The radiant gas fire was checked and a fault found but that was not the cause of outages. A bit of advice was offered on air-admittance and no reported problems after that when I checked back.
Which set me thinking.
Vitiation in that room down to 18.5% oxygen (normally 20.9%) should in theory trip out an appliance which I believe was the cause. Ultimately vitiation would lead to poor combustion and CO production. However the ODS kicked in first as it should, long before CO production.
That assumes of course vitiation occurring in the presence of a safe working appliance. But what if vitiation is being caused by an already unsafe working appliance. That is a whole different ball game. Danger lurks at the outset as it did with the focal-point flueless gas fire in Wales that killed that young lady and her dog.
The gas fire was spewing toxic CO in abundance. That was rising due to the by then high temperatures being generated by the uncontrolled appliance (no thermostat). The inspector's report also alluded to the open-plan design contributing to the tragedy but again I reject that. In the context of rising gases, all homes are open-plan when doors are open, even slightly.
We do know that the house was full of CO with the young lady upstairs killed and her dog in the lounge also killed. But still the air was not sufficiently vitiated in the proximity of the ODS, to cause the appliance to trip out. No good pointing fingers at the ventilation or open-plan.
Even then it assumes the ODS will trip at a precise 18.5% which again I would now need to see proof of. Other experts already have a different opinion.
A popular trick on ACS assessment (everyone knows now) is to stuff a wee bit of grandpa's vest down the burner tube on a radiant gas fire, thus causing incomplete combustion on a few - not all - burner ports. That can and does happen with lint, fluff and pets hairs being pulled in with primary air. It does not require 'pilot error' to cause a problem.
One of the best boilers ever made is the Baxi Solo and the natural-draught one I used to have filled with soot at nine months old. A heat-only boiler, off in the summer, when a red-admiral butterfly entered the flue and blocked a primary air port on one of the seven burners. The boiler (room-sealed) was working for a good while before it shut down. The butterfly remarkably was unmarked.
The point being made is that an appliance can develop a fault at any time.
All considered the ODS will obviously never work when it is most urgently needed and flueless gas fires therefore cannot comply with CE guidelines.