Results matching “DCQC”

Quick Chargers: Ignore The Charge Percent!

Electric vehicle drivers are excited to see the first DC Quick Charge stations coming online. Oregon and Washington have done their part to power up the West Coast Electric Highway allowing electric vehicle drivers to travel I-5 from the Canadian border to the Oregon-California border and take advantage of stations that can charge a Nissan LEAF or Mitsubishi iMiEV from empty to 80% in about half an hour. This greatly increases the usable range of electric vehicles for longer trips and also provides a safety net for rare situations when drivers unexpectedly need more than their normal overnight charge.

Unfortunately, there's a problem that is causing a lot of confusion that can result in a driver getting less charge than needed. Even though the stations are working properly, drivers may think something went wrong because of a user interface issue.

AV-DCQC-Screen.pngThe above is the screen from an AeroVironment DC Quick Charge station in Tumwater, WA, as shown while charging our Nissan LEAF in June. The screen shows the driver two pieces of information: the amount of energy delivered to the car and a charge percent.

The problem is the displayed charge percent: it is not the car's state of charge (SOC) and should not be treated as such by a driver to decide when to end the charge.

It's pretty well known that it's difficult to determine the exact SOC of a car's battery. Even the best estimate of the battery's SOC may be off by a few percent. That's not what's going on here. The SOC value reported to the station is completely artificial and differs significantly from the car's estimate of the true SOC.

In addition to showing the invalid SOC value to the driver, Blink quick charge stations also require the user to choose a station-controlled charge limit. This has two big problems. First, the LEAF wants to control the charge and will stop the charge at either 80% or near 100% based on the battery state at the start of the charge, so even if you choose 100% on the station the LEAF will terminate the charge at 80% if the car was at 50% or less when the charge started. Second, the Blink station doesn't know the real state of charge and therefore cannot know when to stop charging at the point it says it will.

Here's an example. I recently used the Blink quick charge station at Harvard Market in Seattle, WA. I arrived with just over a half charge remaining, which means the LEAF will allow me to do a full quick charge up to near full capacity. After plugging in the car, the screen on the Blink station gave me a choice of charge levels, defaulting to 80%, which was the highest level shown. I had to press a "more options" button to be able to choose a 100% charge. The graph below shows data collected from the resulting 52-minute charge, comparing the car's actual SOC with the SOC shown on the station's screen.

Blink_50_to_100_Graph.pngAs you can see, not only is the reported SOC higher than the actual SOC, the reported SOC rises more quickly, increasing the gap as the charge progresses. Throughout the entire charge, the SOC shown on the station consistently overstates the actual charge level and the problem gets worse later in the charge period. As the car gets to about 80% actual SOC, the reported SOC jumps up to plateau near 100% and just sits there for the remainder of the charge, even though the car is far from fully charged.

Had I left the default 80% setting, the charge would have stopped when the reported SOC hit 80%, but the car was really only at 73% at that time. A requested 90% charge would have stopped around 80% actual.

Any driver who sees this behavior and doesn't know that the charge percent value on the station is not the SOC would see it jump up to 97%, perhaps watch it sit there for a few minutes, and likely decide that it would be a waste of time to spend any longer waiting for that last 3%. If the driver ends the charge at that point, the car will be missing perhaps 10% of the potential charge. If that last 10% is needed to finish the journey, this could result in a very unhappy EV driver.

It's not clear where this value comes from, but displaying this invalid SOC on the quick charge stations has created user interface problem with unfortunate consequences for LEAF owners, and perhaps iMiEV owners as well.

So to any EV driver using a CHAdeMO quick charge station that shows an SOC percentage:

1. Ignore what the station shows. Put a sticky over it if you have to. Only look at the car's representation of the SOC.

2. If the station offers you different charge levels, choose 100% charge so that you get the car's best available charge level. If you want to stop the charge early for some reason, do it based on the SOC shown by the car.

I'll contact the quick charge station manufacturers to make sure they are aware of this problem. In the meantime, please help spread the word so EV drivers can get the maximum benefit from these highly valued stations.

For charts of two AeroVironment quick charge sessions, see Cathy Saxton's report. More tips for using quick charge stations are available on our Avoiding Quick Charging Pitfalls page.

US 2 DCQC Inaugural EV Rally

On Saturday, June 16, 2012, a dozen electric vehicles made the inaugural drive along US Highway 2 over the 4,061-foot summit at Stevens Pass utilizing the newly-installed quick charge stations. Most of these vehicles recorded data for driving and charging; this blog is a summary and analysis of that data.

Cars charging at the DCQC and Level 2 stations in Skykomish, WA.

US2qc-skykomish.jpg

We have analyzed this data as well as measurements from other driving and have created pages with information on planning an EV road trip, including guidelines for predicting energy use based on drive conditions and tips for avoiding quick charging pitfalls.

