How GridReady Tests

Power-gear marketing is built on headline numbers — 2,000 watts of "output" that lasts 3 minutes, "13,000 mAh" capacity that delivers 8,000 mAh of real energy, "60 dBA" generators that hit 74 dBA under load. Our testing focuses on the numbers that determine what a unit actually does in a real outage, on a real campsite, or in a real backyard.

Test environments

Every product we recommend has been used in at least one of three environments:

• Home backup simulation. We trip the main breaker on a 1,800 sq ft single-family home and run essentials (refrigerator, Wi-Fi router, lights, sump pump, CPAP) off the unit under test. We measure runtime to depletion, fan-noise dB at 1 m, and how the unit handles repeated motor-start surges from the fridge compressor.
• Off-grid camping. Weekend trips with the unit as the only power source — laptops, phones, lights, fridge cooler. We measure realistic solar recharge rates under varying sky conditions (full sun, partial cloud, overcast).
• Job-site / workshop use. Power tools (circular saw, drill, table saw, air compressor) plus shop lights. Tests inverter quality under heavy motor-start loads and waveform purity for sensitive electronics.

What we measure (the numbers behind every review)

Battery capacity

We discharge each unit through a constant-resistance load and measure delivered watt-hours with a Hopi HP-9800 power meter at the AC outlet. Manufacturer-stated capacity is at the battery cell level; we report what actually reaches your appliance after inverter losses. The gap is typically 10-15% on AC output.

Sustained output (not headline lumens or watts)

Marketing specs love peak numbers that last 3 minutes before thermal protection kicks in. We run units at 50% of their rated output for 30 minutes minimum, and report the actual sustained wattage. For flashlights, we measure lumens after 5 minutes and again at 30 minutes — the brightness that matters during real use.

Recharge speed

Timed from 0% to 80% and 0% to 100% via the unit's fastest supported input method (typically wall AC). We use a clamp meter on the input line to verify actual draw matches the manufacturer's claimed input wattage.

Solar input under real sky conditions

Manufacturer "1,200 W solar input" specs assume STC (Standard Test Conditions): 25°C, no clouds, panel perpendicular to the sun. Real-world MPPT input typically runs 65-80% of the rated panel wattage. We measure actual delivered watts at solar noon with a calibrated pyranometer, with panels at our location's latitude angle, and report the gap.

Noise level

Decibel measurement at 1 meter and at 7 meters (23 ft, the standard generator-industry reporting distance), at 25%, 50%, and 75% load. We use a calibrated Class 2 sound level meter. Generator manufacturers usually report only the 25%-load number; we report all three.

Cycle life claims

We can't run a 10-year cycle test in real time, so we report manufacturer cycle ratings honestly and apply industry-standard derating: LiFePO4 (LFP) batteries typically retain 80% capacity after 3,000-4,000 cycles; NMC chemistry typically retains 80% after 500-1,000 cycles. We weight cycle ratings into our recommendations because daily-cycle off-grid users care, weekly-cycle home-backup users care less.

What we don't do

• We don't accept paid placements. Brands cannot pay to be reviewed, to appear in roundup guides, or to receive favorable rankings. We have turned down sponsored-review requests from multiple major brands in the portable-power space.
• We don't accept pre-approved drafts. Brands do not see our review content before publication and have no input on our verdicts.
• We don't recommend products we wouldn't buy. A 3.5-star rating from us is not a 4.7-star Amazon review. We will say a unit is overpriced, underbuilt, or behind the segment when it is — see the Goal Zero Yeti 1500X review for an example of a brand-name product we explicitly do not recommend at its current price.
• We don't run AI-generated reviews. Our reviews are written by named human editors with hands-on time on the units. AI tooling helps with research, transcription, and editorial polish, but verdicts and recommendations are made by people who have used the gear.

Sources we cite

Where we use third-party data (long-term reliability reports, large-sample-size cycle testing, certification documents), we cite the source inline. Primary sources include manufacturer spec sheets, UL/CSA certification databases, FCC public filings, and independent testing labs (when results are publicly available).

Affiliate relationships and editorial independence

GridReady earns affiliate commissions when readers buy through our links. This funds the testing program (units are expensive, especially in the 2-6 kWh tier). It does not influence our verdicts. Commission rates do not affect which products we recommend — we have repeatedly recommended products with lower commission rates (e.g., Honda generators) over higher-paying competitors when the lower-paying product is better. Full details on our affiliate disclosure page.

Corrections

If we publish an incorrect spec, an outdated price, or a factually wrong claim, we correct it inline and add a dated note at the bottom of the article. Significant corrections (e.g., a spec error that changes our recommendation) trigger a separate notification at the top of the article for 30 days. See our Editorial Standards page for the full corrections policy.

Questions

Methodology questions, source requests, or test-protocol clarifications: [email protected].

Last updated: May 2026