What is an efficient ebike?

[jkbrigman]

Seriously....what is it that makes a very efficient ebike?

We've got a unit of measure that works: watt-hours per unit distance, commonly expressed as Wh/mi or Wh/km. That's offers us an excellent starting point to define meaningful efficiency ranges for ebikes. However, the question remains: "What constitutes an efficient ebike, and what Wh range are we talking about?"

Under 25Wh/mi for an upright mountain bike?
Under 20Wh/mi for an upright road bike?
Under 15Wh/mi for an unfaired recumbent?
Under 13Wh/mi for a faired recumbent?

Note that these numbers are "rule of thumb" from threads here on E-S and from recumbent ebike personal experience. But just because this is what people typically experience, is this how it has to be? Can we do better? Can we see 20 mph on 200 watts or less?

The reason for the question is because there doesn't seem to be a recognizable definition (even a loose one) of measurement and expression of what an "efficient" ebike is. There are many statements which we on E-S have begun to accept as true, like:

- pedaling more improves efficiency (measured motor power consumption, not whole-system efficiency)
- tires with higher inflation pressure are more efficient
- running ebike motors cool and at lower wattages is more efficent
- little or no headwind is more efficent
- larger diameter wheels are (generally) more efficient
- full recumbents are (generally) more efficent
- speeds below 20mph are (generally) more efficient
- fairings (generally) contribute to efficiency

This is great. These ideas are a direct result of what E-S is all about: experimentation and collaboration. But I am still left with questions like:

1) Have I achieved the best possible efficiency with the bike I converted? (an unfaired SWB recumbent)
2) How thin and high pressure can I make the tires, and how much of an effect can I have on efficiency?
3) Can I expect any kind of performance improvement by lowering the weight of the bike and the drive system?
4) Is there any way to "rank" the effectiveness of modifications and thereby set a priority?
5) Are there limitations due to weight and incident wind that set a "best practical efficiency"?

I'm getting these ideas out on the table because we are seeing, more and more, threads about efficiency and range. One of the things you DO know if you have built an ebike and been on E-S, is that "Efficiency is King" if you want range out of your ebike. There's a point where it makes no sense to carry more LiPo: it becomes too heavy and too draggy. Why carry all that LiPo if you don't have to?

Here's an example of the questions I'd like to explore in this thread: A recumbent is more efficient than an upright bike.

Of course, we accept this as a true statement. I've experienced it myself - we all know that a "typical" ebike build on a recumbent can do 20mph on about 250W of power and 15-18Wh/mile, no pedaling. We all know that a "typical" upright bike can do 20mph on 350W (more varied with incident wind) and about 25Wh/mile.

But what more is there to know about this? Is the sum total of the difference always attributable to wind drag? Can an upright bike benefit from a fairing? And if so, what kind of fairing and by how much? Can a tailcone help on an upright as well as a recumbent? Do panniers change the equation any, and if so, how? Is one or the other design (upright vs. recumbent) more or less affected by added weight? Can I ever approach the efficiency of a recumbent with an upright bike?

Please Note: This is not a speculative thread, but a solicitation of measured data and experience. I've started it in the "E-Bike Technical" section in order to gather experienced technical theory tested by experience, not speculative opinion untested by experience. Minimum standard is a photo of a CA display and a photo of the bike in question, preferably at the time of data collection. Information on terrain and wind conditions as well. Datalogger information would be even better, and graphically interpreted logged data even more welcome.

Your Experienced Posts Welcomed!

[d8veh]

You can't use wh per mile as a measure of efficiency without putting it into context. You could have the most efficient bike in the world capable of 100mph, but it won't do 25wh per mile at that speed. Secondly, to achieve the lower values usually requires pedalling, and as soon as you pedal all meaningfulness of values goes out the window, especially at low speed. I've heard loads of guys boasting about how great their bikes are on Pedelecs UK forum, but they're pedalling and only doing 10mph, so how much work is the motor actually doing? Rather than worry too much about efficiency numbers, it's better to think about choosing the right motor for the speed and type of journeys that you do. What other people do is often irrelevant.

