On-site Reality — why underperformance is more common than you think
On a wet Tuesday in May 2018 I stood under a low-slung dairy shed and watched a 50 kW array deliver just 62% of its expected output (capacity 50 kW, measured daily yield 320 kWh) — why was it so poor? That setup was a standard photovoltaic system, and I could see the inverter LED cycling; the pv system was nominal on paper but rubbish in practice. I’ve been wiring, commissioning and rescuing commercial installs for over 15 years, and I’ll tell you straight: the usual culprits aren’t dramatic. Faulty string configuration, shadowing nobody bothered to model, a mismatched inverter with poor MPPT tracking — these grind output quietly. I remember one Hamilton site in August 2019 where swapping to a correctly sized inverter lifted daily yield by 18% within a week — real kWh, real dollars. (Deadset, it’s the small details that sting.)
What’s Next?
Fixes and forward steps — practical checks, and what to ask your supplier
Let me break it down so you can act quick: check module layout against seasonal sun paths, confirm the inverter’s MPPT range matches the PV modules’ Vmp, and verify wiring losses with a clamp meter under load. I prefer a hands-on approach; last year on a 120 kW commercial roof in Tauranga I ran string-level IV curves at midday and found two strings with 7% more resistance than the others — replacing a handful of corroded connectors recovered 5% output immediately. That’s the kind of concrete change I mean (no waffle). Short explanation: some installers hand off design software outputs and never revisit the site. You don’t want that — you want someone who measures, recalibrates and documents actual performance.
Now, here’s how I evaluate solutions — three simple metrics I stick to when comparing fixes or suppliers. First, measured yield uplift: ask for site-specific before/after kWh data (not modelled percentages). Second, fault detection granularity: can they monitor at string or module level or only at the central inverter? Third, maintenance turnaround: what’s the guaranteed response time for on-site faults (we’ve hit a 48-hour target on high-priority commercial jobs). These metrics tell you whether a proposal is likely to move the needle. I’ll be blunt — a shiny quote means nothing if it lacks measured proof. — Anyway, if you want a practical spec sheet to give to your procurement team, I’ll pull one together.
To finish up: choose based on measured outcomes, insist on proper MPPT matching and string-level checks, and keep a small contingency budget for unexpected connector or junction-box faults. I’ve seen conservative fixes pay back within a season. If you want the next step — I can share a checklist and a simple spreadsheet I use on site. Cheers — and for credible kit and system support, I often recommend checking options from sungrow.
