Prius Plug-in and Volt Pros and Cons

Sorry Mike, there are quite a few of us that owned both (a regular Gen3 Prius and a PiP) and can tell you that the PiP will easily do 2-3 MPG better on the interstate than the Gen3. Please don't listen to anything CR says. There are threads and threads on this site about how much they routinely screw up hybrid news.

 

Honestly though, I prefer to not make decisions based on anecdotal evidence.

 

First, the "not true" had nothing to do with my post about blending using plug-supplied electricity.

Second, it's the CR data that's anecdotal. Having owners driving for years in a Gen-3 then switch to a PiP driving the same routes the same way is far more informative.

 

This is what CR says about how they test mpg

"Road tests. Consumer Reports’ fuel-economy tests are conducted on our track and on public roads. Testers splice a precise fuel meter into each test car’s fuel line to measure how much gas is consumed. Each car is then run through highway and city drive loops, with each performed multiple times by two drivers.

The city test is conducted on a loop that’s set up on our track to reflect driving in a suburban area. It’s marked so that a driver must maintain specific speeds in certain sections and stop the car at specific points for set idling times. Highway mpg is measured by driving on a particular stretch of sparsely used freeway near our test track at a steady pace of 65 mph. Each driver runs the test in both directions to compensate for wind and the slight difference in grade.

Our raw results are corrected for temperature using a formula established by the Society of Automotive Engineers. But we don’t test if it’s too hot, too cold, too wet, or too windy. Our overall mpg is a weighted composite of city and highway mpg measurements."

If they indeed tested the PIP this way, this is not anecdotal.

 

How about arbitrary or approximation? That's the same problem with EPA.

Their attempt to simulate real-world is never as good as actual owner driving.

 

CR test cycle is very aggressive. EPA highway for PiP is 1 MPG more than a regular Prius. Drive it easy and PiP may shine even more. That big lithium pack helps a lot with regen capture.

So, it depends on how you drive.

 

I just stumbled over an Idaho National Laboratory document with the following test results for a 2013 Volt:

Fuel Economy at Steady-State Speed, 0% Grade

15 mph gets 39.8 mpg
30 mph gets 63.5 mpg
45 mph gets 72.8 mpg
60 mph gets 44.8 mpg
75 mph gets 38.1 mpg

 

FWIW the ANL data does seem fairly consistent with wider observations from what I've seen. In the "Top 20 MPGe" PiP thread there are folks who are able to acheive 170s and 180s EV MPGe in what I believe are DC based calculations. I've not seen numbers this high out a Volt, the best numbers I've seen were in the 130s to maybe 140s EV MPGe DC.

Here is the PiP thread: Top 20 MPGe | PriusChat

And here is one thread where Volt owners are sharing their real world mi/kWh or kWh per 100 miles data: Calculating Miles / kWh [Archive] - GM-Volt: Chevy Volt Forum

The best number quoted from the Volt thread is 23kWh per 100 miles. This is the onstar reported number, which I believe is AC based. That would equate to about 145MPGe, which is quite good. Its important to note though that the owner is good enough to provide the caveat that he has a long commute and selectively uses the EV in high efficiency parts of his commute. A more reasonable number probably comes from the post by "PC's own" F8L who quotes 26kWh per 100 miles, with no indication of special use during certain parts of the commute. That would be 128MPGe AC which is still quite good relative to typical. I believe some of the numbers reported on the thread are DC from the in car mi/kWh reporting, but typical seems to be in the 3-4 mi per kWh or 30-40 kWh per 100 mi or 84-111 MPGe.

Another way to look at it is the much more commonly tracked EV range. ANL puts the typical used DC capacity of the batteries at about 2.3kWh for the PiP and 10.9kWh for the Volt. At 200 DC Wh/mi (20kWh/100mi or 167MPGe) the PiP would have an EV range of 11.5mi, and the Volt would have a range of 54.5 miles. From what I have seen, 11.5 miles is a pretty pedestrian range for the PiP, while 54.5 miles is an achievable but pretty exceptional range for the Volt. This seems like another indication that the PiP would have to be considerably more efficient under electric drive to match typical observations.

Rob

 

The top MPGe list entries are supposed to be AC based numbers. For my top 10 Volt list entry (133.5 MPGe EV averaged over 1019 miles), I mostly got 130-135 MPGe EV but on one day I drove 100% EV for my 96 mile round trip to work (mostly on the highway at 55 mph) and got 141 MPGe. That includes a 500 foot elevation drop from my home in the hills of San Francisco and back so it wasn't an entirely flat road.

