What If the Harvester Shed 1,000 lbs and Got Better? The Case for LTO + EA211-ERV

[ritterf]

e90 M3 / Jeep LJ Rubi owner · Preorder #3933623624 · replacing an X5 with this

I've been sitting on this for a while. Already sent it to Scout directly — got a very polished "thanks, noted" from support. So I'm bringing it here, where the people who actually care about this stuff are.

The Harvester has a real shot at being the most capable EREV ever built. But I think the current architecture direction is about to repeat the same mistake every other EREV has made. Here's what I mean — and here's the fix.

---

THE PROBLEM: EREVs ARE JUST BLOATED EVs WITH A GENERATOR BOLTED ON

Every current EREV design makes you haul 1,200–1,500 lbs of low-C-rate NMC or LFP just to provide range. That weight spiral kills towing dynamics, destroys payload headroom, and wrecks off-road agility. It's engineering compromise stacked on engineering compromise.

And the cruel irony — you're carrying all that pack weight specifically so the generator doesn't have to work as hard. The tail is wagging the dog.

---

THE PROPOSAL: A HIGH-RATE BUFFER ARCHITECTURE

Instead of a 80–100 kWh pack, pair two modular 10 kWh LTO (Lithium Titanate) packs with a purpose-built turbo generator. Total buffer: 20 kWh. Total pack weight: ~350 lbs vs ~1,400 lbs conventional.

For the generator: the EA211-ERV — the exact engine VW just put into production for the ID. Era 9X. 1.5L, Variable Turbine Geometry, Miller cycle, ~105 kW sustained output. Real production engine. Not a concept. Already validated in a large-platform application.

[ INSERT: Architecture diagram ]

---

WHY TWO 10 kWh PACKS INSTEAD OF ONE 20 kWh PACK

This is the part I'm most excited about. Two modular packs changes everything:

Modularity across the lineup. Run one pack in the lighter Traveler build. Run both in the Harvester towing config. Same skateboard platform, two different buyers. Scout sells both without redesigning the chassis.

Serviceability. Replace one pack at a time if one degrades. ~$4,500 instead of $9,000+. A dealership swaps one in an afternoon. This is the crate engine philosophy applied to batteries.

Weight distribution. Two smaller packs gives chassis engineers real flexibility — one under each axle for 50/50 balance. Try that with a 1,400 lb monolithic slab.

VW group platform play. Here's the part that should get a VP's attention: a validated 10 kWh LTO module proven in worst-case truck duty is a group-wide asset. Audi, Porsche, Cupra, VW ID. Buzz — every EREV in the portfolio gets lighter, more durable, more serviceable batteries. Scout doesn't just ask VW for an engine. Scout hands VW the business case for their next battery standard.

"Scout isn't asking VW for parts. Scout is the validation platform for VW's next battery architecture."

---

THE REAL-WORLD CASE: HIGH ALTITUDE MOUNTAIN PASS TOWING

This is where the architecture earns its money. The EA211-ERV's VTG turbo compensates for altitude power loss that kills naturally aspirated engines. The LTO buffer absorbs full regen on descents without thermal gating — consistent, predictable braking feel all the way down. And on the climb back up, the buffer delivers burst power while the generator sustains.

At 10,000 ft towing 5,000 lbs, a conventional EREV with a weak NA generator hits turtle mode. This architecture doesn't. The generator never stops. The buffer never gates. The truck never quits.

---

WHY LTO WINS THE LONG GAME

[ INSERT: Cycle life / cost math graphic ]


20,000+ verified charge cycles — Toshiba SCiB production cells. At one full cycle per day that's 54 years. The pack outlasts the chassis by decades.

Full KERS capture. LTO absorbs 100% of hard regen energy on a mountain descent without thermal gating. NMC and LFP gate regen when the pack gets warm — you feel it as inconsistent brake response. LTO doesn't. Ever.

Long term replacement cost. When your 100 kWh NMC pack degrades in year 8–10, you're facing a $15,000–20,000 replacement that may not even be available. With two 10 kWh LTO packs doing 20,000 cycles — that conversation essentially disappears.

Tire life and payload. 1,050 lbs off the chassis is real tire wear reduction over a 15–20 year ownership cycle. Multiple sets of tires. On a truck people actually keep, that math matters.

