Monday, January 25, 2016

SparkEV range

What is the driving range for SparkEV? Well duh, it's 82 miles. EPA tells us so. But is it really? When I was driving gas cars normally, I was almost always getting less MPG than EPA figures. When I "hyper miled", a term for driving technique for maximum fuel efficiency, I was getting far more than EPA figures. In this post, we'll explore some aspects of driving range and why they are not simple.

Maximum range

A common question asked when people see me driving an EV is "what's the longest range you get?" The standard answer is 82 miles per EPA, but the true answer is more complicated. It is so very easy to hypermile with EV and get far more range, especially under favorable conditions. Below is a screen shot of a short drive under very favorable condition.


Excuse the dusty image; dog hair and their dust are hard to keep off plastics. SparkEV has 19kWh battery of which some less is usable. Let's assume 90%, or 17kWh. Then the maximum range would be

51.1 mi/kWh * 17 kWh = 869 miles
51.1 mi/kWh * 33.7 kWh/gal = 1722 MPGe

I probably could've done better if not for couple of stop lights. It shows that 82 miles EPA range depends on conditions with which the EV is driven. How did I achieve this phenomenal efficiency? It was mostly down hill, though I kept up with traffic.

You might argue that same condition would result in infinite miles of range in gas cars due to downhill coasting, but that isn't the case. You still have to come to a stop, accelerate on green light, and keep up with traffic. Even if gas car coasts, they still have to apply the brakes eventually, giving up energy as heat while EV with regenerative braking would add energy into the battery upon braking. In fact, it would be possible for EV to have more energy at the end of "coasting" whereas gas car would always lose energy, even for tail light when the brakes are applied. While infinite miles on gas car is not possible, "greater than infinite" miles is possible with EV given enough long downhill.

Therefore, maximum range on SparkEV is infinite miles!

No really, what is the maximum usable range per battery charge?

Obviously, we can't always drive downhill. Then how far can SparkEV drive on a charge? Someone actually drove for entire battery charge at favorable driving condition (flat road, low speed), and achieved almost 140 miles!

http://www.digitaltrends.com/cars/spark-ev-world-record/

So the practical upper limit on 2014 SparkEV is about 140 miles per charge. What the driver did was to drive at 24 MPH, lowest possible speed for cruise control, for about 6 hours. At low speed, EV does extremely well. Why not gas cars, too? We'll explore that later in this post.

Something interesting from his experiment is the miles/kWh achieved: 7.3 mi/kWh. That translates to 246 MPGe. To drive 140 miles at 7.3 mi/kWh would need

140 miles / 7.3 mi/kWh = 19.2 kWh (2014 SparkEV usable capacity)

But he was driving 2014 SparkEV with bigger battery (21 kWh) than 2015 model. The total usable battery of 2014 SparkEV is 91.3% of full battery capacity of 21 kWh. If we assume the same percentage of battery capacity is used in 2015 SparkEV with 19 kWh battery, the usable battery capacity would be

19 kWh * 91.3% = 17.4 kWh (2015 SparkEV usable capacity)

The assumption I made above of 17kWh of usable battery capacity isn't far off.
 
If we assume same power is needed to drive 24 MPH for 2015 model that has 17.4 kWh usable battery, the maximum range would be

17.4 kWh * 7.3 mi/kWh = 127 miles

But 2015 SparkEV is lighter by about 80 pounds as well as having lower gearing that's better for low speed (3.17 in 2014 vs 3.87 in 2015). While the usable battery capacity is just an estimate, it does give you an idea that the maximum driving range of SparkEV is around 130 miles when driven at 24 MPH, probably closer to 140 miles.

What range do you get for your driving?

Above examples are under pretty optimistic scenarios. Typical driving would be far different. But what is typical? We can classify freeway and combined. Living in SoCal, it's hard to avoid the freeway and just drive local, so local only figure isn't available in my experiment.

