This story was first published on June 3, 2017. On November 30, police told the families of victims they would not prosecute anyone in relation to the CTV building collapse.

Kulpe held the thought as the shaking started with a violent upward jolt that lifted her clean off her seat. She ran for the doorway. It was a rickety looking thing with an aluminium frame, but it was better than nothing. Since the magnitude-7.1 earthquake that hit Canterbury in September 2010, Kulpe had made a habit of bolting for the nearest doorway during an aftershock. It was a bit embarrassing: she was a counsellor with Relationship Services and a number of her clients had suffered trauma. Whenever the shaking, and doorway run, interrupted a session, she used to joke that they were handling it better than she was.

This was by far the biggest aftershock. Magnitude-6.3, shallow, and centred in the Port Hills, about five kilometres from the Christchurch CBD. After the initial jolt the building lurched to the east. Kulpe’s colleague Angela Osborne joined her in the doorway. Several metres away, Elizabeth Cammock had been thrown against the wall while in her chair, glass exploding and metal twisting around her. She looked over at the doorframe and saw Kulpe and Osborne taking cover. Good luck, she thought.

At the north end of the building, in the Relationship Services waiting room, Kendyll Mitchell was early for her one o’clock appointment. Her 3-year-old son, Jett, had been terrified of sleeping in his own bed since the September earthquake. As soon as the shaking started, she grabbed Jett and tucked him in close. With her free arm, she held the pram where her 10-month-old daughter Dita sat strapped in place.

For a moment, nothing happened. A calm, lasting a second or two, enveloped the building. Then, more shaking. The floor under Kulpe and Osborne lifted up and the ceiling collapsed, bringing panels, lights, wiring and Pink Batts insulation down on top of them. Then everything went to hell. The fifth level of the building – the one underneath Relationship Services – appeared to give way first, triggering a concertina effect on the floors below. Kulpe sensed the floor was sloping south and the building twisting anticlockwise as it collapsed into itself. When it was over there was another jolt, like an elevator reaching the ground floor. Kulpe was still standing. She could see daylight through a small hole and was able to clamber out after rescuers opened it wider.

The CTV building was born in 1984, when development company Prime West Corporation bought three adjoining sections on the corner of Madras and Cashel streets. Two years later it approached a building company, Williams Construction, about leading a speculative design-build on the site. It wanted a structure that was functional, presentable, and as cheap as possible. Williams managing director Michael Brooks came up with the idea of putting the lift shaft outside the floor plan, maximizing the lettable floor space. Christchurch architect Alun Wilkie was engaged for the design. His client liked the look of a four-storey building on the corner of Durham and Armagh streets, with an offset core, precast concrete spandrel panels, circular columns, and inset glazing. It would be good if the new building looked similar.

To engineer the building, Williams turned to Alan Reay Consulting Engineer (ARCE). Brooks and Williams’ quantity surveyor Tony Scott had worked with the company on another project and liked the standard and presentation of the drawings. The firm’s principal, Alan Reay, would also provide preliminary drawings for costing purposes at no charge.

Reay was a man with a growing reputation. He held a Bachelor of Engineering and a PhD in civil engineering from the University of Canterbury. By the mid-1980s he was prominent designer of low-rise tilt-slab buildings and their production systems. His standing grew further in the 1990s, when he became the first non-American to be awarded for engineering achievement by the Tilt-Up Concrete Association of America. In 1995, he was the engineering advisor to the commission of inquiry investigating the collapse of a viewing platform at Cave Creek on the West Coast, which killed 14 people.

The CTV job would fall to an engineer named David Harding. Reay didn’t have much experience in designing multi-storey buildings and the only engineer at his company who did, John Henry, had abruptly left. Harding had previously worked for Reay in the late 1970s, but left to do civil engineering work. He had never designed a multi-storey building but was keen to broaden his knowledge in the field. Here was a way to do it. He rejoined ARCE in November 1985. “They say you should never go back,” Harding said later, “But I made the mistake of going back.”

Harding started work on the CTV project in March 1986. His design, based on the client’s wishes and Wilkie’s drawings, was a shear wall protected gravity load system. This meant that in an earthquake the lateral forces from the shaking would be absorbed by a shear wall, or walls - the strongest part of the building - while the columns that held up the structure were designed to do pretty much just that. The columns were attached to reinforced concrete beams, which in turn supported the floor slabs. In engineering parlance, it was an “eccentric” design because the centre of rigidity and strength - the shear wall - was not at or near the centre of the building. The shear wall was on the north side, outside the floor plan altogether, along with the lift shaft and stairs, just as Brooks had suggested.

