Energy recovery, how does it work? Which maneuvers?

[meepmeepmayer]

So I've been planning (dreaming of) some wild EUC tours (mountains and all), and was pondering how much energy you can recover and how that affects your range.

My first intuition was downhill = battery gets loaded, but after reading around and thinking about it that makes no sense. Breaking needs power too, after all (right?).

Adding to the confusion are the companies stating you "recover energy from breaking".

--

Logically, if the motor is more or less off (= it does no work) but the wheel is turning, you should get energy back. But when does that actually happen? I'm still confused about the entire physics behind that.

So here are some questions:

• When can one actually recover energy while riding? Why? The only situation I can think of theoretically is somehow rolling down a slope while balancing the EUC upright (so the motor does nothing, until you've gotten too fast and it tries to stop you)
• What actual maneuvers, if any, allow one to recover energy? Are there "best practices" to recover energy or is it simply about conserving energy (aka slow acceleration and deceleration and whatnot)? Is there best way to ride down a mountain? Is downhill "free"? Does it depend on the slope (moderate slope = good, steep slope = breaking needed = bad)?
• How much does energy recovery matter in reality? Do you have any practical experiences in which you notice a difference in battery depending on possible recovery?

I guess this can be summarized as "how does one get the most range out of the battery?". But I'd like to have a mental model of what's going on, right now I'm lacking that.

Crazy bonus question:

• Could you charge your battery while the EUC is off (or on and upright) by moving the wheel? I'm thinking about a tiny waterwheel you could attach to your EUC during breaks

TLDR: How does this energy recovery thing work? In theory and practice?

 

 

 

[dbfrese]

You can thank the First Law of Thermodynamics for this, or the law of conservation of energy. The energy that you stored up in you by going up hill (or being uphill already)  must go somewhere when you go downhill. The motion is transformed into electrical energy.

Why? Because an electric motor and an electric generator are exactly the same thing, and what you call it at the moment depends entirely on the direction of flow of electricity either to or from the battery. As a generator (this only works when the circuit is complete, i.e. your wheel is "turned on") when you are going downhill (or braking) the energy flow is reversed, flowing back into the battery.

Note that when your wheel is turned off, the motor offers little resistance to turning the wheel. When it is on, the motor (now a generator) offers a great deal of mechanical resistance to turning. This is the braking action of your wheel. That is why a tiny waterwheel likely would not have the ability to charge up your battery. You'd have to have a very large (non-portable) water wheel!

As far as efficiency, there are losses. As electricity flows through metal, resistance turns some of the electricity into heat. No energy is lost, just transformed. The circuitry in the battery management system (BMS) and on board computer transforms more to heat. Storing and releasing energy in the battery lead to more losses to heat. As to what percentage is lost, I don't know, but it is substantial. Perhaps others could chime in on their real-world experience.

If you started out at high elevation, with uphill or straight runs strategically placed to use stored battery energy when needed (the BMS will turn off the circuit to protect the battery from over charging) you could theoretically go all day on a single charge. What would be interesting would be to find the ideal terrain shape for all day riding!

[meepmeepmayer]

Thanks for the answer, but I'm confused.

Is any downhill ride going to charge the battery? Any braking?

I was assuming it can't be too steep - because in that case the motor would have to brake (aka counteract, aka use power) constantly (to stop you from accelerating) and only moderate declines would charge the battery?

Similarly, I thought slow braking would charge the battery while hard braking would cost energy.

Basically, have the friction of the wheel on the road keep you in a stable/constant position/situation while the motor does not have to do much.

 

As far as the optimal terrain (excluding only downhill, so maybe same elevation at start and finish is a good constraint) that is a nice mathematical problem I might have a look at once I know how the use and regain of energy works exactly.

[Bob Eisenman]

"Note that when your wheel is turned off, the motor offers little resistance to turning the wheel. When it is on, the motor (now a generator) offers a great deal of mechanical resistance to turning."

