Annual Peak Load Research Articles

A hybrid data mining driven algorithm for long term electric peak load and energy demand forecasting

Load forecasting is one of the main required studies for power system expansion planning and operation. In order to capture the nonlinear and complex pattern in yearly peak load and energy demand data, a hybrid long term forecasting method based on data mining technique and Time Series is proposed. First, a forecasting algorithm based on the Support Vector Regression (SVR) method is developed. The parameters of the SVR technique along with the dimension of input samples are optimized using a Particle Swarm Optimization (PSO) method. Secondly, in order to minimize the forecasting error, a hybrid forecasting method is presented for long term yearly electric peak load and total electric energy demand. The proposed hybrid method acts based on the combination of Auto-Regressive Integrated Moving Average (ARIMA), Artificial Neural Network (ANN) and the proposed Support Vector Regression technique. The parameters of the ARIMA method are determined based on the autocorrelation and partial autocorrelation of the original and differenced time series. The proposed hybrid forecasting method prioritizes each forecasting method based on the resulted error over the existing data. The hybrid forecasting method is used to forecast the yearly peak load and total energy demand of Iran National Electric Energy System.

Comparing individual and coordinated demand response with dynamic and static power grid tariffs

This paper investigates cost-optimal operation of flexible electricity assets with a capacity-based power grid tariff involving power subscription. The purpose of this research is to identify the characteristics of a subscribed capacity-based tariff that promotes efficient network development through demand response. Using historical load data, we compare two consumers with flexible assets being billed by their individual load versus their combined and coordinated loads in a two-stage stochastic program. The frequency of adjusting the subscribed capacity level (weekly versus annually) influences the effectiveness of the tariff in terms of reducing loads that dimension the grid. The results show that weekly subscription on average provides 5−6% cost savings, while annual subscription on average provides 3% cost savings. A combined annual peak load reduction of 15% occurs when the combined subscription level is adjusted weekly. We also find that when the subscription level is adjusted weekly, the load reduction is cost efficient even when capacity is not scarce, which ought to be avoided. Depending on where a bottleneck in the grid is located, the price signal should be based on the combined load of several consumers rather than individual loads if combined peak load shaving is to be cost-optimal.

Predicting the demand flexibility of wet appliances at national level: The case of France

The manual shift of residential wet appliances (dishwashers and washing machines) in French households that have subscribed to a Time of Use (ToU) tariff represents a flexibility potential available already today and at zero-cost. In this paper we quantify, for the first time, this potential by building a national representative synthetic population for France and using a novel bottom-up model. This includes a stochastic occupant behaviour model for predicting the use of the wet appliances and an overlying agent-based model for predicting their manual shifting under a ToU tariff system. The two models are calibrated and validated using empirical data collected over two recent monitoring campaigns, each carried out in more than 100 French households. By using the new model we observe that a 100% penetration rate of the ToU tariff together with an improved (but practically feasible) households’ flexibility could lead, between 19:00 and 20:00, to a mean hourly load reduction of 0.48 GW, which represents around 0.5% of the typical French annual peak load. This flexibility potential, which is today only partially exploited, could be achieved with zero additional costs for new infrastructures, by encouraging eco practices and/or making the ToU tariff more economically attractive for the households.

Forecasting Long-Term Electricity Demand Time Series Using Artificial Neural Networks

Accurate forecast of electricity load are increasingly important. We present a method to forecast long-term weather-dependent hourly electricity load using artificial neural networks. The fully connected dense artificial neural networks with 5 hidden layers and 1,024 hidden nodes per layer are trained using historic data from 2006 to 2015. Input parameters comprise calendrical information, annual peak loads and weather data. The results are benchmarked against the method to forecast electric loads used in the current mid-term adequacy forecasts published by the European Network of Transmission System Operators (entso-e). For validation year 2016, our approach shows a mean absolute percentage error of 2.8%, whereas the common approach as used by entso-e shows an average error of 4.8% using peak load scaling. Further, we conduct forecasts for Germany, Sweden, Spain, and France for scenario year 2025 and assess parameter variations. Our approach can serve to increase prediction accuracy of future electricity loads.

