New South Wales Deputy State Coroner HCB Dillon has handed down his findings following an inquiry into the devastating bushfires of 2013 near Coonabarabran in northern New South Wales. This inquiry follows a previous Parliamentary enquiry into the same fires. Before both the parliament and the coroner a group of affected landowners known as the Coonabarabran Property Owners’ Alliance raised issues of concern with the response by the National Parks and Wildlife Service and the New South Wales Rural Fire Service. In the Parliamentary inquiry it was clear that they were alleging that there had been negligence in the response to the fire.
A coroner’s inquiry may be established to conduct an investigation into the cause and origin of a fire. A coroner may also hold a general inquiry into a fire which includes an ‘examination of all of the circumstances concerning the fire …’ (Coroners Act 2009 (NSW) s 30). This inquiry was a ‘general inquiry’ and so extended beyond the issue of cause and origin and included a review of the planning for and response to the fire.
It has been reported that at least one member of the Coonabarabran community had hoped that the coroner would make some finding of negligence in respect of the fire. Stephen Lill, whose farm was damaged in the fire is quoted as saying “It’s put an unbelievable strain on the people of Coonabarabran. Unfortunately we didn’t hear the negligence word mentioned there in court” (Sophie Tarr, ‘Bushfire findings ‘no comfort’ to victims’ 7News (Online) 28 September 2015). It should be no surprise that the word “negligence” did not appear in the coroner’s report. As the coroner says (at page 10):
A coronial fire inquiry is a search for truth. It is intended to be an independent, objective, fair examination of the available evidence relating to the fire. It follows that a proper inquiry is neither a witch-hunt nor a whitewash. One of the general purposes of holding coronial inquiries such as this one is to address the suspicions and fears of members of the community who have been harmed by events such as this one. Civilised societies are concerned about catastrophic harm caused to their members and try to understand those events. This may help those harmed but also enables the community to learn from their traumatic experience.
It is critical to emphasise from the outset that these proceedings have been an inquiry, not a trial or a contest. Although arguments have been put on both sides of various questions and issues, and the Alliance has not resiled from its criticisms of the NPWS, these proceedings are by definition inquisitorial in nature, not adversarial. The primary concern and focus, therefore, in this (and all coronial proceedings) has been to ascertain the facts. For this reason, there are no formal “parties”, as such, in these proceedings, with rights and interests to be adjudicated upon after they present their respective cases. Nor is it the function of this inquiry to apportion blame, although, of course, it is necessary to make findings as to the appropriateness of what was done.
Despite these comforting words it is often the case that inquiries such as this do begin to resemble adversarial trials, a most notable example being the coronial inquiry into the 2003 Canberra fires. In this case, at least judging from the final report, that does not appear to have happened here though that, no doubt, may help explain the frustration of Mr Lill and other affected property owners.
Even so, the coroner was aware of the concerns of the property owners and was conscious of the need to address them. He set out their concerns and summarised the most important ones (at page 9) as ‘… allegations or grievances [that] relate to the preparation for outbreak of the fire; the response to the fire; the back burn operation; and community notification’. The coroner addressed each of those in the course of his report.
The coroner’s conclusions
With respect to the cause and origin of the fire the coroner found there was insufficient evidence to establish either the cause or the exact starting point of the fire.
With respect to prior preparation and hazard reduction in the Warrumbungle National Park the coroner noted that there had been plans to hazard reduction burns dating back to 2012. This had not occurred for various reasons including inappropriate weather conditions. “With the benefit of hindsight” it was found that there probably was an opportunity for the hazard reduction burn back in May 2012 but:
It is difficult … to be critical of the individual officers who, at that time, did not seize it. In their professional views, the conditions were not sufficiently favourable to carry out a successful burn during the whole period. It is true that they showed no particular urgency in searching for windows of opportunity. This was not carelessness on their part. Their approach, in the circumstances as they perceived them to be, is explicable on the basis of their aggregated experience as NPWS officers with local knowledge that catastrophic fires had not previously broken out in the park. Nevertheless, with the benefit of hindsight, it seems that any hazard reduction burns carried out in the window of opportunity … may have been beneficial…
The gist of the coroner’s finding was that the initial response to the fire was inadequate to control it but not an unreasonable response given current practices and experience. The coroner said (at p 36-37):
A lesson to be learned from this and other fires is that history is not a reliable guide to fire prediction or behaviour in the 21st century. Experience and developments in the science are undermining conventional wisdom based on history.
That is a lesson from the fire, but not a criticism of the first responders who applied their prior knowledge and experience of fires in this area.
The next issue was the response to the fires. The coroner noted that the fire was initially dealt with as a “Class I” fire which meant that was controlled locally by National Parks and Wildlife Service staff. It appears the fire was treated as a class I fire until 11am on 13 January 2013 when it was upgraded to a class 3 fire and a declaration was made under s 44 of the Rural Fires Act 1997 (NSW). The effect of this declaration was to put the Commissioner of the Rural Fire Service in charge of all fire fighting operations. A concern raised was the delay in upgrading the fire with the belief that had the fire been upgraded to a ‘Class 2’ fire, or a s 44 declaration made earlier, that would have brought more resources to bare and improved the outcome. The coroner agreed that, with hindsight, it ‘seems a common sense proposition’ that a s 44 declaration should have been made earlier (p 29) but whether this ‘would have made a difference to the outcome is very difficult to say with certainty’ (p 31).
