1. Attachments are working again! Check out this thread for more details and to report any other bugs.

Thermal performance of gen 2 Prius HV battery

Discussion in 'Gen 2 Prius Technical Discussion' started by 2009Prius, Jul 18, 2012.

  1. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    510
    63
    Location:
    USA
    Vehicle:
    2009 Prius
    There have been several discussions on the cooling efficiency of the HV battery pack in various threads and I have posted data here and there. I hope to collect all of my data concerning the thermal performance of the gen 2 Prius HV battery pack in this thread.

    It's in the middle of the summer now and people are concerned about the cooling of the HV battery. Some argues that turning on the A/C helps since the cooled air in the passenger cabin is drawn through the back seat vent to cool the battery. I did an experiment turning on the A/C in one particularly hot afternoon and compared with another hot afternoon with no A/C. This graph shows the outside air temperature (OAT) and the min and max battery temperatures (blC and buC) for each of the two trips:
    [​IMG]
    All temperatures are plotted at the same scale and offset so they can be directly compared and read off the two vertical axis in degree C or F. As seen in the graph, under my driving conditions the HV battery T only increased by 2 ~ 3 degree C over ~ 50 minutes of driving, regardless the A/C was on or off. I did set the A/C set point conservatively, initially at 75 F, then gradually moved up to 84 F, then settled down to 80 F.

    One big effect of the A/C is the loss in fuel economy. MPG dropped from high 80s to mid 70s. This can be attributed to the added load from the A/C running, as shown in these graphs:
    [​IMG]
    The data shows that in the beginning of the trip the idle current draw was about 1.7 A without A/C, and turning on the A/C moved the current draw up to 8 ~ 9 A (lower two graphs). Further into the trip the idle current draw was lowered to about 1.2 A without A/C, and in the A/C on case it didn't have to work as hard as in the beginning, cycling the current between 1.7 and 5 A (upper two graphs).

    Here is another example, with the same A/C on data and a newer A/C off data:
    [​IMG]
    The outside air T tracked very well between the A/C on and A/C off cases, so did the HV battery Ts. Again the battery T increased only 3 degree C over 50 minutes of driving, even with the A/C off. (The MPG comparison may not be entirely fair in this case: For the A/C off case I had run some errands 3 hours before. So the engine may be warmer than usual. Also the rare event of no S1 happened to occur as well.)

    From the same set of data I also found something peculiar about the control of the cooling fan for the HV battery:
    [​IMG]
    Look at the HV battery cooling fan voltage (hvFanV) traces in the graph. It seemed to go through a few full on / full off cycles, before going into a more reasonable variable voltage control. What was the control circuit/algorithm thinking?

    Now back to the HV battery temperature change. Why doesn't it depend on A/C use and why is the change so small? I had a hypothesis that the thermal mass of the battery is huge compared to the available surface area for cooling, resulting in the small difference in temperature rise between A/C on and off cases. Now I have made a back of the envelope estimation that agrees well with my speculation: The total heat generated in the battery is calculated by integrating the instantaneous joule heating power (I ^ 2) R, where I is the HV battery current and R is the sum of the internal resistances of the 14 battery blocks. I got 193 and 165 kJ for the trips with A/C on and off, respectively (each trip ~ 50 minutes). Dividing them by the mass of the battery (53.3 kg) and a rough estimate of the heat capacity (1 kJ/kg/C), we get 3.6 and 3.1 degree C, respectively. So we predict that, if there is no cooling, the battery T will rise 3.6 C with A/C on and 3.1 C with A/C off. These values are not that far off from the measured T rise of ~ 3 degree C. So under my driving conditions the battery simply does not generate much heat during the 50 minutes of driving, and much of the heat generated by the battery does not get removed by the cooling fan, resulting in the T rise of ~ 3 degree C.

