Underwater, sunlight is rapidly attenuated by absorption and scattering in water, with infrared light in particular being lost even at shallow depths, causing the solar spectrum to skew toward the blue-green range.
In response to this constraint, a new theoretical and experimental approach has emerged that seeks a solution through the geometric strategy of curving solar cells. According to a theoretical model developed by a team at the U.S. Naval Research Laboratory's Electronics Science and Technology Division, an optimally curved amorphous silicon (a-Si) solar cell—evaluated based on device footprint—reached a maximum conversion efficiency of 59.7% at a depth of 2 meters. Furthermore, a calculation model spanning depths of 2 to 50 meters, based on the optical properties of South Pacific water, showed that the optimized curved solar cell's daily cumulative power output was 9.4% to 15.9% higher than that of a conventional flat-panel design. The core of this research lies in abandoning the assumption of flat light collection and instead quantifying the optical gains achieved through curvature.
A Geometric Structure That Captures Blue-Green-Skewed Underwater Light and Scattered Light
Let's first examine the unique light attenuation characteristics of water and the definition of the efficiency metric used.
Water absorbs and scatters light, and infrared light in particular attenuates rapidly even at shallow depths. As a result, the solar spectrum reaching underwater locations skews toward the blue-green range. It's important to note how the conversion efficiency of underwater solar cells is calculated. This conversion efficiency is not measured against total surface irradiance as its denominator; rather, it is defined as the ratio of output power to the light power incident on the cell at that specific depth.
A curved structure increases the amount of light captured while maintaining the same device projection area. According to the research team, cells with greater curvature capture not only transmitted light but also scattered light from the surrounding environment, allowing them to receive more light than flat-panel designs.
The validity of this geometric approach was also confirmed at the laboratory level. In indoor tests using 50 mm square a-Si cells at a depth of 0.1 meters, the cell with maximum curvature—conforming closely to the outer surface of the underwater device—recorded the highest output. This behavior aligns with the predicted trends from the optical model developed by the research team.
Testing Four Commercial Cell Types: Diverging Characteristics at 20 cm and 1 m Depths
Following the theoretical validation, the research team conducted experimental tests to highlight the performance differences among various materials. The tests took place in a 10 m³ pool filled with artificial seawater at 3.5% salinity. Four commercially available cell types were selected for comparison: crystalline silicon, CIGS, CdTe, and a-Si.
For the crystalline silicon and a-Si cells, which require encapsulation, polymethyl methacrylate (PMMA) enclosures with over 91% transmittance in the 400-1100 nm wavelength range were used. Since the enclosure caused less than 10% attenuation of solar irradiance, the research team corrected for this transmission loss in their spectrum and performance calculations.
Real-world measurement data collected outdoors illustrates the severity of light loss underwater. During measurements taken between 11:30 AM and 12:30 PM, when sunlight peaked, irradiance intensity dropped by 44.9%—from 717.81 W/m² at the water's surface to just 395.20 W/m² at a depth of only 10 cm. This substantial loss resulted not only from absorption by water but also from reflection at the water's surface, causing greater loss in visible light than calculations alone would suggest.
Amid this rapid narrowing of available light, outdoor tests revealed clear divergence in performance characteristics corresponding to each cell's light-receiving properties. At the 20 cm depth measurement point, crystalline silicon, CIGS, and CdTe reached their maximum efficiencies of 20.6%, 24.3%, and 35.9%, respectively. The a-Si cell stood out in contrast: its efficiency continued to rise beyond the 20 cm depth, all the way to 1 meter, never reaching its optimal depth within the tested measurement range.
In outdoor tests conducted in artificial seawater at a depth of 10 cm from noon to 5 PM, the open-circuit voltage of curved cells consistently exceeded that of flat cells.
Navigation Trials for Small Underwater Vehicles and the Depth Limits of Direct Motor Drive
What kind of practical difference does the power gain from curvature make in actual underwater equipment operations? The research team conducted charging tests using a small UUV's lithium battery, followed by continuous underwater navigation tests after full charging.
In a charging test where a curved CIGS cell was connected to a 7,100 mWh lithium battery for a UUV, full charging at the water's surface using pure water took 3 hours—35.7% faster than with a conventional flat cell. In tests using the same UUV, the fully charged 7,100 mWh lithium battery enabled over 2 hours of continuous underwater navigation.
