Mercury Rectifiers

[Prospero]

In olden times mercury was used in rectifiers.

Why was mercury used and how did these rectifiers operate ie was there an ionised gas with a grid voltage between anode and cathode?

 

[Skogsgurra]

Prospero,

I actually worked with some of those several years ago. There were two main types; glass and steel rectifiers.

The glass rectifiers looked like an octopus, hands up. But it had six arms. At the end of each arm, there was an anode and in the bottom of the envelope there was a pool of mercury.

The mercury pool emitted electrons (current direction anode to cathode) but the anodes didn't. They were quite efficient rectifiers, especially at higher voltages (voltage drop in the arcs was around 20 V).

They could be controlled using normal gates close to the anodes and the control characteristics was very similar to that of thyristor rectifiers. In fact, all the necessary theories were already developed for mercury rectifiers when the thyristors arrived in the late fifties.

Many HVDC transmissions still use mercury rectifiers. The first one between Swedish mainland and the Gotland island (started in 1954) used mercury. So did the English channel plant (started in 1962).

I said in the beginning that there were steel rectifiers as well. They were steel vessels with insulating walls and they worked much the same way as the glass rectifiers. Since they could be taken apart for service, they had vacuum pumps that maintained the vacuum. I do not remember seeing any pumps on the completely sealed glass rectifiers.

The glass rectifiers emitted a bluish glow. They were fascinating things.

 

[busbar]

 
A place that mercury valves work along side of their newer silicon equivalents is in bulk-power AC>DC/DC>AC conversion in the western US. In one case, the converter stations consist of six “stacks” of seven 133kVDC-rail mercury valves, and two stacks of 100kVDC silicon thyristors—all three-phase full-wave with commutation. In the process of adding the silicon, the system-DC rating went form ±400kV to ±500kV, at 3100MW. See transmission.bpa.gov/PlanProj/Transmission_Projects/Completed-TransP/Celilo/4-15-03FactSheet.pdf

Aside — Although not directly concerning mercury versus silicon, in earlier days of power conversion, it seems like greater attention was paid to harmonic minimization with more judicious use of passive-LC components. This limited higher-frequency components {and related losses} in associated power-frequency circuits.

Variations of mercury power-switching devices are thyratrons and ignitrons.