 

But...isn't the resistance felt on the bot a description of the relative position resistance of the (pendulum) of the pedals as compared to the balanced, upright and on level ground position ?

Add in the bot software's request for forward or reverse types of current flow and the bots response to gravity. Isn't a high tech regenerative effect dependent on pedal position, gravity, forward velocity, bot software current switching and ? Does a MOSFET experience reverse current flow ? Does an adjacent component like a capacitor feel reverse current flow? Does the battery actually recharge or does it experience reduced demand per unit distance travelled ?

If a bot can be 16% regenerative does it mean that only 84% of the equivalent flat land speed while moving forward battery-outflow current is required while going down hill or does it mean that the battery actually sees a +16% regenerative current flow ? In other words is 16% regenerative an absolute value or a relative figure to the normal current requirement in the self balanced device.

The bot is different (in some ways) than a simple DC motor in the off position which generates current when rotation causes the copper coils to move through the magnetic field of the permanent magnets. 

[Mono]

On 25/01/2017 at 9:56 PM, meepmeepmayer said:

I was assuming it can't be too steep - because in that case the motor would have to brake (aka counteract, aka use power) constantly (to stop you from accelerating) and only moderate declines would charge the battery?

Similarly, I thought slow braking would charge the battery while hard braking would cost energy.

I assume this is correct, but it has been controversially discussed in this forum before. 

[Bob Eisenman]

With all due respect to the topic and rider tests, a previous thread URL is found below.

Another thought is whether or not the vehicle temperature increases significantly (possibly creating a false permanent battery charge value increase )  on the Ninebot or similar App  during a downhill braking run but cools down after the long braking run ? One rider claims an 11%/charge increase over one half mile of decline. Another rider claims equal battery charge level or less during a specific downhill run.

If you put your bot upright on a chair, hold it by the handle and tilt the chair, the self level position is always in the vertical. Going downhill the rider's lean back (away from the vertical) is to apply some  direction-reversing current to control forward speed resulting from gravity. 

 

 

[Chriull]

22 hours ago, meepmeepmayer said:

So I've been planning (dreaming of) some wild EUC tours (mountains and all), and was pondering how much energy you can recover and how that affects your range.

My first intuition was downhill = battery gets loaded, but after reading around and thinking about it that makes no sense. Breaking needs power too, after all (right?).

Adding to the confusion are the companies stating you "recover energy from breaking".

The EUC has two modes for braking - one is recovering energy and the other one consumes energy.

The braking mode which consumes energy is an extreme burden to the electronics, motor coils and batteries and leads to quite immediate overheating. As we discussed this once in a thread long time ago it seems this is ?with some wheels? only used at very low speeds (the last moments before standstill)

The normals braking mode of EUCs is regenerative braking. Efficiency is imho quite low - and i can't remember of anyone reporting some substantial tests. For every wheel there should exist some sweet spot (decline & speed) with max efficiency - knowing this combinations could improve the range for hilly tracks. Should be quite easy to measure this with some app like wheellog or GyroMetrics (formerly known as 9BMetrics) going up and down a steady de/incline with different speeds (and don't forget to let the battery rest a bit before taking the measurements - a loaded battery shows a bit to much and a discharged battery a bit less in the first moments...) With this also the motor/electronics efficiency could be examined going uphill with different speeds.

Some indian looked at energy recovery efficiency in his master thesis with an electric trike (BLDC as generator over a DC/DC converter to super capacitors to the battery) - so not the same setup but not something completely different. As far as i remember he reached something about 15%...

If one finds this energy recovery sweetspots for his EUC and the downward tracks allow one to drive these speeds this 10-20% could be reachable - not a real range booster but noticeable...

The best solution anyhow is take a break in the middle of the tour and recharge your wheel... (for example @1RadWerkstatt in germany offers "fast" chargers (1) for most high capacity wheels with 4A or 8A) so that recharging time stays bearable. (just one or two beers instead of a full intoxication ) 

22 hours ago, meepmeepmayer said:

--

Logically, if the motor is more or less off (= it does no work) but the wheel is turning, you should get energy back. But when does that actually happen? I'm still confused about the entire physics behind that.

