Gordon Richardson had the following to say in the forums
On the 28 August 1999 Cape Town was subject to some rare, strong thunderstorms. Cape Town is not known for their thunderstorms and in fact only a handful occur each year with years passing between severe thunderstorms. In the late night hours of 28 August 1999 thunderstorms wrecked havoc in the Manenburg area. Residents reported that the winds picked up and the sky was buzzing with lightning activity, then out of nowhere the winds picked up to what is estimated by officials to have been near 150 km/h. The official government reports and the media have declared that this storms primary damage was caused by a tornado, an extremely rare event in South Africa, nevermind in Cape Town.
The damage of the storm was extensive with thousands of homes destroyed, 177 people injured and 5 left dead – an extremely high amount of fatalities for a wind event. In the morning hours residents of the area were seen carrying out injured members of the community and surveying the damaged areas. The damage was so extensive that the South African army had to be deployed to deal with those who were taking advantage of the situation to loot homes and stores. Damage was the most extensive in Manenburg, Surrey Estate and Guguletu.
The weeks after the wind event saw the Silverstream Primary School being used as a shelter for those who had been left homeless by the storm. An intensive disaster relief effort was underway with members of the community and government organizations tending to the homeless.
As storm chasers we are extremely hesitant to follow media reports on weather events, with every second news source referring to cold fronts as hurricanes and gust fronts from squall lines as tornadoes – though this time there is quite a lot of evidence supporting the idea that this could indeed have been a tornado. I have been trying to find synoptic information on what conditions were present on the 28th of August 1999 though have had no luck (please contact me if you have this information), but when one looks at the details of the situation it would appear it is possible to put together the evidence to try and get an idea of what was happening. Conditions were reported to have been thunderstorm conditions, though given that it was August, this would likely not have been typical summer thunderstorms formed by a trough and Cape Town cold fronts are not typically accompanied by thunder activity, also with the reports of only a 1km by 2km primary damage path it doesn’t fit in with the idea of a cold front, which has a large area of similar speed winds.
My assumption is that the event was caused by a cut-off low pressure system with strong instability, causing severe thunderstorms to form. While the primary damage was restricted to a small area there were reports of strong winds around the general area. I think that it is most likely that these strong winds in surrounding areas were caused by RFD (rear flank downdraft) winds within the thunderstorm, while a short-lived EF-1 tornado could have caused the primary damage. Reports say that cars were flipped over, something that does not occur typically with cold front winds nor gust fronts or RFDs – but are consistent with a tornado.
It appears that it indeed could have been a tornado in Cape Town, though there is no guarantee on this. If you have any more information, such as first hand accounts, satellite imagery or meteorological maps of that day please contact me with this information.