How to Monitor Solar Battery Health: SoH, Degradation & Warranty Claims

You wouldn't drive a car specifically without an odometer. You wouldn't buy a used iPhone without checking the "Battery Health" percentage. Yet, thousands of homeowners install $15,000 solar batteries and have no idea how much "life" is left in them.

Batteries are consumable assets. Like the tires on your car, they wear out physically every time you use them. This is called Cell Degradation. Most monitoring apps (Tesla, Enphase, SolarEdge) show you the "State of Charge" (Fuel Gauge), but they intentionally hide the "State of Health" (Engine Condition).

Why? Because manufacturers don't want you to panic when you see 98% capacity after one month. They assume you don't understand the chemistry.

This guide pulls back the curtain. We will explain exactly how to find your battery's true health metrics, how to perform a manual capacity test that stands up in court, and how to know if you are eligible for a free warranty replacement.

Part 1: The Two Metrics You Must Know

To monitor health, you must stop looking at the percentage on your home screen. That is just the fuel gauge.

1. State of Charge (SoC)

• Definition: How full the tank is right now.
• Range: 0% to 100%.
• What it tells you: "Can I run my microwave tonight?"
• Analogy: The gas needle in your car.
• Deceptive Factor: A battery with 50% health can still be charged to "100% SoC." It just holds half as much energy.

2. State of Health (SoH)

• Definition: The size of the tank compared to when it was new.
• Range: 100% (New) -> 70% (Warranty Replacement) -> 0% (Dead).
• What it tells you: "Did I get what I paid for?"
• Analogy: The odometer or total engine wear.

The Scary Math: A 10 kWh battery with 80% SoH acts like an 8 kWh battery. It still charges to "100%," but that "100%" represents significantly less electrons.

Part 2: Expected Degradation (The "Normal" Curve)

Before you panic, you need to know what is normal. All lithium batteries degrade. It is a chemical reality of ions getting trapped in the anode (SEI Layer growth).

The "Bathtub" Curve

Battery degradation is not a straight line. It follows a predictable curve.

1. The Drop (Year 1): You might lose 2–3% capacity in the first 6 months as the "Solid Electrolyte Interphase" (SEI) layer forms. This is healthy and expected.
2. The Plateau (Years 2–10): Degradation slows to a crawl (maybe 1% per year). This is the "sweet spot" of the battery's life.
3. The Knee (End of Life): Eventually, internal resistance spikes, and capacity plummets. This usually happens after 15+ years.

Degradation by Chemistry (2026 Data)

Note: These are engineering estimates based on cycle life data. Your mileage may vary based on heat and usage.

Part 3: How to Find Your SoH (Brand by Brand)

Manufacturers hide this data deep in the menus. Here is how to find it.

Tesla Powerwall

Tesla does not show SoH in the basic consumer settings.

• Method A (The App): Go to Settings > My Home Info. Sometimes it lists "Energy Capacity." If it says 13.5 kWh, it's just the spec sheet. If it says 12.8 kWh, that's live data.
• Method B (The Gateway): If you connect to the Gateway's local WiFi (TEG-xxx) and log in as "Customer" (email on back of unit), you can see the raw diagnostic strings. Look for nominal_full_pack_energy.
• Method C (Export Data): Download your CSV data. Sum up the "Discharge Energy" column for a full year. If it is dropping faster than 2%, you have a problem.

Enphase (Enlighten)

Enphase provides excellent granularity if you know where to look.

• Method: Open Enlighten App > Menu > System > Devices > Battery.
• Detail: It lists the "Charge Cycles" count.
• Secret: Enphase batteries are modular. If your capacity drops by exactly 33%, it means one of the three microinverters inside the unit has failed, not the chemistry.

SolarEdge (SetApp)

SolarEdge is transparent.

• Method: Log into the Monitoring Platform (web desktop version). Go to Layout > Right-click the battery icon > Information. It usually lists "SOH: 94%" explicitly.

Part 4: The DIY Capacity Test (The Only Truth)

Do not trust the software. Software lies (or estimates). To know the truth, you must physically measure the energy leaving the box.

Warning: This takes 24 hours and requires you to manipulate your breakers.

Step 1: Force Charge to 100%

• Set your "Backup Reserve" to 100%.
• Allow the battery to charge fully from solar or grid until it stops taking power.
• Wait 2 hours for the cells to "balance" (top off).

Step 2: The Discharge

• Wait for the sun to go down (so solar doesn't interfere).
• Go to your breaker panel. Turn OFF the solar PV breaker.
• Set your "Backup Reserve" to 0% (Self-Powered Mode).
• Turn on high loads (AC, Dryer, Oven) to create a steady draw (e.g., 3-5 kW).
• Watch the app until the battery hits 0% and stops discharging.

