prescribed fire lessons learned - fs.fed.us · prescribed fire lessons learned ... learning from...

Prescribed Fire Lessons Learned

Escape Prescribed Fire Reviews and

Near Miss Incidents

Initial Impression Report June 29, 2005

Prepared by Deirdre M. Dether

Submitted to Wildland Fire Lessons Learned Center

2

Summary of Escaped Prescribed Fire Reviews and Near Miss Incidents What key lessons have been learned and what knowledge gaps exist?

Introduction

This analysis is the first known attempt to take a comprehensive look at escaped prescribed fire reviews and near misses. A total of 30 prescribed fire escape reviews and ‘near misses’ (see Appendix A and B) were analyzed to discover what, if any reoccurring lessons were being learned, or whether they were indicating emerging knowledge gaps or trends. It is estimated that Federal land management agencies complete between 4,000 and 5,000 prescribed fires annually. Approximately ninetynine percent of those burns were ‘successful’ (in that they did not report escapes or near misses). This can be viewed as an excellent record, especially given the elements of risk and uncertainty associated with prescribed fire. However, that leaves 40 to 50 events annually we should learn from. This report is intended to assist in that effort.

Evaluating formal reviews and After Action Reviews (AAR) can be a tool for burn personnel to expand their knowledge and supplement their own direct experiences. When reviews go beyond policy and accountability questions they can provide information that can add to our own direct experiences by broadening exposure to what can occur. Learning from other experiences may help avoid undesired outcomes. The intent of this report is not to point out ‘wrong decisions’, but rather it is to use all these individual ‘events’ to see if there are common themes and/or ‘weak signals’ occurring with these escapes and near miss events. The main focus of the analysis was to look for things prescribed fire practitioners could use as they prepare for future prescribed fires. Are there some factors that prescribed fire planners and/or burn bosses have been repeating in isolation? If so, what should be shared with others involved in the planning and execution of prescribed burns to continue to improve outcomes?

Methods

Three questions drove this inquiry: Can comparing these reviews allow us to glean important or emerging trends? Can these reviews help all agencies to learn? and Are we asking the right questions? Next, I posed three straightforward questions: of the analysis addresses straightforward questions. Are there common reported ‘causes’ contributing to the escapes/near misses? Are there repeated findings and ‘lessons learned’ cited in the reviews? What are these? Lastly, I looked for potential emerging trends or patterns gleaned from collectively evaluating all the reviews. The trends or patterns may indicate a ‘blind spot’ that was not previously apparent without looking at all the reports together as opposed to an individual basis.

1

The Dataset

Formal and information documentation from four federal land management agencies (FS, BLM, NPS and FWS) was evaluated in this assessment. Unlike the BLM, the FS did not appear to have a consistent formal review and documentation process. Both the FWS and NPS had too few samples collected to determine their questions and processes, even though current policies for these agencies provide a standardized process

I reviewed 30 reports written for escapes or nearmisses that occurred between 1996 and 2004. Only documents submitted to the Wildland Fire Lessons Learned Center (LLC), collected from agency websites by agency personnel, or located in personal ‘collections’ were used in this assessment (see Figure 1). It is unknown where and how many other reports may be available. This is an indicator that knowledge is not commonly retained and shared from these experiences so agencies have a greater likelihood of repeating ‘mistakes.’

Of these 30, most were formal escape prescribed fire reviews, several were draft review documents one was in a power point presentation, and two were After Action Reviews (AAR) of ‘near miss’ incidents. Most often the review was conducted at the request of an agency administrator following agency policy. The two ‘near misses’ were not declared escape prescribed fires, but the After Action Reviews provided valuable insights so were included in this analysis. Figure 1 Number of reviews by year used in this analysis from 1996 to 2004

Number of Reviews and Near Miss Incidents N = 30

0 1 2 3 4 5 6 7 8

1996 1997 1998 1999 2000 2001 2002 2003 2004

Escapes Near Miss

Geographic area covered includes: several regions of the Forest Service including Northern Rockies (R1), Rocky Mountain (R2), Southwest (R3), Intermountain (R4), Pacific Southwest (R5), Southern (R8) and Alaska (R10). Reviews from Department of Interior agencies including BLM District and Field Offices; National Park Service units; and Fish and Wildlife Service were also used (see Figure 2). The states in alphabetical order (and number of reviews from each state) include Alaska (1), Arizona (5), California (3), Colorado (1), Florida (1), Idaho (2), Kansas (1), Montana (1), Nevada (1), New Mexico (3), Oregon (1), South Dakota (1), Utah (6) and Wyoming (3). Several vegetationfuel complexes discussed in the reviews including: ponderosa pine, mixed conifer, and sub alpine fir, pinyon/juniper, chaparral, sagebrush/aspen, oak brush, grass, and activity fuels (slash).

NOTE: It should not be construed that the years are problematic years since not all agencies systematically report escapes or near misses.

2

The escapes or near miss incidents span from February to October. Most of the escapes were ignited in May (7) and June (5). Both near miss incidents occurred in September. Due to the many geographic areas represented in the sample it was not possible to evaluate any trends related to season. The amount of acres planned for ignition ranged from less than 5 acres up to several thousand acres for individual burn blocks. Several of the more recent escapes involved ignitions on multiple largescale burn blocks.

Figure 2 Number of reports by agency used in this analysis.

Incidents Reviewed by Agency

0

2

4

6

8

10

12

14

BLM FWS NPS FS

# Es

cape

s/Near M

isses

Escape Near Miss

NOTE: this analysis evaluated 30 total events, 28 of which were escapes over a nineyear period or an average of 3 per year (less than 8% of the estimated 40 to 50 escapes occurring annually on average). Of the 99% ‘successful’ burns we, as an ‘organization’ have no idea about the ‘near misses’ and ‘successful saves’ that have occurred. Many other escapes and near misses have occurred, but it is unknown how much formal or informal documentation exists from which we could gain experience. Of the thirty reports evaluated, several mentioned other escapes and near miss events, but only one report (one of the near misses) could be connected to another escape prescribed fire reviewed during this analysis. However, several escaped prescribed fire reviews involved multiple landscapesized projects ignited either simultaneously or days apart, so they were covered in one review document.

Background

The concepts of a High Reliability Organization fit with the planning and execution of prescribed fire

A High Reliability Organization (HRO) is one that experiences less than it’s fair share of incidents. Being and acting like a HRO should apply to organizations using prescribed fire to accomplish land management objectives. The use of fire is a highrisk business that operates in a highly variable environment yet needs to produce reliable outcomes. The excerpts provided below fromManaging the Unexpected (Weick and Sutcliffe 2001) are here to set the context. This analysis is an initial attempt at applying the practices and principles and practices of how a HRO would evaluate escaped prescribed fires and near miss incidents.

NOTE: It should not be construed any one agency has more or less escapes than another since escapes and near misses are systematically reported.

3

All High Reliability Organizations Heed Close Calls and Near Misses

(from Keller October 2004)

High Reliability Organizations:

Regard close calls and near misses as a kind of failure that reveals potential danger, rather than as evidence of the organization’s success and ability to avoid danger. They pick up on these potential clues early on—before they become bigger and more consequential.

Know that small things that go wrong are often early warning signals of deepening trouble that provide insight into the health of the whole system.

Treat near misses and errors as information about the health of their systems and try to learn from them.

Are preoccupied with all failures, especially the small ones.

Understand that if you catch problems before they grow bigger, you have more possible ways to deal with them.

There are five practices of a HRO that can be grouped into two functional categories (Weick and Sutcliffe 2001).

Mindful ‘Anticipation’ of the Unexpected Preoccupation with Failure Reluctance to Simplify Sensitivity to Operations

Mindful ‘Containment’ of the Unexpected Commitment to Resilience Deference to Expertise

Results

What questions are being asked in an escape review ?

Since reports were often generated to meet agency direction the questions mostly focused on policy and accountability issues. The most common questions were:

A. Is agency policy, guidance and/or direction adequate? Was is followed?

Initial observation –Most of the reviews determined that good or sound policy and guidance existed. However, when review teams looked at whether policy was followed, the answer was not always a yes. In a few cases, review teams also compared the local level direction relative to national direction. In at least four cases the finding was the same. The local direction was not consistent or was outdated relative to existing national direction at the time of the escape.

6

B. Was the burn plan prepared and executed relative to agency policy? Was it a ‘good’ plan?

Initial observation – Few reviews concluded that burn plans did not meet policy. However most listed several weaknesses or noted parts missing in burn plans. Common areas cited as weak within the burn plans included complexity and risk assessments, and thoroughness of the ignition, holding and contingency plans. Another reoccurring issue was the lack of fire behavior calculations. Sometimes the fuel model did not accurately represent the fuels and potential fire behavior of the burn area. The fuel model selected generally underpredicted potential fire behavior. Another reoccurring problem was that all the fuel types within the burn area were not evaluated and incorporated into the burn plan. Some of the reviews also noted that burn personnel failed at times to follow what was in the burn plan. This includes obtaining spot weather forecasts or monitoring weather and other prescription parameters for the timeframes specified prior to ignition. Not all procedures were followed during execution. Reviews noted that required documentation was often poor or lacking.

C. Was the planning and execution of the prescribed burn done by qualified personnel? Initial observation –In only two of the reviews were questions concerning the qualifications of burn personnel an issue. In one case the burn boss was from another agency so the reviewing agency was uncertain if the individual was qualified and needed to verify qualification documentation with the other agency. In the other case, the question was whether proper certification procedures were followed. However, several of the reviews noted that burn bosses, while ‘qualified’, were often ‘inexperienced’ with the fuel type(s), which contributed to the escape of the prescribed burn.

In addition, I noticed a trend towards including questions beyond policy, accountability and qualifications. Such reviews are moving toward ‘lessons learned’ and what needs to be improved and applied to future projects. Two reports included in this analysis did not involve an escaped prescribed fire, but shared near misses which indicates a movement beyond focusing on accountability.