The cost math works. LTO is ~3–4x per kWh vs NMC — but at 20 kWh vs 100 kWh you're buying a fraction of the cells. The delta funds a better generator, better suspension, and still comes out ahead.

---

THE 30-YEAR TRUCK

Scout's heritage pitch is durability. A Harvester with this powertrain is the first EV-based truck that actually backs that up with physics. The battery outlasts the chassis. The generator is a production VW engine with a global parts supply. There's no $20,000 pack replacement hanging over the owner in year 10.

That's what "legendary Scout durability" means in 2025. Not a marketing line. An engineering decision.

"Two 10 kWh LTO packs + EA211-ERV = the first EREV that drives like a truck, lasts like a Land Cruiser, and costs like one too."

I'm not an engineer. I'm a preorder customer who's done the homework. Curious what the technically-minded people in here think — especially anyone who's worked with LTO chemistry or knows the EA211-ERV spec sheet better than I do.

---
Ritter Friedrich · Preorder #3933623624
e90 M3 / Jeep LJ Rubi / replacing an X5

 

[ritterf]

I don't know if this will be as much of an issue as you suggest. Or at least that the other battery supporting more charging rates would increase actual range in stop and go traffic that much.

I say that because my Hyundai Ioniq 9 has a screen where you can see exactly how much energy the regen (or accelerator) is absorbing/using, in real time. And the highest number I can recall seeing for regen, was just the other day. And it was 86kw of regen.

I was coming downhill, at around 50mph. I had it in level 3 regen mode, and lifted off the accelerator. It was a pretty aggressive slow down, as Level 3 regen is the second highest mode (one away from one pedal driving, and the real difference is it won't come to a complete stop, the actual regen level is the same I believe). And again that peaked at 86kw. That same drive, coming down a freeway offramp, I saw a peak of ~68kw.

I don’t think a maximum of 140kw regen rate, is going to prevent any meaningful amount of regen into the battery in normal day to day driving. Just because to reach levels higher than that, you have to be very aggressively come to a stop, which is not something people do regularly with passengers in the vehicle, just because its sort of "neck snapping" deceleration that is pretty disruptive to passengers.

In panic braking situations, on a race track , or coming down mountain passes while towing (and actually having to come to a stop, rather than just slowing down/maintaining speed), sure, I could see that. I just don't think thats a regular occurrence.

For context, 400kw of regen is quite high. That’s 533hp of “reverse thrust”. Think about accelerating in a vehicle with 533hp (even a heavy one). Thats stronger than most people brake regularly. And while 140kw is "only 186hp", I'd wager that is still enough to capture the large majority of braking energy.

All that aside:

I'm curious about the temperature statement. Because, even if the max recharge rate was 140kw... you'd think the system would be designed to handle that during DC fast charging, even at high temperatures. Because if this is an 800v system, I'm sure they're targeting more (ie, faster) than 140kw of charging, and it needs to be able to do this in Arizona summers.

Looking at LFP battery packs of similar size, it seems like most of them peak at between 150-170kw of charging speed (The Tesla Model 3 LFP is 170kw). So maybe the 140kw was chosen as a more average power?

 

[ritterf]

Really appreciate the real-world data — that's exactly the kind of ground truth that matters more than theoretical specs.

Your 86kW at 50mph downhill makes sense — that's a lot of kinetic energy to recover. But here's the thing: kinetic energy scales with velocity squared. At 20mph stop-and-go you have about 16% of the kinetic energy you had at 50mph. So the regen potential per stop in city traffic is much lower — maybe 10-15kW per event, well within what both LFP and NTO can handle without gating.

So you're right that 140kW LFP max regen isn't the constraint in normal city driving. I'll concede that point.

Where NTO wins in city driving isn't peak regen rate — it's two other things. First, the lighter truck means less kinetic energy to deal with at every stop and less energy needed to accelerate back to speed. Second, NTO can use 90-100% of its SOC window vs LFP's 80-20% conservative window — meaning the same 23kWh of NTO is more usable capacity than it looks on paper.

The mountain pass and towing case is where the high C-rate really matters — 166kW of sustained demand means the buffer needs to absorb and release hard, repeatedly, for an extended period. That's where LFP gates and NTO doesn't.

