I was looking for data on ice resistance for these kinds of 1000 litres containers. I believe they are made of polyethylene with an outside aluminium structure. Given the physics of ice expansion, what could be the safe level of water inside if subjected to icing conditions ? Does adding a half inflated bike tube inside could allow relieving pressure from ice ?
Ice has a volume about 10% bigger than water, so an inner tube is too small to do anything - unless it floats, see below.
Containers burst when the ice forms a plug and the trapped water then freezes and expands. One way to avoid that is float a beach ball in the tank. When the top of the water freezes over, the pressure beneath it will crack the ice round the ball and let some of the water get on top of the ice, rather than being trapped underneath.
You may find that a plastic tank is flexible enough so that this happens around the edges anyway - but it does no harm to provide another way out for the water.
You might also consider how much latent heat would be released if the entire tank froze solid. It would need to be in extreme conditions for a long time for that to happen. Presumably, the purpose of the tank is to be repeatedly filled and emptied, not just stay full of water and slowly freeze.
Some of these tanks are also sold with external insulating jackets - they are sometimes used for hot water storage, but the heat insulation works just as well in either direction.
If you leave roughly 10% on top empty it should be okay. Ice increases in volume by approximately 8%.
Leaving a half deflated bike tube will hardly help. Because ice formation starts from surface of the tank and by the time it gets to the tube it is too late.
Most of the plastic containers can flex a bit to accommodate icing if you leave the top empty. But they eventually have a certain number of cycles an will crack after that.