Hacker News Reader

Comment by lebuffon | Hacker News Reader

Electric commercial vans are such an obvious use case with real benefits. I am amazed it took so long. Spoke to a UPS driver who complained about his truck having a ~150 mile range on propane. That's a crime. With regen braking in a stop and start job it's electric for the win. There might even be the possibility to extend the range a percent or two with solar on that big van roof. It's not much, but try extracting gasoline from the sky.

View on HN

Regen braking isn't yet good enough for most "stop-start" applications. It turns out that unless you make big adapations to your driving style, typical braking generates too much energy too quickly to be able to charge the batteries.

We'd need a breakthrough in battery technology, or some improved supercapacitors, or productionization of flywheels, to be able to take advantage of that energy.

I'm surprised. I haven't studied this in detail, of course, but regen braking seems to work fine on my Leaf. There is a(n annoyingly) noticeable transition from regen to mechanical braking. The vast majority of the time, it is using regen.

You may be getting stopping power, but the issue is whether you're stopping gradually enough for a meaningful amount of that power to actually be captured in the battery vs. dumped to the environment as heat. I notice this on my electric longboard as well— longer, more gradual brake actions are way, way better for range than stopping more aggressively.

I think to get the power from more sudden stops, you'd need to have some kind of intermediate storage (like a capacitor bank) that could hold that energy for 20-30 seconds while it's streamed at the appropriate rate back into the main battery.

I've seen a few homemade electric cars from around fifteen years ago that combined batteries with a bank of ultracapacitors for exactly that purpose.

That's neat— I wonder if the gains are simply too small for it to make sense in a mass-market vehicle. Either way, seems like it would be a slam dunk for a delivery truck, city bus, garbage truck, basically any heavy vehicle with a stop-start workflow.

I think it's more that the capacitors are too expensive for mass production - the figures I saw put them at around $10K, for something that contributes a relatively small amount to the performance of the vehicle. Manufacturers would probably prefer to put that towards a bigger battery pack, which would give better overall specs, if a bit less efficiency.

You don't need ultracapacitors, regular ordinary teslas starting with the 2012 model S have had this. It's not quite a decade old.

Do you have a source on that? This article, for example, on the Tesla website is clearly aimed at a lay audience, but nonetheless mentions nothing about an intermediate capacitor bank for absorbing spikes of regen braking energy:

https://www.tesla.com/en_CA/blog/magic-tesla-roadster-regene...

I don't get your question. But that is a roadster, not an S or later. But it presumably doesn't mention intermediate capacitors because they don't use them.

I think you and mikepurvis are both saying the same thing - that Tesla and other electric cars are using batteries only, not capacitors.

I will note that the homebuilt electric cars generally had much smaller battery packs, which could not supply the same peak currents that a 50+kWh battery would have.

Sudden stops are not the majority of my stops nor do I suspect that is the case with delivery trucks. In sudden stops, I care more about immediate safety and less about energy recovery, anyway. I would be very surprised if the battery is not soaking up most of the energy under normal conditions.

How much kinetic energy is your Leaf dissipating compared to a loaded delivery truck?

The energy is roughly proportional to weight, as I imagine the electric motor and battery amperage are.

Perhaps relevant

https://www.youtube.com/watch?v=KQ2Eo6wl5r0

I believe that if people drove exactly the samemas before, they wouldn't get full benefit of regen, but I don't believe people will drive the same as before. I know I don't. Driving without ever using the brakes in my Model 3 is smoother and a bit of a fun mini game. One quickly learns not to accelerate quite as much when there's a visible red light ahead and how much the Regen brakes will slow the car down.

I find my experience to be exactly the same with the couple EVs I've owned. You very quickly adapt to learning exactly when you need to let off the accelerator, and even going from EV-to-EV where the regen curve is different, it only takes a couple tries to adapt to the new vehicle again.

The e-Pedal in the Nissan Leaf is wonderful, if you lift your foot off of the accelerator completely the car will apply optimal regen until the car stops.

Its really freeing not to have to look at the regen gauge at all, I just keep my foot over the brake pedal just in case I need to stop quicker than the regen braking.

What if you’re going down a slight decline where you could normally coast and maintain your speed. Do you have to leave your foot on the pedal and waste battery?

