Abstract
Among the many data sets provided by fiber optic deployments in offset wells, low frequency DAS measurements provide detailed information on the wellbore axial strain during nearby hydraulic fracture treatments. Analysis of strain during a treatment typically consists of magnitude-based detection of fracture hits for location and event time. More detailed post-treatment analysis can tease out at least a few key components of far-field fracture solutions that match the pattern and evolution of measured strain, but the highly non-linear nature of the problem impedes real time solutions. This paper introduces a method that linearizes the problem, then applies a Kalman filter based inversion to determine the proximity of fluid to the offset well as it approaches the fiber during the treatment. The computationally efficient approach enables quick response during a treatment to unexpected fracture growth. The use of a Kalman filter also allows for the incorporation of other data types, such as microseismic locations, into an integrated inversion. Theoretical and field examples demonstrate the technique in action.
