Dynamic Graph Reasoning for Conversational Open-Domain Question Answering

In recent years, conversational agents have provided a natural and convenient access to useful information in people’s daily life, along with a broad and new research topic, conversational question answering (QA). On the shoulders of conversational QA, we study the conversational open-domain QA problem, where users’ information needs are presented in a conversation and exact answers are required to extract from the Web. Despite its significance and value, building an effective conversational open-domain QA system is non-trivial due to the following challenges: (1) precisely understand conversational questions based on the conversation context; (2) extract exact answers by capturing the answer dependency and transition flow in a conversation; and (3) deeply integrate question understanding and answer extraction. To address the aforementioned issues, we propose an end-to-end Dynamic Graph Reasoning approach to Conversational open-domain QA (DGRCoQA for short). DGRCoQA comprises three components, i.e., a dynamic question interpreter (DQI), a graph reasoning enhanced retriever (GRR), and a typical Reader, where the first one is developed to understand and formulate conversational questions while the other two are responsible to extract an exact answer from the Web. In particular, DQI understands conversational questions by utilizing the QA context, sourcing from predicted answers returned by the Reader, to dynamically attend to the most relevant information in the conversation context. Afterwards, GRR attempts to capture the answer flow and select the most possible passage that contains the answer by reasoning answer paths over a dynamically constructed context graph . Finally, the Reader, a reading comprehension model, predicts a text span from the selected passage as the answer. DGRCoQA demonstrates its strength in the extensive experiments conducted on a benchmark dataset. It significantly outperforms the existing methods and achieves the state-of-the-art performance.

A Single Equation to Depict Bottomhole Pressure Behavior for a Uniform Flux Hydraulic Fractured Well

Pressure transient analysis for a vertically hydraulically fractured well is evaluated using two different equations, which cater for linear flow at the early stage and radial flow in the later stage. However, there are three different stages that take place for an analysis of pressure transient, namely linear, transition and pseudo-radial flow. The transition flow regime is usually studied by numerical, inclusive methods or approximated analytically, for which no specific equation has been built, using the linear and radial equations. Neither of the approaches are fully analytical. The numerical, inclusive approach results in separate calculations for the different flow regimes because the equation cannot cater for all of the regimes, while the analytical approach results in a difficult inversion process to compute well test-derived properties such as permeability. There are two types of flow patterns in the fracture, which are uniform and non-uniform, called infinite conductivity in a high conductivity fracture. The study was conducted by utilizing an analogous study of linear flow equations. Instead of using the conventional error function, the exponential integral with an infinite number of wells was used. The results obtained from the developed analytical solution matched the numerical results, which proved that the equation was representative of the case. In conclusion, the generated analytical equation can be directly used as a substitute for current methods of analyzing uniform flow in a hydraulically fractured well.

Flow pattern identification of liquid-liquid (oil and water) in vertical pipelines using machine learning techniques

Given the importance of process control in the petrochemical industry, there is a need to determine the behavior of the fluids inside the pipes. In this work a methodology is developed for the identification of flow patterns in vertical pipes with diameters between 0.01 m and 0.10 m, from the implementation of artificial intelligence techniques, for a liquid combination of two phases composed of oil with viscosity in the range of 792 Kg/m3 to 1823 Kg/m3 and water at room temperature. The predictive models generated in the structuring of the methodology were trained with 70% of data based on viscosity parameters, pipe diameter, volume fraction and surface velocities of the working fluids stored in a database. The remaining information, equivalent to 30% of the total, was used to develop the automatic model validation. The flow patterns identified by the intelligent system for oil and water flow, without considering the predominant substance, are churning, dispersed, very fine dispersion, transition flow, intermittent, and annular

Experimental Study of Gas-Liquid Pressurization Performance and Critical Gas Volume Fractions of a Multiphase Pump

