The effect of horsepower on gas mileage

[Matt Timion]

I've heard this many times, and I can't confirm it. Please help.

When I first started planning to increase the fuel efficiency of my car, some people said that race builds and economy builds are almost identical. Their theory behind this was that more horsepower means the engine works less to get to a certain speed, and this in turn saves on gas.

Made sense to me.... sort of.

I mean, if I threw a turbo in my car, or made some other modification that gave me an extra 30 HP, would that mean I could travel the same speed with lower RPMs? It seems to me that is the same way the long geared transmissions work, which give very high gas mileage.

So it makes sense, if it's true, except for the fact that the guys with super "fast and furious" cars only get like 20mpg. It can't entirely be due to the transmission.

Some people have claimed that higher HP equals higher gas mileage for this very reason. it is the exact thinking behind advertisements on Cold Air Intakes, racing headers (exhaust manifolds), and other performance modifications.

So, is this true?

If so, what is to stop me from getting a junked up CRX (very light) and doing the Individual Throttle Body conversion? (the ITB conversion will use a throttle body from a Honda motorcycle and use it for each cylinder, allowing for more percise airflow and claims to increase HP by 30). Will such a modification give me better gas mileage?

And lastly, if more HP means better mileage, why do all of the gas mileage beasts (prius, insight, CRX HF, Geo Metro, etc.) all have tiny little engines with no power?

[omgwtfbyobbq]

Torque is a function of how much fuel is being combusted and some efficiency, and horsepower is torque at some rpm. Gasoline engine efficiency is, among other things, a function of displacement. A corvette and Insight are not that far off in terms of CdA and weight, but the vette get's crap for mileage due to it's large displacement engine having much more in the way of pumping losses at some given output.

Generally, manufacturers have increased power by increasing displacement, which increases pumping losses at some given output and reduces efficiency. This is why GM adopted cylinder deactivation. It allows increased engine efficiency via decreased pumping losses at some output, while still having the other four cylinders available for peak power. At maximum output a gasoline is only 10-25% less efficient than a diesel, but since the large majority of use is at less than maximum output, efficiency suffers due to pumping losses.

This is why Pulse and Glide is so effective for the Prius. By accelerating up to some speed, the driver is effectively minimizing pumping losses by maximizing output. Faster acceleration implies more fuel, and more importantly, more air, which increases the pressure in the cylinder during the intake stroke and reduces the negative work done by the pressure difference between the cylinder and crankcase. Anyway, by breaking up driving into an acceleration phase, and engine off deceleration phase, with the high/low speed window centered on some average speed, the driver will see an increase in FE compared to using the cruise control at the same average speed.

Now, a question I have is. Why don't hybrids just have 30hp engines with battery packs that allow for ~40-50 miles all electric? They'd get something like 80-100mpg in normal driving, with all the kickass acceleration an electric motor can provide. Well, I think this has to do with the manufacturer's unfamiliarity with battery packs and electric motors, they know ICE motors, and they may have heard that electric motors and battery packs can last for so long, but they don't have the long term empirical data to back this up. Preformance concerns are also present, with only 30hp, the car wouldn't be able to cruise at 90-100mph. They also have to deal with contracts that are years in advance, so they can't up and cancel all those ICE engines they've already ordered, which is why GM is in such a tight pickle. Their ~7 year development cycle is still geared for SUV's in a world where gas will probably never go below $2 a gallon, so they have to wiggle around with their parts suppliers and manufacturing such that they can sell as much as possible, and loose as little as possible.

In any event, given Toyota's short ~3 year development cycle, they should be set to roll out a Prius with more power via supercharging, a bigger, plug-in li-ion battery pack, and ~100mpg in 2008. Now, the only problem with this Prius is that on gasoline, not counting plug-in, it won't show as much of a difference when utilizing Pulse and Glide, but for the average driver, the FE will be much better.

Now that I mentioned it, turbocharging allows for the efficiency of a small engine, and the power of a larger one. It does this by reducing pumping losses at some load via the energy present in the exhaust gasses. A CAI will theoretically increase pumping losses a little bit by bringing in colder air that will decrease cylinder pressure, which is why a WAI increases efficiency slightly. But these are very small changes. Diesels are efficient because they don't suffer from pumping losses since they're compression ignition. They rely on the overlean running condition, and have no throttle plate to restrict air into the cylinder. Otoh, if a diesel ever gets down to stoich, it would probably melt because there would be gobs of fuel the didn't totaly combust and would melt through the rings.

