Search

It is expected worldwide that the increase in renewable penetration will worsen the voltage profile in contrast to the conventional synchronous generators. This is because it is not proper to authorize all independent power producers’ renewable sources to control grid voltages. Although there has been a bunch of research to improve voltage stability, each has the problem of huge computational effort, limited grid reflection, calibration, etc. Thus, the indirect voltage control by droop can be a substantial solution. Then, it is the very one of the critical issues to determine the proper droop ratio, which might be strongly non-linear due to the complexity of the power system. This paper proposes the online nonlinear P-Q droop estimation of distributed generations. It improves voltage stability, which might get worse after the connection of renewable energies or energy storage devices through inverters. First, power sensitivities between multiple P and Q are derived and used to determine the initial state of the linear P-Q droop required for the initial operation that gets data for precise estimation of the P-Q droop. Thus, it enhanced the estimation performance by reducing the required data yet without the extreme P-Q range. Then, the nonlinear P-Q droop estimation is conducted with the Kalman-filter algorithm. The performance is verified by applying it to the real distribution power system of an island in Korea. For the verification, the distribution power system is modeled at the EMT level and simulated using the power system computer-aided design and electromagnetic transient and DC (PSCAD/EMTDC $^{\mathrm {TM}}$ ). The voltage stability was improved by the proposed nonlinear P-Q droop estimation compared to the cases using fixed droop or Q of zero.

The separation of transmission and distribution systems raises a variety of questions concerning the integration of many distributed generators (DGs) into future grids. They are difficult to solve by using current energy management methods. Especially, the cooperative control between transmission system operator (TSO) and distribution system operator (DSO) has been increasingly emphasized to manage the line congestion. This paper proposes a new cooperative control of TSO-DSO based on the generation-load power sensitivity analysis. To minimize the required computational effort and data communication, the information of TSO and DSO is processed separately in the generation-load power sensitivity matrix of power system. The proposed cooperative control is implemented by three-step process, which is aggregation, specification, and local distribution. Firstly, the DSOs aggregate the flexibility area of DGs in their networks, and they inform the TSO of the feasible operation regions (FORs). Then, the TSO sends the power references at the boundary buses to many DSOs. Thereafter, the DSOs use these references as loads to calculate the detailed power references for their generators. As the result, the net power of DSOs satisfies the power references requested by the TSO. The case studies are carried out to verify the effectiveness of proposed control. The results show that when the load is increased by 20%, the overall average of line loading is decreased by 4.34% with the proposed control. Also, when the generator of 634 MW is disconnected, the frequency nadir is increased by 0.1 Hz compared with the $P$ - $f$ droop control.

The parallel inverter-based scaled-up slack bus includes multiple battery energy storage systems (BESSs) to support the slack operation without short of real power. The multiple BESSs require a well-designed state of the charge (SOC) balancing algorithm for maximally effective usage. Until now, however, the SOC balancing algorithms only focus on strong SOC balancing without careful consideration of BESS efficiency during its charging and discharging. This paper proposes a soft SOC balancing algorithm based on the Levenberg-Marquardt (L-M) algorithm considering BESS’s charging and discharging efficiency. The charges or discharges of the BESSs are primarily determined by their efficiencies that indirectly reflect the SOCs. As a result, the proposed method provides some margins to improve efficiency rather than tightly focusing on only the SOCs balancing between multiple BESSs (e.g., two BESSs with 0.7 and 0.4 pu of SOCs discharge rather than only one BESS with large SOC discharges when total output is 0.86 pus). The optimizations and balancing are conducted by MATLAB software. Their related converter interactions are simulated using a power system computer-aided design/electromagnetic transient including DC (PSCAD/EMTDC $^{\mathrm {TM}}$ ), to measure data for optimization.

The increase in renewable-energy-based generations, such as photovoltaic and wind turbines, inevitably leads to an increase in the number and capacity of inverters connected to the power system. This also increases the voltage on the inverter-connected and adjacent buses. By absorbing reactive power appropriately, it can suppress excessive voltage and increase the potential capacity of real power. Currently, most inverters connected to the power grid are set up in a way that does not involve adjusting the voltage at the point of connection. This means that the inverter controller's limiter settings have not significantly impacted the system's stability after a fault. This article examines the impact of reactive power absorption on grid-connected inverters’ stability and limiter values’ effects on stability. Additionally, to set the limiter values for a convenient inverter controller, the causes of instability are explained in a phasor diagram, and a method for setting the limiter values using this information is explained. The stability impact analysis and limiter value setting are carried out through accurate EMT model-based simulations. The infinite bus with the equivalent impedance is used for the stability analysis and limiter values setting, and the determined values are verified on the real power system. The simulation is conducted using the power system computer-aided design and electromagnetic transient including dc.

