As part of a wider smoke hazard analysis, TechArt Enterprises provided detailed engineering analysis of an exposed steel structure of an old power station that is being refurbished. 

The analysis involved modelling a fire in a compartment of the building where there was exposed steel structure.  The simulation was set up to specifically model the incident heat flux on the steel structure. 

Results of the modelling were used to determine a conservative temperature-time profile of the exposed steel structure.  Therefore we were able to determine the likely time to failure for the steel members being assessed.    

This project called for us to modelled the amount that smoke and might travel between adjoining apartments through a tiny hole in their common wall. 

Smoke is not normally suppose to travel between neighbouring units, unless of course they leave their doors open.  However, in this particular case, small holes were found in the common walls between units.  Where possible, the holes were going to be filled in.  However, some analysis was required to determine the amount of smoke leakage that could possibly occur if some small holes were to remain.  In turn, the results from the modelling were used to determine the Available Safe Egress Time (ASET) for the occupants in the neighbouring unit to that of the fire.

As the matter was likely to be before the courts, the analysis conducted had to be rigorous and peer review team was also set up.  We were working as part of a team and together a suitable methodology for the analysis was drawn up, TechArt then carried out the CFD simulation required using FDS.  The whole process was subject to constant peer review.

Due to the small size of the hole and the restrictions of CFD modelling, we decided to use a multi-grid model within FDS in order to get the high grid-resolution required at the location where smoke was expected to flow through the hole.  

When trying to solve fluid flows that contract down from very large volume to a very small hole, numerical instabilities can be a common problem.  By carefully stepping down the grid resolution in adjoining grids we managed to get FDS to solve the problem without encountering any numerical instabilities.  A problem if left unattended, would normally yield unreliable results. 

The modelling was successful and we were able to show that the potential and amount of smoke flow between units did have an adverse effect on the ASET, although probably not reducing it to a level that was unsafe.  

As a result of the peer review process and by maintaining close contact with all the associated parties the results of our modelling were able to be used in legal setting and were not challenged.