Researchers tally real-life mileage for plug-in cars

If you're wondering how the familiar term "gas mileage" translates to a car that runs partially on electric batteries, you're not alone.

Industry group SAE International plans to recommend that the Environmental Protection Agency use "electricity per mile" in addition to the familiar miles-per-gallon rating for plug-in electric vehicles, according to a member of the SAE committee tasked with the job. The EPA is working on mileage ratings for plug-ins, which are poised to enter the market, and reviewing its rules for displaying fuel economy on car stickers.

Because efficiency for gas-electric hybrids is far more tricky than gasoline-only vehicles, the National Renewable Energy Laboratory recently said that it has developed a method researchers say accurately reflects real-world mileage of plug-in hybrids, which can vary greatly with driver behavior.

To get real-world mileage for a plug-in hybrid, researchers have come up with a formula to convert standard tests for a chassis dynamometer, seen here at Argonne National Laboratory, into mileage ratings.

(Credit: Argonne National Laboratory)

Government agencies and automakers have been studying the question of mileage for gas-electric vehicles for years. But the issue rushed to the forefront in August when General Motors said that its forthcoming Chevy Volt will get 230 miles per gallon in the city and "triple-digit" combined city and highway mileage driving based on a draft of the EPA's methodology. The EPA has not verified GM's claims, as the tests have not been completed.

Within six months, an SAE committee plans to recommend to the EPA that plug-ins come with fuel-economy stickers that show both miles per gallon and electricity per mile, said Jeff Gonder, a research engineer at the National Renewable Energy Laboratory and a member of the committee.

"There are two different fuels that are being used so you need to report what the usage is for those two fuels," said Gonder. "If you combine them into one (number) artificially, you can't derive a final output like annual costs" or annual greenhouse gas emissions from a car.

Having a rating for electricity per mile allows a consumer to figure out how much it costs to run a car per mile by using the local per-kilowatt-hour electricity cost, he added.

In addition to cost per mile, there are a number of other proposals to measure fuel efficiency for electric cars. They include an electric car's range--a big limitation of all-electric vehicles--or miles per gallon equivalent based on the energy in liquid fuels and batteries.

Recalibrating your dynamometer
With multiple alternatives and a lot at stake, it's unlikely that the question over how to represent fuel efficiency on a sticker will be resolved quickly. Sedans such as the Chevy Volt, Nissan Leaf, and plug-in Toyota Prius are scheduled for release over the next two years.

But labels aren't the only problems that new auto technologies introduce. The automated tests used to measure fuel economy before vehicles are sold need to be adjusted as well, according to NREL researchers.

That's particularly important with plug-in electric hybrids--essentially the same type of vehicle as today's hybrids with bigger batteries--because actual mileage will vary significantly based on driving conditions and how often a car is recharged.

Plug-in hybrids run almost exclusively on battery power for the first 20 or 40 miles, with the battery working with the gasoline engine after that. Driving mainly off the battery will be cheaper in part because electric motors are relatively efficient. So the fuel economy for a 40-mile drive will be substantially better than when a person drives 200 miles in a plug-in hybrid, since the bulk of the driving will be fueled by the gasoline engine, Gonder explained.

To come up with a mileage rating today, cars run a course on a machine called a dynamometer--essentially a treadmill fitted for cars and trucks--and the results are converted into miles per gallon. The current conversions don't work well because plug-ins operate in two modes--the first 20 or so miles when the car runs mainly on batteries and then in the "charge sustaining" mode for longer rides, said Gonder.

To address that issue, NREL researchers devised a formula to convert plug-in hybrid car performance on dynamometers to reflect actual driving performance, he said.

"We're trying to set appropriate expectations for what vehicles will get over a long period of time," said Gonder. "We're trying to predict the average (mileage) based on how often they drive between recharging."

Researchers found that the expected results matched actual mileage of a fleet of Toyota Priuses converted to be plug-ins operated by Idaho National Laboratory. Gonder said the methodology needs to be tested with other cars, but should be able to be adjusted for different types of plug-in vehicles, including the range-extended Chevy Volt.

The data also made clear that the cost of operating a plug-in hybrid will vary significantly based on driving style and frequency of charging.

The annual fuel cost of Idaho National Labs' plug-in Priuses ranged from $987 a year--in the case of an aggressive driver who never recharges from an outlet--to $478 per year with the driver charging about every 30 miles and seeking to maximize fuel economy. The average came to $789 per year with daily charging, from the equivalent of 55 mile per gallon mileage.

Updated on October 6 at 11:30 a.m. PT with corrected credit on photo caption.