...

Crazy bonus question:

• Could you charge your battery while the EUC is off (or on and upright) by moving the wheel? I'm thinking about a tiny waterwheel you could attach to your EUC during breaks

 

15 hours ago, dbfrese said:

...

 As a generator (this only works when the circuit is complete, i.e. your wheel is "turned on") when you are going downhill (or braking) the energy flow is reversed, flowing back into the battery.

...

The motor powered by an external force (braking) acts as generator no matter if the EUC is turned on or off. The difference is, that with the turned on EUC the firmware operates the mosfet half-bridges in conjunction with the motor coils as DC/DC step up converter so the battery gets loaded even with lower generator voltages.

With the EUC turned off one would have to turn the wheel quite fast so the generator produces enough voltage to "overcome" the body diodes of the mosfets and charge the battery. Imho once was reported that Solowheels can be "restarted" by turning the wheel fast enough while turned off. (imho resetting the BMS after a cut-off as normaly done by plugging in the charger)

(!) Charging high capacity battery packs with 8A is mostly no fast charging but only "normal" 1C (or still below) charging. The chargers provided with these wheels (~2A) are very slow chargers...

[meepmeepmayer]

@ChriullThank you very much for the detailed answer! I'll have to think about it some more. It's always good to hear the technical details (e.g. the electrical connection from the wheel-wires to the battery is effectively severed when the wheel is off). Maybe I'll have some more questions later.

For now, I'm mostly interested in answering which terrain can be ignored when planning a route because you'll recover more energy than you use. So basically "free" parts of a route. I want my legs to give up before my EUC does How effective the recharge is the next step then.

--

About the fast chargers: isn't this how battery fires start? Are they safe? Do they reduce battery lifespan?

I'd love to have faster charging than the ca 10h my 1300 Wh battery will need! I was assuming there's a good reason for the long charge times that every manufacturer has...

If it's only cheap chargers, I'll order a fast charger asap (would need one for 84V though). That would allow notable recharging on common break lengths!

[US69]

1 hour ago, meepmeepmayer said:

About the fast chargers: isn't this how battery fires start? Are they safe? Do they reduce battery lifespan?

I'd love to have faster charging than the ca 10h my 1300 Wh battery will need! I was assuming there's a good reason for the long charge times that every manufacturer has...

4 (or even) 8 Amp for a 1300/1600wh ACM is still not fast...

You have to see you have 6!! parallel Batterie packs which split the amperage...means even on 4 Ampere each pack only gets 0,66 Amp! or 1,25 on 8 Amp

For those Kind of batteries this is still a slow charging! 

 

You are going to get a 1,5 Amp 84 Volt Standard Charger on your wheel....Try to get a 3 Amp charger and you are double as fast and still totally safe ......

 

[meepmeepmayer]

How much Ampere would be critical (starting to be dangerous/bad/whatever) for these kind of batteries and battery configuration? Just asking to get an idea about what  slow and fast charging means.

Any why do the manufacturers have the slow chargers? Are they cheaper to produce?

[Mono]

2 hours ago, Chriull said:

As we discussed this once in a thread long time ago it seems this is ?with some wheels? only used at very low speeds (the last moments before standstill)

I am pretty sure I haven't see any data that contradicts the idea that energy-consuming breaking is also done at high speeds, so do you have a pointer?

 

[Chriull]

2 hours ago, meepmeepmayer said:

@ChriullThank you very much for the detailed answer! I'll have to think about it some more. It's always good to hear the technical details (e.g. the electrical connection from the wheel-wires to the battery is effectively severed when the wheel is off). Maybe I'll have some more questions later.

The connection is not effectively severed - you have the motor connected by two turned off n-channel mosfests to the battery. So once the motor generates a voltage bigger than the battery voltage plus the forward voltages of the two body diodes the battery is charged.

Quote

For now, I'm mostly interested in answering which terrain can be ignored when planning a route because you'll recover more energy than you use.

