QUALITATIVE AND QUANTITAVE RISK EVALUATION ON THE BASIS OF MILITARY AVIATION EVENT ANALYSIS

Transcription

1 Mariusz ZIEJA Henryk SMOLIŃSKI Paweł GOŁDA Air Force Institute of Technology RESEARCH WORKS OF AFIT Issue 38, s , 2016 r /afit QUALITATIVE AND QUANTITAVE RISK EVALUATION ON THE BASIS OF MILITARY AVIATION EVENT ANALYSIS Risk management is understood as a process supporting decision making through systemic assessment of possible courses of action, identification of hazards and benefits, and indication of the best way to carry out aviation tasks. The article presents a method of qualitative and quantitative risk assessment on the basis of military aviation incident analysis. The procedure of the suggested approach is contained in a probability and severity degree estimation algorithm and in subsequent steps: of identifying the risk factor (CZN) of a given category, analysis of risk associated with a given CZN, determination of risk, defining correcting actions, risk forecasting and its verification over the following assessment period. A simplified block diagram presents a proposal of the implementation of a qualitative and quantitative risk estimation procedure with the use of the TURAWA IT system, which function in the air forces. Keywords: risk evaluation, risk management, risk management IT support 1. Introduction Aircraft (AC) flight safety management system is a useful tool for achieving and maintaining an accepted safety level and should ensure that all executed processes are covered by the safety policy, thus, with relevant monitoring and constant care that their implementation does not exceed defined safety standards [1, 6, 9]. Such management requires the use of proactive methods (and strategies) of managing flight safety, aimed at preventing aviation incidents, via accurate gathering of information from different sources, which could be indicators of safety problems [7].

2 86 Mariusz ZIEJA, Henryk SMOLIŃSKI, Paweł GOŁDA This approach is focused on revealing factors leading up to dangerous aviation events and developing preventive measures aimed at minimizing the number of such events. The objective of the proposed approach to flight safety management is to, with the use the analysis of gathered data, show the problems preceding dangerous aviation events and to undertake preventive actions, even before the occurrence of an emergency situation. Activities in that field are referred to collectively as risk management, based on risk analysis (or otherwise, risk control). Military aviation has implemented Risk management methodology in the Polish Air Forces, in order to define (standardize) the procedures in the field of risk management, when making decision in planning and the execution of aviation tasks [4]. Risk analysis, as per the a/m methodology, is carried out for particular aviation tasks (flights) and air operations. Quantitative risk assessment is carried out in the 5M model 1, which is the basis to make a decision about the execution of an aviation task. The basic documents of risk analysis are the risk management sheet and the risk estimation card [8]. The methodology proposed in this article, using the approach to risk management presented in [3 and 5], leads to quantitative and qualitative risk estimation on the basis of flight event analysis and can be used to assess flight safety through risk control (risk management) in military aviation. The procedure is contained in a probability and severity degree estimation algorithm and in subsequent steps: of identifying the risk factor (CZN) of a given category, analysis of risk associated with a given CZN, determination of risk, defining correcting actions, risk forecasting and its verification over the following assessment period. 2. Output information structure It is important for flight safety management to utilize all, available information on events, deviations, comments appearing during flight training. Currently, you can use information on aviation events gathered in SI TURAWA, which functions in military aviation. The number of recorded aviation events (accidents and incidents), over the course of a year, is ca. 2 thousand. We can say that in the future, the collection and scope of information gathered in that system will be growing. 1 Man Methods Machinery Materials Management.

3 Qualitative and quantitave risk evaluation on the basis of military aviation The classification of aviation event causes, according to the areas of the 5M model, determined in Flight safety manual for the Air Forces of the Republic of Poland [2], is adopted. In each area, you can segregate categories according to the principle of indicator commonality, e.g. E1, C1 (Table 4). Environment, natural and technological (e.g. flight logistics without SIL, flight traffic) and FOD (E) category E1; Machine, aircraft, SIL personnel (T); Man (flight personnel, crew) (C) category C1; Management (O): managing flight personnel (OC), managing SIL personnel (OT), managing personnel associated with the technological environment (OE). Potential consequences of the hazard factor impact, in the form of material losses, are presented in Table 1. Potential material losses caused by the impact of the hazard factor [9]. KSi In-flight situation Material losses 5 Catastrophic >1m PLN 4 Emergency 0.1 1m PLN 3 Complex (complicated) k PLN 2 Complication of flight conditions 1 10k PLN 1 No consequences <1k PLN Table 1 3. Risk calculation Risk is considered as a measure of quantitative safety measurement with a combination of the size of losses resulting from CZN impact and the probability of a CZN occurring. Both risk components: probability and severity are numerically expressed degree (rank) indicators of probability P and severity S rank. Risk R in standard units is calculated from the formula [3]. R = P S (1) At the same time, it is understood that the results of any calculations associated with an occurring, disclosed CZN, are considered only as output data for the

