Tarabell - CleanMPG.com - June 30, 2006 Section I - Setup, Driving Routine, and FE Testing 2006 Honda Civic Hybrid The HCH-II - Honda’s most fuel efficient 4-door sedan This article describes my personal experience in adapting basic hypermiling techniques to the HCH-II. More advanced techniques such as FAS and P&G are not covered here but hopefully there will be a future article on these as well. Owners of earlier HCH models might also find some information here applicable to their automobile but this article is really meant to be specific to the HCH-II. I divided the article into 3 sections for better understanding and readability. I hope you find the same. Driving the HCH-II involves much more of your senses than driving other non-hybrid and even some hybrid automobiles: Not only are you watching your instrumentation closely, you are also observing the road, timing traffic lights, noticing terrain, listening to the engine and gear train, and feeling the car's actions from the seat of your pants, as well as through your foot all at the same time! Basic and Advanced Setup As with all automobiles, a good setup can make the difference between good FE and great FE. The HCH-II is no different and in fact, may be more susceptible to an improper setup due to the smaller ICE. The benefits of break-in for example are gradual but as it progresses I can expect slightly longer glides, lower torque requirements for a given speed or terrain, and a slightly smoother running and more efficient ICE. While accumulating miles there are other things you can do: set tire pressure at maximum listed sidewall or higher, use a high quality synthetic oil such as Mobil1’s 0W-20 or 5W-20, and quantity at no more than max height on the dip stick. I currently run 44 psi in my HCH-II’s Dunlop’s, heading up to 50 and switched to the high quality, low kinematic viscosity, Honda 0W-20 on her first oil change. Another more advanced setup solution could include the removal of the DRL (Daytime Running Light) fuse so as to limit the current draw from the 12V which draws off the 158 V pack by another 100 + W. That seemingly small wattage has to be replaced and it will come either from “Regen”, hidden, or forced charging from the ICE. So it should be possible to eke out another ~ .2 + mpg’s during daylight hours with this mod. Below is the fuse to be removed in order to disable the DRLs. A fuse map is also in your Owner’s Manual. I haven't tried this myself and do not believe this mod has been widely used in practice so there may be unknown consequences. However it should not affect your parking lights and headlights. Hopefully others will post their experiences regarding this in the future. Disabling DRL’s Interior fuse panel - Fuse #37 removes DRL’s from service Also, to remove some of the warm up FE hit from initial cold light off, I’ll pass on Xcel’s recommendation that everyone consider installing a block heater, given coolant temperature is a major input into the ECU for when to allow closed loop operation or not. Accelerating from a stop up to given speed Start up: When I start my car after it’s been cold for many hours or after overnight while sitting in the garage or drive, it usually starts with high revs as most cars experience during start up with an initial rich mixture like mode to warm up both the coolant and catalytic converter(s). I find no need to give the HCH-II any gas to get to the first stop sign from my home or to get out of the parking lot at work. Once past the first stop sign or so, the revs come down and I’ll then start using the accelerator pedal for propulsion. This cold startup phase typically lowers my FE the first couple miles out. Acceleration: I personally prefer slow acceleration over fast so typically I try not to go over 2000 RPM’s. That number has been mentioned a multitude of times by hypermilers as an informal rule of thumb, and one of the tests I ran for this article backs this up as a valid benchmark. I give it just enough accelerator to get up to speed with the current flow of traffic. My main aim is to ease up on the gas pedal as soon as possible and get the iFCD close to or at its maximum. The other thing I’ll watch is assist. If I accelerate slowly enough, I can forgo any assist but typically I try not to go over three bars of assist on most accels. The main thing I watch is the tach. If I have to do a high speed accel like onto a freeway onramp, I will get up to speed quickly then pulse my foot off the pedal a few times to lower the tach down below 2000 RPM’s as quickly as possible. This is known as Fake Shifting or FS. I will discuss more about foot to accelerator pressure later in the article. Merging onto a freeway is the only time I have difficulty keeping the tach below 2000 RPM’s as I’m forced to match the speed of cars in the right lane into which I am merging. At speed: Once I’ve eased up on the pedal to a near constant speed, the iFCD is usually > 75mpg if on a level road and I can then hold it there with very little accelerator pedal pressure. The HCH-II can easily maintain speed with a fraction of the pedal pressure another