CATHY FREEMAN AND THE OLYMPIC CAULDRON – WHAT NEARLY WENT WRONG – ROBERT IRONSIDE

It was truly a global engineering moment. Selected to light the cauldron flame at the Sydney Olympics, Australian athlete Cathy Freeman recalled how the event director was speaking to her through her earpiece saying, “Stand still while it rises above you, hold your position for a few minutes.” Then she heard him say, “We’ve got a bit of a problem Cathy, can you hold just a bit longer.” Then all she heard was swearing. The last minute need for reprograming of the Olympic cauldron at the Opening Ceremony of the Sydney Olympics in September 2000 was a close run thing but kept from the public for years. On Monday 23 February 2026, engineer Robert Ironside who was appointed to oversee the final weeks of the Olympic cauldron testing, addressed The Sydney Institute to explain what happened. From his daily journal of the work done, Robert Ironside can reveal how an ingenious piece of engineering came together at the last minute on the day.

CATHY FREEMAN AND THE OLYMPIC CAULDRON – WHAT NEARLY WENT WRONG

ROBERT IRONSIDE

Thank you to Gerard and my cousin Anne Henderson.

A couple of months ago I met up with David Richmond and Hon Michael Knight over coffee in a café in Crows Nest. We discussed the opening night and compared “war stories” in the lead up to the opening ceremony. Both David and Michael have wonderful stories in their own right, but David emphasised that my story needed to be out there. He and Michael both said that the best option for me would be to approach The Sydney Institute to act as the conduit, particularly since I had a bit of a connection there.

On 18 July 2000, as an employee of Rolls-Royce Australia Limited, I received a phone call from Wayne Burgess of Colin Ging and Partners requesting some assistance in the commissioning of the cauldron.

At a meeting with the Olympic Coordination Authority (OCA), I was told that the project was approximately six months late and should have been fully commissioned in March and April. At that stage they had less than two months to go before the opening ceremony and they had been commissioning for about two weeks and had achieved very little.

I went back to my superior and told him that the concept was extremely complex, was six months late and was never going to fly. I was told by him that I had to make it work. Rolls-Royce accepted the order to assist OCA with the commissioning of the cauldron and expected me to step up to the plate.

Initially, I was asked to provide a couple of week’s consulting services on how the thing should be commissioned. Within a week, I was made the Commissioning Manager with the responsibility of directing and coordinating all players to achieve the ultimate objective.

The status of the project when I started was that the system was basically installed with a few minor outstanding bits that had to be done. Commissioning had only just begun, and the control system had yet to be verified and debugged.

Work had to be done at night for security and secrecy reasons. In addition, we had to work in an operating stadium, which had functions proceeding throughout the commissioning period. In general, we would work from about 18:00 until about 05:00. After that I would go home and write up my report, which I would then email to various parties. By the time I had finished my day it was usually 07:30 or 08:30. I was generally up at lunchtime and often in the OCA offices by 13:00, particularly over the last few weeks. On the last three days I worked about 21 hours each day.

Key personnel

There were a lot of players in the game, all with differing responsibilities and agendas. The original idea came from Ric Birch. Michael Scott-Mitchell, after whom the project was named, then further developed it. Richard Wherrett directed it. David Thierney and Partners, in conjunction with Watkins Godwin, did initial engineering design sufficient for specification purposes. Australian Winch and Haulage (AWH) won the order for the main contract and, as part of that, did the detailed engineering design, whose review was done by David Thierney and Partners (civil engineers). AGL had the contract for the Ring-of-Fire (ROF) and FCT won the contract for the main burner. AWH had the job of coordinating the other two main sub-contractors. They also fabricated everything except the main burner and facilitated the installation of the whole thing. Richard M Baird and Associates were engaged to check on safety and reliability of the system. Finally, Rolls-Royce Australia was engaged to look after the commissioning management.

