In traction power supply systems for city electric transport without storage devices, a significant part of the recovery energy is usefully reused for the load in the process of inter-train flows along the contact network. The excess part of the recuperation is dissipated by heat in the braking resistors. The use of on-board storage devices makes it possible to redirect all recovery energy from the car's traction drive to them. The remaining amount of no more than two-thirds of the received recuperation energy is supplied back to the traction drive from the storage device, excluding internal losses, since due to ohmic resistance, the actual efficiency of any type of storage device does not exceed 64 %. The on-board storage device consumes a significant amount of paid energy from the traction substation for its transportation by car, compensation for self-discharge, etc. The total balance of recovery energy returned to traction and losses from additional energy consumption for on-board storage devices is negative. When stationary storage devices operate in the contact network, useful recuperation flows are completely preserved. The excess part of the recovery is redirected to the contact network, where one part of it powers a low-power non-traction load while charging the storage device, and the other goes to the storage device, charging it. After a short period of storage in the storage device, the remainder of the excess recovery is released into the network for the load that appears in it. A stationary storage device has no energy consumption for transportation, and the consumption of paid energy for its own needs and compensation for self-discharge is significantly less than that of an on-board storage device due to its higher load factor. The stationary storage device has a positive balance of recovery energy which allows additional savings of paid energy from the traction substation in the amount of up to 10 % relative to such consumption during useful recovery flows without the use of storage devices.