Better to charger higher or run battery lower?

[Allistah]

Sounds like you have TeslaFi. I guess for each charging level, you will need to leave it charging for some amount of time. I guess I would get TeslaFi's data on how many kWh you add for each charging interval (you'll presumably need to interrupt the charge). But you do need to understand exactly what TeslaFi is reporting from the API (there may be multiple fields for the amount of energy - one for energy added to the battery, another for the power out of the AC-DC chargers - they won't be the same!). It's important to understand what the numbers TeslaFi reports mean.

I expect that the Breaker amps and the amps reported by the car should basically be the same. But you'll have better resolution on your breaker amps probably.

You should record the voltage that the car sees in a column. This is going to change depending on your current setting.

It would be interesting to see the exact overhead amps from your current clamp when the car is sitting there not charging (since it will have better resolution than the in-car display), but that 2A (that's what I observe it to be when the door is open and the screen is on - so 500W) does not last long before the Wall Connector opens its relay. You also have to record the voltage at that 2A to figure out the static power. It is good info to have, but I guess you just need to record it once.

I don't think you need to subtract that 2A or whatever from your numbers, though - it's part of the overhead you're trying to determine. And the number you see on initially plugging in probably is not the same when the car is just sitting there in idle mode with the display off and just charging (that seems to be closer to 200W, just based on the energy consumption of idle mode from other data).

There are a lot of other details I'm not covering here. You'll probably have to gather data and then iterate on the method as you discover the shortcomings of the approach.

Ideally you have measurement of the voltage at the current clamp location and at the Wall Connector input, too, but probably not realistic.

I do have TeslaFi but I'm just going to take readings at the breaker with the clamp and use what the car is reporting to make this quicker since I have about 20 sets of data to record. I'm hoping that it will just give me an idea on where it's most efficient. I'll make sure that it charts the gross and net loss so if people want to see one or the other (or both) the data will be there. I will be measuring the voltage at the car as well. I thought about the voltage at the breaker but I don't have a way to be safe with that one that will make it easy so I just left it off the dataset.

 

[AlanSubie4Life]

I do have TeslaFi but I'm just going to take readings at the breaker with the clamp and use what the car is reporting to make this quicker since I have about 20 sets of data to record.

Yes, that is the challenge with this approach. The problem is your data about how much energy the battery received is going to be so low resolution if you use the number of miles added (and you can't use what the car displays on the screen for "mi/hr" - because that assumes that is correct, which I've found it is not, particularly). You do actually need three significant digits for the energy added to the battery (via the API, presumably) to actually get a good estimate here. So to do that WITHOUT the API, at a minimum you'd need to do 100 miles of charging or so for each setting. That would take a while!

The API would probably allow you to do 10-15 minute charge intervals (might need to be longer for the lower current settings) with some accuracy. But it still might not be good enough.

You're looking for just a few % efficiency change here so the accuracy of the measurements has to be pretty good.

 

[Allistah]

Yes, that is the challenge with this approach. The problem is your data about how much energy the battery received is going to be so low resolution if you use the number of miles added (and you can't use what the car displays on the screen for "mi/hr" - because that assumes that is correct, which I've found it is not, particularly). You do actually need three significant digits for the energy added to the battery (via the API, presumably) to actually get a good estimate here. So to do that WITHOUT the API, at a minimum you'd need to do 100 miles of charging or so for each setting. That would take a while!

I see what you mean. What I'm trying to measure is not what you're referring to, but rather when I charge at a 30A rate and the car reports it's charging at 30A, what is the actual current at the breaker?

For example, if I set the car to charge at 30A and it says it's actually pulling 30A. I measure at the breaker and it may read 33A which would be a ~9% gross loss. I'm going to take those measurements from 2A all the way to 40A in 2A increments and see at what point that loss is the smallest. Doing this so that I gain an insight to a sweet spot of where the most efficient charging rate is. Some say it is going to be 32A due to some components in the car that are most efficient at that rate. <shrug> I guess I will find out tonight when I get all this info and look at the data.

