Large-scale Distribution Research Articles

Evaluation Method for Hosting Capacity of Rooftop Photovoltaic Considering Photovoltaic Potential in Distribution System

Regarding the existing evaluation methods for photovoltaic (PV) hosting capacity in the distribution system that do not consider the spatial distribution of rooftop photovoltaic potential and are difficult to apply on the actual large-scale distribution systems, this paper proposes a PV hosting capacity evaluation method based on the improved PSPNet, grid multi-source data, and the CRITIC method. Firstly, an improved PSPNet is used to efficiently abstract the rooftop in satellite map images and then estimate the rooftop PV potential of each distribution substation supply area. Considering the safety, economy, and flexibility of distribution system operation, we establish a multi-level PV hosting capacity evaluation system. Finally, based on the rooftop PV potential estimation of each distribution substation supply area, we combine the multi-source data of the grid digitalization system to carry out security verification and indicator calculation and convert the indicator calculation results of each scenario into a comprehensive score through the CRITIC method. We estimate the rooftop photovoltaic potential and evaluate the PV hosting capacity of an actual 10 kV distribution system in Shantou, China. The results show that the improved PSPNet solves the hole problem of the original model and obtains a close-to-realistic rooftop photovoltaic potential estimation value. In addition, the proposed method considering the photovoltaic potential in this paper can more accurately evaluate the rooftop PV hosting capacity of the distribution system compared with the traditional method, which provides data support for the power grid corporation to formulate a reasonable PV development and hosting capacity enhancement program.

Spatial Pattern of Large-Scale Agricultural Land and Spatial Heterogeneity of Influencing Factors in the Mountainous Areas of Western China—Wuling Mountains as an Example

The scaling of agricultural land is a trend in land use transformation and is important for modernizing agriculture. Therefore, the reasons for large-scale agricultural land formation should be explored. The spatial distribution of large-scale agricultural land and the factors for its formation vary between different regions. Currently, the exploration of the formation mechanism of large-scale agricultural land from the perspective of heterogeneity is not yet sufficient. Therefore, the main objectives of this article are as follows: first, analyze the spatial pattern characteristics of large-scale agricultural land; second, explore the spatial heterogeneity characteristics of influencing factors from both global and local perspectives; third, explore the mechanism of the formation of large-scale agricultural land from the perspective of heterogeneity. The results indicate the following: (1) The large-scale agricultural land distribution pattern in the Wuling Mountains area was high in the east and low in the west. (2) Natural conditions, production factors, and location conditions all significantly impacted large-scale agricultural land, but with differences in their degree of influence. From a local perspective, the influences of various factors in different regions also exhibited spatial heterogeneity. These two types of heterogeneity can be attributed to the differences in regional development stages. (3) Natural conditions, location conditions, and production factors had negative, positive, and positive effects on the agricultural land scale, respectively, but the influence of the first two decreased with the improvement in the regional development stages. The influence of different factors on production factors was related to the regional development stage, and production factors that were suitable for the regional development stage had a greater impact. The conclusion can provide differentiated policy support for regional land use in practice.

A multi-perspective functionality loss assessment of coupled railway and airline systems under extreme events

The coupled national-scale railway and airline systems (CRASs) have drastically improved inter-city connectivity and national economics, but they remain susceptible to various extreme events. These events, including flooding and typhoon, often manifest as localized events because all direct interrupted components lie in a small region relative to the large-scale distribution of CRASs. Instead of the hazard-specific modeling of each localized event, this paper introduces four types of localized disruption models to simulate localized events with various locations, impact coverages, and time spans. A multi-perspective framework is then proposed for travel time-based functionality loss assessment of CRASs under localized events. Taking CRASs in China as an application, results demonstrate that (1) the functionality loss is highly sensitive to the event location, and events in areas with more population, higher GDP, and larger transport systems tend to cause higher functionality loss; (2) the impact coverage of events has a limited influence on functionality loss, as critical areas identified under circle-shaped impact coverage keep consistent with those under administrative district-based coverage; (3) the functionality loss and critical areas vary largely with the time span of localized events. The findings provide valuable insights in devising mitigation strategies for CRASs against various extreme events.

