How it works · Solar Performance Cloud
How Solar Performance Cloud inspects a plant
Most solar monitoring answers one question: is the plant running? That is the wrong question — a plant with a failed string is still “running”, it just quietly produces less, every sunny day, for months. SPC asks a harder one: is every part of it running? We read each panel string on every inverter, every five minutes, and compare it against the strings beside it — same roof, same weather, same minute. When one falls behind, there is nowhere for it to hide.
Works with every major inverter brand in Pakistan — Huawei · Solis · Growatt · Sungrow · Canadian Solar · GoodWe · Solarman · Solplanet. One login, whatever is on your roof.
How the data gets to you
Every 5 minutes, SPC connects to your inverter manufacturer’s cloud and pulls the readings for every string on every inverter you own — current, voltage, power, energy. Nothing is installed on your site: no new hardware, no site visit, no change to your system. If your inverters are already online, we can see them.
Why current, not voltage. Strings on the same tracker share a voltage, so voltage cannot tell one string from another. Current is the only measurement that isolates a single string — so every health judgement is made on current. It is the difference between “this inverter looks odd” and “string 7 is dead.”
The dashboard — your whole portfolio in four numbers
| What you see | What it tells you |
|---|---|
| Overall Health | one score for everything you own |
| Energy Today | units generated across all your sites so far today |
| Inverters Online | how many reported in the last 30 minutes |
| Active Alerts | how many problems are open right now |
Below that, one card per site — its health, its output, its status — so a portfolio of ten sites is one glance, not ten logins. If Inverters Online reads 22 of 24, you have two machines that stopped talking, and you knew within half an hour instead of at month end.
The plant page — one site, in depth
The string-health donut puts every string on the site in one circle, coloured by condition: a healthy site is one solid colour; a site with a problem is not, and you can see it from across the room. Underneath, every inverter expands to its strings, each with its current, its score and its condition.
| Score | Condition | What it means |
|---|---|---|
| 90 and above | Healthy | performing in line with its neighbours |
| 60 – 89 | Warning | falling behind — worth watching |
| Below 60 | Critical | significantly down, or not producing at all |
Every score is clickable — it opens the working, so you can see exactly how it was calculated. We do not ask you to trust a number you cannot check. What you find here isn’t “your plant is at 94%” — it’s “string 7 on inverter 3 is producing nothing while its eleven neighbours are fine.”
A real finding
A string reading zero amps while every string beside it produces normally — the plant total barely moves, and the owner’s inverter app shows no alarm. We have found this on 57 different inverters in 60 days.
The analysis page — the evidence table
The full numerical detail, string by string and inverter by inverter, in two tables with an export button — the data is yours to take. Strings marked Spare, Stopped, Non-standard or Inactive are labelled, not silently averaged away. The patterns are what you find here: one inverter consistently below its siblings points at that inverter; one string down across several inverters points at shading or soiling on one part of the roof; everything down after 2pm points at something else entirely.
A real finding
An inverter carrying 60 unused string inputs — hardware already installed and commissioned for roughly double the panels actually fitted. Found by reading the equipment rather than the paperwork.
The energy page — what you produced, and what it’s worth
Units generated per site over time, with capacity beside them — and units per kW, the fair-comparison number that finally lets a 100 kW site and a 500 kW site be judged against each other. Where a meter is connected, generation and consumption sit on the same page: your solar, the grid units drawn over the same period, and the split by day — turning “we generated 4,200 units” into the question that matters: how much of what we used did we cover ourselves?
Where no meter is connected, the grid row stays blank, not zero. A blank says “we do not know”; a zero would be a lie — the kind that makes every number above it untrustworthy.
A real finding
A site recorded on paper at 135 kW that measurably produces like a system more than twice that size. Its asset register had been wrong since installation — so every performance figure calculated against it was meaningless.
