Solar monitoring: how to spot underperformance early (without obsessing)

Solar monitoring is useful for one thing: catching real problems early.

It becomes unhelpful when it turns into daily anxiety about normal weather swings.

This guide gives you a simple UK-appropriate routine that finds the faults that matter, without turning you into a full-time solar analyst.

If you want a quick postcode-level baseline for what “normal” annual output looks like where you live, start here: Find your postcode

If you have not read these yet, they are the foundations for interpreting your numbers:

• Cloud cover vs solar output: what actually happens on overcast UK days
• Temperature losses in UK solar panels: how heat quietly eats your kWh

Quick answer: the monitoring habit that works

Assumptions and variability

• We talk about power in kW and energy in kWh (see Glossary).
• We assume a typical UK grid-connected PV system with an inverter and a monitoring app or portal.
• We assume you want practical fault detection, not perfect modelling.
• What varies most between homes: shading, roof orientation and pitch, inverter sizing (clipping behaviour), and local weather patterns.
• If you want the modelling assumptions behind SolarByPostcode pages (yields, rates, and savings calculations), see: Data sources and methodology

What monitoring can and cannot tell you

Monitoring is good at catching:

• the system being off (breaker, isolator, inverter fault)
• a string or module issue that creates repeatable dips
• a long-term step-change down after a change (tree growth, new obstruction, wiring issue)
• export caps or clipping patterns you did not realise were happening

Monitoring is not good at:

• telling you anything from a single day in the UK
• proving your system is “underperforming” because last Tuesday was higher
• replacing a proper investigation when there is a real fault

To keep monitoring useful, you need two frames:

1) shape checks (clear day curve shape)
2) trend checks (a week or two versus an expected band)

1) The fastest diagnostic: curve shape on clear days

On a clear day, your PV power curve is a fingerprint.

You do not need to memorise numbers. You need to recognise shapes.

Figure: On clear days, your curve shape is a quick “sanity check”. A smooth bell curve is what you want. A flat top is usually inverter clipping (not a fault, just a cap). A notch or dip that repeats at the same time on multiple clear days is the pattern worth investigating, because it often points to shading, a string issue, or a specific constraint.

Why “shape” beats “numbers” most of the time

On most UK days, clouds dominate your totals.

That makes a single day’s kWh a noisy signal.

But on a clear day, the curve should still look like physics:

• rises as the sun climbs
• peaks around solar noon (often slightly later)
• falls as the sun drops

When it does not, the reason is often discoverable.

2) The most common “false alarms” in monitoring

Before we talk about real faults, it helps to name the things that look like faults but usually are not.

Cloud “texture” can look like a fault

UK cloud cover often creates:

• jagged curves (fast-moving broken cloud)
• flat-ish plateaus (thin high cloud)
• multiple mini-peaks (sun breaks)

That is normal. It is not “underperformance”.

Read if you want the intuition:

• Cloud cover vs solar output: what actually happens on overcast UK days

Heat can mute peaks

Hot panels produce less power.

In the UK this matters most on bright summer days.

Read:

• Temperature losses in UK solar panels: how heat quietly eats your kWh

Clipping is often deliberate

Clipping sounds bad, but it can be an intentional design choice.

If your inverter is smaller than your panel array, peaks can flatten.

This is often acceptable because the lost energy is limited to the sunniest hours, and the trade can reduce system cost.

If you want to understand the components behind this properly:

• Inverters explained (UK): what matters, what doesn’t, and what gets oversold

3) The weekly check that prevents obsession

The simplest healthy monitoring habit is a weekly check.

You compare your recent days to an expected seasonal range.

You do not chase a single day.

Figure: A simple “don’t obsess” routine is to compare your daily kWh to a normal seasonal range, not to the best day you have ever seen. Weather creates noise. A real issue usually looks like a sustained drop below the band over multiple days (especially when nearby days are clear), not one disappointing day.

How to create your “expected band” without spreadsheets

You have three easy options:

1) Use your monitoring history from the same season last year (best option once you have it).
2) Use your postcode baseline as the annual anchor, then think seasonally (UK winter will always be lower): Find your postcode
3) Use a simple rule: compare the last 7 to 14 days against the previous 7 to 14 days, and look for a step-change, not noise.

The point is not precision.

The point is detecting a meaningful change.

