Household Load Profiles Research Articles

MODELING OF PROSUMER LOAD PROFILES BASED ON BEHAVIORAL APPROACH

The article analyzes various approaches to the modeling of daily electricity load schedules and proposes a methodology for improving the "bottom-up" behavioral model of the load of household consumers. The method is based on averaging generated random load schedules for a given type of household on a particular day of the year. Averaging is carried out by season and length of daylight. Next, working days and weekends are distinguished in each interval. Despite some differences, this work does not separate holidays, Saturdays, and Sundays to avoid unnecessary details. The selected parameters for dividing the averaging intervals allow you to obtain a relatively compact set of model data and, at the same time, preserve the features of the load on different days of the year and hours of the day. Household load profiles with varying steps of day distribution and the main difference of such distribution are considered. Models that more accurately convey the sporadic nature of prosumer consumption compared to typical load schedules are built for two types of households. A detailed analysis of the averaged load graphs was carried out in comparison with the typical graph and the generated load graphs. Averaged load schedules are suitable for modeling the operating modes and control algorithms of the prosumer energy generation and storage system.

Energy Household Forecast with ANN for Demand Response and Demand Side Management

This paper presents a short term load forecasting with artificial neural networks. Despite the great imprevisibility, it is possible to forecast the electricity consumption of a household with some accuracy, similarly to that the electricity utilities can do to an agglomerate of households. Nowadays, in an existing electric grid, it is important to understand and forecast household daily or hourly consumption with a reliable model for electric energy consumption and load profile. Demand response programs required this information to adequate the profile of energy load diagram to generation. In the short term load forecasting model, artificial neural networks were used, with a consumption records database. The results show that the artificial neural networks approach provides a reliable model for forecasting household electric energy consumption and load profile. To do so and using smart devices such as cyber-physical systems monitoring, gathering and computing in real time a database with weekdays and weekend, can improve forecasts results for the next hours, a strong tool for Demand Response and Demand Side Management.

Short Term Load Forecasting for Smart Grids Using Apache Spark and a Modified Transformer Model

Smart grid is an advanced electrical grid that enables more efficient distribution of electricity. It counters many of the problems presented by renewable energy sources such as variability in production through techniques like load forecasting and dynamic pricing. Smart grid generates massive amounts of data through smart meters, this data is used to forecast future load to adjust distribution. To process all this data, big data analysis is necessary. Most existing schemes use Apache Hadoop for big data processing and various techniques for load forecasting that include methods based on statistical theory, machine learning and deep learning. This paper proposes using Apache Spark for big data analysis and a modified version of the transformer model for forecasting load profiles of households. The modified transformer model has been tested against several state-of-the-art machine learning models. The proposed scheme was tested against several baseline and state-of-the-art machine learning models and evaluated in terms of the RMSE, MAE, MedAE and R2 scores. The obtained results show that the proposed model has better performance in terms of RMSE and R2 which are the preferred metrics when evaluating a regression model on data with a large number of outliers.

Effects of Household Battery Systems on LV Residential Feeder Voltage Management

With the advancements of the battery energy storage systems (BESSs), reduction of their manufacturing costs and government subsidies, the BESS uptake is likely to increase rapidly in power distribution networks. This paper investigates the effects of residential BESSs on low-voltage (LV) networks using the actual household load profiles equipped with BESS and solar-photovoltaic (PV) systems. The electricity consumption data collected via smart meters (2200 households with PV/BESS, 1950 households with PV only and 1000 households without a PV or a BESS) at different solar-PV penetration levels and network types are used to simulate real network operating scenarios. A real LV distribution network in Australia is analysed in DIgSILENT PowerFactory under different scenarios, such as, customers with and without a solar-PV/BESS, with a solar-PV but without a BESS, and without a solar-PV, by using both the power flow and quasi-dynamic simulation studies under balanced and the unbalanced network loading conditions. According to the study, customers experience large voltage excursions from solar-PV power exports, which could be resolved by the household BESS, provided that the BESS charging is coordinated with the solar-PV production. Moreover, quasi-dynamic simulation shows that the BESS could reduce the violation of the over-voltage limit during the solar peak hours (midday) by lowering the worst-case feeder voltage by 3%. Finally, extreme-event (high solar PV generation scenario) simulation shows that the implementation of the BESS controller to facilitate charging BESS during afternoon solar-PV export may reduce the negative grid impact and will assist to avoid network upgrades.

