2006 Prius High Voltage Battery Reconditioning Project

On 29 July 2024, when starting the Prius after it sat for a week in my driveway, I got the orange triangle of death on the dashboard, indicating a problem with the hybrid system. When I took the car for a short drive, the Multi-Function Display (MFD) showed that the high voltage (HV) battery was quickly varying between fully discharged and fully charged. When I arrived back home, I connected the old Windows 7 laptop running the Toyota Techstream software to the Prius OBD2 connector. Techstream indicated two Diagnostic Trouble Codes (DTC): P0A80, "Replace Hybrid Battery Pack", and P3018 “Battery Block 8 becomes weak”, meaning that the voltage is low in block #8" of the HV battery. This is a problem that can’t be ignored.

A little background information on the HV battery is warranted. The Generation II Prius (model years 2003-2009) HV battery is comprised of 128 1.2-volt nickel metal hydride (NIMH) battery cells. Six cells are packaged together inside a 7.2-volt module. Twenty-eight modules are connected in series inside the HV battery that has a nominal voltage of 201.6 volts and varies from 180 volts at the lowest allowable charge (40%) and 270 volts at the highest (80%). There are 14 blocks in the battery. The blocks are numbered sequential. The flagged battery block, #8, is located in the center of the battery.

The battery enclosure contains a battery computer that monitors a variety of battery parameters including temperature at 4 locations in the battery, and the voltage and internal resistance of each Block. The nominal voltage of a block is 14.4 volts since they are connected in series. This is the voltage at 50% capacity. For longevity of the battery the battery computer limits the charge to within 40% to 80%.

Techstream indicated that that voltage of block #8 was about 0.35 volts less than the other blocks, that we're all within 0.05 volts of each other. According to Toyota, 187601_Hybrid_Battery.pdf, the malfunction threshold for flagging a block is when its voltage varies from the other blocks by 0.3 volts. The battery modules charging discharge together as a unit. An event is flagged by the computer when one of the blocks goes out of range with its neighbors.

At this point it was clear there was a battery problem that needed to be addressed. It was not a situation that could be resolved by simply clearing the DTC codes, although I did try that. The codes would clear for about 5 minutes and then come back. The battery needed to be replaced or repaired, if possible.

One salient feature of NiMH is that the battery cells can be reconditioned through voltage cycling. The reason a module fails is due to crystal growth, referred to as dendrite growth, on the anode side of the battery. Dendrites increase the cell internal resistance within the battery and diminishes its charge capacity. Cycling of the battery between fully charged and discharged helps to break down the dendrites that are dissolved into the battery, restoring the charge capability. There are a handful of battery reconditioning companies that do just that, take an old battery, recondition all the modules and resell it with a warranty based on the quality of the reconditioned components.

A reconditioned HV battery cost ranges from about $1,000 for one that is covered by a 12-month warranty, to $1,900 or more for one with a 3-year warranty. Perhaps the batteries that come with a longer warranty are comprised of better-quality modules. There are very mixed reviews of reconditioned batteries. The key would be to find a reputable company, and the Better Business Bureau (BBB) provides information about the customer experience with some battery reconditioning companies. While there are over a million Generation II Prius models on the road, it is not known to me how many have had battery replacements and if the number of complaints reported is a significant number of those.

A new HV battery can be purchased from Toyota, minus the battery computer, for about $2,000. Toyota backs new batteries with a 12-month warranty. A new battery from Toyota was enticing if not for the price tag, and the age of the car.

There are many resources on Priuschat and on YouTube about Prius owners who reconditioned their own hybrid batteries. In all cases this involves replacement some failed/degraded modules and reconditioning the remainder. This is certainly a more budget friendly approach, although it means the car will be out of use for longer than replacing the battery directly. I decided that I was up for a DIY battery reconditioning project. I arrived at this decision after considering the options and decided this DIY project could be rewarding for several reasons. 1) I thought it would be an interesting. 2) I liked that challenge of it given the online information resources and the attraction of having to figure it out some things myself. 3) I would be able to assess the overall health of the battery during the process and would better understand where I stood at the end. 4) Installing a purchased reconditioned battery would leave me wondering wonder when it's about to fail because I'd have no knowledge of the actual condition of the components. Hence began my adventure.

Battery Removal and Assessment

The HV battery pack was removed from the car, a straightforward process that takes about 3 hours. The entire back of the car must be disassembled to get to the battery. This includes the interior side panels, and the rear seat. The HV battery has a safety switch that must be removed. This switch effectively disconnects the HV battery. It's still dangerous at this point, because all the battery modules are connected, but is safe to handle with is few precautions such as using rubber gloves and insulated tools. After examining the opened battery more closely I came to realize that it is very logically designed and can be safely disassembled without much problem.

In addition to the high voltage hybrid battery, the Prius also contains a 12-volt battery to power the computer and all the regular auxiliary components such as the lights, radio, etc. The ground wire to the 12-volt battery was disconnected.