Impact of emerging technologies on the electricity load profile of residential areas

Technological progress and the ambition to de-carbonise the energy sector are changing the way electricity is generated and used. In the residential sector, more efficient electric devices, electrification of mobility and heating systems, as well as on-site PV and battery systems are the main technologies driving this change. These changes will impact the characteristics of the electric demand. A comprehensive understanding of these changes is essential to efficiently design the energy system of tomorrow.The target of this research is to explore the possible changes of the electric load profiles with the current technological development, both at an individual household level and at an aggregate level. An inner-city area of 1550 houses is simulated with a time resolution of 1 minute using a stochastic bottom-up approach. Three scenarios considering refurbishment of buildings, efficiency increase in devices and emerging technologies are simulated. Differences in the load profiles are then analysed.Results show that future load profiles will change significantly. A strong seasonality will appear due to the increased penetration of PV, μ-CHP and heat pumps. Intra-day fluctuations will also increase, mainly due to the introduction of electric vehicles and PV. Annual peak load will rise significantly mainly due to the electric back-up heaters of air-source heat pumps. Increased electrification increases annual electricity demand, but results show that this will be partly compensated by better insulated buildings, more energy efficient household devices, and on-site production from PV-Battery systems and μ-CHP.

A Comparison of Households’ Energy Balance in Residential Smart Grid Pilots in the Netherlands

This paper presents an analysis that frames the impact of various smart grid technologies, with an objective to provide a transparent framework for residential smart grid demonstration projects based on predefined and clearly formulated key performance indicators. The analysis inspects measured energy data of 217 households from three smart grid pilot projects in the Netherlands and a public dataset with smart meter data from 70 households as a reference. The datasets were evaluated for one year and compared to provide insights on technologies and other differences based on seven key performance indicators, giving a comprehensive overview: monthly electricity consumption (100–600 kWh) and production (4–200 kWh); annually imported (3.1–4.5 MWh) and exported (0.2–1 MWh) electricity; residual load; peak of imported (4.8–6.8 kW) and exported (0.3–2.2 kW) electricity; import simultaneity (20–70.5%); feed in simultaneity (75–89%); self-sufficiency (18–20%); and self-consumption (50–70%). It was found that the electrification of heating systems in buildings by using heat pumps leads to an increase of annual electricity consumption and peak loads of approximately 30% compared to the average Dutch households without heat pumps. Moreover, these peaks have a high degree of simultaneity. To increase both the self-sufficiency and self-consumption of households, further investigations will be required to optimize smart grid systems.

Reliability, economic and environmental analysis of a microgrid system in the presence of renewable energy resources

The application of renewable energy resources in a power system has received significant attention owing to the environmental impacts and fluctuations of fossil fuel prices. Consequently, renewable energy resources have become important sources to generate power at the commercial level due to their various benefits, coupled with the government incentives and public supports. This research work is focused on the evaluation of the reliability, economic and environmental benefits of renewable energy resources in a microgrid system. The lifecycle analysis of a microgrid system that consists of the photovoltaic, wind turbine generator, electric storage system and diesel generator is implemented in this study to test their commercial prospects in rural communities that have no access to electricity due to economic and technical constraints. The objective of this research work is to minimize the cost of energy, lifecycle cost, the annual cost of load loss and lifecycle greenhouse gas emission cost as well as to improve the overall benefit of green technologies in the proposed microgrid system. This objective is achieved by utilizing the basic probability concept to obtain the reliability performance indicators such as expected energy not served, loss of load expectation and loss of load probability, in addition to utilizing an fmincon optimization tool in the MATLAB environment to investigate the environmental and economic effects of renewable energy resources in a power system. The suitability of the model is tested on six case studies by using the same load profile, wind speed and irradiation of the site and diesel generator power capacity. The market factors such as interest rate and price of diesel fuel as well as forced outage rate, annual peak load variation and distributed generation penetration level are utilized to study their impacts on the operation of a microgrid system. The results obtained in this study demonstrate the optimal feasibility of renewable energy resources in a microgrid system. This indicates that it offers a significant reduction in the values of lifecycle cost, cost of energy, greenhouse gas emission cost and the annual cost of load loss when compared with case study 1. This research work shows that the utilization of green technologies in a microgrid system optimizes the reliability, cost savings, lifecycle cost and environmental impact. The technique adopted in the study can serve as a reference for rural electrification projects and solve socioeconomic problems that are associated with power outages.