Your Ioniq 9 data point actually helps refine the argument — city driving the chemistry difference is modest, towing/mountain duty is where it's decisive.

 

[ritterf]

So you are saying all the required components stay with the battery pack/module?
Sorry - I wrote software and don’t thoroughly understand hardware…

Good question — not all the components stay with the pack, but the voltage architecture shapes everything connected to it.

Think of it like a towing package on a truck. Bigger radiator, bigger brakes, heavier duty transmission cooler — you're adjusting components based on the build requirements, but they're just components. Same basic truck, tuned for the duty cycle.

Same idea here. The battery pack contains the cells and BMS. But the inverter, motor controllers, and charging circuitry all have to be rated for whatever voltage the pack runs at. An 800V system needs 800V-rated components throughout the entire high-voltage chain from pack to motors — more expensive, more complex to service everywhere.

At 460V you're in the mature, cost-optimized range of power electronics. The same voltage class used in industrial drives, forklifts, transit buses — proven, globally serviceable parts. You tune the components to the voltage, just like you tune the brakes to the tow rating.

The other consideration — Scout's mechanics will be servicing these trucks for 20-30 years. 460V high-voltage service is routine for a well-trained technician. 800V adds specialized tooling and safety protocols. On a truck marketed for longevity that matters.

Lower voltage makes every component in the chain simpler, cheaper, and easier to service. The pack and the rest of the truck benefit together.

And on the logistics question of running 800V for the Traveler and 460V for the Harvester — VW already does this. MEB platform is 400V. PPE platform — Taycan, e-tron GT — is 800V. Two separate high-voltage architectures, same group, same supplier base. A 460V Harvester is a third point on a spectrum they already span. And VW's existing 400V component base is close enough to 460V that there's meaningful supplier and tooling overlap. It's not starting from scratch — it's a tuned variant of infrastructure they already run.

 

[Logan]

Really appreciate the real-world data — that's exactly the kind of ground truth that matters more than theoretical specs.

Your 86kW at 50mph downhill makes sense — that's a lot of kinetic energy to recover. But here's the thing: kinetic energy scales with velocity squared. At 20mph stop-and-go you have about 16% of the kinetic energy you had at 50mph. So the regen potential per stop in city traffic is much lower — maybe 10-15kW per event, well within what both LFP and NTO can handle without gating.

So you're right that 140kW LFP max regen isn't the constraint in normal city driving. I'll concede that point.

Where NTO wins in city driving isn't peak regen rate — it's two other things. First, the lighter truck means less kinetic energy to deal with at every stop and less energy needed to accelerate back to speed. Second, NTO can use 90-100% of its SOC window vs LFP's 80-20% conservative window — meaning the same 23kWh of NTO is more usable capacity than it looks on paper.

The mountain pass and towing case is where the high C-rate really matters — 166kW of sustained demand means the buffer needs to absorb and release hard, repeatedly, for an extended period. That's where LFP gates and NTO doesn't.

Your Ioniq 9 data point actually helps refine the argument — city driving the chemistry difference is modest, towing/mountain duty is where it's decisive.

To be clear, I'm really not trying to shout down the proposal/idea, and hope its not coming off that way. I'm just trying to be realistic about what the actual pros-cons are of the other things that you're proposing (where I'm at a handicap, as a non-expert on any of these chemistries). So in this case, I had recent (literally on Wednesday) data, so figured i'd share. And yeah, the knetic energy goes WAYYY up with speed. I'll be curious to see what regen looks like on the at least vaguely similarly sized/weighted Ioniq 9 (or at least in the same ballpark of weight for the Scout, at ~6k lbs), the first time I go down a mountain pass.

I do hope that Scout Engineers are working their hardest to make the solution work the best they can, with the constraints that they have available.

