It is important to forecast the water flow into the dam in order to operate gates safely. The water flow in the river is a complex natural phenomenon, therefore it is difficult to make the high performance forecasting system by conventional mathematical methods.
We have developed a high performance water flow forecasting system into the dam using neural networks and fuzzy theory. To improve forecasting performance and to decide the structure of the forecasting system, we propose two new methods. One is to decide the structure of neural networks and fuzzy parameters by the relation to the data. The other is to use two neural networks to improve forecasting performance when the river condition changes. One neural network is for normal river condition, and the other one is for flood condition. Forecasts obtained with two neural networks are combined by fuzzy theory which infers river condition. The forecasting results show the effectiveness of the proposed methods.

View full abstract

In recent years, cascade use of waste energy, so called “heat cascading”, in the industrial sector is expected to be one of the promising measures for improving the energy efficiency of regional energy system. This paper discusses the feasibility of heat cascading among major energy intensive industries from technical and economical aspects.
The simulation results revealed that (i) the introduction of heat cascading contributes to the drastic energy saving, (ii) the energy transmission by electricity achieves the drastic energy saving when heat transmission loss is large, (iii) heat cascading promotes the efficiency of regional energy system including industrial, residential and commercial sectors by optimizing the energy flow in the heat combinat, (iv) high cost of heat transmission brings about a trade-off relationship between energy-saving and economical potentials, (v) the economical incentive such as reduction of cost of heat transmission and/or introduction of carbon tax relaxes the trade-off relationship.

View full abstract

Distributed generator (DG) has several advantages over the large-scale power plant and is expected to complement the power system. In this paper, we focused on the locational flexibility and short lead time as advantages of DG, and examined the influence of uncertainty of demand growth on the optimal locational configuration of DGs in the transmission network and on the amount of installed DG.
The results indicate that the total capacity of installed DGs is greater under uncertain demand growth than under certain growth. And it is also indicated that under the uncertainty, the installed DG capacity is distributed generally in proportion to the growth of the local loads. The latter finding suggests that the expected load growth at each point could be a useful indicator for screening candidates of locational configurations of DGs.

View full abstract

Transient stability analysis is more difficult and of great importance because the power system becomes larger and larger. Therefore, accurate stability to contingency faults is demanded in power system planning and operation. The energy function method, which is one of the methods for transient stability assessment, can search the stability limit by one simulation. This method has an advantage of obtaining the result at high speed, but there are faults that enough accuracy is not obtained if this method is applied to general power systems. However, when generators are assumed as the classical model, the energy function method is available to first swing stability. Therefore, there is a problem of improving accuracy in the analysis for power systems that include controllers such as AVR and Governor.
In this paper, we consider the problems that occur when the energy function method with the classical model is applied to a more detailed generator model that includes controllers. It is pointed out that the cause of the inaccuracy comes from changes of internal voltages. Because the internal voltage is affected by AVR's action, an analytical technique is proposed for the treatment of the internal voltage. Moreover, it is proposed that the obtained first critical clearing time be revalued to obtain a more accurate critical clearing time.

View full abstract

This paper presents an optimal weekly scheduling method for the complicated hydrothermal systems which involve not only hydraulically coupled multi-chain, multi-reservoir hydro systems, but also pumped-storage plants together with thermal units.
The basic approach of the proposed method adopts an iterative optimization algorithm of scheduling each operation of hydro, thermal and pumped-storage systems sequentially. An efficient aggregation-decomposition algorithm for the multi-reservoir hydro scheduling is developed. That is, to get an efficient scheduling method, hydroelectric plants along a river are aggregated into one equivalent hydro plant and then, this plant is decomposed again to each original plant through a matrix screening method so that constraints satisfied optimal solution is realized. Simulation results show that the proposed method has enough efficiency for practical use in terms of convergence characteristic and computing time.

