How to Monitor Your Solar Production (6 Steps)

Your solar system quietly makes a number every single day, and the difference between an owner who catches a failing inverter in a week and one who discovers it on next year’s true-up bill is nothing more than knowing how to read that number. Learning how to monitor solar panel output is the single most valuable habit a solar owner can build, because the panels cannot tell you when something is wrong, but the data can, if you know what a normal day looks like and what a genuinely bad one looks like. The trap most owners fall into is the opposite of neglect: they check the app on a cloudy afternoon, see a low number, and either panic over nothing or, worse, learn to ignore the app entirely so that a real fault later slips by unnoticed.

This walkthrough fixes that in six plain steps: find and log into your monitoring, learn to read kilowatt-hours, set an expected baseline, compare expected against actual while accounting for weather and season, spot underperformance and find which panel is at fault, and turn on alerts so you know exactly when to call your installer. It builds on our broader solar maintenance field guide, which treats monitoring as the first line of defense, and it leans on the same production math as our solar sizing reference when you want to know what a system your size should make. Keep the companion calculator open as you read to turn your own system size and sun hours into a baseline you can check against.

Before you start: what you need

Monitoring solar production takes almost no equipment, because the tools came with your system, but it does take a few minutes of setup and one honest mental adjustment: you are learning to read a number that swings wildly with the weather, so patience and a baseline matter more than any gadget. The work here is not physical and carries no risk, unlike cleaning or roof inspection, which makes it the safest and highest-value maintenance habit you have. What you are really gathering before Step 1 is access and a reference point, not hardware.

Gather these before you begin:

• Your monitoring account and app. Nearly every system installed in recent years came with a manufacturer app tied to your inverter brand or a small communication gateway on the wall. Your installer created the account at commissioning; if you never got the login, they can resend it.
• Your system size in kilowatts. This is on your installation paperwork or your original quote, and it is the single most important number for setting expectations. If you are unsure, our sizing reference explains how to read it off your equipment list.
• Your local peak sun hours. A rough figure for your region, commonly around 5.5 to 6 in the Southwest and 3.5 to 4.5 across the Northeast and Midwest, lets you estimate a realistic daily target rather than guessing.
• A little patience. The single biggest mistake is judging the system on a cloudy Tuesday. You need to watch across similar clear days to learn what normal looks like.

Difficulty here is genuinely low, an illustrative fifteen minutes to log in and set a baseline, then a couple of minutes a week thereafter. What matters is the habit and the reference point, not constant vigilance. With your app open, your system size in hand, and a realistic sense of your sun hours, the six steps below turn a wall of numbers into an early-warning system that pays for itself the first time it catches a fault under warranty.

Step 1: Set up and log into your monitoring system

Start by simply getting into your monitoring, because everything else depends on seeing your own data. Almost every home solar system installed in the last several years includes monitoring by default, tied to one of three setups: a string inverter with its own app, microinverters or power optimizers reporting through a gateway, or a battery system whose app also shows solar production. Your installer connected this to your home internet and created your account when the system was commissioned, so the first task is to find that account rather than to install anything new. If you have the app already, open it; if not, the login was emailed at handover, and your installer can resend it in minutes.

Once you are in, confirm three things before you trust any number. First, that the system is actually communicating: a monitoring gateway that has lost its internet connection will show stale data or zeros even when the panels are producing fine, and mistaking a communication dropout for a production failure is one of the most common false alarms. Second, that the system size shown in the app matches your paperwork, so you know the app is configured correctly. Third, where the app keeps daily, monthly, and lifetime totals, since you will return to those constantly. As an illustrative habit, note the shape of today’s production curve so tomorrow’s looks familiar.

Watch out for assuming no news is good news. A monitoring system that has quietly gone offline stops warning you about anything, so a dead app is itself a problem to fix, not a reason to relax. Confirm the last successful reading is recent, not days old. If your gateway shows a connection error, the fix is usually as simple as rebooting the router or the gateway, and our maintenance field guide covers keeping that link alive. With a working login and a confirmed live connection, you are ready to actually read what the numbers mean.