[SteveChicago]

This is important info for consumers to understand. I had a friend who bought a hybrid about 5 years ago, but did not realize that it was not that beneficial for him since he did a lot of highway driving, which in his model meant the engine was running the car.

[ikramerica--2008]

And it depends on the car. Prius still gets very high highway mileage, especially in a flat region, but some of the hybrid SUVs don't get any benefit on the highway due to their weight and large aerodynamic footprint.

Ultimately, the rating should be cost per year, like with furnaces and clothes dryers and refrigerators. As long as the cost per gallon and cost per kWh are kept constant (they don't have to be true, just CONSTANT) one can easily compare vehicles based on cost per year.

12,000 miles per year, $3 a gallon for gas, XXX per kWh for electricity (no idea), assumed plugged in 250 nights a year (off-peak). Something like that.

That last part about the XXX is important, because I am sure the vast majority of people couldn't tell you what their kWh rate is (considering it varies based on baseline/overage/off-peak, etc.) A figure such as cost per year is going to be easier for people to understand.

But it has to be made clear to the consumer that this is not the true cost to run it, but for comparison purposes. So even if you end up paying 50% more in your region to run the car per year, the point is that if it's rating was 25% better than the other car you wanted, it will still be 20-30% better than that car run in your region.

[Renegade Knight]

@ikramerica--2008

Cost per year? Heck no. Efficiency is a better and more direct rating. That's MPG or Miles per Unit of Electricity.
The stickers on my appliences have always been hard to use. I can't make a direct comparison. Only "this one is about mid way through the line"

[Joe Real]

@renegade knight:
It is a lot better to express the fuel consumption as gallons per 100 miles and/or kWH per mile. The calculations of fuel costs of different cars and hybrids are much easier depending upon your commuting distances. This argument of using gallons per 100 miles or kWH per mile have been floated around by leading researchers and has been published here earlier.

for example, a big rectangular vehicle is rated at 20 gallons per 100 miles, you'll know right away that it cost $0.60 per mile if the price of gasoline is $3/gallon. If you used 5 mpg, the calculations would be longer.

I liked the idea of using kWH per mile, because it gives you directly how many cents you use per mile while in electric mode when you know how much your electricity cost is. Even if your vehicle is an E-REV like the GM-Volt, you can easily calculate your typical cost of your commuting trips even if you hit the gasoline mode or just pure electric mode or any combo.

[mmormando]

I've had my 2nd gen Prius for almost 6 years now, and I do almost all highway driving. I average 52 mpg winter and 54 summer. I've noticed that I get poor mileage if I take trips less than about 10 miles, so taking my car to get lunch at work tends to make the mileage suffer, but with my 30 mile each way commute I usually do ok. So now I'm really not seeing the advantage of the plugin.

[Renegade Knight]

With some hybrid specific driving skills you can get 50+ in town as well.

[Joe Real]

With GM-Volt, it would be excellent for commuting trips like your 30 mile one way commute. If they allow you to recharge at the office, you won't need any gas at all.

If they don't allow you to recharge at the office, then you'll get 40 miles of electricity power and 20 miles from gasoline at about 50 mpg.

The Volt has slightly lower drag coefficient than the Prius, so it is slightly more efficient in pure electric or pure gasoline mode.

But I reserve judgment, we should await real world data when the Volt comes.

[carlhage]

I'm surprised this SAE proposal is news-- the way to measure plug-ins using the EPA driving cycles has been defined for quite a while now, and there seems to be agreement on the reporting (MPG+Wh/mi). THE SAE keeps it's standards secret, though, so you need to pay for them to read them. There is also consensus that MPG-Equivalence should not be used (don't convert kWh to gallon-gas-equivalent either by BTU/gal-gas or average fuel-BTU/kWh for US electric generation). The details can be found by googling vss_05_duoba.pdf (In a press release, GM used an "SAE" test to come up with the bogus MPG without Wh/mi, but didn't say what the test was.)

I agree with ikramerica-- reporting $/yr or even better, $/150Kmi (divide by 10 to get $/yr) using standardized rates is the best way to help consumers. Refrigerators do it-- why not cars?

[ikramerica--2008]

And if people think this will create confusion of "real" dollars v. theoretical dollars, then create an inverted scale of come kind from 1-20, and call it the "automobile efficiency rating" similar to how AC units get a rating. Set whatever cost per mile a 10MPG combined vehicle gets as 1, and then set 20 as the figure a theoretical 200 MPG (gas only) car would get, based on the same fuel costs.

