: Downhole tiltmeters and microseismicity are the two main direct hydraulic fracture mapping technologies used in the petroleum industry to determine fracture dimensions. Each technology measures different aspects of the hydraulic fracturing process. Performing a joint inversion using both datasets together, either in cases where both types of data sets are available from different observation wells or from a hybrid array which has both tiltmeters and microseismic sensors in one well, can help improve the fracture geometry and reduce the uncertainties in the measured parameters. A joint inversion was previously performed using a distributional approach, in which the microseismic events are distributed around a model fracture, which in turn is estimated from the tiltmeter data. In this paper, we use a statistical technique from multivariate analysis, called Principal Component Analysis (PCA) to handle the microseismic data. Another improvement over the previous joint inversion method is the inclusion of layered velocities in the relocation of the microseismic events. Assuming the P and S wave?s arrival time picks and the direction of the events are correct, the velocity structure can be refined in the joint inversion process. This joint inversion technique is applied to a field case study in Jonah tight gas field, where both downhole tilt data and microseismic data are available. Results show improved fracture geometry parameters over the ones obtained by downhole tiltmeter data or microseismic data alone.
With the increasing worldwide demand for more energy resources, the unconventional hydrocarbon reservoirs such as tight gas sands, shale gas and coalbed methane, are receiving more and more attention from operators and service companies. Due to the limited permeabilities of these types of reservoirs, hydraulic fracturing or some other stimulation technology is required to achieve economic production. For optimal hydraulic fracture design, it is necessary to have accurate information on rock properties, stresses, permeabilities and other parameters which are not usually available. Fracture diagnostics play an important role in providing key information about the created fractures and their complexity, and this information can be incorporated into a fracture simulator to build calibrated fracture models to be used in the field as a predictive tool. The most commonly used direct fracture diagnostic techniques for obtaining far field fracture dimensions are downhole tiltmeter and microseismic mapping. They have been used to optimize well stimulation in the Barnett shale [1, 2] as well as in tight gas sand field development . Each of these mapping technologies measures different aspects of the fracturing process, so they can independently provide some information about the fracture growth process; combining them together, however, should give more complete information about the fracture. Jointly inverting both tilt data and microseismic data was first performed  using a distributional approach, in which the microseismic events are assumed to be distributed along the model fracture plane. This technique works reasonably well but requires a number of distributional parameters that are not well known, leaving considerable uncertainty in its application. In this paper, we adopt the Principal Component Analysis (PCA) method to interpret the microseismic data.