Did you know that a solar panel can appear to ‘lose efficiency’ during a school experiment simply because a cloud crossed the sky? Pupils might compare two voltmeter readings ten minutes apart, declare the second panel weaker and start explaining a defect that has not been demonstrated. The readings may be correct. The comparison is the problem: the sunlight that reached the panels was not the same.
A fair comparison of solar-panel electrical output measures the appropriate electrical quantity while keeping or recording important conditions such as incoming light, panel angle, temperature, connected load and measurement method. When clouds change during a test, a lower output does not by itself establish that a panel is defective or inherently less efficient. The immediate response is to record the changing conditions, distinguish output at that moment from panel performance under matched conditions, and repeat the comparison in a more controlled way.
This article gives parents, learners and teachers a concrete diagnostic, hand-checked voltage–current calculations, a small safe test design, the difference between watts and watt-hours, progressive practice, limitations and useful next steps. All numerical results are invented teaching examples. They do not describe a particular installation, school or pupil.
What is the science behind the changing number?
A photovoltaic (PV) device converts some of the solar radiation reaching it into electricity. The incoming sunlight is not constant: clouds, shadows, atmospheric conditions, time of day, orientation and season all affect what reaches the surface. The US Department of Energy’s overview of solar radiation distinguishes direct sunlight from diffuse radiation and explains how clouds can reduce direct-beam solar radiation. A cloud does not automatically turn the output to zero; diffuse light may still reach the panel.
Solar-panel electrical performance also depends on temperature and the electrical load. The Department of Energy’s PV performance guide explains that power is the product of voltage and current, and that electrical measurements change with load resistance. A panel’s voltage reading by itself is therefore not a reliable full description of useful output power. If different groups attach different loads, the comparison is further confounded.
A student’s first question should be: Were we testing the panel, the weather, or both without realising it? An honest report can answer ‘we observed output under changing conditions’ even when it cannot establish a clean cause for the difference.
A worked example with one panel and one fixed load
Suppose a teacher uses a low-voltage educational solar panel, an appropriate fixed 20-ohm resistive load and safely connected measurement instruments. The apparatus is supervised and rated for the expected voltage and current. On a brighter interval, the observed voltage across the load is 3.6 volts (V) and the current through it is 0.18 amperes (A). During a cloudier interval, voltage is 2.8 V and current is 0.14 A. Assume the load remains close enough to 20 ohms for these hand-checked teaching figures, and that the two readings in each interval were taken together.
- Bright interval: power = voltage × current = 3.6 V × 0.18 A = 0.648 watts (W).
- Cloudier interval: power = 2.8 V × 0.14 A = 0.392 W.
- Difference: 0.648 W − 0.392 W = 0.256 W less power in the second reading.
- Relative reduction against the first reading: 0.256 ÷ 0.648 × 100% ≈ 39.5%.
- Consistency check for the fixed 20-ohm load: 3.6 V ÷ 0.18 A = 20 Ω, and 2.8 V ÷ 0.14 A = 20 Ω; both examples are consistent with the same resistance.
- Independent arithmetic check: P = V²/R. For the bright interval, 3.6² ÷ 20 = 12.96 ÷ 20 = 0.648 W. For the cloudier interval, 2.8² ÷ 20 = 7.84 ÷ 20 = 0.392 W.
These are two different observed output powers under changing outdoor conditions. They are not measurements of the panel’s conversion efficiency, because we have not measured the incoming solar power over its area and other information required for a defensible efficiency estimate. They also do not prove that the cloud was the only cause: panel temperature, orientation, partial shadows and contact issues may change at the same time.
A student who calculates the 39.5% decrease correctly has completed the arithmetic. The more important scientific step is adding the qualification: power under the second set of observed conditions was about 39.5% below power under the first set. Saying ‘the panel became 39.5% less efficient’ would claim more than these measurements establish.
Why volts alone cannot settle the comparison
A voltmeter measures electric potential difference, not automatically electrical power. To determine power delivered to a load at a moment, we need voltage across that load and current through it, or an appropriately justified equivalent relationship. The electrical operating point depends on the load. For a typical small educational panel, a high unloaded voltage with little usable current is not interchangeable with the power delivered when a device is connected.
An especially risky classroom shortcut is connecting a current meter directly across a panel or power supply without the proper mode, range and test circuit. That can create a short circuit and damage equipment. Pupils should use a teacher-approved low-voltage kit and follow its supplied circuit instructions; never connect equipment to rooftop panels, mains wiring or improvised high-energy systems.
