Taycan optimal driving speed chart

[tutis]

I've put together a simple model that calculates the optimal driving speed based on what is the achievable charging speed. The model takes as input the consumption of the car at multiple speeds and generates curve from there. The model can be used for any EV, provided that the inputs are adjusted

The inputs include:

1. battery net capacity
2. total distance of driving trip (as obviously if no charging stops are needed, then optimal driving speed is infinite)
3. and speed consumption at various speeds. For this model I have used the following, although this may vary for winter vs. summer, 22" vs 20" etc. you get the idea

Speed	Consumption
50 km/h	16 kWh/100kms
70 km/h	17 kWh/100kms
90 km/h	19 kWh/100kms
110 km/h	21 kWh/100kms
130 km/h	24 kWh/100kms
150 km/h	26 kWh/100kms

Here are the output curves for given average charging speeds:

The optimal speed for a given curve is where that curve peaks; e.g. if you can only achieve 11kW charging speed, then your optimal driving speed is around 100km/h. For 50kW it is around 175 km/h

For those who want to copy/edit the model it is accessible here

All inputs are in yellow boxes, and to re-run the model first delete everything in the H45:Z60 area, then click on Extensions > Whatif > Refresh Data Tables and wait until it is done working

Let me know if you have any comments or questions

 

[andix]

Nice work, thanks. Will look at that over the WE...

 

[B61]

The optimal speed for a given curve is where that curve peaks; e.g. if you can only achieve 11kW charging speed, then your optimal driving speed is around 100km/h. For 50kW it is around 175 km/h

I don’t understand how chaefing speed is related with diriving speed?

 

[tutis]

I don’t understand how chaefing speed is related with diriving speed?

the total time to travel a given distance is driving time + charging time. If you drive faster then the driving time reduces, but your consumption increases - thus you need to charge more frequently and longer (if the starting charge is insufficient). How much longer to charge is determined by your average charging speed available.

This model finds the minimum total travelling time (shown on y-axis) as the combination of driving time + charging time. That minimum travelling time is represented by the fastest average speed achieved.

there are other improvements I could build in such as overhead time to stop and plug in the charger etc, but those are second-order effects.

Hope that makes sense

 

[B61]

Got it, thx.

 

[Donlam]

Nice work. Thanks. Taycan's supreme charging speed so we could drive like stole it.

 

[ThePaddyWan]

Question, was this optimizer model targeting a fixed distance traveled over a given time period?

 

[McgR]

Looks nice. So basically according to your graph it makes sense to drive faster unless only slow chargers are available. Even with worse eco the travel time will be less.

most reliable to take the average charge speed for the entire charging session. That would we 150 - 175 kw. You won’t get 230 Kw for the entire session.

Another interesting thing I read somewhere. In cold weather driving faster can be more economic because less heating is needed because of earlier arrival.

and like mentioned above. What distance did you take in account? This will determine the number of charge stops and can make a big difference.

driving 120 vs 150 km/h in Germany results in a shorter travel time (higher average speed) in your graph but my personal experience is different. When driving higher speeds needs one extra charge stop it is really difficult to compensate the lost time.

 

[TDinDC]

Very interesting. As a visual person, I would prefer the legend on the right to be reversed in order so that they line up with the curves.

I think the only thing that would be even more useful as a practical standpoint would be to work in distance: The total distance you can go without charging depends upon speed (which is true regardless of battery characteristics).

When you have multiple stops, this becomes really interesting, because slowing down a bit to go from, say, three charging stops to two charging stops can have a big impact, particularly if the difference is only a little, and the impact is magnified if the chargers you have available to you are slower.

