Solar energy has gotten really, really cheap in the last decade and it's causing some unusual changes in the industry. A commodity solar module costs about $0.35 per watt today, or around $100 for a module with dimensions of 1 meter by 2 meters. Given the cost of the aluminum frame, glass covering solar cells, and electronics in the solar module, there aren't a lot of places to cut costs left, which ironically makes costs less important for solar manufacturers.
What's becoming increasingly important is the efficiency of the cells that are in a solar module. Commodity cells made of multi-crystalline silicon can make modules that are about 15% or 16% efficient, but cap out around there. Mono-crystalline silicon modules can bump up efficiencies to around 18% for a commodity product or higher for mono-PERC (passivated emitter rear contact) construction or over 23% for modules made with SunPower's (SPWR -2.17%) back contact construction.
After years of expanding multi-crystalline capacity, JinkoSolar (JKS -0.14%), Canadian Solar (CSIQ -3.90%), Hanwha Q Cells (HQCL), and JA Solar (NASDAQ: JASO) are rushing to upgrade to PERC production because it doesn't add a lot of costs and increases efficiency a couple of percentage points. To keep up, even First Solar (FSLR -2.09%) is upgrading from a solar module that is 16.9% efficient today to one that is over 18% efficient.
Here's why cell efficiency is so important.
Image source: SunPower.
Costs and efficiency collide in solar
It's easiest to explain the importance of efficiency by using some examples. According to the Q2 2017 U.S. Solar Market Insight report from GTM Research and the Solar Energy Industries Association, on average a utility-scale solar system in the U.S. costs $1.08 per watt to build, a commercial system costs $1.56 per watt, and a residential solar system $2.84 per watt. From there, we can back out what the impact of a higher efficiency module will be.
In the example below, I'm going to use module costs of $0.40 per watt because that's what the U.S. SMI report used, but you can adjust the numbers to higher or lower costs if you like. This first table simply calculates the non-module costs of each size solar system.
| Cost Components | Utility | Commercial | Residential |
|---|---|---|---|
| Total cost | $1.08 | $1.56 | $2.84 |
| Module cost | $0.40 | $0.40 | $0.40 |
| Non-module cost | $0.68 | $1.16 | $2.44 |
Data source: GTM Research Q2 2017 U.S. Solar Market Insight report. Calculations by author.
We can then use these numbers to build a hypothetical solar system of each size. In this example, assume that we are constrained to a certain plot of land or rooftop for each system.
| Component | Utility | Commercial | Residential |
|---|---|---|---|
| System size | 10 MW | 1 MW | 6 kW |
| Module cost | $4 million | $400,000 | $2,400 |
| Non-module cost | $6.8 million | $1.16 million | $14,640 |
| Total cost | $10.8 million | $1.56 million | $17,040 |
Data source: GTM Research Q2 2017 U.S. Solar Market Insight report. Calculations by the author.
If we assume that these systems were built using 15% efficient multi-crystalline silicon panels, we can then estimate what the impact of higher efficiency will be.
In the next example, I've assumed that the physical size of the solar system is unchanged, but the solar modules used are 20% efficient rather than 15%; also, the cost of these modules increased to $0.45 per watt to account for some higher costs from higher efficiency. I've left non-module costs unchanged, which slightly underestimates added costs for bigger inverters or other components, but these costs will be leveraged if the number of panels and physical space remains unchanged. Here's what the overall costs and cost per watt look like:
| Component | Utility | Commercial | Residential |
|---|---|---|---|
| System size | 13.33 MW | 1.33 MW | 8 kW |
| Module cost | $6.0 million | $600,000 | $3,600 |
| Non-module cost | $6.8 million | $1.16 million | $14,640 |
| Total cost | $12.8 million | $1.76 million | $18,240 |
| Cost per watt | $0.96 | $1.32 | $2.28 |
Data source: GTM Research Q2 2017 U.S. Solar Market Insight report. Calculations by the author.
You can see that higher efficiency solar modules can lead to lower costs per watt, particularly on the residential and commercial side. Thus, with the cost premium of higher efficiency coming down, we'll probably see more focus on efficiency in the future.
The race to higher efficiency is on
Over the next couple of years, most manufacturers will upgrade equipment to make mono-PERC solar cells or another higher-efficiency construction. They're doing this because solar costs have come down so far that the industry is starting to focus more on squeezing more energy from a limited amount of space rather than just trying to lower costs. Companies that can't keep up with these upgrades risk being left in the dust.