For freeway driving, I have 55 MPH figure from previous post: 84.7 miles driven with 9 miles remaining  = 93.7 miles.

http://sparkev.blogspot.com/2015/05/freeway-efficiency.html

Strictly keeping at 55 MPH with windows closed and no AC/heat, one can expect bit over 90 miles range. Obviously, this is only possible in certain times, and not in summer heat or in rain (wipers, rain drop force, etc).

At this point, let's check our results. Assuming 17 kWh usable battery capacity and 5 mi/kWh at 55 MPH, the maximum range would be

5 mi/kWh * 17 kWh = 85 miles

What!?!? How did 93.7 miles come about? Going the other way,

93.7 miles / 5 mi/kWh = 18.74 kWh

Hmmm. Something must be wrong. Indicated range is based on almost 100% of battery capacity? While it's possible, more likely is that I misread "short section of flat road" power reading; 10% off might be 10kW, and it could be possible that I was going slight uphill. In any case, 17 kWh usable battery capacity for 2015 model will be assumed based on experiment from 24 MPH for 2014 model with 21 kWh.

Then what would be mi/kWh for driving 93.7 miles with 17 kWh?

93.7 miles / 17 kWh = 5.5 mi/kWh

Wow, that seems incredible. I think 17 kWh usable capacity makes more sense. Furiously waving my hands, let's just say that SparkEV range at 55 MPH is about 90 miles.

What range do you get for your combined driving?

More typical would be combined driving over long distance. I have kept my trip meter running since getting the car. The driving includes about a dozen 300 miles per day all-freeway trips using multiple DCFC as well as uphill, downhill, local, country road, city slicker havens, stuck in traffic. Below image shows the current mi/kWh.



At 5.1 mi/kWh, 17 kWh would result in

5.1 mi/kWh * 17 kWh = 86.7 miles

That's slightly better than EPA estimate of 82 miles. But if we assume almost 100% capacity is used as suggested by 55 MPH experiment,

5.1 mi/kWh * 18.74 kWh = 95.6 miles

That's when the battery is fairly new. What happens as the battery degrades? This is just a guessing game, but let's assume 20% degraded after 3 years, with average of 15% driven at degraded state. Because the battery capacity initially diminish as exponential decay, degradation would slow down as time goes on, which means one would spend longer time in more degraded state. But remember, this is just a guess as far as actual numbers are concerned.

19kWh * 85% = 16.15 kWh
5.1 mi/kWh * 16.15 kWh = 82.4 miles

Well, well, what do you know? We're right at EPA estimated range when battery capacity degradation is taken into account. I don't know if this is what Chevy and EPA had in mind when they came up with the range estimate. It sure as heck doesn't make sense to use EPA MPGe figure to derive the range. Using 17kWh as usable battery capacity and 80% charger efficiency, it works out to

119 MPGe / 33.7 kWh/gal * 17 kWh / 0.8 = 75 miles.

Therefore, telling people that SparkEV typically gets 82 miles per charge is accurate, despite the EPA discrepancy.

Experimental range summary

Range can be summarized as follows. All are with new battery with 2015 SparkEV and using actual experimental data.
  1. 85 miles under typical conditions, summer and winter, and mostly with windows down (even in freeway) and dogs sticking their heads out.
  2. 90+ miles at 55 MPH windows up with slight elevation gain (about 500 ft).
  3. 130 miles at 24 MPH flat road without stopping.
  4. 869 miles in some long stretches of downhill, even with some traffic lights.
  5. Greater than infinite miles (?) for very long downhill.
Power vs speed: theory vs practice

It's good to know the range based on actual driving, but it's not possible to experiment at different speeds. For example, we know the range would be less if driven at 90 MPH than at 55 MPH. But I don't have the luxury of long drives at 90 MPH on flat ground; there is vast government conspiracy to take away my driving license, even when I'm not exceeding the speed limit (I was nowhere near 186,000 miles per second!).