Anyway, **** I wrote a lot. Turbocharging your engine probably won't result in any increase in FE, however, replacing your current engine with a smaller displacement turbocharged version with the same peak power will. If you toss a turbo on, you'll still need Xhp to move at Nmph, and the engine will still bring in the same amount of air, so you won't see a decrease in pumping losses. ITBs can increase hp and FE by allowing you to lean out your mixture, which reduces pumping losses and improves efficiency, but a standalone ecu can do the same much more accurately. A taller transmission/taller tires/bigger rims results in the engine turning slower at the same speed, which means it'll need more air/fuel to make the same amount of power, so that can help a bit. If you toss a 200-250hp PDI diesel with different maps in an Insight, you can probably get something like 80-100mpg EPA highway with lots of hp/tq. It'd probably make a really nice swap in a rabbit too, but the Cd's too high and mpg would be in the mid 60's. Unless aero mods really helped a rabbit down to the Insight's .25, then you'd have sports car performance with great FE.

[GasSavers_Randy]

Originally Posted by omgwtfbyobbq

A corvette and Insight are not that far off in terms of CdA and weight, but the vette get's crap for mileage due to it's large displacement engine having much more in the way of pumping losses at some given output.

Vettes are more than 3000 lbs; Insights are less than 2000. You can get 30mpg in a Vette... Chevy fans love comparing them to S2000's when someone says Hondas are fuel-efficient.

Diesel mileage isn't really fair because there's about 12% more energy per gallon. Fuel costs per mile is fair enough, and BTU per mile is fair too, but straight MPG with different fuels is silly. You couldn't compare things like CNG, and ethanol or LPG would be tragic. The lack of a throttle, high compression ratio, and the way those combine with a turbo do make diesels really efficient motors, just 12% less efficient than you think.

There seems to be a lot in common between efficiency and horsepower, but it's down to the individual thing. Like a rich mixture bumps up power, but obviously trashes mileage. Things like indexing plugs or high compression helps both.

I don't know if individual throttle bodies would help mileage or not, but I haven't heard of them used for mileage. They seem to be on things like motorcycles and performance v12's. The big thing being the great throttle response. Tuning is per-cylinder though, so it would be a big pain.

[omgwtfbyobbq]

Originally Posted by Randy

Vettes are more than 3000 lbs; Insights are less than 2000. You can get 30mpg in a Vette... Chevy fans love comparing them to S2000's when someone says Hondas are fuel-efficient.

Yes, but it can't get 50mpg, even though the vette glider is only marginally more inefficient than a insight glider. 1000lbs~500kg only adds ~10N of force to overcome rolling friction, whereas the force needed to overcome fluid friction at that speed is ~200N. Weight does not have that much of a bearing on FE, especially highway FE. A vette glider would be getting ~50mpg EPA highway compared to an insight glider if they had the same engine, but the 6.0L engine results in way too much energy lost when not using the ~400hp available. Go check out the FE ratings for the 2007 Malibu on fueleconomy.gov. They are all the same car, all have automatic 4sp transmissions, but as the engine size increases FE decreases because pumping losses increase.

Diesel mileage isn't really fair because there's about 12% more energy per gallon. Fuel costs per mile is fair enough, and BTU per mile is fair too, but straight MPG with different fuels is silly. You couldn't compare things like CNG, and ethanol or LPG would be tragic. The lack of a throttle, high compression ratio, and the way those combine with a turbo do make diesels really efficient motors, just 12% less efficient than you think.

I think the EPA factors this in because they test the carbon output through the exhaust, so when people buy diesels, they usually end up getting better than the EPA mpg rating. The most efficient gasoline engines are slightly less efficient as the most efficient diesel engines, bsfc of ~235g/kwh (gas) vs ~190g/kwh (diesel), however, the diesel engine is more or less always this efficient, while the gasoline engine is almost always not this efficient.

[GasSavers_Randy]

Originally Posted by omgwtfbyobbq

I think the EPA factors this in because they test the carbon output through the exhaust, so when people buy diesels, they usually end up getting better than the EPA mpg rating.

I thought this was interesting, so I looked it up. Net result is that diesels can put out more carbon for the same mpg. I think they exceed the EPA mpg the same way we do, in addition to diesels being less sensitive mpg-wise to a heavy foot.

They put the actual formulas they use in 40 CFR 600.113. Diesel mileage is equal to 2778 / (.866 * HC + .429 * CO + .273 * CO2), where the values are in grams per mile. The gas one got complicated in 1988, but prior to that it was the same, but with 2421 instead of 2778. It still seems to be within a couple percent of that.

Of course, diesel contains more carbon per gallon. One CA emissions study put the carbon weight fraction at 87% in diesel and 85% in gas, and the density at 840 g/L diesel and 740 g/L gas. So a gallon of diesel should have about 2770 grams carbon, and gas 2380. It's pretty close to the above figures for diesel, and about 1.5% off for gas, but CA implies reformulated gas. So, I'd say the EPA is trying to match the normal idea of mpg for both gas and diesel.