View full abstract

Two kinds of operational methods (A and B) are proposed to a solar-heat/electricity hot-water supply system. These operational methods are based on the forecasted insolation of next day, i. e., after the insolation of next day is forecasted, the heat storage tank of the system is heated with midnight-time electricity (22 o'clock at previous day_??_8 o'clock at next day). In operational method A, the maximum power P_max is set to the midnight-time electricity. In operational method B, the minimum temperature T_Smin is set to the heat storage tank at 8 o'clock of next day in addition to P_max. The value of P_max and T_Smin, are determined using the simulation with meteorological data from 1991 to 1994. Using meteorological data of 1995, the operational methods A and B are compared with the normal operational method which does not consider the insolation forecast. The results show that the operational method A is economical when the forecast error is lower than 15%, while the operational method B is economical when the forecast error is higher than 15%.

View full abstract

Tests included ground and short circuit faults of a 6.6kV small-scale utility, which is capable of independent operation and consisted of dispersed PV, fuel cell, wind generation and power storage systems. Results show that the main source of short circuit current is either the substation or the dispersed power systems operating in voltage mode. Also, the effective operation of the individual fault detection functions of the dispersed systems was verified.

View full abstract

A frequency dependency of PD signal propagation in GIS was investigated at an underground substation. Injecting dummy pulses into a GIS through its internal electrode, frequency spectrums were measured by a spectrum analyzer at various points of the GIS. It is shown that this measurement is very reproducible and the signal propagation phenomena are independent of its direction. It is also clarified that the PD signal reduction phenomena are deeply related with mode transfer from TEM to TE or TM. The major higher mode seems TE11. Pulse waveforms were also measured with a 65MHz bandpass filter. Their magnitudes are compared with the gain of 65MHz component of the frequency spectrums and good agreement is assured. This implies a possibility of PD location by observing a frequency pattern with several bandpass filters.

View full abstract

Observation of lightning current flowing into earth of various power distribution equipment has been performed by using magnetic pieces in summer of 1996-1997. It has been made on two types (with and without a ground wire) of power distribution lines in Gunma Pref. and Tochigi Pref. of Japan. The authors succeeded in obtaining 1048 surge currents in two years. Using these data, we investigated the relation of the number of installed equipment and the number of surges, frequency distributions of surge currents, and the effect of a ground wire. In addition, we compared the calculated rate of damage on distribution lines with the observed rate. As a result, we have confirmed the effectiveness of the lightning protection method of installing surge protective devices. And we have compared our observed results with another observation results of lightning by LPATS. Out of these data, we have investigated coincidence of observed results and the property of surge currents diverted by surge arresters.

View full abstract

A combination grounding electrode of horizontal mesh and vertical conductors is, in many cases, enclosed with iron fence. When fault current flows into the grounding electrode, a person touching the fence is subjected to the touch voltage. If the touch voltage is hazardous, an additional construction work is needed for reduction of the touch voltage. For this purpose metallic rods may be vertically driven into the earth at each fence corner or at each fence center, and electrically connected to the fence.
In this paper, simulation of the touch voltages is executed by the current sources superposition method and the method of images, for homogeneous earth and two-layer earth. Touch voltages are estimated without and with driven rods. Relations between buried lengths of driven rods and touch voltages are described for various dimensions of grounding electrode, earth resistivity ratios and spacings between the grounding electrode and the fence. It is pointed out that, when the upper layer resistivity is greater than the lower layer resistivity, in some cases the maximum touch voltage arises at other positions than the fence corner.

View full abstract

Based on experiments and calculations, this paper describes the influence of the time constant of the DC starting voltage on the no-voltage time required for successful restarting and the time required to complete the restarting.
(1) The 300kV class DC tension insulator assembly tested in this study was composed of disc-type insulator units used for AC transmission lines. The insulation recovery characteristics of the DC tension insulator assembly after DC arc extinctions of 400 and 4, 000 A were obtained by the switching impulse voltage and simulated DC starting voltage. The results showed that both methods obtained the same minimum values for the required no-voltage time. Thus, we verified the validity of this method to calculate the minimum value of the no-voltage time required for successful restarting based on the insulation recovery characteristics obtained by the switching impulse voltage.
(2) Based on the insulation recovery characteristics obtained by the switching impulse voltage after arc extinction, we calculated the relationships between the required no-voltage time and restarting completion time versus the time constant of the DC starting voltage. As the rising time constant of the DC starting voltage increased, the required no-voltage time decreased and the restarting completion time increased. It was shown that reducing the rising time constant of the DC starting voltage was an effective measure to shorten the restarting completion time after the arc extinction.