DC transformers have emerged as essential devices for medium voltage DC (MVDC)-low voltage DC (LVDC) distribution systems. However, conventional step-down single-level converters have limits on the voltage level of the MVDC-LVDC distribution system. This paper proposes a non-isolated step-down (NISD) DC transformer based on discontinuous conduction mode (DCM). The proposed structure can withstand high voltage levels by sharing voltages between energy storage modules dividing voltage levels. The proposed NISD DC transformer determines operational modes based on energy storage modules and performs the voltage conversion process. The effectiveness of the proposed NISD DC transformer is verified based on a case study using a power system computer-aided design and electromagnetic transient simulation engine including DC (PSCAD/EMTDC™).

The high penetration of converter-based distributed generations (DGs) to power system can give rise to the lack of rotational inertia while replacing the conventional synchronous generators (SGs), which provide the primary frequency reserve (PFR) in power systems. As the result, the decrease in PFR aggravates the frequency stability. To overcome this problem, the droop coefficients of governors in the remaining conventional SGs must be re-determined newly and properly. This paper proposes a new solution based on the grey wolf optimization (GWO) method to optimally select the droop coefficients of SG governors in the low-inertia large-scale power system due to the high penetration of renewables. The proposed solution is very effective for reducing the computational effort significantly, and it is able to recover not only the steady-state but also the transient frequency stability. To verify the effectiveness of proposed optimization solution based on the GWO method, several case studies are carried out on the practical Korea electric power system with the penetration of wind power plants of 4 GW.

The recent railway system is a huge microgrid assembling multiplex structure with distributed active loads, sources and storage devices. The active load represents the train. The sources are a microgrid based on renewable energies. The big problem of the most of electrical train is that they can’t recover energy recovery during regenerative breaking phases. another problem of the electrical train is the long charging time from the stations. This paper suggests a techno-economic process for the energy storage by using SCs in the train, with the aim to reduce the energy consumptions. The proposed design of railway station use PV and wind sources, and batteries for energy storage system (ESS). For the train, SCs are implemented to the ESS where they are alimented breaking phases and from stations by a pantograph installed in an air power line in each stop. SCs are distinguished by high characteristics power and a wide number of charge/discharge cycles, they provide low particular energy and a fast charging time. An energy management approach is suggested to control the DC bus by voltage and the buck-boost converter by current. The Sizing of PI controller used for the stabilization of the DC bus of train and station is given. The whole system is modeled in MATLAB-Simulink. Simulations for the train and station show the suitability of the suggested powertrain and control strategy.

Recently, the generalization of P2P (peer-to-peer) technology with enhanced security due to blockchain technology and the expansion of renewable energy-based distributed energy resources have led to blockchain technology being applied in power transactions, thus giving the potential to become a new platform for DC microgrid operation. Meanwhile, the voltage of a DC microgrid represents the balance of energy supply and demand and also serves as a stability index. The balance is represented as a steady state; the stability is represented during and after events. This paper examines the stability of the DC microgrid built on a university campus in Korea and, in particular, the blockchain technology-based power transactions performed in the DC microgrid. The test is based on the pre-planned transaction schedule applied in the DC microgrid. The transaction schedule has used day-ahead and real-time bidding data. Although many technologies are included in the project, this paper focuses on the voltage stability of the DC microgrid. In addition, the DC protection is applied and evaluated. To consider general DC protection, the DC breaker was simplified with several IGBTs, diodes, capacitors, and arrestors and was designed to interrupt the fault current within five milliseconds. The stability was evaluated using a PSCAD/EMTDCTM.

It is expected that the penetration of renewable-energy-based generators will rapidly increase worldwide. These factors also lead to an increase in the portion of inverters in the power system. This paper proposes a method to scale up the slack generator size based on parallel inverters. All of the multiple parallel inverters except one share the measured power from the slack bus to the grid by a predetermined ratio based on their sizes. A single inverter that does not participate in power sharing is allocated to be in charge of clearing the small power gap caused by the control delay and the measurement error (i.e., the physical slack). As a result, the net size of the scaled-up slack generator is $n$ times greater than that of the physical slack when $n$ non-slack inverters are added to the physical slack inverter. The proposed method is validated based on a real island power system model in Korea using a power system computer-aided design/electromagnetic transient including DC (PSCAD/EMTDC™).

