Solar design tips, sales advice, and industry insights from the premier solar design software platform

Andrew Gong
Author

Andrew Gong

Andrew Gong is a Research Engineer at Aurora. He previously worked on two Solar Decathlon projects, and he received his M.S. from Stanford and B.S. from Caltech.

Our Newsletter

Should I Point My Solar Panels West to Optimize for Afternoon Peak TOU Rates?

Andrew GongAndrew Gong

Here at Aurora, we like to tackle hard questions. One question we see often is: “Given that electricity rates are higher from 4 - 9 pm in California, should I place my solar panels facing west instead of south to maximize savings?”

In this post we will dive into why west-facing panels might be more advantageous than south-facing ones, and provide recommendations based on your location and how your peak time-of-use (TOU) rate works.

Context: Moving West in California

Aggregate Demand and the Duck Curve

The three main investor-owned utilities in California have been utilizing peak TOU rates 4 - 9 pm since late 2017; customers are charged more for the electricity they use during the late afternoon and evening hours. Prior to this, peak hours in the middle of the day—between 11 am and 6 pm—were most common.

The peak hours shift was driven by aggregate demand on California’s grid. Prior to the early 2010s, homeowners and businesses used the most energy in the middle of the day and early evening. Midday peak-pricing provided the financial incentive to reduce energy usage or shift loads to non-peak hours. Since then, the falling cost of solar has drastically shifted the total demand on the grid as seen in the infamous duck curve, pushing the peak demand, and moving peak-pricing to early evenings.

Source: CAISO. California’s “Duck Curve” shows the low mid-day net demand on the California electricity grid after solar and wind generation is accounted for. The low mid-day trough and steep afternoon ramp are part of why time-of-use rates have peak pricing between 4pm and 9pm. 

Directional Variation

Taking a look at the energy production of a south-facing system in the Los Angeles area, about 87% of its summer energy is produced during off-peak hours and 13% is during peak hours.

 

Chart showing a PV production curve for a south-facing system; only 12.6% of energy is produced during peak TOU hours.

If we take the same system and rotate it to face a different direction, we estimate that you can get up to 20% of that system’s energy produced during peak hours, but the total production (yield) from the system is lower. The tradeoff is having more peak-hour energy at the expense of having less total energy; the peak yield of a system in this location at an azimuth of 190, just slightly west of south.

Chart showing energy yield (kWh produced per W per year) and the percent of production that’s during peak hours. Maximum energy yield is around an azimuth of 190 degrees. The most on-peak kWh production occurs facing west, but the overall yield is substantially lower.

Net Metering Valuation

In standard net metering policies, energy production from the PV system can be assigned a dollar value based on the utility rate at that time*. For example, producing 2 kWh in an hour with a rate of $0.15/kWh provides a credit of $0.30, and producing 2 kWh in an hour with a rate of $0.35/kWh is valued at $0.70. It’s important to understand the time-dependent production profile of a PV system, not just daily total yields.

Taking our south-facing production curve from above, about 12.6% of energy is produced during peak hours, but that energy is valued at nearly double that of off-peak energy. As a result, 21.2% of the value of the energy from the PV system comes from on-peak production.

Chart showing how the on-peak energy produces a greater value to the homeowner than off-peak energy, thanks to the peak TOU rate.

Back when the evening-peak rates were newly implemented, we reviewed several hundred San Diego projects that had been designed in Aurora and found that the rate change would substantially increase bills for customers who were still large net-consumers after going solar, but that solar was still a solid investment even with the less favorable TOU periods.

Solar software from Aurora makes it easy to model your customer's energy usage and create an optimal solar design and proposal. See how in a live demo! 

Is It Worth Pointing Panels to the West?

After simulating production profiles for systems with various orientations and combining it with residential TOU rates, we created an “energy value” of the PV systems. We calculated energy value by taking the kWh created by a system during each hour of the year and multiplying it by the current retail rate, and then tabulating the sum of all hours. The data is presented in gauge charts below. The wheel color indicates the system yield similar to the irradiance maps in Aurora. The sunshine icon indicates the optimal azimuth with the maximum energy production, and the needle showing which orientation achieves the most “energy value” throughout the course of the year.

