An adaptive physics-informed deep learning method for pore pressure prediction using seismic data

Accurate prediction of formation pore pressure is essential to predict fluid flow and manage hydrocarbon production in petroleum engineering.Recent deep learning technique has been receiving more interest due to the great potential to deal with pore pressure prediction.However,most of the traditional deep learning models are less efficient to address generalization problems.To fill this technical gap,in this work,we developed a new adaptive physics-informed deep learning model with high generalization capability to predict pore pressure values directly from seismic data.Specifically,the new model,named CGP-NN,consists of a novel parametric features extraction approach(1DCPP),a stacked multilayer gated recurrent model(multilayer GRU),and an adaptive physics-informed loss function.Through machine training,the developed model can automatically select the optimal physical model to constrain the results for each pore pressure prediction.The CGP-NN model has the best generalization when the physicsrelated metricλ=0.5.A hybrid approach combining Eaton and Bowers methods is also proposed to build machine-learnable labels for solving the problem of few labels.To validate the developed model and methodology,a case study on a complex reservoir in Tarim Basin was further performed to demonstrate the high accuracy on the pore pressure prediction of new wells along with the strong generalization ability.The adaptive physics-informed deep learning approach presented here has potential application in the prediction of pore pressures coupled with multiple genesis mechanisms using seismic data.

A hybrid machine learning optimization algorithm for multivariable pore pressure prediction

Pore pressure is essential data in drilling design,and its accurate prediction is necessary to ensure drilling safety and improve drilling efficiency.Traditional methods for predicting pore pressure are limited when forming particular structures and lithology.In this paper,a machine learning algorithm and effective stress theorem are used to establish the transformation model between rock physical parameters and pore pressure.This study collects data from three wells.Well 1 had 881 data sets for model training,and Wells 2 and 3 had 538 and 464 data sets for model testing.In this paper,support vector machine(SVM),random forest(RF),extreme gradient boosting(XGB),and multilayer perceptron(MLP)are selected as the machine learning algorithms for pore pressure modeling.In addition,this paper uses the grey wolf optimization(GWO)algorithm,particle swarm optimization(PSO)algorithm,sparrow search algorithm(SSA),and bat algorithm(BA)to establish a hybrid machine learning optimization algorithm,and proposes an improved grey wolf optimization(IGWO)algorithm.The IGWO-MLP model obtained the minimum root mean square error(RMSE)by using the 5-fold cross-validation method for the training data.For the pore pressure data in Well 2 and Well 3,the coefficients of determination(R^(2))of SVM,RF,XGB,and MLP are 0.9930 and 0.9446,0.9943 and 0.9472,0.9945 and 0.9488,0.9949 and 0.9574.MLP achieves optimal performance on both training and test data,and the MLP model shows a high degree of generalization.It indicates that the IGWO-MLP is an excellent predictor of pore pressure and can be used to predict pore pressure.

Effects of confining pressure and pore pressure on multipole borehole acoustic field in fluid-saturated porous media

In-situ stress is a common stress in the exploration and development of oil reservoirs. Therefore, it is of great significance to study the propagation characteristics of borehole acoustic waves in fluid-saturated porous media under stress.Based on the acoustoelastic theory of fluid-saturated porous media, the field equation of fluid-saturated porous media under the conditions of confining pressure and pore pressure and the acoustic field formula of multipole source excitation in open hole are given. The influences of pore pressure and confining pressure on guided waves of multipole borehole acoustic field in fluid-saturated porous media are investigated. The numerical results show that the phase velocity and excitation intensity of guided wave increase significantly under the confining pressure. For a given confining pressure, the phase velocity of the guided wave decreases with pore pressure increasing. The excitation intensity of guided wave increases at low frequency and then decreases at high frequency with pore pressure increasing, except for that of Stoneley wave which decreases in the whole frequency range. These results will help us get an insight into the influences of confining pressure and pore pressure on the acoustic field of multipole source in borehole around fluid-saturated porous media.

