Friday, April 1, 2011

Prius-ing with Hobbit

fyi: non-Prius users will find this post relatively useless, except pretty pictures at the end :-) Prius drivers, read on.

Tuesday morning I headed up toward Boston to meet Hobbit. He had very kindly offered to meet and show me his highway techniques in person, since I had largely failed in the couple highway trips I had tried. I would usually get a little over 50 mpg, and felt like I was constantly jerkily speeding and slowing, speeding and slowing. It didn't feel like the right way, and the poor results confirmed it. From reading others' experiences on the PriusChat forums, I got the feeling this is a common problem.

I got up a little earlier than needed and attempted to hypermile up on unfamiliar back roads. End result: 65.5 mpg over 61 miles or so in clear, low- to mid-40°F weather. I was hoping for 70, but I killed my chances near the beginning getting stuck in a traffic jam on the ascent of a bridge while in stage 3. It was also prime school bus time, and the stops were frequent and many.

After a rather time consuming drive, I pulled into the parking lot where we'd agreed to meet and called Hobbit. He said he was parked at the parking lot perimeter - well, it wasn't too hard to spot him. Here's our Prii together. Even if you're not already familiar with Hobbit's site, you can probably guess which one is his.

We headed out onto route 93/128/1/95 (yes, it's that confusing) in Hobbit's car. Driving with him is something of an experience - I won't spoil the surprise in case a serendipitous series of events ever puts you into this situation, but suffice to say he's very into following distance.

Now for the driving technique. His sweet spot posts (first one, second one) clearly delineate the need to accelerate at an output of pretty much 15 kW. I was a bit surprised to see him jump straight to 25 kW as we merged onto the highway heading uphill. He explained it was important to just get to speed and not interrupt the flow of traffic, and plus you need more juice heading up a hill to keep moving. He eased up toward 15 kW near the crest, and, much to my surprise, didn't go into warp stealth on the downside. He used it to continue accelerating to an ideal speed (in the first part of our ride this would tend toward the low 60 mphs), and then continued a steady, slow acceleration up the next slight hill, too. At this point, even though I was watching kW since that's what he talked about a lot in his article, he was really watching the RPMs. He liked to keep it under 2000 RPM in some of these initial accelerations. The thing that struck me here was that he wasn't concerned if he was accelerating at 30-40 mpg for a good distance; it was all about accelerating at the most efficient state possible for a situation, and not at all babying the accelerator for the instant gratification of high instant mpgs on the display.

He then regularly started using warp stealth when he felt it was appropriate on downhill segments. He would keep burning if needed, like if there was a truck coming up quickly behind us. It's interesting to know that he often would accelerate slightly harder if heading up a hill, to keep gradually accelerating over what felt like a good deal of time - maybe 1 mph increase every 5 seconds, for example. In these cases it wouldn't be unusual to see 17 or 18 kW.

I asked about SHM - he said he doesn't believe in it, and never uses it (!). This is hugely counterintuitive, but we've all seen the 65 and 70 mpg highway photos of his over long trips, which brings up another point. If you drive for 5 or 10 minutes only on the highway each day, you may get some nice highway bars, but your overall mileage will very unlikely be up near 60, barring clement terrain. The mileage will play out over time.

Result: despite feeling like we were burning at relatively low instant mpgs for some time (anywhere from 30-50), the consumption display showed some nice bars of our highway time up near 75 mpg, and all of them were easily over 50.

It was then my turn to try it out in Wentworth. It was rather difficult to keep the feeling of 15 kW in my pedal foot, since a. I wasn't used to it, b. Scangauge has quite the lag on updating kW, and c. if I found myself going at 18 kW, I would let up just a bit, thinking it would bump down to 15, only to find myself at an inefficient 12 or 13 kW. It requires a precision more demanding than a standard glide. Despite this, Hobbit told me several times that it's important to get the feeling in your foot so you don't have to stare at the Scangauge while driving. Typically, if your instant mpg is at 75% the value of your speed in mph, you're right in the sweet spot. On the slightly harder climbs, this ratio would sometimes dip to 66% or so while Hobbit was driving.

I was constantly under the impression that my bars would return 40 mpg given how long and relatively hard the acceleration was, but miraculously all the highway bars were over 50, and there was a nice 75 mpg bar somewhere in there. It's completely counterintuitive to the non-instrumented driver.

Coming back to the parking lot, we found another Prius had sought out a parking space near Hobbit, and the opportunity was used as an excuse for a Prius photoshoot.

After a nice food court meal (Hobbit finally had to put on his shoes to go in the building :-D), I headed out on the highway to drive home, wondering how this would work in normal traffic.

