sward — Your data are interesting. Thanks for taking the time to run these tests. They certainly support the notion that the FE is not grossly different when one forces EV mode, and perhaps there's not much in it either way. With a bit more perseverance you may be able to further refine your results (hint!).

Stan

 

What we _really_ need is some kind of fixed-route competition to get the best FE over the same distance, with no speed requirement (except a minimum time, say), and have something that can record the exact driving parameters (like the black box of a plane, say, recording the conditions in the car--speed, at what point the ICE came on or switched to EV, etc). Get some kind of serious insights into what driving methods really work best, compare similar driving styles among different drivers to find things like the impact of a slightly heavier foot or coasting more in EV mode, and so on. Ideally with 2-3 cars to reduce any potential variability between cars.

If we're going to imagine a testing ground, might as well make it fun and turn it into a TCH rally!!

Of course if you really want to be strict about everything you have to make sure each trip starts with both same charge and same weight (doesn't have to have the same amount of gas but a heavier driver presumably incurs some FE penalty if all other variables are made equal). I could add additional fantastic options like having one pass "record" the system data, and then have a way to "play back" the same driving actions to remove the inconsistency of the variables of each pass even for someone trying to drive the same, but... you can just keep taking it further and further into all of that. I mean it would be interesting to see the results, no doubt (and probably a bunch of us would have fun trying to maximize FE!).

Just a thought.

 

Well, I can rerun the tests to help smooth out the random variations (I don't mind - the driving is fun anyway ). But I'm not going to get a whole lot more precision with the conditions and instrumentation I've got (no long flat roads, no scangauge, no chance of using cruise-control safely).

One difference between my tests and yours is that I spent significant time in EV-mode both when driving for 'sustained EV' and for 'sustained speed' (because my target speed was lower). The difference between those two runs was mostly the duration of the individual EV-mode stretches, and the fact that the battery did not discharge as far on the 'sustained speed' run.

 

I realize this thread is stale and probably should be left for dead, but I will throw in my .02 here (.0192CDN). After reading through all the posts, I see only one time where Stanley has touched on what I consider to be a critically important issue. That is, an ICE is most efficient when under max load. You may be rightfully proud of the 31% efficiency you get puttering around, but my 7.0L hunk of Otto-cycle Detroit iron in my 38 year old Mustang is well over 40% efficient when touring the 1320 in .00383 hours. If we were talking Otto cycle, the EV issue would be a slam dunk as the efficiency improvements from an extra load would be significant. But a Miller/Atkinson cycle improves the low-load efficiency so efficiency improvement from additional loading is less. I have a Ford Escape Hybrid, so the following may not apply to Toyota's (clearly inferior )system:

At 35 miles per hour, with no load (full battery, no a/c, etc) I am using ~10kW to mosey on down the road. I am converting this from my petrol at a 31% rate thanks to Mr. Miller. However, if I charge my HV battery while moseying, I need 15.5kW total. But, my ICE efficiency improves to a whopping 33.6% thanks to the 55% increase in load I've placed on the little ICE. So, I have sequestered 5.5kW into the battery at an efficiency of ..... 39.6%! Wait - how can that be??? Clearly, we're violating the laws of physics here! No wait, we're not. 5.5/(15.5/.336-10/.31)=39.6% Hmmmm...... well, we still have multiple conversion losses if we want to use this energy in EV of ~ 90% X 90% =81% * 39.6% =32.1% So this is a slightly better use of our Dino juice after all! Use EV, get better mileage without violating the laws of Physics! Cool! YMMV, because in reality this all still remains a spherical cow exercise...

This is why series hybrids would be so much better. You can run a small ICE at max conversion efficiency to charge the batteries after the electrons from the plug have done their job. Downside, of course, is that an ICE always running at max load will wear out a lot sooner than one that is mostly lightly loaded. But hopefully it will only run for 10,000 km or so out of 150,00 km of total driving.

 

DesertDog — I hope this thread isn't dead! I'm still hoping that someone will put his/her money where his/her mouth is, and do a good experiment to provide supplementary data or to either refute or support my data. So far, no-one has done so. Your FEH would certainly qualify as a suitable vehicle for such experiments. I encourage you to produce some good data. Just make sure to keep as many other variables as possible constant (round-trips only, NiMH battery SoC the same at the end as at the beginning, no wind, etc.), so that the data can be directly related to the point of this thread — namely, that unless you really can outwit the built-in ECUs (which most people can't, but you are very knowledgeable, and have many supplementary instruments to tell you what's going on inside your FEH, and might be an exception), forcing as much pure-EV operation as possible will not lead to better FE.

