Cheaper modules are (probably) here to stay, and so is the increased use of east-west racking—also known as dual-tilt. At the same time, system engineers are also pushing the limits of their inverters, making sure that they squeeze every last dollar of value from the system. Combined, this raises a question: do these two trends help or hurt each other?
The east-west value proposition
At a high level, east-west racking enables more modules to fit into a roof area, typically an increase in capacity of 20-30%. At the same time, because the modules are not oriented toward the equator, the energy yield per installed watt of capacity drops by 5-10%.
Yet it isn’t just a question of how much less energy these new design approaches yield, but when the differences occur. Specifically, east-west arrays produce less energy in the middle of the day, when the sunlight is greatest.
Figure 1: Energy generated by hour of day, south-facing vs. east-west array
Because the east-west array is not pointed toward the equator, it does not produce as much at the peak production times. As a result, the energy production is more evenly weighted across hours of the day. This can also be seen if we chart how often the array is producing at various power levels (as a percentage of maximum STC rated power)—the east-west array spends significantly less time at significant power levels.
Figure 2: Frequency histogram: number of hours spent at power level
While the east-west array produces 8% less energy overall, it actually produces 75% less energy at the highest energy levels.
Because these high-power times are also when inverter clipping occurs, this means that the clipping losses from east-west arrays will be lower than south-facing arrays. As can be seen in the chart below, a DC/AC value of 1.35 results in just 0.5% over-power clipping losses with east-west, versus 1.5% for a south-facing array.
Figure 3: Clipping losses for east-west vs. south-facing array
Economic value
So, east-west racking can reduce clipping losses (and therefore improve production) by over 1%. There are two ways to think about this benefit.
First, with the same inverter budget, a system designer could simply enjoy reduced clipping losses, on the order of a 1% improvement as shown above. This helps to offset the 5-10% lower irradiance that east-west arrays get.
Figure 4: Clipping loss reduction for east-west array
Alternatively, a designer can target similar clipping losses, and use the east-west production profile to design for a smaller inverter. In the example used in this article, the inverter size could be reduced by 10%, saving $0.01-$0.02/watt in CAPEX cost.
Figure 5: Inverter size reduction for east-west array
In either case, system designers can leverage this knowledge to help reduce the LCOE of east-west racking arrays.
This piece of research is still valuable years later – so thank you.
I was drawn to the EW concept not for any of the reasons above but as a way to reduce soiling losses: a 10 degree slope and rain will help reduce dust. I have no data to back up my hypothesis but Its worthy of consideration should someone wish to take up the research.
Great explanation. We are going to install a SE and NW facing system (two different mppt trackers of course). At least in our neck of the woods the difference between SW and NW wasn’t as big as I thought it would be.
What was the actual kva output difference between the two systems
Ie, 10 kw inverter with 13Kw array vs E6.5Kw & W6.5Kw arrays over a day and over a year
Look forward to your thoughts
By the way good read