Thanks to Ron Johnston-Rodriguez for all his work getting electric vehicle charging stations installed along US 2 and organizing this event.

Route

This event marked the official opening of DC quick charge (DCQC) stations in Sultan, Skykomish, Leavenworth, and Wenatchee, WA. It was also a test of the spacing between stations. Tom had helped with the US 2 electrification process by collecting data for driving this route in our Tesla Roadster in December, 2010, so we had good information on the energy use required for each segment.

With a carefully-orchestrated schedule from Ron, each vehicle was assigned a charging period at each station. This added a unique constraint, as vehicles would not necessarily have sufficient time for a full charge at the DCQC stations. We provided suggestions for a target charge level when departing each location and the expected energy required to comfortably reach the next station. The most demanding segment was the one over Stevens Pass; our guidance included a recommended state-of-charge level at the summit so drivers would know whether they should stop for Level 2 charging at Stevens Pass ski resort.

We got nifty SWAG from Leavenworth and Wenatchee!

US2qc-goodies.jpg

Cars

We have data from 8 Nissan LEAFs, 1 Mitsubishi iMiEV, and 1 Tesla Roadster.

The DCQC spacing worked great for the LEAFs.

The iMiEV was able to make the drive with additional charging for the segment over the pass (charging Level 2 at Stevens Pass in both directions and Level 1 at Nason Creek for the westbound trip).

The Roadster can't use DCQC stations, but with its longer range didn't need much extra energy use. Tom was able to "opportunity charge" at Level 2 while we charged our LEAF and participated in the ribbon-cutting ceremonies.

Data

For those interested in all the gory details, the data and analysis are in this spreadsheet (XLS, 214k).

Drivers recorded information on time, distance, energy, temperature, and driving conditions. The details are in the individual EV# sheets. There are summary sheets for driving and charging that compare the data for multiple vehicles.

The iMiEV (EV1 in the spreadsheet) used an amount of energy similar to the LEAFs.

There were three LEAFs with after-market state-of-charge (SOC) meters that enable more precise monitoring of battery state than the factory instrumentation. These meters show the SOC as a percentage, and also in a unit called a "gid," which represents 80 Wh of energy in the battery. The gid values fell nicely in the range of values based on the LEAF's more coarse SOC bars.

We drove our LEAF (EV4) and Roadster (EV4b) together so that we could compare energy use for the same driving conditions. They turned out to be very similar; there is a summary sheet showing the data for both cars together.

Thanks to everyone who collected and shared data: Patrick, Phil, Lee, Jeff & Mary Lynne, Matt & Laura, Bruce, George, and Mike & Kimm.

EV drivers at the US2 DCQC inaugural ceremony in Wenatchee, WA.

US2qc-wenatchee.jpg

Photo by Jessie Lin, WSDOT. Used by permission.

Quick Charging

DCQC stations made it practical to make this trip (and the return) in a single day. We learned several things about the stations with all the data collected by drivers during this event.

One of the most enlightening was confirmation of an observation during our prior DCQC experience: the station-reported SOC is not a useful indication of the car's charge.

When using a DCQC from under 50% to get to 80%, the LEAF's charge rate averaged 400-500 Wh/minute. When charging from over 50% to "full," the charge rate averaged about 200 Wh/minute.

The charging overhead (energy from the station that didn't make it into the battery) was 10-18%.

More details on DCQC are on our page with tips for avoiding quick charging pitfalls.

Driving

For each drive segment, these are the minimum and maximum kWh (and corresponding gids and bars) used between the DCQC stations. The energy use will vary based on speed and weather conditions.

Trip miles kWh gids bars
Eastbound
Sultan to Skykomish 26.4 7.20 - 8.24 90 - 103 4.5 - 5.2
Skykomish to Leavenworth 51.0 12.24 - 14.88 153 - 186 7.7 - 9.3
Leavenworth to Wenatchee 20.5 2.40 - 2.48 30 - 31 1.5 - 1.6
Westbound
Wenatchee to Leavenworth 22.3 5.60 - 5.84 70 - 73 3.5 - 3.7
Leavenworth to Skykomish 51.0 12.48 - 13.60 156 - 170 7.8 - 8.5
Skykomish to Sultan 26.5 4.80 - 7.04 60 - 88 3.0 - 4.4

Conclusions

The spacing of the DCQC stations along US 2 over Stevens Pass works well for LEAF drivers. Level 2 charging at the pass is either helpful or mandatory for iMiEV drivers, depending on the driving conditions.