Other Observations

Other questions raised were developed based on issues that were specific to that particular event. However, it was interesting to note there were some common themes among those areas evaluated by review teams. At least 10 of the 28 escaped fires burned onto private ground. Therefore, one area added on often looked at how well collaboration, communication and coordination occurred with the public and between cooperating agencies.

Not all reviews evaluated the linkage of the environmental document to the burn plan. Often when this was done there were missing mitigations from the NEPA document that were not incorporated into the burn plan.

All stopped at the point of looking at actions beyond the escape although most noted that there was safe and successful transition from the prescribed fire to suppression actions.

7

Agencies are not yet fully behaving as a learning organization escapes and near misses are not systematically and routinely reported, evaluated, and shared.

What are the Common ‘Surprises’?

The ‘surprises’ came in three areas – Fuels and Fire Behavior, Weather and Communication and Coordination. Many prescribed fire practitioners have already experienced one or more of these types of ‘surprises’ possibly all on the same prescribed burn whether the burn escaped or not. Several of the prescribed fire reviews and near misses expressed ‘surprise’ about the fire behavior they saw from the various vegetation fuel complexes. In some cases, the personnel involved with the burn knew to expect something different than what models predicted, but the fire behavior (either rate of spread and/or flame lengths) was not even imaginable. A reoccurring theme mentioned by the reviews was that many escapes occurred because conditions were not ‘normal’ (e.g. periods of drought, warmer and drier than normal). When burns were implemented, burn personnel ‘failed to adjust operational procedures’ to account for the ‘abnormal’ conditions.

Since surprises, expectations and the ability to manage the unexpected are linked together; therefore it is important to focus on what surprised burn personnel, and why. In Managing the UnexpectedWeick and Sutcliffe (2001) state that ‘surprises’ come in many different varieties. Prescribed burn personnel have experienced many of the ‘varieties of surprise’ listed below. Weick and Sutcliffe describe what can occur if we blindly follow expectations and do not update them with new information.

“The continuing search for confirming evidence postpones the realization that something unexpected is developing. If you are slow to realize that things are not the way you expected them to be, the problem worsens and becomes harder to solve. When it finally becomes clear that your expectation is wrong, there may be few options left to resolve the problem.”

Varieties of Surprise

1 st Form – Something appears which you had no expectation, no prior model of the event, no hint that it was coming.

2 nd Form – The issue is recognized, but the direction of the expectation is wrong.

3 rd Form – Occurs when you know what will happen, when it will happen, and in what order, but you discover the timing is off.

4 th Form – Occurs when the expected duration of the event proves to be wrong.

5 th Form – Occur when the problem is expected, but the amplitude is not.

see Chapter 2 pages 3539

8

Surprises in Fuels and Fire Behavior In numerous reviews, the rate of spread, flame length and resulting spotting caused much of the challenge for burn personnel. Fuels are often the source of unexpected or overlooked sources of trouble. One burn boss related this example of unexpected fire behavior in a fuel type involving standing dead pinyon juniper: the trees had been bugkilled with no needles left on the crown yet during burn operations the fire was able to move into the crowns of the standing dead trees and sustain fire spread through the aerial fuels much like a typical crown fire. In this case, an adequate control line stopped the spread of fire so the prescribed fire did not escape. Another example of unexpected fire behavior came from a small pocket of fuels adjacent to the burn area boundary. The small pocket of fuels was not the dominant fuel type within the burn area and was not noted in the burn plan as a potential source of heat and spotting.

Another reoccurring ‘surprise’, reported in several escapes was greater than ‘normal’ fuel loading due to seasonal variation (greater moisture increased fine fuels) or a change in land management activities (area rested from grazing for 2 years prior to the burn actually being implemented). The change of conditions was not captured or discussed in the burn plan nor noted prior to ignition which would have caused a plan to be reevaluated for complexity, risk and/or adequacy of contingency plans.

What are we not ‘seeing clearly’? Are we not appreciating how complex burns are?

The use of natural barriers – change in fuel or vegetation type, moisture gradient either by changing topography or nighttime recovery often failed to check the spread of fire or put the fire out. In several cases changes in fuel or vegetation types were planned to check the spread of fire. In one case, aspen stands were to be used as a natural barrier to fire spread, but the aspen did not check spread as expected because the burn was not implemented during the planned season. In another case, a ‘swamp’ adjacent to the burn area was planned as a natural barrier. However, when the burn was implemented conditions had changed (i.e. the swamp was dry) and this area did not contain the spread. Prior to ignition burn personnel did not check whether this area would stop the spread of fire. With several prescribed burns, nighttime humidity recovery was expected to stop or check the spread of fire. In each case, burn personnel did not gather onsite information to confirm whether the area did experience sufficient humidity recovery. The review teams recommended that prescribed burns relying on this technique should staff the area through the night and monitor onsite conditions.

Surprises in Weather – Surprises in weather often combined with or caused other surprises to occur. A number of reviews stated that ‘drought’ conditions were believed to be a contributing factor. In several cases changes in moisture (increased precipitation) changed the amount of fine fuels present at the time of the burn which was not accounted for in the burn plan. Unexpected winds (strength, duration, etc) were very common contributing factor to many escapes.

Proximity of thunderstorms to burns may be another emerging knowledge gap or indicates a gap in ‘sensemaking’. Burns that reported strong, erratic winds resulting from thunderstorm development nearby were ‘surprised’ by the effect. One case related that the storm was forecasted, they could see the cells developing miles away from the burn area, but felt that since the storm was approximately 30 miles away, they would be ‘ok”’ and proceeded with ignition. Another prescribed burn escaped due to the effects of a

9

nearby thunderstorm. The weather forecast did predict the development of thunderstorm activity, but burn personnel did not recognize the potential impact to the prescribed burn.

Surprises in Communication and Coordination – This was a common theme among the largest and most notable escapes (Lowden, Cerro Grande, Sanford, Cascade II and North Shore Kenai Lake). Other reviews noted concerns when burning adjacent to non agency lands. Two reoccurring themes were lack of proper notification and recommendations for developing agreements with adjacent landowners. Problems with lack of proper notification occurred prior to burning and/or timely notification once the prescribed burn had escaped burn boundaries. Proper notification of the individuals or agencies affected by an escape was often delayed because they did not know whom to contact. Developing relationships and contacts well before ignition followed by notification just prior to burning was a common recommendation. In cases where the prescribed burn escaped onto private ground an agreement with the landowner prior to the burn would have eliminated the need for declaring an escape. The lack of coordination and communication among key burn personnel and assisting/cooperating agencies or units also appeared to be a reoccurring theme in several escapes.

What are the Common ‘Surprises’?

Two patterns were observed and explored in this analysis. The first pattern was the tendency to underrate overall prescribed fire complexity using the NWCG complexity rating system. A second pattern emerged when chronologies of escaped prescribed burns were examined and evaluated for common causal factors. Although not all reports mentioned the overall complexity rating, most did indicate there were problems with how individual elements of the complexity rating system were addressed (e.g., underrated, missing rationale or reasons a particular rating was selected, or was inconsistent with the agency’s policy). This avenue was not explored further to determine if there were elements consistently underrated.

Complexity All agencies evaluated currently use the National Wildfire Coordinating Group (NWCG) complexity rating system (NWCG January 2004 NFES 2474). The complexity rating system consists of 14 elements to evaluate and determine risk, potential consequences and technical difficulty. The rating system is relatively comprehensive, and is designed to aid in selecting the correct level of difficulty. An initial rating is recommended during project planning and development followed by a final rating, which is done during burn plan development. The overall complexity rating was not systematically reported as part of the escaped prescribed fire review process. Half of the burns did not have the overall complexity rating reported although several within this group did note that elements of the complexity rating within the burn plan needed improvements (e.g. rationale for rating missing and if rated ‘high’ mitigations were not included in the burn plan per policy). There was a fairly even distribution between Low, Moderate and High complexity of those that did report the overall complexity rating of the burn (Figure 3). Review teams concluded that many burn plans with an overall rating of Low or Moderate were ‘underrated’ in complexity. That is, they were actually High complexity instead of Low or Moderate (see Figure 3). In several cases, one noted cause for underrating complexity was due to the preparer not following agency direction.

A separate theme occurred when individual burns were rated Low to Moderate but then were implemented at the same time. Review teams noted burns conducted simultaneously would not warrant the same overall rating as the individual burn. In other words, two Low complexity burns implemented at the same time did not necessarily still rate a Low. Also noted in these cases was that burn personnel should have considered

10

changing the level of management oversight when conducting burns simultaneously (i.e. burn boss level switching from a RxB3 to and RxB2). However, the NWCG system specifically states that the “rating system is for a single prescribed fire project”. Figure 3 A comparison of overall complexity ratings for all projects reviewed. One column shows the complexity rating as determine during burn plan development versus the complexity rating determined during the review process.

Overall Complexity Rating Escapes and Near Misses

n=30

0 2 4 6 8 10 12 14 16

Low Moderate High Not Reported

# of Escap

es/Near M

isses

As Planned Determined After

Chronologies – There were two distinct groups in terms of the length of time from test burn to escape declaration. Prescribed burns either escaped very quickly or escapes occurred several days after the main ignition period while in the patrol and monitoring phase. The chronologies of at least 12 of the 28 escapes indicate that from the time of the ‘test burn’ or ignition it took 6 hours or less before they were declared an escape or should have been declared an escape (see Tables 13). Several more occurred before the ignition phase was complete. Weather changes (winds increased and/or shifted direction and relative humidity dropped over short timeframes) were associated with these events, leading to spotting. However, there are other factors often associated with these events. Many reviews indicated the fire behavior was more than expected or anticipated and burn personnel did not fully realize what kind of fire behavior to expect. Also, the fuel complex either inside or outside the planned boundary caused unexpected fire behavior that was often not addressed in the burn plan. Another connected reason for burn personnel being mislead about what kind of fire behavior to expect was the incorrect selection of fuel model(s) during burn plan development. Fuels models selected for burn plan development often underrepresented fire behavior.