The panel area also matters when making an efficiency claim. Electrical output of 0.6 W from a small panel and 0.6 W from a larger panel are not equivalent conversion-efficiency evidence without accounting for incoming light power and active area. The DOE’s solar performance guidance explains how sunlight resource and temperature complicate comparison with rated performance.
How to diagnose the wrong explanation
Give a learner this fictional statement: ‘Group A measured its panel at 11:00 under bright conditions, and Group B measured a different panel at 11:20 while clouds covered the Sun. A recorded 0.65 W; B recorded 0.39 W. Group B’s panel is therefore worse.’ Ask: ‘What evidence is missing before that conclusion is justified?’
A good answer identifies two separate variables: panel identity and incident light changed together. Other conditions may have changed too. The data show two output readings, not a controlled ranking of the panels. This is a confounding problem: when the conditions differ, we cannot safely assign the observed difference to the panel itself.
An even better learner proposes an improvement: use the same type of load, test panels simultaneously side by side under comparable illumination, log the light conditions and temperatures, swap panel positions where safe and appropriate, then repeat. Simultaneous comparison helps control rapid changes in cloud cover, but it does not guarantee equal incident sunlight if one panel is shaded or tilted differently.
A practical classroom design with fairer comparisons
- 1. Choose one precise question. ‘How does the output of one panel change as illumination changes?’ is not the same as ‘Which of two panels performs better under the same light?’ Write the question before deciding how to measure.
- 2. Use safe, suitable equipment. Select a classroom PV demonstration kit, the manufacturer’s permitted load, rated leads and meters. An adult sets up the circuit; no rooftop, building or mains connections are used.
- 3. Fix the operating conditions that can be fixed. Hold panel orientation and distance from an approved artificial light steady for a controlled model, or use paired panels at matched positions for outdoor observations.
- 4. Record changing variables rather than hiding them. Note cloud cover, sudden shadows, time, orientation, panel temperature where measured, light conditions and the load connected.
- 5. Measure voltage and current correctly. Use the supplied safe circuit to measure potential difference across the load and current through it. Compute power in watts from paired readings.
- 6. Repeat observations. Keep raw readings and label each trial. Do not discard inconvenient cloudy readings simply because they make the story untidy.
- 7. State what the method can support. If sunlight was uncontrolled, report observed associations and limits, not a definitive panel-efficiency comparison.
A light meter can help track changing illumination if it suits the classroom purpose, but an ordinary lux reading is not automatically a solar irradiance reading in W/m². Lux is weighted for human visual response, whereas solar irradiance is radiant power per area. Do not convert between them using a made-up universal multiplier, especially when the light spectrum changes. If the research needs true solar irradiance, use suitable equipment and interpretation.
Second worked example: simultaneous, but still not automatically conclusive
Imagine two classroom panels, A and B, placed side by side under apparently similar light, with each connected to an identical 20-ohm load. At one moment, panel A supplies 3.4 V and 0.17 A; panel B supplies 3.2 V and 0.16 A. For the invented numbers, A’s power is 3.4 × 0.17 = 0.578 W, and B’s power is 3.2 × 0.16 = 0.512 W. Their difference is 0.066 W. Both pairs are consistent with a 20-ohm load.
This does establish that at that measured instant, A delivered 0.066 W more to its identical load than B, subject to the assumed measuring accuracy. It does not automatically prove that A is always the better panel. To investigate that stronger claim, measure more than once, compare illumination and temperature, and consider whether a panel or cable occupies a slightly different position.
A careful pupil can also calculate the relative difference against B: 0.066 ÷ 0.512 × 100% ≈ 12.9%. If the pupil says ‘A was 12.9% more powerful in this observation, using B’s measured power as the baseline’, the baseline is visible. If the pupil says ‘12.9% more efficient’, the missing incoming-light evidence remains a problem.
From watts to energy: a different question
Power in watts is a rate of energy transfer. Electrical energy delivered over an interval depends on how power changes over time. If, in a deliberately simplified experiment, a constant 0.50 W is supplied for exactly 10 minutes, the energy is 0.50 W × 600 s = 300 joules (J). Equivalently, 0.50 W × (10/60) h ≈ 0.0833 watt-hours (Wh). The two values represent the same energy in different units.
But sunlight and output may change second by second during a real cloud passage. Multiplying one brief 0.50 W reading by a long duration can overstate or understate actual energy. Record the time variation and use a suitable interval-based or integrating method if the question is energy rather than instantaneous power. This is a useful reason to separate the science question from the easiest number on the instrument.