OCA was the project management organisation for the Olympics. They managed the contracts that had been let by SOCOG. In this case, SOCOG Ceremonies had let the cauldron contract to AWH.

The main people that I had to deal with were:

1. Brian Hemsworth (CEO and owner of AWH);
2. Ian D’Arth (Engineering Manager for AWH);
3. Joe Cronin (RMB).
4. Peter Tait (Controltec – sub-contracted to AWH – wrote all the control software for the job and looked after electrical design and installation, and was amongst the few who had an appreciation of process engineering and commissioning);
5. David Harrold (sub-contracted to AWH) – He is not on the slide but did the detailed electrical design, worked closely with Peter Tait and was a key member of the commissioning team;
6. Hans Foertsch and the AGL team (AGL);
7. Leon Daych and the FCT team (FCT);
8. Morris Lyda –Ceremonies Technical Director (SOCOG);
9. Peter Labza – Stage Manager (SOCOG),
10. David Atkins – Artistic Director (SOCOG),
11. Colin Ging – Development Adviser (OCA),
12. Terry Andrews – Cauldron Project Manager (OCA),
13. Brett Henderson and Bob Treharne (NSW Police).

Apart from the above there were many others that we had to deal with. These ranged from the accreditation people, the electrical contractors with the Stadium, SOBO, ACA, camera and lighting people, etc. During the last stages of commissioning and on the night, the main team members were:

AWH Team

Ian D’Arth

Peter Tait

David Harrold

Adrian Hammond

Greg Walkden

Jorge Montecinos

Darren Grech

Richard Cannon

Ian Wood

FCT Team

Leon Daych

Michael Steer

Stephen Hill

AGL Team

Hans Foertsch

Andrew Bernotas

Mario Tritsiniotis

Richard M Baird Representative

Joe Cronin

SOCOG Ceremonies Stage Manager

Peter Labza

Athlete

Cathy Freeman

OCA Representative

Rob Ironside

Commissioning highlights

On 28 July we had a successful disengagement and re-engagement of the cauldron from the handle but could have killed two people when some of the cross fencing that had to be installed after the run, came loose and careered down the stand. On 3 August, we faced the third occasion that the cradle rise sequence failed due to an apparent failure of the PLC to pick up a proximity switch that allowed the sequence to continue from ROF low fire to medium fire. This stopped the sequence at the top of the cradle lift and the ROF went out on each occasion. On the same night a new limit switch was installed to stop the cradle on its downward path and to start the trolley on its ride up the ramp. This limit switch and its associated cable and PLC input gained notoriety on the opening night.

This brings us to 7 August when the cradle sequence again failed due to the apparent PLC non-detection of medium fire proximity switch. Two days later, several incidents occurred. The trolley crashed into its end stops, which resulted in misalignment with the cradle on the way down. The communications umbilical snapped and delayed operations until we could reconnect it.

On 14 August, the back-up ignition system could not ignite the ROF. Glitches in the programable logic controller (PLC) were discovered and fixed. David Atkins began to talk about altering the ROF ignition sequence. The pond still leaked significantly.

A day later, there was a near catastrophe. Four proximity probes whose purpose was to detect that the mast had successfully engaged with the cauldron came on with the heat from the ROF. This was only detected when the mast-rise sequence stopped prematurely. These proximity switches acted as permissive conditions for the cauldron to detach from the handle.

We knew that all four were on for sufficient time at one stage for the disengagement process to activate. However, the pins jammed, fortunately. Otherwise, the cauldron would have careered down the stand and probably have finished up in the middle of the field of play after taking out up to ten people and destroying the track.

One of the many issues that we had to deal with was security and secrecy. David Atkins held me personally responsible for ensuring that the guts of the sequence did not find its way to the press and would regularly chew my ear off if anything ended up in the paper. Fortunately, nothing of significance occurred, which was a real credit to the police.