Exploration and application of new distribution system technology for large-scale distributed distribution network digital service access

Abstract Digital technology is leading fundamental innovation and development, and digitalization represents a great opportunity for the energy industry to take off and develop. In this paper, digital and intelligent technologies are used to integrate, analyze, and manage distribution network equipment information. Secondly, a new smart distribution terminal design is proposed and modeled using the IEC 61850 standard to monitor the status of smart distribution grid management equipment. Through the active distribution system structure, based on the decoupled PCC, a model is constructed for large-scale distributed optimal dispatching. Finally, two scenario experiments are taken as examples to simulate and analyze the distributed co-dispatch of the new distribution system. The results show that the operation reward interval of scenario 2 is [194.43, 205.44] when the decomposition convergence criterion is set to 0.05 and the threshold value of the second-order cone relaxation accuracy is set to 10−5. The novel distribution system proposed in this paper is able to achieve optimal decision-making that satisfies the distribution network operation constraints, while the multi-microgrid seeks to maximize its benefits.

Multi-area state estimation for active distribution networks under multiple uncertainties: An affine approach

An enormous challenge for the state estimation (SE) process in large-scale distribution networks lies in how to deal with multiple uncertainties and the increasing computational tasks simultaneously. This paper aims to address the issue and propose a multi-area architecture for perceiving system status with uncertain inputs. The first step involves the introduction of a novel Louvain-flow-based partition structure, which effectively divides large-scale unbalanced distribution networks into several sub-areas. Subsequently, an affine state estimation (ASE) model, employing the Neumann series method, is formulated for each sub-area, taking into account uncertainties in system measurements and line parameters. Finally, the interior-point filter line-search algorithm combined with consensus theory is utilized to solve all local ASE models, and the interaction of boundary state between adjacent sub-areas is focused and completed, to output the global state results of the distribution network precisely. Case studies are presented to verify that the proposed distributed ASE model yields obvious advantages in tracking system uncertainties by contrast with the existing SE model.

Allocation of PV Systems with Volt/Var Control Based on Automatic Voltage Regulators in Active Distribution Networks

This paper presents an optimal allocation methodology of photovoltaic distributed generations (PVDGs) with Volt/Var control based on Automatic Voltage Regulations (AVRs) in active distribution networks considering the non-dispatchable mode of PVDG operation. In the proposed methodology, an intelligent coordinated Var control is activated via controlling the AVR tap position and the Var injection of PV inverters to achieve a compromise between reducing active and reactive power losses and enhancing voltage quality in a distribution network. Also, the scheduled power factor mode of operation is investigated for the PV inverters. Added to that, the proposed allocation methodology is handled on the basis of hourly loading variation under simultaneous control modes of PV inverters and AVR. Moreover, the impacts of the specified number of PVDGs are assessed on the distribution system’s performance. A recent effective optimizer of the slim mold algorithm (SMA) is dedicated to solving the proposed optimization framework. The simulation implementations are executed on a practical distribution network of the Kafr Rabea area related to South Delta Electricity Company in Egypt. Also, the application is conducted for a large-scale distribution network from the metropolitan area of Caracas. The proposed methodology provides superior performance in minimizing the active and reactive power losses and improving the voltage profile.

A multicentennial mode of North Atlantic climate variability throughout the Last Glacial Maximum

Paleoclimate proxy records from the North Atlantic region reveal substantially greater multicentennial temperature variability during the Last Glacial Maximum (LGM) compared to the current interglacial. As there was no obvious change in external forcing, causes for the increased variability remain unknown. Exploiting LGM simulations with a comprehensive coupled climate model along with high-resolution proxy records, we introduce an oscillatory mode of multicentennial variability, which is associated with moderate variations in the Atlantic meridional overturning circulation and depends on the large-scale salinity distribution. This self-sustained mode is amplified by sea-ice feedbacks and induces maximum surface temperature variability in the subpolar North Atlantic region. Characterized by a distinct climatic imprint and different dynamics, the multicentennial oscillation has to be distinguished from Dansgaard-Oeschger variability and emerges only under full LGM climate forcing. The potential of multicentennial modes of variability to emerge or disappear in response to changing climate forcing may have implications for future climate change.

Confident learning-based Gaussian mixture model for leakage detection in water distribution networks

Leakage detection in the water distribution system not only helps to reduce water waste but also decreases the risk of drinking water pollution. To reduce reliance on hardware devices and enable real-time detection, the water utilities are transitioning towards the data-driven based approach that relies on the analysis of the flow and pressure data collected from the supervisory control and data acquisition (SCADA) system. Due to the lack of leakage data, most of these methods are unsupervised methods that rely heavily on assumptions about the distribution of anomalies; whereas, the water utility's repair records contain much valid information about the leakage and normal characteristics. To convert this information into available labels and to address the lack of leakage data, this paper proposed a new leakage detection framework to infer the pressure characteristics under normal conditions based on historical data by combining a label cleaning method—confident learning (CL)—with an unsupervised method—Gaussian mixture model (GMM)—for leak detection. The methodology is validated with synthetic and real measured data. Comparisons with four unsupervised methods demonstrate that the GMM method has superior identification of leak features from pressure data. For a real-world water distribution system in K city containing 91 pressure sensors, the average true positive rate is 0.78 and the average false positive rate is 0.11. This methodology is a promising tool to identify signs of leakage in large-scale water distribution networks (WDNs).

