Some parts of the London underground use passive energy recovery by locating stations nearer to the surface than most of the tunnel between them. Trains start by rolling downhill and when they approach a station, uphill.
> If the line is unreceptive, braking energy is dissipated in on-board resistors
How many watts are dumped into onboard heat-generating resistors on the trains in the most heat-affected lines per week?
Should regenerative braking be disabled in aboveground trains when heat impacts reach uncomfortable levels in belowground tunnels?
> Regenerated braking energy is transmitted to the London Underground high voltage distribution network
If regenerative braking oversupply is inducing higher temperatures belowground through on-train resistors, then only an operational change to aboveground mode would be required to minimize that induced heating during times of thermal need.
(Obviously longer-term solutions with non-zero capital expenditure exist that could be pursued in parallel.)
> Should regenerative braking be disabled in aboveground trains when heat impacts reach uncomfortable levels in belowground tunnels?
It’s not that simple. You can’t just treat the entire line as some kind of perfect conductor that allows to you move unlimited amounts of energy around. In reality there’s issues with both the conductive capabilities of the lines themselves, but there’s also the simple problem that train lines aren’t generally electrically connected end-to-end for a few reasons.
1. You don’t want trains pulling power down more of the line than necessary, when it’s more efficient to draw power from other parts of the high voltage grid.
2. You don’t want a single track fault to cause your entire line to be forced to disconnect completely.
3. You don’t want your line to accidentally become a power carrier for electrical grid, just because the two ends of your line a physically located far enough apart that they experience different grid conditions.
As a result most train lines are broken up into electrically isolated segments, each with its own distinct power supply. So you could turn of regen on overground trains, but unless they happen to be sharing a section of track with underground trains, it doesn’t create any additional capacity to dump breaking energy into.
Even with isolated segments, the measurements performed with Southern found 10% of regenerative power was turned into heat as the line voltage was already at max threshold from other trains; if that holds true today, then there’s still opportunity to consider optimizations for ensuring that 10% is weighted towards aboveground rather than below.
How exactly are you going to “weight” that 10%? Each segment is independent, you can’t “weight” the voltages, because that would require the ability to move power between track segments, which you can’t do, because then they wouldn’t be track segments anymore.
Additionally you’re making the assumption that line voltages are stable and constant, and can be tweaked as needed. They’re not, they’re some of the most noisy, electrically unpleasant power systems in the world, with voltage bouncing up and down as trains accelerate and decelerate, consuming up to a MW each under peak load.
Power distribution companies hate providing power to train systems because you need so much infrastructure to handle the noise and prevent the train line from seriously degrading the local power grid. The substation that connect train lines to the power grid are constantly having their switchgear trip in and out as the huge spikes in demand cause the equipment to momentary disconnect to protect downstream equipment.
So creating headroom for extra regenerative breaking isn’t a simple thing to do. When each train can instantly consume as much power as 10 households all maxing out their power supply, electrical systems stop behaving anything like “normal”.
These are all valid concerns. My goal was to highlight an opportunity for improvement that does not require capital equipment purchases. As you note, analysis may not reveal any opportunity for improving the regenerative controls. Given their inability to address the problem today, I expected they’d rather investigate a low-odds opportunity than disregard it. I was incorrect; I will raise my minimum-likelihood thresholds at HN in the future. ‘Simple’ was meant only in relation only to all other identified options, that each require non-zero capital expenditure and thus invoke the complexities of capital expenditure. It was not my intent to diminish the intricacies or difficulties of rail electrical engineering with my imprecise use of that adjective and I apologize for the disrespect inflicted.
The space on tube carriages is used for ... passengers. Unlike long-distance trains (where the locomotive is a rather sizeable unit of its own and has pretty much an electric substation of its own in it), the parts of a tube train reserved for driver, engines, electrical distribution are comparatively tiny.
How many watts of extra energy would be required to accelerate each kilogram of batteries on the train? Would the additional thermal load introduced by battery charge and discharge effects counteract or worsen the desired cooling outcome?
It may be better to install such batteries somewhere on each track circuit as fixed stations, that can vent their charging and dispersal heat away from passenger tubes, rather than mobile — assuming that onboard emergency power needs are already met.
> How many watts of extra energy would be required to accelerate each kilogram of batteries on the train?
That's not really the problem; the problem is there is _no space_. Ever been on a London Underground train? You certainly wouldn't want the interior getting _smaller_.
I'm sure they'd consider the worst-case scenario, and at present that's probably some serious arcing within a motor causing a fire. Cutting the track power (easily done by the driver) then makes that a fire that can be handled with water and/or foam.
Smoke from large batteries would be very hazardous.
To me, 'disable abovegound regen' feels like not likely to solve the problem, just from a feeling that those systems are not that closely coupled. Otherwise, it seems easy to just keep on doing regen and set up (maybe not even need to: run a cable up to) aboveground dissipation grids.
I will guess that the limit is how much regen current can be passed back from the train into the supply system through the power supply rails / pickup shoes.
If I were making (confess, yes, untrained outsider) suggestions, I'd add water tanks to the trains, use the resistive braking to heat the water (not ambient air) during the trip, then change out the now-hot water for cold at the destination layover points. Not thinking this is a particularly creative solution, sounds like the "pull trains full of ice" already noted. Also this is off-the-cuff, so welcoming critiques!
Speculate: district level heating (wikipedia entry: https://en.wikipedia.org/wiki/District_heating) using heat pumps to draw out the tunnel heat; not sure if that is too complex altogether, maybe it would work as a longterm maintenance process but not as a 'fix the current problem' one...?
The trains in London can be up to 50 years old at this point. Where the technology was available during the building of the trains, you can generally expect regenerative breaking. But it’s far from universally available unfortunately.
