How HJT and TOPCon solar panels work in Indian conditions

From photons to usable electricity

Photons with sufficient energy create electron–hole pairs in silicon. Selective contacts collect the charges, cells in series build voltage, and the inverter tracks the array’s maximum-power point. Some light is reflected or lost as heat, so electrical output depends on the actual spectrum, irradiance and cell temperature.

Light absorption creates mobile charge carriers. Recombination removes carriers before they contribute current, particularly at surfaces and unpassivated metal contacts. A well-designed contact conducts one carrier type efficiently while suppressing recombination or transport of the other. The cell’s current–voltage curve moves with illumination and temperature. Short-circuit current is strongly related to irradiance; open-circuit voltage varies more slowly with irradiance and typically falls as temperature rises.

At any operating point P = V × I. The maximum-power point occurs at the best product of voltage and current, between open circuit and short circuit. The fill factor is FF = Pmax/(VocIsc). Resistive loss depresses fill factor at higher current, and poor shunt resistance can be especially important in low light. After modules are connected to a string, an inverter’s MPPT seeks a useful operating point. Partial shade and bypass diodes can create more than one peak. AC delivered to a house also reflects inverter efficiency, clipping, standby use and wiring loss. PVWatts technical manual

The distinction between energy and power matters. A 5 kWp array is rated at a test point, and a meter reporting 3 kW shows an instantaneous condition. A day’s energy integrates power over time. Specific yield in kWh/kWp makes different array capacities more comparable when all systems use the same reference nameplate convention and measurement boundary. An hourly graph reveals whether low output came from shade, a weak morning sun, inverter clipping or downtime.

HJT, HDT and TOPCon contact structures

HJT commonly places intrinsic hydrogenated amorphous silicon on crystalline silicon to passivate its surfaces, then uses doped selective layers and transparent conductors. TOPCon uses an ultrathin silicon oxide with doped polysilicon. Devsol describes HDT as its heterojunction implementation, with exact construction tied to the model.

In a typical silicon heterojunction cell, the intrinsic amorphous silicon layer reduces surface recombination at the interface. Thin doped layers help create electron- and hole-selective contacts. A transparent conductive oxide (TCO) collects current laterally to the grid and still admits light. The layer thicknesses, hydrogen content, conductivity and optical absorption are jointly optimised. For instance, a front layer that improves passivation but absorbs more short-wavelength light can reduce the collected blue response. The useful balance is quantified by voltage, fill factor and external quantum efficiency. HJT research

TOPCon generally separates the crystalline wafer from a heavily doped polycrystalline silicon contact with a very thin silicon oxide. Carrier transport can involve tunnelling and, depending on oxide formation and annealing, other local transport paths. The oxide can reduce recombination and the polysilicon layer contributes selectivity. A typical n-type TOPCon design uses a rear electron-selective contact and a front junction/passivation stack. Production variations change the exact front and rear response, so a topology name gives an explanation of the concept rather than an exact process recipe. Feldmann et al. TOPCon contact study

Conventional P-type PERC uses rear dielectric passivation with local metal contacts. This improves on a full-area rear contact but leaves a different recombination and contact design than an HJT or TOPCon example. Performance gaps depend on model and test conditions. The Devsol 740 W HDT listing and the Devsol 610 W glass-to-glass TOPCon listing identify two products to compare. Use matching datasheets for watts, dimensions, temperature coefficient, current and installation limits. Devsol 740 W HDT Devsol 610 W TOPCon Devsol HDT technology

MBB, zero busbar and microcracks

Multi-busbar and zero-busbar interconnect designs reduce collection distance or change how current is collected. Multiple contacts can improve tolerance to some cracks. Glass, encapsulant, ribbon and mounting design still determine mechanical durability. Neither MBB nor 0BB makes a panel immune to microcracks.

Cells use fine conductive fingers and interconnects to move current from the semiconductor to adjacent cells. More collection points can shorten the path along a finger, lowering series resistance. Round wires can change optical reflection and effective shading. Half-cut layouts reduce current in each cell segment and can reduce resistive losses, while parallel substrings respond differently to partial shade depending on the circuit and bypass layout. The geometry and assembly process matter more than the number printed in the model name.

Zero busbar usually refers to a cell layout without conventional printed busbars. Interconnect wires or another contact scheme still carry current. Repeated flexing, point loading and transport can crack silicon even if the glass appears intact. A crack that disconnects an electrically active region can reduce output. Electroluminescence imaging, compared under consistent current, is useful for detecting crack patterns before and after transport or mechanical stress. A visual inspection alone misses many electrically meaningful cracks.