The dehydration carousel of stratospheric water vapor in the Asian summer monsoon anticyclone

Abstract. During the StratoClim Geophysica campaign, air with total water mixing ratios up to 200 ppmv and ozone up to 250 ppbv was observed within the Asian summer monsoon anticyclone up to 1.7 km above the local cold-point tropopause (CPT). To investigate the temporal evolution of enhanced water vapor being transported into the stratosphere, we conduct forward trajectory simulations using both a microphysical and an idealized freeze-drying model. The models are initialized at the measurement locations and the evolution of water vapor and ice is compared with satellite observations of MLS and CALIPSO. Our results show that these extremely high water vapor values observed above the CPT are very likely to undergo significant further freeze-drying due to experiencing extremely cold temperatures while circulating in the anticyclonic “dehydration carousel”. We also use the Lagrangian dry point (LDP) of the merged back-and-forward trajectories to reconstruct the water vapor fields. The results show that the extremely high water vapor mixed with the stratospheric air has a negligible impact on the overall water vapor budget. The LDP mixing ratios are a better proxy for the large-scale water vapor distributions in the stratosphere during this period.

Precipitation seasonality and soil pH drive the large-scale distribution of soil invertebrate communities in agricultural ecosystems.

Soil invertebrates contribute significantly to vital ecosystem functions such as the breakdown of organic matter and cycling of essential nutrients, but our knowledge of their large-scale distribution in agricultural systems is limited, which hinders our ability to robustly predict how they will respond to future global change scenarios. Here, we employed metabarcoding analysis of eukaryotic 18S rRNA genes to examine the diversity and community composition of invertebrates in 528 sorghum rhizosphere and bulk soils, collected from 53 experimental field sites across China. Our results revealed that Nematoda, Arthropoda, and Annelida were the dominant soil invertebrate groups in agroecosystems. Among all climatic and soil parameters we examined, precipitation seasonality (i.e. the irregular distribution of precipitation during a normal year) had the strongest relationship with the richness of soil invertebrates, with an increase in soil invertebrate richness predicted with increasing precipitation seasonality. Mean annual precipitation and soil pH were the most important predictors of soil invertebrate community structure, with numerous invertebrate phylotypes showing either significantly positive or negative relationships with these two variables. Our findings suggest that shifts in precipitation patterns and soil pH, induced by future climate change and agricultural practices, will have important consequences for the distribution of soil invertebrate communities, with implications for agricultural ecosystem sustainability.

Regional sets of diverse, specialized ground beetle and plant assemblages recolonise China’s restored temperate forest landscapes

Following the historic destruction of China’s forests, large protected forest landscapes have been reestablishment across the country in recent decades. Currently, the species assemblages inhabiting these landscapes remain widely unknown. Here, we aim to establish whether the new temperate forest landscapes in Eastern China, composed of plantation, secondary and remnant mature forest patches, are dominated by widely distributed habitat generalists, widely distributed habitat specialists, or by sets of highly localized forest habitat specialist species, using vascular plants and ground beetles as case studies. We sampled plants and ground beetles on 159 plots representing different forest types distributed across ten study regions. The Chord-normalized expected species shared (CNESS) dissimilarity for beetles and Jaccard dissimilarity for plants were calculated to establish species’ turnover patterns, while we used structural equation and multiple regression models for dissimilarity matrices to establish the relative importance of location, elevation and biotic interactions in explaining large-scale distribution patterns of species assemblages. Both taxa showed highly regionalized recruitment, with plant communities responding strongly to latitude, and to geographic location in general, while the composition of ground beetle communities appeared strongly governed by both the composition of the vegetation and study plot location. Both assemblages showed a distinct lack of widespread habitat generalists. Our results demonstrate for the first time that local assemblages of forest specialist species appear to have survived last century’s deforestation in unknown refugia before successfully re-colonizing China’s restored forest ecosystems. We argue that restored forest landscapes can play a significant role in the conservation of temperate China’s biodiversity at both local and regional scales.