Because of this, and the fact that the building was more than four storeys high, it was a code requirement to conduct computer analysis to measure earthquake loading. To do this, structural specifications were fed into a program called ETABS, which produced a reading of inter-storey deflection (how much the floors would move laterally in an earthquake). The first time Harding ran the numbers, ETABS predicted deflections greater than what was allowed under New Zealand building standards. Harding increased the wall thicknesses and ran it again. No change. After four or five unsuccessful attempts, he went to see Reay.

The two men discussed the problem and settled on a solution: a second shear wall on the south side of the building. Until then, the design had been relying only on the north shear wall for earthquake strength. Adding a south shear wall was not straightforward. There was already a firewall there, and the upgrade would have to allow for access to the escape doors and stairs. This meant the new wall would be coupled - split in two and connected by steel beams. Harding ran ETABS again with the new design. It came back with deflections inside the allowable limits.

Harding’s difficulties had not been helped by his lack of experience with ETABS. With no history working on multi-storey designs, he had never had to use it. His only experience with the program came a couple of months earlier on a project for which Henry had done the bulk of the work before he left. Because of this, Reay gave Harding the file for Landsborough House - another building Henry had designed that was similar to CTV - complete with ETABS calculations his new engineer could use as a “method template”. Harding did so, but his inexperience led to a crucial error. ETABS was a rudimentary program. The deflection data it produced was for a single point at the centre of mass for each storey of a building, not the corners, where movement would be exacerbated. An engineer running ETABS on eccentric buildings like Landsborough and CTV would have to calculate their centre of rotation - which would be somewhere other than the physical centre of the building - and use that point to work out likely deflections at the corners. Henry’s Landsborough file contained extensive notes, but nothing about calculations for corner deflections. Harding didn’t do any, which meant he had numbers telling him the floors in the building would move less in an earthquake than they actually would.

Once ETABS analysis was complete, Harding and ARCE draughtsmen went to work on detailed structural drawings. By the end of the year Harding had logged more than 300 hours on the project, all the while thinking his work was being reviewed by his boss. It wasn’t. Reay thought Harding was up to the job; Harding thought Reay was shepherding him through his first full multi-storey design project. Neither of those things was true, and the end result was that when the drawings were submitted to the Christchurch City Council as part of a building consent application, the engineer who reviewed them didn’t like what he saw.

The application was lodged on July 17, 1986. It was common practice in the 1980s for developers to submit incomplete paperwork to get the ball rolling on consent for a project, and CTV was no different. The council didn’t receive the structural drawings until August 26, and even then they weren’t finished. City council engineer Graeme Tapper responded in writing to ARCE a day later requesting a copy of the calculations supporting the design and outlining numerous omissions in the drawings. One stood out: “Floor connection to shear wall system and general connection between floor slab and walls.”

Tapper was a meticulous man, but even by his standards this was a long list. Anything less, though, would have been out of the question. Tapper had little tolerance for substandard work. He held strong views, was happy to share them, and reluctant to change them. He periodically had blazing rows with his boss, buildings engineer Bryan Bluck, when they disagreed on a job. Around the ARCE office, Reay referred to him as “Colonel Tapper”. If he ever needed to talk to someone at the council he would deal with Bluck, whom he had known for years.

The additional calculations and amended drawings were provided to the council nine days after Tapper’s letter. Five days after that the city engineer signed off on the application on behalf of the council’s structural unit. During that time, the CTV building had turned into a battle royal for Tapper and Bluck. Tapper wasn’t happy. He broke the habit of a lifetime and told wife Pat of his concerns about the building’s earthquake risk and the “huge pressure” from Bluck to sign off on the design. He went on and on about it. It was “not a question of if,” he said, “but when”.

“When it happened he was concerned the CTV building would not prove to be strong enough,” Pat Tapper said.

News spread through the engineering fraternity. West of the city, Riccarton Borough Council engineer Peter Nichols, formerly a city council engineer, heard about a Tapper-Bluck run-in that was particularly caustic, even by their standards. A few months later, while in the central city, he made a point of visiting the construction site at 249 Madras St to see what all the fuss was about.

As Nichols stood on the Cashel St side of the building, he bumped into his old boss, who was taking his midday walk. Bluck told Nichols the building had a “novel technological approach” with two non-symmetrical shear walls. The design initially worried him, he said, but his fears were allayed after talking with Reay.

Whatever was discussed in Bluck’s conversation with Reay, it wasn’t enough. Despite the extra calculations and drawings from ARCE, the CTV design was still under code, including the connection between the floor slabs and the north shear wall Tapper had mentioned in his letter. The building shouldn’t have got a consent, but it did. Williams quantity surveyor Tony Scott costed the project at $2.45 million.

Construction on the CTV building began in late 1986 and did not get off to an auspicious start. Site foreman Bill Jones found it hard to get good workers. Some were hired on a daily or weekly basis, and retained depending on whether or not they were any good. Jones himself was accustomed to smaller projects - light timber structures up to three storeys - and had never managed a design-build contract. He was used to having a clerk of works on site for guidance.