A machine learning approach for the prediction of pore pressure using well log data of Hikurangi Tuaheni Zone of IODP Expedition 372,New Zealand

Pore pressure(PP)information plays an important role in analysing the geomechanical properties of the reservoir and hydrocarbon field development.PP prediction is an essential requirement to ensure safe drilling operations and it is a fundamental input for well design,and mud weight estimation for wellbore stability.However,the pore pressure trend prediction in complex geological provinces is challenging particularly at oceanic slope setting,where sedimentation rate is relatively high and PP can be driven by various complex geo-processes.To overcome these difficulties,an advanced machine learning(ML)tool is implemented in combination with empirical methods.The empirical method for PP prediction is comprised of data pre-processing and model establishment stage.Eaton's method and Porosity method have been used for PP calculation of the well U1517A located at Tuaheni Landslide Complex of Hikurangi Subduction zone of IODP expedition 372.Gamma-ray,sonic travel time,bulk density and sonic derived porosity are extracted from well log data for the theoretical framework construction.The normal compaction trend(NCT)curve analysis is used to check the optimum fitting of the low permeable zone data.The statistical analysis is done using the histogram analysis and Pearson correlation coefficient matrix with PP data series to identify potential input combinations for ML-based predictive model development.The dataset is prepared and divided into two parts:Training and Testing.The PP data and well log of borehole U1517A is pre-processed to scale in between[-1,+1]to fit into the input range of the non-linear activation/transfer function of the decision tree regression model.The Decision Tree Regression(DTR)algorithm is built and compared to the model performance to predict the PP and identify the overpressure zone in Hikurangi Tuaheni Zone of IODP Expedition 372.

Research on Multi-Wave Pore Pressure Prediction Method Based on Three Field Velocity Fusion

The optimization of velocity field is the core issue in reservoir seismic pressure prediction. For a long time, the seismic processing velocity analysis method has been used in the establishment of pressure prediction velocity field, which has a long research period and low resolution and restricts the accuracy of seismic pressure prediction;This paper proposed for the first time the use of machine learning algorithms, based on the feasibility analysis of wellbore logging pressure prediction, to integrate the CVI velocity inversion field, velocity sensitive post stack attribute field, and AVO P-wave and S-wave velocity reflectivity to obtain high-precision seismic P and S wave velocities. On this basis, high-resolution formation pore pressure and other parameters prediction based on multi waves is carried out. The pressure prediction accuracy is improved by more than 50% compared to the P-wave resolution of pore pressure prediction using only root mean square velocity. Practice has proven that the research method has certain reference significance for reservoir pore pressure prediction.

Rock physics model for velocity–pressure relations and its application to shale pore pressure estimation

Acoustic wave velocity has been commonly utilized to predict subsurface geopressure using empirical relations.Acoustic wave velocity is, however, affected by many factors. To estimate pore pressure accurately, we here propose to use elastic rock physics models to understand and analyze quantitatively the various contributions from these different factors affecting wave velocity. We report a closed-form relationship between the frame flexibility factor(γ) in a rock physics model and differential pressure, which presents the major control of pressure on elastic properties such as bulk modulus and compressional wave velocity. For a gas-bearing shale with abundant micro-cracks and fractures, its bulk modulus is much lower at abnormally high pore pressure(high γ values) where thin cracks and flat pores are open than that at normal hydrostatic pressure(low γ values) where pores are more rounded on average. The developed relations between bulk modulus and differential pressure have been successfully applied to the Upper Ordovician Wufeng and Lower Silurian Longmaxi formations in the Dingshan area of the Sichuan Basin to map the three-dimensional spatial distribution of pore pressure in the shale, integrating core, log and seismic data. The estimated results agree well with field measurements. Pressure coefficient is positively correlated to gas content. The relations and methods reported here could be useful for hydrocarbon exploration, production, and drilling safety in both unconventional and conventional fields.

