Why Does Bifacial Module Current Plateau Under Strong Rear Irradiance

When the rear side of a bifacial module receives more light evenly, Isc should normally rise with it. With 80% Isc bifaciality, every additional 100 W/m² of rear irradiance adds about 80 W/m² of front-equivalent irradiance. For a 14 A module at 1,000 W/m² front irradiance, that works out to about +1.12 A Isc for every additional 100 W/m² of uniform rear irradiance.

If the measured current becomes flat instead, do not assume the cells have simply reached their limit. First find out where the flat line appears: Isc, Imp, the inverter MPPT, or the measuring equipment. Those four situations point to very different causes.

The calculation uses:

G_eff ≈ G_front + φ_Isc × G_rear

This is a first estimate, not a guaranteed field result. IEC TS 60904-1-2:2024+AMD1:2026 covers bifacial I-V measurement and defines bifacial nameplate irradiance at 1,000 W/m² front and 135 W/m² rear.^[1]

Use Isc as the First Check

If you want to know whether extra rear light is actually creating extra current, start with Isc.

Tongwei's TNC-G12R bifacial cell data show normalized Isc values of:

The pattern is clear: when irradiance changes, Isc changes with it.

Take a module operating at about 1,000 W/m² on the front. If uniform rear irradiance increases from 100 to 300 W/m² and bifaciality is 80%, effective irradiance changes from:

1,080 W/m² → 1,240 W/m²

For the 14 A example, estimated Isc changes from:

15.1 A → 17.4 A

That is a 2.3 A difference. If the real measurement stays at 15.1 A, the missing response has to be somewhere: the rear light may not be uniform, the tester may be limiting current, the measurement may be wrong, or the module may have a problem.

Compare Isc With Imp

Isc and Imp are both current values, but they do not tell you the same thing. This distinction matters when a system appears to stop gaining current.

Here:

• Isc rises from 15.0 to 18.0 A: +20%
• Imp rises from 14.2 to 15.5 A: about +9.2%

The cells are still responding to the extra light because Isc keeps increasing. What is flattening is the current available at the maximum-power operating point.

Isc rising + Imp flattening = not intrinsic current saturation.

At this point, the practical checks are Vmp, Pmax, mismatch, inverter control and high-current losses.

Measure More Than One Rear Point

A rear sensor can be perfectly accurate at its own position and still give a poor picture of the whole module.

Consider three readings:

The simple three-point average is:

(250 + 380 + 160) ÷ 3 ≈ 263 W/m²

If the system uses only the center sensor, it treats rear irradiance as 380 W/m² instead of approximately 263 W/m².

That center reading is about:

(380 − 263) ÷ 263 × 100 ≈ 44%

higher than the three-point average.

Three sensors still do not give a perfect area-weighted rear measurement, but the example shows the size of the error one sensor can introduce.

NREL measurements have found rear irradiance gradients across bifacial modules and differences near row edges, confirming that sensor location matters.^[2]

Tongwei's rear-irradiance and rear-current analysis shows the same practical issue: similar rear sensor values do not guarantee similar module current.

Check Rear-Irradiance Spread

The average value alone can hide a large difference between bright and dark areas behind the module.

Both modules average 250 W/m². On paper, they look the same.

They are not.

Module B has an irradiance spread:

340 ÷ 40 = 8.5 times larger

than Module A.

That matters because cells in the same electrical path cannot each run at their own ideal current. A bright area may be capable of more current, but a darker area can hold the shared circuit back.

NREL modeling confirms that rear-side irradiance non-uniformity can create additional bifacial mismatch loss.^[3]

Check the Module Circuit

Rear-light distribution only tells half the story. The other half is how the cells are wired inside the module.

Modern bifacial modules can use half-cut cells, several substrings, parallel current paths and multiple bypass diodes.

Because of that layout, two rear-light patterns covering the same 10% of a module can lead to different electrical losses.

If the weak area sits mostly on one branch, another branch may continue working with less impact. If the same area crosses several connected sections, the current loss can be larger.

For a real troubleshooting job, these are the useful questions:

• Which cells receive the weakest rear light?
• Are those cells in the same series path?
• Does the weak region affect one or both half-cell branches?
• Which cells share the same bypass diode?

NREL's bifacial modeling tools include half-cell layouts, cell-level irradiance and electrical mismatch because module-average irradiance cannot answer these questions.^[4]

Tongwei's analysis of uneven rear irradiance and bypass activation gives examples of how the same rear-side percentage difference can create very different total effective-irradiance differences.

