is to expose what is known and what is still unknown.

The NREL PVWatts Version 5 manual is a useful reference for separating modeled PV-system loss categories such as soiling, shading, snow, mismatch, wiring, connections, nameplate rating, age and availability from the inverter model. A modeled derate is an assumption for an estimate; it is not a measurement of a particular AmpBird system.

Stage 1: PV Array to the Controller or Inverter

The first stage is everything that happens before the selected conversion input. It is often described as “solar loss,” but it includes both physical reduction and energy that was available in principle but could not be accepted by the system at that time.

Site and array conditions

Common factors include:

• Soiling, dust, snow or other surface obstruction.
• Shade from buildings, trees, poles, roof structures or seasonal changes.
• Cell temperature and the resulting change in operating voltage and power.
• Module mismatch, degradation or an array design that does not operate at the expected point.
• DC cable length, conductor sizing, terminations and connection condition.
• Inverter or controller clipping, curtailment or a battery that is already at its permitted charge state.

These factors are not interchangeable. A shaded array has a different corrective action from an undersized conductor. A battery that is full may cause charging power to fall even though the sun is still available; that is not the same as heat dissipated in a cable.

Why PV production data needs context

A daily PV total without time stamps cannot explain whether energy was lost, shifted to a direct load, curtailed after the battery reached its target SOC or simply produced during a period when the household demand was low. Use interval data when possible, and record weather, shade, cleaning and operating mode.

For a planning estimate, the PVWatts manual explains that modeled loss categories combine through a multiplicative treatment rather than a simple addition of every percentage. That is why a quotation that lists several derates should not be interpreted by adding them mechanically, and why an estimate should not be presented as a field measurement.

Stage 2: Controller or Inverter Conversion and Acceptance Limits

After the PV energy reaches the conversion equipment, the system still has to accept and transform it. Depending on architecture, the device may perform MPPT, DC-DC conversion, DC-AC conversion, battery charging control, load sharing, grid synchronization or multiple functions at once.

Conversion is not the same as curtailment

Conversion loss is energy consumed or dissipated while changing voltage, current or waveform. Curtailment is energy that the system does not accept or deliver because of a control limit, full battery, grid rule, thermal condition, load condition or another operating decision. Both can reduce the energy that reaches a later boundary, but they require different evidence and remedies.

For example, adding more PV capacity may improve morning and late-afternoon production when the existing array is below the inverter's acceptance range. It may not increase midday battery charging when the controller, inverter, battery or export limit is already the lowest limit in the chain. A bigger array cannot remove a downstream acceptance limit by itself.

Standby and auxiliary energy

Some systems draw energy while monitoring, communicating, cooling, balancing or remaining ready for backup. Whether that energy appears as a separate loss depends on the meter boundary. If an AC-side meter already includes the equipment's own consumption, subtracting it again would double-count the loss.

This is one reason the answer to “what is the inverter efficiency?” is not enough to calculate usable household energy. You need to know whether the published number is measured at a particular load, voltage, temperature and direction, and whether standby or auxiliary power is included.

Stage 3: Battery Charge, Storage and Discharge

The battery boundary is often where a simple PV-to-load comparison gets mislabeled. A battery can have a charge path, a stored-energy state, a discharge path and an operating reserve. The energy entering the battery is not automatically the energy that can later leave it.

Use round-trip efficiency only with a stated boundary

Round-trip efficiency is commonly represented as:

RTE = E_discharge / E_charge

The numerator and denominator must be measured for the same defined battery cycle. The NREL definition of battery round-trip efficiency describes it as energy retrieved divided by energy input. NREL's Fundamentals of Energy Storage also explains that round-trip efficiency reflects energy output relative to input and includes losses such as heat.

The result is not automatically the whole solar-to-AC-load efficiency. A DC battery RTE may exclude PV-side loss, charge-controller conversion, inverter standby, AC wiring, reserve energy and load-side measurement. An AC-to-AC test may include more of the system, but only if its input and output boundaries are documented.

Capacity, SOC and usable energy are different

Nameplate kWh, usable energy and energy delivered to a load answer different questions. Usable energy depends on the permitted SOC window, reserve setting, temperature, current, cutoffs, battery condition and system controls. A battery may still contain energy while the system refuses to discharge it because a reserve or protection limit has been reached.

This is why the home battery kWh sizing guide and the 16kWh runtime guide should be read as planning frameworks, not as promises that every rated kWh will appear at every AC appliance.

Stage 4: Battery or Inverter Output to the AC Load

The last stage can include DC-to-AC conversion, inverter idle consumption, AC cable losses, protection devices, subpanel distribution and the chosen load meter. The result may be lower than battery-side discharge even when the battery itself is operating normally.

The load boundary also matters. Whole-home energy, a backup subpanel and one appliance may each have a different meter. A system can show a healthy battery discharge value while the selected appliance receives less energy because other loads, conversion equipment or distribution paths are included in the comparison.

Do not treat a short high-power burst and a long low-power discharge as equivalent tests. Conversion efficiency and auxiliary consumption can change with operating point. Record the load shape, not only the daily total.