Influence of Hydrodynamic Pore Pressure Damage on the Performance of Hot-Mixed Renewable Asphalt Mixture

For evaluating the water stability of hot-mixed renewable asphalt mixture(HRM),the traditional methods are all tested under still water conditions.Except for damage in still water conditions,the hydrodynamic pore pressure generated by the tire driving on the surface water has a great impact.Thus,the RAP contents of the HRMs were designed at 0%,30%,45%and 60%with AC-25 gradation.Then,the self-designed evaluation methods of water stability and dynamic modulus were studied.Finally,the mechanism of the influence of hydrodynamic pore pressure damage on HRMs was studied.The results show that the water stability of HRM containing 30%RAP is equivalent to that of 45%RAP,and the water stability of HRM containing 60%RAP decreases significantly.The Contabro test after MIST treatment can be used as an evaluation method for hydrodynamic pore pressure damage on HRM.Low-speed,heavy-load traffic and larger RAP content have greater damage to the mixture after hydrodynamic pore pressure damage.The performance differences between the aged bitumen and pure bitumen,as well as the aged minerals and new minerals,are continuing to be enlarged in hydrodynamic pore pressure conditions,finally affecting the water stability and dynamic modulus of the HRMs.

Pore pressure prediction in offshore Niger delta using data-driven approach: Implications on drilling and reservoir quality

Despite exploration and production success in Niger Delta,several failed wells have been encountered due to overpressures.Hence,it is very essential to understand the spatial distribution of pore pressure and the generating mechanisms in order to mitigate the pitfalls that might arise during drilling.This research provides estimates of pore pressure along three offshore wells using the Eaton's transit time method,multi-layer perceptron artificial neural network(MLP-ANN)and random forest regression(RFR)algorithms.Our results show that there are three pressure magnitude regimes:normal pressure zone(hydrostatic pressure),transition pressure zone(slightly above hydrostatic pressure),and over pressured zone(significantly above hydrostatic pressure).The top of the geopressured zone(2873 mbRT or 9425.853 ft)averagely marks the onset of overpressurization with the excess pore pressure above hydrostatic pressure(P∗)varying averagely along the three wells between 1.06−24.75 MPa.The results from the three methods are self-consistent with strong correlation between the Eaton's method and the two machine learning models.The models have high accuracy of about>97%,low mean absolute percentage error(MAPE<3%)and coefficient of determination(R2>0.98).Our results have also shown that the principal generating mechanisms responsible for high pore pressure in the offshore Niger Delta are disequilibrium compaction,unloading(fluid expansion)and shale diagenesis.

Numerical simulation of dynamic pore pressure in asphalt pavement

For studying the driving role of dynamic pressure in water-induced damage of asphalt pavement, based on the fast Lagrangian finite difference method and Biot dynamic consolidation theory, fluid-solid coupling analysis of the pavement is conducted considering asphalt mixtures as porous media. Results reveal that the development and dissipation of the dynamic pore pressure are coinstantaneous and this makes both the positive and negative dynamic pore pressure and seepage force alternate with time. Repetitive hydrodynamic pumping and sucking during moisture damage is proved. The dynamic pore pressure increases with vehicle velocity. Effective stress and deflection of pavement decrease due to the dynamic pore water pressure. However, the emulsification and replacement of the asphalt membrane by water are accelerated. The maximum dynamic pore pressure occurs at the bottom of the surface course. So it is suggested that a drain course should be set up to change the draining condition from single-sided drain to a two-sided drain, and thus moisture damage can be effectively limited.

