GreenHybrid - Hybrid Cars

GreenHybrid - Hybrid Cars (/forums/)
-   GM Hybrid Trucks, Cadillac Escalade Hybrid, Chevrolet Tahoe Hybrid & GMC Yukon Hybrid (https://www.greenhybrid.com/forums/f69/)
-   -   Analysis of an 11 year old, 165K mile battery pack (https://www.greenhybrid.com/forums/f69/analysis-11-year-old-165k-mile-battery-pack-31796/)

Jaime 01-02-2019 09:56 PM

Analysis of an 11 year old, 165K mile battery pack
 
2 Attachment(s)
I swapped 40 low-mileage modules into our 2008 Tahoe battery pack and now my brother says it drives like a completely different car. I bought Two 2016 Prius C battery packs with 20 modules each. They were removed from salvaged cars with 16K and 20K miles from Sacramento and San Diego, CA.

Here are some of the problems experienced with the original battery pack:
  • Sluggish acceleration from a stop
  • Had to be light on the gas pedal from a stop to prevent the Internal Combustion Engine from stalling during the Auto-stop to ICE transition. Previously suspected the Aux Transmission Fluid Pump but it hasn't happened since putting in the “new” battery.
  • Unable to go very far in electric mode
  • Unable to go more than about 10-15mph in electric mode
Some of the codes that appeared with the original battery:
  • P0AC4 - Powertrain Control Module Requested MIL Illumination
  • P0BBD - Battery Pack Variation High
  • P0C32 - Battery Cooling System Performance
The original battery pack has a build date of November 8, 2007 (based on the serial number) making it more than 11 years old. The original modules inside had a build date of October 31, 2007. They were taken out at 165,000 miles.
Here are the results of the tests I did on the original modules:
https://cimg3.ibsrv.net/gimg/www.gre...9b4cd704b7.png

The first discharge column is the capacity that was discharged from each module to 6V at 6 amps just as the modules came out of the Tahoe. Note that due to using 18AWG 3 foot leads, there was about a 0.3V drop so when the charger was reading 6V, the modules were actually at 6.3V. This results in slightly lower capacity numbers, but in the grand scheme of things, it does not make much of a difference. Thirty-nine of the modules are total junk, with only one module at half the original capacity of 6500 mAh.
The second discharge column is the actual capacity of the module after doing a full charge. This is the number I attempted to improve with reconditioning of some of the modules (as seen in the column on the right). The 3 modules with the highest capacity actually gave me lower overall discharge capacity after doing the reconditioning. The lower capacity modules I did reconditioning on did improve slightly, but they are still junk.

Here are all the capacities in a chart:
https://cimg4.ibsrv.net/gimg/www.gre...fbe97ffdad.png
The best modules are the ones closest to the fresh air intake. Heat = Bad.

Here is the chart of the internal resistances. I'm not too sure I would trust the accuracy of the internal resistance measurements taken by the Reaktor charger.

https://cimg6.ibsrv.net/gimg/www.gre...f23806363e.png

Here's a chart of the module voltages after they were taken out of the Tahoe (minus one I forgot to measure before beginning the discharge process):https://cimg5.ibsrv.net/gimg/www.gre...d54a7c3a58.png

I found it interesting that the modules with the highest and lowest voltages were all junk. The highest capacity modules close to Module #1 all had voltages in the mid-range. Not quite sure how to interpret this, or if it has any significance. Take a look at my post from last year. You can see Blocks 1 and 2 were the two blocks where voltage didn't swing as much under discharge/charge. This is due to them having the highest capacity modules.

I'd like to thank S. Keith for the guidance he provided during the compilation of this data and the installation of the "new" modules into the Tahoe battery pack.

S Keith 01-03-2019 06:18 AM

Re: Analysis of an 11 year old, 165K mile battery pack
 
Nicely done!

Hands down the worst battery I've ever seen built with Toyota prismatic modules. I've often stated that GM does a poor job of battery management, and the above data support my claim.

The capacity error associated with a cut off at 6.3 is pretty small. On modules with far more usable capacity, it's around 300mAh at 20A. It should be notably less on these modules due to their absurdly low capacity. You can always confirm by discharging a module to 5.7V from full.

You are correct. The IR values are untrustworthy.due to your test leads. For a module with healthy resistance with short 12awg leads with ring terminals and a solid bullet connection (they rarely are), it's about 15mΩ.

Lower capacity modules are more likely to be at an extreme particularly when their state of charge is masked by the other module in the block. It's not uncommon for this to occur in batteries that are horribly deteriorated and/or out of balance.

