Dr. Graeme West

Projects

Here are details on some of my current projects.

Fuel Grab Load Trace Analytics and Bore Estimation

When an Advanced Gas-cooled Reactor (AGR) station is refuelled, data called the Fuel Grab Load Trace (FGLT) is recorded. This represents the weight of the fuel assembly as it travels through the reactor core and is used as an input into the protection system which shuts down the reactor should an abnormal load be detected during refuelling. It has been subsequently discovered that this FGLT can provide additional information relating to the bore diameter, and hence condition, of the fuel channel. Manual interpretation of this FGLT can provide indications of the fuel channel condition. Ongoing research is supporting this manual analysis through the application of artificial intelligence techniques, to provide automated decision support to the graphite core engineering team. Initial stages of the work resulted in the development of the BETA: British Energy Trace Analysis system, which was developed from research prototype to industrial system deployed at EDF Energy (British Energy at the time) offering support to their engineers for over 6 years. Quintessa now provide support for this function through their LoTAS software, based on the BETA system.

Further details of some of the research aspects of this project can be found here and here.

This work is funded by EDF Energy through the Advanced Nuclear Research Centre

Advanced Image Processing Techniques for In-Core Inspection Videos

During routine inspection of reactor cores video footage is recorded of the inner surface of selected fuel channels. We have developed the ASIST software system which produces a single 360 degree image of the inside surface of the fuel channel from video footage taken at multiple orientations. The resulting image is known as a channel panorama, or Chanorama. Our software significantly improves the inspection process by reducing the time taken to generate montages of any regions of interest, an activity which lies on the critical outage path and providing full coverage of the fuel channel which aids improved location and quantification of any regions of interest.

Further details of some of the research aspects of this project can be found here and here.

This work is funded by EDF Energy through the Advanced Nuclear Research Centre

Automated Sizing and Classification of Pressure Tube Defects

During planned statutory outages at Bruce Power's reactors, a selected number of fuel channels are inspected using a tool with multiple ultrasonic probes. The gathered ultrasonic data is assessed for the presence of defects and any identified defects and manually sized and classified. We have developed a prototype system, ADAPT (Automated Data Analysis of Pressure Tubes), which aids in the sizing and classification of data of pressure tube defects using codified knowledge elicited from analysts. The ADAPT system aims to provide a robust, repeatable and reliable method for analysing pressure tube inspection data, providing time saving on the critical outage path and freeing analysts time to focus on the difficult, hard to quantify defects rather than the routine, well understood ones.

Further details of some of the research aspects of this project can be found here and here.

This work is funded by Bruce Power through the Advanced Nuclear Research Centre

MAPS:Mosaicing for Automatic Pipe Scanning

The interior visual inspection of pipework is a critical inspection activity required to ensure the continued safe, reliable operation of plant and thus avoid costly outages. Typically, the video output from a manually deployed probe is viewed by an operator with the task of identifying and estimating the location of surface defects such as cracks, corrosion and pitting. However, it is very difficult to estimate the nature and spatial extent of defects from the often disorientating small field of view video of a relatively large structure. This project has developed a new visual inspection system designed for inspecting 3 - 6 inch diameter pipes. The system uses a high resolution camera and structure from motion (SFM) algorithm to compute the trajectory of the probe through the pipe. In addition a laser profiler is used to measure the inner surface of the pipe and generate a meshed point cloud. The camera images are projected onto the mesh and the final output of the system is a photorealistic 3-D model of the internal surface of the pipework.

Further details of some of the research aspects of this project can be found here and here.

This project is funded through Innovate UK and is a five member consortia of NNL, Inspectahire, WideBlue, Sellafield and the University of Strathclyde.