User avatar
TickTock
Posts: 1701
Joined: Sat Jun 04, 2011 10:30 pm
Delivery Date: 31 May 2011
Leaf Number: 3626
Location: Queen Creek, Arizona
Contact: Website

Re: Regen question

Sat Sep 03, 2011 7:25 pm

tps wrote:Yep, the motor continues to spin the same direction. The direction of current flow, however, reverses. Current always flows from higher voltage to lower voltage. When charging a battery, the charger has to put out a higher voltage than the battery's voltage to get current to flow into the battery. The inverter is where this magic happens when the car is moving. You can think of it almost as having a continuously variable voltage conversion ratio between motor and battery. To provide power to the motor, it must adjust the ratio so that to motor will draw the desired amount of current from the battery. To regenerate it adjusts the ratio so the battery will draw current from the motor (acting as a generator). This is not really so much different from down-shifting the transmission in a car to get a bit of engine braking effect when going down a hill, except that an ICE engine just wastes all the braking it's doing as more heat exhausted through the radiator, whereas the inverter puts the energy (minus the conversion and charging loss which is exhausted as heat from the LEAF') back into the battery.


This is an excellent description of how a brushed DC motor controller works. The Leaf, however, uses a brushless motor which is far better for three main reasons 1) no brushes to wear out and 2) the coils are on the *outside* housing (stator) instead of on the inner rotating part (rotor) so you can get the heat out much easier and 3) you get regen for free. Brushless motors use permanent magnets on the rotor. The controller monitors the position of the rotor and simply activates the coils on the outside in sequence. When coasting, it will align the coils exactly with the rotor field. When accellerating it just adjusts the timing of the coil activation to lead the rotor (drawing power from the battery and pulling it along). For braking it changes the timing to lag the rotor causing the rotor to induce voltage in the coil - sending charge back to the battery. There's extra detail to deal with back EMF and limit regen but it's actually a quite elegant solution. 80% efficiency is very doable with good fast switching, low resistance switches - in fact almost required. If you are less efficient then that you have to figure out how to get that heat *out* of the controller.

I have a lot of respect for the engineering in the controller. Some buddies and I designed a smaller (18hp) brushless motor controller for a battlebot a few years ago and we literally blew up one prototype (and others just caught on fire) before we worked out all the kinks - there's a lot of energy moving around in that thing!

User avatar
planet4ever
Posts: 4674
Joined: Sat Apr 03, 2010 9:53 pm
Delivery Date: 02 May 2011
Leaf Number: 1537
Location: Morgan Hill, CA, south of San Jose

Re: Regen question

Sun Sep 04, 2011 12:11 am

TonyWilliams wrote:The 17% (48 of 281 raw data) is when LBW comes on, regardless of Bar 1. The chart is not clear as to when Bar 1 will disappear, since the range planning would be based on LBW from that moment on.

Ah! Thanks for the clarification, Tony. I can now see that is what you were saying, but I was just too dense to catch the subtle point. I agree, the bars no longer matter once you get the Low Battery Warning.

So, to rephrase my point just slightly, Nissan's definition of a QC in "about" half an hour is for 80% - 17% = 63% of the usable battery capacity [to which the lawyers would want to add: with a new battery, standard temperature , standard voltage, maybe some other factors].

Ray
End of April 2013: Traded my 2011 SL for a 2013 S with charge pkg.

User avatar
Nekota
Posts: 616
Joined: Sat Dec 11, 2010 11:16 pm
Delivery Date: 12 May 2011
Leaf Number: 2182
Location: Sunnyvale, CA

Re: Regen question

Sun Sep 04, 2011 12:50 am

planet4ever wrote:
Nekota wrote:The 80% efficiency is the product of the motor acting as a generator (about 90%) and the inverter box capturing the regen AC power back into DC to charge the battery (another 90% operation resulting in 0.9 x 0.9 = 0.81).