Step 2: Learn to read your kWh output

With the app open, learn to read the one unit that matters: the kilowatt-hour, or kWh, which is the actual energy your panels made, as opposed to kilowatts, which is the instantaneous rate at any moment. Your bill is charged in kilowatt-hours, so kilowatt-hours are what you compare against it. Every monitoring app organizes production into the same nested totals: today’s kWh, this month’s kWh, this year’s, and lifetime kWh since the system switched on. The daily figure is the one you learn the feel of first, because it is what swings with weather, while the monthly and annual totals are the ones that reveal real trends by averaging that noise away.

Beyond the totals, learn to read the daily production curve, the graph that plots power output through the day. On a clear day it should trace a smooth bell shape: near zero at dawn, climbing through the morning, peaking around midday when the sun is highest, then tapering to zero at dusk. That shape is your friend, because deviations from it are diagnostic. A curve with a clean bell but a lower peak usually means thin cloud or a hazy sky. A curve with a sharp notch cut out of an otherwise smooth arc often means a shadow, a chimney, a tree, or a vent, crossing part of the array at the same time each day. A curve that flatlines at zero on a sunny day means the system is not producing at all.

Watch out for confusing kilowatts and kilowatt-hours, which trips up many new owners. Seeing “5 kW” on the app at noon does not mean the system made 5 kWh; it means it is producing at a rate of 5 kilowatts at that instant, and the day’s energy total is the area under that curve. As an illustrative example, a 7 kW system might peak near 6 kW around midday yet make roughly 25 kWh across the whole day. Read the kWh totals for judging performance, and use the live kilowatt reading and the curve shape only for spotting shade and faults. Next, turn those totals into an expectation.

Step 3: Set your expected production baseline

Numbers mean nothing without something to compare them to, so the pivotal step is setting an expected baseline: roughly what a system your size should make. The simplest honest estimate multiplies three things: your system size in kilowatts, your local daily peak sun hours, and a derate of about 0.8 that accounts for real-world losses from heat, wiring, inverter conversion, and minor soiling. That gives an average clear-day figure in kilowatt-hours. As an illustrative worked number, a 7 kW system at 4.5 peak sun hours estimates to about 7 times 4.5 times 0.8, or roughly 25 kWh on an average clear day, which scales to somewhere near 9,000 kWh across a year.

The annual figure is the sturdiest baseline because it averages out every cloudy day and every brilliant one, so it is what you compare year over year to confirm the system is healthy over the long run. The clear-day figure is your day-to-day sanity check: when the sky is genuinely clear, output should land in that neighborhood, allowing for season. Set both now, using your own size and sun hours, so that every future reading has a reference. Our sizing reference walks the same arithmetic in detail, and if you are still deciding on a system, how many solar panels you need starts from the same production math from the other direction.

Watch out for treating the baseline as a promise rather than a center of gravity. Real output orbits the baseline: below it on cloudy and winter days, above it on long clear summer ones, and that spread is normal and expected. A baseline is useful precisely because it lets you see when a clear day falls far outside the normal orbit, which is the signal worth chasing. Enter your size and sun hours into the companion calculator to generate your own clear-day and annual baseline, then write those two numbers down. With a baseline in hand, the real skill begins: comparing it against what actually happened.

Step 4: Compare expected against actual output

Here is where monitoring earns its keep, and where most owners go wrong: comparing actual output to expected in a way that accounts for weather and season instead of ignoring them. The cardinal rule is to compare like with like. A cloudy day will always read low, and a December day will always read far below a June day from the identical panels, so comparing across weather or across seasons produces false alarms and false comfort in equal measure. The only comparison that reliably reveals a problem is a clear day against similar clear days: if a bright, sunny day today makes far less than bright, sunny days made last month or last year, something has changed. If today was overcast, the low number explains itself.

