US models / USD · No active affiliate agreements or commission-earning links.
Methodology / Version 2

How we compare power stations

By Fordham Tucker · Updated · US appliance planning

PowerPick is a planning tool built from manufacturer specifications and transparent arithmetic. We have not tested these products in a laboratory or in outages.

What we calculate.

We convert annual refrigerator energy to average power: annual kWh × 1,000 ÷ 365 ÷ 24. Measured mode instead uses cumulative kWh × 1,000 ÷ elapsed hours. We add the other average AC loads entered by the user. The annual figure already accounts for the estimate's cycling behavior; we apply no second duty-cycle multiplier.

Battery-side draw = average AC load ÷ inverter efficiency + battery-side idle draw

Runtime hours = nominal Wh × usable share ÷ battery-side draw

Required nominal Wh = target hours × battery-side draw ÷ usable share

Default usable share is 90%, default inverter efficiency 90%, and default battery-side idle overhead 10 W. These are editable modeling assumptions, not measured product characteristics. Efficiency excludes the separately modeled idle overhead; using an all-in measured efficiency and adding overhead again would double-count that loss.

For worked low-load scenarios, see how inverter idle draw affects runtime. The guide keeps assumed overhead separate from measured product evidence.

Input evidence and sensitivity.

The user identifies an illustrative example, an EnergyGuide label or a personal measurement. This describes the input source; it is not a numerical confidence score. The displayed sensitivity changes total average appliance demand by the user-selected percentage while holding efficiency, usable share and idle draw fixed. The default ±20% is illustrative, not an empirically established error bound. Actual conditions may fall outside this range.

Low runtime = capacity × usable share ÷ (average load × (1 + sensitivity) ÷ efficiency + idle). The high-runtime case uses (1 − sensitivity). Record and repeat your measurement before treating it as representative.

What the screen does and does not establish.

For each listed station we show modeled energy duration and compare maximum running and starting inputs with its published total continuous output. Above-rating startup requires an additional surge-duration check. Other simultaneous maximum demand is entered separately from its average energy demand. Blank running, starting or other simultaneous demand means unknown; it never counts as a compatibility pass. A below-rating numerical screen is not a compatibility certification. The user must verify voltage, individual outlet and bypass limits, appliance behavior and manufacturer instructions.

No solar or other input is assumed. We do not model battery temperature, aging curves, load-dependent efficiency, transient voltage, compressor behavior or actual food temperature. A zero appliance load is rejected rather than returning infinite runtime. Negative and non-finite input values are rejected. Usable share and efficiency must be greater than zero and at most 100%.

Shopping filters and photographs.

The catalog filters only published nominal capacity, observed unit price and approximate weight. These are shopping constraints, not an appliance compatibility verdict. Sorting changes the order without awarding a winner. Selections remain in memory when filters change.

Product photographs are the exact manufacturer gallery assets for the listed models, served locally. They are not shown to a common physical scale. Image and current offer source ledger records the original URLs and November 15, 2026 check. EcoFlow’s approximately 51.5 kg weight is converted to 113.5 lb.

How we select and order products.

The initial set illustrates two roughly 2 kWh options and one larger 4 kWh option. Order follows capacity class rather than commission rate. It is not exhaustive and is not a ranking of tested quality. Product availability is shown separately from calculated fit. A higher capacity product can be unavailable, and a lower-priced product can be appropriate.

Source ledger.

Checked November 14, 2026. Prices are observations in USD on US pages, not live quotes. The merchant's current checkout and terms control.

Jackery Explorer 2000 v2 — US capacity, output, weight, price and availabilityBLUETTI Elite 200 V2 — US capacity, continuous output, weight and priceBLUETTI Elite 200 V2 — US specifications PDFEcoFlow DELTA Pro 3 — US capacity, output and availabilityFTC — Understanding EnergyGuideGoal Zero — manufacturer explanation of inverter loss and model-specific idle draw

P3 International meter manual supports the distinction between cumulative energy and elapsed hours; it does not validate our sampling duration or startup capture.

Editorial next steps.

Fordham Tucker is the publisher. Send source or calculation corrections to hello@powerpicktool.com. Source observations and current stock still need review before relying on a purchase decision. Future improvements should prioritize measured appliance examples and independently collected runtime tests. We will identify tested observations separately from manufacturer claims and calculations.

Put your own numbers to work.

The free calculator turns your annual fridge energy and other loads into a battery estimate, with every assumption visible.

Open the runtime tool ↗