Williams was in expansion mode, and Brooks decided to hire a construction manager. The new man, Gerald Shirtcliff, had an engineering background and had worked in Australia and South Africa. He was expected to provide daily oversight and guidance to Jones and the foremen on Williams’ other projects. In fact, he visited the CTV site about once a month. “He just wasn’t up to the job,” Brooks later said. Council site inspections to monitor standards were sporadic too. Between April and August 1987, there were none at all.

To complicate things further, Williams was the subject of a takeover bid about the time construction started, and was sold by early 1987. Brooks, Scott and Shirtcliff resigned and formed a new company, Union Construction. Union took over the CTV job and completed the build in early 1988. It later became insolvent.

The speculative design-build project sat empty for two years before a prospective buyer emerged in the Canterbury Regional Council (later Environment Canterbury). The council hired engineers Holmes Consulting Group (HCG) to inspect the building as part of the pre-purchase review. Senior engineer John Hare was handed the job and examined architectural and structural drawings to make a seismic analysis. He quickly identified the same problem Tapper did - the ties between the floors and the north shear wall didn’t meet the code. “There appeared to be no connection detailed for the walls on either side of the lift shaft,” he said. The north wall complex was “punctured” by the lift shaft and stairs, leaving few floor-wall connections. In the event of an earthquake, Hare said, the building would “effectively separate from the shear walls well before the shear walls themselves reached their full design strength”.

Hare visited the ARCE offices, now trading as Alan Reay Consultants Ltd (ARCL), to tell them about the problem and the need for potentially expensive remedial work. Harding had left the firm, but Reay and engineer Geoff Banks agreed the connections were non-compliant. “A straight blunder,” Reay said. To double check nothing had been added during construction, Hare and Banks inspected several floors of the building with a tool to locate structural elements. They found no significant reinforcement.

Soon after HCG sent its report describing the connection problem to the regional council, its services were no longer required. A $14,000 quote for remedial work also went no further. The council would not be buying the building. It was left to Reay and Banks to remedy the situation. Prime West was in receivership so they contacted receivers KPMG Peat Marwick to notify the owner of the problem and confirmed the scope of remedial works with HCG. Hare had not identified any other compliance problems. “[The building] generally complies with current design loading and materials codes,” his report read. That was good enough for Banks and Reay, who did no further structural assessment.

In fact, no-one did much of anything for about a year. Reay kept an eye out when he drove up Madras St in case it looked like anyone was moving into the building, but otherwise nothing was done until a story appeared in The Press on February 4, 1991, reporting that the property had been sold. ARCL took legal advice and resolved to contact the new owner - Madras Equities Ltd. It sent the company a letter in April 1991 notifying it of the problem. Then nothing happened for another five months. Finally in October, steel drag bars were fitted on levels 4, 5, and 6, reinforcing the floor-shear wall connection. ARCL should have applied for a permit for their installation, but didn’t. It was a missed opportunity. Although Tapper identified the floor-wall problem at the design stage, a consent application for the drag bars to shore up a non-compliant connection would have brought the issue to the city council’s attention in even starker terms.

In 2001, when the Going Places language school planned to move into the building, a consent application for the fitout was submitted to council on the grounds that the new tenancy was a change of use - a school would house more people than a standard office. That triggered a council inspection, including a structural check. An engineer noted the building was “reasonable modern 1986 [sic]” and concluded the planned extra 20 people on the floor wouldn’t affect the live loading. He could have asked to see structural drawings, but there was no requirement to do so.

In 2008, King’s Education, a school for English as a second language, moved into level 4 of the building. Another school shifting into office space equalled another change of use and the possibility of a structural check, but Madras Equities filed no application. Seventy-one students and nine teachers from King’s died when the building collapsed three years later - a far higher casualty rate than if level 4 had remained an office.

The 2004 Building Act had introduced the notion of an “earthquake-prone building”. Structures less than 33 per cent the standard of the new code were subject to strengthening requirements. The CTV building was somewhere in the order of 40 to 55 per cent.

Hundreds of buildings in the central city were damaged, CTV among them. Cracks appeared in the interior walls on every level, some running from floor to ceiling. On level 2, CTV technical engineer Peter Brown could see more daylight through one of them every time there was an aftershock.

After the earthquake, the city council dispatched building and engineering experts to assess structures in the CBD. Their first task was to conduct Level 1 Rapid Assessments - basic structural checks for obvious signs of damage that might warrant further scrutiny or anything that suggested a building was about to fall over. They issued buildings with green, yellow, or red placards, depending on the urgency of the situation. Three assessors, including a chartered professional engineer, visited CTV on the afternoon of September 5. They found no apparent structural or safety damage and allocated it a green placard - “no restriction on use or occupancy”. They noted on the form that their inspection was brief.