Table 1: Underperformance signals that are worth caring about

Signal	What it often means	What to do first
Several days are far below “normal” for the season	A real step-change (fault, disconnection, configuration issue)	Check inverter status, breakers/isolators, and whether the system is reporting correctly
Repeatable notch at the same time on clear days	Shade, string issue, or a consistent constraint	Compare multiple clear days and note the time window; identify likely shading objects
Flat top at peak on clear days	Inverter clipping (often normal by design)	Confirm your inverter rating and whether clipping was expected
One day is terrible, the next day is fine	Almost always weather	Do nothing; look for repeatability before you act
Long-term slow decline	Often normal (seasonality), sometimes equipment ageing or growing shade	Compare same month year-on-year; consider shading growth and cleanliness only if the drop is meaningful

4) A simple routine: what to check, and how often

The best monitoring routine is the one you will actually keep.

Here is a sensible default:

Weekly (5 minutes)

• glance at daily kWh for the week and compare to your expected seasonal band
• if you see a sustained drop, pick one clear day and check the curve shape
• log one sentence: “normal”, “watch”, or “investigate”

Monthly (10 minutes)

• compare this month to last month (seasonality-aware)
• check whether clipping behaviour is consistent (if you clip)
• check whether any new shading objects exist (tree growth, new scaffold, new structure)

After storms or electrical work (2 minutes)

• confirm the inverter is on and reporting
• confirm generation is non-zero on the next daylight period

Monitoring daily is optional. For most households, it reduces happiness more than it increases value.

5) The top causes of real underperformance in UK homes

Inverter off, tripped breaker, or isolator switched

This is the boring one, and it is common.

If your system shows flat zero at times it should not, start here before you assume something expensive broke.

A string or connection issue

These can show up as:

• a step down in peak power
• a persistent reduction even on clear days
• or a weird repeatable shape

If you suspect this, document:

• 2 to 3 clear days
• screenshots showing the curve
• the time and date

That makes installer support dramatically easier.

Shading you did not account for (especially winter)

New shade sources happen:

• trees grow
• neighbours extend
• a new vent or aerial appears
• panel soiling patterns change near chimneys

If you want the shading mental model:

• Shading and solar panels: when a single tree really does matter

Export caps and grid constraints

Export limits affect export, not generation. But they can still confuse monitoring if you are looking at the wrong metric.

If you are in a situation where export is capped by DNO rules (see Glossary), your system may behave differently to what you expected.

Read:

• DNO, G98/G99, and export limits: why your inverter may be capped

6) What to send your installer if you think something is wrong

Most “my solar is underperforming” messages fail because they are vague.

Here is what actually helps:

• system size (kWp) and inverter size (kW) (see Glossary)
• 2 to 3 screenshots from clear days showing the curve
• the date/time and a note of whether there was clipping or a notch
• a rough description of weather (“clear”, “broken cloud”, “overcast”)
• whether anything changed recently (storm, roof work, electrical work, tree trimming, scaffolding)

That turns your message into a diagnosis aid rather than a complaint.

FAQs

How do I know if clipping is “bad”?

Clipping is often designed in. It shows as a flat top on the sunniest hours. If you only see it on very bright days and the system otherwise looks normal, it is usually fine.

Should I monitor every day?

You can, but you do not need to. A weekly check catches most real issues early enough, and daily checking often just amplifies weather noise.

My curve is jagged. Is that a fault?

Usually not. In the UK, broken cloud causes jagged output. A fault pattern tends to repeat on clear days.

What is the biggest underperformance blind spot?

Repeatable winter shading. People judge solar based on summer, then get surprised by winter behaviour. Shade is more punishing when the sun is low.

Bottom line

• On clear days, curve shape is your fastest diagnostic.
• On normal UK days, weather creates noise. Do not chase it.
• A weekly check is enough for most households.
• Escalate only when a pattern repeats or a sustained drop appears.

Next reads

• Cloud cover vs solar output: what actually happens on overcast UK days
• Shading and solar panels: when a single tree really does matter
• Inverters explained (UK): what they do, what can go wrong, and what “clipping” really means
• Why solar panels never hit their rated output (and why that is normal)
• Solar warranties and insurance in the UK: what is covered, what is not, and who to call first
• Do solar panels increase house value in the UK? The boring truth (and when it backfires)

Run the calculator for your postcode