An INL report says the 2013 Volt gets 187 MPGe AC (180 Wh per mile) at a steady 30 mph and 201 MPGe AC (160.8 Wh per mile) at a steady 20 mph. I know someone on the GM forum recently put his Volt on the minimum 24 mph cruise control and went 67 miles which is roughly consistent with that.

The DC usage for the PiP seems a bit low and the DC usage for the Volt is perhaps slightly high. I've seen national lab reports showing the PiP taking 2.9 kWh AC at 72F at 240v. If 2.3 kWh DC were true that would be a 26% charging overhead which seems too high.

The 2013 Volt INL results appear to show around 10.85 kWh DC and 12.6 kWh AC which is a charging overhead of 16%. The equivalent numbers for the 2013 PiP would appear to be around 2.375 and 2.95.

I don't really understand the DC numbers for the PiP. In any case, if you ignore them and look at the AC charging records showing about 2.95 kWh and divide that by 11.5 miles you get about 0.257 kWh per mile. If you then divide the 2013 Volt's 12.6 kWh AC charge by that you get a Volt range of 49 miles which isn't that uncommon if you remember that 2013 Volts drive farther than earlier Volts (38 vs 35 EPA or 8.5% farther on a charge). I drove my usual 45.5 miles to work this morning and had an estimated 5 miles of range remaining. If I drove a 2013 Volt I would have had 8-9 miles of remaining range.

I think part of what's going on is that PiP owners reporting the really high MPGe results are driving short distances at 20-30 mph. Most Volt drivers (and PiP drivers!) go farther and drive faster.

I'm perfectly willing to believe that the PiP is somewhat more efficient at EV, especially when you include regenerative braking at city driving conditions (because it's 600 pounds lighter). I agree that it appears to be easier to achieve 160-180 MPGe while slow driving a PiP than a Volt in real world conditions. I'm less sure that the PiP is significantly more efficient in EV driving at highway speeds or in a combined driving real world commute.

As I mentioned earlier, my own limited experience got me the same 130 MPGe EV commuting for a week in both cars over the same route although the PiP got 30% better gasoline hybrid mileage. I got several rounds of 10-11 miles of range along with a 14.3 once driving the PiP.

In any case, the Volt can be pretty efficient at EV and do it a much farther distance or make it easier to obtain a high EV% for drivers with longer trip segments between charging.

 

As you are no doubt aware, wall to wheels or tank/battery to wheels efficiency is a cascaded calculation of many losses. This is every bit as much true in electrical drive systems as it is in traditional ICE drive systems. Some of the sources of losses are the same (weight, CdA, rolling/rotational resistances, mechanical transmission efficiency, etc) and some are unique (battery efficiency, charging efficiency, electrical transmission efficiency, inverter efficiency, motor efficiency, etc). Most of these losses are complex, state dependent, and often inter-related resulting in an extremely complex series of design trade-offs between efficiency, cost and performance over a wide range of conditions.

I would say the spread comes primarily from Toyota's significant head start in the work of optimizing these design trades. I don't think there is anything particularly unique to Toyota about this, they simply have a lot more accumulated engineering hours and development dollars put into hybrid and electric drive system development than GM or Nissan, largely by virtue of their 10-15 year head start.

Rob

 

Bill, I understand your skepticism and appreciate your curiosity.

The real answer is no one knows exactly how they do it. This is the most carefully guarded trade secret of the Toyota HSD. Otherwise, anyone could do it. We have some insights because so many people have been trying to figure it out, including our own federal research laboratories in order to try and help our domestic automakers catch up.

IMHO none of the factors you are tempted to dismiss are likely trivial.

Transmission: No one knows for sure, but the HSD has been estimated to be at least 95% mechanically efficient. I believe this is the worst case path for the ICE, the MG2 path is more direct and presumably even more efficient. Contrast that with ~85% for the Honda IMA CVT and perhaps 75-85% for traditional transmissions/transaxles.

Inverter: As I understand it, Toyota not only designed the inverter internally, they designed custom ICs for the driver IGBT because the ones available on the market were not good enough. ORNL found the Gen 3 Prius inverter to be 98-99% efficient at medium to high loads at medium to high speeds. My understanding is typical values are 85-95% in commercial / industrial inverters.

Motor: Also custom designed by Toyota. My recollection was that one of the early focuses of the DOE investigations was into how Toyota had figured out how to mass produce such a small, light, powerful, efficient motor so inexpensively. ORNL data shows 95-96% efficiency at medium to medium high loads / rpms. My understanding is typical is 80-95% in commercial / industrial AC PM motors.