An improved apparent permeability model considering full pore pressure range,variable intrinsic permeability and slippage coefficient

Although the slippage effect has been extensively studied,most of the previous studies focused on the impact of the slippage effect on apparent permeability within a low pore pressure range,resulting in the inability of matching the evolution of permeability in the remaining pressure range.In this paper,a new apparent permeability model that reveals the evolution of permeability under the combined action of effective stress and slippage in the full pore pressure range was proposed.In this model,both intrinsic permeability and slippage coefficient are stress dependent.Three experimental tests with pore pressure lower than 2 MPa and a test with pore pressure at about 10 MPa using cores from the same origin under constant confining stress and constant effective stress are conducted.By comparing experimental data and another apparent permeability model,we proved the fidelity of our newly developed model.Furthermore,the contribution factor of the slippage effect Rslip is used to determine the low pore pressure limit with significant slippage effect.Our results show that both narrow initial pore size and high effective stress increase the critical pore pressure.Finally,the evolutions of the slippage coefficient and the intrinsic permeability under different boundary conditions were analyzed.

Machine learning and data-driven prediction of pore pressure from geophysical logs:A case study for the Mangahewa gas field,New Zealand

Pore pressure is an essential parameter for establishing reservoir conditions,geological interpretation and drilling programs.Pore pressure prediction depends on information from various geophysical logs,seismic,and direct down-hole pressure measurements.However,a level of uncertainty accompanies the prediction of pore pressure because insufficient information is usually recorded in many wells.Applying machine learning(ML)algorithms can decrease the level of uncertainty of pore pressure prediction uncertainty in cases where available information is limited.In this research,several ML techniques are applied to predict pore pressure through the over-pressured Eocene reservoir section penetrated by four wells in the Mangahewa gas field,New Zealand.Their predictions substantially outperform,in terms of prediction performance,those generated using a multiple linear regression(MLR)model.The geophysical logs used as input variables are sonic,temperature and density logs,and some direct pore pressure measurements were available at the reservoir level to calibrate the predictions.A total of 25,935 data records involving six well-log input variables were evaluated across the four wells.All ML methods achieved credible levels of pore pressure prediction performance.The most accurate models for predicting pore pressure in individual wells on a supervised basis are decision tree(DT),adaboost(ADA),random forest(RF)and transparent open box(TOB).The DT achieved root mean square error(RMSE)ranging from 0.25 psi to 14.71 psi for the four wells.The trained models were less accurate when deployed on a semi-supervised basis to predict pore pressure in the other wellbores.For two wells(Mangahewa-03 and Mangahewa-06),semi-supervised prediction achieved acceptable prediction performance of RMSE of 130—140 psi;while for the other wells,semi-supervised prediction performance was reduced to RMSE>300 psi.The results suggest that these models can be used to predict pore pressure in nearby locations,i.e.similar geology at corresponding depths within a field,but they become less reliable as the step-out distance increases and geological conditions change significantly.In comparison to other approaches to predict pore pressures,this study has identified that application of several ML algorithms involving a large number of data records can lead to more accurate prediction results.

Soil pressure and pore pressure for seismic design of tunnels revisited: considering water-saturated, poroelastic half-space

This paper describes a systematic study on the fundamental features of seismic soil pressure on underground tunnels, in terms of its magnitude and distribution, and further identifi es the dominant factors that signifi cantly infl uence the seismic soil pressure. A tunnel embedded in water-saturated poroelastic half-space is considered, with a large variety of model and excitation parameters. The primary features of both the total soil pressure and the pore pressure are investigated. Taking a circular tunnel as an example, the results are presented using a fi nite element-indirect boundary element(FE-IBE) method, which can account for dynamic soil-tunnel interaction and solid frame-pore water coupling. The effects of tunnel stiffness, tunnel buried depth and input motions on the seismic soil pressure and pore pressure are also examined. It is shown that the most crucial factors that dominate the magnitude and distribution of the soil pressure are the tunnel stiffness and dynamic soil-tunnel interaction. Moreover, the solid frame-pore water coupling has a prominent infl uence on the magnitude of the pore pressure. The fi ndings are benefi cial to obtain insight into the seismic soil pressure on underground tunnels, thus facilitating more accurate estimation of the seismic soil pressure.

