Latest Comic : Tuesday, June 25. 2019

<< First  < Previous

Assessing one full year of Bolt EV driving

My Bolt EV has travelled 24,099.4km in its first year

On June 1, 2019, my Bolt EV had its first birthday. As it moved around more than 24,000 kilometers, I kept a detailed log to make continuous analysis of the car's conditions and characteristics. Many people including myself are interested in how an electric vehicle fares over the years, so this should provide some good insights.

Bolt EV's monthly statistics from June 2018 to May 2019 - distance, fuel economy, and battery capacity

I used to drive around 1,000km per month on average before getting a Bolt EV. But you can see that it has doubled since. Cheaper fuel costs was a major factor (less than 1/10 that of gasoline), with some "new car curiosity" thrown in. Efficiency suffered in summer and winter due to the extreme temperatures, which affects battery performance and climate control use. Largely speaking though, fuel economy had been improving because I've been adjusting my driving style to be smoother in order to go further before recharging. This proved to be helpful in long-distance trips.

The three lines at the bottom of the graph depict the battery capacity as calculated by various means. The battery degradation is a major concern for many, so I kept track of this closely as well. Going by the reported values, my Bolt EV originally had 58.63kWh of usable capacity (65.14kWh raw) and had 55.98kWh usable (62.20kWh raw) by 24,099.4km. This is a degradation of 4.52%. Assuming linear progression, the battery would have exactly 70.0% of capacity left after 160,000km. This is in line with the industrial average warranty and shows that my Bolt EV's battery is in a reasonably good condition so far.

So why did I have three lines here? It stems from the fact that the Bolt EV doesn't tell you its battery health outright. One of the Parameter ID (PID) readings from the OBD-II port (#2241A3) correlates directly with battery capacity, but interpreting the number has been up for debate. So I decided to find an interpretation that I was comfortable with.
Continue reading "Assessing one full year of Bolt EV driving"

Bolt EV under ideal charging conditions

Korea's Ministry of Environment (ME) has been aggressively expanding its network of DC Fast Charging (DCFC) stations throughout the country, with more than 1,100 new chargers being installed and operating in pilot mode since early this year, accounting for nearly 40% of total. I was fortunate enough to live close to one of such stations (and a 100kW version at that), which let me test out the charging characteristics of my Bolt EV without costing me a dime.

Although the chargers are supposed to switch to normal paid operation starting mid-May as the firmware updates are deployed in a staggered manner, I was able to observe what the close-to-ideal charging situation would be before this happened to the nearby charger. The following graphs plot the data I recorded.

It should be noted that the ME chargers have either a 40-minute or a 41-minute time-out. This was done to prevent a single person from hogging the charger for too long. Therefore, I did these charging sessions during the early hours in the morning when no one else was around in order to have as close to continuous charging as possible. This led to a bit of "blips" in the graph (64% - 41st minute / 88% - 82nd minute), but it did not affect the overall picture that much.

Bolt EV's charging curve on a 100kW station as a function of Displayed SoC

Charging speed is largely dependent on the battery's State of Charge (SoC), so it helps to see the data as its function. As you can see, the charging current remains more or less constant at a given "zone", then drops down a step after a certain level of SoC is reached.

The actual charging power will slowly increase in a zone because the charging voltage rises. This is a direct reflection of the the voltage of the battery cells themselves, which rise as the energy is filled up. The 288 cells are arranged as 96 groups in series of 3 cells in parallel, so there would be nearly a 100-fold difference between the cell voltage and the charging voltage.

Another major factor in the speed is the battery's temperature. Assessing multiple charging sessions, it became apparent that it should be around 24 to 27°C at the beginning in order for the Bolt EV to enter maximum current (roughly 150A before 50%). If it's colder, it will start out a bit slower, then ramp up to 150A as the battery heats up to about 24°C. If it's too hot (more than 30°C), the charging current caps to 95A to prevent overheating.

Bolt EV's charging curve on a 100kW station as a function of time

For someone who's waiting for the car to fill up, the time it takes for each of those charging zones is also quite important. So this is a graph showing the same data, but plotted as a function of time. Key numbers are distilled into the following table.

To make a quick comment about the displayed and actual SoC, the two meet at around 75% mark, with the displayed getting larger above and the actual getting larger below. At the extremes the two differ by about 4%, showing the buffer for preventing over-charging or over-discharging.

SoC Range (%) Duration (mm:ss) Speed (%/min) Power (kW) Current (A)
10 ~ 50 28:30 1.40 53 149
50 ~ 67 16:00 1.06 39 107
67 ~ 83 24:00 0.67 25 66
83 ~ 92 20:00 0.45 18 46
92 ~ 98 17:00 0.35 11 29
98 ~ 100 08:30 0.24 6 ~ 10 17 ~ 28
The power and current values are as seen from the charger. The values from the vehicle's subsystem were about 96% of these, showing the losses inherent in the charging process. Further losses occur as the energy ends up inside the battery, so we end up with a bit more than 10% loss in total.