Step 3: The Math

• Open the app's "Energy impact" graph.
• Look at the "Discharged" total for that specific time period.
• Formula: (Discharged kWh) / (Nameplate kWh) = SoH.
• Example: You discharged 11.2 kWh from a 13.5 kWh Powerwall.
  • 11.2 / 13.5 = 0.829 (83% SoH).

Part 5: When to File a Warranty Claim

Most warranties have a "70% retention guarantee" for 10 years. If your DIY test shows 68%, you theoretically have a claim.

However, read the fine print.

1. Throughput Clause: Some warranties expire after a certain amount of energy (e.g., "37 MWh of throughput") regardless of years. If you participate in VPPs every day, you might hit the throughput limit in Year 8.
2. Temperature Clause: If the logs show the battery was exposed to >122°F (ambient temperature), they can deny the claim. This is common for batteries installed on south-facing walls in Arizona.
3. Connection Clause: If the battery was offline (no internet) for >6 months, the warranty is often voided because they couldn't push firmware updates to protect the cells.

The Script: How to Talk to Support

When you call Tesla or Enphase, do not say "It feels weak." Say this: "I performed a controlled capacity test on [Date]. From 100% to 0% SoC, the unit only discharged 8.4 kWh. This represents 62% of rated capacity, which is below the 70% threshold in my Warranty Agreement. Please open a Tier 2 ticket for a log review."

Glossary of Terms

• Cycles: One full discharge from 100% to 0% (or equivalent, e.g., two 50% discharges).
• BMS (Battery Management System): The computer inside the battery that balances voltage.
• Internal Resistance (IR): The friction inside the battery to electron flow. High IR = Heat = Degradation.
• Depth of Discharge (DoD): How much of the battery you use. Using only the middle 50% makes it last longer.

Frequently Asked Questions (FAQ)

Reference: Voltage vs SoC Table (48V System)

If you are using a voltmeter, use this table to estimate your battery's charge. Note the difference between NMC (Tesla Powerwall 2) and LFP (Powerwall 3 / Enphase).

• Warning: LFP batteries have a very flat voltage curve. It is almost impossible to guess SoC between 30% and 70% using voltage alone. You must rely on the BMS "Coulomb Counting."

Related Articles

• Tesla Powerwall 3 vs Enphase 5P
• Lithium Iron Phosphate vs NMC
• How to Maintain Your Battery

Summary: Trust But Verify

Your solar battery is an expensive asset. Don't ignore it for 10 years.

• Check logs monthly.
• Do a capacity test annually.
• Keep it cool.

View Top Rated Batteries for Reliability →

Engineering Reality

Battery health monitoring is more nuanced than the monitoring app's simple "System OK" status light suggests. Understanding what the metrics mean — and what they don't capture — is essential for proactive health management.

State of Health (SoH) as reported by the BMS is an estimate, not a measurement. The SoH percentage shown in a monitoring app is calculated by the BMS using cell voltage interpolation and coulombic counting algorithms — not by directly measuring actual capacity. These algorithms assume cells age uniformly, which is not always the case. Cell imbalance within a battery module (where one cell group degrades faster than others) can cause the BMS to report optimistic SoH while actual usable capacity is lower. The only way to confirm actual usable capacity is a controlled discharge test: charge to 95%, discharge to 10% at a known, consistent power level, and measure total kWh delivered. This provides ground-truth data that the monitoring app's SoH calculation cannot.

Cycle count reporting across manufacturers uses different definitions. Tesla Powerwall cycle count is defined using the "equivalent full cycle" (EFC) methodology — charging and discharging 13.5 kWh counts as one cycle regardless of whether it occurred in one event or across multiple partial cycles. GivEnergy reports total charge throughput in kWh. BYD reports a cycle count based on the BMS's internal tally of charge/discharge events. These different methodologies make direct cycle count comparisons between brands meaningless and complicate warranty claim assessment. When monitoring cycle count for warranty purposes, confirm the methodology used by your specific manufacturer.

Monitoring app outages do not mean battery faults. Most battery monitoring apps use cloud-based data aggregation — the battery streams data to a cloud server, which then serves it to the app. A monitoring app that shows no data or an "offline" status may reflect a cloud server issue, a router connection drop, or a broadband outage rather than a battery fault. Before escalating "no data" alerts to an installer, confirm: (1) the monitoring device (Zigbee gateway, ethernet connection) is powered; (2) the broadband router is operational; (3) the app has not been logged out by a software update. True battery faults are indicated by physical fault lights on the inverter or BMS, not by app disconnections.

Round-trip efficiency decline is a leading indicator of accelerated degradation. The round-trip efficiency of a healthy LFP battery system (including inverter losses) should be 90–95% in Year 1. As cells age, internal resistance increases, causing higher heat generation during charge/discharge and a measurable efficiency decline. A system that charged with 10 kWh and discharged 8.3 kWh (83% RTE) is significantly more degraded than one that achieved 9.2 kWh (92% RTE). Most monitoring platforms do not show RTE directly, but it can be calculated monthly from the charge energy in and discharge energy out totals visible in the export data. A sustained decline below 88% RTE should trigger a maintenance investigation before the annual review.