Another factor mentioned several times was lighting at the upper end of the prescription which caused prescription parameters to be exceeded often during the peak of the day. However, it was not always a ‘surprise’ to burn personnel that the prescription parameters would be exceeded during the ignition period. Sometimes burn personnel would start earlier in the morning in an attempt to compensate for the expected trend in conditions. However, either the conditions occurred sooner than expected or delays in implementation caused the ignition phase to be ongoing when the conditions exceeded the prescription parameters. Burns were sometimes lit outside of prescription parameters.

A final factor was the test fire was not conducted in a representative location. Test fires were conducted in locations that were in cooler or moister locations or conducted in fuels

11

with a different kind of fire behavior than the burn area. At the selected locations fire behavior was lower (lower flame lengths or rate of spread) misleading burn personnel.

Commonalities of prescribed fires that were declared an escape the day of ignition The vegetationfuel complex played the biggest role as the immediate causal factor in escapes. Several of the escapes noted that increases in fine fuel loading due to seasonal variation ‘surprised’ them. Burn plans were prepared assuming ‘normal’ fuel loading so during execution burn personnel may not have accounted for this influence. One burn boss did note the changed conditions and made some adjustments to holding forces to compensate. However, other factors including not fully appreciating the influences of the fuel type still resulted in an escape. In several cases, the escape occurred during the ‘test fire’ phase.

Another noted factor was that spotting occurred early and/or frequently during ignition phase, and in some cases during the test fire or blacklining phase. Spotting, according to some burn bosses, is a common occurrence with prescribed burning.

“A thorough recon of the area surrounding a burn unit is invaluable. Think about the worst case scenario, and then imagine the worst case going bad, then go back and plan your contingency”.

Lesson learned from an escape fire review

Commonalities of prescribed fires declared an escape during the patrol and monitoring phase Weather was commonly reported to have gotten progressively warmer and drier prior to escape. The reviews often cited that the weather was known to be ‘more than normal’ for the time of year. Increased winds or a wind ‘event’ that increased wind speeds for a short period of time contributed to the burn being pushed outside of the allowable burn area. In most cases, the burns were patrolled on a daily basis. In burns of longer duration the patrols noted activity increasing in the burn unit (smokes or open flaming of fuels). In some cases the patrols noted ‘smokes’, but thought they would not threaten boundaries. In some cases, personnel knew other prescribed burns had recently escaped within their geographical area. However, in spite of these ‘signals’ there were no changes in actions such as altering mopup protocols or utilizing heat detecting equipment, etc. In some cases, not having someone directly assigned as the lead for a prescribed fire until declared out caused lapses in awareness of the situation and direction to change procedures. One review provided six useful signals that may indicate conditions are not normal and suggest changes to operational procedures.

1) No significant precipitation in nearly two months. 2) Receiving severity funding just prior to igniting a burn. 3) Fire restrictions have just been lifted for your area. 4) Thousand hour fuels are at or below critical levels (or other levels that indicate

they are available to burn) 5) There is a pattern of below normal moisture (precipitation) for more than one

year. 6) Trends for dead and live fuel moisture are they at or below longterm averages.

12

Go/No Go It is easier to light a burn than not to light one.

It is easy to let the ‘pressure to produce’ override the signals (ignore them or don’t look for them) indicating that a burn may not be best executed that day or even that year.

Other Observations of Repeated Lessons Learned and Recommendations.

Many escapes began to take place well before the first spot or slopover. A repeated recommendation for future prescribed burns started with project design and the environmental assessment. Burn block boundaries that were not developed based on known fire behavior characteristics were often a contributing factor in the escape. In some cases, resource specialists selected areas without input for logical or realistic control points. This limits options to successfully implement a prescribed burn or delayed a burn because changes needed to be analyzed and disclosed in an environmental document. Having expertise in fire behavior and practical experience with prescribed fire will help resource specialists to develop ‘logical’ burn blocks.

Review teams often cited one key weakness with overall burn plan development. The weakness noted was burn plans for complex burns that did not have sufficient depth and detail to match the complexity of burn. Largescale burns will likely have multiple aspects, variable vegetationfuel complexes, resource objectives and constraints that require more complex planning and burn organization to implement successfully.

Conclusions and Recommendations

The disparate types of information across the agencies made an assessment of this nature challenging, particularly tracking potential knowledge gaps or identifying developing trends, and especially without being able to access all existing reports. Despite this, the collected data do converge on several important lessons that we as a fire community need to learn. These are outlined below.

Reviews of all escapes and near misses should be consistently conducted, collected and stored in a centralized location to assist the community in learning from its experiences. A consistent framework with common questions and documentation would help make future lessons learned efforts more meaningful and samples more robust. While individual reviews provide opportunities for the local unit to learn, consistency would better assist this learning across Agencies. Such a consistent interagency framework to conduct reviews of escapes and near misses could assist all practitioners and agency administrators to identify knowledge gaps. As Weick and Sutcliffe (2001) recommend, sharing of near misses may tell us more about reliability than escapes, but they also provide us the opportunity to ensure that we take signals as a sign that things are ‘ok’ or ‘not ok’ until proven otherwise

The NWCG Complexity Rating System Guide should be explored to look at how to handle cases where burns are simultaneously lit potentially changing complexity, management oversight and organization structures. Is it true we underrate complexity? If so, why and in what context? Is there a tendency to underrate overall complexity or individual elements – risks, potential consequences and/or technical difficulty because experience with particular situations is low so it is not recognized as a potential source of problems?

13

Is there a tendency to underrate because adequate resources and skill levels are not available to the unit (e.g. a hesitancy to identify a burn as more complex because an RxB2 or RxB1 is not on the unit or readily available).

There maybe a gap between the intended use of the complexity rating system and policy. Per policy, a final complexity rating is to be done and included in the signed burn plan. This freezes the rating for implementation. There is no mechanism or direction to aid burn personnel to reevaluate overall complexity or trigger the need to increase the level of detail in burn plans when multiple plans are implemented at the same time.

Vacancies in key positions were noted in several reviews as having an important impact on fire operations. The recommendation in the reviews was to fill vacant positions to help relieve the stretched workforce experiencing expanding programs.

We continue to be surprised by fuels and fire behavior. Why? Have we lost our knowledge base and skills in identifying the carrier of fuels and selecting the most appropriate fuel model? Or, are we dealing with vegetationfuel types that are more complex or different than we have experiences with or fuels models to represent? Likely, both of these issues are causes as well as others. Due to limitations of the models to reflect reality practitioners may be reluctant to use these tools. Even with the limitations of fuel models and fire behavior modeling these tools assist with identifying sources of problems. However, continued training in usefulness and shortcomings of fuel and fire behavior models may help prescribed fire planners.

Other specific recommendations include:

• Continue to share lessons learned with other fire management personnel to broaden experiences levels.

• Investigate mechanisms to minimize possibility of escape the day of ignition. For instance, checking fuel receptivity outside the unit may be as important as how it burns inside the perimeter. Test all fuel types. If there is more than one fuel inside or outside the planned boundary we should be aware how each of these are going to respond. Holding and contingency forces then can be ramped up or down accordingly.

• Monitor leadership assignments and personnel transitions closely to ensure someone is directly assigned as the lead for a prescribed fire until the fire is declared out.

Next Steps to Become a HRO

Achieving and maintaining high reliability requires not just intellectual understanding, but translation of this into practice. This review of reviews has not revealed any earth shattering weaknesses; all weaknesses summarized above have been known previously. What it has done is to further highlight several trends and points of weakness in our practice. How do we move forward from here? This section is included in the hopes that it will stimulate further discussion…and actual change in our practice.

‘Reviews’ need to be approached as a tool for learning and be clearly separated from disciplinary actions. Did you have a good plan? Did you follow the plan? Did you execute it with qualified people? These are all good accountability questions. However,

14

accountability and learning from undesired outcomes are often at odds with each other. Accountability can lead down a path of blame after which opportunities for learning disappear. If fire use practitioners are going to move toward becoming a learning organization then we have to examine our ‘failures’ as an HRO. Reviews for learning should use questions that help evaluate expectations, assumptions, surprises and blindspots.

Further develop and integrate efforts exploring how to become a learning organization and operate consciously use the principles of a High Reliability Organization. Efforts so far include conducting After Action Reviews and Managing the Unexpected workshops. Both of these have been useful pathways to explore and should be developed to further integrate these concepts into daily practice.

Agencies need to further explore/validate the emerging trends such as a tendency to ‘underrate’ complexity, look at why this is happening and what we need to overcome that tendency. Another area that may need further work is to look at why some vegetationfuel complexes seem to repeatedly ‘surprise’ practitioners especially the more flammable fuels like cheatgrass and pinyonjuniper. Another area that seems to be a struggle is for practitioners to recognize risks and develop burn plans for complex landscape scale burns. One way to overcome this tendency is to encourage burn plan development by a team (prescribed fire planner, burn boss, holding and ignition specialists) because more eyes, and more experiences will be added to the preparation of the plan.

How can we avoid being blinded by burn plans (Weick and Sutcliffe 2001)? Federal land management policy requires burn plans. How can we strengthen their use? One option might include drawing on concepts fromManaging the Unexpected,. A team could explore ways to overcome the trap of expectations and tendencies to seek confirming evidence. For example, increasing awareness through NWCG courses that burn plans could blind us (how and why does that happen?). Another possibility to consider are ways to develop burn plans with a balanced approach focus on what we do not want to have happen as much as we focus on what we do want to have happen. It could be stressed to prescribed fire planners they should focus on ways to look for ‘disconfirming’ evidence that they are not in prescription and put them in the monitoring plan. Since it is the policy of federal land management agencies that burn plans are required this area would warrant further exploration.

Agencies should look into ways to build line officer and agency administrator ‘sensitivity to operations.’ Only two of the most recent reviews explored and acknowledged the role of the agency administrator. One case commented on the active role of the agency administrator, but also noted they had not yet attended a mandatory training (i.e. Fire Management Leadership). The other case praised and acknowledged the active role of the agency administrator who was onsite while the burn was being conducted and played a keyed role once the prescribed fire had been declared an escape. Agency administrators will play a key role in operating as a HRO.