Season shapes the whole picture and must be built into your expectations. Summer days are long and the sun rides high, so output peaks; winter days are short and the sun sits low, so even a cloudless winter day produces a fraction of a cloudless summer day. Snow covering the glass can cut output to almost nothing until it sheds or melts, which is normal, not a fault, and our note on whether solar panels work in winter explains why a low-but-not-zero winter month is exactly what to expect. The healthy pattern across a year is a broad hump, high in summer, low in winter, and your monthly totals should trace that hump every year in roughly the same shape.

Watch out for two opposite errors. The first is panic: seeing a low daily number during a storm and assuming the system is broken, when the sky alone explains it. The second is complacency: shrugging off a genuinely low clear-day reading as “probably just weather” when the sky was in fact clear and the drop is real. The discipline that avoids both is simple: before you react to any low number, ask what the sky was doing, and only investigate when a truly clear day underperforms your baseline. Enter today’s figure in the companion calculator to see where it lands against your clear-day baseline. When a clear day genuinely comes up short, it is time to find out why.

Step 5: Spot underperformance and find the culprit

When a clear day genuinely underperforms your baseline across several similar days, move from noticing to diagnosing, and here the kind of monitoring you have decides how fast you can work. The first question is always scope: is the whole system down, or is one part dragging? A whole-system drop to zero on a sunny day points at the inverter or the grid connection; a partial, steady shortfall points at soiling, shade, or a single failing component. Rule out the free explanations first, in order: confirm the sky really was clear, check whether the glass is visibly soiled or snow-covered, and look for new shade from a tree that has grown or a structure that went up since the system was installed.

This is where panel-level and system-level monitoring diverge sharply. With panel-level monitoring, from microinverters or power optimizers, the app shows a tile for every panel, so a single soiled, shaded, or failed module appears as one low tile in a grid of normal ones, and you can point at the culprit in seconds. With system-level monitoring, from a plain string inverter, you see only one combined number, so a single weak panel hides inside the total and you rely on the daily curve shape, a visual inspection, and comparison to your baseline to infer that something is off. Neither is wrong; panel-level simply shortens the search. The section below on panel-level versus system-level monitoring covers this trade in more depth.

Watch out for chasing a fault that is really just physics or dirt. A notch in the daily curve at the same time each afternoon is almost always predictable shade, not a broken panel, and a gentle across-the-board sag on a hazy day is thin cloud. Before you conclude a component has failed, confirm the drop is real, sustained, and unexplained by weather, soiling, or shade. Our maintenance field guide walks the physical checks, and the lifespan note explains why a small, gradual yearly decline is normal aging rather than a fault. When the drop is real and unexplained, the last step is knowing when the app should page you and when you should page your installer.

Step 6: Set alerts and know when to call your installer

Finish by making the system watch itself, because no one checks an app reliably forever, and a fault that sits unnoticed for months quietly wastes the production your warranty would have restored for free. Nearly every monitoring platform can send automatic alerts, and turning them on is the highest-leverage two minutes in this entire walkthrough. Enable, at minimum, an offline or communication alert that tells you when the system stops reporting, and a fault or low-production alert that flags an inverter error or an unusual drop. With those on, you no longer have to be vigilant; you only have to respond when the system asks for you, which frees you to check the trend casually rather than anxiously.

Knowing when to escalate to a human is the other half. Call your installer when a clear-day comparison shows a real, sustained shortfall you have already ruled out weather, soiling, and shade for; when the app flags an inverter fault or error code; when a string or block of panels reads zero on a sunny day; or when the monitoring goes dark and a router reboot does not bring it back. These are the signals that point at equipment, and most systems carry workmanship and equipment warranties for years, so a genuine fault caught early is frequently repaired at little or no cost. That is the whole financial case for monitoring: it converts a silent, expensive loss into a warranty claim.

