Can Tesla Tame the Duck Curve?

Unless you’ve been in a cave the last 24 hours (or at least off the internet), you’ve no doubt heard about Tesla’s move into the battery storage field. I attended the event last night and reported on it for POWER in the wee hours afterward. (The announcement came at night so Tesla CEO Elon Musk could wow attendees and the hordes watching the live stream with the fact that the new batteries were powering the entire party.)

The most striking element of the announcement was not the technology—which is at best evolutionary, as other companies are already offering similar systems—but Musk’s ambition for, as he said, “the complete transformation of the entire energy structure of the world.” He estimated that 2 billion of the company’s new 100-kWh Powerpacks would be enough to power a 100% renewable globe.

But one has to start somewhere, and that somewhere is California, where Tesla has already deployed several demonstration systems such as at the famed Kendall-Jackson winery in Napa Valley. California is also ground zero for the challenges of renewable integration, at least in the continental U.S. Again, you’ve probably seen this graphic from the California Independent System Operator (CAISO), which has come to be known as the “duck curve.”

duck curve

1. If it quacks like a duck. This chart shows solar-induced ramping on a spring day in California between 2012 and 2020. Source: California Independent System Operator.


The curve represents the challenge of solar coming online in the middle of the day and then dropping off just as the evening peak demand starts ramping up. The shaded area is new solar capacity being added from 2013 to 2020.

The question that comes to mind after last night is, what if a big chunk of those solar panels were hooked up to a Tesla Powerwall? How many would it take to flatten the duck curve by shifting that mid-day trough a few hours later into the evening peak?

The Powerwall, by its limited nature (the company says the 7 kWh version is designed for daily cycling while the 10-kWh version is for backup power), is likely best for shifting power a few hours forward. That’s good, because the ramping period along the left lasts about three hours.

In 2020, the area under the evening peak curve above 20 GW (about where it starts the day on the left) is roughly 22.5 GWh, unless my rusty calculus skills have failed me.

California currently has about 2,400 MW of behind-the-meter solar photovoltaic capacity according to California Solar Statistics, out of a total of just under 10 GW; of that, about one-third is residential. If we assume that breakdown between residential and non-residential storage continues, about 12% of that 22.5 GWh would need to be met by home storage batteries.

That works out to 2,700 MWh of residential storage, which would require about 385,000 7-kWh Powerwalls working at full capacity. Commercial and industrial storage would account for 5,400 MWh, or 54,000 100-kWh Powerpacks. The remaining 20,925 MWh would be met by utility systems, and if we stick with the 100-kWh unit size, that’s another 209,250 Powerpacks.

It probably doesn’t need saying that there’s no way that many batteries (of any sort) are going to be deployed in five years, not when the state itself has a far modest goal of 1.3 GW of storage over the same period.

Bottom line, Tesla’s new batteries will help, but they won’t do the job on their own, and certainly not in the time period envisioned in CAISO’s duck curve.

—Thomas W. Overton, JD is a POWER associate editor (@thomas_overton, @POWERmagazine).