View full abstract

This paper deals with flashover characteristics of a disc type insulating spacer contaminated by oil or grease. Three kinds of conductive and nonconductive grease, and two kinds of vacuum pump oil were examined. They were painted either entirely or partially on one side of the spacer surface. The spacer was exposed to a ramped DC voltage or a lightning impulse voltage. The SF₆ gas pressure was 0.4MPa. When the DC voltage was applied, grease with comparatively lower resistivity decreased the flashover voltage to the value lower than that of a clean spacer even in the partially contaminated conditions. The lightning impulse voltage caused the similar effect when the spacer was totally contaminated, but was less influential in the partially contaminated conditions. These results, together with the result from flashover track observation, suggest that conductivity of a contaminator causes the distortion of the potential distribution, which is responsible for the lower flashover strength of a partially contaminated spacer.

View full abstract

A study was conducted on the prediction of transformer noise based on geometrical acoustic. For near field estimation, a revised radiation method called 2-dimensional noise analysis was applied. It is designed to calculate the noise around a transformer assembly including tanks, bushings and coolers. For far field noise estimation such as a substation boundary, a geometrical method called 3-dimensional analysis was applied. Considering the effect of sound reflection and diffraction, it is specially applicable for site with complicated landscape.

View full abstract

UHF interferometers were operated during winter thunderstorm observations in Japan. Tbe main objective is the investigation of lightning channel progressions inside thunderclouds, especially for positive cloud to ground strikes. Since the time resolution of interferometers is high, observation results are used additionally to confirm the upward initiated leader formation by laser firing. Lightning currents measured by Rogowski coils and electric field changes are compared with UHF radiation to confirm leader formation. All of the observations are consistent with the time accuracy of 17 μ seconds and this paper strongly claims the first success of laser triggered lightning experiment in the field.

View full abstract

This paper discusses the thermal behavior at the quench inception of ac superconducting wires in liquid helium, using the highly time-resolved, synchronized optical/electrical measurement of the quench phenomena The optical observation with the high speed video system of 18, 000 fiames/s enabled us to visualize the quench-induced local disturbances (LD) in the vicinity of the superconducting wires at several tens of μs after the quench-onset, followed by the film-boiling state bubbles around the wires. The LD generation mechanism was investigated from the dependence on the strand configuration as well as the heat flux calculation into liquid helium. As the result, the LD can be a bubble cluster in the transient nucleate-boiling state generated by the drastic and transient injection of heat flux due to quench. The LD was also interpreted to be brought about by the activated weak points on the cooling stability in the strands, which was peculiar to ac superconducting stranded wires for large current capacity.

View full abstract

Accurate modeling of a pole-type distribution transformer is very important in assessing the lightning performance on a middle-voltage power line. In this letter, an equivalent circuit of a pole-type transformer for analysis of lightning surge on a middle-voltage is proposed by modal analysis of measured primary-side admittance. The validity of the equivalent circuit is ensured by comparison with experimental results.

View full abstract

Simultaneous observation of lightning current on a tall structure and associated induced voltage waveforms on a test distribution line showed that the observed current waveforms were influenced by reflection and refraction of traveling current waves at both extremities of the structure. A model of a lightning stroke attached to a tall structure is proposed for analysis of induced voltages. The induced voltage waveforms, calculated by using the proposed model, are ensured by comparison with measurement associated with an actual lightning stroke hitting a 200m stack.

View full abstract