Home        The Boat        Photo Essay        Contact

Fuel Hogs: Why Motorcycles get such poor mileage.

Weighing in at anywhere from 250 to 800 pounds for the bike and rider it would be expected that motorcycles should get great gas mileage. The reality though is that motorcycles usually have a hard time going more than about 50 or 70 miles on a gallon, and many of the more powerful machines routinely go only 30 or 35 miles on a gallon. There are two main reasons that motorcycles use so much fuel, and a whole bunch of other transient situations that add up to unbelievably high fuel consumption. The biggies are transmission inefficiencies with single cylinder engines and the disproportionally high wind resistance of a lightweight vehicle going very fast.

Performance of Typical Motorcycles
Transmission Inefficiencies
Two Stroke Engines
Fast and Light
Heavy Acceleration
Wheel Bearings
The Cost Issue

Performance of Typical Motorcycles - What kind of mileage some fairly typical motorcycles of different types are capable of.

Any motorcycle that can go 50 miles on a gallon has traditionally been considered a rather fuel efficient bike. There are a number of big cruisers that can do this well on fuel with about one to one and a half liter displacement two cylinder engines tuned for fuel efficiency. These large displacement bikes usually have to go pretty fast though to be able to get 50mpg, and idling in traffic sucks down a lot of fuel. Dirt bikes also will typically do at least 50mpg if they are geared high and ridden fast on the open highway. Some modern fuel injected dirt bikes do somewhat better on fuel, a good example being the KTM 690 Adventure which is said to be able to deliver 60mpg on the highway. The 2014 KTM 690 is also said to be the most powerfull single cylinder motorcycle ever sold. Smaller displacement can yield even higher mileage, a good example being the 1980 Kawasaki KZ440 that I rode every once in a while over the course of more than a decade. If it was tuned just right it would do 70mpg either idling along at 2500RPM in higher gears or screaming at 6000 to 8000RPM accelerating hard out of turns and then braking into the turns. If the speed was increased to over 50mph on the highway though the mileage began to drop off and 55 to 65mpg was typical for higher speed cruising on mountain highways. The KZ440 was a pretty hot sport bike for it's day, but by modern standards it is really a slow street bike. The 40hp it was rated to do was enough to accelerate hard on smaller roads and it felt quite fast in third and fourth gears. At 80mph in sixth gear on the freeway acceleration was rather slow even though the engine speed was right in the meat of the power at 7500RPM. The main advantage of the KZ440 was that it was one of the smallest displacement two cylinder motorcycles ever built. The butterfly valve carburetors and points ignition system were unsophisticated, but could be made to work fairly well under some conditions. Smaller displacement bikes can do even better, and things like a Suzuki DR250, a Yamaha XT225 or a new fuel injected Honda CRF250 street bike are often said to do about 70 or even 80mpg if they are ridden at medium speeds. For off road ridding though the mileage drops off a whole lot and about 40mpg is considered very high mileage for fast ridding on tight trails. The old cylinder port two stroke dirt bikes did a whole lot worse, typically delivering around 17 to 25mpg. The Italians claim to have a 125cc cylinder port two stroke street bike that can churn out 33hp and also deliver 70mpg. I don't have any experience personally with the small scooters, but 85 to 120mpg is often claimed. Even with a single cylinder gasoline engine a small two wheeler that is limited to 35mph should quite easily be able to attain 200mpg at 35mph and even higher mileage at somewhat reduced speeds. If a vehicle is run at it's maximum speed problems both with engines that will not run under a reduced load and transmission inefficiencies on single cylinder engines are much less severe. People who have put 10hp direct injection Hatz diesels on full sized motorcycles report around 125 to 175mpg cruising near the top speed of 50mph. Ten horsepower should be enough to push a full sized motorcycle somewhat faster than 50mph, and more like 300mpg should be attainable cruising at reduced speeds. Production diesel powered motorcycles have also been built that deliver much higher performance and slightly better than 100mpg. Fuel mileage record attempt racing where the rules dictated that the vehicle speed remain between 15 and 35mph was stuck for many years at about 800 to 1200mpg using modified Hatz single cylinder diesel engines. The technique was to accelerate to 35mph, then shut the motor off and coast down to just over 15mph before starting the motor again. Then in the year 2000 the record was blown away by the Cal Poly team with a run at well over 2000mpg. Their success was probably due to finally building an entire engine and injection system small enough to cruise continuously at some speed closer to 15mph. A vehicle getting 2000mpg at 20mph would burn just a hundredth of a gallon of number two diesel fuel per hour, which would be less than a quarter horsepower. That quarter horsepower output could be produced by a four stroke engine of as little as one cubic inch displacement spinning at 2000RPM. The major problem with such a small engine is of course getting the piston speed high enough for any kind of descent efficiency. For 2000RPM operation the absolute minimum stroke length would be three inches. A one cubic inch displacement three inch stroke engine would have a bore of only 0.65 inches, which does not leave much room for valves. Since direct injection diesel engines can remain quite efficient down to about a one quarter load the obvious solution is simply to use a somewhat larger displacement engine than would strictly be required. A two cubic inch engine with a three inch stroke would have a bore size of nearly one inch, which would give more room for the valves. Since the engine would then also be running under a substantially reduced load there would tend to be much less trouble with getting enough air through the intake valve (or valves as such an extremely radically undersquare engine at 2000RPM might benefit from a four valve per cylinder configuration).