Recently, renewable energy has received considerable attention worldwide due to environmental problems such as climate change. However, there are still many problems to be solved to increase the penetration level of renewable energy sources in a power system. Above all, a power system with a high renewable penetration level requires an entirely novel operational strategy and control approach to achieve reliable operation by considering the intermittency of variable renewable energy sources. This paper proposes power-sensitivities-based indirect voltage control of renewable energy generators with power constraints. The proposed method realizes seamless control of the voltage by compensating with the reactive power if the real power of the distributed generator (DG) is insufficient due to its intermittency. To achieve the proposed method, the real and reactive power references of DGs are initially operated based on the power sensitivities between generation and loads if all the DGs operate within the power constraints. However, if at least one of the DGs reaches its constraint limit, the references can be modified using other sensitivities between real and reactive power generation enabling indirect voltage control of DGs. Therefore, a high level of renewable penetration in power systems can be accomplished with the proposed method. The proposed method is verified with several case studies that are based on a microgrid with practical data from a real island power system. Verification is carried out using the power system computer aided design and electromagnetic transient including DC (PSCAD/EMTDC™) simulation software.

Wind turbine generators (WTGs) are highly sensitive to the disturbances of the grid and tend to disconnect quickly during a voltage dip (when the voltage value is less than 80% of the nominal voltage) or when the frequency is greatly changed. As an increasing number of permanent magnet synchronous generators (PMSGs) are incorporated into the modern power grid, system operators expect PMSG-WT to play an active role in low-voltage ride-through (LVRT) and primary frequency regulation (PFR). Consequently, PMSG-WTs must be capable of supplying additional active power in response to changes in system voltage and frequency. In this context, a new de-loaded technique enhanced by a fuzzy-logic controller is suggested to develop the PMSG-pitch angle control (PMSG-PAC). The studied MG consists of a wind farm (WF), variable load, and a battery energy storage system (BESS). The WF contains five PMSG-WTs which are considered to be the principal resource. The proposed DT-FLC ensures maximum aerodynamic reserve power for the plant, enhances its capability to regulate the PAC, adjusts the WTG drop in response to the wind speed, and increases the resilience of the PMSG-WT in the presence of low voltage. Moreover, the PFR is significantly improved in terms of controlling the PAC (−0.0007 Hz) which meets the frequency maximum droop recommended by the IEEE Std 1547-2018 and the Moroccan grid code, −3 Hz and −2.5 Hz, respectively.

The implementation of distributed generation (DG) becomes more common while causing a challenge to maintain the voltage stability in a power system. The optimal placement of multiple DGs in a planning stage can effectively solve this problem. In many cases, the best placement has been determined by the small-signal study like the modal analysis. However, it is acceptable only if the linearization of large-scale nonlinear power system is accurate. To overcome this, this paper proposes the new optimal placement algorithm for multiple DGs based on the model-free Lyapunov exponent estimation to maintain the voltage of system stable. In other words, the individual Lyapunov exponents of all buses are firstly estimated to determine the candidates for optimal placement of DGs. Then, the maximum Lyapunov exponent for each candidate is calculated to decide which bus is the best place to improve the voltage stability of system. Several time-domain simulation studies are carried out to verify the effectiveness of proposed algorithm. In particular, its performance is compared with that of the conventional method. The results show that the optimal placement of multiple DGs determined by the proposed algorithm improves the voltage stability of system much more effectively.

Supercapacitors are energy storage devices with high electrical power densities and long spanlife. Therefore, supercapacitor-based energy storage systems have been employed for a variety of applications. The modelling and simulation of SCs have been of great interest to this objective. This paper presents an electrical schema and mathematical modelling of three models of supercapacitors. The first is the RC model, the second is the two-branch model and the third is the multi-branch model. The objective of this modelling is to choose the best model that can respect the same behaviour of the experimental model. These models are compared with an experimental model. This comparison prove that the response voltage of the multi-branch model correctly describes the behaviour of the experimental model of Belhachemi. The disadvantage of this model is the slow simulation duration in MATLAB/Simulink. The RC model represented the faster model in terms of simulation. The choice of 15 branches in parallel in multi-branch models gives good results and correctly describes the reel model. The automatic charge and discharge voltage of SCs reduce by reducing the charge current.