I have to use short sections of flat road to test at lower speed. At 30 MPH, it was between 3kW and 4kW. At 55 MPH, it was almost solid 11kW. At 60 MPH, it was roughly 13kW to 14kW. Using these data points, we can construct a model for SparkEV power vs speed. Then we can figure out how much power is consumed at particular speed, and calculate the range as if it was driven at that speed.

According to Physics, power required to overcome aerodynamic drag is proportional to cube of speed. Power required to drive at 60 MPH would be 8 times the power required to drive at 30 MPH (twice the speed, 2^3 = 8 times the power). But the data I collected show only 4 times the power at 60 MPH. Physics is wrong?

Physics isn't wrong. We must also add other resistance. Tires are not friction-less, bearings, and all the moving parts contribute to some force resisting motion. As a simple model, force required to overcome "rolling resistance" is proportional to weight. Little detour in Physics lesson: That means the energy is also proportional to weight to drive the same distance (Work = Force * distance). Since the weight does not change, energy spent is constant at constant speed. But the rate of change of energy (power) is linearly related. Therefore, power needed to overcome this weight related friction is linearly increasing with speed.

But wait, there's more! When the car is "on", it is using power, whether it's the day time running lights or the radio. This power is constant regardless of speed, unless you start to mess with AC/heat or blasting the radio.

And then even more! At different power levels, current drawn would be different, but it would still go through same diameter wires. Higher current would increase electrical losses. This is not easy to quantify with the data we have. However, one can guesstimate this by looking at other electronics circuits. I use switching regulator as a model for this, which is about 90% efficient in operating range, much less in extreme low and high currents. This is also my motor efficiency (Engine efficiency).

tps54040a (42V in switching regulator)
Now we can go and do bunch of calculations to figure out the relationship between power vs speed by combining the parameters, or we search the Internet to do that for us. If we plug in the right parameters, there are web sites that give us very nice summary of calculations. Here's an excellent web site that does exactly that. The link below has all the parameters to produce the data for SparkEV.

http://ecomodder.com/forum/tool-aero-rolling-resistance.php?Weight=3000&WeightUnits=lbs&CRR=.01&Cd=.326&FrontalArea=27&FrontalAreaUnits=ft^2&FuelWh=33557&IceEfficiency=.9&DrivetrainEfficiency=.95&ParasiticOverhead=500&rho=1.225&FromToStep=5-200-5

Since not all values are available from the <sarcasm>extremely technically informative SparkEV user manual</sarcasm> following are some common values found on the Internet.

Weight (lb): 3000 lb (google)

Coeff. of rolling resistance (Crr): 0.01 (https://en.wikipedia.org/wiki/Rolling_resistance, tire on asphalt)

Drag coefficient: 0.326 (http://media.gm.com/media/us/en/chevrolet/vehicles/spark-ev/2014.html)

Frontal area (sq ft): 8.8 / 0.326 = 27 sq ft (http://www.mychevysparkev.com/forum/viewtopic.php?f=6&t=4003&start=10 gives drag area, divide by drag coefficient gives frontal area.)

Fuel energy density (Wh/gal): 33557 (summer blend drop down option)

Engine efficiency: 0.9 (guess based on switching regulator)

Drive train efficiency: 0.95 (guess; default value)

Parasitic overhead: 500 (lights are about 250 Watts, double it as guess)

Air density: 1.225 (default value)

Then it spits out a nice table. I reproduce just from 0-90 MPH, maximum speed of SparkEV and few more. I added the last column which shows the range assuming 17kWh battery capacity.




Yellow highlights the key data to be discussed. Green cells are actual experiment data that match the calculations. As you can see, 30 MPH is between 3kW and 4kW, 55 MPH is bit above 11kW, 60 MPH is between 13kW and 14kW, close to what was observed. This tells me the parameters found are pretty close to actual values.