car would normally require. I find I can keep it near 90 mpg in city driving with little trouble if no traffic is around and can often nudge it into an EV/Glide mode which I will also describe later. I will only match the flow of traffic if necessary. If there is no traffic behind, I’ll go slower than the speed limit, preferably 25 – 35 mph range on a 35 - 45 mph roadway. For freeway driving I stay in the far right-hand lane, typically in the 50 - 60 mph range while DWL, DWB, and RR. I moved from my favorite 2nd-to-left lane very soon after I started recording segments as I found anything above 60 mph has a definite impact on my FE. I do not use cruise control as a rule for freeway driving, even long ones! Cruise control feels too “passive” to me and since I paid for all these fancy hybrid gauges for feedback I figure I should use them. With CC on, I might as well cover them up. Below I included some CC and non-CC based steady state cruise tests which most should find enlightening and maybe shocking, as to the proper way to drive the HCH-II for maximum FE. Speed and Acceleration Tests Now that I’ve introduced how I come up to speed and maintain it, this is a good time to discuss two very specific tests I performed for this article. The first test looks at what effect speed has on FE. This test compares speeds two different ways - using cruise control and without - which also helps answer the question of which method is likely to allow the highest FE. The second test looks at the effect of different acceleration rates on FE. The question of whether one should accelerate quickly or slowly to get up to desired speed has been a perennial forum discussion and debate and I hope the data helps shed some light, at least as far as the HCH-II is concerned. Road and Test Conditions: Finding ideal roads for these tests seemed like half the battle. First, I wanted to be able to use a single street for all runs for data consistency but there are not many streets one can drive in a range of 30 - 60 mph (at least legally). The route also needed to be a couple miles long to allow for start up and slow down, no traffic signals, hopefully little traffic, and of course as road grade level as possible! Piece of cake I finally found a divided road that met most of these conditions except for one, a “T” intersection traffic light midway. That light rarely changed except when triggered by another car approaching and of course that only seemed to happen just as I’m barreling towards it during a run. Other traffic was a nuisance too as I would get boxed in behind a slow moving truck, a car would slow down in front of me to turn into a driveway, or some other such nonsense causing me to break speed. I expanded my library of colorful expressions and gritted my teeth through at least 2 - 3 runs for every good data point. Then I noticed if I mixed runs going in different directions I got vastly inconsistent data which implied the road was not as perfectly flat as I thought so I ensured that all the speed tests were run in one direction and the acceleration tests were run in the other direction so that each run used the exact same segment. The only test I could not run on this road was the 60 mph test, since the speed limit was 50 mph and it was too far from home for midnight testing. For that test, I picked a distinct, flat freeway segment on my commute home and recorded a run there every day for a few days. Here are two views of the test road: one looking east and one looking west. Mid-speed test road heading East and then West - Suburbs of LA. As far as other environment conditions, the tests were run with a warmed up car with full battery charge, no A/C on, and the windows cracked a couple inches. There was no wind and temperatures were in the high 70’s. I found most of the data to be quite consistent, but sometimes there were surprises between similar runs. Test 1: Speed Effects on FE Test Steps: 1) Accelerate to desired speed (30, 40, 50, and 60 mph). Maintain speed using cruise control, then zero Trip-A odometer and drive until 1.0 mile reached. Record trip FE. 2) Repeat above for all speeds, except without using CC. Trip-A – Constant speed tests - MPG Results Column1Column2Column3Column4Column5With Cruise ControlRun #30 mph40 mph50 mph60 mph184.3 mpg89.3 mpg69.1 mpg56.4 mpg286.5 mpg89.0 mpg64.8 mpg48.5 mpg385.1 mpg90.7 mpg67.3 mpg50.5 mpgAverage’s85.3 mpg89.7 mpg67.1 mpg51.8 mpgWithout Cruise ControlRun #30 mph40 mph50 mph60 mph1109.6 mpg95.4 mpg70.9 mpg57.9 mpg299.2 mpg96.0 mpg71.4 mpg56.0 mpg397.6 mpg87.4 mpg70.2 mpg52.9 mpg4140.0 mpg92.2 mpg68.8 mpg53.8 mpgAverage’s111.6 mpg92.2 mpg68.8 mpg53.8 mpg 30 mph steady state FE w/ CC engaged - 1,075 RPM. 40 mph steady state FE w/ CC engaged - 1,175 RPM. 50 mph steady state FE w/ CC engaged - 1,500 RPM. 60 mph steady state FE w/ CC engaged - 1,800 RPM. Notes: The cruise control tests did not show much variation so I only did 3 runs with CC. The tests without CC were more difficult to run holding a set speed and the results showed it so I added a 4th. I was only able to take photos of the iFCD during the runs with CC, as