Acknowledgements

Intermountain Region (R4) FAM, particularly Dave Thomas, Regional Fuels Specialist for promoting learning organization, supporting and funding a detail program analyst to the Wildland Fire Lessons Learned Center to develop this analysis.

15

Tim Sexton, National Fire Use Specialist, Forest Service and Al Carrier, National Fuels Management SpecialistOperations, Bureau of Land Management for digging into their files and supplying additional escape prescribed fire reviews not electronically available.

This report is intended to further the mission of the Wildland Fire Lesson Learned Center: Collect and analyze observations Share lessons learned and best practices, and Archive knowledge and information.

16

References and Citations

Garvin, David A. 2000. Learning in Action: a guide to putting the learning organization to work. Harvard Business School Press. 256 pages. Keller, Paul. October 2004. Managing the Unexpected in Prescribed Fire and Fire Use Operations – a workshop on the high reliability organization. RMRSGTR137. 73 pages. National Wildfire Coordinating Group. 2004. Prescribed Fire Complexity Rating System Guide. NWCG NFES 2474. 45 pages. Snook, Scott A. 2000. Friendly Fire: the accidental shootdown of U.S. Black Hawks over Northern Iraq. Princeton University Press. 257 pages. Weick, Karl E, and Kathleen Sutcliffe. 2001. Managing the Unexpected – assuring high performance in an age of complexity. 200 pages.

Prepared by D.Dether 1 June 2005

Table 1 Timeline of Escaped Prescribed Burns Involving Fine Fuels (Grass and Grass-like) Vegetation-Fuel Complexes. In three out of five escapes, the fine fuels were not the targeted vegetation for burning.

Burn Date Fuel

Time 0800 0900 1000 1100 1200 1300 1400 1500 1600 1700

Pa-CA

June 22, 1998

Cheatgrass

During initial black-lining

Blacklining at upper end of prescription window. Plan written for normal conditions, but conducted with “abnormal” level of fully cured cheatgrass. Burn boss limited experience with fuel type.

BQ-CA

July 1, 1998 Heavier loading

cured grass

“shortly after

ignition”

Two spots occurred “shortly after ignition. First spot contained. Second spot grew to 4 acres and prescribed burn declared an escape. Holding actions limited due mechanical failure of helicopter.

Me-UT

Sept 9, 2004 Cheatgrass

During “test

“burn”

Had no lookouts posted to look for spots. While more than the minimum holding forces were on-scene unable to respond quickly to slopover due to hindered access. Burn boss limited experience with fuel type.

IL-AZ February 4,

2004 Reed Grass

Less than 2 hours

Main ignition of unit began 15 minutes after test burn. Ignition stopped after 50 minutes. Possible spot across river was reported and confirmed in 5 minutes. Holding forces were unable to locate and control spot immediately due to thick smoke and access was also hindered.

No chronology provided

1235 test ignition Declared ?

1545. 1602 1625

0945..1000…1050 1110 1115


Table 1 (cont) Timeline of Escaped Prescribed Burns Involving Fine Fuels (Grass and Grass-like) Vegetation-Fuel Complexes.

Burn Date

Fuel Type Time 0800 0900 1000 1100 1200 1300 1400 1500 1600 1700

LR-CA July 2, 1999

Invasive weeds

Less than 3 hours

1050 1200 1300-20..1330


Table 2 Timelines for Escaped Prescribed Burns Involving Shrubland/Woodland Vegetation-Fuel Complexes.

Burn Date

Fuel Type Time 0800 0900 1000 1100 1200 1300 1400 1500 1600 1700

Bl-WY June 4, 2003

Pinyon/ Juniper

3 hours

Ca-UT Sept 23,

2003 Heavy, oak

brush

5 hours

Main ignition of one area began 10 minutes after test burn. Approximately 1 hour later, ignition began in a second area away from the first. Multiple spots occurred some with rapid rates of spread. Test burn was not conducted in representative location.

SC— April 10,

2003 Decadent

bitterbrush & P/J

5 hours

Prescribed burn not declared an escape when it should have been.

IB-FL March 2,

2004 Southern Rough & swamp

Less than 3 hours

Not declared an escape unit 5 days after ignition, but review determined the prescribed fire should have been declared within 3 hours of ignition due to fire behavior being outside of prescription parameters. No documentation of test burn.

Est. 1100 1400

1200 1245- 1300..1345………………1500

1230 1240 1320 1430 1515 1630 1700

1130 1200 1400 1640


Table 3 Timeline of Escaped Prescribed in Forested and Slash Vegetation-Fuel Complexes. While most escapes in these vegetation-fuel complexes occurred days after ignition during the patrol and monitoring phase; however, several did occur during the ignition phase with two declared an escaped prescribed fire within six hours.

Burn Date

Fuel Type Time 0800 0900 1000 1100 1200 1300 1400 1500 1600 1700

Pi-NM March 12,

1996 Ponderosa pine, mixed conifer &

P/J

Less than 2 hours

No mention of when test fire was conducted but lit in an area not representative of fuel conditions. Initial spot fires handled by holding forces.

NR-UT June 28,

2001 Slash w/

subalpine fir

Less than 5 hours

LJ-AZ May 5, 2004 Ponderosa pine & P/J

Same day as ignition. Unknown start time.

Ignition was planned for and stressed during briefing the need to be completed by 0900 hours due to observed burning conditions. Ignition was not completed within timeframes due to logistical and mechanical problems.

Unknown 1100 1143 1517

1300 1400 1500 1600 1645 1736

1130 1230 1300 “shortly after 1300”

Prescribed Fire Escape Reviews Appendix A

Project Name Yr Agency Unit State Planned Acres Escape Size and Consequences

Complexity Rating Veg/Fuel Complex

FBO Fuel Model & Modeling

Season of Ig (mo/day)

Length of time from Test Fire or Ig to being declared an Escape

Lizzie Springs 1996 BLM Rock Springs District WY Not Reported

(NR)

350, unknown if these acres were outside unit or total ac burned

Not Reported (NR)

Conifer including heavy subalpine fir timber with aspen. More fine fuel due to 2 yr rest from grazing.

Not Reported (NR) Oct 2-4 8 days

Pinatosa 1996 FS-R3 Mountain Air RD, Cibola NF NM

Total 9,000 in project, approx. 5,100 ac actually burned w/in planned area

2,000 acres unplanned

including 300 ac pvt

High - aerial & hand

Multiple - ponderosa pine, mixed con &

P/J, oak shrub

FM 9 & 10. Used FM 10 for BEHAVE runs.

12-Mar

1.5 hours No mention of test fire. Hand ignition began ~1130 hrs. Spot fires noticed between 1230-1300 by hand ignition crew, but able to control with holding resources. RxB fired main ridge inside unit with PSD; however fire spread downhill rapidly towards perimeter. Shortly after 1300 declared.

Banner Queen 1998 BLM

El Centro Resource Area, California Desert District

CA 650 NR NR

Mature chaparral (chamise) w/ unusual amount of grass due to unusual amt of ppt with el Nino.

Used brush model, but did not account for amt of grass and dead/live ratio in brush

1-Jul

"shortly after" ignition. First ignited (test burn) @ 1235 hrs. Shortly after ignition two spot fires occurred. 1st spot was contained. 2nd spot grew to 2 acres then 4 acres getting into steep terrain and not safe for ground forces.

Prepared by D.DetherForest Fuels Planner, Boise NF 1 6/29/2005







Fox Lake 1998 BLM Klamath Falls Resource Area OR 278

10 acres inside unit plus 29 acres on Pvt - timber lands

NR NR NR 29-Apr

5 hrs. Firing began in NW corner ~ 1130 after briefing. 1230 began having problems w/ spotting. 1300 stopped firing and took holding actions in NW corner. More spots along N-side line @ 1430. 1530 one acre outside unit on pvt ground. 1630 declared an escape.

Pahcoon 1998 BLMCedar City District, Dixie Resource Area

UT Blacklining NR NRCheat grass with heavy "above normal" loading

NR 22-Jun

< 1day no chronology included. Spot fires or slopovers that were more than holding forces could handle was one of noted factors.

Lowden Ranch 1999 BLM Redding Field Office CA 100 2,000 including pvt

+ 23 homes LowGrass, oak brush & riparian veg. Forested outside

2-Jul<3 hrs. Test fire @ 1050 hrs. Multiple spots at different locations. Declared @ 1330 hrs.

Wilson Gulch (missing append) 1999 BLM Burley Field

Office ID NR NR NR Juniper 18-Aug NR < 1 day. No chronology included








Cerro Grande 2000 NPS Bandelier NM NMThree phase project. 1st of 3. Approx. 900

48,000 plus pvt w/ 280+ home/ structures

Moderate - hand Multiple 4-May 1 day

EB-3 2000 BLM Arizona Strip Field Office AZ 3 acres "test"

subunit

197 acres Damaged pvt property

NR - a burn plan was not prepared for this "test" unit

NR NR

Ignited 3/31/2004. Declared 4/13/04

13 days.

Mt. Como 2000 BLM Carson City Field Office NV

Multi-unit project. 1st burn in 1997. EA done 1996

NR

NR - Elements rated as "high" missing rationale and mitigations. Hand ignition

Pinyon with litter/duff & brush that provided ladder fuel

Oct 18 Started Ig 1325 Done by 1700 Oct 20 Declared 1645

2 days Had 11 people & 2 engines to hand ignite area, but no direct access to unit for engines. Ignition went well, but fire burned into the evening & actively backed down hill all night. Next day no new ignition & tried to contain area already lit. A forecasted wind event kick up fire & spotted outside of unit from unburned island inside unit.








Alkali Rim (Near Miss) 2001 BLM Worland Field

Office WY 1,200 to 1,500

N/A. This review was about the need to use escape routes during burn ops.