Range data analysis

Again, let's assume the battery capacity of 17 kWh. Something to keep in mind is that these are just estimates. For example, the motor may not hold 90% efficiency throughout the speed range. Still, it gives some indication of what SparkEV ranges could be.

At 25 MPH (2.7 kW), it would be good for 6.3 hours of driving, which would be 25*6.3 = 158 miles range. In the experiment above, the driver only reached about 140 miles. Was he overweight? Did he make some stops? Did he slow down? More likely, the parameters I use are not exactly what he had. His experiment was with 2014 model which was about 80lb heavier as well as larger battery than 2015 model. To achieve close to 140 miles, one would have to drive between 30 MPH and 40 MPH. Still, that's close enough for this exercise.

We can evaluate EPA figures as well. EPA range is 82 miles. From the table, that range is achieved when driven at constant 55 MPH. Unfortunately, this includes all driving conditions, not just the highway. Then we turn our attention to EPA's MPGe which separates city and highway.

EPA rates SparkEV as 119 overall (between 60 MPH and 65 MPH from table), 128 city, and 109 highway. While city figure is complicated due to multiple stops one has to consider, the highway figure is presumably based on much simpler driving dynamics. From the table, we can see that 109 MPG is reached at speed bit above 65 MPH. I suspect it was closer to 65 MPH with the heavier 2014 model of SparkEV. Also of significance at 65 MPH is the range of about 65 miles. Considering one doesn't drive to empty battery, actual range would be less, maybe 55 to 60 miles.

An interesting aside is the theoretical maximum speed of SparkEV. While SparkEV is electronically limited to 90 MPH (and saving me from getting my license taken away... again!), ecomodder web site shows all the way to 200 MPH. SparkEV is capable of 100kW, and the speed at which 100kW is needed is over 125 MPH. Without electronic limit, that's the absolute maximum speed. At that speed, the range would be only 22.3 miles. Taking some margins into account, usable range would be 12 to 17 miles.

But SparkEV may not sustain 100kW for very long. The longest known sustained battery use is 48 kW at DCFC. 51 kW is used at 100 MPH, 45 kW at 95 MPH. Well, 95 MPH close enough to 90 MPH with some margin, and that's probably why SparkEV is electronically limited to 90 MPH. At 90 MPH, range would be only 40 miles. Taking some margins into account, usable range would be 30 to 35 miles.

In summary, the maximum range for SparkEV on flat road is between 40 miles (90 MPH) and 150 miles (25 MPH, lowest speed limit in most of US). Note that all of these values are with windows closed, no heat / AC, just simple driving at constant speed on flat road. Range will be less for any additional drain on the battery, such as two 150lb dogs sticking their heads out the window. In addition, maximum range should be de-rated to take into account some margin; 10 miles is a reasonable number. Then the usable range is between 30 miles (90 MPH) to 140 miles (25 MPH).

Screw it. It's all just too messy. Just say it's 82 miles.

Is 1000 miles per day using DCFC possible?

Previously, I wrote a blog post that 1000 miles per day using DCFC is possible.

http://sparkev.blogspot.com/2015/05/dcfc-quick-charge-and-driving-from-la.html

But is it really? Now that we have better range estimates, we can do some checks to debunk (or validate) the assertions. I'm using 17 kWh as maximum battery capacity. 80% of that is 13.6 kWh. Knowing at 12% to 89% took 13 kWh in 20 minutes, 13.6 kWh in 20 minutes starting at lower state of charge may be possible. Adding 5 minutes to get off the freeway and and another 5 minutes to get back on, total time for DCFC is 30 minutes.


Driving at 65 MPH, the range is 66 miles. 80% of that is 53 miles. Time to drive is 0.82 hours (53/65). Adding 0.5 hours for DCFC equals 1.32 hours. 53 miles / 1.32 hours = 40 MPH average. 24 hours of this would be 964 miles. It's close, but not quite 1000 miles in a day.

