Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
40 Conclusions and Guidance The risk of fatal aircraft accidents in the vicinity of run- ways is relatively large compared to those occurring in other areas of the airport. The RSA is intended to mitigate the conse- quences of aircraft veering off, overrunning, or undershooting a runway; RSA design standards have proven to provide good protection for these types of events. The demand to operate larger aircraft coupled with higher traffic volumes has often resulted in airport operators being unable to meet airfield standards that were created three or four decades ago based upon engineering judgment. In many cases, adhering to these standards would be cost prohibitive due to physical, economic, and/or environmental restrictions. However, even more compelling is that adhering to the existing standards may not improve the level of safety. This has created the need to reassess the level of safety provided by these stan- dards through the use of risk-based methodologies. One of the current challenges with airfield design is to develop a tool to evaluate the level of safety at airports that cannot comply with current standards for RSA. ACRP Reports 3, 50 and 51 were intended to fill some of those gaps; however, a methodology to evaluate risk in certain sub areas of the RSA was not available to the industry. This study attempts to fill such need with the development of models and analysis tools to allow assessing risk when RSA standards cannot be met. Major Achievements Updated Veer-off Accident and Incident Database The number of aircraft veer-off events identified during this study can be helpful for additional research in this field. The comprehensive database includes 1,144 recorded events in an organized structure to facilitate its use. The database includes veer-off events involving aircraft over 6,000 lbs of MTOW. In addition to basic information on location and date, data about the airport, the operation, and the consequences were collected when available. The veer-off pathway was characterized for approximately 50% of the records, and each report was reviewed in an attempt to identify the causes of aircraft damage to support the con- sequence model. Validated Location Models to Estimate Likelihood of Aircraft Challenging Runway Safety Subareas The objective of the model validation effort was to check if the risk estimates provided by the new models compare to historical veer-off accident rates and that probability distri- butions generated would be similar to those provided by an independent sample of veer-off reports. Results obtained demonstrate excellent agreement between probability distributions given by the location models and that of the independent sample of 91 veer-off events. Also, there was good agreement between historical accident and incident rates for U.S. airports compared to the rate for a sample of eight airports, and the rate estimated from LRSARA analyses for each airport. Some differences were identified; however, these may be attributed to the small sample size and large variations expected when modeling rare events. Finally, the validation effort has helped identify bugs in the LRSARA software and allowed them to be resolved. Sev- eral enhancements were made to the program, resulting in increased protection for inconsistent input data and improved accuracy of modeling runway veer-off risk. The models integrated to the approach were based on evi- dence of worldwide veer-off accidents and incidents collected from 1982 to 2011. The models utilize a transformation to the longitudinal veer-off distance that is based on the RDA for the operation. Two other alternatives were attempted; how- ever, more accurate results were obtained by using the selected alternative. C H A P T E R 8
41 The resulting models provide a way to estimate the prob- ability of veer-offs. The models also provide a way to assess risk when obstacles are present or proposed in the vicinity of the runway. Approach Incorporating Veer-Off Location Models to Estimate Risk In addition to developing the mathematical models for veer-off risk, an approach integrating the models in a step- by-step process was necessary to serve as the basis for the computer algorithm developed for the analysis. The approach concept is similar to that used in previous ACRP studies and the framework is based on a three-part model: event probability, location probability, and veer-off consequences. The event probability (frequency) models are those presented in ACRP Reports 50 and 51. The location probability and veer-off consequences models were devel- oped in this study. The risk-based approach introduced in this report is rational and robust, and it can be used to quan- tify the risk of veer-off accidents and evaluate how obstacles in the vicinity of the runway may impact the risk. Software Tool to Analyze Lateral RSA The standalone analysis software was named Lateral Run- way Safety Area Risk Analysis (LRSARA) and combines Micro- soft Access databases and Microsoft Excel spreadsheets to store and output results generated during the analysis. This concept provides user-friendly interfaces for inputting data, running the analysis, and outputting results. There are two sources of information required for this analysis: general data and analysis-specific data. General data are requirements that apply to all analyses, such as model parameters and aircraft characteristics. Analysis-specific data are information specific to the airport and the lateral RSA to be evaluated. The software records user input data into a project-specific database where the information can be later assessed along with the results of the analysis. Updates to general data, as well as uploads of analysis-specific data, are made through a spe- cific software interface. There is no need for user interaction with the databases; however, the databases are also available as regular Microsoft Access files. Outputs of the analysis results are reported in Excel spread- sheets. These spreadsheets are generated based on a predefined template. One spreadsheet for each runway in a given airport is created. The spreadsheet contains the summary probabili- ties of veer-off incidents and accidents. The probabilities are illustrated with graphs of accumulated risk and probability histograms. The templates for databases and spreadsheets are compatible with MS Office version 2010 and newer. Model Limitations The main challenge to develop the models and analysis tool presented in this report was the availability of reliable information to develop the mathematical probability models. Only approximately 10% of the accident and incident reports for events identified as veer-offs had comprehensive informa- tion about the pathway during the aircraft veer-offs, and 50% of reports contained no information that could be used to infer the veer-off path. To overcome these limitations, it was necessary to use the report narrative and obtain information from additional sources to make inferences to characterize the veer-off path- way. These inferences can certainly have some negative impact on characterizing the exact track during the veer-off and to a lesser degree on the accuracy of the location models developed in this study. Although some accuracy may have been lost due to the assumptions made, the information presented in this report will assist the industry in understanding the mechanisms and the relationship between risk, available safety areas, and the presence of obstacles associated with aircraft veer-off accidents and incidents. The approach will certainly improve the knowl- edge of the relationship between airfield design standards and the risk level involved when standards cannot be met. Guidance Expand the Approach to Lighter Aircraft The prevalence of and risks associated with runway excur- sions have not been addressed for aircraft weighing less than 6,000 lb, in part because no effort has been made to compile a database of those excursions. Little research effort has been spent to evaluate how design standards and non-compliance issues may impact risk of runway excursions at airports. Most aircraft with MTOW lower than 6,000 lb are operated under 14 CFR Part 91 rules, which represent close to 90% of civil aircraft registered in the United States. Many commu- nities benefit from general aviation flights, which generate over $150 billion in economic activity (AOPA). Many of the airports used by lighter GA aircraft are not certificated and have no towers that could report incidents. A research study using approaches similar to those pre- sented in ACRP Reports 3, 50 and 51 would greatly benefit general aviation airports to reduce risks of runway excursions and improve aviation safety in the United States. Improve Veer-Off Reporting The FAA has recently adopted a Safety Management Sys- tems (SMS) approach to aviation safety. In this approach, one needs to identify risks and then take measures to mitigate
42 those risks. FAA Order 8000.369A furthers safety manage- ment by moving towards a more process-oriented system safety approach with an emphasis on risk management and safety assurance. Comprehensive information required to identify the stop- ping location and veer-off path of the runway veer-offs is seldom explicitly available in accident and incident reports, except for major accidents for which a full investigation report was developed. This study identified some gaps in veer-off reporting in existing aviation databases. These gaps are related to infor- mation required to develop risk models for veer-off events and the suggestions provided are intended to enhance data collection to improve accuracy of risk models. The main suggestion is to report information to character- ize the veer-off path. It can be a drawing or a picture showing the veer-off path, or a narrative describing the incident. A template was created and is presented in Appendix A. It helps the reporter identify key information that may be provided in narrative format. The gaps in information made available in veer-off events reported were identified for the three main aviation accident and incident databases available in the U.S. Filling those gaps may be beneficial to improve accuracy of existing models.