[jkbrigman]

You can't use wh per mile as a measure of efficiency without putting it into context. You could have the most efficient bike in the world capable of 100mph, but it won't do 25wh per mile at that speed. Secondly, to achieve the lower values usually requires pedalling, and as soon as you pedal all meaningfulness of values goes out the window, especially at low speed. I've heard loads of guys boasting about how great their bikes are on Pedelecs UK forum, but they're pedalling and only doing 10mph, so how much work is the motor actually doing? Rather than worry too much about efficiency numbers, it's better to think about choosing the right motor for the speed and type of journeys that you do. What other people do is often irrelevant.

d8veh- thanks for your input!

You put it into context, please. You just now visualized "the most efficient bike in the world". What does that mean? Is it a faired recumbent with a gearmotor? Is it a velomobile with a mid-drive? Is it something that doesn't yet exist?

What I'm looking for is pure research science. What I'm asking is What is the least possible amount of electric power it takes to traverse a flat, windless road on a 2, 3, or 4 wheel bicycle? What would that bicycle look like? What is the "right motor for the speed and type of journeys that you do", and HOW do you define that?

I'm begging you: Tell me what you know, what you have seen and measured and done. What is possible? What is your experience? What have you seen and done with your own eyes and hands and bike?

(Note: edited to decrease snarkiness...sorry d8veh and all)

[migueralliart]

I think a recumbent tadpole trike with a frontal fairing and a lightweight frame plus a geared wheel in a 26" or 24" will be pretty darn efficient. You will need a controller capable if loosing the least amount of power so a sensored 36V with 4110 fets or better. Semislick tires for the road and a low CG.


When I say fairing I mean something like this

http://basicallybicycles.com/merchant/5 ... XCombo.gif

[Lock]

Sorry, didn't real the other replies, but the only part missing from your question is "at X speed".
Generally, true to all vehicles, max eff = lightest weight, fully fared with human passengers recumbent.

[jkbrigman]

Sorry, didn't real the other replies, but the only part missing from your question is "at X speed".
Generally, true to all vehicles, max eff = lightest weight, fully fared with human passengers recumbent.

I didn't leave off the speed. I want to know, at ANY speed. At ALL speeds. 15mph. 20mph. 30mph I'm actually trying not to define the criteria too tightly.

But ignore that for the moment. I agree with you, to the extent that my own experience bears out what you are saying. But I'm not ready to say "lightest weight": what if there is an optimal weight? I'm not ready to say "fully faired": what if there's an 80/20 rule, where a "20% fairing" will give "80% benefit"?

You've made me realize, the answers I'm looking for are curves, not dots. Like on the ebikes.ca simulator.

[Kinni420]

shave every hair off your body, grab a pair of speedos (so your fat ass doesnt offend) and nothing else, jump on the lightest recumbant with nearly 90% angle, add a 1s lipo with the lightest motor you can buy, point your ass downhill at 10% grade with a 20 mph tail-wind. 0 wh/mi at 60+ mph and pray you dont crash....sounds pretty efficient to me!

[John in CR]

JK,

I have to agree with d8veh, that it has to be put in context. Not only the speed, but the conditions as well such as stops and starts vs steady state speed, and flat terrain vs hills.

You've left 2 things off the list, which are motor efficiency and variable gearing. I have a more efficient hubmotor than anyone on the forum at 93% peak efficiency, but you'll never see me quoting eye popping numbers, though I have done 30wh/mi at 30mph with no faring and punching a pretty draggy hole through the air. Otherwise I can't talk efficiency, because I ride faster than typical. As far as overall efficiency the motor is using 15-20% less electricity with my everyday riding, and that's over the long haul of 4 months of daily riding, and compared to the 86% peak efficiency motors I've used for years. The way I ride with lots of stops and starts and plenty of hills, a multi-gear system would definitely result in better overall efficiency despite the greater losses in the drive system, probably even if I used a less efficient motor.