On 20 August, we inspected the base of the mast and found that the gas umbilical pipes, which had natural gas at 1000 kPa, had been severely damaged due to fouling at the bottom of the mast when the mast was lowered to its stops. These had to be repaired and re-tested before we could operate the main burner again. All five umbilical pipes were affected, the three pilots, the main burner and the axial burners.

Four days later, the performance tests failed.  We tried to recover the situation when the mast failed to rise properly on the first performance test, but our attempt resulted in pushing the mast up too far into the cauldron, fortunately without causing any damage. These tests were done in front of senior SOCOG, OCA and other personnel and were very embarrassing.

In a repeat of the performance tests done on the previous day, the next day the trolley crashed into the mast due to the mast creeping up without people noticing. This occurred because of a knock-on effect when trying to get the mast to rise more quickly. To try and overcome this, the hydraulics “experts” were brought in, but it wasn’t all that successful. However, we managed to solve the problem by having the mast supported by the hydraulics, rather than by its bottom stops.

By 30 August, after a couple of days of unsatisfactory results with ROF testing, changes were made to the ignition sequence of the cauldron. These were very significant in that the LPG was now to light first while under the water with the athlete’s torch – the low fire setting was set very low and the solenoid valves to the side of the LPG ring were closed; the hydrogen settings were also set very low, and this now did not start until five seconds after the LPG was lit; the cauldron was now held under the water after ignition for a few seconds and when it lifted, the curtain air did not come on at all unless activated manually in an emergency. These settings satisfied the look that SOCOG was after.

The same day and on several other occasions, there were failures with one of the water pump couplings and with the roller bearings on the cradle. In addition, the water leaks from the bottom of the pond were getting worse.

With just two weeks to the Opening Ceremony, the changes that were made to the ignition sequence had another knock-on effect. The air hose quick-release coupling failed to release due to locked in air pressure. Changes now had to be made to the air dumping arrangements to ensure that this now worked.

There were still many outstanding items to complete. These included raising and painting the steppingstones (these for Cathy Freeman to walk through the water), completing the athlete egress system, painting the island, performing a four-hour run on the main burner – until then the burner had only been run for a couple of minutes a time due to security worries. In addition, we needed to replace the igniter detector wiring (the detector signals were so low that they were close to the igniter trip point, and the cable needed to be replaced with a different type and shortened in length), construct a control room humpy at the top of the stand – for Peter Tait and me – and install a remote alarm panel in the security control room for remote flame monitoring during the games.

On 6 September, we completed a successful demonstration for the Hon Michael Knight, Minister for the Olympics. The only incident that occurred was that the umbilical, which is used to drain air from the ROF during its purge, failed to detach properly and travelled up with the cauldron.

After that, we fully commissioned the backup radio link between the PLCs and ran the sequence all the way to the top with the umbilical line disconnected. This proved to be a very fortuitous decision. On the same day, I had a run-in with Brian Hemsworth over operating personnel. I considered that to have new people assigned to the operation of the system at this point would have added unnecessary risk.

I also prevented other last minute attempts to change plans such as AGL making unauthorised changes to the air system to try and use the, now defunct, athlete air as a means of blowing air to billow up the dress and train of Julie Anthony. This was given the royal order.

Three days before the event, the athlete egress system was finally being built and on the eve of the Opening Ceremony, unbelievably, a backup generator, automatic switching system and associated cabling were installed on MCC’s 1 and 2, despite my protests. All this did was add a degree of unreliability. I told them that they had to be run manually on the night. In addition, the stage opening system (consisting of a Heath Robinson affair with manual winches and cabling) was installed at the back of the stage. Curtains were also added on the day. Both Peter Tait and I were very concerned with these late additions.

At this stage, a television camera was installed on the trolley. My belief is that this could have resulted in the glitch seen on the day although Peter believes that it was the curtains.

Athlete testing was conducted on the evening of 14 September. Several tests were conducted whereby Cathy Freeman would ignite the flame and the ROF would rise to the top of the lift. The sequence was deliberately stopped each time immediately before the trolley began to travel up the ramp. During these tests, the water torrent was in operation.