I found it too dangerous otherwise. I want to add that all the above results were still increasing after the 1.0 mile point, so they could have been higher with even longer segments! Conclusions: The FE results with Cruise Control were somewhat different than I had expected. Of the 4 speeds, 40 mph clearly gave the best FE, even 5% better than 30mph. At higher speeds the results show a significant decrease of ~ 25% for each 10 mph over 40 mph. Without CC, the tests show a quite predictable loss of FE with increasing speed of ~ 20 mpg for every 10 mph over 30 mph. Comparing FE both with and without CC, the driver’s right foot easily wins at very low speeds but by much less at higher speeds. I suspect the 2 extreme data points at 30 mph were due to going into “EV-Assist” (but my eyes were glued to the Trip A odometer unfortunately). At 60 mph, the gap between using CC or not narrows significantly. Because the 50 & 60 mph FE results appear low compared to what you would think the HCH-II was worth, I went and clocked an ~ 10 mile segment for comparison at 60 mph on the freeway where I got about 75 mpg (no CC). That is a huge difference which I believe indicates that FE depends to some extent on the segment length you are testing/driving. Test 2: Acceleration Effects on FE Test Steps: Zero Trip-A odometer while at a full stop. Accelerate smoothly until the tach needle reaches desired RPM’s (1900, 2100, and 2300 RPM’s). Maintain RPM’s until speed reaches 50 mph and use CC to maintain speed until 1.0 miles has been reached. Record trip FE. Trip-A – Acceleration tests - MPG Results Column1Column2Column3Column4To 1.0 milesRun #1,900 RPM2,100 RPM2,300 RPM144.2 mpg40.9 mpg41.6 mpg244.2 mpg42.1 mpg42.9 mpg345.4 mpg42.4 mpg42.7 mpg447.7 mpg42.2 mpg41.4 mpg544.2 mpg40.9 mpg40.3 mpgAverage’s45.1 mpg41.7 mpg41.8 mpgTo 0.6 milesRun #1,900 RPM2,100 RPM2,300 RPM138.6 mpg35.3 mpg34.7 mpg238.8 mpg35.3 mpg35.2 mpg3--34.9 mpgAverage’s38.7 mpg35.3 mpg34.9 mpg Notes: In the HCH-II, it’s a bit hard to distinguish between 2100 & 2300 on the tach since the marks are so close together. Smooth acceleration is the key because if ragged, the tach will bounce around harshly Because there was very little difference in results between the 2100 and 2300 runs I wondered if the 1.0 mile distance was too long for this type of test? So I performed a second set of tests using a smaller 0.6 mile segment to see if I did something wrong with the first and to see if the results would spread more. There was little change as noted. Conclusions: Accelerating above 1900 RPM’s results in about a 10% drop in FE. I was surprised to see the 2100 and 2300 rpm results so similar. ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Section II – EV/Glide Modes and Experiences Regen/Glide/EV-Assist Modes The different EV/Glide modes on the HCH-II are important tools in increasing FE. The easiest way to describe them is to go through a scenario I typically use that engages all three of them in sequence. I’m going to use the shorthand “Regen”, “Glide” and “EV/Assist” for the 3 modes and refer to them in this manner throughout the rest of the article. Assume I'm cruising at 55mph on level highway: “Regen” - This is what I always considered “coasting”. While at speed I simply lighten or remove my right foot from the accelerator pedal. The Charge display shows green bars indicating energy is being recaptured and is being used to charge the battery. Less foot pressure, faster deceleration => more bars in the Charge display, more “Regen” charge to the pack. The iFCD will move to the right and peg at 100 mpg. However, “Regen” slows me down more than I normally would in a non-hybrid car with standard ICE-braking. Even if heading downhill, I can feel the slowing effect of “Regen” on the car. In this mode as well as the next two, the ICE is still turning over with the transmission engaged, but the valves and injectors are closed and no fuel is being consumed. I also find the engine very quiet in all three modes. Regen Losing speed quickly with full “Regen”. “Glide” - While watching the Charge display, I gently reapply pressure to the accelerator pedal and the regen bars start to disappear. When completely gone I am still “coasting” except now I’ve eliminated the drag caused by “Regen” and can coast much further without slowing as fast while the iFCD will continue to be pegged at 100 mpg. This is known as “Glide”. I can hold the “Glide” mode for a good distance, often several miles depending on terrain. Glide Still losing speed but at a much slower rate in “Glide”. “EV-Assist” - While the iFCD is still pegged at 100 mpg and all “Regen” is cancelled, I continue gently applying pressure to the accelerator pedal until I see one or more gray bars appear in the Assist display. This indicates the electric motor is now assisting propulsion even though the ICE is not consuming fuel with its valves and injectors closed. I can hold this for at most, half a minute at a time at high speeds and the car will start to slow down even with foot pressure steady as a rock. This requires a fine foot-balancing act because given