It's not wasting battery if you have your pedal in the position where it's coasting, neither regenerating nor accelerating.

Its really intuitive, if you are going a little too fast you just let up the pressure on the pedal. You don't have to think about if it needs to use battery, regen or coast to maintain the speed you want. In that aspect its like using cruise control in other hybrids/ev.

If you want to coast, you have to apply slight pressure to the accelerator. It doesn't use battery. It's a little weird at first, but you get used to it in a couple days, then it's second nature.

a supercapacitor could suck up all that sudden power and send it right back out to get the vehicle moving again.

Say a 0-60 start needs 20KW for 10 seconds. That’s 200,000 joules? How far away are we from a set of capacitors that can store that?

0.5 * 1850 kg * ((60 mph)^2) = 665 483.056 joules

Supercaps are 28800 Joules per liter, so you'd be needing nearly 700 liters of them to store the energy of the car decelerating from 60 mph once.

700 liters is a lot of volume in a car where you already use a lot of the available volume for batteries - it's more than the trunk and frunk combined on a Model3.

That might still be doable for a delivrry van or a bus, even if it is heavier. Natural gas powered already have quite a big "hump" on top for the pressure botzles with gas, I could imagine putting the supercaps there and the regular batteries inside the chasis to keep the center of gravity low.

0-60 certainly is not the important metric for this application, perhaps aside from rural routes. Most residential delivery, you're maxing out at 30mph.

Correct, and given that the velocity is squared in the energy equation, this is an enormous difference.

The more interesting use case for supercaps is in the 60-0 case or in the case of delivery trucks, 30 or 15 to 0. The inefficiency is in capturing the braking energy. If you can use supercaps to capture this inrush then you can trickle it back in to the battery pack. You don't need the capacity to capture all the braking energy nor do you need supercaps to replace the battery. You just need enough buffer to improve your overall efficiency. In something like a delivery truck the battery/cap ratio could even be adjusted with modular packs depending on route. I'm imagining something similar to what Gogoro scooters use but with two kinds of pack.

I don't think commercial drivers are going to be paying attention to regen rates as they are trying to make 300 deliveries on their daily run.

For personal vehicles it makes a lot more sense, particularly when you realize using regen braking means less brake and battery wear so it helps lower maintenance. I do similar things with the fuel economy gauge in my car.

Good insight -- the incentives are on maximizing number of deliveries per unit time. I certainly hear (and sometime even see) UPS drivers gunning their giant diesel vans down my street only to hit a stop sign every block.

So the interesting question is can operating costs make their way into the incentive structure (theoretically it should be able to with good vehicle telemetry).

Whether it's worth it or not to adopt that incentive structure is another question: is the cost of fuel and maintenance greater or less than the marginal revenue from each additional package? Because you're right, currently the incentives are infamous for maximizing number of deliveries. However, the famous UPS-minimize-left-turns routing optimization story suggests operating costs do factor in as well.

> So the interesting question is can operating costs make their way into the incentive structure.

It would likely be easier to just increase vehicle efficiency in the form of better braking, lighter vehicles, higher battery density, etc.

Are you sure about that? I have a regen meter in my Toyota and am often able to get 15-30Wh out of a single braking from 50-70km/h to full stop.

Yeah, but for how long? How long does it take you to come to a full stop from 70kph? A few seconds? Even at 30KW, that's not a lot of juice. But parent's point still stands: a Prius weighs a hell of a lot less than a UPS truck, and if one is trying to counter-act the energy required to accelerate, then that truck is going to try and dump a lot more than 30KW into that battery.

And I'm pretty sure your meter is measuring KW, not Wh. Regardless, I question its accuracy in our Leaf. How thick is the cable taking the 30KW and putting it in the batter, when the on-board charger on that thing only puts out 3.3KW? And maybe the answer is, "as thick as your wrist, but it's very short"), dunno, but color me skeptical.

The onboard charger is going from 240 VAC to pack level. That's a somewhat secondary use case in terms of peak load, and has nothing to do with getting up to speed or regeneration, so they don't spend much resources on it.

The main inverter is going between motor voltage and pack voltage, and depending on the car can handle regeneration at upwards of 60 kW (Teslas) to 265 kW (Porsche Taycan).

The cable that is putting 30 kW into the battery? It's the same one that's designed to conduct 80-110 kW from battery to motor in your Leaf.

