Temperature vs. Energy Use

bolosky · Apr 20, 2010

It's spring here in Seattle, and that means warmer temperatures. With that, I've noticed substantially improved energy efficiency in my Roadster.

On my drive in to work this morning (temp about 60F, about a 50/50 mixture of freeway and surface streets), I used just under one ideal mile per mile driven. During the winter when it was typically 25-30 degrees cooler, this never happened. I don't remember the exact numbers, but I think it was something like 20% worse (1.2 ideal miles/mile driven), and it may well have depended on how cold it was. I generally considered the ideal miles range to be a joke, and often saw 280-300 Wh/mile.

I've heard lots of stories about how Li-ion batteries don't hold as much charge when they're cold. However, that's not the car is reporting. It's saying that the battery capacity is about the same, but that the rest of the car is using more energy. I don't think that running the heat is sufficient to explain the difference (though it's a pretty big drain). So, either the car really is using more energy when it's cold for some reason that I don't understand, or else the VDS is reporting lower battery capacity as higher energy use.

Have other people had this experience? Does anyone have an explanation?

doug · Apr 20, 2010

Besides the cabin heat, there also might be energy used to regulate the battery pack temperature.

Would be interesting to see a graph of efficiency versus temperature. Data which I suppose the car could log.

TEG · Apr 20, 2010

Lots of other factors too...

Check these for instance:

9 reasons why your winter fuel economy bites! - MetroMPG.com

...
2. Low tire pressure
Of course you're smart enough to keep up your tire pressure as the temperature drops, right? A 10-degree (F) change in ambient temperature equates to a 1 psi change in tire pressure (source). Fuel economy declines 0.4 percent for every 1 psi drop (source).
3. Increased rolling resistance
Even if you're completely attentive to proper tire pressure, cold ambient temperatures will still cause your tires to return worse mileage. That's because a tire's shape isn't completely round - the sidewall bulges out at the bottom, and where the tread meets the road the small contact patch is actually flat. As the tire rotates, it constantly deforms to this shape, and this deformation requires more energy when the rubber is cold and hard. Rolling resistance at 0 degrees F is 20% greater than at 80 degrees (source 1, source 2).
4. Crappy road conditions
It's increased rolling resistance of another kind: driving through slush and snow. And then there's its wasteful polar (no pun intended) opposite: no friction at all! (A.K.A. wheelspin on ice.)
... 6. Higher average lubricant viscosity
Engine oil thickens as it cools. So does transmission and differential fluids and even bearing grease. Significantly more energy is needed to overcome the added drag these cold lubricants cause.
Using synthetic fluids can address this problem, since their viscosity changes less at extreme temperatures than traditional mineral fluids.
... 8. Higher electrical loads
In colder temps, you use electrical accessories more often:
- lights (in higher lattitudes it's darker in the winter)
- rear window defroster (because it's easier than using the ice scraper, right?)
- heater blower motor (I don't have a/c, so this isn't balanced out during warm conditions); heated seats/mirrors
- windshield washer pump (because it's easier than using the ice scraper, right? And for frequently cleaning off dirty road spray.)
9. More aerodynamic drag
No, I'm not referring to the layer of snow you're too lazy to brush off the top of the car (though that would hurt mpg too).
A vehicle�s aerodynamic drag is proportional to air density, and the density increases as temperature drops. For every 10 degree F drop in temperature, aerodynamic drag increases by 2% (source).

My NiMH RangerEV gets better range in summer, and my friend's Prius gets better MPG in summer, so this isn't just a Tesla or Li-Ion effect.

AndrewBissell · Apr 20, 2010

Hi Bolosky,

Having just taken delivery of a Roadster here in the UK, I was finding that in my driving I was between 1.25 and 1.5 ideal miles per mile driven. Temperatures here are currently around 15 C which must be close to 60 F.

Your stats are so much better - is there something wrong with my driving, or something wrong with my car? Maybe just first-week over-enthusiasm! But then again a 2.1 mile roundtrip on surface streets (all under 35 mph) used over 300 Wh/mile. Does this seem high, all you experienced Roadster owners?

Bolosky, can you give some typical idea of your speeds and acceleration on your commute?

Thanks

Andrew

bolosky · Apr 20, 2010

AndrewBissell said:
Bolosky, can you give some typical idea of your speeds and acceleration on your commute?