Anisotropic Rock Poroelasticity Evolution in Ultra-low Permeability Sandstones under Pore Pressure,Confining Pressure,and Temperature:Experiments with Biot's Coefficient

This study aimed to show anisotropic poroelasticity evolution in ultra-low permeability reservoirs under pore pressure,confining pressure,and temperature.Several groups of experiments examining Biot's coefficient under different conditions were carried out.Results showed that Biot's coefficient decreased with increased pore pressure,and the variation trend is linear,but the decreasing rate is variable between materials.Biot's coefficient increased with increased confining pressure;the variation trend is linear,but the increasing rate varies by material as well.Generally,Biot's coefficient remains stable with increased temperature.Lithology,clay mineral content,particle arrangement,and pore arrangement showed impacts on Biot's coefficient.For strong hydrophilic clay minerals,expansion in water could result in a strong surface adsorption reaction,which could result in an increased fluid bulk modulus and higher Biot's coefficient.For skeleton minerals with strong lipophilicity,such as quartz and feldspar,increased oil saturation will also result in an adsorption reaction,leading to increased fluid bulk modulus and a higher Biot's coefficient.The study's conclusions provide evidence of poroelasticity evolution of ultra-low permeability and help the enhancing oil recovery(EOR)process.

Tight gas production model considering TPG as a function of pore pressure,permeability and water saturation

Threshold pressure gradient has great importance in efficient tight gas field development as well as for research and laboratory experiments.This experimental study is carried out to investigate the threshold pressure gradient in detail.Experiments are carried out with and without back pressure so that the effect of pore pressure on threshold pressure gradient may be observed.The trend of increasing or decreasing the threshold pressure gradient is totally opposite in the cases of considering and not considering the pore pressure.The results demonstrate that the pore pressure of tight gas reservoirs has great influence on threshold pressure gradient.The effects of other parameters like permeability and water saturation,in the presence of pore pressure,on threshold pressure gradient are also examined which show that the threshold pressure gradient increases with either a decrease in permeability or an increase in water saturation.Two new correlations of threshold pressure gradient on the basis of pore pressure and permeability,and pore pressure and water saturation,are also introduced.Based on these equations,new models for tight gas production are proposed.The gas slip correction factor is also considered during derivation of this proposed tight gas production models.Inflow performance relationship curves based on these proposed models show that production rates and absolute open flow potential are always be overestimated while ignoring the threshold pressure gradients.

Experimental investigation on the wave-induced pore pressure around shallowly embedded pipelines

A series of regular wave experiments have been done in a large-scale wave flume to investigate the wave-induced pore pressure around the submarine shallowly embedded pipelines. The model pipelines are buried in three kinds of soils, including gravel, sand and silt with different burial depth. The input waves change with height and period. The results show that the amplitudes of wave-induced pore pressure increase as the wave period increase, and decay from the surface to the bottom of seabed. Higher pore pressures are recorded at the pipeline top and the lower pore pressures at the bottom, especially in the sand seabed. The normalized pressure around pipeline decreases as the relative water depth, burial depth or scattering parameters increase. For the silt seabed, the wavelet transform has been successfully used to analyze the signals of wave - induced pore pressure, and the oscillatory and residual pore pressure can be extracted by wavelet analysis. Higher oscillatory pressures are recorded at the bottom and the lower pressures at the top of the pipeline. However, higher residual pressures are recorded at the top and the lower pressures at the bottom of the pipeline.

Influential factors of loess liquefaction and pore pressure development

The liquefaction of loess under dynamic loading is studied experimentally with a dynamic triaxial test system. The effects of over-consolidation ratio (OCR), saturation degree and the frequency of dynamic loading upon loess liquefaction are investigated. The development of pore pressure within loess samples is also discussed. Based on the experimental results, the empirical relationship between pore pressure ratio and loading cycle number ratio is established for normal consolidated saturated loess.