Here's a clip from a generation older than mine:

https://youtu.be/KAIzAlgHF2E?t=57s

It measures regenerated Wh. In my case it was in increments of 30Wh.

How thick is the cable taking the 30KW and putting it in the batter, when the on-board charger on that thing only puts out 3.3KW?

As thick as the one putting 100kW into the motor I guess, so at 300V - not very thick.

> color me skeptical.

Skeptical of what exactly?

Even if the benefits of regenerative braking were zero, electric vehicles still make sense for delivery vans. They have known ranges, usually less than 100 miles and the reduction fuel and maintenance costs will be huge. This is why there is so much interest in the Tesla semi and other similar tech where regenerative braking has nearly little effect on milage because the vehicles aren't stopping often.

But existing driving data suggests there will be some non-zero effect on efficiency through regenerative braking so I'm not sure what the point here is.

Yes, you can't regen 100% of the energy of a full stop from 70->0. But why would you want to?

You see the traffic slowing down, foot off the pedal, regen starts and the car slows down. Traffic gets going again, push the pedal.

If you drive a hybrid/PHEV/EV like you would an ICE, it won't be economical.

Regen does pretty well in my Tesla Model 3. I definitely get energy back. Tesla also did a reworking of regen software (in the motor controllers) in 2019 that boosted the fleet's low-speed regen capabilities all the way down to a stop. So it definitely is capturing energy. Also relevant is this great video about using a pickup truck to charge a Tesla, by towing it. They then compare how much energy the Tesla got from the pickup truck, and how much energy the pickup truck burns towing the Tesla. Watch for the very surprising conclusion!

https://www.youtube.com/watch?v=RaGVoB4Zn-Y

What does good enough mean? It might not be capable of recovering 100% of the energy. And it would be a wonderful thing if it could.

But that's not the only way to define good enough that might be relevant here.

Many electric vehicles have a greater range in city traffic than they have on the highway, which is the opposite of how it is for ICE vehicles. So that means the benefit is quite significant, enough to help with cost savings.

Also, stop/start driving creates a lot of wear and tear on brakes. Maintaining brakes is probably a significant part of operating costs. If regenerative braking can reduce those costs, even if it doesn't eliminate them, it could still be significant.

The Tesla Model 3 can charge at up to 250 KW on a 75KWh battery. That does drop significantly at higher charge levels, but the battery in a van like this should also be significantly higher and the period of time required is significantly shorter.

Does it really require a lot more than 300KW to stop a UPS truck on an average start-stop (traffic light or stop sign) situation?

In normal driving 100KW is plenty for regenerative breaking. 300KW = 400HP which is heavy acceleration for most cars below 70MPH.

Heavy acceleration... or a heavy vehicle slowly. F = MA and all that. Tesla's cars are all very heavy for their class though I'm not sure how they compare to a half loaded UPS truck. (On average, a truck is between fully loaded and empty, with emphasis on trying to improve the efficiency in the average case, with an empty truck as the degenerate case.)

My point is that 300 KW is ~400 HP. I can't imagine that these trucks are always stopping vastly more quickly than they accelerate, so this ought to be able to enough to get the majority of the energy, especially taking into account that the regenerative braking is not 100% efficient to begin with.

It just doesn't seem like the issue of being able to disappate the energy with regenerative brakes would be a real problem with a full EV delivery vehicle. I think the people pointing to it as an issue are thinking of either different battery tech or hybrid delivery vehicles that had much smaller batteries (and thus lower peak regen).

This weird claim that evs don't work out in practice because they can't regen is just the latest example of FUD or extreme misunderstanding of the market. Just like people said it was impossible to make a compelling ev, that you could never go a long way, that you could never charge it fast, this latest claim is just silly. Teslas since 2012 (S and later) have had great recharging ability when slowing down. It's not 100%, nothing can be.

UPS trucks have either 165 or 195 HP, so it’s still significantly faster than they accelerate. Even 18 wheelers generally have ~500 HP or less

That's likely true at freeway speeds, but I expect the typical UPS van spends most of it's time running at 20-30mph.

> Regen braking isn't yet good enough for most "stop-start" applications.

Is this something you have numbers on or just seat of the pants math? Most mileage numbers for electric cars are much better in city driving with start/ stop conditions so I'm a bit skeptical about your methods.