But, to address your points, the Toyota/Ford hybrid systems are already continually using MG1, MG2 and the NiMH battery to control the rpm of the ICE so as to produce the required power at close to its greatest possible efficiency at that power. Your numerical example demonstrates how this can be so. You believe that you can do this better than Ford's programming can. It would really be nice to have some good data (round-trip, etc.) to back this up! This usually means lugging the ICE at relatively low rpm's. This could, of course, in principle, also be achieved with a purely mechanical transmission, if it has enough gears (or is a CVT with enough range). Few of them do, however. I believe that Toyota's 'Atkinson'-cycle ICE is considerably more efficient than the 31% you quote, and the mechanical/electrical/chemical (i.e., battery) conversion efficiencies are also much higher than 90% — maybe of the order of 95% based on my experiments.

A series hybrid is always subject to a double conversion loss (including the battery); a system like ours always has some losses associated mainly with the electrical path between MG1 and MG2 (even if the battery isn't involved); whereas a system like GM's 2-mode hybrid (which incorporates mechanical clutches) can, I believe, operate in a purely mechanical mode while cruising, and so avoid any additional conversion losses. That said, "heretical" mode while cruising (where electrical energy actually flows backwards from MG2 to MG1) is, I think, used to achieve greater overall efficiency by forcing the ICE into a substantially more efficient operating region (lower rpm, higher load, and less noise for the same power) in our vehicles.

Stan

 

The only way your arguments would hold water was if the ICE BSFC was independent of load, which is most definitely NOT the case. Get rid of the butterfly and camshaft, and put in that variable moment-of-inertia flywheel, and you'll get closer.

The fact that Prius drivers can get better FE using pure ICE P&G vs. ICE steady state cruise should be a sufficient clue that steady state ICE is not the most energy efficient way to get from point A to Point B.

Lastly, experimental results are pretty worthless unless you have a large sample and equally large control. The physics and math of why pushing EV in a hybrid is better for FE should be convincing enough.

 

Stan- I wish I could say I did this test just for you, because I know I can get better MPG with lots of EV mode driving.... 3 years of practice tells me so, but I also wanted to see how my car was behaving with 30% ethanol.

I filled up with 30% ethanol 100+ miles ago.
So there's been more than 100 miles of ethanol "break-in".
I went from E10 previous 3 tanks to E30 this tank.

Outside air was 28 degrees at the time of my drives.
I went 5 miles at 55 MPH to get everything warmed up.
This drive also got the HV battery to max. stasis of 53.0%.

I found a county farm road (paved) that would have very little traffic.
It was 8 miles between stop signs. Gentle rolling hills, but nothing major. Minnesota prairie and farmland.

I reset my ScanGauge trip meter.
I tried to accelerate at a steady 2400 RPM up to 35 MPH.
At 35 MPH I hit the cruise control and let the car take over.
The car had full control, and 35 MPH is low enough for the car to drop into EV if it wanted to. The cruise control varried from 34 MPH on the uphills to 40 MPH on the downhill. 8 miles down the road, I waited until 50 yards before the stop sign and applied the brakes briskly and stopped. I turned around, and tried to manually accelerate at 2400 RPM to 35 MPH. At 35 MPH I applied cruise control. I followed the same 8 miles of rolling hills in reverse. The car went into EV at the stopsign, but did not go into EV at any other point during the drive. At the end of the return trip, I wrote down the data while the car was in motion at 35 MPH, just before applying the brakes at the initial stop sign, thus only one stop was included in the test, reducing one variable.

Then I repeated the same drive without cruise control. I drove with my foot in control and tried to maintain an average 35 MPH speed. I would need to accelerate at ~2400 RPM to about 39 MPH, then apply the brake for a few seconds to evoke EV mode which would slow me to about 35 MPH, and then I would use battery power to maintain EV at 35 MPH the best I could. On a couple, but not all, of the rolling hills I put the car in Neutral and "coasted" in EV without battery power at 35 MPH plus or minus 2 MPH.