This paper proposes an adaptive droop based virtual slack (ADVS) control for multiple distributed generations (DGs) to improve voltage stability of a practical DC distribution system. Although there have been many researches for optimal sizes of multiple DGs, their solutions are valid only in the particular operating point. Additionally, a previous study proposed a voltage control based optimal operation method, its performance depends on measurement accuracy in practice. The proposed ADVS control operates the system based on the current sensitivities between the DGs and loads, so that it can regulate the system voltages without a large computational effort. This is effective even if measurements are noisy and biased. All DGs contribute to voltage regulation by current control even though they do not directly control voltages. As an additional effect, they effectively share the load. To verify the proposed method, the DC system is modeled based on the real distribution system of the Do-gok area in Seoul, Korea. Then, the Levenberg-Marquardt algorithm determines its operation point. The proposed method is verified based on the electromagnetic transient (EMT) simulation with random loads.

A research was conducted to determine the ac conductivity and dielectric responses of zinc phthalocyanine (ZnPc) thin films as a function of frequency and temperature using the vacuum evaporation method. It was observed that the conductivity reduced with the rise in temperature and that the Debye type of polarization existed in the films.

A heterojunction solar cell design composed of poly(3-hexylthiophene) (P3HT) and intercalated indium-tin-oxide (ITO) donor-acceptor system is explored for the first time. Substantial change in band edge of ITO is noticed after intercalation. Structural and surface morphological studies are reported. Due to tuned band gap of ITO, an increase in short circuit current from 0.0012 to 0.46 mA/cm2, fill factor from 0.39 to 0.51, and power conversion efficiency from 0.000 367 to 0.3% is obtained for heterojunction solar cell when compared to P3HT alone. This novel, room temperature design approach would be of great scientific interest in current solid-state solar cell scenario. [ABSTRACT FROM AUTHOR]

Time series models based on an artificial neural network (ANN) and support vector machine (SVM) were designed to predict the temporal variation of the upper and lower freshwater-saltwater interface level (FSL) at a groundwater observatory on Jeju Island, South Korea. Input variables included past measurement data of tide level (T), rainfall (R), groundwater level (G) and interface level (F). The T-R-G-F type ANN and SVM models were selected as the best performance model for the direct prediction of the upper and lower FSL, respectively. The recursive prediction ability of the T-R-G type SVM model was best for both upper and lower FSL. The average values of the performance criteria and the analysis of error ratio of recursive prediction to direct prediction (RP-DP ratio) show that the SVM-based time series model of the FSL prediction is more accurate and stable than the ANN at the study site. [ABSTRACT FROM AUTHOR]

The article describes the synthesis and characterization of a conjugated polymer with alternating isoindigo (II) and naphthalene (NT) units (PII-NT) and a small molecule. Topics covered include the factors limiting the development of p-type polymers for organic solar cells, the possible reason behind the low molecular weight of PII-NT and the measurement of the oxidation and reduction onset potentials of PII-NT.

The article discusses a study on the synthesis and characterization of phenothiazine-isoindigo copolymers for photovoltaic applications. Topics covered include the chemical structure of phenothiazine II (PPT-II) copolymers confirmed via its proton nuclear magnetic resonance (NMR) spectrum, the polydispersion and solubility of PPT-II, and the onsets of oxidation and reduction potential of PPT-II.

A series of new symmetric tetrathiafulvalene derivatives were designed and synthesized. These compounds have highly ordered lamello-columnar liquid crystalline phases in addition to their crystalline phases at lower temperatures. Charge transport in the organic thin-film-transistor devices is faster in the liquid crystalline phase by two orders of magnitude than that in the crystalline phase due to the decreased defects. [ABSTRACT FROM AUTHOR]

Increase in power consumption can cause serious stability problems in electric power systems if there are no on- going or impending construction projects of new power plants or transmission lines. Additionally, such increase can result in large power losses of the system. In costly and environmentally effective manner to avoid constructing the new infrastructures such as power plants, transmission lines, etc., the distributed generation (DG) has been paid great attention so far as a potential solution for these problems. The beneficial effects of DG mainly depend on its location and size. Therefore, selection of optimal location and size of the DG is a necessary process to maintain the stability and reliability of existing system effectively before it is connected to a power grid. However, the systematic and cardinal rule for this issue is still an open question. In this paper, a method to determine the optimal locations of multiple DGs is proposed by considering power loss. Also, their optimal sizes are determined by using the Kalman filter algorithm. [ABSTRACT FROM AUTHOR]