too much pressure the assist bars disappear, the iFCD drops, and a distinctive lurch happens as the ICE comes back online to help re-accelerate or maintain speed as well as consuming fuel once again. If the iFCD hasn't dropped too much I can sometimes nudge it back into “EV-Assist”, but often I have to start the process all over again with “Regen” to “Glide” to “EV-Assist”. EV-Assist Attempting to maintain a higher speed with “EV-Assist”. Using the Three Modes Regen: I see “Regen” nearly every time I let up on the gas to slow down for a signal or stop sign, exit off the freeway, or go downhill. I can control the amount of “Regen” I want, basically trading off Charging the battery for speed, but more importantly for FE. If I feel I could use Charge more than Speed, I’ll allow the maximum amount of “Regen”. However, I might want to maintain speed on a downhill because I see another uphill coming so then I’ll cancel “Regen”. I will discuss more “Regen” effects on FE and my slowing/braking process later in the article. Glide: The majority of time on the freeways I find myself in a low load ICE-On condition without the iFCD necessarily at 100 mpg but as high as I can make it with no visible Assist or Charge bars. Since I don’t find “EV-Assist” sustainable for long periods at high speeds, I am quite happy in either “Glide” or the ICE-On at a very low load with 75 or higher mpg on the iFCD, otherwise known as DWL. Technically DWL means holding the iFCD constant while allowing your speed to vary as necessary, which is especially useful on smaller hills and overpasses. On level roads I simply try to max the iFCD as high possible. If that is 75 - 95 mpg, I am quite content and leave EV mode alone EV-Assist: In city driving, with few traffic lights, I find it easy to hold the iFCD at > 90 mpg and often cross between low load ICE-On, “Glide”, and “EV-Assist” modes easily. “EV-Assist” is fun and seems to be what most HCH-II drivers want to practice but it is costly to my SoC if I abuse it. I try to be sparing with this EV mode, but again it’s a tradeoff in different circumstances. If my battery’s full and I see a great segment in the making, I’ll use it. If I’m heading downhill, often I will find myself drifting into and out of “EV-Assist” and I’ll maintain this but will not force it. It is great fun for example, getting up a good head of steam then easing into “EV-Assist” with up to 3 bars of Assist, and being able to climb up and over a short hill using that Assist. (I must go for cheap thrills ) In city driving, I will use “EV-Assist” more, especially at low speeds. Which brings me to describe an interesting variant of this mode. DWL DWL - Maintaining 85 mpg on the iFCD at 52 mph Variant of “EV-Assist” mode at 1000 RPM’s Most HCH-II owners are familiar with engaging “EV-Assist” at high speeds in the sequence described above: First coasting with “Regen”, canceling “Regen” with accelerator pedal pressure to allow “Glide” and continuing with even more accelerator pedal pressure until “EV-Assist” appears with bars in the Assist display. However, the HCH-II also appears to be able to go directly into “EV-Assist” mode at low speeds (< 35 mph) and also appears to do this from a dead stop under certain conditions. I say “appears to” because all the reports I’ve read regarding “EV-Assist” for the HCH-II say it does not have this capability, which is one of the main differences from the Prius. I have experimented with this variant at least a couple dozen times now and can find no better explanation for what I am seeing? It only seems to work when my car is well warmed up, such as right after exiting off the freeway. From a stop, it would look like the following: I accelerate very slowly using “feather light” acceleration, with tach at 1000 RPMs heading up to 12-15 mph. It is best done either in very heavy traffic or in light traffic with no cars behind you. I see one or two bars of assist at the same time as I (barely) press the accelerator pedal. At the same time as speed increases and assist appears, I observe the iFCD swing quickly and smoothly to the right. It will stay pinned at 100 mpg, while I level off speed, staying somewhere under 20 mph. The assist bars remain, typically varying from one to three. The car is very quiet and once at steady speed the gauges are indistinguishable from normal “EV-Assist” mode. I feel the tell-tale bump and see the iFCD drop as soon as I raise the tach above 1K RPM. However as soon as it drops back to 1K I can easily pin the iFCD back at 100 mpg, see the Assist bars, and drive along some more in “EV-Assist”. So the 1K RPM tach seems to be the key here. Of course one does not usually want to accelerate this slowly, so it is not that useful a feature to exploit on a daily basis. However it is extremely easy to decelerate into “EV-Assist” mode during normal city driving, which I’ll return to in a minute. I want to go into a bit more detail here about how accelerating into “EV-Assist” looks different from a normal “EV-Assist” mode. 1. I see Assist bars immediately with accelerator pressure. Normally I don’t expect to see Assist at slow accel (and in normal driving I try to accelerate with minimum assist.) 