I'm not so sure anymore about the BTU content though. I got 12% from one source that said 124k BTU/gal for gas and 134k for diesel (it probably included water condensation heat). But the EPA says it ranges from 108 to 117 for actual gasoline. A couple different places put the test fuel (TIER 2 unleaded code 61) at 114.3k (0.743 s.g., 18437 btu/lbm). Also, I found one source that said 129.5k for diesel and 118.3k for B100.

More interesting was the EPA pages on testing. They have full details of the test schedules on one page. And another page has full details of each car test. This isn't the normal car list... they're individual test results. These are unadjusted: multiply by .9 city and .78 highway to get the normal ratings. It'd be interesting to figure out what the dyno settings mean, because it would summarize rolling and air resistance factors for each car. I can't find much info on them, though one (non-EPA) source mentions A as rolling, B as 'intrinsic' (whatever that is), and C as air drag, normally expressed as HP at 50 mph. These numbers don't match that though.

[Ted Hart]

Quote: Now, a question I have is. Why don't hybrids just have 30hp engines with battery packs that allow for ~40-50 miles all electric? They'd get something like 80-100mpg in normal driving, with all the kickass acceleration an electric motor can provide. Well, I think this has to do with the manufacturer's unfamiliarity with battery packs and electric motors, they know ICE motors, and they may have heard that electric motors and battery packs can last for so long, but they don't have the long term empirical data to back this up. Preformance concerns are also present, with only 30hp, the car wouldn't be able to cruise at 90-100mph. They also have to deal with contracts that are years in advance, so they can't up and cancel all those ICE engines they've already ordered, which is why GM is in such a tight pickle. Their ~7 year development cycle is still geared for SUV's in a world where gas will probably never go below $2 a gallon

In any event, given Toyota's short ~3 year development cycle, they should be set to roll out a Prius with more power via supercharging, a bigger, plug-in li-ion battery pack, and ~100mpg in 2008.

Why don't hybrids have 30 HP gas motors & more battery capacity? Well, this "rocket scientist" says...Toyota / Honda don't want to make 'em too good (save room for the improved models for the future)! Besides, shouldn't the hybrids use a little gas...to provide the revenue for the oil giants? After all, giants gotta eat, too!

[The Toecutter]

Let the giants go without. We shouldn't be forced to need them, and they could stand to lose a bit of fat.

[JanGeo]

A high torque engine running at a lower rpm to produce X hp will have less rotational losses than a low torque engine running at a high rpm to produce the same X hp. The higher rpm engine will require more gear reduction to maintain the same speed as the low torque engine which will result in the same torque at the wheel to make the cargo.

[thisisntjared]

wow this thread has a lot of fuzzy logic.... i think im just going to sit this one out.

[The Toecutter]

Torque is a function of how much fuel is being combusted and some efficiency, and horsepower is torque at some rpm. Gasoline engine efficiency is, among other things, a function of displacement. A corvette and Insight are not that far off in terms of CdA and weight, but the vette get's crap for mileage due to it's large displacement engine having much more in the way of pumping losses at some given output.

The Insight has a CdA of 5.125 square feet. This is from a published Cd of .25 and published frontal area of 20.5 square feet.

The Corvette has a CdA of 6.132 square feet, from published .28 Cd and estimated 21.9 square foot frontal area(Car and Driver).

Not only is the Insight 1,100 pounds lighter, it also uses LRR tires with a Cr around .006. The Corvette's tires are very sticky, Cr around .012.

The difference in forces to overcome for the Vette versus the Insight are actually going to be quite large. If you go to fueleconomy.gov, and note the differences in fuel economy between a V6 and V8 Mustang, or the L4 and V6 Camry, the differences for combined mpg are around 20%. The highway mpg difference between the V6 Camry and the hybrid Camry is only 18%!


Were that Vette to have the same CdA as an Insight, same weight, same tires, and keep the V8, it would probably do around 40 mpg highway. As is, the Vette does 29 mpg highway. Likewise, would the Insight have the same CdA as the Vette, were 1,100 pounds heavier, and had sticky tires with the same hybrid drivetrain, it would probably get around 45 mpg highway.



By adressing aero drag and rolling losses, we could have 40-50 mpg hwy V8 musclecars. Say, take a GM Precept concept car with a .16 Cd, and hypothetically shove a Corvette engine into it as opposed to the hybrid-diesel drive. I'd lay down money on the table that it would get better than 35 combined MPG. Having the Corvette V8 engine with a 'mild hybrid' powertrain like Citroen uses to kill idling at a stop would probably bring that number over 40 mpg combined.