Doing the same with 55 MPH (83.3 miles range, 80%=67 miles, 1.2 hours drive + 0.5 DCFC = 1.7 hours, 39 MPH average), result is 934 miles, less than 65 MPH case.

Doing the same with 70 MPH (59.4 miles range, 80%=47.5 miles, 0.63 hours drive + 0.5 DCFC = 1.13 hours, 42 MPH average), result is 1006 miles. So it is theoretically possible to drive over 1000 miles in a day at 70 MPH using 21 (24/1.13) DCFC sessions. I'm not crazy enough to try this, but I'm sure there are loons out there.

Edit: 2016-10-20

Someone drove over 1000 km in one day (16 hours) with SparkEV! That might be one day distance record for SparkEV. Below is the video.



Below is the discussion. It's in French since he's in Canada (yes, they sell SparkEV in Canada, Mexico, and Korea), but you can use google translate to view in any language.

http://menu-principal-forums-aveq.1097349.n5.nabble.com/1000km-en-une-journee-td53319.html

It works out to 1050 km (650 miles) in 16 hours, which is 65.6 km/hr (40.6 MPH) on average. Had he driven 24 hours in same pattern, he would've driven 1575 km (975 miles). Such feat would not be possible with slower charging EV like Nissan Leaf (24 kWh version) or EV without DCFC like Fiat 500e.

So for now, the real world range of SparkEV per day is 1050 km (650 miles) while leaving 8 hours for sleeping and extrapolated 1575 km (975 miles) in 24 hours without sleeping.

SparkEV theoretical worst case

Above analysis is a typical case using values that match the experimental data. Now let's explore the worst case. I don't mean turn up the heater and open the windows. I mean things like changing the tire to less efficient model or the motor/battery has worse efficiency due to age.

Worst case rolling resistance for tire on asphalt from Wikipedia is 0.15, probably when using super wide tires. Who knows? Someone may want to install Corvette rims and tires on SparkEV. Some people are insane as Tesla appropriately name their driving mode.

We guessed engine efficiency to be 90% (bit less than DCFC), but it could be less especially with aging battery and not-as-well lubricated motor and gears. Using 80% as engine efficiency, the same as L1 charging efficiency, we get the following.

http://ecomodder.com/forum/tool-aero-rolling-resistance.php?Weight=3000&WeightUnits=lbs&CRR=.015&Cd=.326&FrontalArea=27&FrontalAreaUnits=ft^2&FuelWh=33557&IceEfficiency=.8&DrivetrainEfficiency=.95&ParasiticOverhead=500&rho=1.225&FromToStep=5-200-5

To save you the trouble, below is the screen shot of the table for worst case.


Well, it's not so bad. At 55 MPH, the range is only 72 miles vs 83 miles above. At 65 MPH, the range is only 59 miles vs 66 miles above. Still, it's only about 10% worse. I can live with that if the replacement tires cost $100 less.

SparkEV Conclusion

SparkEV gets about 85 miles of range per charge, but that's with new battery and without any margin. With margin, it's more like 75 miles range. Under low speed conditions, it would get far more miles per charge, as much as 140 miles at 25 MPH. For city driving and stuck in traffic, it should get much better than EPA estimate. It's simple to have longer range on SparkEV; just keep the speed low, but above 15 MPH. Using multiple DCFC sessions, even 1000 miles per day may be possible at 70 MPH or more.

SparkGas theoretical analysis

For SparkGas, major differences are weight and engine efficiency. Although SparkGas has grill opening that increases drag, we ignore that for this exercise to give benefit to gas car. What we are after is close to EPA highway MPG number of 39 MPG at 65 MPH by tweaking the engine efficiency number. Why 65 MPH? Because SparkEV's highway EPA MPGe figure matched at 65 MPH.