Anybody who has ever driven a Prius, or a Fusion, and looked at their regenerative braking dial will confirm this.

When you hit the brakes on a regenerative-braking hybrid, you will see a bar start to fill up. The bar is divided into two sections.

If you tap the brakes, the bar only fills up a little, and stays in the left section. That means that your regenerative brakes are slowing you.

If you slam the brakes, the bar fills up fully, and is way in the right section. That means that your friction brakes are slowing you.

If you drive 'normally' - like most other people drive, you will never keep the bar in the left section during braking. It will always be half-way in the right section. This means you are wasting ~half of your braking energy, as friction and heat.

If you drive very smoothly, with no sudden braking, you can keep the bar almost constantly in the left section. But this requires you to start braking a lot earlier than most drivers, when you see, say, a red light.

This isn't a bad thing, since smoother acceleration/braking reduces traffic jams, but it's not how most people drive.

In practice, I have observed that the difference between these two forms of driving (Aggressive braking + accelleration, smooth braking + accelleration) is the difference between ~43 MPG and ~48MPG.

Prius and Fusion are bad examples of how much power a BEV can draw back from regen breaking back into the battery as Prius and Fusion have rather pitiful battery sizes and electronics compared to a typical BEV battery at this point. Current BEVs with "fast charging" electronics in ideal conditions can pull in a lot of regen power even in "slam the brakes" scenarios. The remaining trade-off there now is no longer what the battery can pull off of the regen brakes but that in the case of an emergency braking, physical brakes are still safer than regen brakes.

Prius also can only handle regen above 3-5mph (or something like that. Not going to haul out the owners manual for HN)

Well, sure, the Prius does not handle high-torque situations well. You're not going to be pulling a pickup truck out of the mud with it. It's not a great working car - you can't even jump-start another vehicle with it. That's an edge case for most drivers, though.

> If you drive 'normally' - like most other people drive, you will never keep the bar in the left section during braking. It will always be half-way in the right section. This means you are wasting ~half of your braking energy, as friction and heat.

Regenerative braking improves efficiency even without careful driving, additionally you can drive carefully and improve your efficiency even more. This isn't the same as suggesting current regenerative braking isn't "good enough" for start stop applications.

Braking will always involve energy loss, that's just physics. Over time we can improve how much energy is regained during braking but again that it can be better does not prove that it's not worthwhile now.

> Regenerative braking improves efficiency even without careful driving, additionally you can drive carefully and improve your efficiency even more. This isn't the same as suggesting current regenerative braking isn't "good enough" for start stop applications.

I've never said it doesn't help. When the bar is in the right section, that means that your regenerative braking is operating at peak capacity, with extra friction braking applied on top.

> Braking will always involve energy loss, that's just physics. Over time we can improve how much energy is regained during braking but again that it can be better does not prove that it's not worthwhile now.

The energy loss of regenerative braking is some XY%. The energy loss of friction braking is 100%.

The optimal way to drive with regenerative braking is slow and steady braking. So that you are eating XY% energy loss in your braking energy, and not "Partially XY%, partially 100%". Most people drive in a way that would put them in the latter category.

As another poster mentioned, its quite possible for cars with $10,000 batteries to usefully absorb more energy through their regenerative braking.

Yes, more efficiency is possible.

I'm not really sure what your point is or where your comments fit into this conversation at this point.

I think the way my Bolt does regen braking in 'single pedal driving' mode is the way to do delivery trucks. In that mode you almost never use the brakes, it uses regen back-pressure from the motors to stop you. As you let off the accelerator it slows you down. The more you let off, the faster you slow down. If you're stop-starting down the street, I think you'd capture back quite a bit of the power from the little bit of accelerating you're doing between mailboxes or buildings.

In my BMW i3 I hardly ever use the brake pedal. But releasing the "go" pedal all the way seems to add braking to the regen. I know because my brakes are noisy at the moment.

This is not accurate about regen technology. Teslas have been having significant regen since their first model s in 2012. What does "not good enough mean"? You don't have to do anything really, you just drive, when you slow down you can have the car do more aggressive regen or less aggressive. Less aggressive gets you less charging. I've had mine on more aggressive charging for 8 years (wow, has it been that long since I got my S?). I actually traded in my first one for the awd model so I could take it to the ski area so my car is 5 years old, 3 years in the first one. Yeah yeah, first world problems, but the take away is regen works great, almost a million people are using it every day they drive (I used to use it every day, but mostly sit at home now).