My commute's about 17 miles. Roughly 10 miles of that is on freeway and the remainder's on surface streets. Today, I did a little bit of hard accel near my house (showing 480 Wh/mile at 0.7 miles out), then got on the freeway gently, drove a few miles and got off to deposit my son at school. When I got back on, I did a hard acceleration to about 60 mph, just because it was fun (if you're anything like me, you'll find this hard to resist most of the time).

The surface streets are mostly 30-35 mph with a small amount of 20 mph near schools, but with only a little stop and go.

There's a moderate amount of up and down on my trip, but the net elevation change is at most a couple of hundred feet, so I doubt that matters too much.

How are you measuring your energy efficiency? I suggest not using the information screen that comes up when you engage the parking brake. What it calls "net energy" is actually the total energy out of the battery (what I think of as gross energy). To take into account the effect of regen, you need to subtract that number from the displayed "net" number.

Instead, I did my calculation in two different ways: I reset the trip counter at the beginning of my trip and used its special screen (which is off by default, you have to go through setup to turn it on); and, I added miles travelled to remaining ideal miles, and compared that to the 190 that I had when I left the garage in the morning.

AndrewBissell · Apr 20, 2010

I am measuring miles and ideal miles the same way. Thanks for your inputs.

tennis_trs · Apr 20, 2010

AndrewBissell said:
...
I was finding that in my driving I was between 1.25 and 1.5 ideal miles per mile driven.
...
Does this seem high, all you experienced Roadster owners?
...

Having the top off and windows down makes a significant difference, along with speed and acceleration (at 60F I'd have my top off, but I know that's a little cool for many people, especially at highway speeds).

On my longest trip so far (151 miles, nearly all on highways, with the temperature about 75F most of the time), I estimated that I averaged (I started with a full standard mode charge):

1.41 ideal miles per mile driven for the first 96 miles,
in standard mode,
top off, windows down,
speed mostly 60-80 miles/h.

1 ideal mile per mile driven for the last 55 miles,
in range mode,
top on windows up,
speed mostly under 65 miles/h [I put the top on, changed to range mode and stayed slower since I drove more than I had expected and didn't want to chance getting near running out of charge].
I had 27 miles (both ideal and estimated) left when I got home.

I don't have any specific numbers, but I believe that I've also seen tire pressure make a fairly noticeable difference.

scott451 · Apr 20, 2010

Using ideal miles is a bad idea.

tennis_trs said:
1.41 ideal miles per mile driven for the first 96 miles,
in standard mode,
top off, windows down,
speed mostly 60-80 miles/h.

1 ideal mile per mile driven for the last 55 miles,
in range mode,
top on windows up,
speed mostly under 65 miles/h

Using ideal/estamated miles is the not the best way to measure how the car is performing in a specific configuration. The problem is that the battery system is constanly re-calculating the estimated miles based on how the car was driven in the last 30+ miles. I don't know for sure, but there may also be some variation to ideal miles based on how the car was driven.

The correct way to measure the energy required, is to use the trip screen and reset it before each test. The Wh/mile will accurately reflect the total energy used from the battery during the test. It will also allow you to calculate average speed (dist/time), so you can confirm that the speed was reasonably constant between the two test cases.

Also, changing the mode has an impact to the Wh/mile. In Standard mode, the Battery is cooled when it rises above ~30C. This happens alot when you are driving +65mph. In range mode the battery can be as high as 40C. additionally, the motor and PEM are allowed to run hotter. (you can confirm this for yourself, after a spirited drive in standard mode, park the car, but leave it on, and switch the car to range mode. notice that all the fans and AC stop, switch back to standard and they resume)

-Scott

tennis_trs · Apr 21, 2010

scott451 said:
Using ideal/estamated miles is the not the best way to measure how the car is performing in a specific configuration. ...

-Scott

Thanks for the info. I mostly made my post to point out that, based on my experience, the 1.25-1.5 numbers from Andrew probably aren't unreasonable.

My way of looking at the Roadster info is more useful for me since my main concern is total miles I can reasonably expect to get out of the car on future trips. But the method you point out would be more helpful for running "tests", during my normal driving, to know how to best stretch out the miles on future trips when needed.

I thought that the ideal miles were just based on estimated usable available energy in the battery, regardless of how the car has been driven recently.