Boost converter: The prius inverter includes a switching boost converter that can dynamically increase the nominal ~240V battery voltage up to 650V DC. This gives the Prius controller another degree of freedom in optimizing efficiency for varying load/speed conditions. Inverter/motor efficiency can fall off significantly (10-15% perhaps) when operated substantially away from its optimized peak efficiency load / rpm condition. By varying the output voltage, peak efficiency can be maintained over a much wider range of operating conditions.

Electrical Transmission and Net Efficiency: Yes copper is copper, but there is a lot to effective electrical coupling. For a given gauge of interconnect wire a higher voltage will result in lower losses for the same power delivery. This will have the most impact at high load conditions where the Prius is using the highest voltage. A more efficient system overall will result in less current and consequently less loss for the same amount of mechanical work overall. Additionally, there are many factors that can influence the effective coupling of the switched AC power out of the inverter and into the motor including impedance matching, timing accuracy, ripple / crosstalk control, etc. ORNL found the total combined efficiency of the Gen 3 Prius motor and inverter to be 93-95% for medium to medium high loads / rpms, which would seem to me to indicate extremely effective AC power coupling.

Combined mechanical/electrical efficiency: Using the numbers above the Prius would be estimated to run about 88-90% overall efficiency under medium to high loads / rpms, possibly better if MG2 path mechanical efficiency is better than 95%. I have found one obscure reference to the existence of a "Volt Technical Book" that puts the Volt electric motor combined efficiency at 88% peak, and 80% or better under medium to high loads / rpms. However there is no clear indication of what efficiencies this encompasses, no publically available source, and no independant verification that I can find so far.

Drag: Chevy Volt looks to have a CdA of 6.51 compared to PiP of 5.84. This accounts for ~4-8% greater power required to maintain speed depending on speed.

Tires: EVCalc does not estimate any significant difference in tires between 255/55/17 and 195/65/15 assuming both have same RR.

There are more factors to discuss, but I'm running out of steam

Some links:
useful for load estimates: EV Calculator
Gen 2 PSD simulator: Toyota Prius - Power Split Device
Gen 3 ORNL motor/inverter efficiency report: http://info.ornl.gov/sites/publications/files/Pub26762.pdf
Volt efficiency "reference" (saghost): Question for physics nerds [Archive] - GM-Volt: Chevy Volt Forum
Possible Volt efficiency source (S-EL06-54.01WBT): http://acdelcotechconnect.com/pdf/2012_ACDelco_Catalog.pdf
CdA reference: Vehicle Coefficient of Drag List - EcoModder
Typical EV AC drive motor 88% efficiency: Curtis 1239-8501 HPEVS AC-51 Brushless AC Motor Kit - 144 Volt, EV West - Electric Vehicle Parts, Components, EVSE Charging Stations, Electric Car Conversion Kits

 

Here is another source that seems to more or less back up the Argonne test data for DC efficiency, though gives a slight charging efficiency edge to the Volt.

INL has published the following fact sheets as a result of their baseline testing.
http://avt.inel.gov/pdf/phev/fact2013toyotapriusphev.pdf
http://avt.inel.gov/pdf/EREV/fact2013chevroletvolt.pdf

Steady State Speed DC Wh/mi (PiP vs Volt)
10mph: NA / 219.5
15mph: 102.2 / NA
20mph: NA / 160.8
30mph: 120.6 / 180.0 Volt 49.3% higher
40mph: NA / 207.8
45mph: 180.1 / 226.0 Volt 25.5% higher
60mph: 204.5 / 294.2 Volt 43.9% higher
70mph: NA / 361.6
80mph: NA / 411.4

Possible interpretation: It looks like the Volt has a funky operating mode at low speed that compromises efficiency? At 30mph the PiP is probably in a sweet spot of electric and mechanical efficiency (hasn't started spinning MG1/ICE yet?). At 45mph PiP has lost ground, possibly due to spinning MG1/ICE? At 60mph PiP is likely benefiting from boosted drive Voltage while Volt efficiency is falling off?

At 45mph they give us a direct AC compare of 217 Wh/mi PiP vs. 248 Wh/mi Volt, closing the gap from 25.5% DC to 14.3% AC. Hard to say how they would compare at other speeds, as the net AC/DC efficiency appears to vary (mostly due to battery efficiency presumably).

