Yogesh Kumar Bichpuriya

Non-parametric probability density forecast of an hourly peak load during a month

The Load Serving Entity (LSE) requires, for its power procurement portfolio management, accurate peak load forecast in medium term (upto six months ahead). A complete description of the random variable, i.e., load, is provided by probability density function. Hence, we consider the problem of forecasting probability density function of hourly peak load during a month. First, we propose a non-parametric model based on the Alternating Conditional Expectation (ACE) to obtain point forecast. Then, by considering multiple scenarios of the weather variables i.e., temperature-humidity tuples, we obtain probability density forecast of the peak load. Out-of-sample testing is used to demonstrate efficacy of the proposed approach.

Combination approaches for short term load forecasting

Short term load forecasting for day ahead operations is an important task of an electric distribution company. Forecasting errors directly impact the economics of the distribution company in a market scenario. Many categories of methods like, expert system, artificial neural network and time series analysis, have been developed for short term load forecasting. We compare and contrast these methods on a utility data set. It is seen that no method can be said to be consistently better or worse than the other. Therefore, this paper explores the idea of development of a combination forecast from the three individual forecasts. The combination forecast is shown to have better expected performance than any one of the individual forecasts. Different methods of combining forecasts like, in proportion to probability of success, combining with weights calculated from variance minimization, on the basis of eigenvector of covariance matrix and median forecast are considered. Results of an urban electric distribution companys data is used to demonstrate the efficacy of the approaches.

Synergism of MW and Mvar Losses for Contingency Ranking Using Fuzzy Approach

Security assessment of a power system is very important function to detect any violation in the system and to ensure secure operation of the system in deregulated environment. Contingency ranking is the process of indexing the possible contingencies of system on the basis of their severity. The contingencies of higher ranking are further analyzed for detailed analysis. The contingencies based on the real power flow and the voltage deviations should be dealt simultaneously for better ranking. The paper presents fuzzy approach to combine ranking based on MW and Mvar losses and a new rank list is prepared. The proposed approach is tested on IEEE-30 bus test system.

Day ahead load forecasting models for holidays in Indian context

Accurate Short Term Load Forecasting (STLF) is critical for efficient functioning of electricity distribution company. High forecast error may result in non-optimal system operations and financial risk in short term power markets. Load profiles on holidays are very different from that on normal days. In India, holidays can be categorized as Sunday, Public holidays (e.g., Independence day, Republic day etc.) and Festival days (e.g., Diwali, Eid, Christmas etc.). Apart from these holidays, there are a few regional holidays e.g., Ganesh Chaturthi and Maharashtra day in the state of Maharashtra. Sunday is a repeated holiday having weekly frequency while other holidays come once in a year. Also, some of these holidays follows lunar calender and some follows Gregorian calender. Each holiday, excluding Sunday and Public holidays, has different characteristics in terms of activities, lighting load and the number of peoples celebrating the holiday. In such a scenario, predicting the accurate load profile for the holidays is a difficult task. This paper proposes two different models for Sunday and other holidays. Sunday model is used for forecasting load profile on Sundays and Holiday model is used for all other public holidays and festival days. The proposed models have been tested on load data of an urban distribution utilities and the results are illustrated.

Design and testing of heuristic for market splitting in electric power exchange

This paper proposes Taguchis method for design of experiments to test market splitting algorithm of an electric power exchange. Such a mechanism is required because price-volume pairs of a trader are usually bound by confidentiality requirement and hence not available as test data. We also propose a novel and simple market splitting heuristic considering congestion management problem in a meshed network with incomplete information of the network. Simulation results are reported in the paper.

Provision for reactive power market in the reactive power control groups

Implementation of electricity market is facing many challenges in developing countries like India. The market for real power is gaining more interest of investors while the market for reactive power has less incentive due to its technical limitation. But reactive power is of importance for system security and improving the line transmission capacity. The system operator is therefore willing to buy reactive power from different sources to ensure the necessary potential at each node of the system. The paper discusses an approach for market for reactive power in reactive power control groups using the concept of electrical distances. The approach does not involve any optimal power flow (OPF) algorithm for avoiding the complexity and non-convergence for large systems but discusses a simple and economical market provision for reactive power. The approach is illustrated with the transmission system of Uttar Pradesh Power Corporation Limited (UPPCL).