At a time when NSW and Queensland are again being impacted by severe and destructive bushfires, I can report on a NSW deputy state coroner’s findings following an Inquiry into the ‘Fire at ‘Flagview South’, Sir Ivan Dougherty Drive, Leadville February 2017 (the Sir Ivan fire).

The fire started on 11 February 2017 and was finally declared ‘out’ on 6 March 2017 ([1]).  The coroner reported (at [2]) that:

The Commissioner of the NSW Rural Fire Service, Mr Shane Fitzsimmons (‘the Commissioner’), stated that the fire burnt during the worse fire weather conditions ever recorded in NSW. The fire consumed 55,372 hectares of land, destroyed 35 houses and damaged 11 others. People suffered huge losses; together with the loss of their homes, there was also loss of sheds, farming machinery, fencing and agricultural land. Over 4,700 sheep were lost, 440 cattle, 4 horses, 325 goats, 35 poultry, 4 alpacas and 20 domestic animals. Miraculously, no persons were killed or seriously injured as a result of the fire. The emotional, physical and financial impact on the people who suffered losses and assisted in fighting the fire should not be underestimated.

Mr Fitzsimmons had requested an inquiry because (at [5]), in his view, there is:

… there is a need for a broader understanding that in declared catastrophic fire conditions, no home or structure is designed or constructed to withstand such fire conditions. Fires in these rare conditions burn aggressively and simply cannot be controlled.

The NSW Farmers Federation (known as NSW Farmers) also asked the coroner to conduct an inquiry.   They (at [7]):

… expressed the view that farmers affected by this fire have reported feeling a sense of isolation and disengagement from the RFS and that an inquiry would be an opportunity for them to put their story forward and to have their questions answered in a public forum. A particular area of concern was communication breakdowns between the RFS and landholders, which may have played a part in the initial responses and strategic management of the fire.

The coroner reminds readers (at [11]):

A coronial general fire inquiry is intended to be an independent, objective, fair examination of the available evidence that relates to the fire. It is not the function of this inquiry to apportion blame but rather, to make findings as to the origin and cause of the fire, and also to make recommendations that are considered necessary or appropriate as a result of the circumstances surrounding the fire.

The cause of the fire

It was determined (at [99]) that the fire was caused by lightning striking a wooden fence strainer post. This probably occurred some days before the fire leaving the fence post to smoulder until the weather conditions of 11 February ‘were conducive to a fire starting in the grass surrounding the strainer post’ ([93]).

The fire spread and the response

The evidence (at [19] and [20]) was that in the period January to October 2016 there had been significant rainfall in the areas affected by the fires, and this had allowed significant vegetation growth, both pastures and in timber and scrub areas.  At the end of that period there was no opportunity for effective hazard reduction burns prior to the fire in February. An RFS deputy captain gave evidence that there was no opportunity to complete hazard reductions prior to the Sir Ivan fire. He said (at [21]):

I don’t know in what area you would do the hazard reduction, they’re mostly all grazing properties around. No one’s going to … burn the feed, burn the grass. But certainly there’s an operation that needs to take place in the winter months, not spring or summer.

With respect to hazard reduction, the coroner noted (at [18]):

The RFS provides advice to landholders on managing fuel loads on their properties but this is ultimately a matter for the individuals concerned. Healthy pastures are assets in their own right that land holders might be reluctant to reduce in order to try and mitigate fire risk.

Having reviewed the evidence presented, Mr Geoff Conway, an independent court appointed expert in fire investigation said (at [22]):

There was significant variation in the fire prevention and mitigation works undertaken by landholders on those properties impacted or at threat from the fire…some images provided with these statements show areas of fuel reduction around homesteads and farm buildings. These appear to be the exception rather than the rule.

The weather forecast for 11 February 2017 was for extreme fire weather conditions.  The forecast for 12 February was for catastrophic fire conditions.  These conditions affected much of the state, not just the area where the Sir Ivan fire started (see [35]).  The coroner said (at [33]-[34]):

This inquiry is conducted with the benefit of hindsight. Those involved in making preparations in advance of 11 February 2017 had no way of knowing if, let alone where, fires might eventuate under the difficult conditions. Indeed there was an earlier fire within the Castlereagh District that started on Saturday morning before the Sir Ivan fire. This was near New Merrigal, west of Gilgandra and crews were deployed. This fire was successfully contained but was at ‘patrol’ status at the time the Sir Ivan fire commenced.