At 15 kW, I found myself going surprisingly fast - it got me quite easily up to 70 mph on a couple occasions, easily faster than I've driven in four months. I would then bleed some of that speed with warp stealth, and up the mileage went. At first my mileage was in the 40s, it quickly rose sharply toward 50. I maintained 15 kW religiously, with the occasional foray into 17-18 kW territory. The mpgs kept increasing with warp stealth glides - 54, 55, 56, 57. Like any technique, it was also dependent on terrain to some extent. As route 24 got hillier, or at least was heading uphill in general for many miles, the mileage slowly came back down to 54. There were almost no opportunities to use warp stealth. While I've driven route 24 from time to time in the past, especially when I was at school in Boston, I didn't really have an intimate knowledge of the terrain coming up, so I imagine there were plenty of instances where something could have been done "better".

Nonetheless I kept on, and to get home, there were two semi-exits to take and two corresponding major accelerations to merge back onto the highway . Toward the end of the ride, where I knew the terrain well, I knew there was a couple mile long stretch where I could mostly use warp stealth, and was rewarded with the second to last bar you'll see below. The last bar I was off the highway, taking advantage of full hybrid capabilities.

Results of the trip: 58.5 mpg displayed over 51 miles (Scangauge inaccurately higher, but looks nice!), which is much better than the 50 or so I would usually get. Also keep in mind that the 2007 Prius, when adjusted for the EPA's 2008 mileage test adjustments, is only rated at 45 mpg on the highway. Hobbit's technique easily yielded at least 25% higher.

Despite the latter half being the uphill/accelerate onto other highways part, and thus lots of 50 mpg bars that you see in the picture, the first part of the trip had several 60 and 75 mpg bars once I was on the highway. Considering I was inching toward 60 mpg on my first try ever, it's pretty exciting to see how much potential is in this approach, since presumably with practice you'd improve. The one major downside is that your foot will get way more tired than pulsing and gliding, since you need a constant, precise pressure applied, either in the finicky 15 kW pedal position or the slight play you get with warp stealth. PriusChat user pEEf confirmed to me that neutral is safe at any speed with the engine running, so that may be a possibility to give your right foot some exercise from time to time in place of warp stealth (though as a side note, Hobbit was absolutely certain that his gas engine does not start at 52 mph when in neutral from a glide under 42 mph (see post #22). He said he's gone up to 70 mph or so in an engine-off neutral glide. Perhaps this is a difference in model years? Hobbit drives a 2004, pEEf a 2008).

I also want to note I never saw him head up to 3000 rpm once we were up at speed. From reading forum posts some people tried that regularly without good results - my feeling (Hobbit, correct me if I'm wrong) is that it's only to be used if you're at your mph floor and you're heading up a particularly tough hill. It's not something to do all the time (just the opposite!), and I don't think he meant to advocate it, if some read it that way.

In related news, that day, with its near-50°F temps later on and another trip I made with a warmed up engine, made for some great results. Even later at night down in the 30s, I got a record mpg number for a particular trip. Nice day for driving overall!

First time I saw 80 in the middle of an actual commute.

The end result. Admittedly the engine was warm before I left...

Trip record on another route. Notice the temperature!

One day after, guess the crazy mileage of the previous day carried over. From a cold engine start!!


  1. Absolutely amazing Mike. I have two questions:
    a. Can this high mileage be accomplished without a Scangauge or similar device?
    b. Do you have the time/patience/desire to teach me how to do this?
    I have quite a significant commute 5 days a week and could really benefit from MPG like this.

  2. a. Yeah, definitely, but not quite as high as with it. I was doing pretty well, I think, before the Scangauge, and with its information I seem to be doing 10% better or so. I imagine for the ordinary driver who has not yet made any changes for fuel economy that it could help you better your consumption by 20-30%.

    b. I do, perhaps, if you live nearby (who are you? :-D). Also, I'm assuming you drive a Prius? If you're looking for 15 kW on the highway, the simplified answer is that usually instant mpg will show at 75% of current mph.

  3. Nice write-up, and *very* nice results.

    I don't see kw on the SG -- I presume your have two, or switch between parameters ?

    -- SageBrush from PC

  4. Hi SageBrush,

    Thanks for stopping by and reading!

    kW has to be added as an XGauge. I keep it up most of the time these days, along with RPM, BTA, and fWT. It's rather slow to reload, but you start to get the hang of it without looking anyway. Here're the XGauge fields in order, taken from Hobbit's site:

    00 / 4000 8000 0000 / 0000 / 000A 001F 0000 / kw

  5. I just realized what you meant - in these pics I have AVG up, but I just have the one Scangauge, and typically don't use AVG, but rather kw in its place.

  6. 15kW is suggested for a Gen 2. Would you know the kW value for a Gen 3?