Gas entrainment may cause pressurization deterioration and even failure of pumps under conditions of high inlet gas volume fraction (GVF). When the inlet GVF increases to a critical value, an obvious deterioration performance of pump occurs. Air-water pressurization performance and inlet critical GVFs of a centrifugal multiphase pump are investigated experimentally under different inlet pressures and gas-liquid flow rates. To determine the first and second critical GVFs, a new method is proposed by computing the local extreme points of the second derivative of performance curves. New prediction correlations for two critical GVFs are established with relative errors lower than ±10% and ±8%. Boundaries of three different flow patterns and the transition flow rates are determined and presented by critical GVFs on the flow pattern diagram. Moreover, boundaries of maximum pressurization are determined by performance curve clusters and a power function correlation of gas-liquid flow rates when reaching the maximum pressurization is established. With the increase of inlet pressure from 1MPa to 5MPa, two-phase pressurization performance is significantly increased; occurrences of pressurization deterioration are obviously delayed with the first and second critical GVFs increasing by maximums of 8.2% and 7.1%.

Rate/Pressure Transient Analysis of a Variable Bottom Hole Pressure Multi-Well Horizontal Pad with Well Interference

Multi-well horizontal pads are common in unconventional reservoirs. With addition of infill wells and hydraulic fracturing, interference between multiple multi-fractured-horizontal wells (MFHWs) has become a serious issue. Current RTA workflows assume a single MFHW in the unconventional formation. This paper presents a new multi-MFHW solution and related analysis methodology to analyze targeted well rate performance in a multi-MFHW system. In this work, a semi-analytical equation describing multi-well pad in the Laplace domain with well interference is proposed. The proposed semi-analytical model can simulate the rate performance of a multi-well horizontal pad with variable BHP for a targeted well in the pad and different initial production durations for the offset well. From the constant BHP condition and Laplace transforms, we obtained multi-MFHW solutions for transient flow. We used superposition of various constant BHP solutions to study interference among various fractures and MFHWs. The variable BHP of the targeted well is achieved by a variable dimensionless BHP function in the Laplace domain without any convolution or deconvolution calculations. A systematic validation for the proposed method is conducted using a commercial numerical simulator for cases of different initial production times for offset MFHWs, multi-MFHWs with variable BHP. Through the total material balance of the multi-MFHW system, we can analyze a target well in the pad with this multi-MFHW analysis. Interference by offset wells often appears after pseudo-radial flow in the target well's hydraulic fracture. It causes the pressure derivative curve during elliptical and infinite-acting radial flow (IARF) to rise, as does the RNP derivative. The inverse semi-log derivative has the opposite trend. Well interference also makes the rate/pressure drop functions to deviate from initial straight lines in later stages. Sensitivity analysis of well spacing shows that "transition flow" will change from elliptical to formation linear flow between wells as well spacing increases and it can show the transitional flow characteristics in more common cases.

Rill Erosion on Slope of Spoil tips: experimental study of runoff scouring erosion in multiple times

The soil erosion of the spoil tips seriously threatens the safety of people's lives and property and the surrounding ecological environment. Rill erosion is an important cause of water and soil loss in spoil tips. This study was conducted to investigate the process of rill erosion on the slopes of spoil tips, changes in the morphological characteristics of rills and the mechanisms of rill erosion. A Field runoff plot (5 m long, 1 m wide and 0.5 m deep) with three inflow rates (1.6, 2 and 2.4 mm min−1) and three typical slopes (28°, 32° and 36°) was used for runoff simulation experiments. The results showed that, compared with the slope and scouring times, inflow rate was the most important factor affecting rill erosion of the spoil tips. The development of rill mainly goes through three stages: the rill formation stage, the rill development stage and the rill adjustment stage. The overall predominance of parallel-shaped rills at all experiments suggested that the formation of rills was dominated by concentrated runoff. The average rill depth was the best indicator of rill morphology for evaluating rill erosion. The flow regimes under the experimental conditions were supercritical-laminar flow and supercritical-transition flow. The Reynolds number was the best hydraulic parameter for predicting rill erosion. The stream power was the best hydrodynamic parameter to describe rill erosion mechanism. These results contributed to further revealing the rill erosion mechanism on the slope of the spoil tips and provided a scientific basis for its soil erosion control.