Two days later the building was identified as one of three in need of an immediate, more rigorous re-inspection. No-one was exactly sure, though, of the reason for the urgency or how much more rigorous. The three building experts assembled for these inspections did not include an engineer, which was a requirement if they were going to conduct the next logical step - a Level 2 Rapid Assessment (an internal and external review). The trio were sent out without clear instructions on what they were supposed to do, and when they arrived at CTV and saw the green placard on the door the problem was clear: someone had done an L1 assessment already and they couldn’t do an L2. They decided to have a look around anyway.

The three men spread across several floors to look for damage. They noticed cracks similar to those the L1 team found two days earlier and concluded that an engineer should conduct a proper assessment. They didn’t see any damage that posed an immediate problem, but thought the building needed an expert check nonetheless.

Despite this, they filled out an L2 assessment form, issuing another green placard. A box on the form labelled G2 was circled, denoting “occupiable repairs required”, but there was nothing to say there had been no engineer present or that the inspectors agreed that one should be commissioned as soon as possible. This was important, because a green placard from an L2 assessment looked reassuring to management and staff going back to work in the building. CTV manager Murray Wood could be forgiven his mistake when, within an hour of the inspection, he emailed staff: “We have just had an internal inspection of the building from 3 engineers and they have found that this building is in good condition and is deemed habitable.”

The one positive that emerged from the non-L2 assessment was that CTV building manager John Drew commissioned a more detailed inspection. Structural engineer David Coatsworth examined the building on September 29. Having identified the north wall as its structural core, he looked closely for cracks in that area. There were some pointing to minor structural damage, but nothing that would compromise the shear wall and nothing suggesting the floor and walls may have separated. None of the column cracks he found indicated structural damage either. His report concluded there was no evidence of structural failure. The findings may have been different if he had been given the structural and architectural drawings of the building that he asked to see. Drew didn’t have them so put a request into council for a copy. Council, cleaning up its own house like everyone else after the earthquake, wasn’t exactly prepared for archives inquiries. When Drew finally got word the papers were available, Coatsworth had already finished his report. Thinking the engineer must have been satisfied without them, Drew never chased it up. Coatsworth’s visual inspection was thorough - it was unlikely there was any significant damage to the building - but yet another chance at diagnosing its crucial structural flaw was missed.

On December 26, 2010, Christchurch was rattled by a sequence of aftershocks, shallow and close to the city centre. The first and strongest one measured magnitude-4.7. The city council rapid assessment teams swung into action again. The CTV building was inspected on December 27. Another L1 assessment resulted in another green placard and a note about the possibility of unstable windows. Drew, on holiday and out of cellphone range, put a call into Coatsworth when he found out about the aftershock the following day. According to his voicemail, the engineer was away over the Christmas-New Year break. Drew decided not to follow up and pushed ahead with repair work Coatsworth had recommended.

Staff working the building, appeased by the September inspections, were rattled anew by the Boxing Day quake. Some of the existing cracks were longer and wider, and numerous new ones had appeared. At King’s Education, on level four, the crack in the pillar in the video conference room now ran floor to ceiling and was wide enough at the top to put your hand through the wall. When she covered reception, operational manager Margaret Aydon used Blu-tack and rubber bands to keep pens from rolling off the desk. A building demolition nearby on Cashel St periodically sent vibrations through CTV and didn’t make anyone any calmer. Relationship Services national practice manager Jo-Ann Vivian, visiting from Wellington, was concerned about cracks in a column on Level 6 and emailed the city council on January 5 requesting an inspection. She withdrew the request two days later after speaking to Drew, who said engineers had checked the building after September 4 and Boxing Day.

“I remember feeling a sense of relief that had occurred and was also a little embarrassed I had contacted the council before checking with him,” she said.

Drew called a concrete repair contractor, Graeme Smith, who inspected the building three times in early February to scope the damage Coatsworth had highlighted. Smith, a civil engineer, inspected both shear walls, the lift shafts and the interior and exterior of the building. The damage was “unremarkable” and in line with what Coatsworth had identified. He prepared a quote for Drew dated February 22, 2011.

The building lurched in every direction - north, south, east, west - twisting around the north shear wall. It withstood the shaking for several seconds before succumbing to its force. On Cashel St, facing the south side of the building, Hawker and Spencer saw the concrete columns on level 5 - the second-top storey - crumble and start to collapse.

Almost in unison, all the windows in the building shattered and the cladding melted away from the exterior wall. Level 5 pancaked into level 4, into level 3, until the entire building had disintegrated neatly onto itself.

Some cars were crushed, but others parked almost up against the building suffered only minor damage from falling debris. The top level of the building, partly intact, sat atop the rubble pile, having pulled the south shear wall down on top of it as the collapse was complete. The north shear wall stood ragged and alone; stripped of every facet of the structure it was built to support. Within minutes, smoke began billowing from the rubble. The collapse had taken no more than 20 seconds.