ChargeIt! · Apr 21, 2010

tennis_trs said:
I thought that the ideal miles were just based on estimated usable available energy in the battery, regardless of how the car has been driven recently.

That's how I understand it too. And that "usable available energy" is probably based on an "across-the-bricks" calculation/measurement of average real-time voltage in the pack. Then divide that left-over capacity by approx 221Wh/mile (54kWh/244 EPA miles) to get the remaining Ideal Miles. But I do believe the firmware starts being more conservative (or could it be "accurate"?) nearer the bottom of the pack.

AndrewBissell · Apr 21, 2010

Thanks to everyone for your inputs.

I tried being very gentle with the car today - 60 mph with cruise control on fast dual carriageway (slow for the UK but acceptable); gentle acceleration (hard to do!); using cruise even on slower roads in 30 and 40 mph zones.

Result was 279 Wh/mile overall, and 1.17 ideal miles/real mile.

I also observed that on cruise at 60 the kW needle tends to sit just above 12.5 kW. Call it 13-14 kW, suggesting that you can cruise for an hour at 60, thus doing 60 miles for around 13-14 kWh. Multiply by four and you get 52-56 kWh! So a full range mode charge could allow you to cruise at 60 for close to four hours i.e. 240 miles.

My working hypothesis is that (practically speaking) acceleration chews through battery energy and regen recuperates a quite small proportion of that. Maybe it's obvious really: the kW meter goes to 200 on the "spending energy" side, but just 40 on the "recovering energy" side. Also rear wheel electric drive will be less effective at regenerative braking than front-wheel drive would be (due to change in effective "weight distribution" forward a you decelerate).

I am therefore assuming that steady speed is quite efficient compared with the kind of repeated acceleration/deceleration you get in both city driving and spirited country road driving.

I'm doing my first Range Mode charge tonight and longish drive (110-120 miles return, mostly motorway) tomorrow, so hopefully I'll get a good chance to test this all further.

Andrew

PS To stay vaguely on topic today's results were again in conditions around 15C (60F) with a mix of cloud and sun, which caused me to use the soft-top outbound at freeway speed in the morning and come back on slower roads with the top open in glorious sunshine (and easily hearing the sound of birdsong - how fabulous in a car!) on the way back.

Tdave · Apr 21, 2010

Andrew,

kW meter goes to 200 on acceleration because it can do 0.9 g's when accelerating(!), but during regen will only do -0.2 g's. Do the math and you'll see the regen recaptures energy pretty well (more than half). The big energy losses are due to air resistance at high speed, and at low speed the fixed energy use systems in the car. Also regen should be no more or less effective with fwd or rwd or due to weight transfer. Those effects just aren't going to be a factor.

The big factors will be, in this order I believe:
• speed (air resistance)
• top on/off, windows up/down (air resistance)
• a/c and heat usage
• ambient air temperature (what this thread is about)

AndrewBissell · Apr 21, 2010

Tdave,

Thanks for some good inputs.

Tdave said:
kW meter goes to 200 on acceleration because it can do 0.9 g's when accelerating(!), but during regen will only do -0.2 g's. Do the math and you'll see the regen recaptures energy pretty well (more than half).

Yes, I see the point and had considered that. However it's clearly (even if > half) still considerably less than 100%. Lets guess 75%. Accumulate that over many accelerate, decelerate cycles - let's say 10: 0.75^10 = 0.056. So after 10 cycles you lost over 94% of the energy you originally put in.

I think therefore that a constant speed wastes much less energy than a sequence of accel/decel cycles leading to the same average speed.

Btw - 0.9 g! Haven't seen that yet - 0.66 g or so is my max so far. Maybe that is a Sport vs standard Roadster difference.

The big energy losses are due to air resistance at high speed, and at low speed the fixed energy use systems in the car. Also regen should be no more or less effective with fwd or rwd or due to weight transfer. Those effects just aren't going to be a factor.

Thanks for your thoughts. On the areas that drive energy use, I think you are right on the money.

I have some reservations regarding the RWD/FWD regen point. Elsewhere I have read what seemed like fairly convincing analysis on this. I'm going to think on this part more.

Andrew

Tdave · Apr 21, 2010

AndrewBissell said:
Yes, I see the point and had considered that. However it's clearly (even if > half) still considerably less than 100%. Lets guess 75%. Accumulate that over many accelerate, decelerate cycles - let's say 10: 0.75^10 = 0.056. So after 10 cycles you lost over 94% of the energy you originally put in.