4 88 Mariusz ZIEJA, Henryk SMOLIŃSKI, Paweł GOŁDA forecast. Talking about risk of events that happened does not make sense since risk is always the estimation of the uncertainty of future activity. Risk calculations should be carried out for each category; they can be implemented in the analytical module of the IT system proposed in Fig. 3. The basic assumptions of the computational algorithm are presented below. 4. Calculating the degree of severity The severity degree S is calculated for a category. The output data for the calculations are expert estimates of severity (KSi) for each event, according to an adopted scale of in-flight situation assessment (defined for military aviation). Estimation categories are developed by experts for their group. A five-point scale of in-flight situation assessment was assumed for the expert estimation of each KSi event (Table 2). The assumed values provide an approximate exponential relation. Scale of in-flight situation assessment Table 2 KSi Si Each Si value is considered as the implementation of the random value S the severity of occurrence of a group of hazard factors for a given category. With this approach, the upper limit of S assumed for probability, can be calculated using the Chebyshev inequality: DS ( ) P( S MS ( ) α ) (2) 2 α where: M(S) and D(S) S; mathematical expectation and dispersion of the random value S; α deviation S from the mathematical expectation with the probability P. For the working formula, the probability needs to be assumed and the S value obtained as an estimate of severity of the category. It the sum MS ( ) + Cσ ( S) exce- max eds the highest of the estimates S i, it is recommended to assume S = S max i. For example, for the confidence level P=0.75, we have the following relation (rule):

5 Qualitative and quantitave risk evaluation on the basis of military aviation max ( MS ( ) + 2 σ( S); MS ( ) + 2 σ S ) i S = max max ( Si ; M( S) + 2 σ( S) + 2 σ( S) > Si (3) The obtained estimate is considered as a forecast of the events degree of severity in a given category, for the next reporting period. 5. Calculating the degree of probability The estimation of the degree of probability P should be carried out by an expert in each particular case. Of course, the estima50123te will be subjective but based primarily on the repeatability of the events. The dependency of the degree of probability P from the frequency of events F needs to be assumed when estimating risk. Frequency of events for military aviation is presented in Table 3. Frequency of events for military aviation Table 3 Indicator of the degree of probability P Descriptive characteristics P Extremely unlikely Unlikely Slight Sporadic Frequent Average assessments of frequency F according to [5] The relation F = f (P) is roughly approximated with an exponential function (fig. 1). When solving an equation with an exponential tendency in relation to x, we will obtain a formula to calculate the indicator of the degree of probability P: P = Aln( F) + B (4) F can be expressed in units of events per 1000 flights, as adopted in certain airlines and we can assume expert numerical estimation of frequencies in average values according to categories of own sector.

6 90 Mariusz ZIEJA, Henryk SMOLIŃSKI, Paweł GOŁDA The proposed estimation of the degree of probability according to event frequency in fig. 1 may be verified for the adopted categories by military aviation experts. Częstotliwość Frequency Stopień Degree prawdopodobieństwa of probability Fig. 1. Averaged degree of probability estimates

7 Qualitative and quantitave risk evaluation on the basis of military aviation Risk register Code Category Date CZ N E1 Damage of engine compressor blades by foreign bodies Failing to inspect the engine intake duct or inaccurate cleaning of the runway and taxi way. C1 Exceeding acceptable inflight engine torque values Improper control of the in-flight engine torque values. Description of risk associated with a given CZN (CZN group) Possibility of severing a blade of the air engine. Failure to detect damages to the compressor blades in the proper time. Big material losses due to the possibility of damaging the structure of the propulsion unit. R Corrective actions Training on the engine use in-flight, for the whole flight crew. Conducting individual talks with crews. Responsible Training in the scope of damages to the engine by foreign bodies, for all SIL personnel and airport services. Inspection of the technical condition of the runway and taxi way surfaces. Inspection of the runway and taxi way cleaning system. Commander Commander Date of inspection R forecast Table 4 R actual