When This Approach Breaks Down

Standard health monitoring protocols are calibrated for residential systems in straightforward use cases. They require modification for specific deployment scenarios.

Systems installed in unmonitored environments. A battery installed in a holiday home, rental property, or secondary residence with intermittent owner attendance often receives no routine monitoring. Faults that would be caught within days in a primary residence may go undetected for weeks or months. For these installations, manufacturer email alert subscriptions (available for most Tier 1 brands) should be enabled — these provide fault notification without requiring active log review, and represent the minimum monitoring standard for unattended systems.

Systems with multiple battery units. Homes with 2–3 battery units stacked in series or parallel may have BMS systems from the same brand that report aggregated metrics rather than per-unit data. A fault in one unit may not be individually flagged if the aggregate system SoH remains within expected range. For multi-unit systems, request per-unit health data from the installer during annual reviews rather than relying solely on aggregate app data.

Systems in Very Long Winter Dormancy. A poorly understood health monitoring scenario exists for some off-grid and rural properties that experience extended periods where the battery is at very low SoC in cold temperatures. LFP batteries stored at low SoC in cold conditions can experience lithium plating during the next charging cycle — a form of electrolyte interaction that accelerates capacity fade and, in extreme cases, reduces safety margins. Systems in cold climates that regularly discharge to below 20% in winter should be monitored for unusual cell temperature behaviour during the first charge cycle after a cold dormancy period.

Real-World Example

Scenario: A homeowner in Cambridge, UK monitors a Tesla Powerwall 3 (13.5 kWh) installed in March 2024. Monthly monitoring review in January 2026 (approximately 22 months of operation):

App data reviewed:

• Reported SoH: 96.7%
• Reported lifetime cycles: 734 EFC
• Recent daily savings: £2.10/day (on Octopus Flux, consistent with prior months)

Manual RTE calculation (from January 2026 monthly data export):

• Total January charge energy in: 248 kWh
• Total January discharge energy out: 227 kWh
• Calculated RTE: 91.5%

Comparison to Year 1 baseline (January 2025 export data):

• Total charge energy in: 231 kWh
• Total discharge energy out: 216 kWh
• Year 1 RTE: 93.5%

RTE decline: 2 percentage points over 12 months. This is within the expected degradation range for LFP systems (typically 0.5–2% per year of RTE decline in Years 1–3). No action required, but the baseline is now established for comparison in January 2027.

Lesson: The app's "96.7% SoH" metric alone provided limited actionable insight. The manually calculated RTE from raw export data provided a quantitative baseline that can be tracked year-over-year to identify any acceleration in degradation. Monthly data exports typically take 5 minutes to review. See the battery maintenance guide for the full annual review protocol.

Engineering Recommendation

Battery health monitoring should be treated as a predictive discipline, not a reactive one. The value of monitoring is not in detecting faults after they occur — that's the inverter's fault light — but in identifying trends that allow corrective action before irreversible degradation accumulates.

The minimum effective monitoring programme:

• Monthly: Review the prior month's charge/discharge energy data and calculate actual round-trip efficiency. Flag any month where RTE falls below 88%.
• Annually: Conduct a controlled capacity test and compare against prior years. Record results in a simple spreadsheet that tracks degradation rate year-over-year.
• Annually: Review cell temperature peak logs for the prior summer and confirm no readings exceeded 35°C (LFP). Investigate any that did.

Tools available:

• Most battery monitoring apps provide raw data export (CSV or JSON) — enable this data export in your app settings for monthly review
• For Home Assistant users, the LFP Battery Monitor integration provides automated RTE calculation and trend tracking
• Annual capacity test methodology: Charge to 95% at a defined power rate (e.g., 2 kW), hold for 1 hour, then discharge to 10% at the same rate. Measure the kWh recorded by the monitoring app during discharge. Compare against rated usable capacity.

The key decision trigger is a measured usable capacity below 80% of original rated capacity within the warranty period. Document the controlled capacity test result, retain the monitoring export data showing the test, and submit a warranty claim with both the test result and the supporting data. Most Tier 1 manufacturers process warranty capacity claims within 4–8 weeks when supporting data is complete. Review the top-rated batteries guide to understand what warranty terms are standard before making a claim.

Related Reading

• Biggest Mistakes Homeowners Make with Solar Batteries — Ignoring health signals is a costly mistake
• When NOT to Buy a Solar Battery — When battery health signals indicate a poor investment
• Solar Battery Payback Reality: UK vs US vs Global — How degradation affects your payback figures