Reconnaissance datasets for prospective wind-farm developments generally comprise ultra-high resolution 2D seismics, combined with a number of geotechnical sampling locations, spread over large areas. The high volume of data, coupled with strictly defined work-scope requirements, limited time-frames, tight budgets and often limited resources, imposes severe constraints on interpretation geoscientists. This is only exacerbated by highly complex geological sequences. In attempting to accommodate these occasionally conflicting constraints, interpretations can acquire an unacceptable degree of subjectivity. Two example workflows are here examined which, while accommodating key constraints, remove as far as possible overtly subjective interpretation practices and consequent uncertainties and inaccuracies in gridded and exported model data. Two case-studies are presented in which geotechnical data are transformed into the petro-physical domain and used to calibrate quantitative acoustic impedance inversion of seismic amplitude data. Subsequent export of multi-parameter database products represents construction and delivery of ‘useful models’ of reconnaissance survey datasets. These models, while acknowledging the uncertainties inherent in their construction, may provide the bases of future iterations within the project life-cycle.
Francis Buckley (Thu,) studied this question.