Transmission Inefficiencies - Because of large torque spikes from a single cylinder engine motorcycle transmissions have to be strong, high capacity gear boxes.

A typical six speed transmission on a 500 or 600cc dirt bike is so burly that it could easily be used behind a 500hp small block Chevy V8 racing engine. This is a big reason why motorcycles use so much fuel. The transmission is just a really huge high capacity gear box that has high frictional losses. Gear boxes for single cylinder engines can attain rather high maximum efficiencies up in the neighborhood of 85%, but when the load is reduced the efficiency drops off dramatically. A typical four cylinder automotive engine and manual five speed transmission might attain a maximum transmission efficiency of 95%, and when the load is reduced to one quarter the transmission efficiency drops off to perhaps 80%. For a single cylinder engine though that 85% efficiency under a full load drops off to something more like 50% under a one quarter load. That is a dramatic difference! These numbers for transmission efficiency are just out of the hat estimates, but the trends in real transmission are of a similar nature. Transmissions with more selectable ratios tend to have worse problems with the efficiency dropping off under partial loads, and more complex transmissions such as those with reversing gears or power take offs also tend to suffer worse from efficiency fall off under partial loads. One advantage for motorcycles is that they don't require reverse, making the transmission simpler. Anyone who has ever rode a heavy motorcycle though might argue that some form of reversing mechanism would in fact be of great benefit for maneuvering in tight spaces. Two cylinder motorcycles only suffer half as much from transmission problems, although typically the design of motorcycles masks this advantage. Many single cylinder bikes have minimalistic transmissions to deliver the highest possible efficiency under reduced loads, and these heavily loaded minimalistic transmissions wear out rather quickly. Two cylinder motorcycles in most cases use somewhat burlier transmissions that will last for many tens of thousands of miles. What is observed is that two cylinder motorcycles do a slight bit better on fuel, but tend to last a whole lot longer. An electric drive for a single cylinder motorcycle could probably be made to do better than a gear box (see Motors and Generators).

Two Stroke Engines - Great potential, but poor practicality.

Two stroke engines also are easier on transmissions, with a single cylinder two stroke being equivalent to a two cylinder four stroke engine. This is in fact one of the main reasons that two stroke bikes became so popular in the 1960's and 1970's. Not only can a two stroke potentially make twice as much power for the same displacement, but much higher transmission efficiencies are possible. The reason that two strokes have once again fallen out of favor is that cylinder port two strokes just don't work well at all. There are three main problems with cylinder port two strokes: Carburetion, piston melting and narrow power bands. Any carbureted two stroke is going to blow unburned fuel and lubricating oil out the exhaust under all operating conditions, and under some conditions the amount of unburned fuel and lubricating oil that is blown out the exhaust is quite large. Even for applications where low fuel efficiency and dirty exhaust emissions were only small problems such as chainsaws and dirt bikes cylinder port two strokes just did not work very well due to piston melting problems and narrow power bands. These two problems are due to the exhaust port being in the side of the cylinder. With the hot exhaust gasses exiting out a cylinder port the piston, piston rings and cylinder wall get very hot in the immediate vicinity of that exhaust port. The overheated cylinder wall, piston and rings all wear out more quickly than they otherwise would and catastrophic failure has also tended to be a rather large problem. The intake ports do not cause nearly so large of problems for engine longevity and reliability, but any cylinder ports do tend to require long strokes and long pistons. And this brings up the final problem with cylinder port two strokes, the range of engine speeds over which they work well tends to be extremely narrow. For a cylinder port two stroke to run well the shape of the ports and the shape of the exhaust pipe must be tuned to allow a large intake charge to enter the cylinder without too much raw fuel being blown out the exhaust. Generally it is only possible to get this to work well over a narrow range of engine speeds. And to make matters worse this tuning of a cylinder port two stroke also typically only works over a narrow range of engine loads. What this means is that cylinder port two strokes can make big power from small and lightweight engines with small and lightweight transmissions, but all they can do is make that big power at a single engine speed. If cylinder port two strokes are tuned to run better over a wider range of speeds and loads they can't make as much power for a particular displacement and they typically become dirtier and less efficient as well. There is however hope for two strokes. In recent decades direct injection two strokes have been built that blow little or no unburned fuel out the exhaust. The typical direct injection two stroke still relies on oil mixed with the intake air to lubricate the bottom end of the engine though. The lubricating oil mixed with a small amount of gasoline is injected into the intake air to lubricate the main bearings, rod bearings and piston skirts. A portion of this lubricating oil and gasoline still ends up blown unburned out the exhaust under many conditions, but at least the amount of fuel blow unburned out the exhaust can be dramatically reduced. Preventing the lubricating oil from being blown out the exhaust requires a radically different type of engine as case reed induction is no longer possible. Two stroke diesel engines have been used with good results, and they rely on an engine driven blower to force the intake air into the cylinder. A direct injection gasoline two stroke could potentially be built that would use some sort of a blower so that the crankcase would be free to be filled with lubricating oil. As far as I know such an engine has not yet been built.