Because gas car is so messy in terms of efficiency, not only does it vary efficiency through RPM but braking wasted as heat, highway MPG is best we can hope for in this comparison. Multiple iterations of tweaking engine efficiency to make 39 MPG at 65 MPH result in engine efficiency of 30%. Then the following parameters are used.

Weight (lb): 2300 lb (google)

Coeff. of rolling resistance: 0.01

Drag coefficient: 0.326 (same as SparkEV, though probably worse with SparkGas)

Frontal area (sq ft): 8.8 / 0.326 = 27 sq ft (same as SparkEV)

Fuel energy density (Wh/gal): 33557 (summer blend drop down option)

Engine efficiency: 0.3 (trial and error; see above)

Drive train efficiency: 0.95 (default value)

Parasitic overhead: 500 (lights are about 250 Watts, double it as guessed figure)

Air density: 1.225 (default value)


To save you the trouble, below is the screen shot of the table.


One thing that jumps out at you is 107 MPG at 15 MPH. Yeah, this is total bogus. SparkGas will not achieve anywhere close to that. The problem is that gas engine efficiency is not a constant but varying over RPM with only a narrow range tuned for best efficiency AT FULLY OPEN THROTTLE. Typically, the maximum engine efficiency is about 75% of full power. With SparkGas having 98 HP, 75% would be about 75 HP. At that power, it would be going way over 90 MPH, probably even way over 100 MPH. While the engine efficiency might be maximum, it would get awful MPG due to aerodynamic drag.

Therefore, getting meaningful theoretical data is far more difficult with gas car; one has to know the engine efficiency at particular speed (that also varies by gears). Getting that data may be possible, but I won't bother. Gas cars suck!

So how do you maximize efficiency on a gas car? Fully open the throttle (minimize pumping loss) to reach maximum engine efficiency (about 75% of red line) beyond the speed limit, then shut off the engine to coast to below speed limit, then repeat. This is called "pulse and glide", a practice that should be avoided (just get an EV!)

Another is to simply get smaller engine (ie, less powerful). Then it's more likely to have reasonable highway speed matched to efficient engine RPM. This is why old Geo Metro (55 HP) and new Mitsubishi Mirage (74 HP) are so efficient, but they are very boring with 0-60 MPH time of over 12 seconds. Again, just get SparkEV for maximum fun and efficiency.

Hoping for SparkEV 2.0

With GM having used up half the $7500 federal EV tax credit even before first Bolt rolls off the assembly line, it's doubtful they'll continue or expand SparkEV sales. SparkEV costs $12K less than Bolt, and even with same percentage profit margin, Bolt will make more money for GM. Having SparkEV eat into tax credit at same level as Bolt makes no sense. Well, at least that would be the bean counter's argument, which I completely understand.

But I can hope and dream, can't I? Suppose there could be SparkEV 2.0, how would I go about it? Previously, I showed that longer range doesn't have to mean bigger battery. I stated SparkEV with drag coefficient of 0.15 (like 1930's prototype car) would yield close to 180 miles range while charging 140 miles in 20 minutes using existing 50 kW charger.

http://sparkev.blogspot.com/2016/01/mass-market-ev-hoping-for-tesla.html

But of course, that's wrong; that ignored the rolling resistance and static power requirement. So let's explore several cases to see what is possible: Drag coefficients of 0.24 (Tesla S, Prius), 0.22 (half way between EV1 and Tesla S), and 0.15.

First is the case of 0.15 to debunk my own erroneous assertion.


Well, the range is only 110 miles at 65 MPH. That's about on par with 2016 Nissan Leaf's 30kWh battery, but using 19kWh battery. 55 MPH yields 130 miles, similar to what current SparkEV would get at 24 MPH. Not bad, but drag coefficient of 0.15 would take significant work.

Next, let's see drag of 0.22. While this is still lower than most (all?) production cars, it may be more feasible than 0.15.


At 65 MPH, it's only 87 miles compared to 66 miles with current SparkEV. While that's an improvement, it's nothing to get excited about. I'm not too hopeful for 0.24, something that could be if SparkEV had shape of smaller version of Tesla S.