A number of incidents eventuated across the state. Between 10 and 19 February 2017, RFS crews in NSW attended 697 bush/grass fires, including 223 incidents across the weekend of 11-12 February 2017. Over 1,800 fire fighters (the vast majority of them volunteers) were deployed per day between 10 and 19 February, including an estimated 2,500 fire fighters deployed on 12 February 2017.

Around midday on 11 February smoke was observed on the property ‘Flagview South’. Neighbours, RFS brigades and RFS firefighting aircraft attended and extinguished the grass fire however (at [40]-[41]) :

… nearby scrub was already well alight. The fire burnt predominantly east under a westerly wind towards the village of Uarbry. The terrain varied from grassland to undulating rocky hills and thick wooded scrub that is largely inaccessible to vehicles.

The prevailing winds also pushed the fire from ‘Flagview South’ out to the north east. Initial efforts to fight the blaze included attempts to get in front of the fire through a property owned by Mr Coe and mount a direct attack on the fire edge.

The fire behaviour was erratic as it passed through different geography and fuel loads and was affected by wind shifts ([42]-[46]).  At [47]:

Mr Conway, the independent expert, gave evidence that even though the strategies applied in the initial period from detection to about 1900 hours were appropriate, they were not successful. The combination of terrain, vegetation and weather made it unlikely that any strategy would have been successful in containing the fire until weather conditions moderated.

Further (at [50]):

Mr Conway agreed that the strategy adopted on 12 February 2017 was appropriate even though there was a real prospect that it would not be successful because of the terrible conditions.

The fire conditions were indeed terrible and the fire was predicted to form a pyro-convection column ([69]) with the risk of ‘fire thunderstorm events’ ([73]).  At [78]-[83]) the coroner said:

It was the risk of the pyro-cumulonimbus cloud giving rise to a fire thunderstorm over the fireground that prompted a red message from the RFS at about 1705 hours.

Mr Jones [Superintendent and District Manager RFS and appointed Incident Controller] directed the message to all fire fighters working on the fire ground, including those now reporting to the Liverpool Range and Cudgegong Fire Control Centres. In doing so, he was “instructing fire fighters to ensure they had safe refuge, to expect erratic fire behaviour and to work from a safe refuge.”

This is the only red message he has had to send in his 25 year career.

Some RFS personnel then communicated the message to local landholders on the fire ground, with mixed results. Some locals thought that the advice to seek safe refuge was ridiculous in circumstances where they were fighting to preserve their homes and livelihood.

In these circumstances, individual landholders were entitled to assume their own risk and disregard the warnings if they chose. Many with no experience of a fire of this magnitude made decisions based upon their previous experience in smaller fires and thought that the RFS was overreacting.

The RFS warnings were however necessary and appropriate, and the RFS would be subject to criticism if they had issued warnings to their members which were not then passed onto other people on the fire ground.

At [87]:

Mr Conway observed that the implementation of the IAP was:

“…as effective as the circumstances allowed given the resources available. Protection of life was identified as a priority…the protection of property in the fire conditions that were present on the Sunday was always going to be challenging. The efforts of firefighters in these circumstances only plays a small part in the success of asset protection tactics. The ability for firefighters to access properties, fuel loads and fuel conditions around the assets, and any fire protection measures taken by residents prior to the arrival of the fire front contribute significantly to the fire suppression effort.”

Criticism of the RFS response

These were summarised at [100]:

Various local landholders expressed concerns by way of written statements and oral evidence that the RFS is disengaged with the local community. The primary concerns raised were; that when the RFS took control of the fire to Coonabarabran that control was too removed from what was happening on the fire ground and too removed to receive input from local landholders who were also involved in fighting the fire and who had important local knowledge and experience. A particular area of concern was communication breakdowns between the RFS and landholders that may have played a part in the initial responses and strategic management of the fire.

