Do the lowest-cost modules always deliver the highest project value?
Consider the two module options below and how you would go about making a selection for a 49 MW DC solar project.
- Option A — 540W module at 0.426 $/W
- Option B — 650W module at 0.469 $/W
While making a decision strictly based on the dollars-per-watt ($/W) of a PV module may be a general rule for many project developers and independent power producers (IPPs), this approach fails to consider both the effect that module selection can have on balance of system (BOS) costs, as well the long-term impacts of module selection on project revenue.
Maximizing the value of a PV project is a balancing act. Estimated financial returns can change as modules are analyzed based on their prices, energy yield and degradation rates along with BOS costs. Any of these can tip the scale from a low-cost, lower-wattage module to a more expensive, larger-format module — and vice versa.
The secret to understanding how module costs affect project value lies in the formulas used in your financial analysis and your specific installation details. While experienced module buyers are familiar with the steps and considerations to take, even the most seasoned players can benefit from some guiding strategies to navigate the ever-changing procurement process.
Understanding net present value
You can use product data to compare the value of different modules to optimize overall project value. Determining each module’s net present value (NPV) is best. NPV calculates an asset’s costs and projected income and assigns a present-day value, accounting for the fact that money has greater value now than it will in the future, a concept known as the time value of money. NPV is calculated by summing up annual cash flows and discounting them back to today’s dollars. Employing NPV is more valuable than selecting modules based on the lowest cost, or even lowest all-in $/W cost, because it accounts for the total costs and revenue over the asset’s life.
Consider these two optimization levers in your financial analysis:
- CapEx levers: Any deltas in module costs must be weighed against potential effects on balance of system (BOS) costs. For example, larger-format modules will act to reduce the total module count, but their impact on the number of racks and foundations will vary depending on the racking product and the site wind and snow loads. A lower open circuit voltage (Voc) increases the allowable string length, which decreases the total number of strings and the total length of DC source circuit wiring. Frame thickness may necessitate larger, more expensive clamps or more robust racking structures. It is common for the additional costs associated with these countermeasures to be 0.015 $/W or more.
- Revenue levers: The revenue produced by a project, either through the sale of energy or through tax incentives, is primarily based on the system’s production. Module selection can significantly affect energy production due to differences in specific yield and module degradation. Bifacial modules at a project site optimized for rear-side production can easily increase output by 5%. Loss factors related to low light, temperature, incident angle modifier (IAM) and light-induced degradation (LID) can result in performance deltas of 2% or more between modules. New cell technologies will further differentiate the highest and lowest-performing modules.
Real-world proof
While choosing the lowest-cost option or a large format module may seem more attractive, real-world projects prove this is only sometimes true. In reality, you could be leaving money on the table, as outlined in the examples below.
Case Study #1
System size: 49 MW DC
Racking type: Single-axis tracker
Location: Peoria, Ill.
Table 1 – module options summary
| Module | Watts | $/W Module Cost | NPVDelta ($) | CapEx Savings ($) | Revenue Delta ($) | ||
| Module A | 650 W | 0.469 | – | – | – | ||
| Module B | 540 W | 0.426 | -1,109,935 | 293,000 | -1,402,935 |
Even though there was a $2.1M (0.043 $/W) module cost advantage for the 540W module, there was only a $293,000 (0.006 $/W) difference between the two modules in the all-in CapEx costs to build the project. This was due to a $1.8M (0.037 $/W) pickup in BOS cost for the 650W module over the 540W module. The 650W module was also projected to produce revenue that was $1.4 million higher than the 540W module on an NPV basis. When module costs, BOS costs and revenue were combined, the analysis showed a project built with the 650W module would have a higher NPV. In this example, selecting the lower-cost 540W module would have reduced project NPV by $1.1 million.