I set up a bench in my garage and placed the 80 lb battery on top. The battery cover was removed to inspect the condition. All of the battery modules are connected together by copper bus bar plates. They all had a little corrosion on them, some more than others (see Figures 1 & 2). Corrosion isn't surprising because of the dissimilar metals used: nickel battery connectors versus copper bus bars. In this case the copper is the sacrificial metal. I speculate that Toyota chose copper bus bars to maximize current flow even though it poses a corrosion maintenance concern.

I removed all of the bus bars on one side of the battery. With this done the battery is no longer dangerous as a high voltage threat because the modules are no longer connected together.

The voltage of each individual module was measured and found to vary between 7.87 and 7.91 volts, except for module #15 which was 7.36 volts.

The first test conducted for each module was a discharge test that measured the voltage drop over 1 minute after connecting a 55-watt, 12-volt, halogen bulb to each module. The voltage drop associated with module #15 was 1.36 volts, versus about 0.17 for the remainder (Figure 3). Based on this result I decided to order one reconditioned module. I ordered the replacement module from Exclusively Hybrid. The company claims to put each module through three charge/discharge cycles, let them sit for at least 30 days to monitor for corrosion/battery leakage, and retest them before shipping. This appears to be a good approach. I called the company and spoke to an engineer who said that these are the same modules that they used to put together their replacement battery packs. The cost was $35 plus $5 shipping.

In order to recondition the batteries at home I needed a specialty charger that was able to do the charge/ discharge cycles. There are a number of chargers available in the marketplace ranging from about $40 to as high as $1,000 more. One of the complaints about many chargers is that it can take a very long time to put a single module through a cycle. The cycle time is determined by the charger amperage capability for the charge and discharge. The least expensive models could take 12 to 24 hours to do a single cycle. Given that it's necessary to do at least three cycles for each of the 28 modules, this would take a long time, possible months for some of the chargers. Based on several online recommendations, including from the YouTube channel H-EV Tech, I decided on the Tenergy T180 charger. According to the specs, this unit can charge at a rate of 100 watts and discharge at a rate of 20 watts. Two T180 Chargers were purchased at $80 each. The T180 can be used to program charge discharge cycles, up to five in a row, with cool-down time in between and the specifications of maximum and minimum voltage can be preset.

I decided to move the battery from my garage to my basement where it was cooler. The battery was placed on wire shelving to allow air flow underneath, and a fan was set up for battery cooling during the charging process. The batteries expand during charging and must be housed together in their compression pack to prevent damage (Figure 4).

A new battery module has a capacity of 6,500 milliamp hours (mAh). Module #15, the one that I suspected had failed, was put through the first 3 charge/discharge cycles. The first cycle, which represents the condition of the module when it was in the Prius, was only 572 mAh. After three Cycles it was only 749 mAh. Considering that I was shooting to restore these too at least 75% of their original capacity, about 5,000 mAh, module #15 was way off the mark.

A set of three charge/discharge Cycles was begun on the remaining 27 modules. After a few of these I concluded that it was probably best not to do them one after another, but to allow 12 or more hours between cycles so that the batteries had time to dissolve more of the dendrites within the cells after each cycle. This seemed to improve the overall recovery rate between cycles.

The time span for an individual cycle ranges from 4 to 5 hours depending on the capacity of the battery. As the battery capacity increased, so did the cycle time. If the capacity was 5,500 mAh, for example, the total time for one cycle could be up to 5 hours. Fifteen of the twenty-eight modules did not meet the desired capacity after three cycles. I decided to increase the number of cycles to six. After a while it became apparent that some of modules would not recover to the level I desired. At this point I decided to order four more modules from Exclusively Hybrid.

It was becoming clear after the initial few cycles (Figure 5) that the modules at the ends of the battery pack were in better condition than those in the center of the battery. The ones at the ends had a capacity of over 5,000 mAh after two to three cycles, while many of the interior batteries had a capacity in the 2,000 mAh range after the first cycle and for some under 3,000 mAh after three cycles. That many of the interior modules were in worst shape is supposedly due to greater heat stress within the center of the battery pack. Toyota designed a cooling system that blows cabin air over the top of the modules, and between the modules through 1/16” holes located along the sides of each module. The module sides are comprised of metal plates with nubs that help to conduct the heat away. This probably works well unless the holes get blocked. More on this later.

The replacement module arrived after 1 week. I put it through a charge/discharge cycle to assess its condition. It registered a capacity of 5,263 mAh after one cycle, lending confidence to the quality of modules provided by Exclusively Hybrid. After 7 days I completed three cycles of the pack and decided to order four more replacement modules to replace those that did not seem to be adequately recovering capacity. It was possible that the laggards could be recovered given more cycles, but replacements were ordered as a risk reduction considering that it takes a week for replacements to arrive, and the project was getting long in the tooth.

While I waited for the replacements, I continued more cycles of the modules until they reached the desired capacity. Four modules required nine-cycles, but the average was six.