For a large module, verify rail position, clamp zones, mounting loads, wind uplift and handling instructions. Two technicians carrying a large module can still twist it if unsupported. Avoid stepping on any part of an installed module. Require the manufacturer’s approved clamp and mounting conditions for warranty and mechanical loading claims. A module that passes type qualification still needs correct on-site support and installation. IEC 61215 Devsol 740 W HDT

A practical yield formula for dust, heat and rear light

Estimate DC power from front irradiance, usable rear irradiance, temperature and soiling. Convert to AC with the inverter curve and apply clipping, wiring and downtime. Sum over time to estimate energy. Use measured or site-specific assumptions for dusty or humid Indian rooftops.

For a simplified model, let Geff = Gfront(1 − sfront) + φ Grear(1 − srear), where φ is bifaciality and s is fractional soiling loss on each side. Then PDC ≈ PSTC(Geff/1,000)[1 + γ(Tcell − 25)]. Over a time step, EAC = Σ PAC,t × Δt with output bounded by inverter capacity and local operating limits. Detailed simulations can model irradiance non-uniformity, spectral and incidence-angle effects, string mismatch, temperature by mounting style and per-hour inverter behaviour. PVWatts technical manual SAM modelling manual

Illustration for one sunny hour. Assume 5 kWp, 900 W/m² front light, 80 W/m² rear light, 5% front soiling, 10% rear soiling, φ = 0.85, 60°C cell temperature and γ = −0.0026/°C. Effective irradiance is 855 + 61.2 = 916.2 W/m². The temperature multiplier is 0.909. Estimated output is 4.164 kW DC, or about 4.00 kWh AC over one hour at 96% conversion efficiency before other losses. That is a worked example, not an operating guarantee. Removing the rear term or changing the dust factors changes the result.

Fine dust often accumulates between rain events and cleaning. Rain may remove loose dust while leaving residue or non-uniform bands. In humid settings, moisture can change optical conditions and make contamination adhere. Long-term risks such as corrosion, edge ingress and insulation faults need different verification from instantaneous yield. Use a clean and soiled reference or repeatable before/after measurement, and record tilt, last cleaning date and recent rain. Do not apply a single national percentage as if it described every rooftop. IEC PV monitoring

How to run an Indian climate comparison

Mount matched modules in the same plane, log front and rear irradiance, temperature and DC power at the same time, and compare specific yield. Record differences in module size, rear clearance, inverter limits, soiling and shading. A single wattmeter reading cannot isolate technology from test conditions.

Start with two hypotheses. At the same irradiance and cell temperature, the coefficient difference predicts a model-specific DC power gap. At the same site over days or seasons, specific energy yield also includes thermal behaviour, low-light response, rear gain, dust and system losses. These require different comparisons. A matched daytime power test may use synchronised I–V traces; an energy comparison requires logger intervals, an outage policy, cleaning records and checks for inverter clipping. To compare heat, record the cell-temperature proxy at a defined module position and the associated uncertainty.

Specify exact model numbers, serials, power bins, mounting geometry, temperatures and measurement equipment. Use calibrated sensors, synchronise clocks and reject incomplete or shaded intervals according to a written rule before calculating an average. A sensor mounted in the wrong plane can create an apparent performance gap. If the panels share one MPPT, the two I–V curves can interact. Matched separate trackers or an equivalent test circuit reduce that confounding factor. Compare kWh/kWp and kWh/m² when both energy economics and roof area matter.

For procurement, request the test file, not only the chart. Retain sample interval, inverter state, maintenance logs and a reproducible calculation. When expected differences are only one or two percent, measurement uncertainty can be as large as the claimed gap. Compare results within defined temperature and irradiance bins and include enough data to see whether an apparent advantage persists across a usable operating range. IEC PV monitoring

What a reproducible rooftop result should report

Report the model, front nameplate, aperture area, irradiance, cell temperature, test duration and measurement uncertainty. Publish both watts per rated watt and energy per square metre when the question is roof use. A higher raw watt reading can reflect higher rated capacity or different light exposure.

For example, 460 W from a 610 W panel and 550 W from a 740 W panel represent 75.4% and 74.3% of nameplate respectively. The larger module produced more watts in that illustrative instant, while the smaller retained a greater fraction of its rating. Neither observation identifies the cause. An irradiance difference of 3% could move either ratio by more than the gap. Measure each module plane and record temperature rather than applying a single common correction to panels exposed to different conditions.

For a longer comparison, export an interval table with timestamp, front irradiance, rear irradiance where relevant, DC voltage, DC current, cell-temperature proxy, inverter state and exclusions. Report the fraction of missing data, and distinguish a lost interval from a measured zero. Inspect whether conclusions change when clipped, shaded, unusually dusty or recently cleaned hours are analysed separately. With these records, a buyer can reproduce the calculation and decide whether the conditions resemble the intended Indian site. IEC PV monitoring

First-party demonstrations and what they show

These public videos document discussions and selected hardware demonstrations by Devsol founder Deepak Chaudhary. The watch counts are audience metrics. A matched laboratory claim requires identified samples, calibrated instruments, test conditions and raw results.

Frequently asked questions

These answers cover exact models, engineering limits and the checks needed before installation.