One of the concerns about battery electric vehicles is the amount of degradation the battery suffers during normal use. We drove a Chevy Bolt for three years, clocking nearly 30,000 miles on the odometer. We tracked the car’s performance during this time, including changes in the battery’s capacity. In contrast to our experience with the Nissan Leaf, we found only a modest amount of degradation in the Bolt’s traction battery.
We leased a 2017 Bolt for three years and drove it for 29,000 miles when we returned it to GM. I collected miles traveled, state-of-charge, range, and other parameters over the life of the lease. (Yes, I am a nerd. I’ve been doing this for all of the vehicles I’ve owned—gas or electric.)
I used three different methods to monitor traction battery performance or degradation: directly with signals from the battery pack, indirectly from kWh used and percent State-of-Charge (SOC) remaining during a charge cycle, and the car’s estimate of range after a complete charge.
Other Bolt drivers have noted that battery capacity "bounces around" a lot, reflecting seasonal and other effects. There's more battery capacity available during the warm summer months than during the winter. Thus, taking a snapshot of degradation over only a few months may give misleading results. It's best to examine capacity over a few years to account for seasonal swings in capacity and that’s what I’ve tried to do by chronicling battery capacity over the life of the lease.
Chevy advertized the Bolt with a nominal usable capacity of 60 kWh. Nominal, of course, means approximate. For liability reasons, Chevy was unlikely to ship any cars with less than 60 kWh of usable capacity under standard conditions. With typical variations in assembly, the cars were likely produced with a little more than the nominal capacity.
Much has been written about the actual usable capacity of the Chevy Bolt's traction battery. There's been, shall we say, “lively” debate on user forums about what was the Bolt's initial battery capacity. While there's no universal agreement, there's a consensus that the Bolt during the model years 2017-2019 was shipped with slightly more than its nominal capacity of 60 kWh.
The government of Korea reports that the 2017-19 model year Bolt was shipped with a capacity of 60.9 kWh.
OBD Scanner Method
Unfortunately, I didn't begin monitoring battery status directly with an electronic scanner until well into the lease. I began monitoring battery capacity with an OBD (On-Board Diagnostics) scanner and Torque Pro ten months after we began driving the car. (Though I eventually mastered it, using Torque Pro was far more challenging to use than Leaf Spy for the Nissan Leaf. See Peeking Inside the Bolt's Brain Reveals Valuable Secrets.)
Torque Pro linked to the OBD scanner with Bluetooth and codes for interpreting the stream of signals from the car’s computer provides a host of parameters on the car’s performance. This includes battery capacity, labeled “Bat Capacity” and given in kWh.
After driving the Bolt nearly 7,000 miles, the OBD scanner reported that the traction battery had 61.1 kWh total capacity. Within a few weeks, the scanner was logging a capacity of 60.7 to 60.9 kWh. When we returned the car after nearly three years, the scanner reported 58.7 kWh.
As seen in the chart below there is substantial scatter in the data. To make sense of this we need to apply some statistical analysis to the data. In the chart this is presented as a trend line. As expected, there is a gradual decline in battery capacity as the car is driven more miles.
I used two different spreadsheet programs and each provided a slightly different regression analysis. The charts were made in OpenOffice Calc, but the table summarizing the statistical analysis was made in Quattro Pro. In the chart the trend line intercepts the y-axis at 60.3 kWh, Quattro Pro calculates it at 61 kWh. In either case, this fits with our experience and that reported in Korea for new Bolts at the time. The trend line indicates that after we returned the car, there was slightly less than 58 kWh of capacity remaining. Thus, battery degradation was less than 4% from nominal capacity based on 125 observations over nearly 30,000 miles.
However, the variability in the data is quite large and the Coefficient of Determination or R² is less than 0.5.
Percent Used Method
We can infer battery capacity by knowing how much of the battery was used for so many kWh consumed between full charges. For example, on our last full charge, we consumed 31.2 kWh and arrived home with 45.9% State-of-Charge. Thus 31.2 kWh/(1-0.459) = 57.7 kWh.
Disclosure: I worked for GM's Delco-Remy Division 1968-1970 as a cooperative engineering student. I was a member of UAW Local 1981 until the National Writers Union left the UAW in May 2020. The Chevy Bolt is assembled by UAW Local 5960.
