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.
37 RSA standards have changed over the years to improve safety, but many existing airports were built under older, less demanding standards. To comply with the current standards, some airports face enormous challenges due to physical, eco- nomical, and environmental restrictions. Safety levels associated with the protection provided by RSAâs can be different from airport to airport. Two airports with similar runway lengths and RSA configurations may have very different conditions related to operations and weather. Factors like aircraft model, runway elevation, visibility condi- tions, and availability of NAVAIDâs also have an impact on the risk of each operation. When airports do not comply with the RSA standards and there is a need to improve existing conditions, it is necessary to evaluate the alternatives that can be most effective to reduce risk and compare the safety levels achieved and the associated costs for each option. The objective of this study was to develop a software tool that can be used for risk assessments associated with incidents occurring in the RSA. The basis of the approach used in this study was presented in ACRP Report 3. Analysis capabilities were enhanced by improving the risk models to address the analysis of runway declared distances, the use of EMAS, and incorporating the approach to evaluate the presence of obstacles in or in the vicinity of the RSA. In addition, it is now possible to evalu- ate the risk of aircraft veer-off in the lateral sections of the RSA. The result is a powerful tool to help the aviation indus- try perform risk assessments. The major goals of this study can be summarized as follows: 1. Update the ACRP Report 3 accident/incident database to incorporate aircraft overrun and undershoot accidents and incidents occurring after 2006 and include runway veer-off events occurring since 1980. 2. Develop risk models for frequency and location for each type of incident. 3. Develop a practical approach to assess the impact of: runway distance available on the probability of overruns, under- shoots, and veer-offs; the availability of EMAS as an alterna- tive to standard RSAs; the use of declared distances; and the presence of obstacles in the RSA or its vicinity. 4. Develop user-friendly software that incorporates the meth- odology and models developed as a practical tool that air- port stakeholders may use to evaluate RSA alternatives. 5. Field test the software developed and validate the new tool and models based on data gathered according to an airport survey plan. Each of these goals was accomplished, and the major achievements are presented below. Major Achievements Extended Database of Accidents and Incidents The database developed under the study presented in ACRP Report 3 included 459 aircraft overrun and undershoot acci- dents and incidents occurring from 1980 to 2006. The database was expanded significantly to 1414 events with the inclusion of overruns and undershoots occurring from 2006 to 2009, and the addition of veer-off events and information provided by MITRE. Additional events were identified using a manual search of the FAA incident databases and the accidents and incidents in- volving GA aircraft with MTOW between 5,600 and 12,000 lb, which had been excluded from the ACRP Report 3 study. The comprehensive database is organized with editing and querying capabilities, and information is available according to different categories including synopsis of the event, air- craft involved, airport and weather characteristics, level of consequences, wreckage location, and major causal and con- tributing factors. C H A P T E R 7 Conclusions and Recommendations for Further Research
Development of Improved Risk Models The models presented in ACRP Report 3 were improved, and new ones to address veer-offs were developed. Five sets of frequency and location models were developed, including models for LDOVs, LDVOs, LDUSs, TOORs, and TOVOs. These types of events constitute the great majority of aircraft incidents that challenge the runway RSA. New data and new factors were incorporated into the new models. Two of the most important ones were the runway criticality factor and the tail/head wind component. The run- way criticality factor was defined as the ratio between the run- way distance required and the runway distance available for the operation. The higher this value is, the smaller is the safety margin for the operation, and it represents the relationship between the runway and aircraft performance. Development of Approach to Evaluate EMAS EMAS has proved to be a successful alternative when the RSA area available at the runway ends is shorter than the standard. The improved deceleration capability provided by EMAS is an important consideration when performing an RSA analysis. The approach presented in ACRP Report 3 did not address the possibility of using EMAS; ACRP 04-08 filled this gap. A simplified approach based on data provided by Engineered Arresting Systems Corporation (ESCO), the manufacturer of EMAS, was developed and incorporated into the software. The approach used can help airport stakeholders verify the safety benefits of using EMAS beds, even when non-standard configurations are used. Development of Approach to Assess Impact of Declared Distances Statistics were used to demonstrate that the likelihood of landing and takeoff incidents may depend on the safety mar- gin available for the operation relative to the runway distance required by the aircraft. In this project, the estimate of frequency of incidents incor- porates a runway criticality factor defined as the ratio between the runway distance required and the distance available. Al- though the runway distance can only be calculated using the actual aircraft weight, and this information is difficult to ob- tain, other factors may be used for modeling. Some of these factors include the basic distance required for standard con- ditions, the runway elevation, the air temperature, the wind, and the runway surface conditions. In this project, the land- ing distance required is estimated based on each of those fac- tors for the specific type of aircraft. The incorporation of these factors into the frequency mod- els is used to help assess the impact of the declared runway distances on risk of overruns, veer-offs, and undershoots. Development of Software Tool The approach and the improved models were integrated into analysis software for risk assessment of RSA. The tool, called RSARA, is user-friendly and practical, and allows the user to consider each of the factors impacting RSA risk. The software works as a simulation tool to estimate risk for each operation from an annual sample of operations for an airport. The historical sample data include flight operations data, like aircraft model, runway used, and the type of oper- ation, as well as the weather conditions to which each of these operations was subjected. Within the software, the definition of RSA areas is a very simple process based on Microsoft Excel spreadsheets. The procedure is as easy as drawing the RSA in a plan view and defining the RSA surface type: unpaved, paved, or EMAS. The output is comprehensive, and