Watch out for calling on the strength of a cloudy day or a single low reading, which wastes everyone’s time and teaches you to distrust your own instrument. Before you call, have your details ready: system size, the specific dates and kilowatt-hour figures you are seeing, what the sky was doing, and any error codes on the inverter or in the app. A clear, weather-adjusted description lets the installer often diagnose remotely, since many can see your monitoring data too. Alerts on, clear-day discipline in hand, and your numbers ready, you have turned a passive array into a system that tells you the moment it needs attention.

A worked example: a month of monitoring

Take one household through a full month to see the habit in motion. The Nguyen family owns a 7 kW system in a region with about 4.5 peak sun hours, and in Step 3 they set their baseline: roughly 25 kWh on an average clear day and near 9,000 kWh a year. Through the first two weeks of the month, their app bounces between 8 kWh on stormy days and 29 kWh on brilliant ones, and early on they nearly panic at the 8 kWh reading, until they remember the rule and check the sky, which was a downpour. They learn quickly to glance at the weather before the number, and the low days stop alarming them.

In the third week, something different happens. On a genuinely clear, bright day, their app reads 19 kWh, well below the roughly 25 kWh clear-day baseline, and the next two clear days read similarly low. This is the pattern that matters: a real, sustained shortfall on similar clear days, not a one-off cloudy dip. In Step 5 they work the checklist in order. The sky was clear on all three days, so weather is out. They look at the panels and see a chalky film of dust after a long dry spell, and because their system happens to have panel-level monitoring, the app shows two tiles in the lower rows reading noticeably below their neighbors, consistent with soiling settling on the lower panels.

They resolve it without a service call: a careful ground-level cleaning on a cool morning, covered in our cleaning walkthrough, and two clear days later the app is back near 25 kWh, with the two low tiles matching the rest. Had the cleaning not restored output, the flat clear-day shortfall plus panel-level fault flags would have been their cue to call the installer under warranty. Every figure here, the 7 kW size, the 25 kWh baseline, the 19 kWh dip, is illustrative, chosen to show the routine rather than to predict your roof. Run your own size and readings through the companion calculator to build the same instinct.

What a clear day makes by season

Because season swings output so much, it helps to see the shape of a normal year before you judge any single day, since a winter reading that would be a red flag in July is perfectly healthy in January. The chart below shows an illustrative pattern of what the same system makes on a clear day across the seasons and weather, scaled so that a clear summer day is the peak. It is a teaching pattern to calibrate expectations, not a measurement of your roof, and your latitude, roof tilt, and climate will shift every bar.

Read down the bars and the discipline from Step 4 becomes obvious. A clear winter day making under half of a clear summer day is not a fault; it is geometry, the sun sitting low and the day running short. Overcast days making a small fraction are the weather, not the equipment. Snow reading near zero is expected until the glass clears. The only way to see through all this seasonal noise is to compare a clear day to similar clear days, or to compare a whole month against the same month in a prior year, which is exactly why the annual total is the sturdiest health check you have.

Why production reads below expected

When output does fall short, it helps to know how the causes tend to break down, because it keeps you from blaming a fault for what is usually the sky. The chart below splits an illustrative set of below-expected readings by their real cause. The point is not that these exact shares apply to your system, but that ordinary weather and season account for the large majority of low readings owners actually see, while genuine equipment faults are the smallest slice, which is why weather-adjusted comparison matters so much.

The lesson is to investigate in the order the causes actually occur. Check weather and season first, because they explain most low readings and cost nothing to rule out. Check soiling next, since dust and pollen are common and often fixable from the ground. Consider shade, especially a notch that appears at the same time each day as a tree grows or a structure goes up. Only after those are ruled out does a genuine equipment fault become the likely answer, and that is the moment to lean on your alerts and call your installer. Working the causes in this order keeps you from a needless service call while still catching the real fault when it is there.