Personally I don't worry about efficiency much, because I can always toss on more batteries or ride slower to go farther. If your focus is efficiency, then you certainly need to start with defining the conditions. Even then, how are you going to factor in the compromises? eg Sure a skinny high pressure tire will have less rolling and wind resistance, but at a cost of ride comfort, safety, and durability. Of course good farings improve efficiency, and as I understand, the tail is more important than the front. Of course panniers hurt efficiency, because they increase wind resistance.

One thing I can say about wh/mile, is that with a 100lb bike, blasting around my everyday routes using well over 100wh/mile is lots of fun, and I can't wait to push that to 150 or 200.

Being less efficient is more fun, and I'm certainly not concerned about trying to save a couple of pennies per ride, especially when my fun ebikes are the cheapest amusement park rides I've ever been on.


John

[Kingfish]

JKB,
I should like to qualify myself as having the best experience with cross-country, endurance and commuting with MtBs… and no other (so there is no confusion; I’m not going to talk about speed racing ‘round tracks etc.) Allow me to abstractly answer the proposition of efficiency from philosophical perspective:

Things like wind and rain and weight and drag and physical size and shape will color any bike’s efficiency the same, so let’s discard these as being obvious.

Break the bike down into components, and align it towards the goals. We have a frame, a power system requiring integration, a concern for safety, and the will to make it happen. All affect my efficiency at production.

Whether I use 2WD or single, I want the best bang for the buck in power distribution from the batteries to the controller to the motor and back again for regen. Thus I focus on large conductors from A to B to C with quality connections. The controller circuit board has all the beneficial upgrades possible: Typically rated for 100V 40A+ per wheel (enough for my needs). Allow me to re-emphasize: Big phatty traces, big mombo caps, excellent power-in and –out. The last thing I want to worry about is voltage sags when I needs the power. I don’t want to melt phase wires or char connectors. An added luxury, if possible – I’d upgrade the phase wires to the motor (total pain in the butt, but worth it). Now I have improved my drive system to the maximum service potential for the components purchased. System resistance can never be lower. Good placement of my 12-FET controller will ensure that it does not get too hot. I’ve made my cabling as short as possible using phatty wire gauge: Can’t make any better – We’re Done. Anything other improvement we do to this bike and ryder is sheer gravy.

Commuting is an aggressive effort that consumes much power. Many people don’t think of it, but adding good forward lighting, brake lighting, and indicators is adding efficiency and safety by communicating intention. This directly impacts the economy of motion and allows for an ebiker to correctly flow with traffic at lower speeds. There’s no guesswork; drivers understand the language and gestures without interpretation, hesitancy, or fear. In Washington State, bikes are legally allowed (as opposed to required) to mount brake lights and indicators, and it just makes good sense. No riding on the sidewalk or margin (unless traffic is moving faster), no waiting at crosswalks, no dorkin’ around acting like a regular bike. The power is there: Use it and enjoy.

To go the maximum distance means that I pedal always. Maybe after 4 hours, I might coast to give myself a well-deserved break, though for the most part… I’m balls-to-the-walls pedaling hard as I can to push the bike and myself, aggressively. At the end of the ride there is both a great hunger and even greater thirst to be quenched. I will sleep well and soundly too. During these marathon events, when I’m in focus, in stride, channeling all thought, all purpose, all sweat forward onward without end, I think of myself as the embodiment of a thoroughbred stallion built for one function: To Win!

Earlier I said I wouldn’t talk about speed racing, but this, what I am discussing is a personal race against myself and no other (although I certainly enjoy passing others). So being the horse and acting out the part is the most efficient that I can be as an augmented man with his machine, and it cannot be achieved without breathing and focus and diet.