Check Clearance and Tracker Position

Low clearance often makes the rear side harder to illuminate evenly. Rails, torque tubes and the module itself block part of the reflected light.

In one NREL high-albedo rooftop model:

Increasing clearance from 0.15 m to 0.50 m reduced mismatch in that specific model from 1.86% to 0.49%, about a 74% reduction.^[3]

Those values should not be copied directly into another project's loss model. Their value is in showing that mounting geometry can change rear-side mismatch by a large amount.

Tracker systems add another variable: the shadow pattern moves as the tracker moves. If the current plateau appears only at certain tracker angles, torque-tube or structural shading is worth checking. Tongwei's bifacial tracker backtracking analysis covers this angle-dependent rear-side behavior.

Do Not Blame the Bypass Diode Too Early

A large drop in rear irradiance does not always mean a large drop in total cell irradiance.

The rear irradiance falls from 300 to 50 W/m²:

83.3% lower rear irradiance

but total effective irradiance falls from 1,240 to 1,040 W/m²:

about 16.1%

That difference can reduce module power without necessarily turning on a bypass diode.

Bypass conduction becomes more likely when weak cells can no longer support the current flowing through their series path and are pushed into reverse bias. IEA PVPS describes the relationship between shading, reverse bias, hot spots and bypass protection.^[5]

In practice, an extra knee or voltage step in the I-V curve is much stronger evidence of bypass activity than a rear irradiance difference on its own.

Compare Current With the MPPT Limit

When current stops at the same clean number again and again, compare that number with the inverter datasheet.

Example:

• two parallel strings
• available current per string under strong rear irradiance = 18 A
• total available current = 36 A
• MPPT operating-current limit = 32 A

The difference is:

36 − 32 = 4 A

or:

4 ÷ 36 × 100 ≈ 11.1%

of the available current at that operating point.

In this case, the modules may still be capable of more current while the inverter stays around 32 A.

That 11.1% is not an annual energy-loss number. The inverter can change voltage, and the condition may last for only part of the day.

Also keep these two inverter specifications separate:

• maximum operating current — current the MPPT can use during normal operation
• maximum short-circuit current — a separate allowed array Isc limit

An inverter with a 32 A operating limit and a 45 A maximum short-circuit-current rating does not continuously operate at 45 A.

IEA PVPS notes that the higher DC current possible from bifacial arrays has to be considered in system-component sizing.^[5]

Add Parallel-String Current

At the module level, a 2 A increase may not look dramatic. At the MPPT level, several parallel strings add those increases together.

With four parallel strings, +2 A per string becomes +8 A at the MPPT.

This is worth checking during inverter selection because current is already high in modern large-format modules. Tongwei's TNC-G12R 66 bifacial specifications, for example, list front-side STC Isc from 15.94 to 16.18 A and Imp from 15.07 to 15.33 A.

Rear contribution can push field current above these front-only STC values.

Separate Current Limiting From Power Clipping

A flat AC power curve does not automatically mean the DC current has hit a hard ceiling.

Example:

• available DC power = 115 kW
• inverter AC rating = 100 kW

Available DC power is 15 kW, or 15%, above the 100 kW AC rating.

The inverter can keep AC output at its limit by shifting the DC operating point. Voltage, current or both may move.

Check the I-V Tester Rating

A hard current ceiling can come from the test equipment rather than the module.

A 20 A plateau from this tester cannot tell you whether the module would have produced 21 or 22 A.

Before using the result, check:

• maximum tracer current
• maximum electronic-load current
• maximum load power
• current-probe range
• shunt range
• measurement accuracy near full scale

IEC TS 60904-1-2 includes additional requirements for bifacial I-V testing, including rear-side conditions and irradiance non-uniformity.^[1]

Tongwei's I-V test diagnostic article also covers measurement errors that can make a healthy module fail an I-V comparison.

Match the Timestamps

One of the easiest ways to create a false plateau is to compare data that were not recorded over the same time period.

Example rear irradiance:

The five-minute average is:

(200 + 220 + 240 + 300 + 420) ÷ 5 = 276 W/m²

If inverter current is reported as a five-minute average of 15.6 A, but someone pairs it with the last one-minute reading of 420 W/m², the graph creates this relationship:

420 W/m² → 15.6 A

That is misleading. The 15.6 A current value belongs to a period with about 276 W/m² average rear irradiance.

Before plotting current against rear irradiance, match:

• sampling period
• averaging period
• timestamps
• time zone
• logger clocks
• missing-data handling

Correct for Temperature

Temperature changes power much more strongly than it changes Isc.