Sergeant Michael Brooklands, of the Kaiapoi police, was en route to Christchurch to drop a vehicle into the central station for repairs when he got a call to respond to a burglary in the CBD. He apprehended one suspect and, with another officer, was searching for a second when the earthquake hit. When it was over Brooklands walked up Cashel St trying to steer people towards the expanse and relative safety of Latimer Square when he came across the remains of the CTV building and radioed for urgent assistance. It was just after 1pm.

As the senior officer, Brooklands assumed command of the scene. Detective Sergeant Mark Keane arrived soon after, but the two men saw no reason to pass the baton between equal rank. They also decided to split the scene in two - east side and west side - and lead two searches for survivors.

It was soon apparent that the two jobs were very different. On the eastern side, rescuers faced large concrete slabs and beams they couldn’t shift. In the west, some of the rubble could be moved by hand, exposing pockets of space. Kulpe, Osborne, Cammock, Mitchell and her children, and a number of other survivors close to the surface were freed and taken to nearby Latimer Square for treatment. A makeshift morgue was established on an adjoining vacant site.

Brooklands was nervous. The smoke was getting thicker and aftershocks were tearing through and there was no way to know how stable the rubble pile and the still-standing north tower were. He ordered most of the people off the rubble while a handful listened for signs of life. Several more people were rescued this way.

Brooklands had ordered a digger on site to drop dirt and water on the fire, to no avail. Flames and smoke continued to pour forth. The first fire engine on the scene arrived just after 1.30pm, headed by Addington station officer Alan Butcher. Station officer Stephen Warner’s crew arrived soon after. Their options were limited. They didn’t have much water and the site was dangerous and unstable. Two firefighters were craned onto the top of the north tower to look for survivors, but were driven out by the heat and smoke.

By late afternoon the scene was chaotic. There was limited equipment and rescuers in the rubble were yelling for tools. Inefficient human chains were taking things to and from the site. The fire was still burning. The ranking fire officer on the site was senior station officer and urban search and rescue (USAR) member Dave Berry, who many firefighters, but not Berry himself, considered to be in charge. More senior officers had visited the CTV scene but none had stayed and there was no order to put anyone who was there in command.

By 4pm a call was made to deploy heavy machinery and start lifting rubble from the site. The decision laid bare the appalling dilemma rescuers faced. Digger drivers were nervous about crushing trapped survivors if they disturbed the pile, but it was going to be difficult to get people out otherwise. Moving large pieces of rubble would help firefighters get to the seat of the blaze but would also stoke it with fresh oxygen. Dousing the flames from the outside came with a risk of drowning survivors. There were no good options.

Warner came up with a plan. He took everyone off site and lined them up along Madras St, facing the building. They would look for signs of life as the diggers removed layers of rubble. Warner watched the watchers. This method found and saved a woman pinned by a filing cabinet and a man, miraculously uninjured.

As night approached the resources on site increased. Fire, volunteer fire and USAR personnel arrived, including some from outside Christchurch, and civilian contractors supplied gear such as diggers, concrete cutters and listening equipment. At one point a head count of rescuers surpassed 200. Problems persisted. The bisected rescue approach meant people on the east side of the site didn’t know there was listening equipment and a concrete cutter on the west side. Contact was made with a group of six survivors - Tamara Cvetanova, Ezra Medalle, Jessie Redoble, Rika Hyuga, Rhea Sumalpong and Emmabelle Anoba - who had been in the King’s Education school. Two men - Luke Pickering, who owned a listening device, and Usar technician Ian Penn - drew responses to their tapping during a quiet period to listen for survivors just after midnight. Pickering also heard a woman’s voice, but couldn’t tell who it was. None of the six collapse survivors made it out alive. Cvetanova died of massive crush injuries and Anoba from the effects of the fire. The causes of death for the others were never ascertained.

Less than a month after the earthquake, Prime Minister John Key announced the Canterbury Earthquakes Royal Commission would examine building collapses that killed people and make recommendations on building standards.

One of those was the Pyne Gould Corporation (PGC) building on Cambridge Tce, where 18 people died in a partial collapse. The other was CTV. It accounted for nearly two-thirds of the entire earthquake death toll.

The commission ran for nearly a year over 2011 and 2012. Nearly two months was given over to CTV. No other building took more than a week.

By the time the commission began hearing CTV evidence in June 2012 the building had already been subject to much scrutiny. The Department of Building and Housing (DBH) had commissioned a report from structural engineers Clark Hyland and Ashley Smith, which concluded there was little structural damage before February 22 and the collapse was due to substandard design. The commission also convened a panel of engineering experts to discuss the causes and consider a collapse scenario based on computer modelling. For eight weeks, the airy church hall of St Teresa’s parish in the Christchurch suburb of Riccarton considered the worst part of New Zealand’s deadliest disaster in 30 years.