A few other thoughts from this post (and these are open questions, not arguments):

• What happens when the NTO battery fills up on a mountain pass?
  • As you mention, the higher C rate is definitely a benefit on a mountain pass, at least for the peak recovery rate.
  • But if the pack is smaller, and it fills up faster, would it being full actually be better?
    • Doing some quick math, the 23kwh NTO battery would fill from 0-100% in
      • ~3.5min at 400kw
      • ~7min at 200kw
      • ~8min 20 sec at 166kw
    • None of those times are outlandish. You can totally come down a mountain pass at ~60-75mph at those speeds, for that long. So it seems like you "could" actually fill the battery completely, and still have some left to go.
  • the larger battery of an LFP pack, would allow it to be more consistent during these passes (ie, absorb more), which I think would make it easier on the braking system of the vehicle, and could result in more energy being recaptured?
• I know that most PHEV's don't really charge their battery to 100% when they show "100%", because they are frequently kept at 100% and that would be more damaging
  • I wonder if scout will do the same, partially alleviating the "80-20%" topic
  • But to your point, those buffers would both be good from a battery life perspective, but bad from a weight perspective.

Happy weekend all!

 

[ritterf]

Not coming off that way at all — this is exactly the kind of pushback that sharpens the argument. Really appreciate it.

Your mountain pass math is right — a 23kWh pack at 166kW regen input fills in about 8 minutes. On a long descent that's a real scenario worth thinking through.

but like any EV when the battery is full the mechanical brakes take over. So ideally you don't start the decent with full charge..
would be cool if GPS recognized a long decent upcoming and the generator would let the battery get much lower.. like 5% before a decent.
since the LTO doesn't care about low volts or high volts. you can play with SOC more.

On a large-pack EREV or BEV, a full battery mid-descent means the BMS gates regen and you're back on friction brakes — that energy is gone.
and the LFP heats up a lot more on a long decent the regen rate drops

So the sequence on a long pass looks like: climb with generator running + buffer discharging → summit → descent begins → generator cuts back → NTO absorbs regen at high C-rate → pack fills → generator stays off → bottom of descent with a full pack ready for the next climb. That's the loop working as designed.

On the PHEV buffer point — really sharp observation. You're right that most PHEVs hold back 10-15% at each end to protect chemistry. NTO's advantage is that buffer can be much tighter — maybe 5% each end — because the chemistry is stable and doesn't plate lithium at full SOC under fast charge. More usable capacity from the same nominal size than the spec sheet suggests.

On your last point — yeah, I genuinely hope Scout engineers are looking at this. The EA211-ERV exists, the NTO cells are sampling now, the architecture is coherent. Whether the timing works for a 2026 launch is a different question. But the conversation is worth having before the specs freeze.


Happy weekend — this has been the most useful technical exchange in the thread.


My hope is there is an engineer somewhere over there having this discussions with the accountants and marketing team..
and he can say... see the people care... they want Heritage quality not Fleet cycle longevity.

 

[Jkew]

is it that bad being an AI bot? Do tell us all about it.
But as to what you are eluding too..

Yeah I use AI to help research and check my math. So does every engineer with a calculator. The numbers are either right or they're not — happy to be corrected on specifics.

With respect, I use AI nearly constantly now at work. Using these tools well means knowing when to stop, reevaluate and refine the underlying assumptions that generated the output. LLMs are very useful, but at their core they are stochastic parrots. Once you have established a context window which is aligned with particular regions of its model it’s hard to shake it loose from its self-reinforcing feedback loop. This is honestly something I fight with every day.

What many people have pointed out is that your underlying assumptions about product market fit may be… biased, and I suspect self-reinforcing. This not intended to be a dig, it’s just the reality of how these tools are different from calculators.

I would intentionally play devils advocate with the AI and think of ten questions which are antagonistic to its current conclusions. Proceed to ask them in sequence, and ask it if its conclusions are now different about product market fit.

You will, in all probability (literally), see a shift in its conclusions and even its math.

 

[Logan]

I had to go run to the store today to grab some fertilizer (pre-emergent ftw).

And on the trip, I decided to check the regen some more, for kicks and giggles. The short version is that I tried some relatively aggressive deceleration moments (not stopping, as I was on a 2 lane road, I just didn’t have anyone following me). It describe it as a step away from panic braking, but generally a “forceful slow down”. The type that would wake a sleeping spouse on a road trip and have them say “whoa, what was that, what happened?!?”

I did this a few times, including the spot I saw the 86kw from the other day. The highest rate I saw was 141kw.