On the last trial I had noticed that the lift did not start to descend and advised Peter Tait that something was wrong. We then conducted a further test, albeit without the water torrent, and the sequence worked perfectly. We finished at about 0500 on 15 September.

The day of the Opening Ceremony

Our team arrived on site at approximately 0800. While conducting planned pre-checks of the system, Peter Tait noticed that the backup telemetry system was experiencing interference from possible rogue transmitters. He managed to contact ABA and SOBO who then came to see us and subsequently found and shut down a pirate television transmitter that was among the crowd near us at the top of the stand.

The sequence started. Cathy Freeman entered carrying the torch. She skilfully ran up three flights of stairs, the torch held high. She walked through the steps in the shallow water and lit the ROF. All was operating correctly during the initial stages. The cauldron was lit and lifted out of the water and the handle transferred from the cradle to the trolley. The trolley then began its upward path before the cradle moved down and out of the way, which resulted in a collision between the two.

Our problem was that a limit switch, used as a permissive to allow the trolley to begin its travel, was incorrectly indicating that the cradle was out of the way. After this, it would be Peter Tait who saved the day. Peter quickly determined that the limit switch input was on, although the cradle had not begun its downward travel. Because the sequence had moved to the next step, it was not possible to change the step logic to move the cradle down, irrespective of the status of the switch.

What Peter had to do was to break into the final output to the double acting rams that drove the cradle lift and then to input -100 per cent manually – which drove the cradle down.  He achieved this after 3 minutes and 42 seconds of investigation and determining how to fix it, while several billion people around the world were watching. Not a bad effort!

Those watching – as the video of the night shows – could sense there was a hitch. But throughout it, Cathy Freeman remained calm holding aloft the torch as if this was part of the act. Her youth and athletic training backing up a weighty ordeal. Below her, the team watched anxiously as Tait reprogrammed the operation.

The trolley then began to travel after the lift was lowered. The impact severed one of the critical communications cables, which meant that the sequence was then operating entirely on the back-up telemetry system. The sequence performed flawlessly on the back-up system from then on.

The Opening Ceremony was the first time that the cauldron team had been able to recover from an interruption/aberration within the sequence – all previous attempts had resulted in the sequencing being aborted. The problem may have occurred during Cathy Freeman’s training.

The offending limit switch had been checked to work correctly on the day of the Opening Ceremony, as part of the system pre-checks. While the cauldron was held stationary at the bottom of the ramp, we had less than 10 minutes of gas to keep the flame going.

The disciplines that had been put in place worked well and none of the actual cauldron team panicked, although others did. The only person that could get the system going again (Peter Tait) managed to do so.

The aftermath

Peter and I differ in opinion on what caused the switch to activate.  Peter believes that it was because the curtains installed the previous day had flicked the switch when fluttering in the wind.  I think that the cable from the switch might have been damaged when the trolley camera was installed, exposing the conductors which could have short circuited when they came in contact with water. Both scenarios are plausible and both are the result of late changes the day before.

The gas bottles that kept the ROF alight until the main burner was started were found to be empty the following day, which meant that the ROF remained alight on the remaining vapour in the gas pipes.

Due to a dispute between OCA and AWH there was no-one to look after the cauldron during the games. I volunteered to do this at no cost and subsequently kept the flame going throughout the games, having taken annual leave to do so. I also designed, commissioned and conducted the cauldron shut-down sequence for the closing ceremony.

The story of what happened at the Opening Ceremony was kept a secret for a long time. Some details have trickled out but the full story is told here – thanks to the work reports I had made of daily progress as the commissioning of the cauldron was undertaken.

The whole enterprise was a remarkable achievement of engineering design and implementation, but it could only have been achieved in the weeks our gifted team spent testing and correcting the enterprise in preparation for the Opening Ceremony.