Youtube Version: https://youtu.be/aWWViOuSWeM
“Phantom Drain” is the term that the Tesla community has collectively decided on when referring to the mysterious loss of charge from their vehicles’ batteries when parked. It’s a catchy term, but when the cause of the drain can be determined and potentially avoided in the future, the drain is not so puzzling.
Running the Tesloop Shuttle fleet over 2.6 million miles (4.1 million kilometers) over the course of three years, we didn’t see much drain. That’s because we kept the cars moving or charging the vast majority of the time. Now, with the Carmiq community providing data (from many Tesla configurations with many use cases) we have been able to identify major causes of drain and how to prevent them.
We are going to make sure that Carmiq doesn’t impact your battery’s drain rate. As a first step, we automatically pause data collection when your car is parked and not charging. This is the best way to ensure that your car can go to sleep and conserve charge any time of day or night.
It has also been hypothesized by some that certain firmware versions are correlated with higher rates of Phantom Drain, but based on our data we have not yet been able to confirm this definitively.
Some cars will not sleep at all while data is being continuously collected. Carmiq data shows that this issue is dramatically more common in Model S and Model X vehicles than in Model 3 vehicles. This agrees with the rumors that Tesla pushed a software update that improved the Model 3’s sleep behavior as mentioned by our friends at Teslanomics. It’s also possible that this is due to the “Energy Saving” and “Always Connected” controls being available to owners in Model S and Model X vehicles, but not in Model 3.
To find the most effective strategy to minimize Carmiq’s impact on sleep behavior we needed to test. We began with a Model 3 that a friend of the company had given us to manage for him as a rental asset. We knew from analyzing Carmiq data that this Model 3 was not going to sleep at all, so we ran a series of 48 hour tests over the course of a few weeks.
First, we made sure no other apps would be collecting data so we could run a clean test. Next, we paused Carmiq data collection and recorded how long it took for the car to go to sleep. Finally, after the car went to sleep, we resumed data collection to see whether this would cause the Model 3 to wake up.
The Model 3 went to sleep about 30 minutes after we paused data collection, and it did not wake up when data collection was resumed. This result was in line with what we expected based on other Tesla owners’ reports.
We found that there is not a consistent time that it takes for every Tesla to go to sleep. Of the hundreds of vehicles connected to Carmiq, we found examples of vehicles taking anywhere between just 7 minutes all the way up to 6 hours to fall asleep after parking. While our test showed that automatically pausing data collection is effective at allowing Tesla vehicles to go to sleep, it’s not a guarantee. Though we can control data collection, we aren’t in control of other factors that prevent sleep such as interacting with the car when it is parked, Smart Preconditioning, or other third-party Tesla apps.
Looking at usage patterns, we found that the optimal time to wait before pausing data collection is about two hours after the car is parked and not charging. This avoids missing out on collecting data for daytime drives, while also allowing the car to go to sleep while sitting overnight. This only risks losing about two miles of range compared to no data collection at all, based on the average drain rate observed by us and anecdotally by the greater Tesla community.
We will continue to refine and improve how Carmiq collects data, allowing you to use your Tesla to its fullest potential.