One limit not to be ignored will be overheating while going up steeper/longer inclines.

Also maybe going down steeper declines to prevent cut-off from overcharge protection/too fast charging from regenerative braking. Too fast charging should normally not happen, especially with the big battery packs - but can easily be checked by using some of the logging apps. (imho gw and KS exaggerate the current values sent from the wheel by a huge factor! ?2-3?)

Quote

--

About the fast chargers: isn't this how battery fires start? Are they safe? Do they reduce battery lifespan?

http://batteryuniversity.com/learn/article/ultra_fast_chargers

One should not charge above 1C. Most imho recommend charging at about 0,5C for best battery care.

So for an 84V 1300Wh pack this would mean, if it consists of:

-120 LiIon cells with each 2579 mAh (20s6p):

1C charging current would be 15,5A

0,5C charging current: 7,7AThe supplied 1,5A charger charges at 0,1C

-100 LiIon cells, each with 3095 mAh (20s5p):

... gives the same results as above (20s6p)

So with an 8A charger (0,52C) one would not stress the batteries (and charge in ~2h to ~80% constant current phase of charging ?). Whole charging time would be ~5 times as fast as with the 1,5A charger

Only thing to consider is, that normally the charging plugs and the wires from the plug to the battery are not designed for 8A and they should be replaced. Also some BMS of the battery packs are not designed to take 8A at the charge side (some wheels use battery configurations where one pack is connected after the other, so the first pack has to "take" the whole charging current)

Quote

I'd love to have faster charging than the ca 10h my 1300 Wh battery will need! I was assuming there's a good reason for the long charge times that every manufacturer has...

If it's only cheap chargers,

Imho the only point. Also "fast" chargers with ~8A mostly have a fan and they are bigger than the small 1,5A chargers...

Quote

I'll order a fast charger asap (would need one for 84V though). That would allow notable recharging on common break lengths!

I have not seen any ~8A chargers for 84V till now (2 edit) - you could asked @1RadWerkstatt (1) if they can "calibrate" their 67,2V chargers to 84V. They also could assist you with any necessary measures for plug, wires and BMS.

(1) since i mention them already the second time here and a couple of times in other threads - i have no relationsship with them except my reseller got my KS16 from them and i bought from them such an 8A charger for my KS16. ( i also don't get or want any comission;) )

(2 edit) just seen http://www.batterysupports.com/72v-74v-84v-10a-lithium-ion-lipo-battery-charger-20s-20x-36v-p-445.html . So if one likes and knows how to tinker with this there are such LiIon chargers available.

[Chriull]

19 minutes ago, Mono said:

I am pretty sure I haven't see any data that contradicts the idea that energy-consuming breaking is also done at high speeds, so do you have a pointer?

 

 

[Mono]

28 minutes ago, Chriull said:

 

Thanks, I have seen this thread and commented on it. I can't see how this suggests that power breaking is limited to low speeds. In the graphs it seems to set in above 20km/h. 

[Bob Eisenman]

On 1/25/2017 at 4:02 AM, meepmeepmayer said:

So I've been planning (dreaming of) some wild EUC tours (mountains and all)

How about a zippy ride around someplace like  Wuerzburg and just forget about all this talk about regenerative braking ? Maybe have a beer or two ? Then go back home.

[meepmeepmayer]

Yea but I need some kind of plan because that's what I eventually want to do with the EUC (which btw includes a city with "burg" in the name).

Right now I have to learn to ride first so no mountains or Würzburg for quite some time

[dbfrese]

6 hours ago, Chriull said:

The motor powered by an external force (braking) acts as generator no matter if the EUC is turned on or off.

I would still have to say that if there is not a completed electric circuit, there would be no electrical current being generated. Any energy being input into the system is either stored (as in a flywheel) or turned to heat from various sources of friction. Have you ever been to a science museum where there is a hand cranked generator attached to a light bulb? With the circuit closed, (circuit complete) the crank takes quite a bit of work to turn. As soon as you turn off the switch (open circuit) the light bulb goes out and the generator is far easier to crank. It is no longer under electrical load. Electricity is no longer being generated. 