Efficiencies at this point dwell into personal preferences and features. Bike frame, seat, posture, baggage, farings, shoes, sex (not really, I just wanted to see if yer reading), and riding gear. These are intangibles that are unique to each of us. Take identical twins, outfit them the same, spin them up and turn them loose – they will perform differently because we can’t control the mind.

To conclude:

• Mount the best power system possibly onto the bike of choice. It’s better if you do the mods that way when it breaks you’ll know what to fix.
• Illuminate and indicate legally, peacefully coexisting in traffic.
• Get your mind into the game. Take no prisoners.

If you can accomplish these three concepts, it’s given that your bike, frame, components, and garments will match the quality of your commitment. Then will you be as efficient as possibly, and in Nirvana.

Cheers brother. KF

[Farfle]

IMHO, efficiency involves two things, Aero/motor/wiring/drivetrain etc losses, and how much fun you are having, where fun is inversely proportional to efficiency. As for me, im having alot of fun, if I can manage under 100wh/mi, im riding pretty efficiently.

[Alan B]

For me, efficiency is not the primary factor, it is adequacy of range. I need enough efficiency to make my trip with the battery capacity I have. If I don't have enough capacity I need to either increase capacity or efficiency. On the shorter trips it is easy to add battery and brute force it. On a longer trip this is not so easy. In that case it becomes a more interesting tradeoff.

So for the 100 mile Mega Enduro they have a 20 watt hour per mile limit. So this places a 2kwh limit on the battery. Based on my experience with the mountain bike, this is a pretty low number. So for that situation a recumbent is one way to save energy. I didn't go dig up all the numbers which I have in other threads, but there is a considerable improvement from my mountain bike with 9C 2810 to the EbikeE with BMC over my commute run.

There is a thread on ES where a lot of folks posted their energy per mile usage along with info about the conditions and machine. It would be good to mine that for some of the data you seek. But in general it might make sense to define the project goal in some manner. Specify a speed and distance goal, for example, and then designs can be considered.

[jkbrigman]

So, hey: besides Farfle, who seems to eat Kwh for breakfast, the topic DOES seem seductive, eh?

We have some awesome races - Grange proved that. I'm wondering if we could likewise do something like the Shell eco-challenge, but with ebikes?
Answer all these concerns about standardization and fixing variables. Set up divisions by type of bike (and maybe even rider weight) Hold it in a large indoor forum.

And what bikes? 700c fixies? Recumbents? Velos? Geared hubs? Belted mid-drives? Maybe even something more exotic?

No speedos for me Kinni420, only baggy pants as a public service to the world.

Keep going, keep going, I love this topic. I can't decide if it's like football players at the 1 yard line or jumping in a leaf pile, but it's great.

[ddk]

there's no way I can quantify what's happening with my tricycles.
Both are very efficient as far as I can tell.
Both use(d) the same drivetrain/motor but different controllers.
One trike is a standard sit-up-high design, speed-limited by way of controller and it's pedal gearing to ~8MPH
The other trike sits me fully 18" lower in a semi-recumbent position. I've modified the pedal gearing to be able to pedal above 20MPH. No fairing.
I pedal 100% of the time on both trikes.
They both sip about the same amount of power per mile if I ride the semi-recumbent at more than twice the speed.
-So my unscientific assumption is the semi-recumbent tricycle is twice as efficient as the regular tricycle (even though I know it really isn't)

what really begets high-efficiency is keeping the machines tuned up and their tires properly inflated to their max.
(if only I could properly tune me up the same way!) I do this before any ride.
-My neighbors think I'm forever fixing the trikes
A full fairing would probably help, yet hinder the tuning-up of the machine unless super easy to remove.

[Alan B]

There is another type of efficiency. One which I have not achieved yet. Support efficiency. Now I have to set up chargers, plug things in, push buttons, review cell voltages, etc. Low efficiency. High efficiency would be one plug, no buttons, one summary display to review. Someday I'll get there.

[dogman]

If you are going to try to compare one e bike to another, then the maximum number of variables is one. It's hard enough to do that with the same rider on the same course, just because weather varies.

