Amir Hisham Hashim

Backward reduction application for minimizing wind power scenarios in stochastic programming

In order to make informed decisions in the presence of uncertainties, risk management problems of power utilities may be modelled by multistage stochastic programs. These programs use a set of scenarios (or plausible realizations) and corresponding probabilities to model the multivariate random data process, e.g. electrical load, stream flows to hydro units, generation output of intermittent renewable sources as well as fuel and electricity prices. The number of scenarios needed to accurately represent the uncertainty involved is generally large, thus due to computational complexity and time limitation, scenario reduction techniques are often utilized. The paper proposes a new application for recursive backward scenario reduction to establish possible next-day scenarios for wind power generation at Mersing Johor, Malaysia. The algorithm determines a subset from the initial scenario set and assigns new probabilities to the preserved scenarios. The output is intended to assist generation scheduling of power system employing intermittent type renewable sources.

Moving holidays' effects on the Malaysian peak daily load

Malaysias yearly steady growth in electricity consumption as a result of fast development in various sectors of the Malaysian economy have increased the need to have a more robust, reliable and accurate load forecasting for short —, medium-, or long-term. A reliable method for short term load forecasting is crucial to any decision maker in a power utility company. Many studies have been made to improve the forecasting accuracy using various methods. The forecasting errors for the holiday seasons are known to be higher than those for weekends. This paper aims to determine which model would be a better model to estimate the holiday effects and therefore give a better forecasting accuracy for the peak daily load in Malaysia. Some of the holiday effects in Malaysia are from Eid ul-Fitr, Christmas, Independence Day and Chinese New Year. The seasonal ARIMA (SARIMA) and Dynamic Regression (DR) or Transfer function modelling are considered. Furthermore, the final selection of the models depends on the Mean Absolute Percentage Error (MAPE) and others such as the sample autocorrelation function (ACF), the sample partial autocorrelation function (PACF) and a bias-corrected version of the Akaikes information criterion (AICC) statistic. The Dynamic Regression (DR) model recorded 2.22% as the lowest MAPE value for the 2004 New Years Eve and 2.39% for the seven days ahead forecasting. And therefore, DR model is the most appropriate model to be considered for forecasting any public holidays in Malaysia.

Cash-flow analysis of a wind turbine operator

The intermittent characteristics of wind energy seriously affect the reliability and output power of wind turbine generator (WTG). The paper outlines a method to evaluate the distribution of WTG operators daily cash-flow by developing an algorithm based on Monte-Carlo technique. Two-parameter Weibull type probability density function (PDF) is used to model wind profile at two locations. WTGs output is obtained by using power curve while also considering the availability factor derived from the Forced Outage Rate (FOR). The daily cash-flows were obtained by calculating the difference between the revenues from sales of power at feed-in-tariff and the cost of energy. The paper contributes towards a better understanding of wind profiles effect on the expected energy production and the sensitivity of profit distribution to changes in feed-in tariff.

A combined probabilistic and deterministic method for an improved capacity outage probability table synthesis using monte carlo methods

An electrical power system is typically operated at least cost given some technical and reliability constraints. To that end, many utilities use deterministic methods such as n−1 criterion in order to ensure reliability. This method though being used worldwide and for a really long time can be expensive as contingencies occur rarely. In order to evaluate power systems, it must be noted that contingencies are stochastic in nature. Since reliability of supply and economic cost must be balanced, probabilistic analysis seems to be one of the tools which can be applied. In the case of generator dispatch, the most recognized probability tool is the Capacity Outage Probability Table (COPT). However, the synthesis of the COPT using the Recursive Method as proposed by Billinton and Allan can be cumbersome. This paper presents an integration of probabilistic and deterministic methods for outage analysis. It shows an alternative method to develop the COPT using a pseudo-deterministic method combined with a Monte Carlo method. Thus, making the process of developing a COPT simpler and can reduce computation time.

Determination of Auto-Recloser Location Using Cost Analysis in the Sabah Electricity Distribution Network

Sabah electricity (SESB) operates a vertically integrated electricity utility in East Malaysia. It is currently embarking on an initiative to improve the reliability of its 11 kV distribution network which is prone to trippings due to vegetation growth along their long spur lines, Apart from rentice management, another proposed solution is to install auto-reclosers (AR) at appropriate locations in their network. This work deals in developing a methodology to determine the most economic location of ARs and their numbers to be installed. In order to do that, data in the form of customer types, load levels and network topology were collected from site visits at SESB. The study then used an estimated value of lost load (VoLL) to quantify the financial losses that customers suffer given a loss of supply. These losses would then be summed with the cost of ARs for a particular AR placement scenario to form an objective function. The best placement and numbers of AR can be decided by determining the least cost option from the said objective function.