And I paid more attention in typical average braking situations too. I’d say most of those on normal 25-45mph roads, were in the realm of ~25-50kw.

Not that relevant, but, figured I’d share.

 

[ritterf]

With respect, I use AI nearly constantly now at work. Using these tools well means knowing when to stop, reevaluate and refine the underlying assumptions that generated the output. LLMs are very useful, but at their core they are stochastic parrots. Once you have established a context window which is aligned with particular regions of its model it’s hard to shake it loose from its self-reinforcing feedback loop. This is honestly something I fight with every day.

What many people have pointed out is that your underlying assumptions about product market fit may be… biased, and I suspect self-reinforcing. This not intended to be a dig, it’s just the reality of how these tools are different from calculators.

I would intentionally play devils advocate with the AI and think of ten questions which are antagonistic to its current conclusions. Proceed to ask them in sequence, and ask it if its conclusions are now different about product market fit.

You will, in all probability (literally), see a shift in its conclusions and even its math.

That is mostly what I do, I argue with it, most of it seems really dumb, that's why you need to bounce them off each other.
I bounce between 3 of them. check their sources. see if it makes sense. Push back constantly.

these things aren't coming up with this itself.
I read reports ask it questions, have it check grammar and do sanity checks on math.

 

[ritterf]

I had to go run to the store today to grab some fertilizer (pre-emergent ftw).

And on the trip, I decided to check the regen some more, for kicks and giggles. The short version is that I tried some relatively aggressive deceleration moments (not stopping, as I was on a 2 lane road, I just didn’t have anyone following me). It describe it as a step away from panic braking, but generally a “forceful slow down”. The type that would wake a sleeping spouse on a road trip and have them say “whoa, what was that, what happened?!?”

I did this a few times, including the spot I saw the 86kw from the other day. The highest rate I saw was 141kw.

And I paid more attention in typical average braking situations too. I’d say most of those on normal 25-45mph roads, were in the realm of ~25-50kw.

Not that relevant, but, figured I’d share.

it has a lot to do with pack temps..
if you are at 40% with a cold battery it will suck up a lot.. but if you are low or high SOC and the battery is warm, it will Gate the regen down to like 20% of available. That's when the mechanical brakes are used.
the system is being conservative with fragile batteries.

 

[ritterf]

With respect, I use AI nearly constantly now at work. Using these tools well means knowing when to stop, reevaluate and refine the underlying assumptions that generated the output. LLMs are very useful, but at their core they are stochastic parrots. Once you have established a context window which is aligned with particular regions of its model it’s hard to shake it loose from its self-reinforcing feedback loop. This is honestly something I fight with every day.

What many people have pointed out is that your underlying assumptions about product market fit may be… biased, and I suspect self-reinforcing. This not intended to be a dig, it’s just the reality of how these tools are different from calculators.

I would intentionally play devils advocate with the AI and think of ten questions which are antagonistic to its current conclusions. Proceed to ask them in sequence, and ask it if its conclusions are now different about product market fit.

You will, in all probability (literally), see a shift in its conclusions and even its math.

Did you stress test anything I said? Did you get any different conclusions?
at one point it Did use some old numbers I didn't catch... but we corrected that..

They will always make mistakes, the human job is to catch them and challenge them on it.
Prove their logic.

I don't get why you would want to carry around 100 miles of extra battery for 90% of your trips, when you have the range extender in a series hybrid.. the added battery weight ruins the whole concept.

 

[ritterf]

To be clear, I'm really not trying to shout down the proposal/idea, and hope its not coming off that way. I'm just trying to be realistic about what the actual pros-cons are of the other things that you're proposing (where I'm at a handicap, as a non-expert on any of these chemistries). So in this case, I had recent (literally on Wednesday) data, so figured i'd share. And yeah, the knetic energy goes WAYYY up with speed. I'll be curious to see what regen looks like on the at least vaguely similarly sized/weighted Ioniq 9 (or at least in the same ballpark of weight for the Scout, at ~6k lbs), the first time I go down a mountain pass.

I do hope that Scout Engineers are working their hardest to make the solution work the best they can, with the constraints that they have available.