20 MPG = 2
30 MPG = 3
100 MPG = 10
etc.

But, the key here is that it's really based on the cost per mile of a 20MPG car, not the fuel efficiency of that car, so you can place a plug-in hybrid in this same scale, or a true electric vehicle, or a CNG vehicle, whatever you want. If a plug-in hybrid has a theoretical cost per mile equivalent to a 70 MPG non-hybrid car, then it gets a 7. And you can fit diesels in here too, based on the average cost of diesel fuel. All you are doing is replacing MPG with a number that is MPG/10, but taking off the "scale" so that other vehicles can be placed in the same scale.

And the value of using this kind of scale, is that it can be regionally adjusted based on the regional average costs of the various fuels involved. A window sticker can have the national and the regional scale. And that is important because depending on where you live, the value of electric over gas is going to vary.

And for those who believe that CO2 emissions even matter, you can use the same scale for Carbon footprint, where higher is better. A 10MPG car gets a 1, a 100MPG car would get a 10, and then you can figure out where a diesel and plug-in hybrid, as well as all electric, fit in based on the CO2 used for each type of car per mile. Or I suppose, if people are confused by higher being better for this scale, you could invert it x10 or something, so the best is 1-, and the worst is 10+.

[Frank Mlinar]

I totally disagree with any $/anything. It's a measurement at a single instant in time, and means nothing a day or two later. I want to see an energy per mile standard. Energy doesn't change from day to day. Since we are hopefully going away from gasoline, we need a different standard than mpg. I vote for kWh/mile.

[Renegade Knight]

Good Point.

Miles per Kilowatthour. Good.
MPG unplugged. Good.

Cost per mile? What the heck does that mean?

[EvanSei]

it is about darn time I am sick of car companies telling us a car gets 700 miles to the gallon (like the volt, I can't quite recall the exact MPG but it was some ungodly high number) eventually I want to be able to accurately compare hybrids to plugins to all combustion vehicles.

[galeso]

The Prius was advertised at 60 EPA MPG and then they said that you should not expect to get the EPA MPG. I liked that. I once got over 60 MPG because the traffic on I-90 would not let me go fasted than 40 and I seldom had to break. At 75 MPH, I think I get about 30 MPG.

[bobnaumann]

Just tell me the miles on the stored electrical charge and the city and highway mileage in the hybrid mode after the stored charge has been expended. What could be simpler?

[cbscowards]

Well, that's not enough iinformation to know the cost of operating the vehicle. You also need to know how many KwH it takes to recharge the battery after it is depleted.

[Renegade Knight]

@cbscowards

They could do that simply enough with a charging efficiency.

[htomfields]

You can learn more about INL's energy and environmental projects at the lab's facebook site.
http://www.facebook.com/idahonationallaboratory

[Joe Real]

When rating the kWH per mile, I hope that the SAE will define it as the kWH you draw from the WALL outlet of your house to the mileage driven by the vehicle while in electric mode.

With this definition, it will expose how efficient is your charging system, the efficiencies of your batteries, and the electric propulsion motor.

Many manufacturers, especially those of the NiMH types, will surely be inclined to report the kWH per mile based on the stored energy of their batteries. The charging efficiency of NiMH types are a lot lower than today's more advanced Li-Ion polymers and iron-phosphate types of batteries.

So if you base the kWH per mile on the amount of battery energy stored, that is bogus number not to be trusted, for all we know, we could be throwing about 50% of the energy between the wall and your battery during recharging because of inefficient recharger and battery chemistry.

[chash360]

Here is an idea...how about using the range as the metric to determine the numbers. How many average miles (report city and highway driving numbers) will the vehicle go, fully charged and/or fueled, before requiring a recharge/refuel. Then, how much fuel/energy is required to refuel/recharge to full again (this is a fixed number). Then divide.

Regardless of what driving mode the vehicle is using, these figures are fixed, and should not change, and can be reported as purely separate (only on gas, only on electricty) and combined (full runout of both, you will never get better than this number) based upon what operating modes are available in the vehicle.

For a person to know how much it costs, requires the person to use local costs of electricity and/or gas, as well as their own driving habits and needs, something which can not be accomplished with a standardized cross-comparable nationwide metric.

When I look at a car I look for how many miles I get out of a tank, and how much that tank costs. Most people only calculate their own MPG by the tankful anyway, they don't do it by the mile or by the gallon.

So City/Hwy Miles per tank/recharge and Tank/charge capacity (as determined by how much it takes to fill) The subjective numbers are in what is considered City vs. Hwy, not the energy consumed, and from experience they seem to have a good grasp on that already, they just need to publish what it is.

Trying to simplify this metric into other (fewer) terms will not tell me how beneficial a particular hybrid, etc. will be for my commute.