Here's the data reported by the ScanGauge:

Cruise Control set at 35 MPH
HV battery SOC 53.0% at start of drive.
2387 RPM was max. value during entire trip.
16.0 miles round trip
34.0 MPH average speed for round trip
HV battery SOC 53.0% at end of drive.
41.8 MPG ( with 30% ethanol )

Manual Driving with as much EV as possible
HV battery SOC 53.0% at start of drive.
2640 RPM was max. value during entire trip.
16.0 miles round trip
33.0 MPH average speed round trip
HV battery SOC 52.5% at end of drive.
44.6 MPG ( with 30% ethanol )

I was going to repeat the test 3 times, but it takes a long time to go 32 miles at 35 MPH! ( and I had to go pee really bad )

If the results were less than 1 mpg apart, I would have repeated my manual EV driving to try and refine my technique, but I figured the results were obvious enough after one trip. Plus why waste two more gallons of fuel ( can't call it gas anymore! ) that would be hypocritical.


So Stan, does that give you a good feel for things?
-John

 

Stan - Let me first apologize for the tone of my post #116. After cranking through the numbers for the Prius, I was astonished at the results vis-a-vis the Escape. The BSFC curve vs. power for the Prius is much flatter above 10kW than the Escape. There does appear to be no EV benefit at cruising loads >10kW due to the lower energy storage capacity and flat BSFC profile. At lower loads (<8kW) EV does appear to give a slight positive advantage, although the calculations required eyeball interpolation of the constant BSFC contours at the lower power levels. Plus, the MG efficiency contours are not smooth which further muddy the water.
I can not hold EV in my Escape at much more than 35mph, so I didn't look at anything higher than that. At 30mph, the calculations for the Escape show a net EV benefit of about 8% - definitely significant. When you add in regen, you can obviously do even better. But I guess when you're worse to begin with, there's more room for improvement....
When Ford flattens out their BSFC curves, I guess I will no longer be able to outsmart them, xgauges or not.......

 

I apologize for the delay in my replies to recent postings; I didn't have the time to compose a detailed response until now.

DesertDog — No need to apologize! I wasn't offended, and anyway I prefer straight talk. But I think you misinterpreted some of the comments in my post #115. Perhaps I wasn't as clear as I had hoped.

