Taking the Toys to Task

September 13, 2007

Toyota has been on a bit of a PR binge lately, in which they have been taking some shots at GM and the hype surrounding the E-Flex/Volt development. Specifically, one of their hacks in Japan put together an internal presentation where it was claimed that the parallel plug-in hybrid approach was inherently superior to the series configuration. After that presentation was leaked, it caused a bit of a stir in the blogosphere, causing one of Toy’s North American execs the pen the following article as a clarification.

Irv reiterates the Toyota standpoint with the following assertions:

-Lithium ion technology is nowhere near ready for automotive use yet
-GM’s claims of 40 miles on one charge are totally unrealistic
-The series hybrid wastes energy by hauling around a heavy engine that doesn’t directly power the car.
-The parallel plug-in Prius is a super design because it uses a much lighter battery pack and can use the ICE for propulsion, not just electricity.

Now, I work in the sciences, so I always appreciate intelligent skepticism. But Mr.Miller made some sweeping and rather uneducated comments.

I’ve pulled out some legitimate efficiency numbers drawn from real life – not theory – that go directly against Irv’s incredulity over range and efficiency.


There is so much FUD (fear, uncertainty, doubt) propaganda in your post I don’t even know where best to start. But I’ll give it a shot:

1) FACT: The li-ion batteries you describe as “theoretical” are already in mass production from several companies, and fit the bill for what domestic carmakers need to make a Prius-killer. I.e. they have an energy density more than twice that of NiMH (>110Wh/kg), are far safer than prototype li-ion cells that Toyota appears to have abandoned, and are steadily decreasing from an already reasonable price. So there’s your first point completely out the window.

2) The efficiency of the series hybrid genset vs. a parallel hybrid engine. The Prius’ atkinson cycle engine may be efficient, but it is still heavier (1.5L, 4 cylinders) and less efficient than many of the gensets in both past and future series hybrids (e.g. 1 liter turbocharged 3 cylinders in the upcoming volt, and dating way back to the 60-100 mpg microturbines that were used in EV1-based series hybrids.) Even with the advanced CVT and throttle management, the prius’ engine is still working under a non-optimized, variable load. The genset’s constant speed, constant load configuration shortcuts all this. There’s very little energy lost in charging the li-ion battery with the genset, btw – it’s on the order of 98-99.9% efficient, versus 66-88% in NiMH.

3) You get out of a PHEV what you put in – especially when it comes to electrons.

If you’re going to build a PHEV, the entire premise is that the car will be more electrically dominant than a standard hybrid. That means adding a larger battery to provide more EV-only range as well as more EV assistance during hybrid operation. However, a smaller battery pack will be subject to more charge/discharge cycles on a more frequent basis, and thus a shorter lifespan than a larger pack. If Toyota’s engineers have any sense they’ll know that deep-cycling (draining the pack dry) will be extremely bad for the pack’s lifespan. Thus, I suspect that for their prototypes they will have the internal combustion engine kick in *much* earlier than a series hybrid would, thus preventing deep cycling. However, even during normal hybrid driving the smaller battery would still be subject to a *much* higher workload than the standard prius, and the more you use the ICE to compensate for this, the more you defeat the entire purpose of the PHEV. Moreover, if the electric drivetrain is much more utilized, and the battery is only recharged by regen or plugin – and not by any onboard generation- it’s bound to run very low – if not dry – on a long enough trip. And then you’re just left with a very expensive standard Prius with sub-standard Prius abilities.

The bottom line is that Toyota is not immune from having to pursue larger, safer, more energy dense batteries for its PHEV’s, and it needs to stop fooling itself with regards to this issue.

Posted by: AES | September 11, 2007 at 08:26 PM

A few other things before some Toy hack responds:

1) “several firms are working hard on perfecting them for automotive use. But for now, the 40 miles between charges that the Volt’s engineers talk about, and that have so many people fired up, are purely theoretical”

Here are the hard straight numbers from real life that show how Irv’s wrong:

GM’s engineers want the battery pack to have 16kWh and weigh less than 400 pounds (~182 kg). In other words, MINIMUM specific energy of 88Wh/kg. However, the density of modern high conductivity lithium iron phosphate that’s manufactured by firms like Valence and A123 is 110-130 (theoretical maximum is ~140). And that isn’t vaporware technology – I’ve got a cordless drill with A123 cells sitting in my tool chest as we speak. Using those cells would give a pack weight of 123-145kg ( 270- 320 pounds). So weight-wise, the technology is all there. What’s more, the requisite safety features are all there too. Here’s a widely-distributed test video that shows the LiFePO4 chemistry compared to cobalt-based cells (which Toyota tried and then abandoned for use in the next generation Prius):

2) The next critical question is: Can a car really travel 40 miles on that 16kWh of electric power? Better yet, can it do 40 miles on 8kWh, so as not to deep cycle the battery and shorten the lifespan?

Well, to do 40 miles on 8kWh, you’d have to have an efficiency of ~5 miles/kWh (I’ll call it mpk for short). So is 5mpk an impossible efficiency for a full size electric vehicle? Not at all.

If you look at the efficiency data for past electric vehicles (full size cars, not tiny NEVs), 5mpk is actually pretty reasonable. The gen1 EV1 had massively heavy lead-acid batteries, and achieved between 4 and 5.34mpk. The gen2 EV1 could get up to 5.68mpk with NiMH. The EV-1 derived tZero gets 5.88mpk.

This isn’t a total GM show, nor one limited to 2-seat roadsters- Toyota’s own big heavy Rav4 EV gets 4.25, and AC propulsion’s recent Scion XB-based eBox can get up to 5.14mpk.

So ~5mpk is by no means far-fetched for a lightweight hatchback.