I did some envelope math and I don't think we need improved supercapacitors, especially not for trucks: we just need a big bank of the ones we have.

All that's required is storage for ~35mph to 0, and you get 95% regen instead of 10%.

Can still shunt to the battery on long descents, but for UPS, this would be perfect.

DHL was so annoyed that they started producing their own electric vans in Germany (https://en.m.wikipedia.org/wiki/StreetScooter)

I remember the "driving shop" visiting our house in like 1980, it was an electric vehicle.

https://nl.wikipedia.org/wiki/SRV-wagen

> I am amazed it took so long.

It didn't really. The European and Chinese markets are further along:

- Peugeot e-Expert: https://insideevs.com/news/423332/peugeot-electric-expert/

- Toyota Proace: https://insideevs.com/news/425503/toyota-proace-electric-15-...

- Opel Vivaro-e: https://int-media.opel.com/en/04-28-vivaro-e

- UPS Vans: https://www.theguardian.com/business/2020/jan/29/uk-electric...

- Citroën ë-Jumpy: https://www.electrive.com/2020/05/12/citroe%cc%88n-e%cc%88-j...

Some US ones:

- Workhorse C-Series: https://insideevs.com/news/404214/workhorse-deliveries-c-ser...

- Ford Transit: https://edition.cnn.com/2020/03/03/cars/ford-electric-transi...

The Nissan e-NV200 is relatively popular around here (Madrid) too. By popular I mean "I see a couple of them moving packages every other day".

https://insideevs.com/nissan/e-nv200/

Regenerative braking isn't nearly as efficient on something like a UPS truck or semi, compared to your Leaf. They use hydraulic-based regenerative systems instead.

Why wouldn't it work like it does in a car? If you've never driven an EV, then you might not know that there are three modes of acceleration: power from batteries, regen to batteries, and braking from kinetic. If you don't activate the braking from kinetic but you activate regen to batteries instead, then hydraulic is never activated.

I guess in a semi you brake all wheels and braking just the front ones (where the motors are) wouldn't work?

Although with normal cars, most of the braking is done in the front anyway? And vans can be 4x4, electric motors are probably relatively affordable to put on all wheels, considering things like the Toyota SUVs with AWD system that has electric motor in the back instead of shaft from the engine.

There's a difference between emergency braking and gentle/standard braking. Emergency braking you obviously need physical brakes on every wheel. With gentle braking you can use just the drive wheels, wherever they are, since they can obviously slow the truck down just as quickly as they can speed the truck up, and generally that's all you need.

> Regenerative braking isn't nearly as efficient on something like a UPS truck or semi, compared to your Leaf.

I don't get the point here. If it's 40% effective on his LEAF and 20% effective on a delivery van, then it is. Anything that increases range or decreases the cost of the vehicle is a big win. Regenerative braking is one of many reasons why electric delivery vans make a ton of sense, it's not the only reason.

Workhorse a company that recently bought a closed factory from GM in Lordstown Ohio and got financing from GM to do so has had a concept for an electric commercial van advertised on their website for some time. They used to have a concept pickup truck as well. Believe they were hoping to get the contract to replace the postal services fleet before that got canceled.

Problem is they are a bit of a joke. The pickup they were marketing has an advertised range of like 80 miles and the commercial vans they still appear to be working towards have an advertised range of 100 miles. I think the strategy of going into a market Tesla isn't in yet is a loser if the reason you are doing it is because you can't compete with Tesla. Business isn't going to be anymore willing to accept a crappy product than a consumer is.

It makes a lot of sense from technology, environmental and maintenance standpoints, but unless they can get the price and charging speed down significantly then fleet owners will continue to choose gasoline vehicles.

Swappable batteries, like they have for electric warehouse forklifts, would likely help solve the charging speed issue... but it adds an infrastructure challenge. Someone like the Postal Service would be a good candidate since they have many "warehouses" where they could station battery chargers/swappers. AMZL, UPS, Fedex, etc would be more difficult since they tend to have fewer warehouses that are further from many of their delivery endpoints. It's not impossible, and they'll likely be forced to go electric at some point, but it will require additional infrastructure.