Where does floating Sargassum in the East China Sea come from?

Floating macroalgae of Sargassum horneri (S. horneri) in the East China Sea (ECS) has increased in recent years, with ocean warming being one of the driving factors. Yet their possible origins, based on a literature review, are unclear. Here, using multi-sensor high-resolution remote sensing data and numerical experiments for the period of 2015–2023, we show two possible origins of the ECS floating S. horneri, one being local near the Zhejiang coast with initiation in January–February and the other being remote (> 800 km from the first) in the Bohai Sea with initiation in June–November. While their drifting pathways are revealed in the sequential remote sensing imagery, numerical experiments suggest that S. horneri from the remote origin (Bohai Sea) can hardly meander through the strong Yangtze River frontal zone, which may serve as a “wall” to prevent trespassing of surface floating seaweed to the south of the frontal zone, where S. horneri has a local origin. Plain language summarySargassum horneri (S. horneri) is a brown macroalgae (seaweed) abundant in surface waters of the East China Sea (ECS), which can serve as a moving habitat, but can also cause major beaching events and environmental problems. Knowledge of its origins is important to help implement mitigation strategies and understand possible ecological impacts along its drifting pathways. Using high-resolution remote sensing images and numerical experiments, we track floating S. horneri in space and time between 2015 and 2023. Two possible origins are identified, one being far away from the ECS and the other being local, both of which are known to have benthic S. horneri. The study also reveals how S. horneri are transported from their source regions resulting in large-scale distributions previously observed in medium-resolution satellite imagery.

Trends in retail sales of insecticide-treated nets and untreated nets in Tanzania: cross-section surveys

BackgroundThe commercial sector plays a vital role in mosquito net ownership and access in Tanzania. The National Malaria Strategic Plan (NMSP) includes long-lasting insecticidal nets (LLIN) delivery through the commercial sector as a complementary mechanism. The NMSP aims to increase LLIN sales while decreasing untreated mosquito net sales. This survey aimed to track quantities, market share of different net categories, prices, and origins of mosquito nets in retail markets and to engage stakeholders to analyse market trends.MethodsThis mixed-method mosquito net retail outlet survey was conducted in mid-2021 in six and in mid-2022 in eight regions. Field teams identified net-selling outlets in major urban and peri-urban markets and used snowball sampling to identify additional outlets. A structured questionnaire was used, and photos of available mosquito net products were taken. Key informant interviews were conducted with wholesalers and retailers. The relative market share of a product was calculated by using the mean of each sales category as frequency weights. Qualitative data analysis was undertaken by summarizing common themes and observations based on the research question.ResultsA total of 394 and 1139 outlets were surveyed in 2021 and 2022, respectively. More than 96% of distributed brands in both years were untreated nets. The market share for untreated mosquito nets was 99.2% in 2021 and 88.3% in 2022. Bed net sales were seasonal, peaking in the rainy season and at the start of the school year. Leaked LLINs from the public sector comprised 0.3% of the market share in 2021 and 8.3% in 2022. Kigoma markets had the most significant frequency of leaked LLIN products. Legitimate LLINs were rare in 2021 (n = 2) and not found in 2022, despite the presence of a local LLIN manufacturer. A small number (n = 3) of untreated nets fabricated in China claiming to be LLINs were observed in 2022.ConclusionsDespite NMCP’s strategic approach to increasing retail market share for legitimate LLINs, significant challenges remain. Efforts are needed to change the current situation given the context of large-scale public sector distributions of LLINs, the higher consumer cost of LLINs, the lack of bed net varieties. Improvement of registration process is recommended.

Multi-threaded power flow of large-scale active multiphase distribution networks

Multi-threaded power flow of large-scale active multiphase distribution networks

DETisSeg: A dual-encoder network for tissue semantic segmentation of histopathology image

Accurate segmentation of tissue regions in hematoxylin and eosin (H&E)-stained histopathology images is an essential step in computational pathology. It is still a challenging task due to the large-scale distribution, irregular morphological variations, and fuzzy boundaries between different tissues. Semantic segmentation methods with strong feature representative capability for extracting local fine-grained and global coarse-grained features are desired to tackle this challenge. In this paper, we propose a dual encoder network for tissue semantic segmentation of histopathology image, called DETisSeg, which aims to fully fuse the features of global context information by the Swin Transformer branch and local features by the convolutional neural network(CNN) branch. Particularly, we design an improved residual connection to recover the lost spatial information during the patch merging phase in the Swin Transformer branch. In addition, we employ a pyramid architecture decoder to generate a composite feature map with multiple scales. We perform experiments on two publicly available tissue semantic segmentation datasets, including the BCSS and LAUD-HistoSeg datasets, to evaluate the effectiveness of the proposed DETisSeg. Compared with eight state-of-the-art semantic segmentation methods, the results show that the proposed DETisSeg achieves improved performance. Particularly, it outperforms Swin transformer on the mean Dice and IoU with improvements of 1.36% and 1.79% on BCSS dataset, 0.96% and 4.48% on LUAD-HistoSeg dataset, respectively.

Estimation and Mapping of Soil Organic Matter Content Using a Stacking Ensemble Learning Model Based on Hyperspectral Images

Fast and accurate SOM estimation and spatial mapping are significant for cultivated land planning and management, crop growth monitoring, and soil carbon pool estimation. It is a key problem to construct a fast and efficient estimation model based on hyperspectral remote sensing image data to realize the inversion mapping of SOM in large areas. In order to solve the problem that the estimation accuracy is not high due to the influence of hyperspectral image quality and soil sample quantity during the estimation model construction, this study explored a method for constructing an estimation model of SOM contents based on a new stacking ensemble learning algorithm and hyperspectral images. Surface soil samples in Huangzhong County of Qinghai Province were collected, and their ZY1-02D hyperspectral remote sensing images were investigated. As input data, a feature band dataset was constructed using the Pearson correlation coefficient and successive projections algorithm. Based on the dataset, a new SOM estimation model under the stacking ensemble learning framework combined with heterogeneous models was developed by optimizing the combination of base and meta-learners. Finally, the spatial distribution map of SOM was plotted based on the result of the model over the study area. The result suggested that the input data quality of the estimation model is improved by constructing a feature band dataset. The multi-class ensemble learning estimation model with the combination strategy of the base and meta-learners has better predictive effects and stability than the single-algorithm and single-level ensemble models with homogeneous learners. The coefficient of determination is 0.829, the residual prediction deviation is 2.85, and the predictive set root mean square error is 1.953. The results can provide new ideas for estimating SOM content using hyperspectral images and ensemble learning algorithms, and serve as a reference for mapping large-scale SOM spatial distribution using space-borne hyperspectral images.

Digital twin assisted decision support system for quality regulation and leak localization task in large-scale water distribution networks

Effective water resource management is essential in large metropolitan cities. Digital Twins (DT), supported by IIoT and machine learning technologies, provide opportunities for real-time prediction and optimization for effective decision-making in water distribution systems. A framework for the digital twin of the Water Distribution Network (WDN) is developed in this paper to achieve higher operational efficiency using ‘WNTR’, the Python-based library of EPANET. All computational experiments and methods were validated on the benchmark hydraulic C-TOWN network (Ostfeld et al., 2011). The hydraulic parameters and quality parameters of the DT model for the water network were calibrated using the Differential Evolution (DE) algorithm. The calibrated DT served as a real-time proxy to generate simulation data, which is used for two different applications in large-scale water networks: (i) Disinfectant dosage regulation task using booster stations and (ii) pipe leakage localization task. The calibrated DT was utilized to estimate the optimal disinfectant dosing rates, ensuring water quality control within an acceptable range using optimization. The results highlight the effectiveness of the neural network and real-time optimization strategy to achieve the optimal dosing rate. For the leakage localization task, the Graph Convolution Networks (GCN) based neural network trained on the DT was found to predict leakage location very accurately.

Indoor radon interval prediction in the Swedish building stock using machine learning

Indoor radon represents a health hazard for occupants. However, the indoor radon measurement rate is low in Sweden because of no mandatory requirements. Measuring indoor radon on an urban scale is complicated, machine learning exploiting existing data for pattern identification provides a cost-efficient approach to estimate indoor radon exposure in the building stock. Extreme gradient boosting (XGBoost) models and deep neural network (DNN) models were developed based on indoor radon measurement records, property registers, and geogenic information. The XGBoost models showed promising results in predicting indoor radon intervals for different types of buildings with macro-F1 between 0.93 and 0.96, whereas the DNN models attained macro-F1 between 0.64 and 0.74. After that, the XGBoost models trained on the national indoor radon dataset were transferred to fit building registers in metropolitan regions to estimate the indoor radon intervals in non-measured and measured buildings by regions and building classes. By comparing the prediction results and the statistical summary of indoor radon intervals in measured buildings, the model uncertainty and validity were determined. The study ascertains the prediction performance of machine learning models in classifying indoor radon intervals and discusses the benefits and limitations of the data-driven approach. The research outcomes can assist preliminary large-scale indoor radon distribution estimation for relevant authorities and guide onsite measurements for prioritized building stock prone to indoor radon exposure.

