Model for risk and reliability analysis of complex production systems: Application to FPSO/flow-Riser system

https://doi.org/10.1016/j.compchemeng.2008.11.017Get rights and content

Abstract

A model for risk and reliability analysis of complex multifunctional production process systems is presented. The model employs Monte-Carlo and Markov Chain algorithms that uses a weighted index to train and simulate the fuzzy hazard data sets which represents failure outcomes of risk component transient and non-transient systems. Early simulation results shows that hazard rates and the risk of containment loss from typical floating production and storage offloading (FPSO)-Riser system for the risk components in parallel or series increases exponentially with time and decreases as safety ratings fraction increases. The reliability value decreases with time and safety fraction (SFAC) for all fuzzy hazard classifications. The results of the computed mean time before repair (MTBR) show that the minimum computed years before repair range from about 0.5 computed year for worst case (fuzzy class 1, very likely to fail) to almost 5 million computed years for the best case (fuzzy class 5, remote to fail) assuming availability is 80%. This new method for risk assessment would allow users of the technique generate skewed failure data hazard rates to predict actual failure outcomes of multifunctional processes and complex risk systems.

Introduction

There are many methodologies proposed for reliability, risk and safety analysis for most, if not all process industries known today. Among the most popular ones are quantitative risk analysis, probabilistic safety analysis, worst-case methodology for risk assessment and optimal risk analysis. The optimal risk analysis (ORA) has been reported as most suitable as it is fast, less expensive to implement, less time-consuming and more precise than alternative analysis according to Khan and Abassi, 1981, Khan and Abassi, 1998. ANSI/ISA S84.01-1996 is the consensus standard for process safety in the U.S, deemed to meet the OSHA 1910.119 PSM regulation.

According to the Offshore Technology Report 2001/063 for marine risk assessment HSE (2002), the use of risk assessment techniques in major hazard industries has grown significantly in recent years. This is particularly true in the offshore industry in the United Kingdom (UK) where many safety aspects are subject to full risk assessment, notably the Temporary Refuge Assessment which uses Quantitative Risk Assessment (QRA).

Other aspects related to offshore facilities such as marine operations and management of the safety aspects of FPSO-Riser-Pipeline Operations have rely on meetings, regulatory requirements, industry codes of practice, or classification society rules. Marine hazards are diverse in nature, and is defined as any potential accident on an offshore installation connected with its interface with the marine environment, which are (i) loss of position keeping (e.g. mooring, dynamic positioning, rig move); (ii) loss of structural integrity (e.g. hull, ballast tank, support structure failure); (iii) loss of stability (e.g. ballast system failure, cargo loads); (iv) loss of marine-utility systems (e.g. propulsion, power generation, hydraulics) and (v) collision (e.g. shuttle tanker, support vessel, passing vessel).

According to Liu, Yang, Wang, SII, and Wang (2004), the growing technical complexity of large engineering systems such as offshore platforms and offshore support vessels led inevitably to problems of uncertainty in representation. Consequently; the risk assessment and reliability model of large engineering system affected by many factors regarding its design, manufacturing, installation, commissioning, operation and maintenance became extremely difficult to construct because of inadequate knowledge about the basic failure events. An, Wang, and Ruxton (2000) proposed that one realistic approach to deal with imprecision especially for offshore engineering projects is to use linguistic assessments instead of numerical values.

Zadeh (1965) and Zimmerman (1991) uses fuzzy logic approaches employing fuzzy IF–THEN rules, where the conditional part and the conclusions contain linguistic variables used to model the qualitative aspects of human knowledge and the reasoning process without employing precise quantitative analysis.

Liu et al. (2004) argued that fuzzy rule-based systems is severely limited if only fuzziness is taken into account in representing uncertain knowledge due to increasing complexity of many knowledge-based systems.

Other theories of uncertainty in fuzzy rule-base have been proposed to model various types of uncertainty, such as type 2 fuzzy sets theory (Karnik et al., 1999, Mendel, 2001, Mendel and John, 2002; Mizumoto and Tanaka, 1976; Zadeh, 1975) and intuitionistic fuzzy sets theory by Atanassov, 1986, Atanassov, 1999 and De, Biswas, and Roy (2001).

The Decision System (D-S) theory enlarges the scope of traditional probability theory. This theory describes and handles uncertainties using the concept of the degrees of belief, which can model incompleteness and ignorance explicitly. Besides, the D-S theory also shows great potentials in MADA (Multiple Attribute Decision Analysis) under uncertainty, where an evidential reasoning (ER) approach for MADA under uncertainty has been developed on the basis of a distributed assessment framework and the evidence combination rule of the D-S theory (Yang, 2001, Yang and Sen, 1994, Yang and Sen, 1997, Yang and Singh, 1994, Yang and Xu, 2002a, Yang and Xu, 2002b).

The limitations of conventional risk and reliability systems which are (i) complexity of interacting risk events in multifunctional systems from use of conventional risk analysis techniques and models to specific risk systems; (ii) questionable hazard and failure data; (iii) lack of performance based models for reliability and risk analysis in variable hazard rate systems (iv) mostly empirical-based and system-specific models which relies heavily on history data and testing to determine the hazard rate are impracticable and not useful for new designs.