“The concurrence of the eyewitness views on this strongly suggests that the collapse was almost immediate,” counsel assisting the commission Stephen Mills said in his smooth baritone, during opening remarks.

More than 80 witnesses were called. Kulpe, Mitchell and other collapse survivors recounted their ordeal. Many of them asked for TV cameras to be turned away. Coatsworth, the engineer who inspected the building between September and February, fought back tears as he described replaying his role over and over again in his mind. Construction manager Gerald Shirtcliff refused to co-operate with the commission at first, not even revealing where in Australia he was living. He relented and appeared via video link from Brisbane to confirm his professional ineptitude and make a hopeless mess of trying to conceal where he did and did not live under his second identity, William Fisher. The Press later revealed the extraordinary scope of his deception. Shirtcliff had assumed the name of a former colleague to trade off the real William Fisher’s engineering degree and pass as an expert. He was swiftly sacked from his job in Brisbane, had his degree revoked, and his Engineering Australia membership cancelled.

Shirtcliff was a sideshow. The crux of the inquiry arrived when two key witnesses, Reay and Harding, gave evidence. Reay was pensive, deliberate on the stand, allowing long pauses to fill the hall while he formulated his answers. He had dismissed the Hyland-Smith report as “technically inadequate” upon its release, citing problems including concrete strength calculations (later proven), retention of building material as evidence and inadequate computer modelling of the collapse. The cumulative effects of aftershocks should have been considered, he said, along with the extreme vertical acceleration of the February 22 earthquake and the stability of the southern wall. He proffered five alternative collapse scenarios, which, under questioning from Mills, he was forced to concede were “just a series of possibilities that occurred to you”.

“Am I right that they all have in common the fact that none of them attribute any responsibility to you or your firm for the collapse of the building?” Mills asked.

Harding was a solemn witness. Less defensive; more matter-of-fact. He disagreed that several structural elements were non-compliant and cited vertical acceleration as a cause too, arguing the force that day was an impossible strain on the CTV building’s columns.

It was about all he and Ray agreed on. Each’s testimony served only to implicate the other for the building’s flaws. Reay said Harding’s work finishing off a multi-storey design John Henry had started gave him “a very good start” using the ETABS computer program, and he couldn’t recall Harding raising any problems. If his employee was uncertain about that, or any part of the design, he was capable of flagging his own limitations. Harding reiterated his belief that Reay was reviewing his work, as he had no experience on multi-storey buildings. “If I had thought I was doing it by myself I would have bailed out right then,” he said.

The commission unearthed myriad failures in the construction phase. The concrete surfaces where the columns and beams met should have been roughened, but weren’t. This was partly a design issue but could have been easily solved in practice. The oversight was pivotal: one of four key contributors to collapse the commissioners would identify.

Potentially more serious was the curious case of the bars that were supposed to link the concrete beams to the west side of the north shear wall. Instead of being embedded in the wall, the reinforcing bars inside in the beams were bent back onto themselves, coming to rest hard up against the wall. It was an odd defect. To bend the bars manually would have taken huge mechanical strength and substantial counterintuition, so it was more likely a manufacturing error. And for some reason the bars on Level 2 alone were inserted correctly. In any case, the joint was poorly detailed by Harding - the bars didn’t extend far enough into the wall.

Much of the evidence of the building’s flaws existed only because of three men. Between the evening of February 22 and the early hours of February 24, engineers Graham Frost, Rob Heywood, and John Trowsdale were called to the CTV site to advise Usar on the risks involved in working on a collapsed building. The trio soon realised the scene would need to be investigated and were disturbed to see debris being piled onto Madras St, or removed from the site altogether. They set about cataloguing the structural damage, taking hundreds of photographs, preserving debris where possible, and examining building materials after they had been taken to the Burwood landfill. Their investigations identified numerous faults including the smooth surfaces of the beam-column joint faces and poorly-reinforced columns.

“There was no formal system whereby this information was collected,” the commissioners wrote in their findings. “The Royal Commission commends these engineers for their very thorough documentation and assessment of the collapse debris . . . [their] public-spirited initiative . . . created an excellent record of the state of the building and individual elements following collapse.”

The commission retained its own experts to examine what Frost, Heywood and Trowsdale had found, review the DBH report and confer on the computer modelling of collapse scenarios. Panel members disagreed on some of the input data for the program which ran seismic tests - called a non-linear time history analysis - and how to interpret the results, but after two tests were able to put forward a likely collapse sequence based on the findings: First, the drag bars connecting the floors to the north shear wall failed, potentially causing the floors to disconnect from the wall. Then, the inter-storey deflection was too much for the columns to withstand, triggering a loss of “axial load carrying capacity”, meaning the floors started to collapse.