Ah, but you forgot half the story. You then have to convert the stored energy back into mechanical energy, perhaps at 81% efficiency again. 0.81 * 0.81 = ~0.66. I personally doubt the round-trip efficiency is much higher than 70%, though I could buy 85% in each direction, which would get it up to 72%.

Ray


Well the question was the efficiency of regeneration back into the battery which is a one way trip. Of course the system loses energy when going from the battery to the traction motor and your description is what I would expect on slowly climbing a hill with the traction motor and then recharging the battery on descent which would result in your ~70% 'cycle' efficiency. The KW loss going from the battery to the wheels is described by Nissan to have an efficiency range of 87% to 95%
Image


Which leads to the question of what the 80KW and 30KW values represent? Are they the energy at the input of the motor or inverter or battery?
Res 20Apr2010 / Ordered 27Sep2010 / Delivered 12May2011

PV System

Volusiano
Posts: 1461
Joined: Thu Feb 24, 2011 2:41 pm
Delivery Date: 03 Jun 2011
Location: Phoenix, AZ

Re: Regen question

Sun Sep 04, 2011 2:26 am

TickTock wrote:
tps wrote:Yep, the motor continues to spin the same direction. The direction of current flow, however, reverses. Current always flows from higher voltage to lower voltage. When charging a battery, the charger has to put out a higher voltage than the battery's voltage to get current to flow into the battery. The inverter is where this magic happens when the car is moving. You can think of it almost as having a continuously variable voltage conversion ratio between motor and battery. To provide power to the motor, it must adjust the ratio so that to motor will draw the desired amount of current from the battery. To regenerate it adjusts the ratio so the battery will draw current from the motor (acting as a generator). This is not really so much different from down-shifting the transmission in a car to get a bit of engine braking effect when going down a hill, except that an ICE engine just wastes all the braking it's doing as more heat exhausted through the radiator, whereas the inverter puts the energy (minus the conversion and charging loss which is exhausted as heat from the LEAF') back into the battery.


This is an excellent description of how a brushed DC motor controller works. The Leaf, however, uses a brushless motor which is far better for three main reasons 1) no brushes to wear out and 2) the coils are on the *outside* housing (stator) instead of on the inner rotating part (rotor) so you can get the heat out much easier and 3) you get regen for free. Brushless motors use permanent magnets on the rotor. The controller monitors the position of the rotor and simply activates the coils on the outside in sequence. When coasting, it will align the coils exactly with the rotor field. When accellerating it just adjusts the timing of the coil activation to lead the rotor (drawing power from the battery and pulling it along). For braking it changes the timing to lag the rotor causing the rotor to induce voltage in the coil - sending charge back to the battery. There's extra detail to deal with back EMF and limit regen but it's actually a quite elegant solution. 80% efficiency is very doable with good fast switching, low resistance switches - in fact almost required. If you are less efficient then that you have to figure out how to get that heat *out* of the controller.

I have a lot of respect for the engineering in the controller. Some buddies and I designed a smaller (18hp) brushless motor controller for a battlebot a few years ago and we literally blew up one prototype (and others just caught on fire) before we worked out all the kinks - there's a lot of energy moving around in that thing!
Thank you for this enlightening insight into how it really works. It makes a lot of sense now.

User avatar
abasile
Posts: 1922
Joined: Thu Sep 02, 2010 10:49 am
Delivery Date: 20 Apr 2011
Location: Arrowbear Lake, CA

Re: Regen question

Sun Sep 04, 2011 12:03 pm

After searching this site using Google, I found the DC Quick Charge plot and associated post, dated March 5, 2011, that I was basing some of my assumptions on. From viewtopic.php?f=9&t=2919#p66586:

DarkStar wrote:Sounds like you're asking about "DC Quick Charging" since there is no such thing as Level 3 (L3) charging yet...

