Hypothetical Battery Upgrade

Hi Dave, you should be realizing by now that the Prius is pretty complex and highly instrumented. If it were just a big battery, slugging in another would be a piece of cake and easy. But the battery has its own ECU that monitors it and communicates with the hybrid system ECU about its state of charge. The hybrid ECU knows the Prius has one battery. If you put in another and nothing else, it might get very confused if one battery said I'm hungry and the other battery said I'm stuffed. Part of the value of a PHEV kit is that the manufacturer has figured out how to fool the hybrid ECU.

While every thing can be improved, Toyota spent a lot of research time coming up with the balance of efficiency, power, safety, and cost that they sell as the Prius.

I agree with you about PriusChat and most of its members.

 

From Bill Merchant: " If you put in another and nothing else, it might get very confused if one battery said I'm hungry and the other battery said I'm stuffed."

That really wouldn't be the problem. 2 large battery banks of equal capacity age etc, wired in parallel would act as one large battery. I agree that the problem would be with the ecu however. For example if the ecu directs the battery to discharge only until it has put out X number of amp/hours, and to stop charging only when it has put x number back in, the additional bank might not do anything except add dead weight. If you could reprogram the ecu to realize that the bank was twice as large, it could therefore double the amp/hours allowed for discharge etc.

Icarus

 

I've had occasion to deal with a dozen, NHW11 battery modules as part of my battery refurbishment experiments. IMHO, if the electrolyte is not part of one pool, it takes only a few cycles before one battery is weaker than the other and the problem appears to get worse. Over time the individual NiMH properties appear to diverge and this may put a natural limit on how big NiMH cells can be.

I have not done much work with lead-acid batteries and since they have a larger dV/dQ (where Q is their charge capacity,) they may work better in parallel. But the nearly flat charge/discharge curves of these NiMH batteries is maddening and makes simple circuit connections a thing of horror and wonder.

I don't want to put anyone off from their own experiments but I'm trying to suggest go cautiously with a lot of 'backup'. Experimentation is good but these NiMH batteries are something else. My admiration for the Toyota engineers grows higher every day I deal with these modules.

You might consider reading up some of the RC hobbyist work. I've found they have a lot of insights about operating NiMH and other battery chemistries ... just before their models crash.

Bob Wilson

ps. If you have an interest in NiMH module experimentation, send me a PM and we can work out a deal. I have plenty from a replaced traction battery and if someone has a proposal for some experiment lacking only modules, let me know. It is better to have more than one expert in this area.

 

I know less than nothing about NiMh battery chemesty, life etc.

From the solar world, I know enough about flooded lead acid batteries to be dangerous. The cardinal rule about lead acid is that all the batteries in the bank need to be the same size as well as the same age. If you put a bigger battery in the bank, you will effectively give the bank the capacity of the smallest in the bank for example. If you put a new battery into an older bank, the new one will rapidly become the effective age of the older ones.

It seems that the Prius battery is nothing but a series/parallel array of individual cells. As such, in my uninformed mind, if all the cells are the same age and size, then the only difference would be that the total amp/hour capacity of the total would be bigger.

Icarus

PS. I too am not advocating hacking the car for novices. People who want to play, push the envelope, go for it!

 

Its actually been done. The short answer is its pretty complicated and it doesn't improve mileage by any appreciable amount.

The common perception is that more battery = better efficiency. After all, thats what makes a hybrid better than a regular car. The reality is that the people who get the best efficiency out of their Prius do it by driving their car using as little battery power as possible. It seems that if you can accelerate the car with the ICE at its peak thermodynamic efficiency while transferring no power into or out of the battery, and then glide with minimal drag and no power going into or out of the battery you achieve the best possible mileage under normal circumstances. The short explanation for this is that while the hybrid system does greatly improve the efficiency of the ICE when its driven in what would normally be very inefficient modes, the inherent conversion losses of mechanical to electrical to chemical energy and back can never do better than the peak efficiency of the ICE since all of the energy must eventually come from the ICE. So for the everyday driver, the hybrid system makes the Prius more efficient than a regular car. For the hypermiler, using the hybrid system as little as possible is key to optimal results under most conditions. Even for the everyday driver the size of the current battery is more than enough to capture and release all the energy encountered under normal driving circumstances. As an example, the DOE ran a Gen1 Prius to 160k miles and then ran tests on it. They found that the battery capacity was down 50%, but they had seen no corresponding drop in mileage.

Extra batttery capacity doesn't really come in to play until you add a means to charge the battery from some other source than the ICE. This is the idea behind the PHEV conversions. Now you are offsetting gasoline burned by the ICE, with electricity from the grid and the bigger your battery the more you can offset.

If you haven't seen it already, you can find out more about plugin hybrids here:
Prius PHEV - EAA-PHEV

Rob

 

As Rob says,,,,BINGO.

You might be able to make the argument that hypermilers would benefit from removing the battery and the associated weight?

