5 kWp Rooftop Solar Plant Wanted Signed Off in a Hurry
A 5 kWp grid-tied rooftop plant on a school building in Guwahati uses twelve 415 Wp modules in a single string, a 5 kW single-phase inverter with an MPPT window of 120–500 V, and 6 mm² copper DC cable with a 38 m one-way run. On a clear May afternoon at an irradiance of about 950 W/m² the inverter display reads 320 V, 6.5 A and 2.1 kW. The string open-circuit voltage measured at the same time is 390 V, while the module datasheet gives Voc 37.8 V per module. A water tank on the roof shades two modules from about 11:00 to 13:30. Earthing measured by fall-of-potential is 12 Ω. The installer wants the plant signed off and the net meter sealed the same week. You are the vocational student assisting the inspection.
1. Quantify the shortfall against expectation and rank the causes using the measurements. Is the DC cable a significant contributor?
Read the worked answer
Expected DC power ≈ 4.98 kWp × 0.95 (irradiance) × 0.82 (temperature, soiling, mismatch, inverter) ≈ 3.9 kW. Measured DC = 320 × 6.5 = 2.08 kW, about 47% short.
DC cable check: loop length 76 m, 6 mm² copper at ~2.87 Ω/km gives 2.87 × 0.076 ≈ 0.22 Ω; at 6.5 A the drop is 1.4 V, roughly 0.4%. The cable is not the cause.
Ranking of causes:
- Shading of two modules, 11:00–13:30 - in a series string the shaded modules limit the whole string's current, which is exactly what 6.5 A (against an expected ~11 A) shows. This alone accounts for most of the loss and coincides with the time of the reading.
- String voltage too low for the array: expected Voc = 12 × 37.8 = 453.6 V, measured 390 V, a deficit of about 64 V - close to two modules' worth. This is not explained by open-circuit shading alone and points to a bypass diode conducting, a module interconnection open or a module of wrong rating in the string. It must be tested module by module.
- Soiling and mismatch - minor compared with the above but measurable.
A string I-V curve test, plus per-module Voc, will separate 2 from 1.
2. Which findings must be corrected before sign-off, and which are performance issues to be recorded rather than safety blockers?
Read the worked answer
Blockers - correct before sign-off:
- Earthing resistance 12 Ω is far above the usual requirement of ≤ 5 Ω for the array frame, inverter and lightning arrester. Add an electrode with backfill and re-test before the meter is sealed.
- String anomaly (390 V Voc) must be explained and rectified; an open interconnection or failed bypass diode is a latent fault that will also be a shock/fire risk under reverse current.
- Labelling and isolation: DC and AC isolators identified, warning labels at the meter and inverter, anti-islanding test carried out and recorded, net meter test done.
- Cable protection and routing: DC cable in UV-rated conduit, AC cable in conduit with MCB/RCCB and surge protection as per design.
Performance issues - record, inform, do not block:
- The 11:00–13:30 shading is a design defect against the client's generation expectation. Record it in writing, revise the estimated annual yield, and agree on relocation of the modules or the tank; do not let the client believe the plant is faulty.
- Soiling: set out a cleaning schedule.
Signing off with 12 Ω earthing or an unexplained string voltage would transfer a known defect to the school.
3. State the safety rules that govern DC-side troubleshooting on this plant and explain why each matters.
Read the worked answer
- The array side is live whenever there is light, even with the inverter and AC isolator off. Opening the DC isolator does not de-energize the string.
- Cover the modules with an opaque sheet before opening any MC4 connector; this is the only practical way to make the string safe to touch.
- Never open a DC connector under load - a 5 kW array at 400 V can sustain a DC arc that does not self-extinguish at a current zero, because DC has none. Isolate, then separate.
- Use 1000 V DC-rated meters, insulated tools and gloves, check the meter range before every measurement, and never short a string to test it - a short-circuit test applies full array current through the meter.
- Wait for the inverter's DC-link capacitors to discharge (typically 5 minutes after shutdown) before opening the inverter cover; the display going dark does not mean the capacitor is empty.
- Work in pairs, one hand where possible, and keep the work area dry - a rooftop at 400 V DC is not a place for improvisation.
- Record every test (per-module Voc, IR of the DC side to earth) before re-energizing.