Search for: All records

Compressional velocity (Vp) and bulk density (ρb) logs are essential for characterizing gas hydrates and near-seafloor sediments; however, it is sometimes difficult to acquire these logs due to poor borehole conditions, safety concerns, or cost-related issues. We present a machine learning approach to predict either compressional Vp or ρb logs with high accuracy and low error in near-seafloor sediments within water-saturated intervals, in intervals where hydrate fills fractures, and intervals where hydrate occupies the primary pore space. We use scientific-quality logging-while-drilling well logs, gamma ray, ρb, Vp, and resistivity to train the machine learning model to predict Vp or ρb logs. Of the six machine learning algorithms tested (multilinear regression, polynomial regression, polynomial regression with ridge regularization, K nearest neighbors, random forest, and multilayer perceptron), we find that the random forest and K nearest neighbors algorithms are best suited to predicting Vp and ρb logs based on coefficients of determination (R2) greater than 70% and mean absolute percentage errors less than 4%. Given the high accuracy and low error results for Vp and ρb prediction in both hydrate and water-saturated sediments, we argue that our model can be applied in most LWD wells to predict Vp or ρb logs in near-seafloor siliciclastic sediments on continental slopes irrespective of the presence or absence of gas hydrate. 

Abstract Near‐vertical hydrate‐filled fractures are found in subseafloor marine muds, at both advective methane vent sites and at sites without obvious methane and fluid advection (non‐vent sites). At non‐vent sites, the mechanisms that transport methane to the fractures and control how hydrate‐filled fractures form are not well understood. However, these mechanisms are important to establish, as most of Earth's natural gas hydrate is likely bound in marine mud systems. Herein, we focus on understanding the origin of hydrate and how fracture form at non‐vent sites by examining previously hydrate‐bearing fractures in conventional cores taken from Keathley Canyon 151, U.S. northern Gulf of Mexico, drilled by the Gas Hydrate Joint Industry Project in 2005. We combine information from well logs, sediment cores, and science party results and add new X‐ray computed tomography of archival sections and scanning electron microcopy of core samples to develop a conceptual model. We propose that locally generated microbial methane is transported via diffusion from small pores in marine mud into biomineralized burrows with larger pore size in a process called short‐range migration. Hydrate forms in burrows once the methane diffuses into them and the dissolved methane concentration exceeds the solubility threshold. When hydrate fills a burrow, heave from additional hydrate growth places stress on the burrow edges, expands the fracture, and creates additional void space in which methane can diffuse and continue forming hydrate. Fractures slowly propagate in the direction of maximum principal stress. 

Abstract The Pāpaku Fault Zone, drilled at International Ocean Discovery Program (IODP) Site U1518, is an active splay fault in the frontal accretionary wedge of the Hikurangi Margin. In logging‐while‐drilling data, the 33‐m‐thick fault zone exhibits mixed modes of deformation associated with a trend of downward decreasing density,P‐wave velocity, and resistivity. Methane hydrate is observed from ~30 to 585 m below seafloor (mbsf), including within and surrounding the fault zone. Hydrate accumulations are vertically discontinuous and occur throughout the entire logged section at low to moderate saturation in silty and sandy centimeter‐thick layers. We argue that the hydrate distribution implies that the methane is not sourced from fluid flow along the fault but instead by local diffusion. This, combined with geophysical observations and geochemical measurements from Site U1518, suggests that the fault is not a focused migration pathway for deeply sourced fluids and that the near‐seafloor Pāpaku Fault Zone has little to no active fluid flow.