Panel-level versus system-level monitoring

Because it shapes how you diagnose everything above, it is worth understanding clearly which type of monitoring your system has and what each can and cannot show you. The difference traces back to your inverter architecture, a choice made when the system was installed. A traditional string inverter wires panels together in series and reports one combined production figure for the whole array, which is system-level monitoring. Microinverters, one small unit per panel, and power optimizers, one per panel feeding a central inverter, both report each panel individually, which is panel-level monitoring. You did not choose this per se; it came bundled with the equipment on your quote, and our sizing reference touches on how these architectures appear on an equipment list.

System-level monitoring is perfectly capable of catching a real problem: if the inverter faults, a string goes offline, or the whole array underperforms across clear days, the single combined number shows it plainly. What it cannot do is point at which panel is the culprit, because one weak module is diluted across the total and can hide inside normal weather variation. Diagnosis with system-level data therefore leans on the daily curve shape, a visual inspection of the array, and comparison to your baseline. It is less granular, not less trustworthy, and for many owners it is entirely sufficient, especially on smaller, unshaded roofs where there is little to go wrong panel by panel.

Panel-level monitoring shines where shade or complexity make individual panels behave differently. Seeing each panel as its own tile means a single soiled, shaded, cracked, or failed module stands out instantly, and the healthy panels keep producing rather than being dragged down by one weak link in a string. The trade is more hardware and a slightly higher cost, decided at purchase. If you are still choosing a system, this is worth weighing against your roof’s shade and layout; if you already own one, the practical takeaway is simply to know which you have, so you set the right expectations for how quickly you can pinpoint a problem when a clear day comes up short.

Common mistakes when monitoring solar

The same handful of errors account for most of the wasted worry and missed faults in solar monitoring, and every one is easy to avoid once named.

• Judging the system on a cloudy day. A low reading under heavy cloud or during a storm is the weather, not a fault. Always check what the sky was doing before you react to any low number, and compare only similar clear days.
• Comparing across seasons. A December total will always sit far below a June total from the same panels. Compare a month against the same month in a prior year, or a clear day against similar clear days, never winter against summer.
• Confusing kilowatts with kilowatt-hours. The live kW reading is an instantaneous rate; the kWh totals are the energy you actually made and what your bill counts. Judge performance on kWh, and use the live kW only to read the curve and spot shade.
• Ignoring the app entirely. An unchecked system can run faulted for months, wasting production a warranty would have restored for free. A weekly glance plus automatic alerts is the low-effort habit that prevents this.
• Mistaking a communication dropout for a production failure. A gateway that has lost internet shows zeros or stale data even when the panels are fine. Confirm the system is online before assuming production has stopped.
• Calling the installer over normal weather. Escalating on the strength of a cloudy day or a single low reading wastes time and erodes trust in your own monitoring. Escalate only on a real, sustained, weather-adjusted shortfall or a genuine fault flag.

Avoid these six and monitoring becomes calm and reliable: you stop panicking over clouds, you stop ignoring the app, and you catch the rare real fault while it is still a cheap warranty repair rather than a year of lost output.

Troubleshooting: when the numbers look wrong

What if the app shows zero production on a sunny day? First rule out a communication dropout, which is far more common than a true outage: check whether the gateway or inverter has lost its internet connection, and try rebooting your router and the gateway, since a stale connection often reports zero even while the panels produce. If the connection is live and the inverter still shows zero or an error light on a sunny day, that points at a genuine inverter fault or a tripped connection, which is a clear reason to contact your installer, especially since the inverter is usually the component most likely to need replacement over a system’s life.

What if only part of my production seems low? On panel-level monitoring, look for individual tiles reading below their neighbors, which localizes the problem to specific panels and usually means soiling, shade, or a failing module in that spot. On system-level monitoring, study the daily curve for a notch or an early drop-off that recurs at the same time each day, which typically signals shade crossing the array rather than a fault. In both cases, check for a physical cause you can see, dust, a new branch, a nearby structure, before assuming a component has failed, since the free explanations are the common ones.