Heat is the biggest killer of all Gen3 batteries. It seems everybody screwed up the cooling system as they made the packs more compact and left the temperature management the same as the Gen2.

I would encourage you to repeat the data gathering process in-car, so we can get an idea of what a known good battery looks like.

Hillbilly Hybrid should devise a means of controlling the battery fan... :)

SonomaGTLS6 01-03-2019 09:18 AM

Re: Analysis of an 11 year old, 165K mile battery pack
 
This is great data. Did you have any issues getting a balanced voltage across all of the new battery packs, especially coming from two separate Prius packs? Did you charge them up individually to the same voltage before installing into the Tahoe battery pack? Any diagnostics trip as soon as you rebuilt the pack? Inevitably rebuilding our powerpacks like this will become the way of the shadetree mechanic as our vehicles age (or at least for me ;)).

As for controlling the fan, if we could get HPTuners to support PCM or Hybrid controller programming I'm sure the thresholds could be manipulated just like the A/C and e-fans. Someone started a thread on their forum asking for support but it doesn't seem to have gone anywhere (https://forum.hptuners.com/showthrea...brid-support&p). Maybe just go oldschool with a thermostat controlled relay or PWM circuit that's independent of the current fan's control. I am reluctant to try anything at this point as my vehicle is still under warranty.

S Keith 01-03-2019 09:49 AM

Re: Analysis of an 11 year old, 165K mile battery pack
 

Originally Posted by SonomaGTLS6 (Post 267054)
This is great data. Did you have any issues getting a balanced voltage across all of the new battery packs, especially coming from two separate Prius packs? Did you charge them up individually to the same voltage before installing into the Tahoe battery pack? Any diagnostics trip as soon as you rebuilt the pack? Inevitably rebuilding our powerpacks like this will become the way of the shadetree mechanic as our vehicles age (or at least for me ;)).



As for controlling the fan, if we could get HPTuners to support PCM or Hybrid controller programming I'm sure the thresholds could be manipulated just like the A/C and e-fans. Someone started a thread on their forum asking for support but it doesn't seem to have gone anywhere (https://forum.hptuners.com/showthread.php?58429-2008-Chevrolet-Tahoe-Hybrid-support&p). Maybe just go oldschool with a thermostat controlled relay or PWM circuit that's independent of the current fan's control. I am reluctant to try anything at this point as my vehicle is still under warranty.


I'm going to jump in here since I have a lot more experience doing what Jaime did, and I provided him some guidance in this area.

Before I get into the details, the bottom line is that a low mileage pack as-removed from a vehicle is frequently always in the 50-60% SoC range and can be installed without concern. What most don't know is that a HEALTHY battery actually has a tendency to self-balance within about a 5% range.

"balanced voltage" is meaningless, and too many people fall into this trap. I feel rage and pity when I see someone paralleling a bunch of modules together...

The only thing that matters is "Balanced State of Charge (SoC)."

Here's a test I personally conducted:
2 modules of 6000mAh capacity
Both discharged to 6.0V @ 20A.
ONE charged to 4000mAh input @ 20A.
Modules attached in parallel with buss bars at both ends for 24 hours.
after 24 hours:
Charged module, "A", had about 2600mAh in it.
Emptied module, "B", had about 1000mAh in it.

A is at 2600/6000 = 43% SoC (started at 4000/6000 = 67% SoC)
B is at 1000/6000 = 17% SoC (started at 0/6000 = 0% SoC)

When disconnected, they sat for an hour before testing, and their voltages were within 0.01V of each other.

However, when subjected to a load, their voltages were dramatically different and well outside what is desirable, and one had less than 1/2 the amount of charge as the other.

Voltage stabilizes almost immediately due to NiMH voltage funkiness. Current flows according to the voltage difference and the resistance. These batteries have very low resistance, so even a small voltage difference will flow a decent current, e.g., 0.05V difference would flow about 4.2A... problem is the voltage difference narrows to almost nothing. Over the 24 hours, 1000mAh was transferred at an average current of 42mA, which yields an average voltage difference of 0.0005V. What really happens is a lot of capacity transfer very early on and then nothing but a trickle thereafter.

A key thing to understand is that my example is extreme... a normally charged module paired with a completely empty one. Imagine how different it would be if they were only 650-1300mAh (10-20% SoC) difference. There would be almost no capacity transfer.

Additionally, a module that has been sitting for 6 months with a resting voltage of 7.55-7.60V MAY have more retained charge than a module that was partially charged (to about 60% SoC) 7 days ago with a resting voltage of 7.75-7.85V.

NiMH resting voltage exists only to confound you. It is minimally useful.

Now... WTF is the answer to your question?