The alerts page — what needs you
One feed of everything wrong, across every site and every brand, newest first, filterable by brand, severity, site and whether it’s still open. Two kinds sit side by side:
- What we found — dead strings, underperforming strings, silent inverters.
- What your inverter reported itself — its own fault codes in plain language: insulation faults, fan alarms, phase faults, grid loss, PID faults, earth-leakage, grid over/under voltage.
The wording is specific, because vague alerts get ignored:
“String 4 is 36.2% below average”
“String 7 producing near-zero current (0.004A)”
“Inverter has sent no data for 12h”
How we keep it quiet enough to be useful: we poll every 5 minutes, but we won’t raise an alert until a fault has persisted for 2 consecutive readings, and won’t close one until it has looked healthy for 6. That deliberate patience cut alert volume by 76% — the same faults, a fraction of the messages.
The sun gate — why we go quiet at night
SPC knows where the sun is, at your site, right now. When it drops too low to generate, live string monitoring pauses until morning — and the page says so: “Sun is low — 4.2°. Live string readings pause until tomorrow morning.”
Why this matters more than it sounds: at dusk every string on every roof falls to zero. A monitoring system that doesn’t understand sunrise and sunset reads that as every string failing at once, every single evening — hundreds of alerts that mean nothing. People stop reading them within a week, and then the one that mattered gets ignored too. So we don’t judge a string when there’s no sun to judge it by. When SPC says a string is dead, it is dead — not asleep.
How SPC inspects a plant, every five minutes
Everything above is what you see. This is what happens before you see it — the inspection itself, on one plant, every five minutes of every day.
Step 1 · Read every string, not the plant
A 500 kW site isn’t one number — it’s ~120 separate string measurements, taken together, twelve times an hour.
Step 2 · Check the sun is actually up
Calculate the sun’s elevation at your site’s own location; if it’s too low to generate, stop and wait. No string is judged in conditions where it couldn’t perform.
Step 3 · Compare each string against the strings beside it
Not a catalogue figure or last year — its own neighbours on the same inverter, same roof, same weather, same minute. If eleven strings are at 8 amps and the twelfth is at 3, the weather is not the explanation.
Step 4 · Score it, and show the working
Every score (90+ Healthy · 60–89 Warning · <60 Critical) is clickable and opens the calculation — output, neighbours’ output, the gap, the verdict. We don’t ask you to trust a number you can’t audit.
Step 5 · Wait, before telling you
A fault must appear on 2 consecutive readings (~10 min) before we raise it, and look healthy for 6 before we close it. That patience cut alert volume 76%.
Step 6 · Separate "bad" from "gone"
Underperforming (working but behind — soiling, shading, a loose connection), Dead (zero amps while neighbours produce — the one plant-level monitoring can never see), and Silent (the whole inverter absent).
Step 7 · Read what the inverter says about itself
Collect the inverter’s own fault codes and translate them to plain language, across every brand — insulation, fan, phase, grid loss, PID, earth-leakage, over/under-voltage.
Step 8 · Keep the record
Every reading is stored, so “when did this start?” has an answer and “is it getting worse?” is a chart, not an opinion.
Fairness rules before accusing a string
- · ≥2 working neighbours to compare against
- · neighbours must average above 1 amp (dawn is noise)
- · spare / never-produced strings excluded
- · odd orientations can be excluded by an engineer
Then the gap decides
- · more than 50% below neighbours → Critical
- · 25–50% → Warning
- · 10–25% → noted, not alarmed
- · under 10% → normal variation
We check every string on your plant, against the strings next to it, twelve times an hour, only when the sun is up, twice before we call it a fault — and we show you the arithmetic.
For operators, boards and tenants
The NOC console puts every site, inverter and string in one live table for operators managing many sites — plant, inverter, string, provider, score, status and last real data, searchable and refreshing continuously. Sort by score and the worst performers in the entire portfolio rise to the top, whatever site or brand they’re on.