We believe that an SOC meter is a valuable tool when making a trip like this, especially when pushing the range limits of the car. Because we'd projected our energy use for each segment and had a meter providing a higher resolution SOC reading, we were able to minimize the amount of time that we spent charging — including successfully skipping one station — and return home with a comfortable buffer.

Planning a Road Trip Using DC Quick Charging

Hi, this is Cathy; I'm guest-blogging on Tom's page, and he's playing editor this time!

We recently needed to drive approximately 80 miles (one way) to get to clearer skies for viewing the Venus transit. That provided a great opportunity to try out DC quick charging (DCQC) with our LEAF.

We charged up to full in Bellevue using Level 2 charging, drove to the Tumwater DCQC station, charged back to full while we watched the transit, and returned home to Sammamish.

Data Collection

We have one of Gary Gidding's SOC Meters. It shows a state-of-charge (SOC) percent, which it calculates from a raw energy unit reported on the car's CAN bus. The meter can be set to show the raw energy unit, which the LEAF community has dubbed a "gid." It is reasonably well established that a gid represents 80 watt-hours (Wh). Gids are divided by 281 to approximate an SOC %.

Note that the pack kWh is not something that can be directly measured. The car determines this value through sophisticated measurements and calculations, which result in periodic adjustments that are seen as "jumps" in the gids.

Charging

We arrived in Tumwater with 2 bars, or more precisely 61 gids, which translates to 4.88 kWh, or 21.7%.

Since we were under 50%, our (first) DC quick charge brought us to 80%. The car charged for approximately 26:40 minutes, at which point it was showing 226 gids (18.08 kWh, 80.4%). The station reported having provided 14.36 kWh, and the car showed a net gain of 13.20 kWh.

The graph below shows the status each minute during charging. The lower blue bar shows the energy in the car's pack at the beginning of the time interval. The upper red bar shows how much energy was supplied by the station during the time interval. You can see that the total energy (existing + added) closely matches the pack kWh at the beginning of the next time period. (We only logged the station-reported kWh for some of this charge session.)

The SOC % axis is scaled to correspond to kWh values (both graphs).

DCQC-80.gif

It is interesting to note that the SOC % reported by the station (gold dots) does not match the SOC % of the car. We are not sure what causes this discrepancy, but we've seen this consistently in subsequent DCQC sessions.

We expected to need more than 80% charge to make it home, and we wanted to see how to handle getting a full charge after arriving at a station under 50%. So, after the first charging session stopped at the expected 80%, we unplugged and returned the connector to the station. Then we initiated another charging session.

Our second DCQC charge took 36:35 minutes, increasing the car's charge to 266 gids (21.28 kWh, 94.7%). The station supplied 3.609 kWh and the car's charge increased by 3.2 kWh.

DCQC-80-100.gif

At the end of the second charge, the battery temperature was registering 6 bars. Ambient temperature was 61° F.

Driving

trees.jpg

We have monitored gids and miles driven over several recent drives. For freeway and combined freeway / surface-street driving we see typical energy use in the range of 250 to 280 Wh/mi, with extreme readings as low as 230 or as high as 290.

The image on the left is after we had driven 74 miles from Tumwater to Issaquah. Stopping for dinner won out over completing the drive home and growing more trees.

Data from our drives to and from Tumwater:

milesWh / miNotes
65.9 251 Bellevue - Tumwater
79.1 235 Tumwater - Sammamish, slow freeway traffic for several miles

Details on a few legs of the trip from Bellevue to Tumwater.

milesWh / miNotes
23.6 247 Freeway, uneven speeds (traffic, merges, etc), partially rainy
15.7 255 Cruise at 60
7.1 259 Cruise at 60

Gids in Relation to Bars

GIDs.gif

The bars are based on an approximation of the percentage of the maximum available energy. This maximum varies with temperature as well as with battery age as capacity is reduced.

Therefore, there is not a static gid-to-bar mapping; it will vary based on current conditions. The image on the right shows gid readings that we observed as bars dropped during driving. It also shows a simple approximation of the gid values at the bar transition points. Since the gid values tend to vary a little anyway, those values should be an easy way to get a pretty close approximation to the actual kWh (gids × 0.08) remaining in the pack. (The gid range will be between approximately 20 × bars + 20 and 20 × bars + 40.)

Don't forget that there is some amount at the bottom (2%?) that is unavailable; the car will shut down before letting you deplete the battery that far.

When we got to approximately 50 gids, the LEAF warned us of low charge with the following:

  • Warning message ("Battery level is low") on the dot matrix liquid crystal display just above the steering wheel.
  • Illumination of the low battery charge warning light (the icon showing a pump with a plug).
  • Flashing driving range (affectionately known in the LEAF community as the Guess-O-Meter).

We arrived home with 35 gids remaining and did not see the turtle.

1

Find recent content on the main index or look in the archives to find all content.