There's a little more going here. GM derives 50%+ of it's income from China at this point. In China the wealthy commonly ride around in vans w/ a servant driving. This work is intended for that market.

> This work is intended for that market.

I'm sure that market will buy them too, and that market is growing in the US too (mostly NYC and LA) but I don't think that's the intention. The article specifically says, "The GM electric van project is aimed at an important segment of the emerging EV market – commercial delivery vehicles."

It doesn't seem implausible. They're generally outdoors during both operation and storage so they'd get most of the day's sunlight. Though in cities they'd spend a lot of time in buildings' shadows so they'd probably get infrequent direct sunlight.

In pronciple you should be able to do that by extracting carbon dioxide from the atmosphere and combining it with hydrogen from electrolyzed water. It would just need a crazy ammount of energy to get the carbon dioxide (likely by fractional destilation of air) and hydrogen from water, not to mention making the hydrocarbon chains for gasoline.

Using compressed methane could be better (I have seen buses here in Czexh Republic using compressed natural gas, which is basically just methane, very often) or even skipping the carbon and using hydrogen, if you can store it safely and efficiently.

But indeed, if you can store the power directly instead of usinfpg it inefficiently to make fuel, thats the best.

The problem is that battery tech does not scale. To push a bigger, heavier vehicle, you need an even bigger, heavier battery. Its a diminishing returns scenario where to add a little bit more range, you need to add more battery, but the added weight of the battery all but cancels out the range that it adds.

Think of the progression. We have had electric scooters for a while already. We have just started to see electric cars going mainstream. We have seen a lot of interest in large electric vehicles like semi-trucks and commercial vans, but nothing viable yet.

EDIT: Another way of thinking about this problem: Why do Teslas have so much storage space? You could easily fill the frunk with batteries and still have competitive storage space. The reason they don't is because adding more battery really doesn't help.

A big empty frunk makes the car a lot safer and filling it with batteries would make it a lot more expensive.

I agree with your general point, but at the same time, the cost of batteries are getting exponentially cheaper. It's just a matter of time for all transport to be BEV, so it makes sense for companies to want to be ready for the new market realities.

s/frunk/trunk/

I don't think the price of batteries tells the whole story. Until the tech is viable, people are not going to buy an inferior car.

Most roads, at least in America, have weight limits per axle. A van or semi can only be so heavy, and having batteries take up half your cargo budget is a huge disincentive. This is on top of the severly limited range of a big, heavy, full-of-cargo EV.

frunk refers to the front trunk that Tesla's have, further increasing their caring capacity past traditional ICE cars.

I remember when the 60kWh Model S got its EPA rating - it was 208 miles vs 265 of the 85kWh version, so roughly a 11% gain in efficiency.

Bottom line: batteries do scale - at least in terms of weight.

I thought Tesla used the same physical battery pack but used software to limit how much you got.

In return you get way better battery life because there’s loads of excess capacity in the battery compared to nameplate capacity

In the very first edition that was the case only with the 40kWh version.

The 60kWh one was fairly popular and battery prices were much greater back then, so a physically different battery was justified.

Only later did the 60kWh version shrink enough in sales for Tesla to apply this trick to it.

The diminishing returns argument only holds for aircraft, where the battery has to constantly work to lift its own weight.

For ground transportation, the percentage of the weight dedicated to batteries is important, but so is the ratio of wind resistance to battery power.

For low speed start and stop traffic, it is true that if your car is already 90% battery by weight, adding battery won’t help. For freeway driving, this doesn’t hold at all, since most of the energy is being lost to wind resistance.

The reason teslas have so much storage space is that air doesn’t weigh much, and the empty space helps them design for aerodynamics. Filling the space with batteries would make the vehicle too expensive.

Aerodynamic drag dominates at high speeds, but there’s a friction component that’s influenced by weight.

Accelerating more mass takes more energy. While you can recover it with regen, you can never break even.

Yes, but battery vehicles are not that much heavier as ICE vehicles and the efficiency of the electric propulsion and the benefit of regeneration works very strongly in favor of the battery vehicles. While it is not released yet, the Tesla Semi even seems to make electric propulsion work for heavy semi trailers.

Reply on news.ycombinator.com