The Project objective has been pursued by refining methods and accumulating field experience and data on AE attenuation and operation background noise on full-size power units, by developing and validating new effective procedures for on-line localization and discrimination of AE sources in the high background noise environment of a running plant (3-transducer binary time-correlation algorithm) and by simulating in shop tests on full size mockups, preferably with service-induced cracks, the thermomechanical transients experienced by power unit components, in order to investigate and characterize the AE behaviour of cracks under realistic non-stationary loading and thereby improve the AE detection and discrimination capabilities in operating plant conditions.
The results of this extensive preparation work have been applied and have undergone field validation in 4 pilot applications of AE to long term, on-line, continuous structural monitoring, by means of a dedicated AE equipment, tailored to the needs and requirements of continuous monitoring applications (unsupervised long-term operation, real-time diagnostic answer, low number of sensors, effective rejection of background noise and of scattered AE sources, affordable cost). The above applications were related to potential structural integrity problems of important items (2 steam headers, 1 steamline weld and 1 steamline elbow) of full-size power units and have therefore represented a severe test of the viability of AE monitoring, stimulating the effort to define and assess criteria for diagnostic evaluation of the large amount of data provided by on-line AE surveillance in different operating conditions (steady load, load variations, startup / shutdown transients). In-plant experience was also useful to identify most promising AE monitoring applications (namely to SH headers).
No stable crack growth event could be positively identified and demonstrated in the long term monitoring period (over 1 year in the case of SH headers). However, the ability of the monitoring equipment to withstand long term plant operation with no breakdown, providing a continuous stream of meaningful, preliminarly screened and compacted data to both the plant operator and to the remote expert, and to avoid false alarm calls by a judicious definition of data evaluation criteria, has been amply demonstrated.
At the Project outset it has become clear that, due to the infrequent occurrence of relevant on-stream discontinuous crack propagation in well run and maintained main plant equipments, to the understandable priority given, whenever possible, to immediate remedial action, and to the confidentiality associated to structural integrity problems, confirmation of results and building of a widespread consensus about the effectiveness of AE for structural monitoring purposes will be a slow and gradual process, requiring a number of significant, long-term applications on full-size plants, with decisive involvement and cooperation of the end users.
From the insight gained during the Project and, specifically, during in-plant experiences, AE monitoring procedures and equipment can be most readily exploited as a potentially cost-effective "risk management tool": AE is an attractive option when a structural integrity problem exists, that might imply service-induced crack propagation and potential hazards, and there is a need to safely continue power unit operation for some time, so as to fulfil short-term or medium-term power production demand, to delay the required investments and to reduce at the same time repair / replacement costs by careful advance planning of the required maintenance action. AE aims at answering the need for an early and reliable warning of crack growth or structural instability, so that appropriate actions can be timely taken, possibly in a planned way.
A parallel, equally relevant, line of research pursued within the Project was the development of AE sensors based on fibre optic interferometry, as a potentially advantageous alternative to piezoceramic AE sensors in the hostile power plant environment. The main advantages consist in non-invasive broad-band detection of AE signals, intrinsically safe operation and electromagnetic noise immunity. Work was initially addressed to non-contact interferometric sensors and a working prototype was produced and successfully tested in the shop and in the plant. Plant experience, however, showed that a contact, intrinsic interferometric AE sensor could be more performant, more easily adaptable to the constraints imposed by the difficult plant environment, and, last but not least, cheaper and more rugged. Effort was therefore successfully redirected to produce a working prototype of a contact fibre optic interferometric AE sensor, with the capability to withstand temperatures of over 500(C. The use of WDM (Wavelength Division Multiplexing) in order to simultaneously manage 2 or more sensors for AE source location purposes was also successfully demonstrated. The exploitation potential of the fibre optic AE sensor is high, once fully engineered, and is specifically related to its non invasivity (no waveguide welding is required), to its intrinsically safe operation principle (monitoring of structural items in hazardous areas), to its e.m. noise immunity, and to the fact that the above mentioned benefits can be achieved in perspective at a cost comparable to that of systems based on conventional piezoceramic transducers.
The objective of the project is to provide the necessary background knowledge and a systematic approach to the effective use of acoustic emission (AE) techniques in on-line structural integrity monitoring (SIM) of power plant components operated at high temperature (superheater -SH- and reheater -RH- headers, steamlines and boiler membrane panels). For each class of components, three distinct sets of experimental data will be measered : acoustic wave attenuation, plant background noise, and the AE signatures of different damage mechanisms.
Focus is on the possibility to detect and localize pulsed AE sources generated by crack propagation events during plant operation transients (startup, shutdown), and to correlate acoustic activity to the significant generating event. The performance of fibre optic laser interferometer sensors for broadband non-invasive AE signal detection will be compared to that of conventional piezoceramic transducers. Final tests will be carried on operating power plants. Guidelines will be issued concerning the use of waveguides, sensor number/ location/ operation frequency, AE data processing/ presentation/ interpretation.
Funding SchemeCSC - Cost-sharing contracts
PA4 8UW Renfrew