2. Unlike the typical “Regen”-“Glide”-“EV-Assist” sequence with “Regen” appearing first and then “EV-Assist”, “Regen” does not enter into the sequence at all. 3. It is odd to watch my iFCD rise during accelerations. Normally during accel, the instantaneous stays close to zero until you’ve leveled off speed somewhat, then rises. Under this variant, the iFCD rises immediately like an invisible hand pulling it up. 4. This variant stays much more constant than when in “EV-Assist” at higher speeds. At high speeds if I drop out of EV, I usually have to go through the sequence of “Regen” to “Glide” again to re-engage it. At low speeds however, as long as I stay at 1000 RPM’s I can stay in “EV-Assist” a very long time, and re-enter easily, if needed. The Assist bars rise and fall directly with gas pressure. If I let up on the gas enough, the Assist bars drop out but iFCD will stay pinned at max, I’m in “Glide”, ready to go right into “EV-Assist” again. If I’m briefly too hard on the accelerator to where the iFCD falls, I just ease up enough to get the tach back to 1K, get the Assist bars right back, and iFCD pinned at max again. 5. I do not observe much bleed-off of speed, if any, as I do when in high-speed “EV-Assist”. I sure would love to drive around a huge empty parking lot some Sunday morning, with at least 2 bars of assist and test how far my odometer increases, FE climbs, and battery SoC can go driving in this manner. Accelerating in “EV-Assist” at 11 mph - iFCD pegged. Still Accelerating in “EV-Assist” at 12 mph - iFCD pegged. On city streets I find extremely easy to decelerate into the 1K RPM “EV-Assist” mode from higher speeds, as long as I get my tach down and keep it there. I had thought 25 mph was about the fastest I could go at 1000 RPMs, but recently I noticed I could go at > 30 mph as well. It’s very useful when I enter a large mall parking lot or quiet residential neighborhood. In that setting I can hear the difference in quietness of the engine compared with regular driving. Anytime I am city driving in the 20 – 30 mph range I can transition into this mode. In fact, I am quite sure many HCH-II drivers have often been into this mode at this low tach, but perhaps have not realized it? Low RPM - “EV-Assist” at 27 mph. Low RPM - “EV-Assist” at 32 mph. However … this low-speed version of EV-Assist hits my SoC just as hard as the equivalent higher-speed mode does so I try not to over use it. I’ll describe some of the effects of pack overuse later in the article. iFCD “recal” while in “Glide” Green&Blue first reported this unique mode/feature in the thread entitled HCH-II on electric only and I hoped to try and fit it into this article if I could reproduce it. Luckily I have, so if Mark does not mind, I will summarize his description here. Basically what happens is that while in a “Glide”, between 30 - 42 mph and the iFCD maxed, the iFCD will unexpectedly drop to nearly zero and then does a jerky crawl back up towards 100 mpg again and stays there. I have only seen this happen when my engine is starting from cold, and I'm heading down a slight hill. When this happens I do not see the aFCD increase as I would expect from going downhill, but can't be sure if that's due to the cold startup or some effect of the recal. Mark however reports he can maintain this longer than “EV-Assist” mode and at higher speeds too! I can attest that it’s unusual to see the instantaneous suddenly drop for no apparent reason. The typical driver reaction is to wonder WTH happened and immediately lighten your right foot to make it go back to the right. If you’re patient and maintain the exact same foot pressure, the iFCD will climb back up all by itself (with a samba rhythm) and remain pretty steady up there. It appears to be some sort of recalibration although iFCD’s should not be affected by this phenomena knowing what we know today? ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Section III - How it’s accomplished Braking & Managing your SoC The topic of slowing and/or braking to a stop will be a good lead into “Regen” and SoC management. On city streets, my goal is to try and maintain a large buffer in front of me (DWB) and anticipate traffic and traffic signals. Of course there are times I have to brake and depending on my SoC, I’ll try to make braking work for me by coasting into the stop without “Regen” in a “Glide” for as long as possible in order to maximize FE. In the recent past, I had not consistently done this as I believed maximizing “Regen” was more a cost to speed, not to FE. Then I realized the drag I felt must be costing me FE and Xcel gently confirmed that yes, charging during slowing and braking “Regen” can be just as costly to FE as a “Forced Charge” when the SoC dips too low. Now when exiting the freeway, I’ll glide to the off-ramp at 50 mph, just maintaining enough accelerator pedal pressure to cancel any bars of charge, then brake when I have to at much slower speeds about 100 feet from the stop or less if no traffic is behind. Recently I was driving LA. city