8 92 Mariusz ZIEJA, Henryk SMOLIŃSKI, Paweł GOŁDA 6. Risk acceptability calculations and estimation The degrees of severity and probability are calculated from the formulas (3) and (4) and the degree of risk indicator is calculated from the formula (1). Next, the degrees of risk are determined and the risk is estimated, e.g. negligible R<5; unaccepted under the condition of a reduction to 5 R 20; unacceptable R >20. The calculated risk values R are compared with determined levels and preemptive or correcting actions are developed. R R unacceptable R acceptable R insignificant 0 E1 C1 T OC OT Category Fig. 2. Graphical presentation of the risk of aviation event categories The computational algorithm and values of acceptability levels can change as data are gathered and the as the analytical model is upgraded. Fig. 3 presents a simplified block diagram of an IT system supporting risk management. The suggested concept requires an expansion of the TURAWA IT system currently functioning in the air forces with: an analytical module, which enables calculating the severity degree of a category, the probability degree of a category and the category risk. An analysis group should be created to work with the system. The analysis group (experts) should be formed from experienced specialists. Their duties would include estimating the importance of events and developing reports. The expert should enter characteristic CZNs into the system according to each category of increased risk and, if need be, once again review the events, which he or she estimated (assessed). For each CZN, undertakings are recommended, which, after agreeing with relevant supervisors and after designating a responsible person, are

9 Qualitative and quantitave risk evaluation on the basis of military aviation entered in the register (Table 4). The effectiveness of the actions is assessed on the basis of results of calculating risk in the next reporting period. Graphs are drawn up for the graphic presentation of risk, with an example of such a graph presented in Fig. 2. Module for entering data on aviation events Event severity assessment module SI TURAWA database Report formulating module Calculating severity degree of a category Calculating probability degree of a category Risk registers, graphs Calculating category risk Analytical module Correcting (pre-emptive) actions regarding risk management Monitoring the effectiveness of corrective measures Fig. 3. Simplified block diagram of an IT risk management system 7. Conclusions The presented proposal of risk management in military aviation, using the qualitative and quantitative risk estimation approach, implemented in some civil airlines, can be a significant supplementation for the current method of risk management, on the basis of Risk management methodology in the Polish Air Forces.

10 94 Mariusz ZIEJA, Henryk SMOLIŃSKI, Paweł GOŁDA An advantage of the presented approach is the possibility of quantitative risk estimation, using IT support in risk management, calculating risk in the next reporting period and its monitoring (determination of trends), on the basis of which, the effectiveness of correcting actions is assessed. The suggested concept requires an expansion of the IT systems currently functioning in the air forces with an analytical module, which enables calculating the severity degree of a category, the probability degree of a category and the category risk. Literature 1. Babak V.P., Caraczenko V.P., Maksimow V.O. et al.: Bezpieka Awiacji. [Aviation Security] Technika, Kijów Instrukcja bezpieczeństwa lotów lotnictwa Sił Zbrojnych Rzeczypospolitej Polskiej [Flight safety manual for the Air Forces of the Republic of Poland]. MON, Kulavskiy V.G., Sharon V.D., Kudryavtsev A.A.: Zarządzanie eksploatacyjnym bezpieczeństwem w liniach lotniczych [Managing operational safety in airlines], Problemy Bezopastnosti 2011, Nr Metodyka zarządzania ryzykiem w lotnictwie Sił Zbrojnych RP [Risk management methodology in the Polish Air Forces], Warszawa 2010, MON. 5. Podstawy zarządzania ryzykiem w lotnictwie. [The basics of risk management in aviation] Dowództwo Sił Powietrznych, Warszawa Safety Management Manual (SMM), Second Edition. ICAO, Zieja M., Smoliński H., Gołda P.: Proactive methods new quality in aircraft flight safety management. Journal of KONBiN 2015, t. 4 (36). 8. Zieja M., Smoliński H., Pigłas M., Gołda P.: Zarządzanie ryzykiem w procesie eksploatacji wojskowych statków powietrznych. [Risk management in the aircraft operational process] Prace Naukowe Politechniki Warszawskiej Transport 2015, z Zintegrowany System Bezpieczeństwa Transportu. [Integrated Transport Safety System] Tom I: [Volume I] Diagnoza bezpieczeństwa transportu w Polsce [Diagnosis of transport safety in Poland], ed. R. Krystek, WKŁ, Warszawa 2009.