Running costs of EV's?

sorry for being so late to the party, & thank you for explaining how you choose to work in theoretical maximums. 1st - No . . . coal is not the primary electricity producer. In the United States, less than 40% of electricity nation wide now is manufactured by coal, & it's dropping steadily. Next, you need to consider another maximum. Most electric vehicles are sold & operated in California where the Lion's Share of electricity comes from cleaner natural gas & other sources, much of which are renewables. Consider our ev for example; after generating enough juice for our entire house & car charging, we over generate a $150 - $200 yearly PV surplus {2.9¢/kWh) that we either take as a pay check or as a credit. We charge off our own pv panels. We've averaged 5 miles per kWh for almost 75,000 miles. Maintenance? 2 sets of wipers & 2 sets of tires. Maybe by 100,000 miles we'll need to swap the traction pack for ~ $5,000. Our pv panels have already paid for their self (comissioned end of 2008). So tell me, how does that stack up against your European diesel which BTW uses a very over-fluffed mpg formula. It would be like us claiming our EV gets 6miles per kwh ( the equivalent of 180 eMPG ... or about 2X the theoretical European diesel best ... until you throw in diesel maintenance which really kills the diesel number) . And of course, if you're 100 MPG diesel is one of the cheat vehicles, then I don't know how you want to massage those numbers to factor in cost of medical/ respiratory disease, because we are talking theoretical maximums, right?

.

 

unfortunately, it is more expensive to drive ev in some parts of the u.s.a.

 

Absolutely!

It looks like the Gen-4 Prius without road taxes can achieve parity with our BMW i3-REx. But everything is in motion and the best we can do is a snapshot in time. Which brings up this recent posting in a BMW i3 Forum:
Charging overhead - why L1 is so inefficient. - Page 3 - BMW i3 Forum

I did the following by monitoring the average speed. If it was below 26 mph (41,6kmph) from a stop, I would resume at a higher speed 30-35 mph (48,0-56,0kmph) until it was back to 26 mph (41,6kmph). No heating or A/C with cabin temperature controlled by partially opening windows as needed. There were two segments:

• Lat 34.652870 Lon -86.571707 Elevation 183m Temp 66 F (18,9C)
• Lat 34.714168 Lon -86.669213 Elevation 201m Temp 61 F (16,1C)

Here is the source data:

• 14.6 mi (23.4km)
• 25.4 mph (40,6kmph)
• 6.8 mi/kWh (10,9km/kWh or 91,7kWh/km)

• 21.4 mi (34,2km)
• 26.3 mph (42,1kmph)
• 7.9 mi/kWh (12,6km/kWh or 79,4kWh/km)

• Segment 1 - 91,7kWh/km * 23.4km = 2145,8kWhr
• Segment 2 - 79,4kWh/km * 34,2km = 2715,5kWhr
• Vehicle = 4.9 kWhr, EVSE = 5.8 kWhr ~84% efficiency, ~0.9 kWhr lost
• 5.8 kWhr / (14.6 + 21.4) = 161.1 kWhr / mi. ($0.58 @$0.10 / kWhr)

My EVSE can handle up to 40 A charging but the BMW i3-REx only draws 30-31 A, ~7.2 kW/hr.

In my normal commuting, speeds are significantly higher. I estimate:

• ~0.5 mi @25 mph, ending in long stop light delay
• ~6 mi @55 mph, ending in long stop light delay
• ~1.5 mi @45 mph, ~1 stop light delay
• ~1 mi @65 mph, 1 stop light delay

I can replicate this benchmark and set the EVSE maximum charge rate to 12 A. This will simulate using the L1 charger that comes with the car. However, I suspect we won't see a significant change in the round-trip efficiency because the built-in charger is likely the greatest source of loss. Converting AC to DC is difficult to do efficiently.

...

ps. I used 1.6 km/mi for conversions. Standard units used "," for decimal point; nothing used for 1000 unit marker, and no " " between quantity and units. Results were rounded to nearest 0,1. If non-USA numbers use "," as the decimal place maker, what do they use for 1000 unit marker. For example, does 2145,8kWhr becomes 2,145,8kWhr?​

I did this test to take another look at our BMW i3-REx efficiency and replicate what was reported by someone else who did a Volt vs. BMW i3-REx test.

Bob Wilson

 

bob, for us simpletons, are you saying L2 isn't more efficient than L1 if you're onboard charger is the bottleneck?

 

L1 is ~ 12% less effivient than L2

 

I have to complete the test which I can do tonight. Part of the problem is we don't have enough detailed metrics but this is what we know:

• Built-in AC-to-DC charger - depending upon which source, no one is claiming over 92% and even then under special conditions and loads.
• Full-wave bridge - if silicon, 0.7V x 2 ~= 1.4V - so there is a minimum 12A * 1.4V = 16W vs 30A * 1.4V = 42 W BUT this won't work because of something call 'power factor.' The distortion of a full-wave bridge induces a lot current distortions that can cause distortion in the power lines and inefficiency.
• Power-factor circuit - this circuit 'chops' the current so it follows the AC voltage profile. Then this power is voltage adjusted to match the DC load of the battery.
• L1 (120VAC) source - the actual peak-to-peak voltage is 120VAC * 1.414 = ~170VAC. But the battery voltage, 360VDC is so much higher. So the input power has to 'boosted' to put a charge in the battery and this is not 'free'. Typically starting at 12A, there is an (I**2)*R loss, the square of the current.
  