 

[AlanSubie4Life]

For example, if I set the car to charge at 30A and it says it's actually pulling 30A. I measure at the breaker and it may read 33A which would be a ~9% gross loss. I'm going to take those measurements from 2A all the way to 40A in 2A increments and see at what point that loss is the smallest. Doing this so that I gain an insight to a sweet spot of where the most efficient charging rate is. Some say it is going to be 32A due to some components in the car that are most efficient at that rate. <shrug> I guess I will find out tonight when I get all this info and look at the data.

I don't think this is what you'll find from your data, but as you say, you will find out tonight! I would suspect your car and the breaker clamp should record currents within about 1A of each other (probably within the resolution of the in-car display). But we will see.

What you are looking for is a much simpler measurement, but I don't expect it will tell you anything about what the most efficient charging rate is. That can only be determined by the ratio of how much energy goes into the battery to how much comes from the wall, plotted vs. the current limit you set in the car.

 

[Allistah]

I don't think this is what you'll find from your data, but as you say, you will find out tonight! I would suspect your car and the breaker clamp should record currents within about 1A of each other (probably within the resolution of the in-car display). But we will see.

What you are looking for is a much simpler measurement, but I don't expect it will tell you anything about what the most efficient charging rate is. That can only be determined by how much energy goes into the battery vs. how much comes from the wall, plotted vs. the current limit you set in the car.

Yeah, I suppose I could do all this and the loss will be flat across the board and it could tell me nothing. I guess I'd rather know nothing than to wonder if it's nothing. lol

 

[Tam]

I'm having a difficult time understanding the last line there. kWh is a measure of watts over one hour of time. The part that is confusing me is in the scenario of the car. If I set the car to a 30A draw and going from 90-100% takes 2 hours, but going from 50-60% only takes 1 hour, the amount of power consumed (and thus the amount I would pay for) would be double on the 90-100 portion. Thats right, right?

At home you pay for what the electric meter says and not by the clock.

I can use an 1,000 watt iron for 1 hour that moves the electric meter 1kWh.

I can also use a 100 watt electric bulb for 10 hours that moves the electric meter 1kWh.

Whether 1 hour or 10 hours that moves the electric meter for 1kWh, that's the kWh I pay, not the time duration.

But as @AlanSubie4Life said, that's false because I ignore the inefficiency while charging: I can get a 1kWh charge into my battery but it may take 1.200 kWh to do that. It's an inefficiency loss that I didn't take into the account.

So, if there's an inefficency loss of 200 watts per hour, I don't want to additionally prolong that to 10 hours (200watts x 10 hours = 2kWh loss).

 

[Tam]

...going from 90-100% takes 2 hours, but going from 50-60% only takes 1 hour...

From 50 to 60% is about 7.5 kWh out of 75kW battery.

From 90 to 100% is also about 7.5 kWh out of 75kW battery.

So if we don't talk about the loss for inefficiency, the amount we pay at home is 7.5kWh no matter how fast or how long it takes.

However, if we take inefficiency into the account, I doubt there's a huge difference.

You may save a few cents for each session to minimize the inefficiency but at the expense of battery health.

Put it in another way, would I:

1) Charge 90% and run it down to 30% (which cost a few cents more due to inefficiency)

or

2) Charge 70% and run it down to 10% (so I can save a few cents due to more efficiency)?

My answer would be: Avoid the lower State of Charge and choose number 1!

The Rules of Model S Road Tripping

 

[Allistah]

I’ve only done a 100% charge once so far. I went back and looked at that charge on Teslafi and the current was constant at 20A until it was near 100% and slowly tapered off the current draw for the last ~20 minutes of the charge. Huh, I would have expected a longer taper at the end but who am I? Lol

 

[Tam]

Here's my data from my 20A charging:

Charging from 46 to 70% took 4:03. It added 17.49kWh but I have to pay for 18.8 kWh or 93.1% efficiency and costs $2.44 (SCE Prime rate = 0.13/kWh).

Charging from 75 to 100% took 4:28. It added 16.1kWh but I have to pay for 20.8 kWh or 77.4% efficiency and costs $2.70.