The printable report is client-ready, on letterhead, generated from live data — something you can send a board, a landlord, a bank or a tenant without rebuilding it in a spreadsheet. And access fails closed: each organisation sees only its own sites, enforced on every page and request; a user with no assigned site sees nothing at all.
The founder and the team
Wattey Energy was founded in 2018 by Engr. Reyyan Niaz Khan — a UET Lahore electrical engineer with 14+ years in Pakistan’s energy sector and the pioneer of digital energy auditing in pakistan. Solar Performance Cloud and WattEY were designed and built by Ali Ahmed, Head of Energy Systems & Product Engineering; the company’s engineers handle installation, rollout and training at client sites.
Meet the founderWhat this is really for
Solar is sold on a number: this system will produce X. Almost nobody ever checks whether it did. A failed string does not announce itself — it does not trip the inverter, it does not show on the bill, and it does not move the plant total enough to notice. It simply produces nothing, every sunny day, until someone looks at the right level of detail.
That is the entire job of this platform: to look at the right level of detail, every five minutes, on every string you own — and to tell you the day something changes, not the quarter.
How SPC works, your questions
- Does SPC need new hardware on my site?
- No. Every 5 minutes it connects to your inverter manufacturer’s cloud and pulls the readings for every string — nothing is installed on site, no visit, no change to your system. If your inverters are online, we can see them.
- How does SPC find a dead string the inverter app misses?
- It compares each string against its own neighbours on the same inverter — same roof, weather and minute. If eleven are at 8 amps and the twelfth is at 3, the weather isn’t the explanation. It reads current (not voltage) because current is the only measurement that isolates a single string. We’ve caught this on 57 inverters in 60 days.
- Why doesn’t it flood me with alerts at sunset?
- The sun gate: SPC calculates the sun’s elevation at your site and pauses live string judgement when it’s too low to generate — so it never reads “every string failed” at dusk. And a fault must persist 2 consecutive readings before it’s raised, and look healthy for 6 before it’s closed. That cut alert volume 76%.
- Can I check how a score was calculated?
- Yes — every score is clickable and opens the working: the string’s output, its neighbours’ output, the gap, and how the verdict was reached. We don’t ask you to trust a number you can’t audit.
- Does it show my electricity use too, not just generation?
- Where a meter is connected, generation and consumption sit on the same page — your solar, the grid units drawn, and the split by day — so you can see how much of what you used you covered yourself. Where no meter is connected, the grid row stays blank (not zero).
- Is Solar Performance Cloud independent of my installer?
- Yes — that independence is the point. SPC is not owned by the hardware or the installer whose work it grades, so it has no reason to explain away a fault. It reads every major brand rather than one, and it stays with you long after the installer has moved on.
- How fast does SPC catch a fault?
- It polls every 5 minutes and confirms a fault across 3 consecutive cycles before raising it, so an alert always means something — typically 15–20 minutes from fault to a verified, ranked action. Across the systems we monitor we hold 98% uptime, and the other 2% is seen the day it happens.
- Which inverter brands does SPC read?
- It is multi-brand: Huawei, Solis, Growatt, Sungrow, Canadian Solar, Solarman, GoodWe and more, all on one screen. It connects to the inverters you already own — no new hardware and no lock-in.
Go deeper on solar performance inspection
The method, the standards and the faults behind what Solar Performance Cloud shows you.
- What string-level monitoring catchesThe faults SPC surfaces that a single plant total hides.
- How solar performance inspection worksThe engineer-verified method behind every score and alert.
- Independent inspection vs monitoring softwareWhat a dashboard alone leaves on the table versus verified findings.
- How to find a bad solar stringIsolating a dead or weak string down to the source.
- IEC 61724 & 62446 explainedThe performance and commissioning standards SPC measures against.
- Why is my solar producing less?The real reasons output drops in Pakistan — and how to tell which is yours.
Find the string that’s costing you
If your solar is watched only by the inverter app, you don’t yet know whether every part of it is running. Let our engineers show you the string-level picture.