Direct PV-to-Load and PV-to-Battery-to-Load Are Different Paths

A hybrid system may serve a load directly from PV while also charging the battery. Another operating period may send PV through the battery before it reaches the same load. The first path can avoid a charge-and-discharge cycle; the second path may provide time shifting or backup value but crosses more boundaries.

Use a path table when reviewing a system:

Without this path distinction, a daily report can combine direct PV consumption, battery cycling, grid charging and export into one number that does not answer the buyer's question.

A Simple Calculation with Explicit Assumptions

Suppose an illustrative test records 10.0 kWh at a defined PV-side input boundary. Assume, only for this example, that the non-overlapping charge-and-discharge path retains 95% of that energy and the later inverter/AC delivery path retains 94%.

Estimated AC load energy = 10.0 kWh × 0.95 × 0.94 = 8.93 kWh

The illustrative difference is 1.07 kWh. It is not a universal loss allowance and it does not say where the difference occurred. To allocate it, the test would need additional measurements: PV-side input, battery charge input, battery discharge output and AC load output, all with stated boundaries and the same relevant operating period.

If the 95% already includes the inverter path, multiplying by 94% would count the same conversion twice. If the first 10.0 kWh is a PVWatts modeled value and the last reading is a physical AC meter, the result mixes an estimate with a measurement. That may be useful for a preliminary scenario, but it is not a measured system efficiency.

Why Loss Percentages Should Not Be Added Blindly

Consider three non-overlapping retention factors of 0.96, 0.95 and 0.94. The combined retention is:

0.96 × 0.95 × 0.94 = 0.857

That is about 85.7% retention, not a shortcut created by treating each factor as an independent additive loss without stating the boundary. In real systems, the factors may also vary by current, temperature, SOC and operating mode, so a single product of brochure numbers may still be a rough planning model.

The important rule is not a specific percentage. It is to use factors only when their boundaries do not overlap and to label every value as modeled, specified, tested or measured. The NREL PVWatts documentation is helpful here because it separates loss categories and treats the combined modeled loss as a system calculation rather than a casual sum.

How to Measure a Real System without Double-Counting

Use a repeatable test window. A short checklist is more valuable than a long list of unverified efficiency claims.

1. Freeze the comparison boundary

Write the input and output points in plain language. For example: “PV inverter DC input to backup-subpanel AC meter, 09:00–21:00, solar-only charging, no grid charge.” If the system changes mode, split the interval.

2. Align timestamps and units

Use the same time zone, interval length and energy unit. A counter that reports local time and another that reports UTC can create a false loss or gain around midnight. Confirm whether each reading is cumulative, interval-based, import-only, export-only or netted.

3. Record operating state

At minimum, record battery SOC at the start and end, reserve setting, charge/discharge power, battery temperature if available, inverter mode, grid state, PV conditions and significant load changes. A result with no operating-state notes is hard to reproduce.

4. Keep direct PV and battery paths separate

If the system reports PV-to-load, PV-to-battery, battery-to-load and grid-to-load flows, preserve those categories. Do not use a whole-home daily total as if every kWh followed the same path.

5. Reconcile a boundary before assigning a cause

If energy entering a boundary is 12.0 kWh and energy leaving is 10.8 kWh, record a 1.2 kWh difference first. Assigning that difference to battery cells, inverter heat or wiring requires a narrower test and additional measurement points.

Practical measurement table

What a Larger Solar Array Can and Cannot Fix

A larger PV array can increase the energy available during weak-sun hours, improve the chance of reaching a desired daily energy target and reduce the time spent below a downstream power threshold. It cannot automatically remove conversion losses, improve a poor connection, change the battery's permitted charge current or make a full battery accept unlimited power.

Use the existing solar-panel sizing guide when the question is “how much PV should I install?” Use the 16kWh solar charging-time guide when the question is “how long might charging take?” This article owns a different question: “where did the energy go between my chosen input and output boundaries, and how can I test it?”

If the battery charge rate, inverter input limit or battery reserve is the lowest limit, a larger array may produce more potential energy without producing more stored energy at that moment. A buyer should ask for the complete chain of limits rather than comparing only PV nameplate power.

How the Ledger Changes a Product or Configuration Request

When you request a quotation, include the information needed to define the energy path:

• Daily energy demand and important peak-load periods.
• Whether the system is grid-connected, backup-only or off-grid.
• PV module or array power, expected orientation and known shade conditions.
• Inverter model or required AC output, including charging and export limits.
• Battery chemistry, nominal voltage, capacity target, usable-energy target and reserve requirement.
• Whether the comparison is DC-side, AC-side, solar-only, grid-inclusive or whole-home.
• Required evidence: data-sheet conditions, test records, meter locations and acceptance criteria.

For a 12V, 24V or 48V DIY route, the LiFePO4 cell collection and DIY battery kit collection are starting points for product research. A complete home-storage route can be reviewed through the home battery systems collection. Product selection still needs to be matched to the inverter, BMS, enclosure, installation and operating requirements; a collection page is not a substitute for that system check.