Pore pressure fluctuations of overlying aquifer during residual coal mining and water-soil stress coupling analysis

Three test models and a simulation model were constructed based on the prevailing conditions of the Taiping coalmine in order to analyze pore pressure fluctuations of an overlying aquifer during residual coal mining. As well, the relation between pore pressure and soil stress was evaluated. The model tests show the vibrations of pore pressure and soil stress as a result of mining activities. The simulation model tells of the response characteristics of pore pressure after mining and its distribution in the sand aquifer. The comparative analysis reveals that pore pressure and soil stress vibration are activated by unexpected events occurring in mines, such as collapsing roofs. An increased pore pressure zone always lies above the wall in front or behind the working face of a mine. Both pore pressure and vertical stress result in increasing and decreasing processes during movements of the working face of a mine. The vibration of pore pressure always precedes soil stress in the same area and ends with a sharp decline. Changes in pore pressure of sand aquifer are limited to the area of stress changes. Obvious changes are largely located in a very small frame over the mining face.

Analysis of wellbore stability by pore pressure prediction using seismic velocity

Borehole problems result in the wastage of huge expenditures in petroleum industry.One of the major challenges is to keep the borehole intact at a particular location and prevent it from falling in reactive shale in high-or low-pressure zones.To convert seismic velocity to pore pressure,the modified Eaton's equation is widely used.A guideline was prepared using both stochastic and non-uniform parameter distributions.Pore pressure and fracture pressure were calculated using a modified version of Eaton's and Mathew and Kelly's methods.To validate the results,geomechanical models were developed based on good correlation.Different regions in the eastern state of India at Tripura were taken into consideration and pore pressure was estimated for the regions of Kathalchari and Ambassa by plotting pressures.The actual pore pressure predicted in Tripura was calculated using seis P and seis P3,the indigenous software which is a modified version of Eaton's method and Mathew and Kelly's method,and the pore pressure from our method matches the calculated Repeat Formation Test(RFT)data from wells.

Effect of pore pressure on deformation and unstable snap-back for shear band and elastic rock system

Fast Lagrangian analysis of continua(FLAC) was used to study the influence of pore pressure on the mechanical behavior of rock specimen in plane strain direct shear, the distribution of yielded elements, the distribution of displacement and velocity across shear band as well as the snap-back (elastic rebound) instability. The effective stress law was used to represent the weakening of rock containing pore fluid under pressure. Numerical results show that rock specimen becomes soft (lower strength and hardening modulus) as pore pressure increases, leading to higher displacement skip across shear band. Higher pore pressure results in larger area of plastic zone, higher concentration of shear strain, more apparent precursor to snap-back (unstable failure) and slower snap-back. For higher pore pressure, the formation of shear band-elastic body system and the snap-back are earlier; the distance of snap-back decreases; the capacity of snap-back decreases, leading to lower elastic strain energy liberated beyond the instability and lower earthquake or rockburst magnitude. In the process of snap-back, the velocity skip across shear band is lower for rock specimen at higher pore pressure, showing the slower velocity of snap-back.

In Situ Observation of Silt Seabed Pore Pressure Response to Waves in the Subaqueous Yellow River Delta

The in situ pore pressure response of silt under wave action is a complex process.However,this process has not been well studied because of limited field observation techniques.The dynamic response process is closely related to engineering geological hazards;thus,this process must be urgently explored.A long-term in situ observational study of the silt sediment pore water pressure response process under wave action was conducted in the subaqueous Yellow River Delta.The response characteristics of pore water pressure are affected by tidal level and wave height.Tidal level affects the overall trend of the pore water pressure response,while wave height influences the amplitude of the pore water pressure response.This study revealed a significant lag effect in the pore pressure response.The transient pore pressure in the seabed did not respond immediately to the wave-induced pressure stress on the seabed surface.This phenomenon may be attributed to the change in soil permeability.The maximum response depth was approximately 0.5 m with a 2 m wave height.A concept model of silt soil pore pressure response under different types of wave action was developed.The accumulation rate of the pore pressure is less than the dissipation rate;thus,the developed model highlights the oscillation pore pres-sure response mechanism.The highlighted response process is of considerable importance to transient liquefaction and the startup process of pore pressure response.