Yes, that is true. That's offset somewhat by the less air resistance you incur during time spent at the slower speed. During stop and go traffic in rush hour where I vary between 10 - 40 mph continually, I use considerably less energy than steady 60 mph, I've noticed.

AndrewBissell said:
I have some reservations regarding the RWD/FWD regen point. Elsewhere I have read what seemed like fairly convincing analysis on this. I'm going to think on this part more.

Please prove me wrong. I like being educated when I'm wrong.

vfx · Apr 21, 2010

scott451 said:
...
Also, changing the mode has an impact to the Wh/mile. In Standard mode, the Battery is cooled when it rises above ~30C. This happens alot when you are driving +65mph. In range mode the battery can be as high as 40C. additionally, the motor and PEM are allowed to run hotter...

Try running 35MPH in hot weather! Everything gets hot.

Bradleybang · Apr 21, 2010

When cruising in the 55-75 mph try drafting other cars. On freeways (motorways for right hand drivers) other cars using cruise control will not mind you getting on their tail. Some have actually waived at me after a 20 minute ride behind them. They notice and after a couple lane changes they get your purpose.

T@oo · Apr 21, 2010

AndrewBissell said:
I have some reservations regarding the RWD/FWD regen point. Elsewhere I have read what seemed like fairly convincing analysis on this. I'm going to think on this part more.

Andrew

I think what he was trying to imply is, that the regen efficiency not dependend whether it is done FWD or RWD. But due to weight distribution during braking going forward, a FWD (or AWD) would be able to do a harsher regen. Maybe up to beeing able to regen even a full emergency stop. On RWD, regen reaches the tire slip limit fairly soon esp. in bad weather conditions.

Summary: The amount of kinetic energy that is regen-able is a lot higher for FWD or AWD than RWD. I expect the Model S to be able to regen more than the Roadster, because of higher weight and wider rear tires (spec say 295 right now ...). Despite still beeing RWD.

TEG · Apr 21, 2010

Just keep in mind that a vehicle with a strong rear weight bias (such as the Roadster with the battery pack in back) retains more traction on the rear wheels than a vehicle with more weight on the front wheels. The limitation of regen being rear wheel only is less dramatic on the Roadster since it has such a strong rear weight bias.

T@oo · Apr 22, 2010

TEG said:
The limitation of regen being rear wheel only is less dramatic on the Roadster since it has such a strong rear weight bias.

That helps, I'm sure. I'm really curious how this bias changes to front while doing a hard full stop. Maybe I can dig out a formular for this.

bolosky · Apr 22, 2010

T@oo said:
I think what he was trying to imply is, that the regen efficiency not dependend whether it is done FWD or RWD. But due to weight distribution during braking going forward, a FWD (or AWD) would be able to do a harsher regen. Maybe up to beeing able to regen even a full emergency stop. On RWD, regen reaches the tire slip limit fairly soon esp. in bad weather conditions.

Summary: The amount of kinetic energy that is regen-able is a lot higher for FWD or AWD than RWD. I expect the Model S to be able to regen more than the Roadster, because of higher weight and wider rear tires (spec say 295 right now ...). Despite still beeing RWD.

I think that the real limit on regen isn't the tire traction, it's the rate at which the battery can accept charge. This doesn't seem to be anywhere near the limit of the tires. Recall that as far as the tires are concerned, speeding up and slowing down at the same rate and initial speed requires applying the same amount of power (ignoring air drag), and the regen power is way lower than the accel power. Both only go through the rear wheels.

Also, I disagree with the statement that "The amount of kinetic energy that is regen-able is a lot higher for FWD or AWD than RWD." I think you're confusing power (rate of energy recapture) and energy (total watt hours put back into the battery) here. If the limit really was the tires rather than the battery, then you'd be able to put it in faster with RWD, but it for a shorter amount of time (because you'd be stopped).

Temperature vs. Energy Use

Member

Administrator / Head Moderator

Teslafanatic

Member

Member

Member

2010 2.0 Roadster Sport

KWH-PWR#1349Sprt,S Sig#96

2010 2.0 Roadster Sport

Member

Member

Member

Member

Member

Well-Known Member

Member

Member

Teslafanatic

Member

Member

Similar threads