# How does a two dimensional elevator work?

If you watch this video, you can get a sense of how an elevator can operate in two dimensions. But, what I can't figure out is how it is driven.

If you consider a conventional elevator, there is a car and a counterweight connected by steel rope. The motor only has to oppose or resist the difference in weight between them (at most half the load, because it can be counterweighted to the car plus half its rated load). This limits the work the motor has to do when lifting the imbalance, and limits the energy it has to remove when lowering an imbalance. Also, the work is done by fixed equipment (typically at the top of the shaft) which doesn't contribute to the weight of the car.

But in the two-dimensional ropeless system, there is no counterwieght, the motor is part of the car (and adds to its weight), so has to drive all that weight going up, and get rid of all that energy going down. It also has to do it without risk of runaway/freefall. And, it shouldn't just burn off all that energy as heat (terribly inefficient). So my question is, how is it done?

• Please note that while the counter balace makes the steady speed of the elevator consume nearly no energy, it does nearly double the effort to accelerate. Which is why nearly all elevator companies are looking into magnetic rail systems. Apr 17 '18 at 4:52
• Weren’t these discussed by Douglas Adams? :) and the Sirius Cybernetics Corp.?? Apr 17 '18 at 16:52

Further to vidarlo's answer, these lifts will be controlled by some sort of variable frequency drives (VFD).

In a typical VFD deceleration causes regeneration and current is fed back onto the DC bus. This will cause the bus voltage to rise and at some point the voltage will reach the safe bus limit and the energy will have to be burnt off in a braking resistor.

Where multiple motors are accelerating, running and decelerating asynchronously it is common to use a common DC bus for a group of drives. This is relatively simple compared with feeding back into the grid which requires an inverter built into the drive. The idea is that the regenerated power is consumed by the driving motors and the overall effect is to reduce overall power consumption.

It also has to do it without risk of runaway/freefall.

Correct. This implies the use of brakes or motors with zero-speed holding torque. The least-wear / least power option would be to decelerate to zero speed and then apply the mechanical brake. The safety brake is, presumably, a mechanical centrifugal or standard lift / elevator brake.

And, it shouldn't just burn off all that energy as heat (terribly inefficient). So my question is, how is it done?

I think we've covered that.

Thanks for the video link. It all looks horrendously complicated with a lot of maintenance.

I tried to think of an application for the system. A couple come to mind:

• With a simple two-car / two-shaft system one could have the cars travelling in a loop. The left door is always going up. The problem is that there is some lost time while the car traverses horizontally.
• It may be possible, as I think is shown in the video, to have multiple cars in each shaft. If the shafts operate as one-way streets with linking cross-roads every so-many floors it seems that a much higher capacity system could be created for a given pair of lift / elevator shafts. In a hundred-story building each lift shaft saved could create another 100 rooms. The economics then start to look rather different!
• This is a very good answer. If you were to remove the "Further to vidarlo's answer" since.. well, it doesn't really add value any more.. and I could vote you up. Aug 15 '19 at 18:26

Modern VFDs can feed energy back to the grid when braking a motor. Some energy will be lost, but 70-90% may be recovered, depending on sizing and available loads.

The alternative to feeding power back to grid is to burn it off in braking resistors. This is a common practice in applications where feeding it back to grid is not feasible.

You could even do it without a VFD or active braking. The grid is rather stiff; it wants to stay at whatever frequency it is. If you power up a standard asynchronous three phase motor, it will rotate at the same frequency as the grid, minus some slip. If you however power the axle, so it want's to move faster than grid, the slip will become slightly positive.

But as the grid provides a large inertia, it will not change the frequency of the grid. Hoists using normal three phase motors without further braking will go slightly faster down than up, but will hold load perfectly fine.