Think of it like a battery. When the circuit is completed, electric current flows through the circuit. When the circuit is open, there is no current.

This is a real-life engineering problem for generators. For example, a hydroelectric dam, generating electricity for a city experiences a total drop in load after the transmission lines to the city are destroyed by an airplane cutting through the lines. The generators instantly go from a state of "load" to "no load" because of the open circuit. The water, however, is still flowing through the turbines. What happens? They begin to spin faster and faster. Equipment in the dam is designed to stop the water flow through the turbines in just such a situation, but it takes time, sometimes a matter of minutes. What happens to all this energy in the meantime? It is absorbed by the generator by spinning faster and faster, and may start to cause enough friction and vibration that can only be dissipated through destruction of the bearings and eventually the turbines themselves. Hydro-generators have in the past experienced such destruction, and engineering on them has improved, and they must be designed to spin quite a bit faster than when under load without excessive vibration and friction damage. All thanks to the law of conservation of energy.

[steve454]

9 hours ago, meepmeepmayer said:

I have to learn to ride first

Have you found any good videos for learning to ride?  Solowheel has some very good ones and so does Ninebot.  Funny, though, I have not seen any tutorials for Gotway or Kingsong.  Or any other wheel for that matter.

[Bob Eisenman]

10 hours ago, meepmeepmayer said:

Yea but I need some kind of plan because that's what I eventually want to do

How about incorporating a Google maps trip distance between points of interest into your learning curve. Pick a destination from your current bot position and then check your battery mileage depletion vs the Google maps trip distance. Try some up and down routes. Get an intuitive feel for what the map distance vs your bot expectations are. Google Street view (bike route) in the states where available) can preview if a sidewalk exists or if you need to try to squeeze into the road with traffic.

I use Google maps for distance and route pictures when 'scouting out' a bot trip at a knew destination reached by mass transit like a commuter rail. I wanted to ride my Ninebot from the commuter rail stop in Rockport, MA to Halibut Point State park. Google maps said the one way trip was 2.6 miles so the round trip distance was no problem for the battery. Using the Google directed route for following safer non-main roads to the same destination is fun (until the Android phone battery level or data use charge (for maps display) becomes an issue.

 

Starbuck's is always a source of power regeneration in my area with its customer provided AC outlets. A fresh charge (1 hour helps tremendously ) avoids the dilema of BMS (battery management system) limits on speed availability. Searching the map finds these Starbucks locations.

For those big mountains trips (part car part bot) check ahead to see for example if the cable car to the top of the Zugspitze allows bots or if you need to store it at the cable car base.?
 

[meepmeepmayer]

9 hours ago, steve454 said:

Have you found any good videos for learning to ride?  Solowheel has some very good ones and so does Ninebot.  Funny, though, I have not seen any tutorials for Gotway or Kingsong.  Or any other wheel for that matter.

Tons of them, so many I can even say which ones are bad It really is a practical problem now. When that's done I can think more about power stuff.

@Chriull The detailed info is appreciated. Didn't think chargers could be a cost issue.

Thank you @Bob Eisenman for the tips. But in the end I still want to know what's going on "inside" my EUC so this is why I made this thread. Even if it's not really needed. Right now I have to learn to ride first so that has priority anyways.

And I think the Zugspitze is a bit steep for a EUC, but there are definitely cable cars I might use - even though going up the mountain is half the fun and there's no reason to miss that if possible.

[Bob Eisenman]

5 hours ago, meepmeepmayer said:

Right now I have to learn to ride first so that has priority anyways.

Given the described range in miles (Gotway ACM16) and your posts suggesting your skill level it might worth focusing on padding for yourself and maybe the bot as well. The described EUC will provide (up to 60 miles) you with sufficient range for the learning experience that lies ahead whether going up or down hills. Ride safely.