Another concern was the use of a back burn to try and contain the fire. It was suggested that the back burn was put in when conditions were inappropriate and this had led to further spotting and the spread of the fire. The coroner found however that the back burn made no difference to the outcome, neither controlling nor increasing the size of the fire (p 39).
Another issue was community warnings.
One of the major grievances of the Alliance and others who were affected by the fire in the region was that emergency notification came far too late to enable them to take action to save their property. (p 41).
Reference was made to the expert evidence that
… emphasised the importance of the public being pre-prepared during the fire season. [The expert witness] … noted that “during major fires, the time period between fires starting and/or breaking containment lines may be very short. That means that there is typically insufficient time to do fire preparation (eg remove fuels from around assets and, in some situations, relocate to safe zones) once the emergency notification has been made.”
Even so, the coroner did agree that a higher ‘watch and act’ warning should have been given earlier. The problem for the IMT was that the fire prediction modelling did not suggest that the fire would cross a major road in the National Park and threaten private assets. Unfortunately the models were inaccurate. This was the first time the RFS had used the fire prediction system and fire ground managers were not familiar with it or its limitations. The coroner recognised the difficult position that caused for firefighters:
… people on the ground then are caught between two stools: do they rely on their own judgment or do they rely on the prediction? What if the prediction casts doubt on their own judgments? In this, as in other fields, such as psychiatric risk assessment tools, the presumed predictive value of the tool may lead to over-reliance or complacency or, on the other hand, may lead to complete distrust of such tools. Defining what the predictive tools can and cannot do is of critical importance, especially in extreme situations. One of the vital lessons to be learned in this inquiry is that the limits of the predictive tools that are available to fire-fighting services in this country need to be clearly understood by all involved in using them.
In this case when the fire started behaving in a way that was not predicted, eg when it jumped over the road, emergency managers should have considered what they would do if they lost control of the fire and warn potentially affected communities. He said (at p 43):
In my view, once it became clear to the controllers that a direct attack could not, for safety reasons, be carried out the next day, and that an indirect attack would be made in conditions unfavourable for back burning generally, that raised the risk of the fire breaking out. Given the rapidity with which Australian fires can develop into catastrophes, it would have been prudent to issue a Watch and Act alert on the Sunday morning.
It is interesting to compare this finding with that of the inquiry into the 2013 Tasmanian fires (Malcolm Hyde, 2013 Tasmanian Bushfires Inquiry: Volume 1 (Government of Tasmania, 2013)). In that fire the modelling using the Phoenix-Rapidfire system was amazingly accurate. It ‘predicted [the fire] would run south east reaching Dunalley at 3.00pm on 4 January. The fire actually reached Dunalley following this path at 3.25pm on 4 January’ (p 53). This type of modelling was also new for the Tasmanian IMT’s and was not used as a basis to warn the affected community. In that case the Special Investigator was critical of the IMT for not relying on the modelling.
Those two findings can be reconciled. If the modelling suggest that a community is at risk then it is appropriate to warn them. If the modelling does not suggest that a community is at risk but the fire starts behaving in a way that is not predicted, then experience fire managers have to apply their own knowledge and understanding to consider where the fire may go in a ‘worst case’ scenario and warn communities accordingly.
A coroners findings are intended to identify what happened and to make recommendations for the future. In this case the coroner made 23 recommendations. Whether they will be useful I will leave for others to judge.
As a legal commentator I can say that the coroner did not make findings that would give comfort to anyone seeking evidence of negligence. Even if there was an unreasonable failure to properly identify the scale of the fire, to seek a s 44 declaration or to properly manage the back burn, the coroner found that these things did not contribute to the final losses. There may have been an opportunity to issue earlier warnings but the coroner made no finding on whether that would, or would not, have made a difference. Where he found that things could have been managed better by, for example, have pre-positioned resources given the high fire danger (Recommendation 18) , responding to the fire with resources taking into account the fires possible growth rather than current size of the fire (Recommendation 8) and taking a more active approach to warning (Recommendation 11), he found that the actions were understandable, reasonable and done in good faith.
This is how the coroner summarised the situation facing the IMT (at p 39):
On the night of 12-13 January, the NPWS incident controllers had (if they had only known it) the invidious choice of doing virtually nothing or attempting what turned out to be the impossible. Had they pulled off a miraculous save, they would hae been local heroes. Had they attempted nothing, the criticism that would have undoubtedly have followed would have been blistering and perhaps justifiably so.
And further (at p 45) they:
… were facing a situation very few people in the world ever have. Although, with the magnificent 20/20 vision hindsight gives us, it appears to me that a number of aspects of the fire could have been managed better, I have no doubt that the NPWS officers involved in this fire acted with utmost good faith and to the best of their ability in the most difficult of circumstances.
These findings are not binding on any subsequent court should the property owners proceed with a claim for compensation. If that were to occur the evidence would need to be tested again and a court would reach its own conclusions on both the facts and the law.
There is no doubt that property owners who believe the fire was mis-managed and that there was culpable fault by NPWS will be disappointed by the measured nature of this report, fire managers on the other hand, may take some comfort from the understanding of the difficult choices and decisions that had to be made in a dynamic environment.