The example below is for a customer in Southern California Edison territory near Los Angeles, using the TOU-D-4-9-PM rate. The ideal orientation for energy production is around 190 degrees azimuth, and because of the late peak hours, the ideal orientation for solar value is slightly further west at 200 degrees azimuth.Screen Shot 2020-03-12 at 5.46.13 PM

In other locations, this trend still holds true. The optimal orientation for system production value is slightly west of the optimal orientation for overall energy production, but not substantially so. Nearly all sites studied (see below) have an optimal orientation of 190 or 200 degrees azimuth, just slightly west of south. The rate that favors the furthest west-facing panels is San Diego Gas & Electric’s TOU-DR-SES for solar energy systems, which features an extraordinarily high peak price and lower off-peak rates.

What Happens If…

In the previous scenarios, there aren’t strong arguments to face your panels west when you have a choice to point them south. However, west-facing panels might be the way to go in some situations, e.g., a return to 3 - 8 pm peak pricing or a higher on-peak price. We will look at a couple of these what-if cases next.

Higher Peak Pricing

In some of the current TOU rates, the peak rate is roughly double that of the off-peak rate ($0.16 vs $0.31 per kWh in the SCE example). Adjusting this ratio from 2:1 to 3:1 or 4:1 by either increasing the peak rate or reducing the off-peak rate will provide more relative values to the on-peak energy production. We tested this out with the SCE rate, and got the following results:Increasing the difference between on-peak and off-peak prices to such an extreme level is not expected, but doing so would favor southwest-facing systems.

The energy production (color scale) remains the same, but the needle shifts west indicating that there’s an advantage for southwest-facing systems when there is an extreme difference between on-peak and off-peak pricing. We don’t expect rates like this to show up frequently, but an off-peak to on-peak price difference of $0.15 vs $0.45 would be enough to favor systems that face more west.

Earlier Peak Pricing Hours

For most PV systems, whether they face south or west, there is more energy production between 3 - 4 pm than 4 - 5 pm. What if peak TOU hours began at 3 pm, would there be an advantage to face further west?

Yes, if the rate is E-TOU-A in PG&E territory, which has a 3 - 8 pm peak pricing period. E-TOU-B is similar to E-TOU-A, but features a 4 - 9 pm peak rate. Shown side-by-side here, there’s no change in the optimal orientation.

Non-Standard NEM Scenarios

The above analysis applies to utilities that have standard NEM policies, or have export rules that are still very close to retail rate. In markets such as Nevada or Hawaii, in which the gap between the purchase rate and the credit for excess energy is substantial,, there is additional value in designing PV production to coincide with home loads (i.e., self-consumed energy offsets the bill at the retail rate while exported energy is credited at a lower amount, or not at all). To correctly model these scenarios, it’s important to have a good measure of the customer’s home energy usage (such as Green Button Data) and to make sure your modeling tools support advanced NEM rules.

Conclusions

If you are planning a ground-mount system or have a flat roof surface, you might be able to boost the value of your solar PV system by pointing it slightly to the west. If your roof is south-facing, the actual difference in produced energy value between due south and the optimal azimuth was between 0.3% and 0.7% in all the cases we looked at. Other factors, such as shade on the site, which roof surfaces are available for solar, and even the efficiency of the inverter, can have a much larger impact on the value of the system than picking between south and slightly southwest.

A choice with greater consequence is picking between a southeast-facing (135 degrees azimuth) and a southwest-facing (215 degrees azimuth) roof face. The southwest-facing surfaces typically have about 99% of the maximum energy value while southeast surfaces were around 95% across the board. Both are great, but if all other aspects were equal, the southwest surface would be the better choice.

Note

Our data models here ignore non-bypassable charges, which are 1.5-3 cents per kWh depending on the utility company and rate. Models also simplify non-peak costs to be an average of off-peak and shoulder-peak prices, since these typically only differ by a few cents. We used 4pm-9pm hours for TOU, and used May, June, July, and August as the summer-pricing months. Production profiles were simulated using unshaded panels sloped at 20 degrees. Actual energy yield and optimal orientation will depend on site conditions including shading, and the actual optimal-value orientation may vary by tilt, shade conditions, and the customer’s load profile. You can complete a more in-depth analysis of system performance and utility bill savings using Aurora’s tools.

Andrew Gong
Author

Andrew Gong

Andrew Gong is a Research Engineer at Aurora. He previously worked on two Solar Decathlon projects, and he received his M.S. from Stanford and B.S. from Caltech.

View Comments