That's a difference of 26 cents in saving which I don't think it is worth running down my battery low.

 

[darth_vad3r]

In some states that prohibit electric reselling, Tesla would charge you my minutes and not kWh. In this scenario, quicker make sense.

However, I assume you are talking about home charging. Utilities charge you by kWh and not by the time duration.

So, the last 1 kWh may take 10 hours or 1 minute at home, I still only have to pay for 1kWh.

Right, technically the "last 1 kWh" (that you draw from the wall) will cost you the same, no matter how long it takes to use that, but very important point is:

This incorrect, due to overhead of about 200W (this is approximate, don't assume it is exactly correct) when actively charging.

For your 10 hours (extremely unlikely) example, that would be 3kWh to add 1kWh, vs. ~1kWh for your (impossible since it would require 60kW AC) 1-minute example.

So your "last 1 kWh" taken in 1 hour has a higher percentage of energy delivered to the battery where you want it, as opposed to the same 1 kWh drawn from the wall spread over 10 hours, which in this example would actually not charge the battery due to overhead losses but you get the point.

 

[darth_vad3r]

Here's my data from my 20A charging:

Charging from 46 to 70% took 4:03. It added 17.49kWh but I have to pay for 18.8 kWh or 93.1% efficiency and costs $2.44 (SCE Prime rate = 0.13/kWh).

Charging from 75 to 100% took 4:28. It added 16.1kWh but I have to pay for 20.8 kWh or 77.4% efficiency and costs $2.70.

That's a difference of 26 cents in saving which I don't think it is worth running down my battery low.

Good data, thanks!

If cost alone were the only issue ya, no sweat ... but I think avoiding "low battery" isn't worth going up to 100% unless "low battery" is extremely low and in danger of totally running out. i.e. 20% or lower is less important than avoiding 90% or higher for battery health reasons.

I'd say 90% to 20% usage is fine, and I wouldn't charge to 100% only because I wanted to drain down to 30% instead of 20%.

If your low-end SoC is borderline "I might run out" then sure, charge to 100%.

 

[darth_vad3r]

I’ve only done a 100% charge once so far. I went back and looked at that charge on Teslafi and the current was constant at 20A until it was near 100% and slowly tapered off the current draw for the last ~20 minutes of the charge. Huh, I would have expected a longer taper at the end but who am I? Lol

Some L2 charge tapers just got posted in another thread after the question got raised and I asked for some input.
See these 2 posts:
newbie TM3 SR is doing only 3kWh on L2
newbie TM3 SR is doing only 3kWh on L2

The first chart (from @wwu123) shows SoC ticking up from 96% to 97% when the taper started from 240V/32A. To me that means "96.5%" SoC for 7.7 kW charging is where the taper starts.
The second chart is harder to read precise info, but the poster (@srs5694) says the taper started when TeslaFi showed "precisely 96%", but from looking at the chart's 100% line and going back (left) 3 dips to what seems to be 97%, it looks to me like the taper starts when the 96 ticked up to 97.

Also 32A, so anyways ... safe to assume 32A charging tapers around 96.5% (maybe 96.0%).

How "near 100%" was your "near"? 96%? ... I don't think anyone will see L2 taper below 96%, and then depending on how high power you are charging at you can start to see it at around 96.0-96.5%. If you are charging at 3kW or lower, or at L1, you might never see taper until the display shows you "100%" and you are just trickling that last 1/2 percent from 99.5% to as much as it can cram into the pack

 

[Allistah]

Got all the data.. Seems like it's a bit more efficient at around 12A charge rate, but after that things flatten out so...it doesn't really matter. But it was fun to check out and to learn. I went back and double checked the rates at 12A and they're accurate and wasn't a fluke. Not sure what that's about.. but it doesn't matter because I won't ever charge that low anyways. I guess I'll charge at different rates and see what I can learn from TeslaFi.

 

[darth_vad3r]

Got all the data.. Seems like it's a bit more efficient at around 12A charge rate, but after that things flatten out so...it doesn't really matter. But it was fun to check out and to learn. I went back and double checked the rates at 12A and they're accurate and wasn't a fluke. Not sure what that's about.. but it doesn't matter because I won't ever charge that low anyways. I guess I'll charge at different rates and see what I can learn from TeslaFi.

