The Canterbury Television (CTV) building's design was "undesirable" for resisting earthquake shaking, an engineer says.
Structural engineering professor Nigel Priestley this morning told the Canterbury earthquakes royal commission the building's key support wall, or shear wall, would ideally not be on the exterior.
"There certainly have been other buildings designed in such a fashion with an external shear wall but primarily in non-seismic regions.
"It's quite clear that from a seismic point of view that this is an undesirable building configuration and [that makes it] very difficult to make it perform well [in earthquakes]."
Priestley peer reviewed a Department of Building and Housing (DBH) report into the CTV building, which collapsed in the February 22, 2011 earthquake, killing 115 people.
The report found the building failed to bend sufficiently during the quake and had poor strength and layout of its key shear walls.
It did not meet the building or design codes for when it was built, the report concluded.
Alan Reay, director of Alan Reay Consultants Ltd (ARCL) which designed the building, labelled the report as "technically inadequate" when it was released in January.
Reay will give evidence today, and will be followed by engineer John Mander, who will testify the CTV building complied with all applicable design and building codes.
Priestley dismissed a conclusion from Mander that the building was "innovative".
"It might be innovative but in a very undesirable sort of fashion. I cannot accept that this is an innovative structure in a desirable form for seismic resistance."
Priestley also gave evidence on the joints that connected the CTV building's concrete columns with horizontal beams.
The strength of the joints has come under much scrutiny at the commission.
It could not be known whether better steel reinforcement in the joints would have kept the building standing, Priestley said, but theoretical tests suggested it could have.
"I can't say that the structure would have survived but I can say the...time-history analysis would not have predicted failure."