From what I've been told, here is a typical DC Quick Charge profile:

Image

This is a Nissan LEAF with over a 50% SOC. It starts out at 48 KWh and 5 minutes later starts decreasing.

Hopefully once I get my LEAF, I'll be able to get a charging profile graph from our DC Quick Charge station here in Portland and we'll find out how quick it goes for sure. My understanding is that from 0% SOC it's fairly linear to 80% SOC and that takes just about 26 minutes. This jives given a rate of 48 KWh. The last 20% is what you see decreasing there on the graph, taking about 20 minutes. So in theory, the first 80% is charge at a rate of 2.46 miles per minute of charge, then it decreases to about an average of 1.25 miles per charge per minute.


Now, I am going to question whether the subject vehicle actually had over 50% SOC at the start of Quick Charging; perhaps DarkStar was making a faulty assumption. If this is "a typical DC Quick Charge profile", then perhaps it actually reflects charging from roughly 25% SOC up to 80%. If that is the case, then at 80% SOC the maximum continuous charge rate would be about 12 kW. That could explain why I am finding continuous regenerative braking to be so limited as the SOC approaches 80%. In the latter portion of each of our 5000 foot descents, perhaps we actually are hitting the battery as hard as the tail end of a "typical" QC.
2011 LEAF at 71K miles, pre-owned 2012 Tesla S 85 at 98K miles
LEAF battery: 9/12 bars and < 49 Ah (-28% vs. new)
Tesla battery: 250+ miles of range (-5% vs. new)

KeithFrechette
Posts: 71
Joined: Mon Jun 27, 2011 9:45 am
Delivery Date: 16 Sep 2011
Location: Duvall, WA
Contact: Website

Re: Regen question

Sun Sep 04, 2011 7:45 pm

Did the SOC value for the "low battery" warning increase with the newer firmware revisions (e.g. to provide extra buffer)? If so, then maybe the 0-80% quick charge in 1/2 hour (as documented in the original LEAF manual) actually gets you a bit more charge than you would get if the manual reflected the current firmware revision.
Keith Frechette [Microsoft], http://keithfrechette.wordpress.com/
-----
Reserved: May 10th, 2011
Ordered: May 10th, 2011
Received: September 16th, 2011
Configuration: Super Black 2011 SL + QC port
, VIN: 80xx

User avatar
planet4ever
Posts: 4674
Joined: Sat Apr 03, 2010 9:53 pm
Delivery Date: 02 May 2011
Leaf Number: 1537
Location: Morgan Hill, CA, south of San Jose

Re: Regen question

Sun Sep 04, 2011 10:33 pm

The SOC levels for the twelve bars were definitely changed, but I think the general consensus is that SOC for Low Battery and Very Low Battery were not moved.

Ray
End of April 2013: Traded my 2011 SL for a 2013 S with charge pkg.

tbleakne
Gold Member
Posts: 983
Joined: Wed Jul 28, 2010 12:05 pm
Delivery Date: 03 Jun 2011
Leaf Number: 2400
Location: Claremont, CA

Re: Regen question

Wed Sep 07, 2011 5:54 pm

abasile wrote:you have to subtract aerodynamic and friction losses before arriving at a figure for the amount of kinetic energy available to be captured via regen. If you do this, then it is not unreasonable to expect 80% efficiency. Coming down 5000' from my home in the San Bernardino Mountains, I think I am getting close to this, but it is hard to get a precise number without fine-grained SOC information.

Since I completed my SOC meter last week, I have done one quick regen test up and down a steep street here in Claremont. For those who know Claremont, it was the North end of Mountain Ave. Google Earth shows an altitude rise of 508' over distance of 4175' = 12% grade. I was in ECO mode with no A/C.