Table 2 – module, row, foundation & string quantity comparison
| Module | Watts | Modules (Qty.) | Modules per String | Modules per Row | Rows (Qty.) | Foundations (Qty.) | Strings (Qty.) | DC Source Circuit (ft) |
| Module A | 650 | 75,240 | 30 | 90 | 836 | 15,884 | 2,508 | 752,400 |
| Module B | 540 | 90,500 | 25 | 100 | 905 | 17,195 | 3,602 | 1,086,000 |
A project built with the 650W module requires installing far fewer modules, tracker rows and foundations. This is largely due to the form factor of the 650W module, but the Voc of the module also plays a role. The 650W module has a Voc of 45.0 vs. a Voc of 49.8 for the 540W module. This allows a maximum string length of 30 for the 650 and a maximum string length of only 27 for the 540s. When considering the physical length limitations of the tracker and the best practice of installing a whole number of strings per tracker row, it’s possible to maximize the number of modules per row by using three strings of 30 for the 650W, but it’s necessary to drop to four strings of 25 for the 540W. As a result, using the 650W results in more watts per row (4.5kW) and requires almost 1,100 fewer strings and over 300,000 less feet of source circuit wiring.
Running performance models for these two modules shows that the 650W module has a yield advantage of 2% over the 540W module. This yield advantage is a result of lower IAM, low light and temperature losses. Based on the energy rate for this project, the project term and the assumed discount rate, the revenue produced by this 2 percent increase in production has a net present value of $1.2M.
The 650W module also has a warrantied annual degradation rate of 0.45 percent, compared to the higher 0.5 percent annual degradation rate of the 540W module. In this case, the additional lifetime energy production that results from the lower degradation rate increases the NPV of the revenue by an additional $200,000. With both factors combined, the 650W module is projected to increase revenue by $1.4M based on NPV when compared to the 540W module.
Case Study #2
System size: 5 MW DC
Racking type: Fixed tilt
Location: Central Massachusetts
Table 3 – module options summary
| Module | Watts | $/W Module Cost | NPVDelta ($) | CapEx Savings ($) | Revenue Delta ($) | ||
| Module C | 540 W | 0.443 | – | – | – | ||
| Module D | 650 W | 0.485 | -147,178 | -155,000 | 7,822 |
Based on the first case study, it’s easy to assume that the best option would be the large-format 650W module. However, this project uses different modules than the first example, so the NPV of the 540W module was higher by $147,000.
Table 4 – module, row, foundation & string quantity comparison
| Module | Watts | Modules (Qty.) | Modules per String | Modules per Row | Rows (Qty.) | Foundations (Qty.) | Strings (Qty.) | DC Source Circuit (ft) |
| Module C | 540 | 9,234 | 27 | 22 | 420 | 1,680 | 342 | 102,600 |
| Module D | 650 | 7,685 | 29 | 18 | 427 | 1,708 | 265 | 79,500 |
In the first case study, the 650W module had an advantage in effectively using the tracker rows and foundations. Looking at Table 4, the 650W module requires slightly more tables and foundations for this fixed tilt project. This is based on the physical limitations of the racking product and the number of modules that can be installed per table. There is also a difference between the two options in the “modules per string” metric, which is not as large. The result is that the BOS savings for the 650W module are only 0.011 $/W, compared to the 0.037 $/W savings in the first project. Performance models for this project show an energy delta of 0.12 percent, in favor of the 650W module. This, combined with the fact that both modules have warrantied annual degradation rates of 0.5%, results in only a $7,822 (0.002 $/W) revenue delta between the two modules on an NPV basis.
In this case, the BOS improvements and revenue advantage of the 650W module are not enough to offset the low cost of the 540W module.
Don’t leave money on the table
During module selection, it can seem easy to narrow your focus to a simple comparison of the $/W cost of the modules. However, if your view is limited to module costs, you will be unable to account for changes to total project cost and revenue that a more comprehensive analysis would be able to demonstrate.
Gathering data and performing financial analysis can be time-consuming and laborious. It may require hiring additional procurement or engineering staff, contending with limited access to supplier data that quickly grows out of date and vetting the results. The good news is that new technology solutions have come to market to compare available inventory from top-tier suppliers. Such technology that can help perform module technical analysis enables users to make informed decisions in a timely fashion and secure contracts quickly based on pre-negotiated supply agreements.
However you choose to proceed, a proper analysis that goes beyond module cost is critical to making informed decisions to optimize revenue and realize better project returns.
Authored by Tarn Yates, Director of Product Management at Anza