Icarus

 

NimH are a very different beast unfortunately. The biggest problem is that their voltage is not a good indicator of SOC. As they get full, they will start to heat up, their internal pressure rises, and their voltage will flatten out or even drop. You have to either monitor dv/dt and look for the flattening or peak, or monitor temperature and look for dT/dt (preferably both). You pretty much have to do this on a single cell or small string of cells basis, as cell to cell variation tends to be significant and trying to look at too many at once will mask the effect you're looking for. Another problem here is that due to these issues NimH cells do not like to be paralleled.

The Toyota pack is I believe made up of a single string of 6.5Ah, 7.2V prismatic packs similar to these:
Plastic Case Prismatic Module | Panasonic EV Energy Co., Ltd.

Even in a single string the charging and discharging is complicated enough that you need a BMS (battery management system) to babysit the batteries and keep them in their safe operating ranges. Since the battery voltage is not a good indicator of SOC, the ECU employs a coulomb counting scheme. Basically, what ever charge you take out of the battery it puts back. This is one of the big challenges to the PHEVer, is that the controller doesn't care how big the battery actually is. It knows how big the battery is supposed to be and only uses that amount of energy. Cal-cars, and others that use parallel packs get around this by taking the pack voltage higher than normal to trigger a recalibration event referred to as SOC-Drift. In this way you can get the car to use the extra battery, but you introduce a lot of additional challenges by using a higher voltage secondary battery. This is why most of the commercial conversions pitch the prius battery all together and install their own primary battery and a new BMS/controller. Thats way beyond the capabilities of any DIY conversion yet though.

This is a big part of the reason that most PHEV work being done has either focused on sealed PbA or Li-ion. Li-ion has its own issues, but large capacity cells and appropriate BMSes are fairly available. One exception is the Nilar NimH, which have a built in pressure sensor to indicate the onset of charge saturation. They're decent capacity, and the charge sensor is pretty easy to monitor. I think you still have to charge them in parallel strings and monitor every pack but its at least feasible for the DIYer.
Plug-In Conversions and Nilar Team Up

Rob

 

Hi Icarus,

This is not correct. The Prius traction battery is a series array of individual cells:

• Six 1.2V cells per module
• 28 modules per traction battery
• 201.6V nominal voltage rating (6 x 1.2 x 28)

If you try to connect battery modules in parallel, you'll end up with one module hogging most of the charging current due to small internal resistance differences.

 

Patrick

Got it!

T

 

MG2 is used to generate power, that power is then sent to MG1. MG1 becomes the load resistance as it attempts to spin the ICE (with fuel and spark cut) against it's internal resistance. IF this isn't enough stopping power, then braking is supplemented with the brake pads as needed.

 

Long climbs and downhills could be a special case where more battery could be useful. You burn a lot of extra gas getting to the top, and ideally you'd want to capture as much of that energy back as possible when you come back down. In my experience anything more than 4-5 minutes of significant downhill grade will usually fill the battery, at which point you are no longer able to recapture energy. If thats a situation you face regularly, you could get some benefit. IMHO there are two ways you could consider doing this.

1. Add a second Prius pack, as Steve did on the page below. The results were not impressive in general, but again if you regularly go up and down very long grades you could be in a unique situation. Steve has now moved on to a full Li-ion plugin conversion in the second link.
DISCOUNTED INSURANCE SALVAGE
Prius Aux Battery preliminary

2. Do a cal-cars open source "PriusPlus" lead acid conversion as described below. This is about the cheapest plugin conversion you can do, at least from an upfront cost perspective. The Nilar NimH version may be cheaper in the long run due to much better battery life. One nice thing about the simplicity of the cal-cars conversion is it allows regen energy to be stored back in the secondary pack. When the secondary pack is depleted it stays attached to the primary battery which has the benefits of increasing battery efficiency by lowering effective internal resistance, and increasing capacity to store regen energy. This is not true of most of the more sophisticated conversions that use big dc:dc converters. Its probably questionable whether dragging the added weight of the lead acid up the mountain would burn more energy than you could recover coming back down. You would also need to plan your battery usage such that the secondary pack is empty before you get to the top so that you will be in this paralleled "charge sustain" mode coming down. Theoretically I can imagine that if you were using the secondary pack for hybrid assist on the way up, giving you lots of battery capacity ready to be filled up on the way down that could certainly be a good thing. Just not sure if the extra battery capacity would be a help, or mostly just the extra energy stored from the grid as with a traditional phev.
PriusPlus - EAA-PHEV

Rob

 

I to was thinking about the possibility of adding a second battery in parralell as the route to work is quite hilly in sections and the battery drops right down and I have to slug up the hills using the ice. The fact that the ecu knows the battery size seems to kill the idea completly.
Nick


"Since the battery voltage is not a good indicator of SOC, the ECU employs a coulomb counting scheme. Basically, what ever charge you take out of the battery it puts back. This is one of the big challenges to the PHEVer, is that the controller doesn't care how big the battery actually is. It knows how big the battery is supposed to be and only uses that amount of energy."