I set the SOC meter to display what I call "clicks" (the integer units the SOC meter converts to % by doing a division), since these are the finest-grain measurement units available. Going up I expended 12 clicks (what I call the units the SOC meter converts to %) driving up, and 5 clicks of regen returned coming back down. I was never going over 30 mph, so aerodynamic losses were low. <<Edited - added new data below>>

To measure the loss due to motion, I did a separate test on near-level ground at similar slow speed, on a circular route so that gravity effects would average-out. After two complete trips around the course, I had lost 7 clicks over 2.2 miles;
2.2/7 = .31 miles/click motion loss
BTW I use a tire pressure of 39 lbs/in^2.

Applying this result to the regen test course,
(4175'*2/5280 = 1.58mi)/(.31miles/click) = 5.1 clicks loss due to motion (rolling + aero).
Energy available for regen recovery: 12 clicks - 5 = 7 clicks.
Regen efficiency = (5 clicks/7 clicks) = ~73% +- ~7%.

I was hoping for higher efficiency, but this course was a little too short for an accurate calculation.

<<End Edit>>>
Because I had to apply light brake pressure most of the time on this steep grade, I probably lost some energy to friction braking. If I can find a steady grade not quite so steep, with a higher safe speed, where I don't need to apply any brake pressure, I can hopefully get a higher regen efficiency reading.

Whenever I have my foot on the brake, even lightly, I worry that some friction braking may be engaged. I am not sure how to “feel” whether there is friction braking, at least at low brake pressures. At higher pressures I am sure friction is engaged.

Assuming one has plenty of spare battery capacity, so that all double circles show, I have found two tricks for maximizing my regen. One is to keep the speed up, at least 20 mph. The other is to apply a brief strong braking pulse. This will often help kick the regen into 4 or 5 circles. Of course this also briefly engages friction braking, but often most of the higher regen level will remain after I have removed most or all brake pressure. Have others observed this ?
Last edited by tbleakne on Fri Sep 09, 2011 5:49 pm, edited 1 time in total.
LEAF Ocean Blue SL, "100 % Electric" decals, Delivered June 3, 2011
Sold June 2014 27K miles, 18% capacity loss, 1 bar, 5.0 mi/kWh.
Solar 4.6 KW DC with both string and micro-inverters.

tbleakne
Gold Member
Posts: 983
Joined: Wed Jul 28, 2010 12:05 pm
Delivery Date: 03 Jun 2011
Leaf Number: 2400
Location: Claremont, CA

Re: Regen question

Wed Sep 07, 2011 5:57 pm

Nekota wrote: The KW loss going from the battery to the wheels is described by Nissan to have an efficiency range of 87% to 95%
Image


Which leads to the question of what the 80KW and 30KW values represent? Are they the energy at the input of the motor or inverter or battery?

I have also wondered about this.

This is a very nice graph. Do you have a link to the report from which it was extracted ?
LEAF Ocean Blue SL, "100 % Electric" decals, Delivered June 3, 2011
Sold June 2014 27K miles, 18% capacity loss, 1 bar, 5.0 mi/kWh.
Solar 4.6 KW DC with both string and micro-inverters.

User avatar
abasile
Posts: 1922
Joined: Thu Sep 02, 2010 10:49 am
Delivery Date: 20 Apr 2011
Location: Arrowbear Lake, CA

Re: Regen question

Wed Sep 07, 2011 8:10 pm

tbleakne wrote:If I can find a steady grade not quite so steep, with a higher safe speed, where I don't need to apply any brake pressure, I can hopefully get a higher regen efficiency reading.

The road that we use to reach our mountain town (CA-330) seems pretty good for that purpose. When there is not much traffic and the SOC is sufficiently low, it is possible to use ECO mode regen to keep one's speed down with little to no use of the brake pedal. You'll have to come visit! :D
2011 LEAF at 71K miles, pre-owned 2012 Tesla S 85 at 98K miles
LEAF battery: 9/12 bars and < 49 Ah (-28% vs. new)
Tesla battery: 250+ miles of range (-5% vs. new)

Return to “Range / Efficiency / Carwings”