UPMSP High School Solar Repair common mistakes
UPMSP High School Solar Repair · Understand the traps and how to avoid them.
Choosing a battery capacity for a small solar home system from the load list alone, without dividing daily energy demand by depth of discharge and then by the system voltage to get ampere-hours, so a 12 V system ends up with a battery far too small for a 3-day autonomy requirement.
Candidates memorise the formula for daily energy in watt-hours but stop there, because autonomy days, depth of discharge and inverter losses feel like optional refinements rather than mandatory steps in the sizing chain.
What to do instead
Practise the full chain in one line: daily Wh ÷ inverter and wiring efficiency = delivered Wh; divide by allowable depth of discharge (typically 0.5 for flooded lead-acid, 0.8 for lithium) to get required storage Wh; divide by autonomy days' multiplier and by system voltage to get Ah; then round up to the nearest standard battery size and check the charge current stays between roughly 5 and 15 percent of capacity in Ah.
Treating the charge controller as a pass-through and setting only one voltage threshold, so the controller never disconnects the load at low battery or never equalises a flooded battery, and the candidate cannot say why a battery that reads 12.6 V at rest fails to run a load.
Candidates learn nominal voltages (12 V, 24 V) as labels and never build a mental table of real rested open-circuit values, so they cannot distinguish a surface charge, a healthy rested battery, and a deeply discharged one.
What to do instead
Memorise the rested open-circuit table for lead-acid: about 12.7 V full, 12.2 V at 50 percent, 12.0 V at 25 percent, 11.8 V or lower effectively empty; and for a 12 V system the low-voltage disconnect typically sits near 11.5 V under load and reconnect near 12.5 V. Rehearse the three controller setpoints as a set: bulk-to-absorption, absorption-to-float, and low-voltage disconnect, and state what each protects.
Diagnosing a low-output array by replacing panels or the inverter before measuring open-circuit voltage and short-circuit current of each string separately, with the array isolated and the irradiance noted.
A single failed bypass diode, a partial shade patch, or one corroded MC4 contact produces the same symptom (reduced system output) as a failing inverter, and candidates jump to the expensive component because it is the one with a display.
What to do instead
Always follow the divide-and-conquer order: measure string open-circuit voltage against the expected value from the module datasheet times the number of series modules, corrected for cell temperature; then measure short-circuit current per string with a clamp meter and compare strings to each other under the same light. Only after strings are matched and verified do you test the inverter input.
Sizing and wiring a module-to-controller run by current-carrying capacity only, ignoring voltage drop over the round-trip distance, so a 12 V system with a 20 m cable run loses several volts and the controller never reaches absorption voltage.
Candidates default to the ampacity table because that is the table they practised, and voltage drop is treated as a refinement for mains wiring rather than a hard constraint in low-voltage DC systems where a 1 V drop is nearly 10 percent of the source voltage.
What to do instead
Compute voltage drop explicitly: drop = 2 × length × current × resistivity ÷ cross-sectional area, and keep it under about 3 percent for the array-to-controller and controller-to-battery runs, tightening to 2 percent on 12 V systems. If the calculated area exceeds the ampacity-based choice, the voltage-drop result governs; then re-check that the chosen area still satisfies ampacity and the fuse or breaker rating.
Describing earthing as one wire and skipping the distinction between equipment earthing, system or neutral bonding, and lightning or surge protection, so the answer conflates a DC negative bond with an AC earth-neutral link.
The three functions share the word earth in everyday speech, and candidates have often seen only one of them installed, so they generalise from that single example.
What to do instead
Write the three functions as separate labelled items before answering: protective earthing of exposed metalwork to a rod or mat; system bonding of the DC negative or the AC neutral at exactly one defined point; and surge or lightning protection via a dedicated device with the shortest practical path to earth. State that multiple neutral-earth bonds on the AC side cause circulating currents and nuisance tripping, and that a single bond point is the rule.
Reading module ratings as if they were delivered power, so a 100 W panel is assumed to give 100 W to the battery, and a candidate accepts a measured 60 W from a new panel as normal without checking the rating conditions.
The nameplate wattage is a single number with no visible conditions attached, and standard test conditions (1000 W/m², 25 °C cell temperature, AM 1.5) are abstract until the candidate works a real derating example.
What to do instead
Build a derating chain and quote it: real irradiance below 1000 W/m², cell temperature well above 25 °C costing roughly 0.3 to 0.45 percent per degree for crystalline silicon, soiling, mismatch, wiring, and controller and battery charging losses. Expect roughly 70 to 80 percent of nameplate at the battery in a well-designed small system, and compare field measurements against that figure, not against the nameplate.
Answering mounting and site questions by tilt angle alone, without mentioning azimuth, shade survey across the whole day and year, roof structure and wind loading, or the clearance needed for airflow behind the modules.
Tilt angle is the one number candidates can recall quickly, and the other factors are invisible in a written question unless the candidate actively lists them.
What to do instead
Use a fixed checklist for every site question: latitude-based tilt with seasonal adjustment noted, true south or north azimuth depending on hemisphere, shade survey at the low-sun months and at the hours that matter, structural fixing into rafters or purlins rather than just roof sheet, wind uplift, and a rear clearance that lets air move. Mention that a shade survey at noon in summer misses the worst winter shading.
Testing an inverter or controller with the DC input live while holding probes by hand, or working on a string without confirming that the DC isolator has been opened and the capacitors have discharged, and describing the array as safe because the AC breaker is off.
Candidates transfer the domestic AC habit of switching off the breaker to DC systems, where the array cannot be switched off at source and the inverter holds stored charge after isolation.
What to do instead
State the isolation sequence every time: open the AC breaker, then the DC isolator, verify with a meter at the terminals, wait the manufacturer's discharge time (often 5 minutes) before opening the inverter, and treat the array side as permanently live in daylight. Add the personal protective equipment and insulated tool requirements, and note that DC arcs do not self-extinguish at a zero crossing the way AC arcs do.
Writing a preventive maintenance schedule as a single annual visit, with no distinction between cleaning frequency in dusty versus coastal versus bird-heavy locations, no torque check on terminations, and no recorded baseline for performance comparison.
Maintenance is the least technical domain, so candidates compress it into a generic sentence and lose the marks attached to interval reasoning and record keeping.
What to do instead
Structure the answer by interval and trigger: visual and cleaning checks at a frequency set by soiling rate, with a performance-ratio check after each clean; torque and thermal check of terminations at commissioning and then periodically; battery electrolyte and specific gravity checks on a schedule matched to the battery type; and a logged baseline of daily yield so degradation and soiling losses can be separated from faults.
Explaining a small solar appliance fault, such as a solar lantern or a solar water pump that will not start, by blaming the appliance itself, without first checking the panel orientation and shading, the battery state of charge, and the switch or connector between them.
Small appliances look like sealed consumer products, so candidates treat them as a single unit and skip the system-level checks they would apply to a larger installation.
What to do instead
Apply the same source-to-load sequence at small scale: measure panel open-circuit voltage in good light, then battery voltage at rest and under load, then continuity through the switch, connector and cable, and only then test the appliance on a known-good supply. Note that a pump with a low battery may hum and stall rather than fail outright, which points at the battery or controller rather than the motor.