A 3 kWp Rooftop Plant Producing 30 Percent Less Than Promised
In Kolkata, at about 22.6 degrees north, a housing society installs a 3 kWp grid-tied rooftop plant: 8 modules of 375 Wp and a 3 kW string inverter. The modules are tilted at 12 degrees and face 25 degrees west of true south. A 2 m high overhead water tank stands about 6 m to the east-south-east of the array and shades it in the early morning. The quote promised 4,200 kWh in the first year, about 1,400 kWh per kWp. The inverter log shows no error codes, and the highest output ever recorded is 2.6 kW. The modules have not been cleaned since installation. The metered generation over 12 months is 2,930 kWh. The installation cost was Rs 1,80,000 and the net-metering tariff is Rs 6.75 per kWh.
1. Calculate the specific yield and capacity utilisation factor, and use the peak-output record to decide whether this is a hard fault or a distributed loss.
Read the worked answer
Specific yield = 2,930 kWh / 3 kWp = 977 kWh per kWp per year.
Capacity utilisation factor = 2,930 / (3 x 8,760) = 2,930 / 26,280 = 11.2 percent.
Reference for Kolkata: a clean, unshaded, optimally tilted array gives roughly 1,250 to 1,350 kWh per kWp, that is a CUF of about 14 to 15.5 percent. So this plant is running about 25 to 30 percent low.
The peak record is the decisive clue: 2.6 kW out of 3.0 kWp is 87 percent of rated capacity. A plant with a dead string would peak at about half of rated, a plant with one shorted bypass diode or one badly mismatched module at perhaps 60-75 percent, and a permanently derated or clipping inverter would flat-top at a fixed value. Reaching 87 percent proves that every module is connected and the inverter is converting normally.
Conclusion: this is not a hard fault and not a warranty case. The losses are distributed and steady: soiling on uncleaned glass, the morning tank shadow, a shallow 12 degree tilt that lets dust and water sit and gives a poorer winter angle, and possibly some cable and mismatch loss. The next step is cleaning the array and logging again before calling the installer.
2. Quantify the shortfall against the promise and judge how much of it is an over-promise rather than a site loss.
Read the worked answer
Shortfall = 4,200 - 2,930 = 1,270 kWh, which is 30 percent of the promise.
First, test the promise itself. 1,400 kWh per kWp for Kolkata is at the very top of the realistic band; a fair design figure would be about 1,300, giving 3,900 kWh. So roughly 300 kWh of the 1,270 is an optimistic quotation and should not be used to judge the hardware.
Now attribute the remaining 970 kWh:
- Soiling: modules never cleaned for 12 months in a dusty urban area typically lose 5 to 10 percent, say 8 percent, about 310 kWh.
- Morning tank shadow: the tank is 2 m high at 6 m, so it blocks the sun below about 18 degrees elevation, which is the whole early-morning period. A shadow on even one module in a series string drags the string current down. Say 6 percent, about 230 kWh.
- Tilt: 12 degrees instead of about 22 degrees costs roughly 2 to 3 percent, say 90 kWh. The 25 degrees west azimuth is close to neutral for annual yield.
- That leaves about 340 kWh, roughly 8 percent, unexplained. Candidates: cable and connector losses, module mismatch between 8 modules, inverter efficiency at low load, grid outages when the inverter cannot export, and meter accuracy.
So about a quarter of the gap is an over-promise and the rest is real, recoverable site loss, mostly cleaning and shading.
3. Recommend corrective actions in order of cost-effectiveness, and recalculate the simple payback before and after the fixes.
Read the worked answer
Payback now: 2,930 kWh x Rs 6.75 = Rs 19,778 per year, so Rs 1,80,000 / 19,778 = 9.1 years.
Order of action:
- Clean the modules with water and a soft brush in the early morning, never a dry cloth or a hard scraper. Cost is nearly zero and it should recover 5 to 8 percent, roughly 200-250 kWh. Do this first, then re-measure for a full month before spending on hardware.
- Deal with the shadow. Either relocate or raise the tank, or fit module-level power electronics (optimisers or micro-inverters) on the affected modules so a shaded module does not throttle the whole string. Relocating the tank is cheapest if the plumbing allows; optimisers cost money but recover the morning hours.
- Raise the tilt from 12 to about 22 degrees, at the same time as the cleaning, and check that the array is not self-shading at the new angle.
- Review the inverter and generation logs monthly against a nearby reference plant, so any drop is caught in weeks, not a year.
After the fixes, using a conservative 3,500 kWh: 3,500 x 6.75 = Rs 23,625 per year, payback Rs 1,80,000 / 23,625 = 7.6 years. If cleaning and shadow removal lift it to 3,800 kWh, payback falls to about 7.0 years.
Note to the society: a 30 percent shortfall against an inflated quotation is not a manufacturing defect, so a module warranty claim is very unlikely to succeed. Keep the IEC test flash reports and the commissioning report, and argue the quotation separately if the installer claimed 1,400 kWh per kWp as a guaranteed figure.