Tongwei's TNC-G12R 66 bifacial module data list approximately:

• Isc temperature coefficient = +0.046%/°C
• Pmax temperature coefficient = -0.28%/°C

For a 25°C cell-temperature increase, using those coefficients alone:

Isc:

25 × 0.046% = +1.15%

Pmax:

25 × -0.28% = -7.0%

The numbers show why a hotter module can still gain Isc while the increase in usable power becomes much smaller.

Temperature therefore fits this pattern:

Isc still rises + Pmax gain becomes smaller

It is a poor fit for this one:

Isc suddenly stops at one exact current value.

Check High-Current Losses

Higher current makes every resistive part of the circuit work harder because resistive loss rises with the square of current:

P_loss ∝ I²

If current rises from 14 to 18 A:

Current increase:

(18 − 14) ÷ 14 × 100 ≈ 28.6%

But the resistive-loss ratio becomes:

18² ÷ 14² ≈ 1.65

So the same resistance produces about 65% more resistive loss.

That can reduce Vmp and Pmax gains even while Isc continues to rise. Sandia's single-diode PV model includes series resistance as a parameter controlling I-V shape.^[6]

External wiring follows the same rule. With an illustrative 0.05 Ω total DC loop resistance:

Normal cable resistance usually does not produce a clean Isc ceiling. Poor connectors and high-resistance joints, however, become more costly as bifacial current rises.

Use the Plateau Shape

The shape of the plateau often tells you where to look first.

A hard plateau such as 16.00 A repeated at every higher irradiance point is a strong reason to check the inverter, tester, sensor range or software cap.

A soft plateau is different. Current still rises, but each extra 50 W/m² adds less:

+0.30 A → +0.20 A → +0.15 A → +0.10 A

That pattern fits mismatch, operating-point changes, temperature, resistance or measurement conditions much better than a simple electronic current ceiling.

Use the I-V Curve to Confirm the Cause

The full I-V curve helps separate problems that look similar in a single current graph.

Check These in Order

1. Confirm what current is being measured: Isc, Imp, string current or MPPT current.
2. Compare the plateau value with equipment limits: inverter, tester, electronic load and sensors.
3. Check front irradiance: rear irradiance alone does not describe total effective irradiance.
4. Measure rear irradiance at several positions: especially near rails, torque tubes and module edges.
5. Match timestamps: current and irradiance must represent the same time period.
6. Compare Isc with Imp: rising Isc confirms that the cells still generate additional current.
7. Check Vmp and Pmax: these show whether temperature, resistance or mismatch is reducing usable power.
8. Inspect the full I-V curve: look for a rounded knee, extra step or several possible operating points.
9. Compare a standalone module test with inverter data: this separates module behavior from downstream limits.

Suspect the Module Last

A module fault becomes a stronger possibility only after the easier system and measurement explanations have been ruled out.

The case becomes more convincing when:

• one module behaves differently from comparable modules
• the behavior repeats under stable irradiance
• another verified I-V tester gives the same result
• rear irradiance is reasonably uniform
• the inverter is no longer part of the test
• the I-V curve remains abnormal

At that stage, check for:

• cracked cells
• damaged cell interconnections
• high-resistance joints
• junction-box faults
• connector faults
• bypass-diode faults

If the electrical problem is accompanied by local heating, Tongwei's rear-side hot-spot analysis provides examples of using rear irradiance, I-V behavior and thermal evidence together.

NREL reported roughly 1–2% agreement between several indoor and outdoor bifacial measurement approaches under controlled conditions.^[7] A repeatable 10–15% current shortfall therefore calls for a much deeper check than a small 1% difference.

Finally

With 80% Isc bifaciality, every 100 W/m² of uniform rear irradiance adds about 80 W/m² of equivalent irradiance. In the 14 A example, 300 W/m² rear irradiance raises estimated Isc from 14.0 to 17.4 A. If current stops at one exact value, compare that value with the MPPT or tester limit first. If Isc rises but Imp does not, the cells are still producing more current. If one rear sensor reads 380 W/m² while a three-point average is only 263 W/m², correct the irradiance measurement before judging the module. In practice, Isc, Imp, Vmp, Pmax and the full I-V curve together tell you far more than a single current reading.

Safety: illuminated PV modules and strings can remain at hazardous DC voltage. I-V testing, short-circuit-current measurements, connector inspection and inverter-side electrical work should use correctly rated equipment and be performed by qualified personnel.