Low, but did not follow agency policy. Should have been Moderate

Juniper with sagebrush/ grass intermixed

N/A 20-Sep N/A

Cordgrass (escape named Pin Oak) 2001 FWS Marais Des

Cygnes NWR KS 80 acres 13.6 acres w/ 4 acres on pvt Low cordgrass

FM 3 inside FM 3 & 9 outside

28-Oct 11 days

Navajo Ridge 2001 FS-R4 Dixie NF UTProject 199, mult. units. 26 ac w/ 2 units previously lit

8.2 including pvt NRLopped slash w/ subalpine fir and aspen outside units

28-Jun

4.5 hrs. Test burn conducted @ 1300 hrs w/ no holding problems. Ignition on unit occurred from 1400-1500 hrs. Ig stopped due to spotting, holding crews contained and Ig resumed @ 1700 hrs. At 1730 hrs, more spotting occurred and several slash piles were ignited. Declared at 1736 hrs. Contained 2000 hrs.

North Shore Kenai Lake (Draft) 2001 FS-R10 Chugach NF AK 1,000 NR

Reported "complex" no doc. of analysis - aerial ignition

Bug-killed spruce Not Reported (NR) 5-Jun

10 days. Area was monitored & patrolled according to the plan. Evening of 6/25 spotfires outside of unit discovered. Increasing winds caused spotfires to spread beyond the capability of local resources.








Pot Creek (DRAFT) 2001 BLM Craig Field Office

CO

Black lining with 261 ac area

250 - 300 acres. Suppression cost $50.0 ths.

NR Sagebrush NR 4-Sep

~3 hours. Test fire completed by 1315 hrs. Ignition began around 1330 hrs. A "significant wind event" occurred 1525 hrs. w/ winds out of the west at 30-40 mph driving the fire across the east line. Winds reported to begin dying down around 2300 hrs.

Rock 2001 FS-R5 Tahoe NF CA 200 201 NR Long needle pine/white fir

FM 9 in, FM 10 outside. Continuous frost-killed brush @ escape

May 9th part of unit successfully ig. May 10 proceeded to complete ig

5 hrs. (2nd day of Ig) Test fire @ 0715 hrs. Ignition started @ 0800 hrs. RH drop dramatically bwtn 0800-0900 hrs, but still in prescription. 1030 hrs. RH drops to 19%. 1100 hrs RH 12%, winds increase to 5 mph. Contingency forces called. 1130 numerous spots, more resources ordered. 1222 hrs more spots which grew, more resources ordered. Declared 1230.

Wilson 2001 FS-R1 Lolo NF MT 706 none outside planned boundary

High - aerial ignition.

Dry timber types, mostly northerly aspects

FM 2/10/9. FM selection not issue. Smoldering in Lrg 1000 hr fuels was

Four different days for ignitions. April 25-26, May 9 and May 11. Declared an escape May 24th.

29 days. Declared an escape due to "logistical and financial" concerns. The burn was NOT outside planned boundaries.

Anderson/ Danskin (this review covers

2 separate burn project areas)

2002 FS-R4 Boise NF ID 3,500 405 High - aerial PP/DF intermixed w/ sagebrush/ grass 15-May

4 days Two different landscape scale burns ignited the same day. Both units had previous ignitions weeks before. Snow remained on northerly slopes and checked surface spread of fire. High winds (30-35 mph) from a passingsystem caused both burns to escape planned boundaries at the same time.








Sanford (this review covers 2 separate

burn project areas)2002 FS-R4 Dixie NF UT 3,500 total in 2

separate RxB 78,000 NR

Two separate days of Ig on two different burns 11 miles apart. April 22, May 13

30+ days

Blanco 2003 BLM Albuquerque Field Office NM NR NR Low Ponderosa Pine but

w/ significant P-J

Incorrect FM selection - used FM 8. Review recommended using a combination of FM 4 & 6

4-Jun

3 hrs. Test fire 1200 hrs - 2-4' FL with occasional torching. Ig stopped 1245. 1300 hrs spot fire reported on top. 1345 hrs burn personnel unable to contain. Declared 1500 hrs. ~1600-1700 hrs more resources ordered (engines, tanker, T-1 crew).

Cascade II 2003 FS-R4 Uinta NF UT1,000 (in BP) Ignited 400 not in BP

7,828 HighHeavy, mature oak brush intermixed with aspen

23-Sep

5 hrs. Test burn @ 1230 hrs. Second ignition in different location 1320 hrs. Spots occurring in 3 different locations @ 1430 hrs, 1515 hrs & 1630 hrs. Declared 1700 hrs. 1720 hrs holding forces retreat to designated safety zones.

Cherry 2003 FS-R3 Prescott NF AZ Project 8,141 2nd of 3

150 ac slopover + 992 ?

NR - hand & aerial Chaparral 16-Jun

1 day. Ignited 6/16. "Unpredicted wind event" shifted wind direction. 150 ac spot occurred and contained. 6/17 no new ig., mopping spot. New spot 6/17 @ 1300 hrs.








Petty Mountain Near Miss 2003 FS-R4 Ashley NF UT 1,000 Not declared (Total

ac burned 2,593) NR Primarily Sagebrush w/ aspen

Fuel Model 8 covered 80-90% of area with isolated pockets of FM 10.

24-Sep Was not declared an escape.

Puma 2003 FS-R2 Black Hills NF SD

Project 800 w/ 11 units. Completed 231 acres on 2 units prior to lighting another 2 units that escape.

NR

Low - hand BB. Reviewers questioned whether RxB3 was correct "management oversight"

Ponderosa pine

Blackline Unit 3 March 14 & 25, Ig on Unit 3 & 5 March 31. April 1 Ig & declared

1+ days? Had been "blacklining" on one of units during two previous times in March. March 31st continued blacklining on one and started on another unit. During interviews, burn personnel mentioned holding/spotting problems. No holding or monitoring overnight. Declared April 1st after fire behavior increased & made significant runs inside and out of target units.

Sink's Canyon 2003 BLM Lander Field Office WY NR 12.76 acres outside

MMA on PVT

Rated Moderate -WUI. According to BLM 9214 all WUI should be rated High

Majority NR, decadent bitterbrush & P-J

Majority FM 2, FM 6 not used in BP to account for brush inclusion.

10-Apr

5 hrs. Test ig @ 1130. 1200 hrs blacklining started. 1400 hrs ig of interior. 1640 hrs winds switch, RH drop & spots across line. Not declared, but later determined outside of MAA.








Impassable Bay (Compartments 16

& 117)2004 FS-R8 Osceola RD,

NF's of Florida FL1,500 in two units 16 = 1,000 117 = 500

32,000 + including state & pvt lands

Individual burns rated low, but burned together (adding helicopter & other mixed resources) indicated should have been Moderate to High. No documentation on "final" complexity rating.

Long-leaf pine & "swamp"

FM 7 (southern Rough) and FM 4 (dry swamp)

2-Mar

5 days before "officially" declaring an escape, but approx. 3 hours after ignition the burn was out of prescription & should have been declared. A test burn was conducted, but not documented. Estimated from report to have been around 1100 hrs. At 1400 hrs obs flame lengths of 20', (prescription called for 4-6') which exceed prescription.

Island Lake 2004 FWS Imperial NWR AZ 630 acres Estimated 200 acres Moderate

Cattails (8-10 ft), phragmites (common reed, 8-10 ft) also intermixed with salt cedar

Fuel model 3. Appropriate model selected but under - estimated potential fire behavior

4-Feb

< 2 hours. Test fire @ 0945 hrs. Fire behavior and smoke Rx met. Ig begins @ 1000hrs. At 1030 hrs. east flank widen w/ aerial ig. At 1040 hrs remaining unit lit in 2 passes. 1050 ig halted. 1105 hrs possible spot across river. Confirmed 1110 hrs. Declared 1115.

Long Jim III 2004 NPS Grand Canyon NP AZ Project 5,050.

1st of 3 for 1,618

NR - one element of analysis was raised as an issue

Ponderosa pine w/ P-J FRCC 3 5-May

Ig time NR, but aerial ignition completed 1100 hrs. Spotting reported @1143 hrs. Declared @ 1517 est. 6+ hrs. Ignition was to be completed by 0900 hrs due to observed burning conditions during blacklining ops a week before and during test ignition that day.








Meadow 2004 BLM Fillmore FO/ Fishlake NF UT

200 acres. Prior burning in spring on same project. 500 ac in project.

524 acres. Only 2 inside unit. Included PVT & state land. Closed hwy due to smoke. Burned power poles, cut power to town.

NR - RxB2 with hand ignition.

Sagebrush with cheatgrass & thinned P-J. Increase fuel loading of cheatgrass at time of burn.

9-Sep

< 1 hour. Test fire Ig 1545 hrs. Two engines released to respond to wildfire 1550 hrs. Decision to "shutdown" and suppress test fire. Spot fire noticed @ 1602, within ~8 minutes spot grows to 30 acres. Declared @ 1625 hrs. Winds increased & shifted with T-storm activity. Spotting

Mitchell & Hot Air (this review covers


2004 FS-R3Clifton RD, Apache-Stigreaves

AZ Planned 5,000 to 7,000 acres 1,537

Individual burns Low - Mod. Ig together should have raised complexity. Also, content of complexity analysis found to be lacking.

Multiple types including ponderosa pine. Others not mentioned.

Fuel model 8 was used for complexity, but prescription parameters used more volatile fuel models 4 & 9. Spotting distances were not calculated.

12-May

6 days Concurrent ignition on two separate projects. First planned ignition delayed due to high wind assoc. with weather front passing. Begin ground ignition next day. Following day proceeded with ignition by hand & aerial.


Prescribed Fire Escape Reviews Appendix B

Project Name Anatomy of the Escape - What were the "weak signals" Key Lessons Learned / Emerging Best Practices / Knowledge Gaps

Lizzie Springs

Appeared to review team to be "well planned and properly executed, but inadequately documented." Knew the head of Lizzie was problem area - fuels and no natural barrier. Two other burns "successfully completed" after this burn, but several days after this burn was ignited two other burns in the state escaped and local personnel knew of these escapes. During 1996 "drought" conditions existed affected heavier fuels more than normal, area was known to have heavier fuels. On the 8th day, the fire escaped "due to high wind event", but it seems that many signals were potentially missed.