These problems have led to the search for better predictive models based on qualitative and quantitative descriptions. This paper reports the development of a risk and reliability model using weight index structures. The weight index incorporates contributions of the failure mode, effect analysis, and reliability of safety systems of complex production and process systems in risk and reliability modelling. Transient failure hazards rate was simulated using Monte Carlo simulation fuzzy sets classification as well as the Markov Chain process. This allowed the possibility of studying the interaction of complex risk system with intrinsic and extrinsic safety systems or devices. Typically conventional models for risk or reliability analysis whether exponential, Poisson, Weibull, Binomial are all considered to be non-transient failures; that is; the hazard rate remains constant with time. However, the Markov Chain algorithm allows the analysis of risk and reliability systems when the hazard rate state varies in transient modes. In this paper, the weighting concept introduced allows the hazard rates to be considered under the limitation of risk systems constrained within safety systems used to safeguard the process equipments and systems; considered for transient and non transient failures in parallel and series risk systems. The results of the computed mean time before repair (MTBR) show that the minimum computed years before repair range from about 0.5 computed year for worst case (fuzzy class 1, very likely to fail) to almost 5 million computed years for the best case (fuzzy class 5, remote to fail) assuming availability is 80% which falls within expected thresholds.

Section snippets

Risk modelling using fuzzy class weight index

In this section, the concept of weighting function as it applies to risk and reliability modelling is introduced and discussed. Weights that are assigned to risk variables in series is obtained by the product sum of the hazard rate λi raised to a weighting function ω, for in, interacting hazards. Different probability models have been used to evaluate several types of failures systems, such as exponential, binomial and Poisson distribution. In this chapter, the exponential risk and

Model applications to FPSO-flow line Riser pipeline system

Fig. 1 shows a diagrammatic representation of a typical FPSO with riser systems connected. Typically the main production flow lines transports fluids from producing wells to the FPSO system. The maximum pressure for a typical FPSO facility has tubing head pressure in wells approximately 5000 psi. The flow line transports produced fluids from the manifold to the FPSO—with an inlet separator pressure of 10 bar, downstream of the surface choke. Risers connect the flow lines to the FPSO system

Data presentation and simulation

Historical hazard rate and failure data are not always available for new FPSO design projects. Thus in this paper, failure data were derived from a Monte Carlo algorithm using concept of fuzzy class. Table 3 was adapted from article reported by Carter, Hirst, Maddison, and Porter (2003). Five classes of failure are identified, very likely (fuzzy class 1), likely (fuzzy class 2), unlikely (fuzzy class 3), very unlikely (fuzzy class 4) and remote to fail (fuzzy class 5) presented in the risk

Discussions of simulated results

A computer program was developed using a structured program to simulate the hazard and risk profile, prepared using MS Excel Spread Sheet.

Conclusion

A new model for analysing risk and reliability systems based on the weighting function concept has been presented. The method allows the flexibility to assess, MTBR, MTBF, risk and reliability potential in analyzing complex interacting hazard and risk events under different safety ratings. It has been applied to model the risk and safety potential of flow line-riser system in a deepwater environment. The results of simulation shows that risk potential for containment loss from the floating

Acknowledgements

The author acknowledges and expresses profound thanks to the University of Lagos; National Petroleum and Investment Management Services (NAPIMS), DPR (Directorate of Petroleum Resources) and Shell Nigeria Exploration and Production Company Limited (SNEPCO) for providing support for this research work.

References (25)

  • B.S. Dhillon et al.

    “Engineering reliability: New techniques and applications

    (1981)
  • HSE (2002). Marine Risk Assessments Offshore Technology Report 2001/063. Health and Safety Executive, HSE...
  • Cited by (21)

    • Risk assessment of FPSO topside based on generalized Stochastic Petri Net

      2021, Ocean Engineering
      Citation Excerpt :

      Santos et al. (2020) investigated the Performance Shape Factor (PSF) that affects the safety operation, based on the decision-making model of Analytic Hierarchy Process (AHP), and then analysed the related accidents of FPSOs. Abhulimen (2009) proposed a risk and reliability analysis model for a typical FPSO riser system, based on the Monte Carlo method and Markov chain algorithms. The fuzzy risk data sets of transient and non-transient failure results of system equipment were obtained through weighted index training and simulations.

    • A systems based approach for financial risk modelling and optimisation of the mineral processing and metal production industry

      2016, Computers and Chemical Engineering
      Citation Excerpt :

      Some methodologies to quantify these risk indices are exemplified in (Abhulimen, 2009; Reniers et al., 2011). In Abhulimen (2009) various hazard data have been classified into fuzzy sets which correspond to the failure outcomes of the risk components. A Monte Carlo and Markov chain algorithm is then used for training and simulation to give a weighted risk index.

    • A reliability evaluation method based on the weakest failure modes for side-by-side offloading mooring system of FPSO

      2016, Journal of Loss Prevention in the Process Industries
      Citation Excerpt :

      The comprehensive assessment of hydrodynamic response is essential at the design stage, whereas it is incapable of predicting the safety levels of specific components. Therefore reliability analysis of SBS offloading mooring system has practical significance (Lars Tronstad et al., 2009; Abhulimen, 2009; Stiff et al., 2003). Reliability analysis is an appropriate way to address these problems, which is implemented by classifying the reliability of components, searching the weakest links of a system, and providing proper security measures.

    View all citing articles on Scopus
    View full text