    Of course a different driving condition can heat up the HV battery a lot more, for example driving up and down lots of steep hills. In this paper
    http://www.nrel.gov/vehiclesandfuels/energystorage/pdfs/2a_2002_01_1962.pdf
    the authors studied gen 1 Prius and found the battery T rose ~ 18 degree C in 60 minutes under the US06 cycles. So as we all say: YMMV. :)

    Links to other posts:
    battery cooling fan mod | PriusChat
    HV battery fan Voltage vs. Temperature | PriusChat
     
    epa5epa, dave77, planetaire and 2 others like this.
  2. bwilson4web

    bwilson4web BMW i3 and Model 3

    Joined:
    Nov 25, 2005
    27,665
    15,664
    0
    Location:
    Huntsville AL
    Vehicle:
    2018 Tesla Model 3
    Model:
    Prime Plus


    Nice charts and analysis. Let me share something from my testing, the effect of a forced charge. Charging our NHW11 batteries is exothermic:
    Sign in to Yahoo!
    [​IMG]

    I found a 7-9C increase in battery temperature doing a forced charge. If the battery is already warm from the ambient temperature, this would be a bad thing.

    Although harder to quantify because grade, altitude, and speed are needed, I am sure that rapid driving over hilly country can and does 'heat pump' the traction battery to a higher temperature. What happens is at high speed, climbing a hill at or over 65 mph, leads to traction battery energy 'borrowing.' On the reverse slope, the battery gets a charge proportional to the climb deficit. But ohmic heating on the next hill climb keeps the temperature up.

    Bob Wilson
     
    dave77 and 2009Prius like this.
  3. SageBrush

    SageBrush Senior Member

    Joined:
    Jun 4, 2008
    11,627
    2,531
    8
    Location:
    Southwest Colorado
    Vehicle:
    2012 Prius v wagon
    Model:
    Two
    2009Prius,
    A question and a comment

    1, I wonder if the battery fan turned on.

    I am interested in battery cooling when the AC is set to recirc vs circ. Any data ?
     
  4. PriusCamper

    PriusCamper Senior Member

    Joined:
    Mar 3, 2012
    11,330
    4,614
    0
    Location:
    Pacific Northwest, USA
    Vehicle:
    2007 Prius
    Model:
    Two
    Interesting... Makes me wonder why Toyota doesn't just put an A/C vent that runs directly into the HV cooling system? I mean this whole internal air intake concept seems flawed... I mean telling people to keep the rear intake vent unobstructed? Come on! Glad the GenIII, V and C... have eliminated that air intake problem. But why not just run an A/C system direct to the Battery Pack? If well designed it seems it could offer a significant boost in gas mileage?
     
  5. bwilson4web

    bwilson4web BMW i3 and Model 3

    Joined:
    Nov 25, 2005
    27,665
    15,664
    0
    Location:
    Huntsville AL
    Vehicle:
    2018 Tesla Model 3
    Model:
    Prime Plus
    Cooling the battery just means the thermal effects are minimized. I've found in my charging/discharging experiments that if the battery is getting warm from the exothermic charging, it is also generating internal gas. If left alone for 12-24 hours, the gas recombines as the module cools. But I don't see this improving mileage but minimizing traction battery stress and aging effects, especially in the older NWH11. The other models have much improved batteries.

    Avoiding heating is the best course so:
    • avoid forced charge in warm weather
    • climb hills without borrowing energy from traction battery (55 mph works great up an 8% grade.)
    GOOD LUCK!
    Bob Wilson
     
    dave77 and NortTexSalv04Prius like this.
  6. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    510
    63
    Location:
    USA
    Vehicle:
    2009 Prius
    I haven't thought about the heat generated/absorbed by the chemical reactions. I suppose for a long drive where the starting and ending SOCs are similar the heat from the chemistry cancels out. The only force charging data I have is plotted below:
    [​IMG]
    I did it twice. First time from 65% to 77% SOC and the 2nd time up to 81.5% SOC when the car refused to charge any more. The battery T did increase 3 ~ 4 degree C during the whole process of about 7 minutes. The rise would be even higher if the starting SOC was lower. And the rate of T rise is much faster than my usual drive; as one would expect from an exothermic reaction. It's also interesting to see the SOC quickly dropped back down to 80.5%.

    The battery fan did turn on. See the last graph from the OP where the fan voltages are plotted.

