German WW2 Hydrogen Peroxide Powered Submarine

Neal Stephenson's book Crytonomicon mentions a German WW2 hydrogen peroxide powered submarine carrying a shipment of gold from Japan to the Philippines, or something equally bizarre. I've found a few mentions of this German secret weapon, but nothing as expansive as this very good brief I found on RC models by Robert Holsting.

1940
German scientist Helmuth Walter demonstrated a prototype for the first true submarine – a boat which in theory could operate submerged for an indefinite period, unlimited by battery capacity or the need for atmospheric oxygen. V.80 was powered by the decomposition of highly-concentrated (95 percent) hydrogen peroxide, H2O2, known as Perhydrol. In essence: when the chemical breaks down, it releases superheated steam to drive a turbine, along with oxygen to support conventional combustion or for respiration by the crew.

The hull-shape of V.80 was optimized for submerged operations, and the boat indeed demonstrated exceptional speed – 28 knots submerged. It also demonstrated exceptionally high fuel consumption, 25 times that of a diesel engine, at exceptional cost. According to one source, one 6.5 hour trial run consumed $200,000 dollars worth of Perhydrol.

The design showed great promise. However, Hitler thought his war was won, and plans for the production of a series of Walter boats were put in limbo.

The 1943 experimental 250-ton Type Wa-201 Walter boat, U-792,
which hit 25 knots, submerged, on sea trials.

Research continued. Perhaps eight, in several variations, 250 and 300 tons, were put into service, 1943-44

The Type Wa-201 Walter boat, U-793, here partially dismantled at the end of the war.

Collapsible hydrogen peroxide storage bags being removed from the 300-ton Type XVIIB Walter boat U-1407 after the war. With the type of storage outside the pressure hull, fuel could be consumed without appreciable change in trim – seawater simply replaced the depleted volume.

Update June 2019 replaced pictures that disappeared after Blogger 'fixed' a problem it detected when I updated the labels.