The coroner said (at 102]-[104]):

I am satisfied that once the Sir Ivan fire had escaped into scrub in difficult terrain on the afternoon of 11 February 2017, no one on the fire ground had a clear view of the circumference of the fire. It was too big to see from a single vantage point. Decision makers needed to communicate with each other to try and appreciate what was happening at different parts of the fire ground.

The Fire Control Centre needed to be big enough to provide (amongst other things) a number of fixed phone lines, a backup exchange in the event of power loss, radios, computer terminals and internet access, display boards for maps, electronic display facilities (including as at the time of hearing, footage showing a birds eye view over the fire ground), room for representatives from Police and Local Government, and room for volunteer members organising food and facilities for RFS crews (including out of area crews).

A fire of this size, where crews were operating out of line of sight from other crews, where there were considerable communication difficulties across the fire ground because of a ridge that effectively separated the area into a northern and southern division on 11 to 12 February 2017, where there were out of area crews fighting in unfamiliar territory, where there were unknown numbers of local private crews working on the fire ground, and where conditions were predicted to become much worse on 12 February 2017, the people on the ground were not necessarily the best informed to make decisions about anything other than the fire immediately in front of them. There needed to be a hub that could collate information coming in from different parts of the fire ground and step back to look at the bigger picture including resources and predicted weather.

Whilst local knowledge is important it can be influenced both by the limited perspective and the need for those with local knowledge to protect their own interests.

This is demonstrated [at [113]-[115]] by the evidence of Mr Fergusson [a Dunedoo resident], where he said that at the time when he was asking for a back burn to be lit on his property he was unaware, for instance, of the location of fire crews on the ground. Mr Fergusson said “I was doing my own little patch. I wasn’t in charge of the fire. I didn’t want to know what everyone else was doing.”

When asked whether, in making a decision about introducing a back burn, it would be important to understand where the crews were Mr Fergusson said “not from my perspective.”

This illustrates the importance of having someone other than local landholders making decisions about when to introduce a back burn on a fire of this scale. That decision maker needs to be impartial to the extent he or she makes decisions without looking to minimise, for example, loss of country belonging to family or friends. That person needs to have situational awareness across the fire in terms of where crews are located and how terrain might hamper or assist efforts. That person needs information about predicted weather and feedback from the fire ground as to actual conditions.

The local landholders say that the decision maker should be present on the fire ground. The RFS say that in a fire of this size, that person should be in the Fire Control Centre.

With respect to a back burn, the IC was using graders to establish wide fire breaks before authorising local firefighters to light a back burn.  At [155]-[157] the conflict between landholders and the RFS is described:

Mr Fergusson wanted the RFS to “light it up”, including lighting up “the whole scrub …with an incendiary helicopter at midnight.” This was a view apparently shared by other local landholders with their own personal history in fighting fires on this country.

According to Mr Conway “burning out operations should be commenced as early as possible once control line construction commences and weather conditions are suitable. It should be undertaken slowly and incrementally, keeping the fire well back from the point where control lines might still be under construction.”

However, Mr Conway did not support the use of incendiaries or “lighting it up” because “introducing more fire into the landscape in these conditions is a strategy that must be very carefully considered, and only undertaken with control lines in place and sufficient resources to deal with any fire that may start outside the control lines due to spotting or failure of those control lines.”

Another concern was the communication between RFS brigades and locals.  At Cassilis ([160]-[162]):

A common theme … was that there was a lack of any RFS presence at important points.

Local landholders alleged that on occasions, RFS crews were present but did not offer adequate assistance or engage with locals to tell them what was planned and the potential ramifications of those plans.

One matter of obvious significance to the people at this end of the fire ground was a RFS decision on 12 February 2017 to prioritise protecting the village of Cassilis at the expense of deploying appliances to properties that stood between the approaching fire and the village.

The fire did not reach Cassilis but that outcome could not have been known when the decision was made to prioritise the village. Fire spread predictions did indicate that the village was likely to be impacted by the fire. At [170] the coroner said:

The decision to deploy the strike teams to Cassilis took into account that:

a. People had been earlier advised to evacuate to Cassilis. Precise numbers are not known but Mr Jones estimates around 40 people travelled there for that purpose, in addition to any residents.

b. There was greater infrastructure to protect there (including a greater number of homes) than on surrounding properties.

c. Even if the fire front didn’t reach Cassilis, the main concern was that spot over fires could put the village at risk spotting over the fire breaks that landholders had established on their properties in between the fire front and the village.

d. Cassilis later lost power which meant that residents lost the ability to monitor fire progress via ABC or other communications. This became more difficult once the Telstra tower was lost. Mr Jones sent a third strike team to the town once told that the power was down.

e. A local government representative then advised that loss of electricity would mean that the pumps were out and no reticulated water would be available. Mr Jones sent a fourth strike team to the village in response to this development.