3) Someone is bound to ask “well won’t the added weight of “hauling” around that battery AND the engine AND the fuel lower the Volt’s miles/kWh efficiency below that of pure EV’s that only have to carry the battery?”

Well, the batteries in all those past pure EV’s were a LOT heavier and had less energy in them than lithium ion – as in 2 to 3 times heavier. So by comparison to the past, a modern series PHEV, even with the small 3 cylinder engine, fuel tank, electric motor, and battery pack, still has a VERY competitive weight compared to the old EV’s in terms of miles/kWH.

Here’s some approximate math to show that:
Weight of battery pack: 270-320 pounds
Weight of 1.0L Ecotec engine: 182 pounds
Weight of AC motor: 70-100 pounds
Weight of 12 gallons gasoline: 78 pounds
Sum= 630-680 pounds (286-309kg) for the vast majority of the drivetrain. Tack on a little more for the diff, and transmission, but not much.

Now compare that gross weight to the weight of just the Rav 4 EV’s battery pack of 550kg or 1,210 pounds! And then consider that it’s an SUV! Granted it had more range than the Volt is planned for, but the point is that despite that massive weight, it STILL had a miles/kWh range of 4.25 (combined cycle, according to US Dept. of Energy).

So all in all, thinking that 5miles/kWh and a 40 mile range are “theoretical” is a rather ignorant opinion.

The informed public is NOT convinced, Irv. I await any response you may have.

PS – charging time of 6 hours too long for you? That’s how long most adults sleep in a given night. And with that thought in mind, I bid you goodnight.

Posted by: AES | September 12, 2007 at 02:22 AM

PPS before I forget:

According to various sources, the prototype Toy Prius plug-in apparently has a 2.6kWh secondary battery (in addition to the main NiMH cell?) that delivers 7-8 miles per charge, max. If you assume that’s deep discharge, that’s 3miles/kWh. If you assume that’s at 70% discharge (more likely), that’s ~4.39 mpk.

After that 7-8 miles I guess the car reverts to regular Prius mode, and due to the extra 58 kg of battery weight (127 pounds – calculated using standard Prius pack specific energy of 45 Wh/kg) the car gets less than its real-world milage of 48 mpg.

Also, assuming that the secondary battery is in addition to the Prius’ standard 39.52kg battery, and is of the same cell type, the gross weight of all the batteries would be 97.5 kg, or ~215 pounds. Considering that the Volt’s pack is only just up the street at ~270 pounds, the plug-in Prius doesn’t have much ground for claiming a significant weight advantage. You might want to include a footnote on Okamoto’s diagram saying “not to scale”.

Good job Toy.

Posted by: : AES | September 12, 2007 at 03:21 AM”



  1. Some more data that provides a certain measure of support for the arguments I presented:

    According to Google.org’s RechargeIt, the PHEV Prius conversions made using A123 lithium ion technology (the same chemistry that’s been contracted for use in the Volt) yields 8.44 miles/kWh in electric-only mode, implying a 21 mile range on 50% DOD of its 5kWh battery pack. These are the conversions that Toy strongly disapproves of, by the way. As well as the technology that it denies even exists.

    So just another piece of real world data confirming the efficiency of plug-ins with larger battery packs. This was, admittedly, using a parallel hybrid as a starting point, but the all electric modes for parallels and hybrids is virtually identical. If anything, the larger motor on a series hybrid would deliver superior efficiency to the Prius’.

    “Hybrid” mode is a bit more murky of a concept. Google.org lists 68.4mpg for its conversions, but its unclear if this is true miles per gallon, or miles per gallon EQUIVALENT. More likely the latter.


    A Toy rep has since gone on record stating that they want the plug-in Prius to have an all-electric range of 30-60 km (18.6- 37.28 miles). In other words, roughly the same performance goal as GM’s E-flex system.


    Now this begs two obvious questions:

    1) How can you even think about doing this without using lithium ion technology? GM already has dibs on the A123 technology, so Toy must be looking at other possible suppliers to make this claim. Panasonic has obviously failed to deliver safe or powerful Li-Ion thus far.

    Regardless, the shift towards larger all-electric range, and thus a larger battery, pretty much invalidates Okamoto’s diagram showing smaller battery size as an advantage to Toyota’s approach. Seems like a case of “say one thing, do another”.

    2) If the Prius plug-in is going to be capable of highway speeds running on electricity alone, isn’t it going to need a more powerful electric motor that won’t need ICE assistance at higher speeds and loads, which negates the entire premise of the HSD system?

  3. Toyota’s battery supplier is Panasonic, who are the ones who have had problems with thermal events in Li-Ion batteries. So, it’s no surprise that they are touting the parallel hybrid design that they’ve been forced to invest so much in. They also want people to buy their parallel hybrids today, rather than wait for the series hybrids to arrive. No doubt the exciting promise of future electric vehicles have put a damper on sales of existing hybrids.

  4. Last comment before I get off my soapbox –

    A plug-in HSD (prius) system would actually be more of a “compound” hybrid than a true parallel hybrid, given that the ICE can both recharge the battery and drive the wheels. This is good versatility, but the motor’s efficiency will suffer under the load of performing both tasks. It will have to be made larger to handle the peak load, and its efficiency will suffer as a result.

    Meanwhile, a series hybrid with battery buffering could be built to an exact size, weight, and power, and take maximum advantage of valve timing, throttle management, and forced induction. Even with the losses of transmitting that electrical power through the generator, batteries, and inverter, it could still be engineered to be as efficient (or more) than a larger ICE running through a complicated and wasteful CVT.

  5. Hi Futuredrive,

    The information you have provided in support of your argument is too good. After reading this, I am better informed and more positive about GM’s e-flex technology.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: