Rotary Steerable System (RSS) is an advanced directional drilling system and research focus in the field of oil and gas exploration and development. In the RSS drilling process, accurately predicting well path changes and dynamically adjusting the steering force, tool face angle and other parameters are the keys to achieving accurate steering and efficient drilling. However, traditional mechanism models for bottom-hole assemblies struggle to accurately describe RSS drilling due to the complexity of its working principles and downhole environment. Conventional machine learning methods have shortcomings such as delayed prediction or insufficient accuracy. To address these challenges, a method based on deep learning is proposed to predict the well path of the RSS drilling process. This method combines graph representation learning with time series forecasting method, which uses dynamic graphs to store drilling parameters and graph attention networks to mine external correlations between non-Euclidean spatial features to improve prediction accuracy. Long short-term memory (LSTM) networks capture the sequential dependencies within features to obtain a certain range of well path prediction values and solve the prediction hysteresis problem in real-time drilling processes. The inclination change rate, azimuth change rate and other indicators were selected as targets of the RSS well path prediction in the experiment. The results show that the proposed method achieves high accuracy and stability in instantaneous prediction tasks, with good applicability and robustness across different prediction targets and ranges. Under the condition of effective data constraints (e.g., measurement while drilling data), the model is capable of maintaining reliable well path construction capability during long-term recursive processes, avoiding significant error accumulation. Based on this finding, we conducted a further investigation into the factors influencing the model's predictive performance to effectively enhance its capability and provide a theoretical reference for Intelligent applications in drilling process.
The swirling abrasive water jet (SAWJ), with its large impact area and strong rock-breaking capability, is currently the optimal method for efficiently creating radial boreholes in hard rock reservoirs such as deep oil/gas and hot dry rock. However, the temporal-spatial evolution of stress and damage in rock under SAWJ impingement remains unclear. This paper establishes a numerical model for SAWJ impacting granite, analyzes the temporal-spatial evolution of stress and damage in granite, and discusses the rock-breaking mechanism of SAWJ. The average error between the model-predicted and experimentally measured outer diameter, inner diameter, and width of the damage zone is 10.28%. Under SAWJ impact, the damage distribution on the granite impact surface is annular, with the internal damage peak occurring at a depth range of 3.2–4.0 mm. Within the annular erosion pit on the impact surface, stress waves exhibit a trend of convergence and superposition toward the jet axis, whereas outside the annular erosion pit, stress waves show a trend of divergence and superposition outward from the annular erosion pit. The primary energy for SAWJ damaging granite comes from the impact of abrasive particles. These findings contribute to a deeper understanding of the interaction process between SAWJ and rock, providing guidance for SAWJ radial drilling in hard rock reservoirs.
This paper presents a comprehensive theoretical and analytical study of a novel distributed pressure sensing (DPS) cable architecture that seamlessly integrates dual pressure and temperature monitoring functionalities within a single fiber-optic structure. The proposed design utilizes a bi-material diaphragm-lever mechanism to physically amplify structural strain at localized sensing zones, enabling highly efficient transduction of external fluid pressure into measurable optical fiber strain. A rigorous analytical model was developed to characterize diaphragm deflection, elastomechanical strain transfer, and geometric amplification, with the theoretical framework thoroughly validated via 3D finite-element simulations. To bridge the mechanical domain with optical signal processing, the simulated strain distributions were translated into expected Rayleigh frequency shifts utilizing established calibration coefficients from commercial tunable wavelength coherent optical time-domain reflectometry (TW-COTDR) instrumentation. Results demonstrate a high-pressure sensitivity of 4.23 με/MPa corresponding to an optical shift of approximately 0.635 GHz/MPa exhibiting an exceptional linear correlation (R2 = 0.995) across the applied pressure range. Although physical prototype fabrication and experimental validation remain as future work, this study firmly establishes the elastomechanical viability of combining localized strain amplification with Rayleigh-based interrogation to engineer scalable, high-resolution DPS cables. Ultimately, these findings lay the foundational groundwork for deploying robust, continuous pressure and temperature monitoring systems in high-pressure, high-temperature (HPHT) downhole environments.
Strong subsurface heterogeneity leads to substantial uncertainty in numerical predictions, thereby motivating the adoption of history-matching approaches to constrain model parameters and reduce predictive uncertainty. While well-based monitoring data provide time-series observations at fixed well locations, they are often insufficient to reduce uncertainty in model parameters far from well locations. Time-lapse seismic surveys enable the interpretation of reservoir-scale dynamic properties, such as phase saturations. However, effectively integrating seismic interpretations into history matching remains challenging because of factors such as grid-resolution differences between flow models and seismic interpretations, as well as spatial error correlations in seismic data. In this study, we develop a history-matching workflow for geological CO2 sequestration that jointly assimilates well monitoring data and seismic-derived CO2 saturation fields using an ensemble-based data assimilation approach. The proposed workflow explicitly accounts for resolution mismatches between seismic interpretations and flow simulation results and incorporates both independent and spatially correlated observation errors for seismic data. A series of numerical experiments is conducted on an aquifer model to investigate (i) the benefit of incorporating seismic data, in addition to standard well-monitoring data, into the history-matching workflow; (ii) the impact of different error levels in seismic-derived data on history-matching performance; and (iii) the effects of seismic acquisition timing and the use of multiple seismic surveys. The results demonstrate that incorporating seismic data significantly improves the accuracy of CO2 plume predictions and permeability estimation, while reducing uncertainty in both short-term and long-term forecasts. Joint assimilation of well and seismic data reduces the median mean squared error (MSE) of late-time CO2 saturation predictions by more than 70%, from approximately 0.07 for the prior predictions to below 0.02 for the posterior predictions. In addition, the proposed history-matching workflow remains numerically stable and effectively reduces posterior uncertainty when the seismic-data error level is set to 30%. Regarding seismic acquisition timing, seismic data collected near the end of the injection period provide a greater reduction in posterior uncertainty than data collected too early, when the CO2 plume is not fully developed, or too late, after the CO2 plume has migrated to the top layers. Overall, the proposed workflow provides a robust and practical framework for history matching with integrated well and seismic data in geological CO2 sequestration.
Sand production monitoring is essential for production safety, equipment integrity, and sand-management decision-making in oil and gas wells. However, existing monitoring methods target different physical quantities and stages of the sand-production process. Sampling captures transported sand particles, acoustic and vibration methods record impact-induced dynamic responses, distributed fiber-optic sensing mainly provides distributed acoustic/vibration-related anomalies, and ER/erosion probes measure local material loss. Therefore, direct comparison using a single performance indicator may lead to misleading technology assessment. This review develops a measurement-target-based framework that links sand generation, transport-impact-erosion processes, signal formation, monitoring outputs, and scenario-based technology selection. Downhole, surface, and subsea monitoring routes are compared in terms of measured quantity, calibration requirement, uncertainty source, field maturity, and decision suitability. Particular attention is given to practical limitations, including DAS interpretation uncertainty, acoustic calibration transferability, ER-probe locality, and subsea reliability constraints. An uncertainty-aware target-specific evaluation framework and a scenario-based technology-selection matrix are proposed to support method combination and decision-oriented sand management. Finally, the review discusses how monitoring data can be coupled with sanding prediction, erosion-risk assessment, safe operating-envelope updating, and choke or intervention decisions. The analysis indicates that reliable sand management benefits from calibrated, cross-validated, and decision-oriented integration of physical sampling, dynamic-response monitoring, distributed diagnostics, erosion assessment, and production data.
The ocean current disturbance is a critical factor accelerating the wear rate of the inner liner in non-metallic flexible pipes used for deep-sea mining, significantly impacting the pipe’s service life, thus making it essential to consider in pipe design. In this study, a coupled dynamic model integrating the bent pipe section of a non-metallic flexible pipe, seawater, and non-spherical particles was established under ocean current disturbance conditions. The key parameters affecting the distribution characteristics of particle dynamics, including particle sphericity, pipe oscillation frequency, and oscillation amplitude, were systematically analyzed. The investigation focused on the evolution of wear on the inner liner under these influencing factors, revealing the mechanisms of particle-wall collision and wear progression. Furthermore, the wear life of the flexible pipe was predicted and evaluated. The results indicate that under the influence of oscillation frequency and amplitude, the geometric characteristics of non-spherical particles and the pipe oscillation parameters exhibit a pronounced nonlinear relationship with the wear rate. High-frequency, large-amplitude oscillations tend to suppress the lateral propagation of the wear area, thereby intensifying localized wear in the wall thickness direction. In the wear life assessment incorporating a safety factor of 1.5, the minimum service life occurs under the conditions of a maximum particle size of 20 mm, sphericity of 0.8, oscillation amplitude of 15°, and oscillation frequency of 1 Hz. These findings can provide solid theoretical support and technical guidance for the wear life assessment and structural design of non-metallic flexible pipes in deep-sea mining.
Classical staggered-grid finite-difference (C-SFD) methods for elastic-wavefield extrapolation can achieve high-order spatial simulation accuracy but only second-order temporal accuracy. Consequently, these operations may suffer from significant numerical dispersion and instability when large time sampling intervals are used. Constructing combined discretization operators by adding auxiliary grid nodes to the C-SFD stencil can enhance the temporal simulation accuracy considerably. However, the existing temporal high-order SFD operators are often structurally redundant and computationally expensive. Therefore, to synchronously improve temporal and spatial simulation accuracy without sacrificing computational efficiency, we develop a novel radial multi-axial SFD (RMA-SFD) stencil. This stencil comprises the C-SFD scheme and a few additional grid nodes along four radial directions for each partial derivative component. By incorporating the plane-wave assumption and a linear optimization algorithm, we derive time-space domain high-order SFD coefficients for this modified stencil. To generate reliable P-, S-, and converted-wave responses, we perform elastic-wave simulation by combining wavefield decomposition with the proposed RMA-SFD scheme. In our approach, the separated P- and S-wavefields are extrapolated using their respective dispersion-relation-based SFD coefficients. The performance of the modified RMA-SFD method is evaluated against the C-SFD method through dispersion analysis and several synthetic examples. Numerical results demonstrate that the proposed method can generate higher temporal and spatial accuracy than the C-SFD method. Moreover, our RMA-SFD method can attain satisfactory computational efficiency by permitting larger time steps. In general, our proposed scheme can be regarded as an effectiveness wavefield continuation tool for elastic-wave imaging and inversion.
The bottom-hole flow around hybrid drill bits with a recessed central structure involves strong interactions among impinging jets, local recirculation, and cuttings transport, which directly affect cleaning efficiency. In this study, the bottom-hole flow and particle migration of a recessed hybrid drill bit with a central conical cutter and multi-blade structure are investigated using a coupled computational fluid dynamics and discrete phase model (CFD-DPM), with validation against a published particle image velocimetry (PIV) flow-field benchmark and full-scale experimental cuttings removal results. The baseline flow field shows that nozzle jets strike the bit face and generate a strong radial sweeping flow. At the same time, recirculation develops in the recessed center and in the regions between adjacent blades. These flow structures enhance local agitation, but they also create low-velocity zones that trap particles and delay their removal. Parametric analyses are carried out to examine the effects of pump rate, concave height, and nozzle angle on cuttings transport. The results show that cleaning performance depends on the balance between jet-driven entrainment and recirculation-induced retention. Increasing pump rate improves cuttings removal at low and moderate levels. However, excessive jet strength intensifies recirculation and promotes particle trapping. Within the investigated parameter range, the most favorable cleaning performance is obtained at a pump rate of 20 L/s, a concave height of 35 mm, a polycrystalline diamond compact (PDC) nozzle angle of 25°, and a central cutter nozzle angle of 7°. These results clarify the flow features governing cuttings transport in recessed hybrid bit geometries and support hydraulic design for deep and ultra-deep drilling.
Sandstone acidizing effectively removes mud and particulate blockages, dissolves acid-soluble minerals, and enhances porosity and permeability, thus increasing reservoir productivity. However, traditional acidizing methods face challenges such as poor adaptability, uneven effectiveness, increased costs, and risks in complex reservoir environments. Underwater high-voltage electric pulse discharge technology can instantaneously release high-energy shockwaves, improving rock porosity and permeability conditions. Combining electric pulse and acidizing technologies reduces acid usage, mitigates reservoir damage, and enhances overall treatment effectiveness. In this study, a high-voltage pulse discharge platform was used to investigate fracture propagation, permeability evolution, and microstructural changes in sandstone cores under different discharge conditions. Results indicate that shockwaves preferentially damage weak rock zones, with fracture propagation influenced by pre-existing fissures. Damage to structurally intact rock primarily manifests as mechanical erosion and fatigue fracturing. Microscopically, pores enlarge and interconnect into fractures, significantly improving permeability. Acid dissolution experiments revealed that a mixed acid solution of 12% hydrochloric acid and 3% hydrofluoric acid achieves optimal dissolution. Composite technology experiments demonstrated that for permeability enhancement, the optimal procedure involves electric pulse application followed by acidizing. Conversely, for deep acid-soluble blockages, acidizing before electric pulses is most effective; for shallow, acid-insoluble blockages, electric pulses followed by acidizing yield better results. This study confirms the efficacy of the combined electric pulse-acidizing technique, proposing a novel, complementary method for reservoir stimulation and blockage removal.
Non-metallic unbonded flexible pipes, known for their significant advantages such as lightweight design, high wear resistance, and excellent load-bearing capacity, show broad application prospects in deep-sea hydraulic pipeline lifting mining systems. A finite element model with unidirectional fiber layup is established to investigate the effects of external pressure and interlayer friction coefficient on bending stiffness under monotonic and cyclic bending. Nodal stresses at the rectangular cross-section are analyzed. Results show that external pressure alters interlayer contact pressure, increasing the curvature required for strip slippage; after slippage, strain energy growth slows. Cyclic bending induces significant hysteresis from interlayer slip. Higher external pressure and friction coefficient enlarge the moment-curvature loop, increase the friction force for slip, and enhance energy dissipation. Additionally, external pressure modifies the strain state of helical strips in the tensile reinforcement layer, leading to hysteresis in nodal stresses, which vary with strip coordinates.
Global ultra-deep oil and gas resources are abundant and possess tremendous development potential. However, ultra-deep reservoirs are characterized by ultra-high temperatures exceeding 200 °C and extremely high formation fracturing pressures, which pose significant challenges to reservoir stimulation. In this work, a novel hydrophobically associating acid thickener (PCHAP) with a viscosity-average molecular weight of 1.067 × 107 was synthesized via aqueous solution polymerization. PCHAP retains over 50% of its viscosity in ultra-high-salinity brine and maintains high phase angle and consistency factor under high shear, demonstrating excellent thickening and fluid-loss control that effectively reduces the acid–rock reaction rate. The weighted thickening acid, formulated with ZnCl2 as a weighting agent, exhibits good shear resistance, corrosion inhibition, and compatibility at 200 °C, along with a low acid–rock reaction rate and a low H+ mass transfer coefficient. The developed system meets the technical requirements for weighted pre-acidification in ultra-high temperature deep reservoirs, and has been successfully applied in an 8,900 m ultra-deep carbonate well in the Tarim Basin. This study provides reliable technical support for the safe and efficient development of ultra-deep oil and gas reservoirs.
The Eocene organic-rich lacustrine shales in the South China Sea were deposited in fresh-brackish environment under tropical climate including the early Eocene Climatic Optimum (EECO, 53–49 Ma) and the Middle Eocene Climatic Optimum (MECO). In the Eocene Liushagang Formation (LF), biomarkers diagnostic for cyanobacteria and algae, together with organic petrology and inorganic geochemical data from ninety-three shale samples in different sedimentary environments were comprehensively analyzed to investigate algal and bacterial responses to the EECO. Cyanobacteria-derived C28–C34 2-methylhopanes during the EECO in shallow lake and during the subsequent 49.0–48.5 Ma in semi-deep lake exhibit higher abundance than in other Eocene time intervals. In contrast, there are remarkable decline of C28–C34 2-methylhopanes and increase of C30 4-methylsteranes from periods of 48.5–47.5 Ma to 47.5–46.47 Ma when the lower LF second member was deposited. Under the inferred terrestrial mean annual temperature > 25 °C during the EECO and the subsequent 0.5 Ma, cyanobacteria were thought to have competitive advantage over dinoflagellate-dominated algae in the shallow water and semi-deep lake. During 48.5–46.47 Ma when the lower LF second member was deposited, there were declined temperature and significantly enhanced monsoonal hydrodynamics and nutrient supply, dinoflagellate-dominated algae had higher abundance than cyanobacteria in the deep-water lake. During 46.47–39.2 Ma, there were not much cyanobacterial or algal contribution under the sustained temperature decline and decreased nutrients supply under global cooling and weakened monsoonal hydrodynamics and. These are partially supported by various telalginite and lamalginite, plant-derived vitrinite content in macerals and inorganic parameters indicative for climate and sedimentary conditions in the upper LF third member and the LF second and first members. They collectively suggest that there is higher cyanobacterial contribution relative to dinoflagellate-dominated algae during the EECO and the subsequent 0.5 Ma in shallow and semi-deep lakes. This finding provides analogues for the coeval lacustrine organic-rich shales deposited during the EECO and the subsequent 0.5 Ma in the tropical South China Sea and other ancient tropical lakes during the same time intervals.
Rocks containing inclined fractures can be approximated as tilted transversely isotropic media with a tilted symmetry axis (TTI). Azimuthal differences in seismic data enable quantitative estimation of fracture-related parameters in TTI media. However, the inversion accuracy is limited by weak azimuthal differences in seismic amplitudes and the empirical selection of regularization parameters and constraint forms. In this study, an approximate PP-wave reflection coefficient for TTI media is derived in a form that is directly parameterized by P-wave velocity, S-wave velocity, density, and fracture weaknesses. The elastic and seismic responses are analyzed to examine the effects of fracture, fracture fluid, fracture dip angle, and background medium parameters. An adaptive AVAZ inversion method is then developed. A stepwise inversion strategy for TTI media is adopted to alleviate the ill-posedness of multiparameter inversion, in which background elastic parameters are first estimated, and fracture parameters are then inverted from azimuthal difference data. To improve inversion accuracy, a kurtosis-based adaptive Lp-norm constraint is proposed for the objective function, with the sparsity level adaptively adjusted based on the statistical characteristics of the model parameters. The resulting non-L2 inversion problem is solved using the Iteratively Reweighted Least Squares method. The generalized cross-validation (GCV) approach is introduced to determine the optimal regularization parameter. Synthetic tests indicate that AVAZ inversion accuracy is affected by the regularization parameter selection, the regularization constraint, and the input fracture dip angle. The tests also show that the proposed method improves the inversion accuracy of fracture-related parameters and remains robust to noise. Application of the method to field seismic data further demonstrates its effectiveness.
Severe erosion of hydraulic fracturing sandblasting nozzles commonly occurs under high pumping-rate and high proppant-concentration conditions, posing a serious threat to the safety and reliability of fracturing operations. However, previous studies have rarely considered inter-particle interactions, and the associated erosion mechanisms remain insufficiently understood. In this study, the Dense Discrete Phase Model (DDPM) coupled with the Kinetic Theory of Granular Flow (KTGF) was employed to investigate erosion behavior in hydraulic fracturing sandblasting nozzles. The O’Rourke stochastic collision model was incorporated into the DDPM-KTGF framework, and the particle collision mechanism was systematically analyzed. Furthermore, an information exchange strategy between particles and fluid-phase grids was developed, establishing a numerical methodology for erosion prediction in hydraulic fracturing sandblasting nozzles based on the DDPM-KTGF coupled with the O’Rourke stochastic collision model. A side-wall throttling sandblasting nozzle was selected as the study object. Numerical simulations and erosion experiments were conducted under high pumping-rate and high proppant-concentration conditions, and good agreement between the simulation and experimental results was achieved. The results demonstrate that particle-particle collisions significantly influence erosion morphology. Frequent impacts of high-velocity particle flows on the wall surface, together with secondary erosion induced by particle rebound, are identified as the dominant mechanisms governing nozzle erosion. The erosion rate is primarily controlled by pumping rate, followed by fluid viscosity and proppant concentration, whereas particle size, operating pressure, and fluid density exert only minor effects. Response surface optimization results show that the maximum erosion rate of the optimized nozzle decreased by 77% compared with the original structure. The present study provides theoretical support for the anti-erosion design and optimization of hydraulic fracturing sandblasting nozzles and offers valuable guidance for improving the operational reliability of fracturing equipment.
This study investigates Archean metamorphic buried-hill reservoirs in the Bohai Bay Basin, focusing on their formation mechanisms, multi-stage fluid evolution, and geochemical characteristics. Protolith composition, chemical index of alteration (CIA), formation water chemistry, and carbon-oxygen isotopes were analyzed for three blocks: BZ19-6, BZ13-2, and BZ26-6. The protolith is dominated by sedimentary rocks including claystone, mudstone, siltstone, feldspathic sandstone, and argillaceous limestone, with minor intermediate igneous rocks and tuff. CIA values vary among blocks and stratigraphic intervals, reflecting differential weathering controlled by overlying lithology and thickness. Higher CIA corresponds to stronger meteoric leaching and better reservoir porosity. Formation water types differ systematically. BZ19-6 is dominated by Na-HCO3/SO4-type water, indicating meteoric infiltration and silicate weathering. BZ13-2 and BZ26-6 are dominated by Na-Cl-type water, suggesting residual brine or seawater influence. C-O isotopic data distinguish three calcite types: meteoric, hydrothermal, and organic-acid-modified. Their spatial distribution shows that BZ26-6 experienced the strongest hydrothermal overprint and organic acid alteration, while BZ19-6 lacks the organic acid signal. These results clarify the tectonic-fluid coupling that shaped the reservoirs. Indosinian to Yanshanian fracturing created pathways for simultaneous meteoric leaching and hydrothermal baking. Himalayan transtension enhanced differential dissolution and prolonged thermal alteration. In hydrocarbon-charged blocks, thermochemical sulfate reduction added late-stage corrosive fluids. The proposed isotopic framework provides a practical tool for identifying fluid sources in similar ancient basement reservoirs.
To improve the rheological properties of polyacrylamide-based fracturing fluids, a common approach is to increase the amount of copolymer; however, this inevitably leads to greater damage to the reservoir. In this study, a self-degradable polyacrylamide (DS-PAM) with high rheological performance was synthesized, featuring disulfide bonds and a dual-network structure (physicochemical structure). The as-prepared fracturing fluid had excellent temperature-shear resistance (> 50 mPa·s) and viscoelasticity (G′ > G″) at 100 °C and 40,000 mg/L salinity. More importantly, kerosene facilitates the “competitive capture” of hydrophobic groups, causing dissociation of the hydrophobic associative network. Concurrently, hydrolysis of disulfide bonds triggers cleavage of the covalent network, while, high temperature provides sufficient activation energy for the degradation reaction, accelerating the scission of molecular chains. The degradation process follows the first-order kinetic model. Under the synergic effects of physical network dissociation and chemical network cleavage, degradation products with molecular weight less than 200,000 are achieved within 48 h. The scaling rate (< 0.5%) and core damage (19.8%) of the degradation fluid are substantially reduced, which is highly significant for reservoir protection.
Drill cuttings serve as a fundamental source of geological information in oil and gas exploration and development. However, traditional analysis remains largely manual, resulting in low efficiency and high subjectivity. While recent studies have introduced artificial intelligence, most focus on single-attribute tasks and purely on data-driven approaches, often neglecting geological domain knowledge. This highly limits model generalization and interpretability in real-world applications. To address these challenges, we propose a multi-attribute intelligent identification model based on multi-task learning, which integrates geological prior knowledge and introduces a novel collaboration classifier to enhance the comprehensiveness, robustness, and interpretability of drill cuttings analysis. Specifically, the model constructs structured inter-attribute correlations as prior knowledge and employs a collaboration classifier designed based on the law of total probability, enabling collaborative modeling and cross-attribute information exchange. These innovations significantly improve recognition performance across multiple attribute dimensions. Experimental results demonstrate that the proposed model achieves a mean accuracy of 93.90% and a mean F1-score of 93.08% in identifying cuttings attributes including color, shape, and grain type. These results validate the effectiveness of the structured prior-guided collaborative modeling strategy and highlight its potential for complex multi-attribute analysis tasks in petroleum geology.