The advantage of using a 100kW (500V x 200A) charger is apparent only for the first 50% of charge, and is not a huge one at that. 50kW chargers in Korea supply either 110A or 120A maximum current, so the charging speed of 1.06%/minute should extend to below 50% SoC when you use them. Hence, you'll shave about 10 minutes off the session with a 100kW charger instead of a 50kW one if you're starting from 10% charge left. You can thank Bolt EV's highly conservative charging regime for this.

So what's the takeaway from all these information? Probably a good basis for forming a charging strategy during a long-distance trip. To minimize charging times, you should keep the car's SoC between 10 to 20% minimum and 70% to 80% maximum, with each charging session lasting about an hour at most. The last 25% alone takes an hour to charge, so a full charge is not a good strategy unless you're going to a place where the chargers are sparse. Meanwhile, the ambient temperature during charging should be as close to 20°C as possible. Hopefully, you can find a charger within a building or under a shade.
Defined tags for this entry: , , ,

The two satellites

Night sky is home to various satellites, both natural and artificial. I've taken photos of these objects in the past few days and here are a couple them for you to enjoy.
Iridium 97 streaks downward in the northern sky on February 17, 2019

The Iridium satellites are known for their flares caused by the interesting shape of their antennae. I had a chance to observe Iridium 97 moving down the northern sky with my iPhone XS. While the satellite shined noticeably for about twenty seconds, this merging of a 1-minute, 610-photo session reveals that it was still dimly visible for some time before and after that. The rest of the satellite reflects the sunlight, just not as effectively.

Device: iPhone XS
Settings: 26mm - ISO 2500 - 1/15s - f/1.8
Filters: None
Time: 2019-02-17 19:04:46-19:05:46 KST
Location: Naju, Korea
610 photos merged with Startrails 2.3

The Super Moon as the first Full Moon of the Year of the Pig on February 20, 2019

This lunar year's first Full Moon (Jeongwol Daeboreum) was coincidentally a Super Moon. A bigger one would not appear until December 24, 2026. I took this photo just moments after the phase reached its peak, with a visible size of 34' 02.37" and a distance of 350,840km. As a result, it appears nearly 4,000 pixels wide (3,955 pixels, 0.516"/pixel). Unless I keep using the P1000 seven years later, this would be the largest photo of the Moon this camera would ever take.

Device: Nikon P1000
Settings: 3000mm - ISO 100 - 1/400s - f/8
Filters: None
Time: 2019-02-20 01:11:49, 01:14:07 KST
Location: Naju, Korea
2 photos merged with Pixelmator 3.8.1

Clearest view of ISS yet from Nikon P1000

International Space Station seen on the evening of February 11, 2019

Animation of the ISS overpass
Previous tracking of the ISS with Nikon P1000 was alright, but I felt that it could have been better. Another good tracking opportunity came up yesterday, so I got my equipment ready and had another shot at it. Suffice to say, the results were quite satisfactory. Many of the shots came out quite clearly, you could see the division of each section easily. I did not need to resort to stacking - only the brightness and the sharpness were adjusted here.

The Space Station came closest to the observer on 18:48:49 (third photo) at a distance of 428km. Altitude from the ground was 411km at the time. You can see that the shots before that had the Zvezda module (lowest point in the second photo at 18:48:19) pointing at the observer, while the shorts after that had the Kibo-Harmony-Columbus modules (lower part of the middle section in the fifth photo at 18:49:49) doing that. Another thing to note is that I was looking at the general direction of the Sun, which had just had set below the horizon, before the space station made the closest approach. As the solar panels are always facing the Sun, I would be looking at the back of them in the first and the second photos, which is why they aren't illuminated and visible there.

Device: Nikon P1000
Settings: 3000mm - ISO 100 - 1/400 to 1/640s - f/8
Filters: None
Time: 2019-02-11 18:48-18:50 KST
Location: Naju, Korea
Defined tags for this entry: , , , ,

Interviewed by KBS for the regional news

My family appeared on the KBS Daegu News 9 on February 3, 2019

During this year's Seollal (Korean Lunar New Year) holidays my family traveled around the country. We visited Daegu on Sunday, February 3, as I read that its National Museum had some interesting activities for the kids in celebration of the year of the Golden Pig.

One such activity was actually colouring wooden decorations in the shape of a pig. While my daughters were busy painting them, a camera crew from KBS (Korean Broadcasting System, a national public broadcaster) entered the room and started recording the area. Then they asked me if they could film us taking part in the activity and subsequently do a brief interview. I told them that it was fine.

The whole thing seemed to be for a regional news segment, so I checked to see if it aired that night. Sure enough, that was the case and my family appeared on the News 9, as you can see above. You can click on the captured frame or [here] to see the entire segment - relevant part starts around 1:30.

Copyright (C) 1996-2018 Wesley Woo-Duk Hwang-Chung. All rights reserved.