[Bob Eisenman]

On 1/25/2017 at 4:02 AM, meepmeepmayer said:

What actual maneuvers, if any, allow one to recover energy? Are there "best practices" to recover energy or is it simply about conserving energy (aka slow acceleration and deceleration and whatnot)?

Have you ever thought about becoming a slope soaring pilot?

 

http://www.gettyimages.com/detail/news-photo/flying-is-a-long-tradition-in-the-rhön-the-wasserkuppe-is-news-photo/551138539#flying-is-a-long-tradition-in-the-rhn-the-wasserkuppe-is-considered-picture-id551138539

I went for an 8 mile ride earlier on a Ninebot One E+. This bot's battery takes me up to 17 miles on a full charge in warm summer weather. The BMS goes from the green zone (17 to 7 miles) to the red zone (6 to 1 miles) at the 6 remaining miles mark. Beeps and pedal tilt back occur at 5 miles and less if the rider goes too fast. In the green zone the rider can go at the Ninebot limit of about 12 mph.

In the 40 degree weather the bot entered the red zone during the last mile of the 8 mile ride. Upon bringing the Ninebot indoors after the ride the battery registered a green zone mileage left of 7.

Since the ride is a round trip no change in PE (potential energy) occurs. The return trip path was recorded on a FitBit. The Google terrain map is shown separately. Since the ride follows streets no maneuvers other than safely following the sidewalk and crosswalks are required. The worst event was a motorist honking their horn as I stood motionless at a crosswalk light waiting to press the button. I passed two pedestrians. One smiled and the second yielded to my use of the sidewalk. Intermittent use of the roads was done where handicapped access ramps were missing.

Battery use: 10 miles after room temp warming but 11 miles noted at 40 degrees

Ride distance: about 8 miles.

Spilts: as shown on Fitbit (split #4 is near the end of a downhill ride -see terrain map)

[Chriull]

On 26.1.2017 at 11:55 AM, Mono said:

I am pretty sure I haven't see any data that contradicts the idea that energy-consuming breaking is also done at high speeds, so do you have a pointer?

 

 

On 26.1.2017 at 0:41 PM, Mono said:

Thanks, I have seen this thread and commented on it. I can't see how this suggests that power breaking is limited to low speeds. In the graphs it seems to set in above 20km/h. 

You are right - there is no proved (at least that i know) contradiction to power braking at higher speeds.

We'd need methodical wheel logs (which would still be no prove if they show no power braking - one could have missed the case power braking occurs) or someone with some insider knowledge...

A strong hint is the (huge) power dissipation occuring for the mosfets (and motor coils) while power braking. Was imho mentioned by one member in either this linked thread or somewhere else?

The one kind of power braking (shorting the motor coils by the mosfets) generates power to be dissipated by the mosfets (and motor coils) proportional to the square of the generated voltage (~rpm) (Or is it in this case proportional to the square of the flowing current which is proportional to the braking torque? Or leads both just to the same result?)

The other kind of power braking (?plugging braking? - where the ?commutation sequence is changed? so the battery voltage is put on the coils with reversed polarity) generates even more power to be dissipated by the mosfets (and motor coils and this time also by the internal resistance of the batteries)

My reasoning for the existance of power braking (mainly/?only?) at lower speeds is the ineffectiveness/impossibility of regenerative braking at (very) low speeds and the (huge) power dissipation for the mosfets at higher speeds.

"Slight" power braking could occur by "PWM'ed" Mosfets anytime? (for example for balancing)

So - no prove or contradiction till now that i'd know of - only assumptions. But i'd be very interested to know the "truth".

On 26.1.2017 at 5:02 PM, dbfrese said:

I would still have to say that if there is not a completed electric circuit, there would be no electrical current being generated.

Yes. I agree with you.

But as written before, also with a turned off EUC the electric circuit can be completed if the generated voltage by the motor gets above the battery voltage plus two times the forward voltage of the body diode of the mosfet.

Then the mosfet body diodes start conducting and current flows (recharging the battery)!