Due to the lack of labeled data in field seismic data acquisition, unsupervised methods have gained widespread attention in recent years. Most existing unsupervised interpolation methods rely solely on the L2 norm for loss calculation on the sampled data, lacking appropriate constraints on the interpolated data. This limitation leads to poor performance for regularly subsampled seismic data with strong spatial aliasing. To address this issue, we propose an unsupervised interpolation method based on the Soft Dynamic Time Warping Divergence (SDTWD) distance. The proposed method constrains the interpolated data by minimizing the SDTWD distance. This is a differentiable misfit measurement between the interpolated trace and its neighboring sampled traces. We enhance the network’s feature extraction capability by incorporating a Simple, Parameter-Free Attention Module (SimAM) into the conventional UNet architecture. Moreover, we integrate the reinsertion step of Projection Onto Convex Sets (POCS) algorithm into the network’s iterative process to further improve interpolation quality. Tests on both synthetic and field datasets demonstrate the superiority of the proposed method compared to the traditional f-x prediction filtering method and unsupervised methods based on deep image priors.
Industrial symbiosis networks, characterized by complex inter-industry resource exchanges, propagate localized disruptions across multiple sectors, thereby generating greater systemic uncertainty than single-industry systems. This inherent complexity undermines reliable energy conservation and emission reduction (ECER) planning in promoting industrial symbiosis and heightens the risk of ineffective policies. In this study, we develop a probabilistic framework integrating Latin hypercube sampling and Sobol sensitivity analysis to assess the impacts of 389 uncertain factors on China's iron and steel–thermal power–cement symbiosis network from 2020 to 2030. Our findings indicate that: (1) Nationwide industrial symbiosis could achieve energy savings of 47.1–61.2 Mtce and significantly reduce emissions of SO2 (170.3–248.9 kt), NOx (166.7–280.3 kt), and PM (41.8–96.7 kt) by 2030, with required investments ranging from 92.8 to 122.1 billion USD. By balancing ambition and feasibility, we establish policy-actionable ECER targets for the nationwide industrial symbiosis network based on a 75% feasibility threshold across 100,000 uncertain scenarios. (2) Conservation supply curve analysis shows that over 95% of the symbiotic technologies exhibit negative energy conservation costs under uncertainty, yet their cost volatility spans several orders of magnitude. Prioritizing 27 high-potential, low-cost technologies can capture 76.9% of the network’s total ECER potential. (3) Technology penetration rates—rather than intrinsic efficiencies—account for 84.5% of the variance in ECER outcomes, underscoring the importance of policies that foster widespread adoption of symbiotic technologies over incremental technological efficiency improvements. This study redefines industrial symbiosis planning from static roadmaps into dynamic adjustment mechanisms, offering robust support for promoting industrial symbiosis under uncertainty.
The remediation of petroleum hydrocarbons (PHs) contaminated marine sediments has attracted worldwide attention. Sediment microbial fuel cells (SMFCs), which provide inexhaustible electron acceptors, offer a promising approach by integrating contaminant degradation with bioelectricity generation. However, the remediation efficiency is often hindered by the slow formation and weak activity of electroactive biofilms on the anode. This study proposed a novel synergistic strategy that combines bioaugmentation with exogenous quorum sensing (QS) signals (C4-HSL) to enhance the removal efficiency of PHs. In addition, an open-circuit control (SMFC-OC) was employed to illustrate the contribution of electrochemical current. The closed-circuit SMFC enhanced total petroleum hydrocarbon (TPH) removal to 45.9% ± 1.7% compared to 25.3% ± 1.6% in the SMFC-OC system, underscoring the essential role of a complete electrical circuit. The combined treatment (SMFC+B+C4) further achieved a superior TPH removal rate of 73.7% ± 1.5%, significantly outperforming the SMFC-alone system. Concurrently, it generated a maximum power density 5.35 times greater than the control SMFC, underscoring a critical synergy between electrokinetic stimulation, bioaugmentation, and QS-mediated regulation. Electrochemical analyses revealed that C4-HSL facilitated the catalytic activity of anodic biofilm, and markedly reduced charge transfer resistance to 35.3 Ω. Microbial community analysis demonstrated a selective enrichment of PH-degrading bacteria (e.g., Planococcus, Sporolactobacillus) and electroactive bacteria, coupled with increased abundances of functional genes associated with PH degradation and extracellular electron transfer. This selective restructuring fostered a more modular and stable ecological co-occurrence network, which directly contributed to enhanced bioremediation and power generation. This work provided a novel strategy for syntrophic remediation of petroleum-contaminated marine sediments, and fundamental understanding of managing complex biofilms in bioelectrochemical systems through exogenous QS manipulation.