Burn Plan based on "normal conditions" yet burn personnel proceeded with plan and did not make "adjustments" to burn plan or operational procedures even knowing that there was "heavier than normal" fine fuels and with drought conditions that the heavier fuels could be effected even though they knew the area had heavier fuels. They knew the Wx was "unusually, warm and windy". Patrolled for 5 days and even with smokes still in unit assignedresources to other burn projects. There was a shortage of "qualified" personnel and this contributed to inadequate mopup & patrol. The key people involved were "fully trained & experienced", but their numbers were barely adequate to execute a single burn let alone multiple projects.

Pinatosa

Shades of Cerro Grande - Snow delayed initial ignition. Spring time with public "knowing" about strong, erratic winds. In this case, technically lit outside of prescription elements and/or shortly after ignition the Wx conditions exceeded prescription. The report mentions that the test fire location used during hand ignition was not representative of fuels or environmental conditions (more shaded and sheltered than unit). Many problems with burn plan and prescription development. Burn plan preparer not qualified nor very experienced. RxB not experienced with burning in these fuel types. Trust with Wx Service was low because of earlier "blown" forecast that did not predict snow storm which caused a delay.

In this state (NM), there may be a need to better understand seasonal patterns of winds or wind events during transitional seasons. Wind and Wx patterns in general need to be understood locally, seasonally and what effects larger Wx influences can have on a burn. This burn is approx. 80 air miles from Cerro Grande. The review noted that the burn boss (although qualified) did not have experience of the with fuels types & may have contributed to practices leading to the escape.

Banner Queen

One of two burns involving use of a contractor - other burn was Fox Lake. The review noted that the burn plan was well developed and prepared. The execution of the burn was done well and going well. Two things directly linked to escape although was not giving enough consideration to fire effects of the grass component in the predominate brush models used (with high live to dead ratio). The grass component repeatedly came up in interview (a surprise to burner?). Knew it could be a problem, but focus was on the brush not fully appreciating what the added grass would do. Experiences with wildfire that year may have misled personnel into believing spread would NOT be a problem. The burn plan did not reflect changes in fuel structure so not planned for with the test fire. 2nd Helicopter malfunction was also mentioned as a direct cause of escape because it was not able to help ground crews suppress second spot when it got into steep ground.

One lesson learned raised here concerns mechanical failures. While redundancy in human systems may not work well, redundancy for mechanical devices does work. When mechanical equipment is key to burn operations and/or holding capacity it may an area to consider and make sure your contingency plan covers this such as having another helicopter with suppression capability on standby-by for quick response. The complexity of the vegetation-fuel complex surprised burn personnel. Also, previous experience with wildfires that year may have "signaled" to burners that this fuel combination was more "flammable" than believed. Local manual direction DID NOT reflect current national direction. Supplemental burn plan was NOT signed by agency administrator.




Fox Lake

This is the only escape that involved using a contractor which prepared and conducted the burn. The review determined the burn plan was adequate and all people including the contractor were qualified. However, contractor did not sample fuel moistures outside of unit on PVT where it was more open therefore drier. Lit on "drier" end of Rx. Lighting too much too fast was noted as a contributing factor. Contractor also did not get spot Wx forecast as required in burn plan. The review noted that since contractor is "paid by the acre" that this may have played a role in choices to proceed with ignition. There was "prolific" spotting which combined into 3 "major slopovers" outside unit all onto pvt ground.

The review noted that the contractor did not transition quickly enough from firing to holding. Did not take on-site 10-hr dead fuel moisture samples nor requested a spot weather as required. Used only general zone forecast. Weather obs. from on-site indicated parameters were near the drier end of the window when ignition was started & conditions seemed to be out of prescription quickly. All ignition had been stopped and in holding mode when out of prescription parameters. The morning of the burn unit was selected over another because the other unit was "too wet" (unknown if this influenced burn boss assumption that is unit would not cause concerns with control, not raised as an issue in report). On-site weather obs. were not taken in a representative area of conditions (winds taken in sheltered area, area that was source of spotting was in near an exposed ridgeline). Also, recommended that 10-hr needed to be monitored at least a day prior to ignition and if conditions outside of unit are different those areas should also be monitored.

Pahcoon

The review concluded that there was a "failure to adjust operational procedures" to reflect the additional on-site fuel loading and an attempt to black line on the "extreme upper end of the prescription". The plan was written for "normal" conditions, but that year there was "above normal" ppt so "heavy" loading from cheat grass was created. RxB did note the extragrass and added engine and "pumpkin" to compensate. RxB had little experience in this type. Brush component was above "optimum" (i.e. wet or high live fuel moisture), but grass fully cured.

Review noted that all burn personnel were qualified for their position, but noted that experience in Rx fire and in this fuel type varied considerably. The lack of expereince in this fuel type of the burn boss was noted as a contributing factor. Burn plan was written for "normal" conditions, and noted it is up to burn personnel to adjust when conditions are different than planned. Recommended that for burns with considerable "blacklining" operations to write seperate prescriptions parameters (Note: this is currently in BLM policy and burn plan templates). Inter-office coordination also needed to improve as well as communication & cooperation with other agencies. Burn was adjacent to an Indian reservation, but when escape occurred they were not notified of escape.

Lowden Ranch

Two firing crews. 1115 RxB concerned abut fire getting into trees, but holding actions mitigated concern. 1145 RxB (T) reported "good progress" and asked if should continue. Given Ok from RxB. 1200 hrs RxB(T) noticed spots. Ig stopped, holding action contained and 100% mopup was proceeding. Ig resumed @ 1220. At 1300 hrs spot reported in area of initial concern and it was also contained. 1320 hrs 3rd spot spread quickly and exceeded holding capability.

There is extensive documentation on this escape so it was not summarized here at this time.

Wilson Gulch (missing append)

Burning in juniper necessitate wide parameters (RH, temperature, live fuel moisture, and wind speed/direction) that would be considered "high". Based on some of the recommendations it appears this burn used a combination of blacklining and "nighttime" recovery for containment lines. The review only mentioned that one night did not get expected recovery. Also, based on review team recommendations it seems that burn personnel were not involved in the upfront (during NEPA) planning and selection of the area to be burned. Also, the NEPA document and analysis did not allow for use of dozers which limited options for the burn personnel to implement.

Missing documentation may have had additional information mentioned in report. Need to have all pertinent information in reviews (this review talked more about what the review team talked about than providing the content and context of their discussions). The review seemed to indicate testing a newer approach to burning that would have benefited others to have more information and share results with other burn personnel. Importance of having prescribed fire experience at the planning table to help with project design, allowable burn areas and to ensure activities or practices that will likely be employed during burn operations (e.g. use of dozers) are appropriately analyzed and disclosed in the NEPA document.




Cerro Grande There is extensive documentation on this escape so it was not summarized here at this time. There is extensive documentation on this escape so it was not summarized here at this time.

EB-3

The review attributed the escape to "failure to adjust operational procedures (holding and mopup) to reflect the changing weather conditions. Available fuels during ignition included only the upper needle cast which was dry & consumed. Duff layer was wet and not consumed. Large fuels (logs) were not consumed. Fine fuels (<3") were reduced by 60-70%. 1st and 2nd weeks after ig, units each checked each day. Smokes were obs, but minimal mopup occurred due to " test" nature of the burn. Into 1st week after ignition, agency's unit suspends any new ignitions because weather conditions "became too dry". The day prior to the escape, patrol notes smokes and open flame in unit, but did not think there was a risk of escape. The day of escape patrol noted smokes in unit and also did not think it posed a significant risk of escape. Seven hours later "significant smoke" was reported in the area. Leadership (Officer manager, FMO, FCO out of town), the RxB in office, but doing other work. General weather forecast for that day call for temps in mid-70's, RH's in teens and SW winds 20-30 w/ gust to 40 mph.

Small burn, testing a "technique" for larger application. Did not constantly staff burn nor had an individual directly responsible for burn until declared out. Were personnel lulled into complacency since it was a small (3 acres) test unit? Trust (lack of confidence) in weather forecasts was mentioned in report, but not cited as a reason contributing to the escape. The "burn plan" used (the test unit was not included in it) did not have "triggers" identified for mopup or contingency actions. Technically, there was no burn plan for this test burn.

Mt. Como

There were no natural barriers or line constructed in places to contain burn area. Potential fire behavior was noted in the complexity analysis as a significant issue, but no specific mitigations were identified. Burn containment relied heavily on nighttime recovery to control burn spread. This coupled with 6 listed "indications" noted in review that conditions were not "normal". 1) no significant ppt. from 8/30 to 10/20. 2) this agency's unit had been receiving "severity" funding until just prior to burn. 3) fire restrictions had just been lifted for this area. 4) 1000-hr TLF @ 6% 5) in 2nd year of "drought". 6) at established monitoring sites Pinyon pine live fuel moisture was EQUAL to long-term avg. & for shrubs substantially BELOW long-term avg. Also noted contributing to problems were fuels - "a substantial portion of the stand had a litter/duff layers & ladder fuels not "normally" associated with pinyon stands.

Review recommended that for burns which rely heavily on weather conditions for containmentshould consider using portable RAWS. Also, no nighttime Wx obs were taken and provided to NWS. Another recommendations noted concerning coordination w/ NWS and on-site Wx obs procedures that should be taken. One other burn (WG-ID-99) also had similar problems with a prescription that relied on nighttime recovery to contain burn, but did not get recovery nor took measures to record & work with NWS to improve forecasts. (Note: this burn review might be a good example to illustrate the concept of "less decision making and more sense-making".)




