In practice various methods are used to demolish structures, ranging from simple ‘brute-force’ type approaches, to highly technical progressive demolition techniques, such as outlined by the JISF (2015). Multiple codes of practice have been produced to assist contractors during demolition [e.g. HK Bldg. Dept. (2004), Indian Standard 4130 (2002)]. Demolition associations have been formed, such as in the UK, USA and EU, to try assist demolition contractors and regulate the industry. However, these guidelines, codes and organisations typically only provide construction related guidelines without giving information to assist structural engineers who have to ensure that structures remain safe during the demolition process. Minimal research has been conducted and literature published up to this point on predicting the load capacity of weakened structures (van Jaarsveldt and Walls 2016). The techniques and analysis methods used in industry rely primarily on experience obtained from previous projects, and are less focused on strict theoretical principles, although some practitioners may use simplifed design equations. It is necessary that sound engineering and scientifc principles be applied to this problem to produce verifable and safe design techniques. Furthermore, with safety requirements becoming much stricter worldwide, contractors may now be required to show by calculation that their procedures are safe, which currently cannot easily be done. One of the most efective methods available for demolishing structures involves the pre-cutting of steel columns of a structure at a certain height, using diferent types of cuts. This weakens the columns of the structure so that the least efort can be used to demolish the structure by means of pulling out some of the columns or by using explosives (NDA 2014). The pre-cutting of the columns is done manually, which causes this method to have safety implications. If a structure has been weakened too much it could collapse during the weakening process (Walls 2017). Conversely, if a structure has not been weakened enough it will not collapse when required, leading to a partially collapsed structure, which is also a very dangerous scenario. This paper will focus on determining the capacity of weakened columns through experimental testing and fnite element modelling. 2 Background Steel structures are demolished for diferent reasons, such as: removing buildings that have been damaged by fre or other events, to make way for new developments, or because buildings are no longer safe to use (NDA 2014). The demolition video screen-shots in Fig. 1 illustrate a furnace structure that was demolished by Jet Demolition (Pty) Ltd. For a detailed description of the collapse mechanism refer to Walls (2017). The structure was weakened through applying pre-cuts to the columns and the beams, as shown in Fig. 1a. However, the furnace can be seen to remain in its original position, whilst demolition staf were operating in the area. The middle columns of the structure were then pulled out using steel cables as shown in Fig. 1b. This caused the structure to fail as seen in Fig. 1c. This process illustrates one of the current industry approaches used for the demolition of steel structures.