Estimation of fracture height growth in layered tight/shale gas reservoirs using flowback gas rates and compositions–Part II: Field application in a liquid-rich tight reservoir

Abstract

While hydraulic fracturing is the key to unlocking the potential of unconventional low-permeability hydrocarbon resources, challenges remain in the monitoring of subsurface propagation of fractures and the determination of which geologic intervals have been contacted. This is particularly challenging for wells that are completed in multiple hydraulic fracture stages (multi-fractured horizontal wells or MFHWs) where fracture spacing may be very close and fracture geometry complex. Understanding the fracture extent is important not only for assisting with hydraulic fracture design, but also for mitigating unwanted fracture growth into non-target geologic intervals that do not contain hydrocarbons (e.g. zones with high water saturation). Popular current technologies used for hydraulic fracture surveillance include microseismic (surface and subsurface monitoring) and tiltmeter surveys. While these methods have proven useful for characterizing the extent of created hydraulic fractures, they do not necessarily lead to an understanding of what portions of the geologic section (bounding and target intervals for MFHWs, for example) are in direct hydraulic communication with the well.A solution for establishing the extent of hydraulic fracture growth from target to bounding zones is to first obtain a fluid composition fingerprint of those intervals while drilling through them, and then compare these data with fluid compositions obtained from flowback after hydraulic fracturing. In the current work, a MFHW completed in a liquid-rich tight reservoir is used to test this novel concept. Gas samples extracted from the headspace of isojars® containing cuttings samples, obtained during drilling of the MFHW well, were used to geochemically fingerprint geologic intervals through which the well was drilled. The cuttings samples were collected at high frequency in the vertical, bend and lateral sections of the well over a measured depth range of 4725 ft (1440 m). A compositional marker was identified in the bend of the horizontal well above which the average methane to ethane (C1/C2) ratio was 15.7, versus 2.6 below it. The flowback gas compositions were observed to be intermediate (average C1/C2 = 7.4) between the reservoir above and below the marker, suggesting fracture height grew above the compositional marker.In order to estimate fracture height growth from the geologic interval and flowback compositions, a compositional numerical simulation study was performed. An innovative approach was used to estimate recombined in-situ fluid compositions, on a layer-by-layer basis, by combining the cuttings gas compositional data with separator oil compositions. The resulting numerical simulation model, initialized through use of the layered fluid model and a detailed geological model developed for the subject well and offset drilling locations, was used to history match flowback rates, pressures and gas compositions. The gas compositions of the fingerprinted geologic intervals were therefore employed as a constraint on fracture height growth, estimated in the model to be 175 ft (53 m, propped height). However, because of the uncertainty in model input parameters, a stochastic approach was required to derive a range in hydraulic fracture properties.The current study demonstrates for the first time that it is possible to constrain fracture height growth estimates from flowback data, combined with gas compositional data obtained from cuttings data, provided that the geochemical fingerprints are distinct.