Alkali Rim (Near Miss)

Review determined that all personnel were qualified for their positions, but for some it was their first time on a burn and where not directly supervised at the time of the incident. Topography where incident occurred was "steep" (30-40%) with thick fuels - closed canopy juniper. Crew was instructed to light at the bottom of this drainage to generate heat to get the unit to burn better. The burn crews had been having a difficult time getting area to burn and had been experiencing low to moderate fire behavior during 2 previous days and up until 1400 hrs when incident occurred (wind shifted and RH's dropped quickly). However, burn boss did not outline his expectations for lighting crew well so they were unsure of what they were doing. The lighting crew was not directly supervised at the time by an Ignition Specialist. The assigned Ignition Specialists were in another area of the burn using a terra-torch. Visibility of the lighting crew was hampered by terrain and the burn boss could not seethis area directly either. No lookouts were used during this operation and weather was being taken sporadically throughout day.

This is a good example why evaluating and sharing "near misses" are important and valuable tools. Good reminder that prescribed fires are just as difficult as a burn out operation on a wildfire and LCES should also be in place for Rx fires. This had the potential to turn out very badly with possible injuries or fatalities. The review team noted that the previous fire behavior lulled the burn team into complacency. They were also operating "blind" by not using lookouts nor taking weather to maintain situational awareness. Need for good and thorough briefings pointed out especially when dealing with lesser experienced personnel. Complexity was underrated relative to national policy (review also noted that with 3 ignition teams should have been rated a moderate complexity). Unanswered question - if this were a wildfire would these people would have behaved the same way as on this prescribed burn??

Cordgrass (escape named Pin Oak)

Day of ignition contained slopover during blacklining operations. 2nd day blacklining continues, but not burning well so burn boss changes from backing fire to strip head fire. Fire activity picks up greatly with flame lengths reaching 7-10 feet. Slopover onto pvt ground, but contained. Patrolled every other day including the morning of the escape. At 1600 hrs the manager receives report of a lot of smoke coming from unit. Resources respond and extinguish, but burn 3 more acres of pvt. ground. Day of escape changing winds and trend of dry weather allows fire to become active. Part of the chain of events leading to escape were - exhaustion from previous burns & wildfires, underestimating low fuel moistures & inadequate patrolling.

Information not from a review document. Information is from a prepared slide presentation about the escape. Again, without documentation we be losing lessons learned and knowledge beyond audiences that have seen this. Mentioned in presentation that this incident (due to mopup procedures & methods) was the 2nd occurrence in this area in 3 years. This burn also relied on "nighttime recovery" to contain burn, but environmental conditions were not adequate in this case to control burn. No monitoring was occurring to verify nighttime forecasts.

Navajo Ridge

Documentation not clear on start/stop of ignition and spots. Not sure how many times ignition was stopped then started due to spotting. Good response from contingency forces noted. Test burn indicated "no holding problems". The report indicates but does not elaborate that on-site observations indicated there was for the potential for escape identified prior to ignition.

Finding in this report did not have sufficient detail to add to this section.

North Shore Kenai Lake (Draft)

This draft review seemed mostly to focus on policy and procedural review as well as the need for coordination with cooperators and public. The report did note that the burn was implemented on a day with weather predictions on the "dry and warm" end of the prescription. Warm temps and low RH's were predicted and continued to exist until the fire was declared an escape. Decision to burn was based on success of burn in same area 2 years prior although conditions were not the same in both years. The report did not elaborate on enough details.

Regional policy outdated relative to national direction. Insufficient dedicated resources were available to conduct burns. Having good interagency/cooperative partnerships were noted asbeing key to successful fuels program. The importance of relationships with the public (appropriate coordination, involvement and awareness of prescribed fire activities and fuels treatment planning) is a reoccurring recommendation from other escape reviews.




Pot Creek (DRAFT)

Burn Plan was for a different area originally done in 1999 and may have been part of a ongoing landscape burn effort starting in '92. Spot forecast predicted favorable conditions for burn ops, but T-storm were also forecasted high with winds reaching 30-40 mph "near" cells. Burners thought they would not experience these winds since they were 20-30 miles away from cell. T-cells were not visible yet when ignition started on the unit. Had tough time burning or getting sage to go during ignition. Had more holding resources than BP required.

No single document of this escape. Some information in documents vary from each other. The review document was only a draft and unknown if every finalized. The role and issue concerning the potential of burning when thunderstorms or cells are forecasted near burning not fleshed out and resolved. One area noted in the review needing improvement was communication and coordination of the cooperating agencies involved in this project.

Rock

Review has detailed chronology/narrative. Could be used as case study. Nov. 11, '99. An attempt to burn unit. During test fire, spotting into brush occurred above a road. Two other attempts were postponed. 4th attempt - 1st day of burning (50 Ac) completed what was considered the "toughest" part due to frost-killed brush above road. 1st day Test fire started ~0636 hrs. Ig began ~0738 hrs. 1300 hrs RH (12%) drops out of prescription. 5th attempt - escape. Burn boss(es) kept on attempting to compensate for conditions (contingency resources, starting ig early, etc). While burn personnel took many steps at what point should the "signals" triggered a "no go" action? If you have to work that hard to pull a burn off and you are at the upper end of prescription then maybe need to re-evaluate.

Review noted some confusion with interpretation and use burn personnel had with new burn plan template. This was noted by review team in trying to determine if burn was conducted "within prescription". Burn personnel did lots of weather analysis especially looking at wind speed and direction & RH. Much attention was focused on the diurnal wind speed and direction. However, the RAWS station trends for winds did not show the larger scale, "notorious" pattern that was cited as the main cause for the escape. The "notorious" pattern was one known for rapid drops in RH resulting in low RH's and higher winds speeds. This burn was conducted on the "hot" end of the prescription so the time of day RH would drop outof prescription was carefully considered. Weather analysis showed that RH would drop out of prescription between 1200-1300 hrs each day.

Wilson All personnel assigned to Rx fire were qualified and experienced. Holding and mopup plans were "weak" in that they did not contain triggers for actions or contingency.

Noted in report "with the shift in the fuels program toward more urban interface prescribed burns, more thorough documentation for all phases of project implementation may become useful."

Anderson/ Danskin (this review covers


These burns were part of two ongoing projects which had prior successes burning other blocks with no escapes. Active monitoring of both burns and weather was occurring. A "high wind" event caused both burns to escape. A local weather station recorded sustained winds of 30-35 mph which helped move fire from surface to crown fire which burned over snow.

Main recommendation for improvement was for increasing communication with the weather service (NWS).




Sanford (this review covers 2 separate

burn project areas)

The burn plan did not include all mitigations from NEPA document including sensitive fish species and smoke monitoring. Both burn plans lacked sufficient detail (deficiencies) in contingency, holding, mop-up & patrol. This review mentions the potential problems that can arise with long-duration, landscape scale Rx burns. This is one of several reviews that mention when burners are relying on vegetation (fuel) or fuel moisture changes need to have other indicators (e.g. measurable drought indicator such as ERC, etc) to signal when they willno longer work. (Note: this was a reoccurring theme in numerous reviews)

Programmatic burn plans - this review indicated that these could be ok IF burn personnel had enough detail regarding the area being ignited. Three main areas which needed detail are 1) what constitutes an escape 2) specific mop-up & patrol standards and 3) detailed maps which display planned ignition areas, locations of holding forces & critical holding points, location of sensitive areas and/or resource values & human safety considerations. The review also noted that training in preparing "good" burn plans was lacking (Note: Rx300 Burn Boss is only course at this time. NWCG is in the process of creating a course which focuses on burn plan preparation). Public involvement and awareness was mentioned as another keyarea that was lacking and becomes important when things do not go well or as planned. (Note: Puma review also touched on problems with using programmatic burn plans).

Blanco

Area previously burned ('93) with less than desired results (did not burn crowns of p/j). Test fire & primary ig occurred in mixed p/j community below rim of mesa top. Blacklining had been completed in April. Test fire & ignition rapidly built into a crown fire and spotted across blackline (est. spotting distance .5 - .75 mi) All burn personnel committed to spot. Unable to contain. Again, another problem is appropriate selection of fuel models and addressing mixed vegetation-fuel complexes within the burn area. In this case they used only one fuel model, but was not reflect the primary carrier of fire which under predicted potential fire behavior caused an underestimation of holding force needs. The fire behavior in P/J again was the veg-fuel complex that caused "surprise". Long-term drought and effects to fuels/fire behavior was also mentioned as a contributing factor not adequately considered.

Burn plan "technically" adequate. However, review team note a need for "further refinement and additional information for clarity" to improve the plan. This review like others addressing escapes with landscape-scale burns noted the challenge in burn plan preparations to seek the balance between sufficient detail to changing situations over time while allowing sufficient flexibility for the burn boss and line officer to manage the project. Burn area design was noted as a problem - did not develop a unit boundary based on fire behavior characteristic. The review noted a contribute factor to problem of unit design was area was originally designed to have mechanical treatment so needed archeological clearance, but clearance was not completed so area was excluded from burning at this time. Burn plan also did not include all mitigations from the NEPA document.

Cascade II Test burn was not conducted in a representative location. Test was done in a cooler, more protected area than area to be ignited. Low rates of spread were observed with test burn.

Numerous findings in report. Key was burn plan weaknesses. Repeat finding that the burn plan did not have enough detail for a high complexity burn. Along with other publicly visible escapes - Cerro Grande, Lowden, Sanford (and those that surprise I.e. impact the public) the importance of public involvement and awareness before, during and after prescribed fire implementation was noted in the review.

Cherry

The burn was set for implementation a week prior to actual execution. The burn was delayed to get "hotter" prescription needed in this fuel type. Anticipating this areas were pre-treated (2,500 acres) which did increase the ability to protect the community during the escape. Burn plan ok, but improvements were noted in the areas of risk assessment including threat identification and mitigations. Complex burn plan execution requires increased "situational awareness", appropriate resources and monitoring. It was unclear to review team that fire behavior was monitored and what if any role it played in decision making.

Review provides numerous good recommendations on how improved burn plans and specific areas that need attention when addressing complex, landscape scale burns. One area noted was developing regional guidance to have an objective method for assessing risks including potential for escape, evaluation of threats to the boundary, communities, developments and sensitive resources; identification of critical holding points; & mitigation actions to mitigate or eliminate threats. This is different than the NWCG complexity rating. Recommendations for use of "management action points".