Application of Lightning Performance Analysis for a Tropical Climate Country

A tropical climate country experience high lightning related activities. This can be represented in terms of the keraunic level which lies within the region of 100 to 150 thunderstorm days per year. For a utility company, the high keraunic level would contribute to high possibility of power interruptions. These outages would ultimately lead to loss of revenue and reduced network reliability. These lightning related interruptions may be in terms of direct strikes to the line. Apart from that a consequent back flashover may result from an indirect lightning strike near the vicinity of the line. The aggregate of the direct and indirect lightning strikes would contribute to the overall lightning performance of the line. This paper outlines the application of a lightning performance analysis tool based upon the IEEE Std 1410-2004 conditioned for a tropical climate country. Results and analysis indicates factors that would affect the lightning performance of the line. The contributing factors found from these analyses would be dependent upon the environment at which the line is located, e.g. presence of tall objects which ultimately acts as a lightning shield would improve the lightning performance whereas if the lines are located in an open field, the probability of lightning strikes would be high. Based upon these factors, a list of possible mitigating actions is presented in improving the lightning performance of the line.

Profit-based optimal generation scheduling of a microgrid

The paper proposes a formulation for profit-based optimal generation scheduling by a microgrid (µgrid). Current methods normally assume either islanded operation, or utility-grid connected µgrid but lacking in market participation elements. The paper addresses this gap whereby the formulated objective function allows for autonomous decision-making to determine the hour by hour optimal dispatch of generators subject to system constraints including market parameters. The distributed generations in the modelled µgrid consist of wind turbines, microturbines and photovoltaic arrays while the system inputs are based on tropical conditions. A case study on the difference between grid-connected and islanded operation is presented. The results demonstrate the efficiency of using genetic algorithm to solve the optimization problem.

Adaptive protection for voltage instability mitigation scenario

Modern power system network are usually operated to its maximum capabilities. This in turn would expose the system towards possible voltage instability condition. Coupled with peak load condition where lines can be loaded to its limit, probable scenarios relating to voltage instability could lead to tripping of overload lines which would then lead to possible cascading tripping; ultimately lead towards voltage collapse condition. However, possible mitigating action is possible which could reduce the line loading and ultimately push the system back towards a voltage stable condition. This paper demonstrates a possibility of this occurring utilizing Fast Voltage Stability Index (FVSI) approaches to quickly assess the network voltage stability. Results show that these mitigating action does improve the stability of the system hence lead to a possible Wide Area Protection application which is synonymously linked to a smart grid system.

Malaysian peak daily load forecasting

Time series analysis has been applied intensively and sophisticatedly to model and forecast many problems in the biological, physical and environmental phenomena of interest. This fact accounts for the basic engineering problem in forecasting the daily peak system load to use time series analysis. ARMA and Regression with ARMA errors models are among the times series models considered. ANFIS, a hybrid model from neural network is also discussed as for comparison purposes. The main interest of the forecasts consists of three days up to seven days ahead predictions for daily data. The objective is to find an appropriate model for forecasting the Malaysian peak daily demand of electricity. The pure autoregressive model with an order 2 or AR (2) has the minimum AIC statistic value compared with other ARMA models. AR (2) model recorded the value for the mean absolute percentage error (MAPE) as 1.27 % for the prediction of 3 days ahead from Jan 1 to 3, 2005. Besides AR(2) model, Regression model with ARMA errors and ANFIS were found to be among the best forecasting models for weekdays with MAPE value from 0.1 % to 3 %.

Stratification & Sampling of Electricity Supply Customers for Outage Costs Survey

In todays demanding business environment, determining the value of lost load (VoLL) is important in order for utilities to make the right decision when it embarks on any form of asset expansion or assesses risks in power system operation. The VoLL is the aggregated or average value of outage costs across the whole range of customers in the electricity supply industry (ESI). This paper covers the initial process in determining the VoLL from customer survey data. It discusses the data stratification method used to categorize customers according to a set of criteria and then the normalization process of customer survey data. Finally, a demonstration of the stratification method on domestic customers is shown and evaluated.