This would result in surges that are not smooth (more like steps), but might be even more useful

 

[lcarron]

I've put together a simple model that calculates the optimal driving speed based on what is the achievable charging speed. The model takes as input the consumption of the car at multiple speeds and generates curve from there. The model can be used for any EV, provided that the inputs are adjusted

The inputs include:

1. battery net capacity
2. total distance of driving trip (as obviously if no charging stops are needed, then optimal driving speed is infinite)
3. and speed consumption at various speeds. For this model I have used the following, although this may vary for winter vs. summer, 22" vs 20" etc. you get the idea

Speed	Consumption
50 km/h	16 kWh/100kms
70 km/h	17 kWh/100kms
90 km/h	19 kWh/100kms
110 km/h	21 kWh/100kms
130 km/h	24 kWh/100kms
150 km/h	26 kWh/100kms

Here are the output curves for given average charging speeds:

The optimal speed for a given curve is where that curve peaks; e.g. if you can only achieve 11kW charging speed, then your optimal driving speed is around 100km/h. For 50kW it is around 175 km/h

For those who want to copy/edit the model it is accessible here

All inputs are in yellow boxes, and to re-run the model first delete everything in the H45:Z60 area, then click on Extensions > Whatif > Refresh Data Tables and wait until it is done working

Let me know if you have any comments or questions

Great job. I think that the x-axis should be kilowatt not km/h.

 

[tutis]

Question, was this optimizer model targeting a fixed distance traveled over a given time period?

AND​

and like mentioned above. What distance did you take in account? This will determine the number of charge stops and can make a big difference.

The charts target an arbitrarily long distance (basically the limit function for a v long distance). In the spreadsheet, there is an approximation to calculate the average speed achieved for an arbitrary distance (e.g. 600kms) but that isn't part of the chart yet, I will add another chart and post it here. See this approximation in C33

For the arbitrary distance model, the model takes a simplistic approach and splits the travel over that distance into 2 segments; 1st segment is achieved on the initial charge, and the second segment is achieved at the best possible speed from the infinite distance calculation.

The true optimum would be a little more complicated, as it would keep the last segment of n segments in total to be charged just enough to get to the destination. I need to think about how to implement this in the model without turning this into a much more complicated algorithm

most reliable to take the average charge speed for the entire charging session. That would we 150 - 175 kw. You won’t get 230 Kw for the entire session.

Completely agree on average charging speed. One can also adjust the battery net capacity to use only 5%-85% when recharging as no one goes in with 0% battery and will recharge to 100% when trying to maximise average charging speed. This is another little improvement that could be made to the model, but this is likely 2nd order at best.

EDIT: the point about having 230kW charging speed was for me to assess what improvement to µ speed achieved we would get if Taycan v2 suddenly charged at 350kW peak, and averaged ~230 kW; the benefit seems very marginal at was is effectively a ceiling around 170km/h of real-world average achievable on the autobahn anyway. Even a petrol car cannot average 220km/h; it's nearly impossible because of traffic

driving 120 vs 150 km/h in Germany results in a shorter travel time (higher average speed) in your graph but my personal experience is different. When driving higher speeds needs one extra charge stop it is really difficult to compensate the lost time.

That's probably because I haven't calibrated properly the speed vs. consumption table which is at the heart of the model, we can make the consumption grow faster with speed and the curves will look different.
However, having driven a few times on the autobahn, I find empirically that the sweet spot is around 170 km/h to have the shortest travel time for c. 400 km travel distance, leaving with a full battery. So that matches somewhat my charts.

When you have multiple stops, this becomes really interesting, because slowing down a bit to go from, say, three charging stops to two charging stops can have a big impact, particularly if the difference is only a little, and the impact is magnified if the chargers you have available to you are slower.

This would result in surges that are not smooth (more like steps), but might be even more useful

See my reply to the quote in this same message as this one. Will think about how to build in multi-stops, it's not super simple

 

[WasserGKuehlt]

Will think about how to build in multi-stops, it's not super simple

You may end up with ANTSBRP (A New, Taycan-specific Better Route Planner). ;-)
Good stuff, keep it coming - although those of us based in the US will lose interest above 160 km/h. :-(

 

[DerekS]

TLDR: drive as fast as you want, pull over and charge when needed. Got it.

 

[bsclywilly]

If I'm ever stopped for speeding, I'm going to try to explain to the officer that it's because the superior charging rate of my Taycan that made me do it. And point to this thread.

 

[McgR]

Wouldn’t it be better to take 800-1000 km? With 400 km I think you will always have only one charge stop at any speed. At 50 kmh you can do just under 400 km and at 200 kmh probably a bit more than 200 km.