CD UDDS (urban cycle) DC Wh/mi vs Temp (PiP vs. Volt)
20F: 144.1Wh/mi+46.1mpg / 270.3Wh/mi + 46.4mpg = 38.5mpge / 33.8mpge (Volt 14% higher)
72F: 201.6 / 258.5 (Volt 28.2% higher)
95F: 281.9 / 337.4 (Volt 19.6% higher)

CD HWFET (highway cycle) 72F DC Wh/mi (PiP vs. Volt)
213.5 / 261.9 (Volt 22.7% higher)

CD US06 (aggressive driving) (PiP vs.Volt)
97.4Wh/mi + 57.0mpg / 364.8 Wh/mi = 48.9 mpge / 92.4 mpge (PiP 89% higher)

All Electric Range (mi) for consecutive cycles (PiP vs. Volt):
UDDS: 15.7 / 50.7
HWFE: 20.1 / 47.8

The INL data also gives a good series of charge sustain mode comparisons.

CS Mode MPG for various cycles (PiP vs. Volt):
SS15mph: 103.8 / 39.8 (Volt 161% higher)
SS30mph: 92.9 / 63.5 (Volt 46.3% higher)
SS45mph: 75.3 / 72.8 (Volt 3.4% higher)
SS60mph: 63.7 / 44.8 (Volt 42.2% higher)
SS75mph: 51.5 / 38.1 (Volt 35.2% higher)

IMHO this is _very_ interesting data. It certainly seems from this data that the Volt transmission is optimized for about 45mph, but much lossier elsewhere? Interestingly the electric drive data seemed to indicate a similar peak in efficiency at 45mph.

CS Mode MPG for various cycles (PiP vs. Volt) Continued:
UDDS CS 20F: 35.7 / 27.8 (Volt 28.4% higher)
UDDS CS 72F: 57.2 / 40.7 (Volt 40.5% higher)
UDDS CS 95F: 44.9 / 29.2 (Volt 53.8% higher)
UDDS 20F: 56.2 / 40.7 (Volt 38.1% higher)
UDDS 72F: 69.6 / 46.5 (Volt (49.7% higher)
UDDS 95F: 49.2 / 31.7 (Volt 55.2% higher)
HWFET 20F: 64.4 / 49.2 (Volt 30.9% higher)
HWFET 72F: 70.6 / 49.2 (Volt 43.5% higher)
HWFET 95F: 62.6 / 41.5 (Volt 50.8% higher)
US06 72F: 48.1 / 35.3 (Volt 36.3% higher)
SC03 95F: 41.8 / 24.0 (Volt 74.2% higher)

Rob

 

I certainly don't understand the Volt well enough to really know what its doing, or why these curves would look this way. The PiP ones make sense to me, but maybe just because that's what I'm used to.

Rob

 

The CS Volt curve shows series hybrid below approximately 37 mph that apparently gets more efficient when generating at higher RPM and load and then at speeds above 37 mph it curves more like the PiP (but at lower efficiency) when the Volt switches to power-split mode.

The EV Volt curve looks roughly similar to the PiP except at speeds below 20 mph but this is mostly an artifact of the inconsistent test speeds in the dataset for the 2 cars. The PiP is very likely less efficient at 10 mph than 20 mph but there just wasn't any data collected for that in the case of the PiP. This efficiency drop under 15 mph is also true of the Tesla Roadster and Model S.

Here's the graph from Tesla for the Roadster and Model S although it is in Wh per mile rather than MPGe so the Y axis goes from more efficient to less and the resulting data curve is flipped upside down from your national labs graphs.

 

INL also has data for C-Max Energi and some for Leaf, added to graphs.
http://avt.inel.gov/pdf/fsev/fact2011nissanleaf.pdf
http://avt.inel.gov/pdf/phev/fact2013fordc-maxenergi.pdf

 

Here's a timely and relevant article....

This guy drove his 2012 Volt with 52,000 EV miles under his radial belts (smaller battery like my 2011, and similar EV miles) around the perimeter road of a sports stadium and got 81.8 miles on a single charge on 9.8 kWh DC.

He doesn't say what the AC charge was but it's probably about 11.5-12 kWh. He drove about 20-24 mph and pumped up the tires a bit excessively. End result was around 230-245 MPGe or somewhere around 143 Wh per mile (AC).

How Far Can A Volt Go In EV Mode? How About 81.8 Miles

 

miscrms,

Thanks for putting together these very informative posts. The quality of engineering that went into the PiP is really amazing. Hopefully they will keep the lead in the next generation too.

 

Really cool. Relevant efficiency check would have the Fusion Energi and PiP around the same belt and the same driving style. My guess is that the PiP could possibly get 30 while the Energi maybe 40-50