Setting the Smart Solar Standard: Collaborations Between Hawaiian Electric and the National Renewable Energy Laboratory

Driving through many neighborhoods of Hawai'i, it is hard to miss the nearly ubiquitous rooftop solar photovoltaic (PV) systems that have popped up during the past eight years or so. Relatively high electricity costs associated with island grids, coupled with various incentives, have made it cost-effective to install solar over the last eight years, as evidenced by the PV-deployment chart in Figure 1. On the most populous island, O'ahu, the PV nameplate acgenerating capacity of 502 MW totals nearly half of the annual peak load for the entire island, which is 1.1 GW . Of that 502 MW of PVs, 54% is from private rooftop solar-nearly 50,000 residences or roughly one of every three single family homes. But Hawaiian Electric, the local utility, has no way to monitor or control the PV generation, even for most nonresidential systems. This means that on sunny days, up to approximately half of the PV generation is outside of the utility's control. This poses many challenges for utility planners and operators-challenges that Hawaiian Electric has been working diligently to address, along with various partners, notably the U.S. Department of Energy (DOE), the National Renewable Energy Laboratory (NREL), and its Energy System's Integration Facility (ESIF). This article describes how Hawaiian Electric has worked with engineers in NREL's Power Systems Engineering Center to improve the way its grid operates with very high levels of distributed PVs, largely by changing the way the PV inverters are operated.

Quantification of the uncertainties produced in the construction process of a building through simulation tools: A case study

This work considers the uncertainties in building energy performance caused by deviations introduced in the construction process. It proposes an approach based on sensitivity analysis by simulation, which allows for the quantification of the impact of these deviations isolated from other possible ones and independently from one another. This information could be useful for practitioners, helping them to identify which deviations with respect to design could be critical regarding energy savings and giving criteria to optimize the construction process. The approach is applied for a case study in which the effects of such deviations on the heating and cooling energy demand of a building located at the Solar Platform of Almeria (Spain) are quantified. The results support that possible deviation from the design specifications during the construction process can affect the building annual energy consumption and peak loads considerably in the Mediterranean area. According to the work reported in this paper, the studied parameters sorted in descending order of influence on the annual load are: infiltration rate, insulation thickness, insulation class, cement mortar thickness and frame-wall gap filled in with cement.

A spatial electricity market model for the power system: The Kazakhstan case study

Kazakhstan envisions a transition towards a green economy in the next decades, which poses an immense challenge as the country's economy and energy system depends heavily on hydrocarbon resources. Here, it lacks inclusive and transparent tools assessing technical, economic, and environmental implications resulting from changes in its electricity system. We present such a tool: our comprehensive techno-economic unit-commitment model determines the hourly least-cost generation dispatch, based on publicly available data on the technical and economic characteristics of the system. It accounts for particularities of the Kazakh electricity system by representing combined heat and power, and endogenously determining line losses. Model results examine two typical weeks: winter (annual peak load) and summer (hour of lowest annual load) presenting regionally and temporally disaggregated results for power generation, line utilization, and nodal prices. In an application to market design, the paper compares nodal and zonal pricing as two possible pricing schemes in Kazakhstan for the envisioned strengthening of the day-ahead market. The model analyze the current Kazakh electricity system and can be easily expanded to assess the sector's future development. Possible applications include investment in generation and transmission infrastructure, policy assessment for renewables integration, carbon pricing, emission reduction, and questions of market design.

