Development of mixed field radiation detection techniques for oil and gas well logging.

点击翻译按钮获取中文摘要

The NuSec-mediated workshop in January 2018 produced a report in which the need to develop alternative detection and data interpretation methods in well logging was highlighted.The detector development group at the University of Sheffield has a long track record of detector development for a range of particles including muons, neutrons and gammas. Our recent STFC-supported work on the development of neutron-sensitive detectors for water content monitoring has resulted in new expertise in the fabrication of neutron- sensitive foils being developed in house. This new capability sits alongside previous expertise in prototyping borehole detectors for CCS monitoring. Furthermore, due to a long-standing relationship with LabLogic we have access to a scintillator fabrication facility enabling short turnaround prototyping and development cycles. Along with our range of neutron sources (pulsed DT, AmBe, 252Cf) this places us in a unique position to be able to develop low-cost plastic scintillator and neutron-sensitive foil based detection systems for application in borehole configurations.Via the studentship, we intend to explore the possibility of using scintillator blocks doped with different materials to produce finely segmented radiation detectors that can fit down a standard borehole. There are several advantages to this approach over standard methods, e.g.:- use of a single detector that covers mixed fields using position-sensitive detection/ coincidence. On-site expertise in many different detector types is thus not required;- deployment of a finely segmented detector made of many similar stacked "modules"or "blocks" means the system has built in redundancy;- the time from initial neutron generator pulse to detection of neutrons/gammas canprovide additional fine grained information not available when using radiological sources.A further issue with current well logging technologies, as noted in papers, is the analysis of data. Gamma logs, e.g., are used as almost a qualitative measure. Similarly, neutron porosity measurements require further calibration to be effective. The inclusion of pulsed generators and associated particle detection yields supplementary information, however, this is at the cost of additional processing time and training. A solution for this is to use machine learning to extract quantities of interest, e.g. the total detected neutron/gamma counts from a borehole detector observed in a gated window after the initial generator pulse would serve as suitable input to a multivariate analysis (MVA) method such as a neural network. Such an MVA could be trained, via simulations, to extract relevant quantities of interest. These simulations could then be experimentally validated using a scenario in which a detector is placed in parallel with the Sheffield DT source and surrounded by volume of material (e.g. soil and/or rock) of known properties.Whilst the borehole detector is the primary hardware focus of this application it is acknowledged that the fusion of plastic scintillator with neutron-sensitive foils may afford opportunities in other application areas such as environmental monitoring and the nuclear industry, where alternative format detectors employing the same underlying technology may be of interest.The project output is therefore expected to comprise an operational borehole detector prototype with empirical comparisons with "standard" (e.g. 3He) methods as well as MVA-based software tools for signal discrimination.