Nice.

You are saying "more efficient" in terms of lost current only? (I think your efficiency number is net current loss divided by breaker current?) ... but you need to consider that there is lost energy in the car as well (before making it into the battery).

 

[Allistah]

Some L2 charge tapers just got posted in another thread after the question got raised and I asked for some input.
See these 2 posts:
newbie TM3 SR is doing only 3kWh on L2
newbie TM3 SR is doing only 3kWh on L2

The first chart (from @wwu123) shows SoC ticking up from 96% to 97% when the taper started from 240V/32A. To me that means "96.5%" SoC for 7.7 kW charging is where the taper starts.
The second chart is harder to read precise info, but the poster (@srs5694) says the taper started when TeslaFi showed "precisely 96%", but from looking at the chart's 100% line and going back (left) 3 dips to what seems to be 97%, it looks to me like the taper starts when the 96 ticked up to 97.

Also 32A, so anyways ... safe to assume 32A charging tapers around 96.5% (maybe 96.0%).

How "near 100%" was your "near"? 96%? ... I don't think anyone will see L2 taper below 96%, and then depending on how high power you are charging at you can start to see it at around 96.0-96.5%. If you are charging at 3kW or lower, or at L1, you might never see taper until the display shows you "100%" and you are just trickling that last 1/2 percent from 99.5% to as much as it can cram into the pack

I'll take a look at those threads. Here is the log entry from the ramp down... and the ramp down only lasted ~20 mins before the charging was done.

 

[darth_vad3r]

e.g. if there is 200-250 W usage by the car in total overhead (car "on" overhead, charger losses, battery charging inefficiencies), then there's about say 1A of current 'lost' inside the car. In terms of overall wall-to-battery efficiency, you want to include that.

 

[Allistah]

Another chart from TeslaFi showing the tapering off as it reached 100% SoC.

 

[Allistah]

With the AC off and such, the car appeared to pull about .08 Amps which is about 19.12 watts. The screen was on and that was it so it sounds reasonable.

It seems with my setup, between the breaker and the car, there is about 6.4% loss with anything above a 14A charge rate. 14A or 40A, it's about ~6.4% loss.

So from a power out of the breaker perspective, the charge rate doesn't seem to matter much in regards to "loss" since it's pretty consistent. Taking this a step further, would be interesting to split the hairs even further and see if I can get more data out of TeslaFi somehow.

 

[darth_vad3r]

Another chart from TeslaFi showing the tapering off...

View attachment 438655

On that chart, the taper starts 3/4 of the way along the 2nd highest blue battery level line. Is that the 99% line? I think it is.

If so I'd guess that's about 3/4 of the way between 98.5 and 99.5, or about 99.25% (based on the theory that the SoC display battery_level that the API reports to TeslaFi is rounded to nearest integer and the tick up from 98 to 99 is at 98.5%, and same with 99 to 100 being at 99.5%.

That was at a 20A charge, so 4.8 kW ... so we are looking at a taper where ~5 kW is allowed at 99%. If you are charging at 7.7 kW or 11 kW you'd see it taper a bit earlier, but still well into the 90's.

 

[darth_vad3r]

With the AC off and such, the car appeared to pull about .08 Amps which is about 19.12 watts. The screen was on and that was it so it sounds reasonable.

It seems with my setup, between the breaker and the car, there is about 6.4% loss with anything above a 14A charge rate. 14A or 40A, it's about ~6.4% loss.

So from a power out of the breaker perspective, the charge rate doesn't seem to matter much in regards to "loss" since it's pretty consistent. Taking this a step further, would be interesting to split the hairs even further and see if I can get more data out of TeslaFi somehow.

I'm not sure it will pull any extra shore power while not actually actively charging... so that 0.08 A is not an accurate gauge of the total consumption by the car while not charging.

We know that Sentry Mode consumes ~1 mi/hr ... so that's in the 200-300 Wh range.