I've kind of got used to this kind of sht. If it seems to be too good to be true then it probably is or there's a catch.
My luck has always been questionable so I should probably have left it be but curiousity and all that.
It would appear that the top if the failed arc tube was poorly made and this is what burst and ejected molten Na into the outer jacket causing instant short circuit and cracking of the stem and rendering the whole thing nothing but a useless piece of glass and sodium oxide / sodium peroxide.

That’s actually heart breaking.id heard of these breaking. Someone said to me once that if I was to ever get one, to never light it. Sorry to hear that happened. I wonder if one arc tube was colder than the other and gave way due to thermal shock!

One of the two arc tubes ruptured after a couple of minutes while running up for the first time.
Here's the lamp.
Biggest waste of money I've ever spent. Still pees me off thinking about it. 

What way has it failed? @Andy

The only exception I think to the top quality lamps produced at Shaw was the 280W SOI lamp.
I had a very expensive and disappointing sudden failure of this lamp and from what I have since heard, several other examples of this lamp have failed or partially failed.
A real shame when all the other lamps were superb.

For 5 hours it indeed took some time to carefully write and make it the best if you ask me. I kind of like how you had a lot to say about Schrèder and AECE. Are these like the 2 biggest Streetlight manufactures or do you just like them manufactures and just had a lot to say. And for the last question, are all those 8 manufacture from Belgium or are there more than 8. Would like to know if So.

And here's a plus, it's not every day you see someone write a overview of a country's manufactures that make streetlights.

A just a quick question, after all you do seem to be quite interested overall from my Polish streetlights posts, do you think I should write the history of polish streetlights? I don't think anyone did it before, but I could do it if you want to 👍

And plus, I think it's a good way of doing this so you don't go through websites and try find some informations when already there is an information on LG's post. Maybe you do spend a few hours, days to write an excellent write up from your country's streetlights, but at least you feel like it worth it and some other may do as well reading it

If you mean break the glass, then I would guess so...

But breaking the filaments and letting that to shut the ballast down is very common method on many cheap ballasts and CFLs.
The higher voltage for maintaining arc at EOL tube means the ballast has to be boosting the voltage by the output curcuit resonance. And when the resonant capacitor is connected via the lamp filaments, the resonance is feeding them by large current, so melting them off. And once broken, the resonant capacitor is not connected to the circuit, so the thing can not boost the voltage anymore, so the circuit shuts down.
The main problem with this method is, some filaments are way tougher than anticipated, so actually refuse to break even at the overcurrent (lamp maker uses thicker filament "to make it stronger", but also a leak into the tube causes excessive filament cooling, preventing them from melting), with the consequence of the ballast not shutting down on EOL and continuing to overhat the tube end assemblies.

It will hit 3 years of dusk - dawn use on 2nd July 2026 and shows little signs of wearing up!

@Danny How many years (or an approximate number of hours) has it done?

I best be patient with the 55w osram from 1996 that I have on test then lol