    Yes I can do another experiment with A/C in re-circulation mode.
     
    bwilson4web and dave77 like this.
  7. NortTexSalv04Prius

    NortTexSalv04Prius Active Member

    Joined:
    Nov 14, 2009
    915
    115
    0
    Location:
    Texas
    Vehicle:
    2004 Prius
    Model:
    N/A
    Well summer is here in N Texas. Few trees and little shade . We are hitting the triple digit temps too. Daytime high's will easily find 103-107 range in July-Sept. Today hit 103 and my hv cooling fan has come on. Question about MFD and color codes Green(full charge cell) Blue(partial charge) Red(scary color) . What are leading signals of HV battery/cell failure.............................
     
  8. bwilson4web

    bwilson4web BMW i3 and Model 3

    Joined:
    Nov 25, 2005
    27,665
    15,664
    0
    Location:
    Huntsville AL
    Vehicle:
    2018 Tesla Model 3
    Model:
    Prime Plus
    Based upon NHW11s:
    • uneven temperatures
    • uneven voltages in the module pairs, above 0.3 V difference can throw a code
    • other battery codes
    You'll notice my temperature increase was a little larger due to:
    • the internal heat takes a minute or two to reach the surface and show up in the external temperature probe
    • my goal was to measure traction battery capacity so I did not wait for temperatures to plateau but started backing up the hill, the altitude change would be proportional to the traction battery energy capacity
    Bob Wilson
     
    NortTexSalv04Prius likes this.
  9. GasperG

    GasperG Senior Member

    Joined:
    May 13, 2011
    1,168
    598
    1
    Location:
    Slovenia
    Vehicle:
    2018 Prius
    Model:
    N/A
    Is it normal for the battery to reach 48 °C, I have tracked the values with Torque app and on a steep downhill when SOC was 79% it reached very high temperatures. Ambient temperature was 18°C in car temp wasn't more than 25°C. Here is the data:

    Welcome to Google Docs
     
    2009Prius likes this.
  10. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    510
    63
    Location:
    USA
    Vehicle:
    2009 Prius
    GasperG thanks for the interesting data. As we can see in the chart below that the battery T rose initially but then turned around and decreased:
    [​IMG]

    even though the joule heating continues to accumulate, as seen by two separate estimates, one from delta SOC and the other from battery current:
    [​IMG]

    Part of this may indeed be the result of the endothermic reaction when the battery discharged from 80 to 68% SOC.
     
  11. nh7o

    nh7o Off grid since 1980

    Joined:
    Dec 10, 2011
    1,686
    340
    0
    Location:
    Hawaii
    Vehicle:
    2009 Prius
    Model:
    II
    According to Seilerts' posts, that is about the upper limit of the acceptable range. I use B mode extensively to avoid these high SOC downhill situations. Turning up the A/C also helps. Roll down the windows for more drag.

    Interesting data.
     
  12. PriusCamper

    PriusCamper Senior Member

    Joined:
    Mar 3, 2012
    11,330
    4,614
    0
    Location:
    Pacific Northwest, USA
    Vehicle:
    2007 Prius
    Model:
    Two
    Once the battery is fully charged the car takes care of not overcharging/overheating, so you don't have to worry about that. (unless you live at the bottom of a steep hill and always turn the car off before it can fully cool the battery). Point being, once the battery is fully charged on a long downhill there is value in putting the car in neutral and getting some extra distance with as little friction on the wheels as possible... Then once it's time to brake or accelerate again don't forget to put car back in drive! :)
     
  13. tdelker

    tdelker Junior Member

    Joined:
    Feb 6, 2012
    46
    4
    0
    Location:
    Colorado
    Vehicle:
    2004 Prius
    Model:
    II
    Isn't there an issue with Gen2 in neutral above a certain speed?
     
  14. PriusCamper

    PriusCamper Senior Member

    Joined:
    Mar 3, 2012
    11,330
    4,614
    0
    Location:
    Pacific Northwest, USA
    Vehicle:
    2007 Prius
    Model:
    Two
    There's lots of threads on here about using neutral at speed... Nothing definitive though... Long ago you could run into problems with a car in neutral at speed, they even past laws against it. But those problems are no longer likely. One dude on PriusChat claims to be using his Prius as a guinea pig for doing neutral coasting as often as possible, he's yet to notice any problems... You might want to find and follow that thread to see what happens!
     