Then if you introduce different riders, on different terrain, even more variables you don't controll creep in. Nevertheless, it does become possible to see a general trend in wh/mi numbers. A great number of typical mtb commuting bikes deliver a commute at 20-30 wh/mi. So if you get into the lowest 20's, likely you are doing pretty good for that type of bike. I tend to do worse, because of a largish altitude change on my commute, very upright rding position, and large panniers on the bike. I pedal, but my 50w has little effect other than to help my aging heart and lungs.

Comparing your wh/mi numbers to somebody elses without a typical speed of travel added are fairly meaningless though. One guys 25 wh/mi could be horribly inefficient if he's traveling 15 mph, while 25wh/mi at 35 mph would be outstanding.

The biggest way to throw a monkey wrench in comparisons is to let pedaling play a role. IMO efficiency of the ebike system is not increased by pedaling miles and miles with the motor off, or at 50w. It's much more efficiency overall obviously, but it's not an efficient ebike system per se. It becomes like saying you get better mileage because you park five miles from work and walk the rest of the way, and counting it into your MPG. That sounds totally stupid, but some how the same effect is counted by many as part of thier bikes efficiency. Especially by vendors of ebikes!

I started out gathering only range data, at wot with my first aotema motors. That was at 25 mph on flat or nearly flat ground. I set a route that tried to minimize the weathers effect by having the route go about 3 miles in each compass direction, and the route climed 500' vertical at a grade less than 1%, then dropped back to the starting point. This is still my test route for gathering better data on any motor changes or new bikes I build. The same thing can be done by riding a set commute route, and lots of good data can be gathered to improve your personal efficiency.

So if your bike hasn't changed it's specs, it's efficiency hasn't changed. But once you have a good data set, you can add varaibles to the bike, such as different riding positions, with or without tuck, flappy coats or skin tight clothes, etc. Then you know what the effect on your bike and rider combined efficiency those panniers have, or whatever.

Or you can test pedaling hard vs light pedaling, or no pedaling, pulse and glide vs steady motoring, or any other "how you ride" stuff to see how it effects your rider+ bike efficiency.

Similarly, you can vary something in the motor system, and make comparisons. Varying the battery type, or voltage, different controller settings, etc can be tested by limiting the variable to that one thing as much as possible.

But once you start changing multiple variables, you get into junk science. Too many variables, and you are only guessing which one had the effect. You can't judge the motor battery controller systems efficiency at all if rider variables are allowed in, such as pedaling. Once you include the rider variables, it's astounding the differences in wh/mi numbers that show.

I have always tried to discuss expected range in terms of little to no pedaling, and travel speeds close to max speed of a given combination or kit. Generally, that winds up being range at 24-27 mph with the typical dd ebike kits. Many of us will travel wot quite a bit, so I evolved my range rules of thumbs for about 25 mph, on an mtb. Anybody can slow down and pedal briskly, and just about double my rule of thumb ranges.

[Kinni420]

I have to agree with dog on this one.

Last weekend I decided to go for a ride where I 'pedal assisted my bike at around 150w up a 1% slope for 5 miles. I used less than .2 Ah for the entire ride on 18s. On the return ride I did pretty much the same down the hill, used less than .1 Ah. At over ten miles .3 Ah is pretty damn efficient but really neither much fun nor fast.

Compare that to my normal commute to and from work without pedaling, 11 miles @ 1% grade (down then up) both directions. I normally use around 3 Ah out of the 5 I own, somewhere in the 25 Wh/m range @ around 80% throttle (24-25 mph). Pretty efficient on the 2810 DD motor, upright MTB in jeans and a heavy ski jacket. I think on a recumbant with lycras (god forbid) Id drop that number to around 2.5 Ah. Wind resistance really is much more important than weight.

