EcoFlow App & PowerInsight 2: Read Your Backup Data

Most owners treat the EcoFlow app as a glorified battery percentage — glance at the number, lock the phone, move on. That wastes the most useful part of the whole system. During an outage, the difference between a household that rides it out calmly and one that panics is almost always information: knowing exactly how fast the pack is draining, whether the panels are keeping up, and how many hours of runtime are actually left at the current load. All of that is sitting in the app already. A reviewer who lived with the software put it plainly: the EcoFlow app turns the power station “from a dumb battery into something you can actually manage in real time” (BackupPowerHub).

This post is a working guide to the five readings that actually predict your runtime, what EcoFlow’s new PowerInsight 2 monitor adds on top of the phone app, and how to set alert thresholds that give you hours of warning instead of minutes. The goal is to stop staring at a single percentage and start reading the dashboard the way an operator reads a control panel — as a live picture of whether you are net-positive or quietly losing ground.

The dashboard is the most underused part of your system

A power station is a reservoir with a tap and a fill pipe, and the app is the gauge on all three. The reason people ignore it is that under normal conditions nothing on the screen changes much — the unit sits at 100%, sipping from the wall. The dashboard only earns its keep when conditions get dynamic: a grid drop, a cloudy afternoon, a heat wave where you are trying to run an air conditioner intermittently off battery. In those moments the numbers move fast, and reading them correctly is the difference between a controlled drawdown and a dead pack at 2 AM.

The mistake is treating state of charge as the whole story. State of charge tells you where you are; it tells you nothing about where you are headed. A pack at 80% that is draining at 600 W net is in worse shape over the next few hours than a pack at 45% that is holding steady because solar matches the load. To know which situation you are in, you have to read the flow numbers, not just the level. That is the shift this guide is built around, and it pairs directly with the sizing logic in our breakdown of usable watt-hours versus nameplate capacity.

The five metrics that actually predict runtime

Start with state of charge (SoC) — the headline percentage — but immediately pair it with output watts, the real-time total your loads are pulling, and input watts, the total arriving from grid, solar, or car. The relationship between input and output is everything: subtract one from the other and you have your net flow. Positive net flow means SoC is climbing and you can run indefinitely; negative means you are on a countdown. This single subtraction is the most important habit the dashboard can teach you.

The fourth metric is time-to-empty (or time-to-full when charging), the app’s own runtime estimate based on current flow. Treat it as a live readout, not a promise — it recalculates the instant a compressor kicks on or a cloud passes, so a fridge cycling will make it swing by hours. The fifth is cell temperature, the early-warning light for derating: LFP packs slow charging above ~45 °C and stop accepting charge below 0 °C. Watch it whenever the unit is in direct sun or a cold space, because a thermal throttle quietly erases the charge rate you were counting on. For why your real working capacity is smaller than the label in the first place, see the derating explainer.

PowerInsight 2: what the dedicated monitor adds

The phone app is complete, but it has one flaw as an outage tool: it lives behind a lock screen and an account login, on a device whose own battery you may now be rationing. EcoFlow’s answer is PowerInsight 2, the touchscreen energy monitor that shipped with the 12,288Wh DELTA Pro Ultra X and surfaces the same telemetry on an always-on wall display (9to5Toys). For a whole-home install, that means any member of the household can read state of charge, net flow, and per-circuit draw at a glance without unlocking anything.

Where the dedicated monitor pulls ahead is circuit-level visibility. Paired with a Smart Home Panel, PowerInsight 2 can show consumption broken out by branch circuit — the fridge on one line, the well pump on another, the office on a third — so you can see exactly which load is eating your reserve and shed it deliberately. On a single portable unit you can infer the same thing by watching output watts jump when an appliance cycles, but the panel makes it explicit. Whether you need that depends on scale: a one-station setup is well served by the app, while a multi-circuit backup benefits from the wall display.

Setting alert thresholds that give you lead time

Default low-battery alerts are usually set too low to be useful. An alert at 10% tells you the emergency has already arrived; what you want is an alert that fires while you still have room to act. The way to set it is to work backward from your discharge rate. If your essentials baseline — fridge, router, a few lights, phone charging — pulls a net 250 to 300 W from a 3,600 Wh pack, you are draining roughly 8% per hour, so a 30% threshold buys you close to four hours of lead time. That is enough to stage panels, start a generator, or call it a night and shut things down gracefully.

Set a second alert for the charging side if your unit supports it. A “charge stalled” or temperature-warning alert catches the failure mode where you think solar is topping the pack but it has actually throttled — a panel knocked out of the sun, a thermal cutback, a tripped breaker. Without that alert you discover the stall only when SoC has quietly slid for two hours. The principle behind both thresholds is the same: alerts should be leading indicators that hand you time, not trailing indicators that confirm a problem you can no longer fix.

Reading watts-in vs watts-out to know if you’re net-positive

The most consequential number on the screen during a daytime outage is your solar input, because it is almost never what the panel sticker implies. A “400 W” panel on a clear but imperfect day might deliver 230 to 300 W after angle, heat, and wiring losses; a “800 W” array might land at 450 to 600 W. The app shows you the real figure, live, and that figure is what you should plan around. If the panel reads 480 W and your loads are pulling 520 W, you are net-negative by 40 W and slowly losing — a small shed, like cycling the AC less aggressively, flips you positive.

This live input reading is also how you size charge time honestly. Take your remaining watt-hours, divide by the real input the app is showing, and you have a grounded estimate of time-to-full — one that already bakes in today’s weather and panel placement. We work that arithmetic in detail, including why real charge time runs well over the nameplate estimate, in the solar charge-time math. And if your input seems capped no matter how many panels you add, the limit is probably the unit’s solar controller, which we cover in the solar input ceiling and MPPT limits.

From data to decisions: load-shedding by the numbers

Reading the dashboard is only useful if it changes what you do. The decision loop is simple. Check net flow first: positive, and you relax; negative, and you check time-to-empty. If time-to-empty is comfortably longer than the expected outage, hold. If it is shorter, find the biggest line on your output total and decide whether that load is worth its share of the reserve. An air conditioner at 1,100 W against a 250 W essentials baseline is more than four times everything else combined — cycling it intermittently rather than running it continuously is almost always the highest-leverage shed you can make.

Over a multi-day event the same loop runs on a longer clock. Each morning you are net-negative until the sun clears the horizon, net-positive through midday if your panels out-produce your loads, and net-negative again after dark. The job is to bank enough surplus during the positive window to carry the negative ones. The dashboard makes that visible in real time, so you can pre-cool, charge tools, and run heavy loads during the surplus hours instead of guessing. For the household-wide version of this scheduling discipline, see weather-aware pre-charging before a heat wave.