Common Mistakes in Solar-Battery Loss Calculations

Mistake 1: Calling every difference “battery loss”

If PV energy is measured before array-side effects and AC load energy is measured after several conversion stages, the difference belongs to the entire comparison boundary. It cannot be assigned to the cells without battery-side measurements.

Mistake 2: Mixing modeled and measured data

A modeled daily PV estimate can help with planning. It should not be presented beside a measured AC total as if both were direct readings from the same system boundary.

Mistake 3: Treating rated capacity as delivered energy

Rated capacity does not state the SOC window, reserve, cutoff, temperature, current or downstream conversion path. Use usable-energy and runtime assumptions that state their conditions.

Mistake 4: Adding overlapping efficiency numbers

An AC-to-AC efficiency may already include battery, charger and inverter effects. Adding a separate battery RTE and inverter efficiency without checking the test boundary can count the same stage twice.

Mistake 5: Ignoring the time interval

Comparing a sunny midday PV counter with a full-day load counter will produce a meaningless difference. Align the interval and separate direct PV, battery and grid paths.

Mistake 6: Assuming a larger array fixes a downstream power limit

PV energy, battery charge power, inverter conversion power and load demand are related but not identical. The smallest active limit can control the result during a given interval.

FAQ: Solar-to-Battery Losses and Efficiency

Does a solar battery always lose the same percentage of energy?

No. Losses can change with power, temperature, SOC, operating mode, standby time, battery condition, PV conditions and the selected measurement boundary. A controlled test may produce a useful result for one operating range, not a universal percentage for every day.

Is battery round-trip efficiency the same as solar-to-AC-load efficiency?

No. Battery RTE normally describes energy retrieved from a defined battery boundary divided by energy accepted at that boundary. Solar-to-AC-load efficiency may also include PV-side conditions, charge conversion, inverter conversion, standby, wiring and load-side measurement.

Can 20 kWh of PV production deliver 20 kWh to my home?

Not automatically. Some energy may be used directly, stored, curtailed, exported or consumed by conversion equipment. To answer for a real system, define the PV and load meters, time interval and operating paths.

Does a larger solar array remove conversion losses?

No. It can provide more available energy and may improve production in weak-sun hours, but it does not remove losses in wiring, converters, inverters or the battery. It also cannot override a downstream charge or export limit.

Should I use the efficiency number from a product data sheet?

Use it as one input, but check its test conditions and boundary. Ask whether the value is DC-to-DC, DC-to-AC, AC-to-AC, peak or weighted, and whether standby, temperature and auxiliary consumption are included.

Why does my battery dashboard show more energy in than I can use?

The dashboard may be reporting a different boundary, a nameplate or estimated SOC value, or energy that includes charging losses and reserve. Compare battery-side charge, battery-side discharge and AC load readings over an aligned interval.

Does a full battery mean all additional solar energy is lost?

Not necessarily. Depending on system architecture, additional PV may serve a load directly, export to the grid or be curtailed. The correct conclusion requires the system's flow records and export/curtailment behavior.

Can I calculate losses from state-of-charge percentages alone?

Usually not with enough confidence for a procurement decision. SOC is often an estimate and may use a battery-specific calibration. It can support a test record, but it should be paired with measured energy and a stated SOC window.

What is the best first meter to add?

The best first meter is the one that separates the most important unknown boundary in your system. For some designs that is PV input versus inverter output; for others it is battery discharge versus backup-panel load. Choose based on the question you need to answer, not on a generic meter list.

How should I compare two battery quotations?

Ask both suppliers to state nominal capacity, usable-energy assumptions, SOC window, charge/discharge limits, test conditions, round-trip boundary, auxiliary consumption and the measurement point used for each efficiency claim. Then compare like with like and keep unknowns visible.

Final Checklist

Before accepting a solar-battery loss claim, confirm:

• The input and output meter locations are named.
• The time interval, units and time zone match.
• Direct PV, battery, grid and export paths are separated.
• The battery SOC window, reserve and operating mode are recorded.
• Modeled assumptions are labeled separately from measured results.
• Each efficiency factor has a non-overlapping boundary.
• The calculation states whether it is DC-side, AC-side or AC-to-AC.
• The result is tied to the actual PV, inverter, battery, wiring and load configuration.

If you are deciding between loose cells, a DIY kit or a complete home-storage system, send the PV size, inverter model, daily load, peak load, target backup hours and intended measurement boundary to AmpBird's contact page. The more clearly the energy path is defined, the more useful a product recommendation and quotation can be.

Technical references

• NREL PVWatts Version 5 Manual - modeled PV-system loss categories and inverter treatment.
• NREL HPXML: Battery RoundTripEfficiency - definition of retrieved energy divided by input energy.
• NREL Fundamentals of Energy Storage - energy-storage efficiency and loss context.
• U.S. Department of Energy: Solar Integration, Solar Energy and Storage Basics - storage efficiency and conditions affecting solar output.