streets in heavy weekend traffic and found myself in “1K RPM EV-Assist”. I was able to maintain this mode consistently for some miles. Also as it happened, I was DWB through heavy traffic and residential side streets for maybe 5 - 8 miles. My FE was great and even rising on city streets. I was so proud of myself thinking that with this low speed “EV-Assist” and DWB combination, I can do just as well in town with my HCH-II as the Prius II can! Well … guess what happens when you rarely stop or allow “Regen” AND you have fairly constant “EV-Assist” appearing? It is a killer combo hit to your battery and that is exactly what happened. My SoC dropped down to 4 bars (8 is max) and on the way home I experienced “Forced charging” which brought my previously excellent FE back down to earth again. I had thought using EV at only higher speeds was the battery killer. As it turns out, it is not for me because I do not maintain “EV-Assist” for very long at freeway speeds. It is the city driving where I have to be very careful with my reliance on “EV-Assist”. I find the SoC can vary from 5 to 7 bars in normal driving and even stay down at 5 bars without much effect on FE. However, when I get down to 4 bars or less, I can almost guarantee some “Forced charging” will rear its ugly head. If I am on the freeway, it lasts a short period of time over maybe a 3 - 5 mile period where I will see 4 - 5 Charging bars appear. During that time, I can expect maybe a 10 - 20 mpg hit to my then current instantaneous FE. When it subsides, my battery is fully charged again. More frequently than this major type of “Forced charging”, I’ll see just one green line of charge that shows up when I get down to 5 bars on the SoC. This single line appears while driving for a much longer period - sometimes for 8 - 10 miles. That also has some effect on my FE but not as strong as a major “Forced Charging” Event w/ 4 + bars appearing in the Charge display. Either way, the single charge line can persist even when my SoC has increased up to 7 bars and can return for several days which can be somewhat irritating. Forced Charging 1-bar - Light Forced Charging In most instances, hidden charging instead of the 1-line kind as shown above happens and the SoC will slowly tick back up to 7 bars as I drive with no visible green lines occurring. I do not have a hard rule but generally when the SoC drops to 5 bars I’ll take advantage of “Regen” more to “manage” my SoC. In this case, if I am coasting to a stop, I will also apply light braking in order to maximize “Regen”. I will also avoid “EV-Assist”, stay more in ICE-On mode, maximize DWL, plan my accelerations to avoid Assist and hope I do not drive into a hilly area. Techniques for hills The HCH-II has a built-in tool to help the driver maximize FE when going up and down hills - the iFCD. Normally in a non-hybrid when driving hills, I used to maintain a constant speed, so I would spend more gas going up a hill and less gas when going down. Using DWL, instead of holding my speed steady, I hold the iFCD steady at a given target and allow my speed to vary as needed. This results in a sort of roller coaster effect where my automobile slows as I approach the crest, rebuilds speed on the down slope and repeat as necessary for the next small hill or overpass. My other objective is to use the least possible Assist during the climb. One of the ways I’ve found to do that is by accelerating on the flat approach to the hill, then more or less coasting partway up the hill with less pressure on the accelerator pedal which keeps the iFCD up. That way I am using assist on the flat portion and not on the hill itself. I might increase speed from 55 to 60 mph on the flat portion, then allow it to droop to as low as 45 mph while climbing. On the descent, I find myself in “Regen” and try to cancel charge bars to maintain speed, unless there happens to be a stop at the bottom in which case I take all the “Regen” I am given with full "Regen" and possibly light mechanical braking to a stop. The HCH-2 Accelerator Pedal Here I would like to mention the accelerator pedal’s unique features, where I like to apply pressure, and the different ways I apply pressure depending what type of terrain I am traversing. Foot Placement The accelerator pedal is designed differently in the 8th generation Civics, as it’s attached to the floor with a bottom pivot. This may explain why I find I get the most control then, pressing at the very tip-top of the pedal. Same as what your basic physics textbook says about levers & fulcrums. Applying pressure at the top of the pedal lets me vary those minute increments of pressure more easily. This also means I have to hold my leg out higher so I’m hitting the top of the pedal with the ball of my foot. Sometimes my foot will shift around a bit as I drive but after about five thousand miles I’ve gotten used to holding it there. I have tried a wide variety of foot placements and this one may be a little tiring for some but I do believe it gives me the best FCD control. Foot Pressure Being able to apply the right pressure at the right time makes a big