risk estimates are pro- vided by type of incident, runway, and RSA section challenged. Risk results are provided in terms of accidents per number of operations or the expected number of years to occur an acci- dent, and are compatible with the criteria set by the FAA. Histograms of risk help users identify the percentage of operations subject to risk levels higher than a desired TLS. Model and Software Validation The risk models were developed and calibrated based on a worldwide dataset of accidents and incidents. A second effort was conducted to verify and validate the models using NOD and RSA conditions for eight airports that were not used to create the NOD used to develop the models. The verification was a key step to demonstrate the applica- bility of the innovative approach and models developed in this research. The comparison between estimated and actual fre- quency and risk rates showed excellent agreement, despite the small sample of airports used in this study. Analysis output for the eight airports and their historical records of accidents and incidents helped to prove the validity of the approach and analysis software. The volunteers selected to test the models provided feed- back to the research team that was used to improve software and eliminate bugs. Limitations Although an intensive effort was made to develop a very comprehensive tool, there are some limitations that users should be aware of. Some of those limitations are related to data availability, and some are related to the computer time to perform an analysis. One important limitation is that the tool is helpful for planning purposes only. Neither the models nor the software should be used to estimate risk during real-time operations. 38
Only aircraft manufacturers can use actual aircraft data dur- ing operations to estimate actual aircraft performance. The models and the approach were developed using actual data from accident and incident reports, and the models are simply based on evidence gathered from that type of informa- tion. For example, to estimate the runway distance required, a basic distance for the type of aircraft and model was used and corrected for temperature, elevation, wind, and surface charac- teristics. Wind corrections are considered to be average adjust- ments, and surface conditions are estimated based on weather conditions only, rather than relying on actual runway friction. It was not possible to incorporate the touchdown location or the approach speed during landing. These are important factors that may lead to accident, but there are no means to account for these factors, except for assuming average values with a certain probability distribution that will lead to some level of model uncertainty. Additional simplifications were necessary to address the in- teraction of incidents with existing obstacles. In many cases, the pilot is able to have some directional control of the aircraft and avoid some obstacles. The approach simply assumes that the aircraft location is a random process and the deviation from the runway centerline follows a normal probability dis- tribution and that, during overruns and undershoots, the air- craft follows a path that is parallel to the runway centerline. One major limitation to obtain more accurate models and estimates is the difficulty in accounting for human factors. Some type of human error was present in the majority of the events reported, and this factor is reflected as a component of the model error. Also, obstacle categories were defined according to the max- imum speed to avoid an accident with substantial damage to the aircraft and possibly injuries to its occupants. The classifi- cation was defined in this project using engineering judgment and assuming that consequences depend only on the collision speed and the area of the aircraft that has collided with the ob- stacle. Again, only engineering judgment was used to classify different types of obstacles according to the categories. Recommendations for Future Work Extend Analysis for Non-RSA Areas Even with its limitations, the approach presented in this report is quite robust for the analysis of RSA. It takes into consideration many local factors and specific conditions to provide estimates of risk. Still, the analysis presented can only cover the areas in the immediate vicinity of the runway. The development of a risk- based methodology to evaluate land use compatibility and third-party risk could be very helpful to support State require- ments and planning efforts. The approach can be similar to the one presented in this studyâusing evidence of aircraft ac- cidents in the vicinities of airports to develop risk models based on causal and contributing factors to aircraft accidents. The study should address the risk of accidents in areas within a 10-mile radius of the runway. The methodology should consider local factors, historical operation conditions for the airport, and the type of land use for specific regions near the airport runways. The recommended study would improve the capability of land use committees and airport operators to assess third-party risk associated with air- craft accidents in the vicinity of airports. The approach should be rational, non-prescriptive, and provide a quantitative assessment of third-party risk associ- ated with aircraft operations at a specific airport. The study should associate aircraft operations with existing runway and environmental conditions, and aircraft type for a specific air- port. Thus, the results of such a study would help decision makers to evaluate alternatives and associated safety benefits. Development of Risk Tool for Airspace Analysis in Vicinity of Runways The RSA analysis methodology and software presented in this study can only address the ground roll phase of op- eration; however, aircraft have both lateral and vertical de- viations from their nominal flight path during landing and takeoff operations. Currently, the aviation industry still relies on the Collision Risk Model (CRM) developed in the 1960s with very limited data to evaluate risk during instrument approaches during the non-visual segment and missed approach phases. The CRM has many limitations and does not cover all phases of the flight and types of approach. Only data for precision approach Cat- egories I and II can be evaluated using the existing model. There is a need to have an updated CRM that can be used to prioritize risk mitigation actions associated with obstacles in the vicinity of the runway. Currently, the FAA is developing a tool called Airspace Sim- ulation and Analysis Tool (ASAT) that has comprehensive capabilities and accounts for aircraft performance, NAVAIDs, environmental conditions, terrain, wake turbulence, and hu- man factors. However, the tool is not available to other air- port stakeholders. The improved tool should have the capability to assess risk associated with fixed or movable obstacles when they are very close to the runway. It should address all types of approach (visual, non-precision, precision, and possibly global position- ing system [GPS] approach technology). Many airports would benefit from such a tool for safety management associated with the presence of obstacles. 39