Progressive practice and independent transfer
Stage 1: multiply the correct quantities
An educational panel supplies 2.0 V across an appropriate load and 0.10 A through it at the same moment. Its delivered power is 0.20 W. Ask the learner to show V × A = W rather than give ‘0.20’ with no unit. Then change the voltage to 2.4 V and current to 0.12 A; power becomes 0.288 W. If the load is fixed at 20 ohms, both pairs are consistent with that resistance.
Stage 2: identify the changing condition
A fictional group tests panel A during sunshine and panel B during heavy cloud. Even if B’s output is smaller, ask for one sentence that separates observation from explanation. An acceptable answer is: ‘Panel B gave a lower output under the measured conditions, but the sunlight differed, so the trial does not isolate panel quality.’
Stage 3: evaluate a ‘fair test’ proposal
A student suggests testing A, then B, then A again, with long gaps outdoors, and taking only the highest reading for each. Ask for two weaknesses. Weather and temperature may drift between measurements, and selecting maxima after seeing results can bias the comparison. Improvement: predefine reading intervals and summary rules, log illumination and temperature, and use simultaneous paired trials where practical.
Stage 4: fresh independent case
The same fictional panel with the same approved load delivers 0.36 W before a cloud and 0.27 W afterward. Calculate the observed drop: 0.36 − 0.27 = 0.09 W; relative to the first reading, 0.09 ÷ 0.36 × 100% = 25%. The learner’s conclusion must include the changing conditions and avoid calling the 25% an efficiency loss. Requiring this explanation checks transfer, not memorisation.
Mistakes that look convincing in a report
‘One number proves a panel rating.’ A momentary power reading is not a laboratory rating. Standardised measurements involve controlled irradiance, device temperature and operating conditions. A school outdoor observation should not be compared uncritically with a manufacturer’s standard-condition specification.
‘Cloud cover is an exact number.’ A label such as ‘partly cloudy’ is a broad observation. It cannot be treated as a precise measurement of solar power arriving at the panel. Use appropriate sensors if quantitative irradiance is required.
‘Lower voltage means lower efficiency.’ Voltage alone does not give delivered power or conversion efficiency. Connected load, current, incoming light and temperature matter.
‘Repeating the same uncontrolled setup makes it fair.’ Repetition helps reveal variation and may strengthen a descriptive account, but if key conditions remain different, repeated observations still do not isolate the cause.
What adults can observe as learning progresses
A parent can ask the child to describe the question in plain language: ‘Are you comparing weather moments or comparing panels?’ A teacher can inspect a pupil’s table for paired voltage-current measurements and contextual notes. A tutor can ask the learner to compute a second case without a worked formula prompt, then justify a cautious conclusion from unfamiliar conditions.
Progress is visible when a learner no longer mistakes volts for watts, can reproduce the fixed-load calculations, recognises cloud cover as a changing condition, and writes a conclusion limited to the evidence. There is no promise of examination scores here. Good science learning is demonstrated by accurate reasoning, sensible measurement and a willingness to revise an unsupported claim.
Frequently asked questions
Does a cloud always stop a solar panel from producing power?
No. Diffuse light can reach a panel under cloudy conditions. Output depends on the amount and type of light reaching it and on other operating conditions; a cloudy sky is not a universal ‘off’ switch.
Can students calculate efficiency from a voltmeter reading?
No. A voltage reading alone is insufficient even for delivered electrical power. An efficiency estimate additionally requires appropriate input-light power and panel information, with a method suited to the device and conditions.
Is simultaneous testing of two panels enough?
It is often an improvement over tests many minutes apart when clouds are moving. Still check matched loads, panel orientation, shading, temperatures, measurement precision and repeatability before drawing a strong performance conclusion.
Why not simply discard the cloudy trials?
If the study’s rules specifically require a defined sunlight range, specify that criterion before collecting results and keep a transparent log of exclusions. Otherwise, discarded cloudy data may hide the very variability the investigation was supposed to understand.
Can we do this safely at home?
Use only a suitable low-voltage teaching kit with clear instructions and adult supervision. Never access installed solar arrays, rooftops, mains wiring or damaged batteries for a child’s experimental task.
Continue the science
Read the related guide to observational studies versus fair tests and how to check carryover between successive experiments. The Core Aim of Education learning hub collects ways to turn examples into independent judgment, while the Clementi tutorial reference explains why diagnosis, clear reasoning and progressive checking support durable learning. A better science report does not merely announce the bigger number. It explains what was held steady, what changed, and exactly what the numbers can support.