Petty Mountain Near Miss

Aspen stands & roads expected to contain fire. Did test fire, and results determined to be acceptable. Ignition proceeded, but conducted at the upper end of prescription parameters. No one monitored the fire overnight. The morning following ignition found that the fire had burned through the aspen stands during the night. The original plan was developed for a spring burn which should have stopped fire spread as planned, but burn was implemented in fall instead so aspen area drier than originally desired and anticipated to act as a natural barrier.

The AAR suggested that rather than "amend" a spring burn plan it was better to write a new plan to deal with current conditions. The amended plan was confusing because of the mix of information. Issues arose over project design and adequate NEPA to cover both a spring and fall burning scenario. The allowable burn area did not cover the logical burn area for fall burn. Also recommended improvements to establish clearly stated and measurable objectives in the NEPA document that could be carried right into the burn plan. If the objectives are not measurable and cannot be monitored then they need to be written into a burn plan so they are.

Puma

The review dealt with numerous policy and procedural issues. The review team looked at a number of areas, but boiled it down to 4 areas - policy, prescribed fire plan, safety and fire program leadership. The review determine the plan was not a "good" plan and they did not follow the plan even the "good" parts of the plan - and amendments were not reviewed and signed. Review determined that ignitions occurred in units that were outside the burn prescription and approved burn areas. Used only the general fire weather zone forecast, did not use spot wx forecasts or local RAWS data to determine if in prescription or more site specific forecasted weather info. Snow was suppose to be used as a containment for this type of burn/burn plan, but not enough snow to contain.

Regional policy outdated and not consistent with national policy (this had come up in other reviews). Burn plan did not contain all elements required to be in a burn plan. Had been using a "master" burn plan and did not contain sufficient site-specificity. Many key elements of the burn plan listed as missing the site-specificity including maps of approved burn areas, holding, patrol, mop-up and contingency. This report brought up an issue not evaluated in other regarding qualifications of the line officer and delegations of authority. This report indicated that delegations to District Rangers need to be based on the "qualifications and experiences" not to the position. Also, although the burn boss was qualified as RxB 3 concerns were raised about the complexity of burn (i.e. higher than "low") and should have been implemented with a RxB2. Review stressed need for burn personnel to maintain "situational awareness" even for prescribed burns.

Sink's Canyon

Fireline was 5' brush line w/ 18" handline. Winds shifted and RH dropped to 15% which moved fire to burn into decadent bitterbrush in a "bowl" and jumps line. Fire continues following heavier fuels (bitterbrush & Juniper) cross-slope above rock cliffs & snow banks. Fire spread checked at rock outcrop. 2000 hrs winds die down and so does fire. 3 days later after mopup and girding smoke reported @1500 hrs. 1730 hrs spot fire was actively spotting & torching. Spotting 200' from rock outcrop that stopped fire spread before. Burns 4 more acres.

This was another review in which the issue of "no substantial definition of an escape" was rased as a problem. The review recommended that there needs to be a solid definition so there is no question of what (and when) there is an escape. Allowable burn area and sign of boundaries again raised as an area that needed to be addressed. Need to design burns with more logical burn boundaries. Review also recommended to better evaluate fuel "inclusions" as sources of trouble. If the un-modeled fuel pockets had been evalauted with burn day conditions it would have indicated it was outside of prescription parameters. A spot weather forecast was requested/received, but forecast did not accurately predict actual conditions (wind shift later in day). Review team interviewed another burn boss in same area & was toldof similar "problems" with forecast accuracy. Recommendation from review was for burn personnel to gather more local weather data, provide feedback & work with NWS office to increase accuracy.




Impassable Bay (Compartments 16

& 117)

Numerous items were of concern in this review. Burn boss made numerous changes to complexity and burn operations not in plan and without getting line officer approval. Due to changes made, the burn plan was deficient (e.g. added aerial ignition, but did not aerial org chart to plan) relative to the burns as implemented. Also, change in complexity to high also raised questions about RxB qualifications. Burn boss was RxB1(T) and was only coached 1st day. Review noted failures of the RxB coach to ensure things were done properly. One direct causal factor noted was prior years of drought & low spring ppt levels made swamp levels lower than "normal". The "swamp" was suppose to provide natural barriers, but due to drought it was dry and did not check fire spread. There was no documentation of this being checked although in one of the burn plans it specifically stated "ensure adequate water level in swamp to contain fire." Only, one FM was used to fire behavior, nothing for swamp & transition zones which would have much higher rates of spread.

These burn blocks were originally prepared and planned to be implemented separately, but implemented simultaneously which then created numerous other inconsistencies relative to policy and procedures. Another interesting note was the confusing direction between national, regional and Forest level manual and LMP/FMP. The issue of lack of consistent direction/policy below the national level came up in three of the other reviews. As in some other reviews this burn was not declared an escape when it should have been. There seems to be a gap in putting clear "triggers" for when an escape should be declared. Management of burn personnel - work/rest ratio exceeded & personnel arriving on burns without being part of assigned resources (this was a "normal" operation) caused safety issues. Management and leadership for the burns "changed hands" many times from the initial RxB (T) with trainer to others not currently qualified. Transitions were not well conducted.

Island Lake

Overall the review noted that this burn was well planned and prep. Speed of ignition (ring-fire of interior) probably contributed to escape - too much fire, too fast. Winds (increase in speed and direction) in river canyon noted at a "localized" event. Spotting distances with highRH's over a water body unexpected. Another "surprise" to burn personnel was the floating, flaming mats of vegetation although not the like source of ignition causing the escape. The potential for spot fires was included in the burn plan with mitigation measure; however, with low water conditions made river-based holding forces ineffective. Thick smoke also was noted as something that hampered the use of the helicopter for bucket work. The lack of communication with all burn personnel due to radio dead spots and background noise from equipment was also an issue noted.

The AAR noted that it was important for the burn boss to get "the big picture", but in this case it was hampered by terrain and access. Recommendations to compensate for this was to putobservation points on burn maps and utilize lookouts. Also, noted strategies to deal with tricky, transition times of the day when igniting were river canyons can be difficult. This keeps on being a repeat "lessons" -- know you local weather. Also, noted in this review as a reoccurring recommendation to have MOU or agreements with land adjacent landowners. The review also noted that slower ignition is warranted in these fuels and in this topography (river canyon). One lesson learned recommendation about contingency was a thorough recon is invaluable. Think about the worst case scenario and then imagine the worst case scenario going bad, then go back & plan your contingency.

Long Jim III

Burn plan ok & had thorough technical review. Burn personnel all qualified. Weather was suppose to be monitored 7 days prior to burn plan, but only taken day prior to burn. ERC were "above average", but below 90th percentile. 1000-hr TLF were "below normal", but within prescription. More active fire behavior than anticipated was observed during blackling ops and test ignition and cited as the reason why ignition of the unit needed to be completed by 0900 hrs. A serious of mechanical and logistical problems & delays caused the ignition to go 2 hrs beyond planned completion timeframe. The RxB1/ICT3 indicated not having seen this kind of crown fire behavior & consumption in P/J in nearly 2 decades at this particular unit.

Critical timelines that are established to ensure success of operations need to be tracked and reassessed frequently so that adjustments or mitigations (I.e. ignition pattern or method) can be made early enough in the operation period to secure successful completion.




Meadow

All burn personnel were qualified for their positions. Once again, the review cited an "above normal" loading of cheat grass was not anticipated to be a major influence on the increased fire behavior by the burn organization. Even with policy in place as part of the "Go/No Go" which specifically address this situation did not provide an "mental check". First phase of this project conducted in the spring with poor results and conditions that did not burn well. In the fall, fine fuel conditions changed from that of the prior spring by adding a "short, but dense" layer of cheat grass. This review noted similarities with another burn on a different unit that has similar experiences with short, dense cheat grass and creating spotting/holding difficulties. Other "procedural" and/or "operational practices" were noted in this review.

Review again stressed the need to gather on-site data relative to prescription parameters prior to burning. Incorrect use of fire behavior modelling may have misled burn personnel to underestimate potential flame lengths and probability of ignition. The potential for thunderstorm development was forecasted (LAL 4) and burn personnel reported seeing T-storm to south but proceeded with burn. Signing Agency Administrator had not attended "mandatory" training for Fire Program Leadership was noted by the review team. This is first time issue of administrator's qualifications raised.

Mitchell & Hot Air (this review covers


1st attempt at landscape scale RxB. Evaluated complexity only for each burn individually. One helicopter for both, but had to stop lighting on one to do bucket work on other. Other factors that contributed to the escape - lower than expected RH recovery, insufficient depth of blackline with unburned islands of fuels, concurrent operations on two units 15 miles apart & burn block boundaries located at mid-slope & canyon bottoms with continuous fuels. Complex ignition patterns to meet resource objectives & multiple fuel models created large unburned areas w/in boundary were allowed to burn "freely". As weather became warmer & drier fire behavior created spotting problems that challenged holding resources daily.

Several good lessons for landscape prescribed fires. When designing burn block boundaries consider factors such as terrain influences on fire behavior, natural barriers, and fuel changes. Drainage and mid-slope lines in continuous fuels should be avoided if possible. During multiple ignitions with long travel times separating them support for each block must include separate holding forces to deal with the potential for multiple escapes. Also, related to igniting multi-blocks when aerial ignition is used and the same helicopter is also key to holding/contingency on both units then should consider either having another helicopter as contingency or ignite only one block at a time. With complex, mutli-block burns may need to consider both day & nighttime staffing. Considerations for containment of burns is a reoccurring issue in reviews. In this case, blacklines were not sufficient in depth & extent to hold this complex fire. The review cautions burn bosses to take time to assess & evaluate conditions to assure proper reductions in fuels has or is occurring before ignitions.


prescribed fire lessons learned - fs.fed.us · prescribed fire lessons learned ... learning from...

Documents