  15. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    510
    63
    Location:
    USA
    Vehicle:
    2009 Prius
    One definitively plausible method to damage the car is to shift to neutral at the beginning of a long steep descend at a speed lower than 40 MPH and let the car accelerate to a high speed like 70 or 80 MPH while remaining in neutral. The reason is that the engine is stopped when shifting to neutral at low speeds, and remain stopped when the car glides to high speeds, causing MG1 to over rev. On the other hand we haven't heard of any report of MG1 damage caused by this practice. Any volunteer living in the mountains willing to experiment with his/her car and report the result here? (As always any damage incurred will be the sole responsibility of the driver, not any other users, staff, or sponsors of PriusChat.);)
     
  16. SageBrush

    SageBrush Senior Member

    Joined:
    Jun 4, 2008
    11,627
    2,531
    8
    Location:
    Southwest Colorado
    Vehicle:
    2012 Prius v wagon
    Model:
    Two
    ^^ The same volunteer is asked to set the car rolling, and watch from the top of the hill. We want to check the steering.
     
  17. GasperG

    GasperG Senior Member

    Joined:
    May 13, 2011
    1,168
    598
    1
    Location:
    Slovenia
    Vehicle:
    2018 Prius
    Model:
    N/A
    Another data from Gen 3 Prius:
    Welcome to Google Docs

    I started on a hill pass on 1350 m (4.400 feet) car was parked for 2 hours on the sun outside temperature around 20° C (68° F). I started A/C to cool down the car, when I started driving SOC was at 34 %, I was using B mode. It took only 5 min of steep downhill driving and 211 m (656 feet) of elevation drop to fill the SOC to 80% (600 Wh in 5 min?). I took a 3 min pause in the middle (white line on the graph phone was off for 2 min), with car on and A/C running on very low setting (400 W). Notice how the temperature was still rising, I think there is 5 min delay of actual battery temperature and temperature sensor (sensor 2). After the pause the battery was filled again from 70% to 80% and temperature was going even higher to 47°C (116 °F).
    chart_1.png

    The most aggressive regen charging was 78 A at 266 V (20 kW).
    My previous data (couple posts above) was similar, with a little higher temperature but it included driving up the hill, the interesting part is that my wife didn't use B mode down that hill, and max regen charging was 72 A at 252 V (18 kW) and when the SOC was 80% regen stopped and engine would spin just like in B mode. This suggests that there is very small difference between D and B mode, of course B mode brakes more by default.

    Edit: There is difference between D and B mode, B mode will spin the engine even when the soc is bellow 80% while in D mode engine starts spinning when it reaches 80%. From B mode I would expect even more engine braking (when SOC < 80%) and not that aggressive regen.
     
    nh7o and 2009Prius like this.
  18. Grahams Number

    Grahams Number Junior Member

    Joined:
    May 5, 2021
    14
    4
    0
    Location:
    Douglasville,Ga
    Vehicle:
    2005 Prius
    Model:
    II
    I have a 2005 with 134000 miles. Battery pack replaced last year - first pack lasted 15 years!
    I learned a lot about batteries after my pack failure and every battery whacker agrees that heat and time are the real enemies of battery life/performance.
    I monitor the battery temperature sensors. I have found that middle sensor temperatures can easily reach 130 degrees F on a warm Georgia day in stop and go traffic. Any appreciable regen causes the temperature to climb. The fan, at speed 1 or 2 seems to have little affect on the temperature rise in stop and go traffic. I have seen the pack temperature at 125 F before even driving if it has been parked in the sun for a hour or two. 20 minutes of heavy regen and battery draw got me to almost 140 F before getting home.
    The fan is clean. I didn't rebuild the battery pack when it was replaced so I'm not certain all the air dams were properly installed when the Toyota tech rebuilt the pack but the fan does drop the temperature when driving at a steady speed until it reaches 90 F. Fan then stops and temps return to slow climb. Fan is more effective in the winter months of after it rains.
    I have searched and searched but no one seems to be able to quantify an acceptable operating temperature range for the battery pack. What are others finding to be the typical or normal range?

    Tom