Also, Compare to the 30 mph winds in my face returning from work last week, 3.5 Ah used just coming home (5.5 miles up a 1%). No fun

I changed my front tire from a street 1.95 x 26 to a Knobby 2.2 x 26 last week and dropped the tire pressure from 85 psi to 40 psi (not looking forward to doing this on the motor! ) without any real appreciable change in efficiency that Ive noticed although I know its there (maybe 1 wh/mi). Givin the fact Ive gone from jeans to shorts and ditched the jacket (yay spring!) I DID notice a little less strain on the bike, but just a touch. Ill start saving numbers a bit more.

[jkbrigman]

There is another type of efficiency. One which I have not achieved yet. Support efficiency. Now I have to set up chargers, plug things in, push buttons, review cell voltages, etc. Low efficiency. High efficiency would be one plug, no buttons, one summary display to review. Someday I'll get there.

Alan - while that's not what I originally had in mind with the thread, you are 110% dead-on. An ebike has to be as practical and easy to use as a car for it to even get used. I intended to ride mine in on Friday, but I ran out of time charging and fidgeting with the battery, so I had to take the car. That's the third time I've lost a perfect ride-in day because I had to fidget with the batteries, chargers, plugs, cell voltages, etc. When your ride takes 1hr 20 minutes each way, you really can't spend any time 'tending to the battery or fretting with the voltages....so it's way past time to get those LiPo battery enclosures made...

[jkbrigman]

If you are going to try to compare one e bike to another, then the maximum number of variables is one. It's hard enough to do that with the same rider on the same course, just because weather varies.
...
I have always tried to discuss expected range in terms of little to no pedaling, and travel speeds close to max speed of a given combination or kit. Generally, that winds up being range at 24-27 mph with the typical dd ebike kits. Many of us will travel wot quite a bit, so I evolved my range rules of thumbs for about 25 mph, on an mtb. Anybody can slow down and pedal briskly, and just about double my rule of thumb ranges.

Good deal, good deal. K420, I figure this is the gist of what you were agreeing with, right?

[Kinni420]

yup

[John in CR]

For me Alan B's point on efficiency is one of my highest priorities, ie absolute reliability, minimum maintenance, and extreme convenience. After almost 3 years living with Konions with the only effort on the front end to build packs, and no balancing worries (plug in when I got home, charge unsupervised, and disco sometime later) I'm totally spoiled. Too bad the faulty Makita packs aren't nearly the same quality and easy to build a pack as they were 2 and 3 years ago. Ebikers really a source of 2p strings of those Sony 18650V or VT cells. As far as maintenance a nice soft force regen braking plus hydraulic disc brakes is great for reducing brake maintenance to about as close to zero as possible. The regen is also an important safety factor, because the discs are always fresh and cool, ready for emergency stops.

I like Kingfish's point too about lights and signals, and I never considered the efficiency standpoint before.

John

[jkbrigman]

Thought I would include Kirk's comments from a non-related thread. Note that he's getting 13.8Wh/mile at 32.4mph....that beats the heck out of what I can do on my recumbent, around 13.1Wh/mi at 20-22mph.

I'm not sure what they meant by saying that 48V is more efficient, unless it's a sweet spot of some kind for batteries, controller and motor...comment welcome...

You might be forced to go with a less expensive trike, like a Terra Trike to keep in budget but I found my Catrike Expedition to be just about the most efficient and quickest ways to get around. I use it a couple of times a month to commute 70 mile round trip thru Springfield IL and 108 mile round trip thru Peoria, IL. The traffic in both cities is pretty nuts but they see my trike just fine and give me plenty of room. I use Zippy 6s1p 8000mAh 30C packs in a 12s6p configuration, for a 48V 48 Ah battery, with a GM 48V 1000W motor laced into a DT/Swiss F630 DH rim. The cruising sweet spot on old Route 66 is 32.4 mph, 13.8 Wh/mile. I've thought about going to 72V 48 Ah but icecube and I kept running the numbers and 48 Volts is a bit more efficient. If you decide to change bikes, please keep recumbent trikes in mind. They look too low for traffic but people see them just fine.