At 65 MPH, it's only 82 miles compared to 66 miles with current SparkEV. So much for my theory that only aerodynamics is needed for significantly longer range. 55 MPH range is 101 miles. Oh well, at least it can break 100 miles freeway range by looking like a miniature Tesla S.

Still, gaining 20% range with bit of sheet metal work isn't bad. This means similar reduction in drag area (coefficient * frontal area) compared to Bolt could reduce battery size while retaining the 200 miles range. At least there's some hope for Tesla to use smaller battery with 200 miles range EV, that's good.


Edit Feb. 1, 2016

I shrugged off discrepancy between theoretical 158 miles range at 25 MPH and actual 140 miles in experiment by hand waving that efficiency could be different. Now I have new data. I used 500 watts as a guess, but "bicycleguy" from SparkEV forums managed to extract more data using Arduino and CAN bus.

http://www.mychevysparkev.com/forum/viewtopic.php?f=10&t=4166&start=10

He shows 372.32 Volts total and 3.33 Amps when "on". That would be 1240 Watts, far more than 500 Watts that I guessed. At 55 MPH power of 11 kW, 1240-500 = 740W represents 7%. That's not much off, especially considering that static power doesn't even show on power display.

For lower speeds, however, it would be significant. Using 1240 W as static power at 25 MPH, the range would be

2690W + (1240W  - 500W) = 3430 W
17 kWh / 3.43kW = 4.96 hours, round up to 5 hours
25 MPH * 5 = 125 miles

125 miles for 17 kWh usable battery on 2015 SparkEV is closer to what I estimated based on 19 kWh usable battery capacity on 2014 model.

140 miles * 17 / 19 = 125.3 miles

So maybe the issue at 25 MPH range was due more to static power being different than electrical efficiency at low power levels.

For higher speed (power), static power would be less of an issue, because 0.75kW would be even smaller percentage. But motor efficiency could be different. Unfortunately, I don't have experimental data for high speed to see who far off I might be at 90 MPH, including decreased electrical efficiency. Oh well, maybe someone in Germany could import SparkEV and test it on Autobahn to see how close the theoretical values are.

Below is the new table using 1240 W as static power instead of 500 W.


http://ecomodder.com/forum/tool-aero-rolling-resistance.php?Weight=3000&WeightUnits=lbs&CRR=.01&Cd=.326&FrontalArea=27&FrontalAreaUnits=ft^2&FuelWh=33557&IceEfficiency=.9&DrivetrainEfficiency=.95&ParasiticOverhead=1240&rho=1.225&FromToStep=5-200-5

It's only 78 miles range at 55 MPH, which is even worse in not matching experimental data of 90+ miles at 55 MPH. Oh well, the parameters are still off, but until I get better tools to measure, this is it.

Edit Feb. 9, 2016

I normally leave 1 or 2 bars remaining before disconnecting the charger so I can regen as I'm going down the hill. Recently, I made a mistake and fully charged the car. The range showed 92 miles at Guess O Meter (GOM)!


Charging was after long drive home that was 3.8 mi/kWh. If we assume GOM bases miles on latest drive pattern, 92 miles after 3.8 miles/kWh would be usable battery capacity of 24 kWh. Obviously, this is wrong. Using 5.1 mi/kWh, battery would be 18 kWh. So maybe SparkEV uses 18 kWh out of 19 kWh? I don't know, maybe, maybe not. But 82 miles EPA could be said to be conservative estimate.

