The air compressor comment was an illustrative tougue-in-cheek comparison, not a serious suggestion.
Imagine the amount of power needed to run an air compressor with a displacement in the 5-6 liter range. That’s exactly the amount of power we’re talking about. Because that’s literally what an engine-braking truck is doing, it’s driving its engine as an air compressor.
Your “tens of thousands of watts” estimate is probably just about right. And that’s not anywhere outside the realm of doable. That’s probably well within the abilities of regenerative system in the drivetrain this vehicle will have, but even if we assume it isn’t, a 10kw resistive load is a $100-$200 part, off the shelf.
Sports cars dissipate several hundred kilowatts in their friction brakes (for reference, the Porsche Taycan which can almost do with only regen braking can regen ~270kW). A hard-braking truck will exceed this significantly, but of course distributed over many more brake discs.
However, for the trailer scenario, I assume that if you go up a certain slope using N kilowatt of propulsion to maintain a stable speed, you'd need somewhere in the ballpack of N/2 killowatt of braking power when going down during the full duration (unlike hard braking, which is only for a few seconds).
Also, peak braking performance is much different than effective regen potential, since you shouldn't need to do hard braking very often.
There are 3 things that limit regenerative braking in its braking capacity at normal speeds:
1. Battery capacity, as you mention. Mostly a concern if you started high, as you mention.
2. Battery charge rate (thermal and lifetime concerns), as you're within or exceeding fast-charge charging rates. Especially important as the battery is likely already operating hot from pulling the load uphill. To give an idea of battery wear, note that a Tesla Model S only allows you to fast-charge a fixed amount on a given battery before you are permanently locked out to not further deteriorate the battery.
3. Charge capacity from the motor controller, which limits total regenerative braking capacity.
The first two are unique to batteries, and become an issue with continuous regenerative braking (such as a long downhill slope with a heavy trailer). Number 2 is likely to be the biggest issue.
Optimal regenerative braking sinks take whatever you throw at them: Either a brake resistor for rheostatic braking, or a connection to the grid which from the perspective of the vehicle is an approximated load of infinite size. Rheostatic braking is only limited by cooling of the brake resistor, which can both operate much hotter than a battery and is much easier to cool.
So why do cars not have brake resistors? Because normal vehicles do not have problems with excessive regenerative braking. Even going downhill, their weight is unlikely to cause severe battery load (although it may fully charge). However, gravity is a bitch when your total weight exceeds 10 tons.
> Also, peak braking performance is much different than effective regen potential, since you shouldn't need to do hard braking very often.
Exactly. The reason I mentioned this is that you noted peak brake numbers, which have no meaning in relation to continuous load capacity, which is much, much lower.
A resonable auxillary air compressor won't make a difference. With a heavy trailer going down hill, you'd need to at least dissipate tens of kilowatts, maybe even touching triple digits if you also need to slow down.
Nowhere to dissipate the power → no regenerative braking.