difference in FE. When I first tried EV mode for example, I found practicing a few times barefoot helped to get the knack of it. Then I drove with shoes as usual (but thin-soled) as naturally I felt safer with shoes and the stability the sole provides. Recently this summer I happened to take off my shoe again and was amazed at the difference. It seems that due to my foot having "learned" the pedal over 8 months with shoe-on, now with shoe-off, its sensitivity to the pedal is simply phenomenal. It’s as if I'd been typing with gloves on and finally took them off. Plus the sensitivity is two-way: Not only is my foot better able to fine-tune application of pressure, it can sense it from the pedal as well. For example, while pressing the pedal in EV-assist, there is a certain point of stiffness that tells me where to stop or I’ll get dumped back into ICE-on mode again. If you have been driving your HCH-II for a while now, and weather still permits, I highly recommend trying shoe-less at least once just to observe the difference. The following are some of the different foot pressures I use every day: Short steep hills - Pulse-off During periods of intense climbing I will “Pulse off” my foot from the pedal a few times in short bursts, quickly returning to original accelerator pressure each time. Each back-off lowers the RPM’s and up shifts me into a higher gear more quickly. This is basically GaryG’s “Fake Shift” (but I didn’t connect this till just recently). I do not do it for battery charge though, just for the upshift-lowering of RPM’s effect. Of course this slows Acceleration a bit so I have to watch the traffic behind me. Inclines - Constant slight foot lift When on a fairly lengthy incline (over a mile for example), I constantly lift my foot just slightly off the pedal every few seconds. Enough to keep the iFCD to the right as much as possible. What I see is the iFCD slowly falling say to let us say 60 mpg, sharply rise up to 75 mpg, and slowly falling again. Hopefully with little to no decrease in speed. DWL - Light foot Basically let your iFCD display guide your right foot. You may think your pressure is light enough already, then the iFCD tells you it could be even lighter. During DWL, I have to continually stay alert that my iFCD is indeed maxed. Often it will drop somewhat when I swear neither terrain nor foot pressure has changed or I suddenly find it can be nudged slightly higher. So adjust foot pressure accordingly as you drive. It’s not called the “game gauge” for nothing “EV-Glide” - Feather pressure Getting a feel for this one takes some time, however once my right foot "got it", it has become habitual. It is the same pressure I would use to engage “EV-Glide” at high speed, or “1K RPM EV-Glide Mode”. The best way I can describe it is this: rest your foot closely against the pedal without actually trying to push the pedal in. To increase pressure, continue pressing against (but not pushing) the pedal. If you are new to this or have difficulty, I recommend trying it shoeless. Again, your gauges will tell you if you’re doing it correctly. In a non-hybrid this sort of foot pressure would probably have no effect whatsoever, however it is quite clear that the HCH-II can respond to more subtle variations in pressure than you could possibly imagine. WOT (Wide Open Throttle) - Only to avoid a potential accident, otherwise never Factors that Most Improve FE Following are some thoughts regarding what I've found seems to have the most impact on improving FE. The first one could apply no matter what car you drive, the second one is addressed more specifically to the HCH-II. Keep a daily segment diary If you have a reliable daily driving routine, I’ve found keeping a diary of your segments a most valuable tool. It works on the same principle as recording what you eat when you go on a diet. Having a record helps you correlate your recorded FE with your drive that day so you can figure out what factors might be working for or against you on the drive and learn to work with them. It also gives you a daily benchmark to beat and keeps you focused on improvement. The best way is to record repeatable segments such as your drive to work and home (without including extraneous trips). After recording these for just a week or two I started to notice where my FE was rising or falling. I used my Trip-A odometer for recording daily segments, and my Trip-B odometer for keeping track of the tank mileage. For example, I know that I get better FE on my drive to work because it’s slightly downhill and my FE coming from work is worse because it is (naturally) slightly uphill. Working with that knowledge I know that on the way home, DWL is what I want to maximize as much as possible. On the way to work, I want to take advantage of all the opportunities for “EV-Assist” or “Glide” that I can on the downward slopes. You will also start to guess fairly accurately what your FE should be at different stages of your segment. For example, I know mine should be nearing the 40’s by the time I get to the freeway. After the first interchange, I should be in the high 40’s. After