Edit Feb.23, 2016

There was a test done on 2014 SparkEV by driving in a big loop in San Diego freeway by Tony Williams (is he the same guy who owns quickchargepower.com?). It was found to have 97.8 miles range (92.8 driven + 5 miles remaining) when driven at 62 MPH (100 km/hr) with 5 mi/kWh average consumption.

http://insideevs.com/real-world-test-shows-chevy-spark-ev-has-substainally-more-range-than-nissan-leaf-62-mph-wvideo

That would make the usable battery capacity for 2014 SparkEV to be

97.8 mi / 5 mi/kWh = 19.56 kWh
19.56 kWh / 21.4 kWh = 91.4% usable

This is different than 24 MPH test performed by digitaltrends (see above) where the range was 140 miles and 7.3 mi/kWh (19.19 kWh, 89.6% usable). What this points to is that there could be substantial (under 2% is substantial?) difference in range and expected battery capacity. But it's only two samples, so the actual variability is unknown.

Tony also did a similar test using 2015 SparkEV, and found the range to be less, though still much longer than EPA figure. He wrote "shhhh", so you have to search for it yourself if you want to know the actual value. It's actually less than my 55 MPH test (85 miles with 9 miles remaining = 94 miles)

My 2015 SparkEV was getting "roughly" 5 mi/kWh when driven at 55 MPH, though I did not record the number after the trip was over. Due to different gearing, it's possible that 2015 SparkEV is bit less efficient at highway speed than 2014 model (higher motor RPM). In addition, 2015 has 18.4 kWh (some say 18.3 kWh) battery instead of 21.4 kWh of 2014 model.

This COULD make the range on 2015 SparkEV to be less than 2014 model WHEN BOTH ARE NEW. I have to stress "COULD" and "NEW", because the batteries are made by different vendors, LG for 2015 model, and A123 for 2014 model. The way different battery chemistries perform as well as aging would make the comparison difficult today. Suffice it to say, using the low figure of 2015 range is a good conservative estimate; if the range is exceeded, it'd be a pleasant surprise rather than the other way around.

There was also some speculation that 2015 SparkEV uses far more percentage of battery than 2014 model. Some speculate as much as 100%. The range of 92 miles from my recent 100% charging session is consistent with this speculation (18 kWh / 18.4 kWh = 98%). However, going by GOM (guess-O-meter) is not entirely accurate.

Below is a summary table of range in miles for various "simulation" models. First column is using 17 kWh and low rolling resistance tires, the first table I have above (Crr=0.01, 90% efficient, 500W static power). Second column is the same with Crr=0.015 and 80% efficiency (second table from above). Third column is the same as first column with 1240W as static power. Fourth column is the same as first column using 18.4 kWh as usable battery capacity. Fifth column is the same as third column (1240W static power) using 18.4 kWh as usable battery capacity.


Green highlights the closest to data I obtained using 55MPH. I could actually go for 4 more miles, so even this is a conservative estimate WITH NEW BATTERY.

Last columns are average and standard deviation. After all, central limit theorem says it (estimates) will all turn Gaussian eventually, I just skip the middlemen and use only 5 samples (it's a math joke). Because static power and other "issues" would play a bigger role at lower speeds, one should expect large variability at low speeds; turn off your lights and the radio to maximize efficiency at low speeds.

Some ask what the worst case range might be. With windows open, AC, heat, stereo, all of them at full blast, worst case is hard to define. Heat could be substantial in very cold weather (4kW all the time = 20% cut in range). One could have inefficient winter tires + heat + open the window for maximum energy waste. The battery could also have been abused so the capacity is far less. One could be stuck in dead-stop traffic for hours using static power (radio + heat/AC) without actually moving. Without knowing the parameters, it's impossible to say what the worst case range might be. Well, I guess we know: worst case is 0 miles range.

More reasonable case is to go by average figure as relatively safe estimate when operating on flat road with windows closed and no heat / AC. With a new battery, windows down and even little bit of AC or heater use would be in range. However, always leave few miles (I use 10 miles) as a buffer to be able to get to a charger in case the intended charger is out of order. As the EV gets older, it will have degraded battery, so the range should be derated appropriately.

And, of course, a pretty little plot of the data table, too.


No comments:

Post a Comment