Long-term forecasting of annual peak load considering effects of demand-side programs

The main purpose of this research paper is to investigate the long-term effects of the proposed demand-side program, and its impact on annual peak load forecasting important for strategic network planning. The program comprises a particular set of demand-side measures aimed at reducing the annual peak load. The paper also presents the program simulations for the case study of the Electricity Distribution Company of Belgrade (EDB). According to the methodology used, the first step is to determine the available controllable load of the distribution utility/area under consideration. The controllable load is presumed constant over the analyzed time horizon, and the smart grid (SG) infrastructure available. The saturation of positive effects during intense program application is also taken into account. Technical and economic input data are taken from the real projects. The conducted calculations indicate that demand-side programs can bring about the same results as the energy storage in the grids with a strong impact of distributed generation from variable renewable sources (V-RES). In conclusion, the proposed demand-side program is a good alternative to building new power facilities, which can postpone investment costs for a considerable period of time.

A Spatial Electricity Market Model for the Power System of Kazakhstan

Kazakhstan envisions a transition towards a green economy in the next decades which poses an immense challenge as the country heavily depends on (hydro-)carbon resources, for both its economy and its energy system. In this context, there is a lack of comprehensive and transparent planning tools to assess possible sustainable development pathways in regard to their technical, economic, and environmental implications. We present such a tool with a comprehensive techno-economic model of the Kazakh electricity system which determines the hourly least-cost generation dispatch based on publicly available data on the technical and economic characteristics of power plants and the transmission infrastructure. This modeling framework accounts for the particularities of the Kazakh electricity system: i) it has a detailed representation of combined heat and power, and ii) line losses are endogenously determined using a linear approximation. Model results are examined for a typical winter week (with annual peak load) and a typical summer week (with the hour of lowest annual load) presenting regionally and temporally disaggregated results for power generation, line utilization, and nodal prices. In an application to market design, the paper compares nodal and zonal pricing as two possible pricing schemes in Kazakhstan for the envisioned strengthening of the day-ahead market. In general, the model can be readily used to analyze the least-cost dispatch of the current Kazakh electricity system and can be easily expanded to assess the sector's development. Among others, possible applications include investment in transmission lines and in the aging power plant fleet, scenarios and policy assessment for emission reduction, and questions of market liberalization and market design.

Long-Term Peak Load Estimate and Forecast: A Case Study of Uyo Transmission Substation, Akwa Ibom State, Nigeria

This paper presents an approach for long-term estimation and forecasting of electric peak load. A 10-year peak load forecast is performed on Uyo transmission substation in Akwa Ibom State, Nigeria. The peak loads of the past ten years (from 2006 to 2010) are used as input data used to develop the model for forecasting the peak load demand in Uyo metropolis. Particularly, Multiple Linear Regression (MLR) method is used to model the annual peak load. The explanatory variables, namely, temperature, population and gross domestic product are used in the analysis. The peak load model parameters are estimated using only the data of the year 2006 to the year 2012, which accounts for 70% of the entire dataset for training and 30% (that is, 2013 to 2015) of the data are used for cross validation. The results show that with respect to the training dataset the prediction model has Mean Absolute Percentage Error (MAPE) of 0.00613%, Mean Absolute Deviation (MAD) of 0.277743 and Coefficient of Determination (R2) value of 0.99184 which shows that about 99.184% of the peak load are explained by the explanatory variables used in the prediction. Furthermore, with respect to the validation dataset (2013 to 2015) the prediction model has RMSE of 1.038042 and percentage error of less that 2% which shows that the proposed peak-load-demand model can effectively predict the peak load demand for Uyo.