“We have concluded that a primary reason for [the non-compliance] was that Mr Harding was working beyond his competence in designing this building,” the commissioners’ report read.

“He should have recognised this himself, given that the requirements of the design took him well beyond his previous experience.”

In evidence, Harding had described his fate - confronting the collapse of a building he designed - as a “worst nightmare scenario” for an engineer.

He was not alone. Reay should have known better, the commissioners said:

“[We] . . . consider that [he] was aware of Mr Harding’s lack of relevant experience and therefore should have realised that this design was pushing him beyond the limits of his competence.

“Dr Reay should not have left Mr Harding to work unsupervised on the design or without a system in place for reviewing the design, either by himself or someone else qualified to do so.”

There was also the matter of the building permit, approved despite council engineer Graeme Tapper flagging a critical design flaw almost immediately. Reay emerged the key figure. Soon after Tapper raised concerns, the commissioners said, he involved himself in the permit process.

“It is likely that there was a meeting called . . . which resulted in Dr Reay convincing Mr Bluck that the concerns over the building were unfounded. We note that we cannot make any finding about the extent to which Mr Bluck required convincing and, therefore, about the extent to which Dr Reay was influential.”

When he gave evidence, Reay said he had no memory of any meeting or the CTV permit process. The collapse was a “terrible tragedy”, he said.

“I really feel for those who have lost their loved ones . . . I have spent my life working in engineering and have always tried to maintain the highest standards of the profession. I apologise to all the families affected as this building did not meet my standards.”

The 40-year-old mother-of-two was working part-time as a nurse at The Clinic, a medical practice on level 5 of the CTV building, when the quake struck. She and 17 other staff and patients perished. Her father, along with more than 50 victims’ families, has spent the past six years seeking accountability for the 115 deaths.

Tim Elms described his daughter as "one of life's great tryers". A "cheerful person always willing to help others". When she couldn’t get into nursing school because of her grades, she took remedial maths classes to improve her marks. Her children, Henry and Thomas, were aged 2 years and 8 months respectively when she died.

At 12.51pm on February 22, 2011, Elms was delivering census papers in north Canterbury. He felt the shaking, but had no clue about the drama unfolding in the city. He later heard the CTV building had collapsed, and hoped it was one of the days Teresa wasn’t working. When she hadn’t been accounted for by the end of the day, he feared the worst.

Elms made the 160km round trip from his home in Hawarden, North Canterbury, to attend almost every day of the eight-week CTV hearing. He became convinced human error had caused the building to collapse.

“People needed to stand up and say, ‘I got it wrong’,” he said.

“When people don’t do that it makes it worse.”

Elms is eagerly awaiting the outcome of a four-and-half-year police investigation, expected within weeks, and for which it is understood only manslaughter charges may be laid.

Investigation head Detective Superintendent Peter Read has said the case is complex and cannot be rushed. Police hired engineering consultants Beca and ordered extensive soil and foundation testing and computer modelling. They had to determine that a crime had been committed by an individual, Read said, as there was no charge of corporate manslaughter under New Zealand law.

Police finished their report earlier this year. Read said an updated version was delivered to the Christchurch Crown solicitors last week. The next step was Crown Law.

“Once [Crown Law comes back] we’ll have the opinion and we’ll be able to release it. Hopefully before the end of June.”

The CTV collapse, and the Pike River mine disaster that killed 29 men on the West Coast just three months earlier, sparked renewed calls for a corporate manslaughter charge. In 2015, Justice Minister Amy Adams asked Workplace Relations Minister Michael Woodhouse to add the provision into beefed up health and safety legislation, but WorkSafe confirmed it was not included when the act was updated last year. Labour leader Andrew Little, while his party’s justice spokesman in 2012, put forward a private member’s bill for corporate manslaughter to be included in the Crimes Act. It was later withdrawn.

After the royal commission found flaws in Harding’s design and a lack of oversight by Reay, Elms and a group of bereaved CTV families sought to have the pair stripped of their engineering licences.

Harding resigned as a member of the Institution of Professional Engineers New Zealand (Ipenz) in July 2014, shortly before its disciplinary committee was due to consider the families’ complaint against him. The action proceeded and the committee found Harding "misrepresented his competence" in his work, but he escaped censure. Harding, who did not attend, said through his lawyer that his mental and physical health were at an "extremely low ebb". His more than 40-year engineering career was "all but over". He hadn’t undertaken new design work and retired that year.