On the other hand, calculations examining direct-drive scenarios—where a 6V motor is powered solely by solar cell output without a battery intermediary—revealed severe constraints imposed by water quality and depth. The maximum depth at which direct-drive operation remains viable varies dramatically depending on water clarity.
| Water Environment | Flat CIGS Cell Depth Limit | Curved CIGS Cell Depth Limit |
|---|---|---|
| Pure Water | 1.0 m | 4.6 m |
| Clean Seawater | 0.9 m | 4.1 m |
| Lake Water | 0.5 m | 1.11 m |
As the table shows, the depth limit increases from 1.0 m (flat) to 4.6 m (curved) in pure water, and from 0.9 m (flat) to 4.1 m (curved) in clean seawater. However, in lake water conditions, the depth limit remains at just 0.5 m for flat cells and only reaches 1.11 m even with curved cells. This calculation demonstrated that the maximum depth at which a 6V motor could be directly driven varied significantly depending on water quality.
In outdoor tests conducted in pure water, both flat and curved a-Si cells were able to directly power an LED beacon down to a depth of 1 meter.
Prior GaInP Research's High Efficiency and Aerospace Manufacturing Cost Challenges
Prior to this study, research on underwater solar utilization using GaInP cells had already been reported.
In preliminary results published by the U.S. Naval Research Laboratory in 2012, GaInP cells generated 7 watts per square meter of area at a maximum depth of 9.1 meters. This earlier research further reported that, under light conditions simulating relatively clear seawater off Key West at depths of 2 to 9 meters, RHJ-GaInP cell efficiency exceeded 51% across the entire range, reaching approximately 54% at a depth of 8 meters.
However, superior material physical properties don't necessarily translate into a technology ready for widespread adoption. The GaInP cells used in the prior research relied on specialized, low-volume aerospace manufacturing processes, making them expensive. Achieving cost parity with CdTe would require mass production and reduced substrate costs.
The current study evaluated commercially available materials such as a-Si and CIGS, examining improvements in light capture achieved through curved structures. Meanwhile, the GaInP cells used in the prior research have been reported to face manufacturing cost challenges.
Rapid Light Attenuation by Water Quality, Simulated Light Source Evaluation, and Real-World Marine Environment Constraints
Translating the figures and test results presented in this model into practical field operations requires careful consideration of several underlying assumptions and limitations.
First, there is significant disparity in light attenuation depending on the environment. The amount of visible light absorbed by seawater or lake water varies according to salinity and organic matter concentration. According to the research team's environment-specific model, theoretical output at a depth of 2 meters decreases by approximately 40% in pure water and clean seawater, but by as much as 70% in turbid seawater or lake water.
This disparity becomes a decisive divide as diving depth increases. In this model, for the selected examples of turbid water and lake water, no usable power can be obtained beyond a depth of 5 meters. In clean seawater, however, usable power remains available even at depths of 40 to 50 meters.
Second, there is uncertainty inherent in the evaluation methodology itself. The prior research's comparison at depths of 2 to 9 meters, which reported the high efficiency of approximately 54%, was not based on actual cell submersion tests but rather on measurements conducted at room temperature in air, using LEDs to simulate the underwater light spectrum. The prior research also acknowledged that the reference cell correction under the 9-meter depth condition could be off by up to a factor of approximately 25—too large a margin for certified efficiency measurements—and thus characterized its results as reasonable efficiency estimates rather than definitive measurements. There remains room for careful verification between estimates derived from simulated light sources and actual behavior in real water environments.
Third, there are physical barriers associated with long-term operation in actual marine environments. Long-term deployment in real ocean settings presents challenges including corrosion, high water pressure, interaction with ocean currents, and reduced light transmission due to biofouling.
The maximum efficiency of 59.7% at a depth of 2 meters achieved through curvature, along with the 9.4% to 15.9% increase in daily cumulative power output compared to flat-panel designs, are results derived from a theoretical model. Practical implementation will require verifying long-term operational challenges, including corrosion, high water pressure, interaction with ocean currents, and reduced light transmission caused by biofouling.