A few other thoughts from this post (and these are open questions, not arguments):

• What happens when the NTO battery fills up on a mountain pass?
  • As you mention, the higher C rate is definitely a benefit on a mountain pass, at least for the peak recovery rate.
  • But if the pack is smaller, and it fills up faster, would it being full actually be better?
    • Doing some quick math, the 23kwh NTO battery would fill from 0-100% in
      • ~3.5min at 400kw
      • ~7min at 200kw
      • ~8min 20 sec at 166kw
    • None of those times are outlandish. You can totally come down a mountain pass at ~60-75mph at those speeds, for that long. So it seems like you "could" actually fill the battery completely, and still have some left to go.
  • the larger battery of an LFP pack, would allow it to be more consistent during these passes (ie, absorb more), which I think would make it easier on the braking system of the vehicle, and could result in more energy being recaptured?
• I know that most PHEV's don't really charge their battery to 100% when they show "100%", because they are frequently kept at 100% and that would be more damaging
  • I wonder if scout will do the same, partially alleviating the "80-20%" topic
  • But to your point, those buffers would both be good from a battery life perspective, but bad from a weight perspective.

Happy weekend all!

Holy crap.. the battery in the Hyundai Ioniq 9 is 1400lbs..
the max Regen they report is 150kw

Ok.. lets compare 150kw braking from 60mph.. to 400kw you could do from 23Kwh of NTO..

 

[Logan]

Holy crap.. the battery in the Hyundai Ioniq 9 is 1400lbs..
the max Regen they report is 150kw

Ok.. lets compare 150kw braking from 60mph.. to 400kw you could do from 23Kwh of NTO..

View attachment 14519

View attachment 14520

View attachment 14521

Did you lookup the pack weight of the Ioniq 9, or did you estimate it using ai math? And where did you find the max possible regen? I haven’t looked it up, but I would have guess it would have been closer to 100% of the rated motor output. So that’s interesting to learn about.

Mostly curious, because it sounds about right, but you also call it an LFP battery pack, which it is not (it’s NMC). But the Ioniq only has 303hp, so it is relatively low on power compared to many others, but I would be surprised if it was because of the battery chemistry with such a large battery. I always thought it was just because they didn’t care about the maximum power output race (mine isn’t a rocket ship, but it’s plenty enough for my needs, with a low 6 second 0-60).

People don’t realize how heavy battery packs are.

For the record, the Hummer EV battery pack just about 3000lbs. The Kia EV9 100kwh pack is 1250lbs.

 

[ritterf]

Did you lookup the pack weight of the Ioniq 9, or did you estimate it using ai math? And where did you find the max possible regen? I haven’t looked it up, but I would have guess it would have been closer to 100% of the rated motor output. So that’s interesting to learn about.

Mostly curious, because it sounds about right, but you also call it an LFP battery pack, which it is not (it’s NMC). But the Ioniq only has 303hp, so it is relatively low on power compared to many others, but I would be surprised if it was because of the battery chemistry with such a large battery. I always thought it was just because they didn’t care about the maximum power output race (mine isn’t a rocket ship, but it’s plenty enough for my needs, with a low 6 second 0-60).

People don’t realize how heavy battery packs are.

For the record, the Hummer EV battery pack just about 3000lbs. The Kia EV9 100kwh pack is 1250lbs.

well they didn't publish the ioniq 9 but they did release the specs on the smaller battery in the Kia EV9 which is 99.8 KWh and 1248 lbs..
the Ioniq 9 is 110.3... so... scaled it up...
totally could be wrong.. but they were supposed to be similar construction.

Charge and discharge are two very different thing when it comes to batteries.. usually can discharge 2x what it can charge..

the 150kw number I got from a google search..
"The system stops the car completely using friction brakes.
Drivers can also adjust the regenerative braking strength when lifting off the accelerator. There are three levels of regenerative braking to choose from, allowing you to customize the driving experience to your preference.

The maximum regenerative braking power is 150 kW."


your right... I used 70kwh of LFP for the braking scenario while you have NMC... but the calculations were based off the 150w max regen.

Hyundai IONIQ 9 AWD Performance – Full Specs, Features, Range & Comparison | EVKX.net

Hyundai IONIQ 9 AWD Performance is an all-electric SUV offering a WLTP range of up to 570 km (354 miles). EVKX.net has all the details you need.

evkx.net