I had another look at it and I think it's not really necessary to prevent this positive feedback loop.
Currently this loop is only possible with eff modules unlocked, and even with this it can only generate a very small amount of energy. The advanced chem plants are not cheap, and if you use higher tier modules it's even worse. This setup won't generate as much energy as solar of equivalent space, and is much more expensive.
The biggest problem is pollution. The fluid boiler is super polluting if you want to power your base with it. At this point you already have access to nuclear plant + dirty water uranium, which is a much better form of infinite energy, and generates very little pollution.
The only real reason to block this loop is realism. From realism, I think the biggest problem is 'electrolyzer' interacting with modules. The energy effciency of electrolysis should be maintained in a fixed range, but modules can vastly change the effciency. In nullius (another overhaul mod), the electrolyzer is designed to be a practical energy storage, and no positive loop. It doesn't accept any modules.
The different way may be to make electrolysis much more energy demanding, but as mod doesn't have dedicated electrolysers, the only way to do this is to make electrolysis process much slower.
Not really. There are still multiple alternatives:
1. You can require water to be pre-processed before electrolyzing. In nullius there is a pretty quick steam electrolysis recipe. I think we can do the same here. Modules won't interact with the steam input, so the effciency is almost fixed. You can change the recipe to like this: 20 steam -> 10 H2 + 10 O2, taking 1s. 2 steam have 60kJ of heat, so 1 hydrogen can also have up to this much energy. This will certainly break existing setups though.
Increasing the fuel value of hydrogen will make voiding by burning harder. So I think there should be a void protium recipe in the air filter. But if there was such a recipe, why are we here talking about this fuel value thing??? LOL
For more realism I think the recipe should be 20 steam -> 20 H2 + 10 O2, taking 1s, and hydrogen 30kJ.
2. Maybe you can use the plasma heater machine? It has a scary 20MW base energy cost. It only has 1 output, so you need to discard the oxygen in the recipe. The main problem is it's too late in the tech tree.