In this article we have reported the nano structure formation in ZincPhthalocyanine (ZnPc) thin films coated on glass substrates by thermal evaporation method. The structure of the films was analyzed by an X-ray diffractometer (using Cu Kα radiation with λ = 1.5418 Å). It reveals that the vacuum evaporated ZnPc thin films are having nano particles in its structure. It is confirmed by the scanning electron microscopy (SEM), X-ray diffraction (XRD) and optical absorption studies. The results of the XRD, SEM and AFM studies have been discussed in this article. A metastable α to the stable β-phase transformation has been observed when the films are coated at higher substrate temperatures. The core structure of the ZnPc macrocycle is formed by four isoindole units endowing the molecule with a two-dimensional conjugated π electron system. [ABSTRACT FROM AUTHOR]

We report that high efficiency blue polymer light-emitting diodes obtained by adding a thin interlayer between a poly(3,4-ethylene dioxythiophene)-poly(styrene sulfonic acid) (PEDT:PSS) hole transporter and emissive polymer. The one of OLED hole transporters, N,N'-bis(4-methylphenyl)-N,N'-diphenyl-1,4'-phenylenediamine (PDA), was modified with cross-linkable trichlorosilyl groups and used for the interlayer material. The devices, with configuration of indium tin oxide (ITO)/PEDT:PSS (65 nm)/interlayer (10-20 nm)/emitting polymer layer (70 nm)/BaF2 (2 nm)/Ca (50 nm)/Al (300 nm), were fabricated by spin coating and thermal evaporation. In this device structure, the cross-linkable interlayer is more adherent and mechanically robust as well as impervious to spin coating of next emitting polymer layer and shows a better stability. In addition, the devices with the interlayer exhibit a higher luminescence and current efficiency than those in devices without the interlayer. [ABSTRACT FROM AUTHOR]

In this paper, we describe a new approach for fabrication of high efficient polymeric light-emitting diodes (PLEDs). In the device configuration of ITO/HTL/EBL/EML/BaF 2 :Ca:Al (ITO: indium tin oxide, HTL: hole transport layer, EBL: electron blocking layer, EML: emitting layer), EBL contains cross-linkable moieties in order to make the layer which is insoluble to layering of an additional emitting polymer. The devices with EBL exhibit strong blue emissions and higher efficiency values than those in devices without EBL. The synthesis, characterization, device fabrication, and electroluminescence (EL) properties of devices will be presented. [ABSTRACT FROM AUTHOR]

Study region: The Songjiho coastal lagoon, located in the Gangwon Province, Korea, has a unique geographic setting surrounded by mountains and the ocean. Its distinctive hydrological configuration consists of two smaller sub-lakes interconnected by a narrow channel. Its hydrodynamics is significantly influenced by an inlet and outlet, which are connected to the inflowing stream and tidal current, respectively. Study focus: We examined the effect of pumping on the spatiotemporal dynamics of the fresh-saline water mixing zone (FSM) to gain insights into the underlying response mechanisms by pumping. We conducted a series of systematic pumping tests at three fully screened wells installed in the lagoon-aquifer. We monitored electrical conductivity and water temperature with depth. Additionally, we collected and analyzed comprehensive sets of relevant hydrogeological field data. New hydrological insights for the region: Our investigation has pinpointed the dispersion resulting from the concentration gradient as a primary governing factor of the behavior of the FSM. Furthermore, the results also underscore the crucial relationship between the permeability of the layers and the mixing process, which is affected by the inherent heterogeneities in the geological structure and hydraulic properties of the aquifer. In light of these findings, we propose a conceptual model that comprehensively represents the spatiotemporal variations of the coastal lagoon-aquifer system. The model will serve as a valuable tool for understanding and predicting the complex interplay of factors that influence the behavior of this complex hydrogeological system.

