Types of solar panels: which one to choose

The most common confusion when comparing panels is putting on the same level things that answer different questions. "Monocrystalline" or "polycrystalline" say how the silicon was made. "PERC", "TOPCon" or "HJT" say how the cell is built. "Bifacial" says whether the module also captures light from the back. And "half-cut" or "half-cell" says how the cells were cut and connected.

That is why one panel can be monocrystalline, TOPCon, bifacial and half-cell at the same time. It is exactly what the four datasheets reviewed for this comparison are: LONGi's Hi-MO 7, two JinkoSolar Tiger Neo models and Trina Solar's Vertex N. All four are monocrystalline n-type silicon, bifacial and dual glass.

So the useful question is not "mono or bifacial?" but three separate ones: which cell it uses, whether the rear face will receive light in your mounting, and what its temperature and warranty numbers say.

Both are silicon; what changes is how that silicon was formed. In monocrystalline, each cell is cut from an ingot that is a single continuous crystal, with its whole atomic structure aligned. In polycrystalline, the silicon is melted and left to solidify in a mould, forming many small crystals with boundaries between them.

Those boundaries are the source of the difference. They act as obstacles for the electrons freed by light, and part of the energy is lost there. That is why, for the same surface, monocrystalline delivers more power. It also explains the appearance: the single crystal reflects light evenly and looks uniformly black, while the many crystals of polycrystalline produce that characteristic bluish, grainy look.

The price difference that justified polycrystalline disappeared once making monocrystalline ingots stopped being the expensive process it used to be. Today, finding it in a new proposal is more a sign of old stock in a warehouse than of a design decision: ask why it is being proposed and what the manufacturing date of the batch is.

Within monocrystalline there are several cell architectures, and that is what usually appears on the datasheet as an acronym. The first split is the type of silicon in the wafer: p-type, the basis of PERC technology, or n-type, the basis of TOPCon, HJT and each manufacturer's own variants, such as LONGi's HPDC or Trina's i-TOPCon.

PERC was the standard for years: it adds a reflective layer on the back of the cell to make use of the light that passes through it. N-type cells are the next generation. JinkoSolar sums up in its datasheet the advantage that matters most in the long run: lower light-induced degradation (LID and LeTID). The four n-type datasheets reviewed also guarantee a loss of only 0.4% a year for 30 years.

You do not need to master the physical differences. What is useful is knowing that the acronym describes a generation, and that what really changes between generations —temperature coefficient, first-year degradation and annual degradation— is written on the datasheet. Compare those numbers instead of the acronym.

A module's efficiency is the percentage of the solar energy reaching it that it converts into electricity, measured under standard laboratory conditions. In the four datasheets reviewed it ranges from 21.7% to 23.2% depending on the model and power. Between brands in the same range, the difference is one point.

The usual mistake is to read it as a measure of quality. It is not: it is a measure of density. A more efficient panel delivers the same power in less surface, and that is it. If your roof is large, a slightly less efficient panel can give you exactly the same kWh per year for less money.

Where it does rule is on tight roofs. With these datasheets, the modules take between 4.3 and 4.6 m² per kWp, not counting walkways, edges or spacing between rows, which on a flat roof push the figure up considerably. Gaining one point of efficiency cuts that area by around 4%. The right question is not "which one is more efficient?" but "do I have roof to spare or not enough?".

A bifacial panel also generates with the light that reaches it from behind, the light that bounces off the surface below it. The four datasheets reviewed state a bifaciality factor of 80 ± 5%: the rear face, lit like the front, would give 80% of its power. But it is never lit the same, and what reaches it depends on two things that have nothing to do with the panel.

The first is albedo: how much light the surface underneath reflects. A white roof, light gravel or a concrete slab return quite a lot; a dark tile or a brown metal sheet, very little. The second is clearance: if the panel sits right against the roof, hardly any light reaches its rear face however light the surface is.

The datasheets do give gain figures, and they should be read for what they are. LONGi tabulates power with rear-side gains of 5% to 25%, and Trina advertises "up to 10-20%" additional power. These are the manufacturer's reference scenarios, not a forecast for your roof. On a conventional residential roof, with a structure a few centimetres high and dark tile, it is reasonable to expect the low end or less.

Where they do shine is in ground mounts, elevated structures, parking carports, light-membrane roofs and agricultural installations with tall structures. If your installer proposes bifacial panels and shows you a gain, ask them to state what albedo and mounting height they assumed to calculate it.

The datasheet power is measured in a laboratory with the cell at 25 °C. On a Colombian roof at midday the cell runs well above that figure, and every extra degree takes power away.

The temperature coefficient quantifies that loss. The four n-type datasheets reviewed state between −0.28% and −0.29% per degree. With the cell at 65 °C, 40 degrees above the reference, a −0.29%/°C module loses around 11.6% of its power at that moment; a −0.35%/°C one would lose about 14%.

Compare that with what efficiency brings: between a 22% panel and a 23% one there is one point of difference in area needed. Between a good temperature coefficient and a mediocre one, in a hot climate, there are a couple of points of production every sunny day for 30 years. On the coast and in hot lowlands, that parameter deserves more attention than nominal efficiency.

"30-year warranty" means nothing on its own, because a panel carries two different warranties.

The product warranty covers defects in materials and workmanship: in the four datasheets reviewed it is 12 years. The performance warranty is something else: it guarantees a minimum percentage of power over time, expressed as a curve. In these datasheets it is 30 years, with a maximum loss of 1% in the first year and 0.4% a year after that; LONGi translates this into 87.4% of the original power after 30 years.

Two panels advertised with "30 years" can have different curves, and that difference is energy and money. Ask for the curve, not the headline. And check that the manufacturer has representation in Colombia: a warranty that requires shipping the module abroad to claim it is not a practical warranty.

Photovoltaic solar panels of 100 W or more are products covered by the RETIE, whose text in force is that of Resolución 40284 de 2026. That means the specific module installed for you must have a product conformity certificate issued by an accredited body, and the regulation takes the IEC 61215 (design and durability) and IEC 61730 (safety) standards as its reference. All four datasheets reviewed state both.

There are two additional tests that are not mandatory but are relevant depending on where you install. IEC 61701 measures salt-mist resistance and is worth demanding on the coast. IEC 62716 measures ammonia resistance, which is what is in the air of stables, poultry sheds and composting plants. JinkoSolar and Trina state both tests in the datasheets reviewed.

Being on BloombergNEF's "Tier 1" list replaces none of this: it measures how banks accept the manufacturer, and BloombergNEF itself asks for it not to be used as a measure of module quality.

Panel technology affects efficiency and the area you will take up, but the decision also depends on the power per panel, the manufacturer's warranty and the quality of the inverter it will work with.

A very efficient panel with a mediocre inverter or a poor installation will not perform as expected. Evaluate the whole system, not just the panel: the difference between a well-ventilated mounting and one stuck to the roof, or between a well-designed string and one with shading, easily exceeds the difference between two panel models in a similar range. Weight counts too: the dual-glass modules in the datasheets reviewed weigh between 31 and 38 kg each, and the roof has to carry them.

These are the ones that most often lead to paying too much or producing too little.