• Yes, an ICE is usually more efficient at powers closer to its maximum capability than it is at low power output. That's one reason why a full hybrid (like the Toyota and Ford designs) downsizes the ICE and uses the battery and MGs to supplement its torque/power for short-duration peak demands. This way, the ICE can operate closer to its peak efficiency when power demands are smaller, such as when the vehicle is cruising at a steady speed on level ground.
• I didn't mean that you had actually re-programmed Ford's ECUs! What I meant was that you have knowledge of your route and of the workings of the FEH that allow you (using your instrumentation) to achieve better results than Ford's built-in programming can achieve.
• Bear in mind that I started this thread because of the widespread opinion that, the more pure-EV operation one could coax from the TCH, the better one's FE would be. I contest the correctness of this belief, since all the energy used by the car (including electrical energy recovered by regenerative braking) is ultimately traceable to energy derived from the gasoline via the ICE. Due to the mechanical-to/from-electrical conversion losses in the MGs, and especially due to electrical-to/from-chemical conversion losses in the NiMH battery, I seriously doubt that an average driver can outsmart the Toyota ECUs in general. I laid out conditions for a controlled test to verify the correctness or otherwise of my belief, and I even produced data that supports my contention.
• I believe that "heretical" mode does increase the overall efficiency of the system by forcing the ICE to run at a more efficient operating point (generally by forcing the engine's rpm down and its efficiency up while keeping its power output constant). I don't agree that energy storage (i.e., the NiMH battery) is needed for this. Indeed, in steady-state cruising on horizontal ground, no net energy flows in or out of the battery — the electrical energy flow is between the two MGs only. The planetary-gear set plus MGs together act as a very wide-ratio CVT, enabling the ICE's operating point to be better optimized for efficiency over a wide range of speeds. Some energy is lost in this reverse energy flow because of conversion losses in the MGs, but if you meant losses other than this in your statement: "All the MG1/2 can do to increase the load of the ICE in steady state cruise is by throwing away energy", I don't agree at all.
• I implicitly excluded the use of pulse-and-glide and other similar methods of increasing FE (e.g., coasting in neutral with the ICE "off") in my challenge. Such methods can indeed get greater FE, but they aren't, shall we say, "normal" driving techniques. I intended the vehicle to stay in 'D' or 'B' (i.e., 'L') all the time. Look how consistent my test results were in post #48. However, not being able to measure the battery's SoC, it's likely that after the two "maximum-EV-mode-driving" runs, it wasn't actually back to the same level as at the start, even though I couldn't tell this from my non-NAV TCH dash's simplified display. If this is so, then the test would have in fact shown a decisive FE penalty from those runs that forced the maximum amount of pure-EV-mode operation.
• I don't have any ICE efficiency information specifically on the TCH. If you have such, I would be very interested in seeing it. In fact, I don't have such data for the Prius either, and would be likewise very interested in seeing it. All I have is what is shown on page MO-18 in the TCH "New Car Features Guide." I am attaching it hereto for your interest.
• I have only generalized (possibly Prius-related) data from Toyota's US patents. Their patent #6 131 680 (attached), for example, shows such curves in Figs. 12 and 13, with a good description of their general strategy in the text column 25, lines 25 - 58. (Where they say "constant energy curves" they obviously mean "constant power curves.") The strategy, when the required power is say that represented by the constant power hyperbola 'C3,' is to force the ICE's speed down and its torque up by moving the engine's operating point to the left along the curve C3-C3 towards the maximum-efficiency point 'A3.' This yields the greatest possible efficiency from the ICE at that power. (It is, of course, a lower efficiency than that represented by the maximum-efficiency point 'A1' in the torque-rpm plane.)
• I find your comments re the efficiency of the FEH versus that of the Prius very interesting. I wouldn't have expected such a big difference. No doubt Toyota has learned a great deal that they haven't divulged in the ten years that they've been making hybrids!
• I have two SGII's, and I intend to get them upgraded to XGauge capability. But I'm not aware of any SG codes for reading such things as SoC, etc. in the TCH. The Prius SoC code does not work in the TCH. Can you direct me to anywhere that I might find such information? I have had extensive e-mail correspondence with Ron DeLong at Linear-Logic about various issues (such as fuel-cut) with the SGII over the past year, but I'm unaware whether he has any such XGauge information for the TCH.

gpsman1 — It's ironic that you, as an FEH owner, have conducted (almost!) the very test that I've been trying unsuccessfully for more than a year now to get other TCH owners to conduct! But well done, and thank you! Your FEH data is welcome here. Your experiment seems to have been very carefully done, and I think that you have produced good data. I say "almost," however, because you used 'N'-mode coasting during part of your EV-mode run, and this takes your experiment outside the controls that I had envisaged. (See my bullet point #5 above.) By the way, I have also been reading your comments in the "Ethanol" thread with interest. I chose a somewhat higher speed of 65 km/h (~40 miles per hour) for my cruise-control runs, so as to ensure that I got no EV-mode operation at all on the relatively flat road that I used. This introduced a significant difference between the average speeds in my two driving modalities. You didn't have this difference in your experiment, and this is good. You have found a significant difference between the two driving modalities in your FEH. I wonder how much of this is attributable to your use of 'N'-mode coasting? Please be so kind as to repeat your test without use of 'N'-mode, if at all possible! I (and I would expect you too!) am most interested to learn how much (or little!) of the FE difference then remains!

Stan

 

Stan - No repeat of my test(s) is necessary.

Shifting to "N" or not will have zero effect on the results.
In the case of the FEH there is no mechanical advantage to moving the gear select to "N". The car, on its own, does a "neutral" glide anytime there is no demand for power or regen, even in D or L position.

There have been lengthy discussions on this, and this fact has been well illustrated. The "N" selection is a matter of user convienience, nothing more. The car will in fact "neutral" glide when cruise control is set, if the downslope exactly matches wind resistance, etc.

You have to take "Neutral" with a grain of salt, hence the quotes.
Nothing mechanical happens in this car, only by software, regen is disabled. The transmission and MG1/MG2 continues to spin when in "N".... even while towing ( which I proved in another thread ).
-John