If you procure two comparably aged packs (within 10K miles of each other, and they've both been sitting for a period within about 3 months of each other), if all modules WITHIN A PACK are within a 0.03V range (will likely be 0.01V or 0V) AND if total variation between packs is less than 0.10V, just build the dang thing.

SonomaGTLS6 01-03-2019 11:31 AM

Re: Analysis of an 11 year old, 165K mile battery pack
 
Good explanation S Keith, thank you. Helps me understand the fundamentals a bit more.

Doing a little more reading on the Prius Gen 2 vs Gen 3 individual packs, I see a main difference is the internal resistance of the Gen 3 pack is lower. Would you expect this to render any performance advantages in the Tahoe application?

To the OP - Jaime - can you share any comments on the improvement in distance traveled in autostop mode? How is the acceleration to 25-30 MPH?

S Keith 01-03-2019 11:54 AM

Re: Analysis of an 11 year old, 165K mile battery pack
 

Originally Posted by SonomaGTLS6 (Post 267058)
Good explanation S Keith, thank you. Helps me understand the fundamentals a bit more.

Doing a little more reading on the Prius Gen 2 vs Gen 3 individual packs, I see a main difference is the internal resistance of the Gen 3 pack is lower. Would you expect this to render any performance advantages in the Tahoe application?

To the OP - Jaime - can you share any comments on the improvement in distance traveled in autostop mode? How is the acceleration to 25-30 MPH?



I have personally measured the IR of thousands of modules... While only a few hundred have been Gen3, there is no data to support that the internal resistance is in any way meaningfully lower than Gen2. There was a massive difference betweeen Gen1 and Gen2, and there may indeed be some measurable difference using laboratory grade instruments, but I have personally seen 10+ year old Gen2 modules measure the same as brand new Gen3 modules.

From a battery builder's perspective, I would take a 2004 Gen2 pack with 200K on it OVER any 2010-2014 Gen3 pack with 100K+ on it. Period. Hands down. Not even a question. I wouldn't even take the 2010-2014 for 1/2 the price.

Of principal concern is the age and mileage. Gen2 is getting old in the tooth given that they are 10-16 years old and high mileage. If you find a low mileage one, that's not good either as they've likely sat a lot, and you may find it won't hold a charge well.

Your best bet will be to find the lowest mileage packs ( < 60K ) FROM A MILD ENVIRONMENT (upper 1/3rd of the country IMHO) that your budget will accommodate.

EDIT: Concerning your request of Jaime, while it may be relevant to the perception of performance, efforts to extend EV operation only serve to damage the battery and shorten its life. It also frequently causes an reduction in overall fuel economy because the battery must be recharged using the gas engine. Simple conservative driving is the best approach. To give you perspective, the usable ENERGY in the hybrid battery is approximately the SAME as the total energy stored in the 12V battery under the hood. I don't think you'd expect much range or economy operating it solely off the 12V. :)

Jaime 01-03-2019 07:05 PM

Re: Analysis of an 11 year old, 165K mile battery pack
 

Originally Posted by SonomaGTLS6 (Post 267058)
To the OP - Jaime - can you share any comments on the improvement in distance traveled in autostop mode? How is the acceleration to 25-30 MPH?

I didn't take any exact measurements when I did this with the new battery, but I'd estimate the distance to be about a quarter mile and at speeds over 20mph. The old battery would never have achieved those speeds and it would have turned on the gas engine after a few hundred feet. The acceleration is night and day. Previously we couldn't just step on the accelerator and go fast from a stop, since it would have stalled most of the time when the gas engine attempted to start. Had to slowly accelerate and give the engine time to start properly and then step on the gas more to accelerate.


Hillbilly_Hybrid 01-04-2019 12:09 PM

Re: Analysis of an 11 year old, 165K mile battery pack
 
"As for controlling the fan, if we could get HPTuners to support PCM or Hybrid controller programming" The service scan tool can control that fan.

My advice is re-wire it to full blast all the time.

S Keith 01-04-2019 12:15 PM

Re: Analysis of an 11 year old, 165K mile battery pack
 
I assume it's a 4 wire PWM controlled fan? +12V, ground, PWM and tach?

Could you ground the PWM and force full blast? That would be my approach on a Honda Civic Hybrid (1 and 2).

If that won't work, can you provide guidance?

SonomaGTLS6 01-04-2019 04:12 PM

Re: Analysis of an 11 year old, 165K mile battery pack
 
Would there be any concern with the battery not getting up to temp properly in winter?


All times are GMT -7. The time now is 01:20 PM.


© 2019 MH Sub I, LLC dba Internet Brands