Fast and Light - Regardless of what sort of an engine and transmission are used motorcycles tend to seem like they use a lot of fuel at higher speeds.

None of the production motorcycles mentioned at the beginning of this article have engines or transmissions that do a great job at any speed or load, but it can still be seen that higher speeds use a lot more fuel. With the smaller displacement bikes the mileage drops off dramatically at speeds above 50mph and at freeway speeds the displacement of the engine often seems totally insignificant for how much fuel it uses. The reason for this is that the frontal area of a motorcycle is quite high compared to the light weight of the bike and rider. A small well shaped car really does not take all that much more power to push through the air up at 70 or 80mph. Of course sport bikes with fairings and hunched over seating positions do somewhat better at high speed, but it is not fair to compare them to the upright seating positions that are universally preferred for comfort and ease of ridding.

Heavy Acceleration - Most motorcycles have a lot of power, and heavy acceleration with braking does tend to use more fuel.

Most motorcycles get just as good mileage if they are ridden fast and aggressively with hard acceleration and some braking. The main reason for this is that gasoline engines run much more efficiently up at over 6000RPM, and motorcycle engines typically attain rather good maximum efficiencies at wide open throttle around 5000 to 8000RPM. For more on why gasoline engines must spin so fast see Combustion Properties of Fuel. The other big reason that heavy acceleration does not appear use more fuel is that the transmission only attains high efficiency under a heavy load. The thing is though that the heavy acceleration does contribute to poor mileage even if the bike is not capable of doing better when ridden more conservatively. Any time the brakes are used into corners energy is being wasted, and motorcycles typically do more braking than any other type of vehicle. Heavy acceleration also tends to lead to high maximum speeds even on smaller roads that would normally be considered to be low speed roads. These higher maximum speeds get the motorcycle into that situation where air resistance is disproportionately high for the size and weight of the vehicle.

Wheel Bearings - Universally inefficient, but unequaly significant.