• L2 (240VAC) source - 240VAC * 1.414 = ~339VAC, still to low to charge the 360VDC battery. Again the input power has to be 'boosted' to put in a charge. Starting at 30-31A, the common resistance is subject to (I**2)*R. To put it in perspective, 12*12 = 144A(**2) vs 31*31 = 961A(**2)
  
• Cooling - the heat from the charger and battery has to be removed or it will 'let the smoke out'. The cooling is actually a regular A/C cooling system. If you read the original thread in the BMW forum, you'll find the other guy was in 50F weather which would reduce the cooling load and improve his efficiency. My test was 10F warmer in the 61-66F range.
  
• Battery hysteresis - you don't get out the same charge you put in due if nothing else from internal resistance losses.

So my car measured electrical load over the EVSE charge power in at ~85%

There may be other losses but sad to say, we don't have any metrics except vehicle load and EVSE charging power. These other losses can be measured but this is a little beyond this synopsis. Since my car is still under warranty, I would prefer NOT dig too deep for now.

Bob Wilson

 

thank you so very much.

 

This is possible because the efficiency of a boost converter is a function of the voltage gain. Stepping up from 120VAC to 360VDC is a bigger change than a 240VAC to 360VDC. Unfortunately, I don't have the specific circuit and metrics to do an efficiency analysis and there are several variants. But a generic version can help explain what goes on.

A 'boost' converter typically has a 'big nice person inductor', a Schottky diode, and one serious power MOSFET. The MOSFET turns ON and the current flows like a short into the inductor limited only by the speed the magnetic field expands. Then the MOSFET turns off and the collapsing magnetic field induced one heck of an inductor kick-back. The Schottky rectifies this pulse into a capacitor and the load. None of this is 'free' and leads to heat and power losses with more loss if the A/C has to remove the heat. But as the voltage 'boost' ratio increases, more current needs to flow into the inductor and we're faced again with (I**2) * R losses.

Bob Wilson

ps. I study this stuff for 'fun.'

 

So, what is the primary electricity producer nation wide?

 

Google: US Energy Information Information Administration

Bob Wilson

 

Nationwide I believe it is natural gas.
On the east coast it is still coal as of 2014. In the rest of the country, it is natural gas.

This varies quite a bit from region to region. Minnesota used to be well over 50%, in 2015 coal was about 44% and it continues to drop. We have been very agressive at closing down coal plants though, and have less and less every year.

 

Preliminary results from 12A charging (aka. Level 1 amps but at 240 VAC):

4.1 kWh / 4.8 kWh = ~85%
4.2 kWh / 4.8 kWh = ~88% (I forgot to use the more accurate trip meter distance)

Not a whole lot of change compared to the 84% seen earlier. But I really need to use the L1 EVSE with a KilloWatt meter at 120VAC. The lower voltage should show any significant effect associated with L1 vs L2 charging.

Bob Wilson

 

Ok, we have three data points at 240VAC:

	mph	miles	mi/kWh	kWh	EVSE kWh	Eff %	max amps
1	25.4: 26.3	14.6: 21.4	6.8: 7.9	2.1: 2.7	5.8	84.5%	30
2	26.1	22.2	5.4	4.1	5.8	70.7%	12
3	27.5	33.9	5.3	6.4	7.9	81.0%	6

1. The first test consists of two segments without the 'down time' between the two segments.
2. The other tests are total segments including down time between segments.

This data suggests there is a fixed overhead in the BMW i3-REx charger. Regardless of how much power is provided by the EVSE, the overhead reduced this vehicle charging efficiency. This was a surprise but understandable.

A switching power supply has an oscillator that drives the switching MOSFET. Plus the current limitation means it has to run longer in total time and per-cycle to create the magnetic field whose collapse generates the output pulse. So there is no free lunch. Worse, you don't know unless you test and measure.

Bob Wilson

 

Assume the mi/kWh is taken from car report and kWh calculated?
Check numbers in tests 1, 3 (e.g. 21.4/7.9=2.7).

 

Thanks,

#3 was a mistake, corrected. However, test 1 comes from: Running costs of EV's? | Page 2 | PriusChat

Bob Wilson

 

So I've dropped back to L1 charging at home:

• SOC - 45.5%
• Usable kWhr - 18.8 kWh
  
• Need 100% - 45.5% = 54.5% # Needed charge %
  
• Needed charge 18.8 kWhr * 54.5% = 10.25 kWhr needed
• Efficiency 85%, 10.25 / 85% = 12.1 kWh expected
• Measured kWh = 14. 02 kWh
• Updated efficiency 10.25 / 14.02 = 73.1% (new efficiency number)

Now we did have temperatures drop to 40 F last night so there may have been some battery pack heating. Just need more metrics.

Bob Wilson

 

Having loaned out my L2 EVSE for an experiment, I have gone back to L1 charging at home:

	metric	L1	L2
1	EVSE kW	1.3	7.2
2	EVSE effic	75%	75%
3	EVSE net kW	1.0	5.4
4	charge hrs	12	12
5	daily kWh	11.7	64.8
6
7	mi/kWh	4.4	4.4
8	mi/day	51.5	285.1
9
10	Avg speed mph	25	25
11	Drive hrs	2.1	11.4
12
13	Usable Bat. Kwh	18.8	18.8
14	Bat range (90%)	74.4	74.4
15	charges sessions	1	4

An OpenSource spreadsheet, hopefully the formula are self-evident. If not, I'll repost with with them if needed. Note, these are BMW i3-REx metrics and model but the lesson learned:

1. L1 EVSE at home - for EV dabblers who really just want to drive a science experiment. Parasite on public L2 to make it practical.
2. L2 EVSE at home - for people who want a car.

With my BMW i3-REx, the engine extends the range to over 1,000 miles in a single, 24 hr day. But if I want to pay $0.25 / 10 miles for as many miles as I want around town, I have to use a 30A, L2 EVSE.

Bob Wilson