The 4 additional replace modules had capacities of 5,212, 4,926, 5,076, and 5,194 mAh after one cycle. Note that Exclusively Hybrid already ran it through 3 cycles. One more cycle was applied to the weakest, bumping it up to 5,004 mAh.

With all the data now gathered, I decided to replace modules #5, #6, #12, #15 and #23.

My experience shows that NiMH module recovery is possible. A total of 14 days was required to recondition the modules. This involved getting up in the middle of the night to start a new cycle on different modules. In the end the cycling improved the module capacity considerable, often from 40% to over 80%.

There are a variety of opinions of how to rearrange cycled modules in the battery. Some suggest simply moving the outside ones to the inside in sequence. Other suggestions include matching module pairs by similar recovered capacities and arranging by internal resistance (Ohms) of the recycled modules. I took a different approach.

All of the modules were recovered to a similar capacity, but the number of cycles to achieve that capacity varied greatly. Some had a much harder life than others, i.e., required more cycles to achieve the target capacity. I sorted the modules by capacity and by the number of cycles needed to achieve the target capacity. These were then distributed in the battery with those that had the hardest life placed on the outside, and those with the easiest life placed in the center. (Figure 6). The logic for this is that those on the outside will have an easier life going forward and their capacity will probably not decrease as much over time as those in the middle that are subjected to greater heat stress. Therefore, distributing the cells in this way will better balance the age experience over time and will hopefully extend further the lifespan of the battery.

After the cycling was completed, the modules were reassembled with all of the negative end on one side and the positive ends on the other. This was done in order to charge the battery as a unit to the same voltage. The terminals on each side of the modules were connected together by a copper wire (Figure 7) and the battery unit was charged to 8 volts, which is about 60% capacity. This involved the use of one Tenergy charger set at 5 amps for 4 hours. The charger was then removed, and the battery was left to sit for 20 hours to balance the voltage of each cell. At the end the voltages of all the modules agreed to 0.01 volts.

One problem did develop with one of the Tenergy chargers. One of them would sometimes not complete a cycle. I found that by unplugging it for a few minutes it resumed normal operation. This happened several times, wasting about 8 hours effort.

Reassembly and installation took place on 17 August, 20 days after the problem was detected, and 15 days after commencement of the battery restoration. It took about 5 hours to assemble and install the battery in the vehicle, plus a few more hours to install the rear seat and side panels.

After the HV battery was installed, the negative lead of the 12-volt battery was reconnected, and then the HV safety plug was installed. I climbed onto the driver seat and pushed the start button on the dash. The dash lit up with triangle of death as well as some other new trouble lights. This was a disappointment, but I realized that the car computer might be in a strange state since the 12-volt auxiliary battery had been disconnected for several weeks. Therefore, I shut off the car and again pushed the start button. This time I heard some familiar clicking in the from the HV battery relays that happens every time the card is started. This time the dash did not display any trouble lights and after the normal 5 second or so system check, the gas engine started. I was off and running. I was anxious to take it for a ride but first I hooked up the Techstream computer to check the condition of the battery. No DTC codes were detected.

I then took the car for a drive to run the battery through its paces. This included about 7 miles at slower speed (< 40 mph) and 8 miles at highway speeds (65 mph). Everything seemed to be working well. When I got home, I used Techstream to display the individual batter block voltages and internal resistance. Figure 8 shows the battery readings measured about 10 minutes apart. While there are small variations in the reported parameters, the battery modules voltages were now moving in unison.

I put the car out on two more local trips throughout the day, logging 39 miles total for the day. I took a longer drive a few days later, traveling by back roads. After putting about 100 miles on the car I had a good sense of the condition of the battery. For the current stretch the Prius averaged over 60 mpg. I haven't seen that level of performance in years. It is exhilarating to be up and running after such an endeavor, and to once again be at the wheel of my most favorite car.

Final notes…

I’ve pondered what might have cause the battery failure almost 4 weeks earlier. It appeared that leaving the car to sit in the hot sun for a week was the straw that broke the camel’s back. However, the car was not garaged much in 18 years and has been exposed to hot and cold temperatures for its entire life. One thing I noticed when I first took the battery cover off was that there was a small layer of dog hair covering the battery cooling holes. Figure 9 shows these holes after the dog hair was removed. It’s possible that the dog hair contamination contributed to the battery’s demise. Given the small size of the holes, it doesn’t take much debris to compromise the air flow. Figure 10 shows the battery cooling system air flow. Cabin air is taken in from a vent on the top of the passenger side rear seat. The vent is connected to a fan that forces air over the top of the battery, through the battery cooling holes, and is exhausted at the back of the car. There is a flapper vent on the side of the car by the 12-volt battery to allow the air to exit to the outside.

Going forward, periodic cleaning of the battery would make sense. However, it’s a big effort to remove all the items that are in the way of unbolting the battery cover. One change that I implement was to cover the seat vent with a screen to reduce the ingestion of dog hair (Figure 11). I also decided to buckle my dog on the other side of the rear seat, away from the high voltage battery cooling intake vent. Time will tell, but the car is off to a new start.