The angled cylindrical roller bearings universally used in all road vehicles for the past 80 years or so are really a very inefficient type of bearing that has a rather large frictional drag under all conditions. Because these have been the standard type of wheel bearings for so long they are simply known as roller bearings, even though there are a number of other types of roller bearings that could be used. Why these inefficient angled cylindrical roller bearings are significant for motorcycles has to do with the low speeds at which motorcycles typically attain their best fuel mileage. The heavier and faster a vehicle the less significant the losses from the inefficient angled cylindrical roller bearings are. This is however a bit of a tricky distinction, because it might be thought that a slower and lighter vehicle could simply use smaller bearings that would cause less frictional losses. This would tend to be true to a certain extent, but the reality is that even slow and light vehicles are required to have rather burly wheel bearings to be able to hold up to the aggressive high speed driving that they are likely to encounter at times. In practice it is the more efficient vehicles able to go farther on a gallon of gasoline at slower speeds that suffer more from the inefficiencies of angled cylindrical roller bearings. Some motorcycles have used ball bearings for wheel bearings, and many older automobiles from the early 20th century did not use angled cylindrical roller bearings for the wheel bearings. The older cars and trucks used a combination of parallel cylindrical roller bearings to handle the radial loads and two sets of ball bearings to handle the thrust load. This was a more efficient bearing system, but it was fraught with difficulties from poor placement of the sets of bearings and poor adjustment procedures. The main design problem with the early 20th century bearing systems had to do with only one parallel roller bearing being used for each wheel. With two widely spaced parallel roller bearings per wheel these parallel roller bearings can handle both the weight of the vehicle as well as the torque load from cornering and running side hills with the ball bearings only needing to support the thrust load from cornering and running side hills. If only one parallel roller bearing is used then the ball bearings have to support both the thrust loads as well as the torque loads, and this means that the ball bearings either end up overloaded and prone to failure or the ball bearings have to be much larger in size. The other big problem with the early 20th century parallel roller bearing systems had to with the adjustment of the ball bearings. If the ball bearing sets were adjusted too tight then they would end up carrying the full weight load of the vehicle with the parallel roller bearings doing nothing. To best ensure that the ball bearing sets carry only the pure thrust load it is be best if the two sets of ball bearings are close together. This can tend to seem a bit backwards, because a wheel bearing that uses only ball bearings works best when the two sets of bearings are very widely spaced. Likewise a bearing system that uses only one parallel roller bearing works best if the two sets of ball bearings are very widely spaced so that the torque load on the ball bearings is as low as possible. The ideal location for the two closely spaced sets of ball bearings would be in the center of the bearing system, that is between the two widely spaced parallel roller bearings. Angled cylindrical roller bearings quickly became the popular choice because they are cheap and easy to manufacture as well as being easy to install and adjust. Angled cylindrical roller bearings are also used in gear boxes of most types, and ring and pinion sets always use large angled cylindrical roller bearings. Motorcycle transmissions often make extensive use of parallel cylindrical roller bearings and large ball bearings because getting the efficiency of a transmission for a single cylinder engine up is difficult and of great importance for the performance of the machine. With a four six or eight cylinder engine getting the transmission efficiency up high enough for good performance is easier to accomplish, and automotive and truck transmissions typically use mostly angled cylindrical roller bearings. Regardless of what type of vehicle or where the angled cylindrical roller bearings are located the efficiency problems will be much more severe at reduced speeds and when carrying lighter loads. Since motorcycles are typically able to operate at speeds well in excess of 100mph over rough terrain yet attain their best fuel mileage down at about 25 to 40mph the angled cylindrical wheel bearings are a particularly irksome problem. Because of their overall light weight motorcycles could easily use all ball bearings for the wheel bearings, although the cost of these bearings would still be considerably higher than the cheap angled roller bearings currently used. For heavier vehicles the combination of two parallel roller bearings to handle radial loads and two sets of ball bearings to handle the thrust loads still has great potential. There is also the possibility of another type of roller bearing which has the high load carrying and high thrust capabilities of angled cylindrical roller bearings with the high efficiency of parallel roller bearings and ball bearings. This other type of roller bearing would best be called a tapered roller bearing, and would tend to be rather complex and expensive to produce. With the rollers tapered to match the angle on which they run it is a pure roller bearing without the frictional sliding of an angled cylindrical roller bearing. The tapered rollers would however require ball bearings at their large diameter ends to handle the small thrust load of the tapered rollers being forced out from between the two races which are not parallel to each other. It is these small ball bearings at the ends of each roller that make the tapered roller bearing complex and expensive. The advantage of the tapered roller bearing over just using ball bearings though is that the high load carrying capability of a roller bearing allows more compact bearing systems that last longer under heavy loads. In fact the tapered roller bearing would have the potential to be even more compact than angled cylindrical roller bearings since angled cylindrical roller bearings are problematic in that they wear out and fail spectacularly if they are not quite a bit oversized.

Who Cares? - The Cost Issue.

One of the reasons that motorcycles use so much fuel is that the cost of that fuel tends to be very low compared to all the other costs of operating a motorcycle. With fast wearing tires needing to be replaced every few thousand miles as well as drive chains, sprockets, chain sliders and brake pads that all wear out pretty quickly motorcycles tend to be rather expensive to operate. In addition to the high cost of periodic replacement of minor wear parts the purchase price of the motorcycle is typically very high compared to the rather small number of miles it is ever likely to wrack up. The reality is that the cost of the fuel tends to seem small and insignificant compared to all the other large costs of operating a motorcycle. What this leads to is a situation where motorcycle riders can only justify the expense of motorcycle ownership for the high performance thrill ride that is possible. It is then higher performance that makes a motorcycle a better value, and fuel mileage usually is only significant for the length of ride that can be accomplished on a single tank. Since just one or two hours is usually a long enough thrill ride the distance that the bike can go on one tank of fuel is usually not much of an issue. Longer range would be nice, but lighter weight, higher performance and lower purchase price tend to seem more important for the thrill seeking motorcyclist. If most motorcycle purchasers just don't really care how much fuel the bikes use there is little incentive for manufacturers to invest in development work towards more fuel efficient bikes. This is of course not always the only reality, because many people are interested in lower fuel consumption simply from an ideological perspective. When the total cost of motorcycle ownership dwarfs the cost of the fuel though it is inevitable that this will influence the industry in general in some significant ways.

Back to Index to Technical Articles

©2014 Michael Traum