For the first time, highly efficient, corrosion resistant, and chemically stable rock barnacle carapace-derived chitosan (CH) supported non-noble nano aluminum-tungsten bimetallic-alloy based bifunctional electrocatalyst has been developed for waste pork lard-derived glycerol (WPLG) electrolysis in NaOH medium, ascertaining sustainability. An indirect bio-glycerol oxidation mechanism has been well executed to produce industrially important glyceric acid (6.65 mmol cm−2 h−1, 79.92% selectivity, ∼90% FE%) and formic acid (1.67 mmol cm−2 h−1, 18.07% selectivity) at the anode at ambient temperature and pressure. However, among the various prepared unsupported and supported nanocomposite electrocatalysts, the optimal aluminum-tungsten alloy anchored over chitosan renovated carbon nanotube (Al:W = 1:2 mole/mole) electrode rendered the lowest onset potential (0.95 V vs. RHE) and highest peak current density (119.61 mA cm−2) for the WPLG electrooxidation at the anode, coupled with H2 generation (28.55 ml cm−2 h−1) in the cathode in a membrane free electrochemical-cell by replacing sluggish oxygen evolution reaction. Comparing the electricity consumption for H2 production from various prepared catalysts, the supported optimum catalyst consumed 9.3% less energy (only 2.9 kWh/m3 H2) and 67.91% less corrosion current density as compared to the unsupported Al/W electrocatalyst, exhibiting excellent stability and durability (>200 cycles) due to good alloy-support interaction. Finally, a comprehensive life cycle impact assessment and life cycle costing of the overall WPLG electrolysis demonstrated 12.29% lower global warming potential (GWP100) while saving 11.03% of the cost compared to commercial glycerol electrolysis, which could procreate an effective approach toward inexpensive catalyst fabrication and its subsequent application for sustainable glycerol electrolysis.

A new primary frequency controller in power grids undergoing massive wind power penetration is the focus of this paper. The inescapable problem in a largely wind penetrated power grid is to ensure the maintenance of its frequency in the nominal band prescribed by the power system operator (PSO). However, with the massive arrival of wind farms with conventional control schemes, the operation of maintaining and restoring the frequency to the regulatory regimes remains very complicated. In order to overcome the above problem, this paper proposes a new strategy for primary frequency control in power grids using model predictive control (MPC) for a multi-cluster doubly-fed induction generator (DFIG)-based wind farm (WF), with a main objective of reducing the frequency nadir (FN), eliminating the second frequency dip (SFD), and providing the optimal support during wind speed variations. In this approach, a rolling prediction and optimization control strategy is developed based on the dynamic power system model to ideally predict the additional power to be provided. Moreover, in order to avoid second frequency dips, the wind turbines (WTs) are not allocated to extract additional power from the grid during the frequency event, the rotor speeds are not recovered to the maximum power point tracking (MPPT) operating points during the primary frequency control. The performance of the proposed controller was evaluated using a two-zone electrical system in MATLAB/Simulink®. The obtained results disclose that the frequency nadir is enhanced with more than 6.1% compared to the conventional schemes. In addition, the frequency response settling time has been improved with more than 10.51 s.

Groundwater levels on the monitoring wells on Jeju Island were monitored, which were caused by the M 5.4 earthquake in Gyeongju City area, South Korea, at 11:32:54 on 12 September 2016 (UTC time) and the Kumamoto earthquake, Kyushu, Japan, at 16:25:06 on 16 April 2016 (UTC time). The groundwater levels changed after 2–3 min by the generation of the Gyeongju and Kumamoto earthquakes and exhibited spikes or oscillations depending on the magnitude of the earthquakes. The groundwater level change caused by the Gyeongju earthquake (M 5.4) was mostly larger than that caused by the Kumamoto earthquake (M 5.4). The reason is explained by that the energy of the Kumamoto earthquake with high attenuation could not be effectively transmitted to Jeju Island since the earthquake took place in low Q region whereas a higher energy of the Gyeongju earthquake with low attenuation arrived on Jeju Island because the earthquake occurred in the south-eastern part of the Korean peninsula belonging to the high Q crust. Besides the seismic energy from the Kumamoto earthquake was scattered and reflected on the Tsushima-Goto fault zone between Kyushu and the Korean peninsula, with a strike in the ENE-WSW direction that elongates to the east of Jeju Island.

In the present study, we designed time series models for predicting groundwater level fluctuations using an artificial neural network (ANN) and a support vector machine (SVM). To estimate the model sensitivity to the range of data set for the model building, numerical tests were conducted using hourly measured groundwater level data at a coastal aquifer of Jeju Island in South Korea. The model performance of the two models is similar and acceptable when the range of input variable lies within the data set for the model building. However, when the range of input variables is beyond it, both the models showed abnormal prediction results: an oscillation for the ANN model and a constant value for SVM. The result of the numerical tests indicates that it is necessary to obtain various types of input and output variables and assign them to the model building process for the success of design time series models of groundwater level prediction.