the main freeway portion, I should be in the 60’s. Finally, once I near work on surface streets, I expect a drop of several mpg because that part is more uphill. If I don't reach these benchmarks (or achieve even higher FE), I notice and wonder what happened and/or where I can improve the next drive to or from work. It could be for example there was a traffic jam that slowed me by 5 mph more than I would usually have gone. The light bulbs start coming on like going 5 mph slower can raise my FE THAT much? What if I purposely go slower tomorrow, can I repeat the result? The shorter segments you record, the more clearly changes in FE along the way will jump out at you. That is just because the trip meter is more sensitive to changes over a few miles, than over many miles. If you are a good data taker and really obsessed with raising FE, you could break a particularly long segment down into several short ones (as far as recording from your trip aFCD). This is like hitting the problem with a 10-ton hammer but higher FE is virtually guaranteed. If you had say a 50 mile drive from home to work where you take several freeways, you could break it into a segment for each. You may need a small pad and pencil, or better yet, a hands-free voice activated recorder so you can note your Trip-A mpg each time you reset it for the next segment. Breaking large segments into smaller ones in this manner will give you great feedback as to what speed and basic tools work best. Afterwards, you can work on optimizing each individual segment. This part is pretty exciting and it is literally impossible not to learn fairly quickly what works and what doesn’t. Eventually, your FE will plateau along with your learning curve and once you start seeing your data stabilize you will want to aggregate it back to 1-way segments, and eventually to round-trip segments only. Finally, be sure to take full advantage of any early high FE in a segment. Most of my record segments started with unusually high FE almost from the beginning. It may have been a result of either skill, luck or a good tailwind but if handed a gift in your first few miles, use it to build upon. However, the only way to see this happen is if you are recording segments and zeroing your trip meter. The point here is do not discount high FE at the beginning of a segment as short-lived. Yes a good stomp on the accelerator can easily undo all the good but if you maintain your normal good driving habits from that beginning high point, you can maintain that margin and your segment will end up even better than expected! Slow, Even Driving and Longer Segments My impressions after 8000 + miles are that the HCH-II likes nice, slow, and constant speed segments. That is one reason I have found that DWB and reduced speed is so helpful to my own FE. The speed test results - especially at the low end - also convinced me of this. Second, I would have to acknowledge it is more difficult to beat the EPA if you only drive your HCH-II a few miles a day in the city. Certainly not saying it cannot be achieved, just that one has to work harder to do it. My take is that the HCH-II likes having room to build FE, so the more miles you give it, the more mpg’s it can achieve. As mentioned in the speed test results, there was a big difference from the results of the one-mile test at 60 mph which averaged 55 mpg -- and my informal 10-mile test that came in at 75 mpg. That tends to make me a believer that short segments just are not going to give you the same FE as longer ones. In Closing ... To recap, try to use the accelerator pedal very lightly during any acceleration, keep as low and steady a speed as possible for maximum FE, DWL and DWB to maximize the HCH-II’s potential in traffic and while traversing low hills and overpasses, use your various EV/Glide modes with a light use of “EV-Assist” when and where it can be exploited for maximum gain, maintain a good record of FE at various check points along your daily route so as to monitor progress to your target FE, and keep your HCH-II well set up. There are other even more advanced driving techniques that can and should be exploited which I haven’t touched upon in this article. With them, the FE possibilities of the HCH-II may even breach those available from the Insight 5-speed at low speed highway cruise or the Prius II under a low to mid-speed P&G scenario. Even without the advanced hypermiling techniques, the HCH-II should be good for ~ 50% above the combined EPA using these basic techniques and exploiting the HCH-II’s built-in features when appropriate. I hope you find a number of these techniques useful to improve upon your own HCH-II’s FE. Finally, I would like to thank Xcel for the huge compliment of asking me to write up this HCH-II FE article considering how little I know about cars, and that there are many better hypermilers out there. I have found asking Wayne about FE techniques can be a little like walking in front of a fire hose however I’ve found his input priceless and his patience and support invaluable. The 2006 HCH-II - Capable of Insight like FE with a little help from you.