[c|net Reader]

The ratings should include different distances per trip, too. That is, as the story notes, some cars do really well for the first 20 or 40 miles after which their efficiency wanes. Thus, a rating that shows the efficiency at, say, 20, 40, 60, and 80+ miles per trip would allow a consumer to find an efficient car for the driving that consumer does. If one only drives 20 miles per day, the other ratings aren't important. If one drives 40 miles per day, the 20 miles per trip rating doesn't matter so much.

As for the discussion of how to rate these vehicles, there should be a common set of numbers, expressed in similar ways, for gas only, hybrid, and electric only, and for city versus highway driving. They should be expressed in terms of distance per unit of energy (or the inverse, it really doesn't matter). For gas only, the unit of energy is a gallon of gasoline (and it should probably be rated for 20% ethanol, not pure gasoline). For an all electric, there should be mileage tiers as described above. A hybrid should probably be treated like a gasoline-only vehicle.

You can well imagine web sites appearing with calculators that permit entering ratings for different sorts of vehicles, local costs per KWh and gallon of gasoline, and distances traveled to compute annual costs for different vehicles.

[weegg]

Leave it to the gov't to mess up the standards.

Electric only vehicles should have the following:
1) Freeway miles w/o amenities
2) Freeway miles w/ AC/Wipers/Lights/stereo on
3) City miles w/o amenities
4) City miles w/ AC/wipers/lights/stereo on

Hybrids:
Electric Range mileage in 4 categories like above - no gas motor assist
Gas assist with same categories for gas tank usage

[Jammer14]

As mentioned by others, a key factor in determining the cost of running any kind of rechargeable vehicle is of course the cost of electricity. The following link contains an excellent summary for the US on a state-by-state basis.

http://www.eia.doe.gov/cneaf/electricity/epm/table5_6_a.html

As can be seen, the average cost per kwh for the entire US in July 2009 was about 12 cents per kWh for residential customers, 10.2 cents for commercial customers, and 7.1cents for industrial customers.

However, there's a wide spread from state to state, as shown by these examples:

US average: 12 cents (res), 10.2 cents (com), 7.1 cents (ind)
Connecticut: 20.2 cents (res), 15.6 cents (com), 14 cents (ind)
Texas: 12.7 cents (res), 9.7 cents (com), 6.5 cents (ind)
Washington: 7.9 cents (res), 6.9 cents (com), 5.4 cents (ind)

Now, according to GM the battery pack in the Chevy Volt has a useable energy storage capacity of 8kWh, which is good for a range of about 40 miles. Plugging in some of the above numbers, that 40 miles would cost a residential Connecticut customer $1.62, an average US residential customer $0.96, and an industrial Washington customer $0.43.

Look at it from another perspective though, taking the average US residential customer as an example. Assume they drive 12,000 miles per year, get 25mpg from their standard gas-powered family car, and pay $3 per gallon for gas. They will buy 480 gallons of gas, costing $1440.

If they buy a Chevy Volt and never drive more than 40 miles between recharges (very unlikely) they'll use no gas at all. The total energy cost for 12,000 miles @ $0.96 per 40 miles will be $288.

More realistically though, let's assume they do 6,000 miles on gas and 6,000 miles on electricity. GM claim roughly 40mpg for the gasoline engine in the Volt, so the gas cost @ $3/gallon would be $450. The electricity cost for the other 6,000 miles would be $144, bringing the total energy cost for the year up to $594.

That's an annual saving of $846 over the 25mpg gas-powered family car. However, if that gas-powered car costs $25k to buy and the Volt comes in at it's projected cost of $40k it would take almost 18 years to recoup the $15k cost differential. Even if a (taxpayer-subsidised) rebate of $7.5k brings the Volt's price down to $32.5k it would still take almost 9 years to break even, assuming that the new technology in the Volt doesn't have any other long-term costs or savings compared with the gas-powered car.

Compared to an existing Prius hybrid (or a diesel-powered car for that matter) costing $25k and getting 50mpg the numbers are worse. Assuming diesel also costs $3/gallon, the annual fuel cost for both these vehicle types will be $720. The annual cost saving for the $32.5k Volt using 50:50 gas:electricity would then be only $126, and it would take almost 60 years to break even on that basis.

Having said all that, I'm still looking forward to the Volt coming out, and I'll seriously consider buying one if the price and technology look good enough. It isn't all about fuel costs, and we have to work through the first, second and third generations to get to the fourth and fifth.

Also, I'd expect the electricity pricing structure to become an issue if/when plug-in cars become common. I know that gas prices vary significantly from state to state (Alaska = $3.35 per gallon today versus $2.47 in Arizona = 36% higher), but a spread of 19.7 cents per kWh for a residential customer in New York versus 7 cents for an industrial customer in nearby Pennsylvania (181% higher) seems likely to generate some kind of energy-smuggling market.