The coroner (at [172]) accepted that ‘the decision to prioritise Cassilis over individual properties was appropriate based upon what was known at the time’.

The coroner’s concluding comments and recommendations

The coroner made a number of concluding comments.  Two that I rate as most significant were his comments at [197]:

Communication and co-ordination between the RFS, the volunteer fighters, and the local landholders proved to be difficult at times. At the conclusion of this inquiry it was agreed between the RFS and the local landholders that improvements could be made. I will make a number of recommendations that the RFS and NSW Farmers agreed upon, given that those two bodies will have primary responsibility for facilitating attempts to improve communication and co-ordination.

And at [199]:

During the course of fighting the fire, it appears that possibly an unauthorised back burn resulted in the unexpected need to relocate firefighting resources and forced the unintended amendment of the Incident Action Plan. The evidence in this inquiry strongly supports the RFS concerns that unauthorised back burning can significantly impact the effectiveness of firefighting operations, the duration of a fire and safety on the fire ground. The RFS would like to see a legislative review of penalties and offence provisions in this regard. That is a law reform proposal that sits beyond the scope of this inquiry. I do however take this opportunity to stress that back burns cannot be undertaken during a total fire ban without the knowledge and consent of the RFS and any unauthorised back burns cannot be condoned.

My summary of those comments is that communication is essential for everyone to understand the other’s perspective, but landholders should not and cannot engage in their own, uncoordinated firefighting efforts.

The coroner made 5 recommendations to the Commissioner of the Rural Fire Service including (at [208]):

That the RFS undertake a community engagement campaign (including information specifically targeted at farming communities) to reflect any revision of the Fire Danger Ratings system following the current review by the National Social Research Project. Such a campaign to include notice that in large fire events the RFS cannot guarantee that every landholder will receive assistance from the RFS and such a campaign to be repeated (even in a modified form) prior to the start of each statutory bush fire danger period.

A further 4 recommendations were made to both the Commissioner and NSW Farmers.   These included:

That the RFS and NSW Farmers consider a joint approach to the Bushfire and Natural Hazards Cooperative Research Centre (or similar organisation) to conduct social research into best developing a “shared responsibility” to hazard reduction, community engagement outside of bushfire season, information sharing around predictions for more extreme fire behaviour, and the delineation of decision making responsibilities on the fire ground when RFS and private vehicles respond to a fire.

That the RFS, in consultation with NSW Farmers, extend and expand primary producer engagement strategies to include a focus on how private landholders within farming communities can work with the RFS, including a focus on information sharing outside of bushfire season, fire ground communication during a fire, fire ground management structure and firefighter safety.

Conclusion

This summary does not do justice to the Coroner’s report, and the Coroner’s report does not do justice to the more than 2000 pages of evidence and submissions and the oral testimony of witnesses.  That process demonstrates the complexity of reviewing these events particularly when there is conflict between those involved.   As the coroner said (at [10]):

Due to the volume of the material, I have referred in these findings only to the issues, evidence and submissions that I consider most significant. This inquiry and to some extent the investigation focussed on examining the fire’s cause and origin and the circumstances in the first two days of the fire that highlighted the issues of concern raised by both the RFS and the local landholders.

In this blog I have further reduced that to refer ‘only to the issues, evidence and submissions that I consider most significant.’

If there is any ‘take home’ message it seems to be the importance of communication both prior to and during the fire season and during active fires, but communication is difficult due to time constraints, the ‘fog of war’ and technological limitations.  Exploring these issues has been the subject of numerous prior findings (see https://www.bnhcrc.com.au/utilisation/ddr) and will no doubt continue to pose complex and difficult problems for fire fighting agencies and those affected or potentially affected by bushfire.