What if my annual total is a little lower than last year? A small year-over-year decline is normal and expected, because panels degrade gradually, commonly a fraction of a percent a year, so a slightly lower annual figure is usually aging rather than a fault, as our lifespan note explains. What is not normal is a sudden, large annual drop, which points at soiling, growing shade, or an equipment problem worth investigating. Compare the size of the change, not just its direction: a percent or two is the slow, expected fade, while a double-digit drop deserves a real look.

What if the monitoring data and my electric bill disagree? Remember they measure different things: your monitoring shows gross production, while your bill reflects net usage after your home consumes some of that solar directly and after any net metering credits. A mismatch is usually the difference between what you generated and what you exported or offset, not a monitoring error. To reconcile them, learn to read both, and our note on reading your electric bill explains how solar production and billed usage relate, so a normal accounting difference does not read as a problem.

Your solar monitoring checklist

Use this as the save-and-act summary. Work it top to bottom and you will catch real faults early while ignoring the weather noise that trips up most owners.

• Log into your monitoring app and confirm the system is online, with a recent successful reading, not stale or zero data from a dropped connection.
• Know your two reference numbers: your clear-day baseline (roughly size in kW times peak sun hours times 0.8) and your expected annual total.
• Read the kWh totals for performance, and use the live kW reading and daily curve shape only to spot shade and faults.
• Before reacting to any low number, check what the sky was doing; never judge the system on a cloudy day.
• Compare similar clear days against each other, or a month against the same month last year, never one season against another.
• Glance at the app weekly, compare the monthly total when the bill arrives, and investigate only a real, sustained clear-day shortfall.
• Rule out causes in order: weather and season first, then soiling, then shade, then a genuine equipment fault.
• On panel-level monitoring, look for individual low tiles; on system-level, study the curve and inspect the array.
• Turn on automatic offline and fault alerts so the system watches itself between your checks.
• Call your installer, details in hand, for an unexplained clear-day drop, a fault or error code, zero output on a sunny day, or monitoring that stays dark.

The bottom line

Monitoring solar production is the cheapest, safest, and most valuable habit a solar owner has, because it is the only thing that turns a silent array into a system that tells you when it needs help. The whole skill reduces to a few disciplines: set a baseline for what a system your size should make, read the kilowatt-hour totals rather than the instantaneous rate, and above all compare like with like, similar clear days against each other and each month against the same month a year before, so that weather and season never masquerade as a fault. Do that, and the wild daily swings that panic most owners become background noise you can read straight through.

The figures throughout, an illustrative 25 kWh clear-day baseline on a 7 kW system, a near 9,000 kWh year, a 0.8 derate, and the seasonal and cause shares in the charts, are teaching examples chosen to build the instinct, not measurements or promises for your roof, and your climate, tilt, shade, and equipment will move every one. What does not change is the payoff: with automatic alerts on and clear-day discipline in hand, you catch the rare genuine fault while it is still a cheap warranty repair rather than a year of quietly lost production. Set your own baseline in the companion calculator, read this alongside the broader maintenance field guide and the sizing reference, and your monitoring app becomes exactly what it was meant to be: an early-warning system you can trust.

WattBarn publishes this walkthrough to help you read your own solar monitoring with confidence, not to serve as electrical, engineering, or warranty advice. The baselines, derate, kilowatt-hour figures, and seasonal and cause shares above are illustrative examples chosen to teach the six steps, not measurements or guarantees for your system, and your own climate, roof tilt, shade, equipment, and monitoring platform will produce different numbers. Solar equipment carries live voltage and its faults are best diagnosed by professionals, so treat your monitoring as an early-warning tool and let a licensed installer, rather than these worked sketches, confirm and repair anything involving your inverter, wiring, or panels.