DGM (1, 1) model optimized by MVO (multi-verse optimizer) for annual peak load forecasting

A large number of renewable energies and uncertain power load accessing electric power system make the power load forecasting more complicated and face more new challenges. This paper presents a hybrid annual peak load forecasting model [namely MVO-DGM (1, 1)], which employs the latest optimization algorithm MVO (multi-verse optimizer) to determine two parameters of DGM (1, 1) model, and then uses the optimized DGM (1, 1) model to forecast annual peak load. The annual peak load of Shandong province in China from 2005 to 2014 is selected as the empirical example, and the analysis results demonstrate that the MVO algorithm for parameters’ determination of DGM (1, 1) model has significant superiority over the least square estimation method, particle swarm optimization and fruit fly optimization algorithm in terms of annual peak load forecasting. In addition, the proposed MVO-DGM (1, 1) peak load forecasting model has more excellent forecasting performance than other non-optimized forecasting techniques and other optimized DGM (1, 1) models due to its ascended local optima avoidance and better convergence speed. The hybrid MVO-DGM (1, 1) model proposed in this paper is feasible and effective in annual peak load forecasting, which can improve the forecasting accuracy.

Modeling and simulation on the performance of a novel double shape-stabilized phase change materials wallboard

The structure of a new wall system was that of a three-layer sandwich-type panel with external/internal layer consisting of shape-stabilized phase change material (SSPCM) wallboard and middle layer consisting of conventional concrete. The external layer was active in hot seasons and the internal layer was active in cold seasons. An office room integrated with the novel double SSPCMs wallboard with a splitting air-conditioner was used for modeling and simulation test platform. Studies were conducted to investigate the effects of the novel double SSPCMs and different parameters on the energy and thermal performance in typical climate area with hot summer and cold winter (Wuhan city, China). Test results showed that the recommended thicknesses of external and internal SSPCM wallboard were 30–60mm. With the recommended thickness, the annual energy savings for cooling were 3.4–3.9%. The peak load reductions for cooling were 3.1–3.8%. The annual energy savings for heating were 14.8–18.8%. The peak load reductions for heating were 8.6–11.3%. The external layer was effective on reducing annual energy demand and peak load of cooling in hot seasons. The internal layer was effective on reducing the indoor temperature fluctuation in cold seasons.

A probabilistic load shedding concept considering highly volatile local generation

Automatic load shedding is the ultimate countermeasure against imbalance in a power system and can effectively help preventing large blackouts. Taking into account a high penetration of renewable energy sources (RES) in the distribution grid, a clear distinction between load and generation at the PCC becomes increasingly more difficult. For that reason an adaptation of frequency relay parameters and their locations of installation are necessary. In Europe this is rest on a multi-step plan based on values such as the yearly peak load. In this paper a novel probabilistic method for automatic load shedding is presented that uses the average values instead of peak values for load shedding. Its applicability is verified by a dynamic power system model that was developed to compare the classical and novel probabilistic load shedding principle. The method is verified using data from a German TSO.

Energy Storage Needs in Interconnected Systems Using the Example of Germany and Austria

The share of renewable energies on the net electricity consumption is rising steadily. Especially intermittent, non-controllable sources like wind and sun are gaining importance. With an installed amount of non-controllable power that exceeds the yearly peak load, situations can occur with a surplus of energy in electricity supply systems. This surplus will rise strongly with the share of fluctuating renewable energies on the net electricity consumption. A lot of studies and experts come to the conclusion that energy storage will be needed to handle these surpluses. The questions that still have to be answered are when and how. Especially the German electricity system will have very high storage needs because of a very strong and fast development of wind and solar power. There are a lot of technologies and approaches to overcome this problem. However, Pumped Hydro Energy Storage (PHES) systems are up to date the most efficient and economic bulk energy storage technology. On the one side there are no high potentials in natural sites for the installation of PHES schemes in Germany. On the other side the Austrian PHES system has still a very high potential. Up so far, the prospects are used just to a small amount. Especially the seasonal Hydro Energy Storage (HES) still holds a very high potential. In this paper the combination of the Austrian and the German energy supply system will be investigated with respect to the development of renewable energies. The overall energy storage needs are evaluated for each country as well as in the combined system, taking also into account the development and the influence of the transmission system capacity between the two countries.