Ipenz dropped a similar complaint against Reay when he, too, resigned his membership of the professional body. The Government applied to the High Court seeking a judicial review. Building and Housing Minister Nick Smith said at the time the decision was "flawed". "We owe it to the memory of those who lost their lives to hold those responsible to account for the building's flawed design, and to ensure every possible lesson is learnt by the engineering profession to minimise the future risk." A decision on the review has not yet been made. Ipenz has since updated its rules to allow investigatory matters to continue after a resignation.

A separate disciplinary hearing was held over whether or not Reay sufficiently disclosed his involvement with the building when he applied for Chartered Professional Engineer (CPEng) registration in April 2011. The committee dismissed the complaint, saying Reay had not misled his assessors, but noted he had no duty to flag his connection. Reay maintained he disclosed the link, but the committee found any reference “must have been in such vague or obscure terms as to deny the assessors any meaningful opportunity to understand the true position”.

Alan Reay Consultants was renamed Engenium in 2013, while retaining the same premises, contact details and many of the same directors - including Reay’s wife, Barbara. Reay quit as a director but remained on staff.

Just after midnight, a magnitude-7.8 earthquake centred in North Canterbury caused major damage in South Island coastal tourist town Kaikoura and hit Wellington hard. More than 80 buildings in the capital were damaged.

Statistics House, a five-storey office block on the Wellington waterfront, partially collapsed despite being just 11 years old and, its owners said, 90 per cent of the new building standard. A post-earthquake report found elements of its design, coupled with the unusual characteristics of the earthquake, caused it to fail. It was a stark reminder of the effect prolonged shaking can have on even modern buildings, especially given the earthquake wasn’t on one of the myriad faultlines running directly under greater Wellington.

The Canterbury Earthquakes Royal Commission recommended a raft of changes to building standards and in the engineering profession. Many guidelines had already improved in the years since the CTV building’s construction through greater knowledge of the effect of earthquakes on concrete structures and better design and review processes.

Apart from CTV and PGC, Christchurch’s modern structures performed as expected. Although many were later demolished, they were designed to protect life.

The greater danger lies in the thousands of older, unreinforced masonry buildings - many of which are heritage listed and therefore either expensive to upgrade or in need of consent to demolish. Unreinforced masonry is a construction of clay brick, concrete block or natural stone units bound with lime or cement mortar, without any reinforcing elements such as steel reinforcing bars. Falling bricks claimed 40 lives in the February 2011 quake.

The Building (Earthquake-prone Buildings) Amendment Act 2016 comes into force in July. It will clarify the definition of, and criteria for, earthquake-prone buildings and provide for a national register. It will also give councils discretion to halve the timeframe for either strengthening or demolishing earthquake-prone buildings. In Wellington, most must be remediated within 15 years.

Ipenz chief executive Susan Freeman-Greene said the organisation acted swiftly after the royal commission findings to improve engineering industry standards. Member engineers can no longer resign to avoid disciplinary action, the assessment standards for structural engineers have been strengthened and a new code of ethical conduct took effect in July last year. “Individually, each of those areas raises the bar, but together it really strengthens the framework around professionalism, expectations, accountability and standards,” she said. Despite failures like CTV, Freeman-Greene believes the quality of buildings in New Zealand is high. “The infrastructure and buildings did hold up well, both in Christchurch and as we saw in recent events [in November 2016], [but] there are opportunities to learn from every event.”

Emergency services have also improved disaster management procedures, after failures during the CTV rescue effort were identified. Eight people who survived the collapse died before they could be pulled from the rubble.

Because there was a fire in the CTV rubble shortly after collapse, the Fire Service was the lead agency in the rescue. A coroner’s inquest criticised the service’s lack of managerial direction but cleared it of causing the deaths. Transport difficulties meant Auckland and Palmerston North Usar staff were not at disaster sites until about 12 hours after the earthquake. The equipment belonging to the Palmerston North Usar taskforce, who were assigned to the CTV building, did not arrive until 4am on February 23.

Coroner Gordon Matenga made eight recommendations to improve New Zealand's response to any future search and rescue operations, while the Fire Service commissioned two reviews of its own. These resulted in improved training and equipment, and a shakeup of the leadership team in Canterbury. Former national commander Paul Baxter said at the time: “The families want to know that the people they lost were not in vain."

Perhaps the sorriest part of the story of the CTV building is to read the chronology of missed opportunities to save it from itself: poorly designed, poorly managed, poorly built, partially reinforced, subject to unconsented changes of use, limited earthquake inspections and, after it finally, inevitably, collapsed, a mismanaged rescue. A royal commission of inquiry, two coroner’s inquests and numerous institutional reports and reviews catalogued the failures and brought about improvements in building, engineering and emergency response standards, but cast no ultimate blame for the tragedy. Criminal charges from the police investigation would be the closest thing.

Tim Elms wants responsibility apportioned for the death of his daughter and 114 others to be taken. He has resolved to fight for accountability “until every avenue is exhausted”.

“We don’t want to see another tragedy like this.”