Battery capacity varies from charge to charge whether using the car's OBD signal or the inference method above. As seen in the chart below, the variability of battery capacity using this method is much greater than that using the signals directly from the car via the OBD scanner.
As before, it’s necessary to calculate a trend line to make sense of the data from 135 observations. Here, the line intercepts the y-axis at 61.3 kWh in both OpenOffice Calc and Quattro Pro. The trend line shows that when we returned the car the battery had 56.4 kWh remaining for a loss of 6%.
Unfortunately, the Coefficient of Determination is less than 0.4--even worse than using the signals directly from the car with the OBD scanner.
Estimated Range Method
Newbies to EVs are often overly concerned with the car’s range estimates as a measure of how well their battery is doing. EVs estimate range based not only on the amount of battery capacity but also on past driving style or efficiency.
As seen in the charts above, there is quite a bit of variability in battery capacity regardless of the method used to monitor it. Introducing another fluctuating factor into the equation—driving style—increases variability even further. This suggests that determining battery capacity by estimates of range over time will be unreliable and that is indeed the case.
Unlike most EVs, the Chevy Bolt provides three estimates of range: low, mid, high. This is a very useful feature of the Bolt. Experienced drivers, such as Eric Way, learn how to use each estimate to more carefully judge how much battery capacity they have left when they are on a road trip.
Consequently, the chart below reports three range estimates over time during our three-year lease. Again, to make sense of this it’s necessary to apply a statistical analysis, shown here as a trend line.
Batteries degrade over time. They don’t get better with age. The trend lines in the chart show an increase in range over time. This doesn’t make any sense and I have no explanation why this is the case.
Based on nearly three years of monitoring range estimates on the Bolt, it is obvious that this method cannot be used to calculate battery health. It also suggests why newcomers to EVs should take range estimates with a grain of salt.
Extrapolating Linear Degradation
Experience with Teslas’ Model S suggests that battery degradation isn’t linear. Maximum degradation occurs in the early years and tapers off with time. There’s no evidence yet of this with the Bolt.
If we assume linear degradation (the worst case scenario), we can estimate our Bolt’s remaining capacity after 100,000 miles, the limit of the car’s drive-train warranty in California.
Our Bolt would lose 10% of its capacity—or 6 kWh—after it had been driven between 45,000 and 66,000 miles. By 100,000 miles the car will have lost between 16% and 25% of its capacity. Even so, the car would still have a range on a full charge of from 180 to 200 miles—more than adequate for most uses.
My analysis is based on our experience, that is, one car out of some 70,000 Bolts manufactured to date. This anecdotal data, however, does compare well with at least one other source.
Geotab is a firm that specializes in telematics for fleet operators. They claim to monitor two million vehicles on their platform, including 6,300 EVs in fleet and consumer use. As a promotion of the company’s services, Geotab provides an interactive graphic on its web site they dub an EV battery degradation tool.
Using Geotab’s tool, I found that after two years and eight months Chevy Bolts in fleet use lost about 4% (3.9%) of their battery capacity. (There’s no indication from Geotab how many vehicles this represents.)
According to Geotab, this is comparable to the same model year for a Tesla Model S. (Teslas' Model 3 was introduced in 2018.) Teslas are often held up as an example of how little battery degradation can be expected from a well-designed battery pack. (For more details, see Geotab's report on EV battery degradation.)
The traction battery on our 2017 Chevy Bolt degraded slightly (less than 4%) over the nearly 30,000 miles we drove it. Using direct signals from the battery monitoring system via an OBD scanner is the best way to monitor the battery’s capacity. Inferring battery capacity by tracking kWh consumed and state-of-charge can also be used, but is less accurate than using direct signals from the battery because of greater variability between charge cycles. Estimating capacity by tracking the car’s range estimates cannot be used. Based on data translated by the OBD scanner, our Chevy Bolt should have at least 200 miles of range—and possibly more--after 100,000 miles on the odometer.
Battery Degradation Comparison Chevy Bolt EV and Nissan Leaf
Tracking Chevy Bolt EV Battery Capacity
Peeking Inside the Bolt's Brain Reveals Valuable Secrets
Battery Degradation Two-Year Status Report: 2015 Nissan Leaf
2020 Chevy Bolt EV Battery Capacity Anecdotal Observation
Slight Chevy Bolt Traction Battery Degradation after Nearly 30,000 Miles