Mid-term forecasting of urban electricity load to isolate air-conditioning impact

Abstract Demand Side Management (DSM) is often one of the most cost-effective approaches toward energy conservation and efficient electricity infrastructure utilization. Identifying total air-conditioning load for assessing targeted interventions is a difficult task given the transient thermal response of buildings, the coupled interaction of multiple sub-systems and the high correlation of demand with weather and other perturbations. An hourly regression-model of electricity consumption was developed for the city of Abu Dhabi, UAE, using measured hourly substation-level data. The fit is exceptionally good, even in a prediction context: Root Mean Squared Error (RMSE) equivalent to 1.54% of the annual peak load and Mean Absolute Percentage Error (MAPE) of 2.01% for the training period (full-year 2010), RMSE of 1.84% and MAPE of 2.64% for the testing period (first-half of 2011). The regression-model was combined with information from Abu Dhabi Urban Planning Council and the National Central Cooling Company (Tabreed) to derive an accurate estimate of the urban cooling load, the main driver of electricity consumption in the region. It was determined that, although only 30% of the annual load is directly weather dependent, air-conditioning explains no less than 57% of the total annual electricity load and 75% of the peak summer demand.

Levenberg-Marquardt Recurrent Networks for Long-Term Electricity Peak Load Forecasting

Increasing electricity demand in Java-Madura-Bali, Indonesia, must be addressed appropriately to avoid blackout by determining accurate peak load forecasting. Econometric approach may not be sufficient to handle this problem due to limitation in modelling nonlinear interaction of factors involved. To overcome this problem, Elman and Jordan Recurrent Neural Network based on Levenberg-Marquardt learning algorithm is proposed to forecast annual peak load of Java-Madura-Bali interconnection for 20092011. Actual historical regional data which consists of economic, electricity statistic and weather during 1995-2008 are applied as inputs. The networks structure is firstly justified using true historical dat a of 1995-2005 to forecast peak load of 2006-2008. Afterwards, peak load forecasting of 2009-2011 is conducted subsequently using actual historical data of 1995-2008. Overall, the proposed networks shown better performance compared to that obtained by Levenberg-Marquardt-Feedforward network, Double-log Multiple Regression, and with projection by PLN for 2006-2010.

Time-of-Use Tariff Program in Saudi Arabia: Design, Implementation, and Evaluation

In Saudi Arabia, electricity consumption increases substantially during the summer months with residential, commercial and industrial customers accounting for the majority of the high summer load. Although the daily peak period occurs between 1 PM and 5 PM, loads exceeding 95% of the annual peak load only last for 0.3% of the total hours of the year; therefore, they are very expensive to serve. Large industrial and commercial customers are participating with the Saudi Electricity Company (SEC) in a Time-of-Use (TOU) tariff program. The objective of the program is to reduce SEC peak load growth and the associated additional generation and network capacity, and to provide the participating customers with financial and other benefits. SEC started its TOU tariff program in June 2006 in its Central Operating Area (COA) with 26 industrial and 9 commercial customers. In 2007, the program was expanded to its Eastern and Western Operating Areas to include an additional 89 industrial and 20 commercial customers.This paper presents an overall evaluation of the TOU program implementation in the COA of Saudi Arabia through an analysis of the results of field visits to selected customers and technical data obtained from SEC. The paper includes the TOU program description, energy and demand impacts, economic benefits to SEC, and customer bill reductions. Success cases and problems encountered are addressed, and recommendations to strengthen the program are made.