B-4, Basement, Hans Plaza Ambedkar Road Ghaziabad, U.P. - 201001
Mon-Fri 09:30 AM - 06:00 PM
Short answer: under-performing rooftop solar is almost always a stack of
specific, findable losses — soiling, unmodelled shading, a dead or mismatched string, inverter
clipping, temperature derating, undersized cabling, grid outages during peak sun, and genuine
module degradation. Each one is measurable, and most are correctable.
The difficulty is that none of them announces itself. The plant runs, the inverter shows
green, and the monthly number is simply lower than the projection in the proposal. Here is how
to work through it in order of likelihood.
Monthly kWh is not a diagnosis — irradiation varies. Compare specific yield (units generated
per kW installed) against the same month last year, and against the plant’s performance ratio,
which accounts for available sunlight. A sudden drop points at a fault; a gradual decline
points at soiling, degradation or a developing problem.
| What You Observe | Likely Cause | How to Confirm |
|---|---|---|
| Gradual decline over weeks, recovers after rain | Soiling — dust, pollution, bird droppings |
Compare output before and after a cleaning cycle |
| Output dips at the same time each day | Shading from a structure, pole, parapet or new construction |
Site walk at that hour; string-level data |
| One inverter or MPPT consistently lower | Failed string, blown fuse, connector fault, module failure |
String current measurement and IR testing |
| Output flat-tops at midday | Inverter clipping — DC capacity exceeds inverter limit |
Check DC-to-AC ratio against the inverter rating |
| Lower generation in peak summer | Temperature derating — module output falls as cell temperature rises |
Compare winter and summer specific yield |
| Loss worse at high output | Undersized DC or AC cable — losses rise with current |
Voltage drop measurement at full output |
| Zero generation in blocks during the day | Grid outage, inverter trip on grid parameters, or anti-islanding |
Inverter event log |
| Small, steady annual reduction | Normal module degradation | Year-on-year specific yield trend against warranty curve |
Dust, construction activity, industrial emissions and bird droppings accumulate quickly, and
the loss is continuous rather than dramatic. Cleaning frequency should be set by site
conditions, not by a generic schedule — a plant next to a cement unit or an unpaved road needs
far more attention than one in a clean coastal area. Bird droppings matter more than their area
suggests, because a fully blocked cell can force a bypass diode to conduct and knock out a
whole sub-string.
Shadow analysis is done at design stage, against the site as it was. Then a water tank
arrives, a neighbour builds a floor, a chimney is extended, or a row of panels is added too
close to the one in front. Because strings are series circuits, a small shadow on one module
can affect output far beyond its own area — which is why shading complaints often sound
exaggerated until the data is examined.
A plant monitored only at the inverter output can lose a string for months without anyone
noticing — the number simply looks a little low. String-level or module-level monitoring
converts that invisible loss into an alarm. If your plant was built without it, periodic string
current measurement with a clamp meter under stable sun is the manual equivalent and should be
part of the O&M routine.
A performance assessment should end with measured string currents and voltages, IR test
results, thermal imaging of modules and junction boxes to find hotspots, inverter event log
analysis, a soiling loss estimate from a cleaning trial, and a loss breakdown that assigns the
shortfall to specific causes with the cost of correcting each. That turns a vague complaint
into a work order.
VTPL executes solar EPC projects and plant rectification works as part of its renewable
energy portfolio, alongside the electrical distribution and protection works these plants
connect into. With 33+ years of electrical EPC experience across 19+ states and clients
including PSUs and industrial groups, our teams approach solar as a power system, not a panel
installation.
Frequency depends on local dust, pollution and bird activity rather than a fixed rule. The
practical method is to measure output before and after a cleaning cycle and set the interval
where the recovered generation justifies the cleaning cost.
The ratio of actual generation to what the plant should have produced given the available
irradiation and installed capacity. It allows fair comparison between months and between
plants, unlike raw kWh.
Module output falls as cell temperature rises, so peak-heat months can yield less per kW
than clear, cool months despite longer sunlight hours. Poor ventilation under the array worsens
the effect.
A faulty string reduces that MPPT’s contribution, and mismatched modules within a string
drag the whole string down. Without string-level monitoring, the loss can continue unnoticed
for months.
Often yes — particularly where the loss is from undersized cabling, failing inverters, weak
surge protection or a lack of monitoring. An audit that quantifies each loss tells you which
corrections pay for themselves.
VTPL delivers solar power
plant EPC, rectification and grid interconnection works across India. Plant not meeting
projections? Request a project consultation
→
