View non-flash version
April 2012 www.sname.org/sname/mt research in this area, with the methods developed showing speed increases of up to 20 times when compared to traditional methods, without loss in diagnostic ability. In addition, these methods are fully automated; human interaction in the process is not required. BlueFlow Ltd., a commer- cial spin-out entity for the exploitation of this technology is already in existence, and future work will focus on developing a fun- damental research program at the university to deliver software solutions that give short term gains, as well as meet the future needs of the industry. ese techniques have been successfully applied to a number of di erent elds, including econometrics (the study of nancial markets), bioinformatics (in par- ticular genomic and proteomic analysis), engineering problems, and the analysis of seismic data. Some of these new technologies being developed to solve current problems also apply to the signicant upsurge of data in the oil and gas industry. Automated methods for the analysis of such data are imperative, because without them it will not be possi- ble to e ectively analyze the data at all, as currently happens with the huge amounts of data generated in this sector. e devel- opment of technology in this area has two benefits. The first is that the cost for data analysis will be significantly reduced in terms of the expensive software required, sta training, sta costs, and time taken for the analysis; the second is that the data will be properly analyzed, meaning that cost benets can be identied through the better use of the available data. ese automated techniques will have potential in many areas outside of the oil and gas sector and will feed into reliability modeling of subsea assets. The inspection, maintenance, and eventual replacement of deteriorating assets are a vital part of responsible asset ownership. e safety, environmental, and commercial impacts of an unplanned cat- astrophic failure are considerable. Subsea infrastructure needs to perform over a long time period, with minimal opportunities for corrective action during its operation. The timing of inspection, maintenance, and renewal is of critical importance, and must reect the condition of the asset and the consequences associated with its fail- ure. Advanced reliability techniques make use of a variety of data sources from sensor measurement to visual inspection to char- acterize material properties, geometries, process conditions, and loads as random variables. ese can be used in conjunction with existing or new limit-state functions, which describe the system condition events of interest including failure. The deterioration of assets via physi- cal and electrochemical mechanisms, together with changes in reservoir pres- sures and temperatures, mean that both time and the spatial variation of deteriora- tion must be incorporated into the reliability models. Research into the use of advanced reliability techniques capable of using data from subsea sensors is required to develop models capable of predicting how the assets reliability changes with time and local- ized damage phenomenon. e results can be used to make risk-based decisions on interventions that minimize the combined costs of operating expenditures, capital expenditures, and risk associated with infra- structure failure. Traditional lifecycle models have used historical failure data to provide insight into future performance. The use of more advanced reliability modeling using a com- bination of limit state-functions and an assessment method based on either simu- lation (monte-carlo based techniques) or analytical solutions such as the rst/second order reliability methods have received some attention by the pipeline and the structural engineering communities. eir success in prediction of system performance has seen them formalized in a number of design codes. However, the incorporation of time- varying quantities such as material properties, geometries, pressures, and cor- related spatial variation of deterioration mechanisms are not commonly in use and new methodologies are required. e n ran- dom variables contained in the engineering problem must be replaced with stochastic processes and random elds, changing the solution methods required. Incorporating changes with time and spatial position is vital to the accurate description of subsea system reliability and will form the focus of the research e ort. Cross-industry application e energy sector relies heavily on technol- ogy to deliver its products, which operate in extremely harsh and challenging environ- ments. With the move to deep water, reliable monitoring of the integrity and remaining life/reliability of subsea systems has become even more critical. Although the thrust of the work on new sensor, communication, data mining, and reliability modeling technologies examined here is of primary interest to the oil and gas sectors, there are potential benets to other sectors, such as sheries/aquaculture and onshore process industries. MTFor information about the authors of this article, see the Feature Contributors section on page 5 in this issue. Further Information To learn more about the issues described in this article, check out the following resources. Kiefer, J., Seeger, T., Steuer, S., Schorsch, S., Weikl, M.C., and Leipertz, A, Design and characterization of a Raman-scattering-based sensor system for temporally resolved gas analysis and its application in a gas turbine power plant,? Measurement Science & Technology 19 (2008). de Angelis, M., et al., Precision gravimetry with atomic sensors,? Measurement Science & Technology 20 (2009). Sputh, B.H.C., Faust, O., and Allen, A.R., RABP: Point-to-Point Communication over Unreliable Components, Communicating Process Architectures? (2008). Davidson, S., Starkey, A., and Mackenzie, A., (2009) Evidence of uneven selective pressure on di erent subsets of the conserved human genome; implications for the signi?cance of intronic and intergenic DNA,? BMC Genomics 10 (2009). Straub, D. and Faber, M.H., Risk based inspection planning for structural systems,? Structural Safety 27 (2005).