Alkalinity biases in CMIP6 Earth system models and implications for simulated CO2 drawdown via artificial alkalinity enhancement

Abstract. The partitioning of CO2 between atmosphere and ocean depends to a large degree not only on the amount of dissolved inorganic carbon (DIC) but also on alkalinity in the surface ocean. That is also why ocean alkalinity enhancement (OAE) is discussed as one potential approach in the context of negative emission technologies. Although alkalinity is thus an important variable of the marine carbonate system, little knowledge exists on how its representation in models compares with measurements. We evaluated the large-scale alkalinity distribution in 14 CMIP6 Earth system models (ESMs) against the observational data set GLODAPv2 and show that most models, as well as the multi-model mean, underestimate alkalinity at the surface and in the upper ocean and overestimate it in the deeper ocean. The decomposition of the global mean alkalinity biases into contributions from (i) physical processes (preformed alkalinity), which include the physical redistribution of biased alkalinity originating from the soft tissue and carbonates pumps; (ii) remineralization; and (iii) carbonate formation and dissolution showed that the bias stemming from the physical redistribution of alkalinity is dominant. However, below the upper few hundred meters the bias from carbonate dissolution can gain similar importance to physical biases, while the contribution from remineralization processes is negligible. This highlights the critical need for better understanding and quantification of processes driving calcium carbonate dissolution in microenvironments above the saturation horizons and implementation of these processes into biogeochemical models. For the application of the models to assess the potential of OAE to increase ocean carbon uptake, a back-of-the-envelope calculation was conducted with each model's global mean surface alkalinity, DIC, and partial pressure of CO2 in seawater (pCO2) as input parameters. We evaluate the following two metrics: (1) the initial pCO2 reduction at the surface ocean after alkalinity addition and (2) the uptake efficiency (ηCO2) after air–sea equilibration is reached. The relative biases of alkalinity versus DIC at the surface affect the Revelle factor and therefore the initial pCO2 reduction after alkalinity addition. The global mean surface alkalinity bias relative to GLODAPv2 in the different models ranges from −85 mmol m−3 (−3.6 %) to +50 mmol m−3 (+2.1 %) (mean: −25 mmol m−3 or −1.1 %). For DIC the relative bias ranges from −55 mmol m−3 (−2.6 %) to 53 mmol m−3 (+2.5 %) (mean: −13 mmol m−3 or −0.6 %). All but two of the CMIP6 models evaluated here overestimate the Revelle factor at the surface by up to 3.4 % and thus overestimate the initial pCO2 reduction after alkalinity addition by up to 13 %. The uptake efficiency, ηCO2, then takes into account that a higher Revelle factor and a higher initial pCO2 reduction after alkalinity addition and equilibration mostly compensate for each other, meaning that resulting DIC differences in the models are small (−0.1 % to 1.1 %). The overestimation of the initial pCO2 reduction has to be taken into account when reporting on efficiencies of ocean alkalinity enhancement experiments using CMIP6 models, especially as long as the CO2 equilibrium is not reached.

Investigation on large-scale 3D seepage characteristics of a pumped-storage power station: a case study in Zhejiang Province, China.

Pumped-storage power stations (PSPSs) have higher requirements for anti-seepage compared with regular power stations. As a result, investigating the seepage distributions of PSPSs is particularly important. However, existing researches remain limited in assessing engineering needs such as ensuring the efficiency of a power station. Taking the Qingyuan PSPS as a typical case, this study aims to investigate the large-scale seepage field distribution while exploring the efficiency of the anti-seepage system. Considering the geological characteristics and structural location, a 3D finite element model is established. Based on the continuous medium model while combined with seepage control measures, the change in leakage while the anti-seepage system failed is further assessed. It is concluded that the operation status of anti-seepage measures will have a certain impact on the leakage volumes of each part. Using a comprehensive assessment, anti-seepage measures can effectively prevent seepage. When failure occurs on anti-seepage curtains, the leakage volume at the corresponding position will show an obvious growth. In summary, the findings of this study highlight the significance of avoiding excessive leakage caused by anti-seepage structure failure, the effective operation of anti-seepage measures must be ensured. The abovementioned results can provide scientific support for the seepage optimization design of PSPSs.