Hybrids Beware? EPA Revises Mileage Standards

Shivetya writes "The federal Environmental Protection Agency announced a new system for determining the fuel economy of many cars and trucks. Hardest hit will be hybrids as all-electric driving is not considered. At the same time, many medium-duty vehicles will get rated, but not have to be published until 2011 This move to more realistic ratings will severely reduce the high numbers some cars have posted."

Comments

Which cars are overrated?

[toadlife]

I have Hyundai Sonata and the mileage quoted on the sticker at the lot is *exactly* what I've gotten. Aside from the hybrid variety, are certain cars more likely to get lower mileage than the EPA estimate?

air conditioning effects mileage?

[Cylix]

I thought MythBusters covered this one.

The final thoughts were that no modern air conditioning system should vastly impact gas mileage.

They even tested it on some SUV and came out with very similar gas mileage. (Windows down actually caused slightly more loss).

I'm sure someone will chime in here and clear this up a bit. I was just a bit confused when the article claimed air conditioning was a gas hog. (Note, on an older car I had when I kicked in the AC I really did feel the engine jump to compensate, but this was ages ago.)

won't change my car

[Mountain_Man87]

My little 93 Geo Metro XFi would still get pretty much the same mileage as the old EPA ratings 51/58. I currently get 57 MPG driving it like a nut. There are a few metros on the road getting 70+mpg on the road right now.

Re:GOOD.

[Anonymous Coward]

Typical drive train loss is 15-30% depending on the configuration. Smaller front wheel drive 5 speeds have less drive train loss then a rear wheel big v8 automatic with OD. I've found some comparisons with Google in the past when I was reseaching figures for my own car and setup. I have 450 rear wheel HP as indicated by various dyno runs and was trying to estimate engine HP.

Even more important then the markeeting driven "HP rating" should be a simple graph showing a dyno run with peak torque and HP noted. Oh, the graph might confuse consumers! Well we are even more let down, fooled, and confused by the peak HP claim that companies use now.

One of my compact cars is rated at 140HP. My mini van that weighs at least 1500lbs more is rated at 165HP. My van will blow that car off of the road even while pulling a 1000lb trailer. The peak HP are almost meaningless. Torque is more important for determining real world output and neither alone are as informative as looking at a dyno run sheet would be. Hell, I guess you could skip the dyno chart and include a 60ft, 1/8 mile and 1/4 time with the trap speed.

Looks like GM is none to happy

[plopez]

http://money.cnn.com/2006/12/26/news/companies/gm_ fuel.reut/index.htm?cnn=yes

Personnaly I am sort of happy to see GM get thier lunch eaten. They've been asleep at the switch for too many years.

Here an interesting article as well. http://www.cnn.com/2006/BUSINESS/12/22/toyota.reut /index.html

A choice quoute from the head of Toyota: '"The important thing is to be a leader in car-making, and that's done by improving products," he told a year-end news conference, adding that vehicle quality will be Toyota's top priority at a time of rising vehicle recalls.'

An American manager would have spoken some crap about "leveraging synergys for value added customer delight", in other words not admitting to a problem and just engaging in window dressing. American management seems to have lost thier way, focusing on image without addressing fundamentals.

95mph, 250mile, 10 minute charge electric vehicle

[Colin Smith]

5 Passengers and a load as well...

http://phoenixmotorcars.com/models/fleet.html

An electric vehicle has almost no parts which require servicing; no valves, no spark plugs, no oil to change, no air filter, no piston rings. Basically it'll last as long as the chassis is structurally sound and the bodywork remains reasonable. The only bits which'll wear out are the consumables, the battery and bearings. With a battery which can last for 20 years, there's no real reason the vehicle shouldn't do a million miles with bugger all servicing.

The battery:

"In addition to high power the Altairnano NanoSafe
batteries deliver:
  Long life - potentially up to 20+ year life
  Very fast charge - rechargeable in minutes
  Extremely wide operating temperature range
from -50C/-60F to +75C/165F
  Inherent safety - no risk of thermal runaway"

Re:Beware of what?

[JakiChan]

Blame your favorite money-losing American car company, and its support for and from Big Oil, for the lack of diesel options in US cars. I don't know what the deal with the car companies is, but yeah I do blame the US oil companies for not giving us ULSD until recently. Now that we have it, though, I hope to see nice and advanced diesel engines from the European car companies to show us stupid American'ts what modern diesel is like.

Maybe what we really need...

[istartedi]

...is something like a Nielsen rating for mileage. Pay some people to put a black box in their car that records the mileage. For new models, you just publish the EPA "laboratory" mileage. For cars with a year or more of real-world driving, they could post "actual" mileage. One big problem however, is that you might not be able to get enough people to sign up. You need enough people to sort out the lemons (although if mileage lemons are produced, that's important to know).

Kudos to the EPA for taking this a step closer to the real world.

Now, it wouldn't carry the same weight as a controlled data-gathering or testing effort, but is anybody aware of a mileage website, where people just enter their mileage for various makes and models? Sounds like something GasBuddy could add as a feature.

Just my honest opinion here.

[jd]

MPG ratings are notoriously inaccurate because it's a single figure derived from an unscientific selection of conditions that an unscientific selection of cars is exposed to, with all other variables and parameters extrapolated and/or ignored.

It would seem much more logical to expose a truly random selection of cars to exhaustive tests over a wider range of conditions for longer periods of time. Instead of averaging, you plot against a distribution and take the average of the distribution. This, however, is not the quoted figure for any car. It's merely the baseline for that model. Each car has to have some nominal testing - at least to see if the engine will start. Assuming that the distribution will be the same with merely the offsets being different, you then derive the effective MPG from the distribution and where that specific car is believed to be on it.

You now have an MPG per car, but it's still a single value and single values are useless. I'd therefore do the above with nine distributions, not one. One for 0-25 mph, one for 25-50, one for 50-75, and each of those for smooth traffic flow, heavy traffic and stop/go traffic.

Consumers tend to drown out lots of stats, though, and nine numbers - trivial to any geek - would be murderous on your average couch potato. On the other hand, colours tend to be workable. Simply do a rainbow spectrum, where violet is so far above average that driving round the planet uses less fuel than a typical hummvee uses to get out the parking lot, and where red is where you're escorted to the grocery store by an oil tanker. Nice and visual, though with hard data for those who actually want hard data to work with.