Intelligent wild geese algorithm with deep learning driven short term load forecasting for sustainable energy management in microgrids

Energy management in power grids becomes essential to reduce the cost for the consumer and improve the power supply reliability. The microgrid is a vital part of the smart grid and it requires intelligent power management approach for effective functioning. Presently, delivering demand load and sustaining energy are two major challenges that exist in the power system. To resolve these problems, short-term load forecasting (STLF) models have been presented as an effective management and energy supply mode in power systems. The recently developed deep learning (DL) and machine learning (ML) models can be employed for accurate STLF in microgrids. In this view, this study presents an intelligent wild geese algorithm with deep learning driven short term load forecasting (IWGADL-STLF) model for sustainable energy management in microgrids. The proposed IWGADL-STLF model intends to accurately and rapidly predict the STLF in the microgrids. To accomplish this, the IWGADL-STLF model uses attention based Bi-directional long short term memory (ABiLSTM) model which involves the input parameters as formation of household and commercial load profiles with commercial load profile of the microgrid as output. The proposed IWGADL-STLF model identifies the behavioural patterns of parameters and models the behaviour in short time period for effective prediction process. Since hyperparameters play a vital role in the DL models, in this study, WGA is applied as a hyperparameter optimizer of the ABiLSTM model. The IWGADL-STLF approach has shown effective results with low MAE, MAPE, and R2 values. A comprehensive experimental analysis reported the enhanced performance of the presented model over the other existing approaches under several aspects.

Dataset on electrical single-family house and heat pump load profiles in Germany

This paper describes a dataset of residential electricity household and heat pump load profiles, measured in 38 single-family houses in Northern Germany. We provide data per household of apparent, active and reactive power (W), voltage (V), current (A) and the power factor (no unit) in 10 seconds to 60 minutes temporal resolution from May 2018 to the end of 2020. We validated the dataset both in itself, comparing different measurements that should produce the same results, and externally to standard load profiles and found no major inconsistencies. We identified an average consumption per single-family house with 2.38 inhabitants of 2829 kWh for the household and an additional 4993 kWh for the heat pump. The dataset can support the understanding of patterns in electrical load curves and can help to estimate the additional load on distribution networks induced by heat pumps.

MultiCycleNet: Multiple Cycles Self-Boosted Neural Network for Short-term Electric Household Load Forecasting

• Proposed a deep learning framework for household load forecasting. • Used correlative data in multiple cycles to explore power consumption patterns. • Used contextual information around the data to improve predicted accuracy. • The proposed framework outperforms the benchmarks with fluctuating load profiles. Household load forecasting plays an important role in future grid planning and operation. However, compared with aggregated load forecasting, household load forecasting faces the challenge of the uncertainty of prolific load profiles. This paper presents a novel multiple cycles self-boosted neural network (MultiCycleNet) framework for household load forecasting, which aims to solve the uncertainty problem of household load profiles through the correlation analysis of electricity consumption patterns in multiple cycles. The basic idea of the proposed framework is that the predictor can learn customers’ power consumption patterns better by learning the features and contextual information of relevant load profiles in multiple historical cycles. We use two real-life datasets: 1. the household load consumption dataset from Low Carbon London project led by United Kingdom (UK) Power Networks and 2. the UK Domestic Appliance-Level Electricity (UK-DALE) dataset to evaluate the effectiveness of the proposed framework. Compared with the state-of-the-art methods, experimental results show that the proposed framework is effective and outperforms the state-of-the-art methods by 19.83%, 10.46%, 11.14% and 9.02% in terms of mean squared error, root mean squared error, mean absolute error and mean absolute percent error, respectively.

Classification and characterization of intra-day load curves of PV and non-PV households using interpretable feature extraction and feature-based clustering

Load pattern categorization plays a significant role in enhancing the understanding of demand characteristics of different cohorts of energy customers on a distribution network. For a distribution network with embedded photovoltaic (PV) systems, it is also desirable to enumerate these premises since unreported PV installations can have a detrimental impact on voltage control, frequency regulation, and incur reverse power flows. In this paper, a holistic smart meter data analytics approach is proposed for classifying and characterizing the intra-day load curves of PV and non-PV households. Unlike existing studies based on raw time-series data, a series of interpretable and discriminating global and peak-period features are first developed to extract the physical information of load patterns. A two-step feature-based clustering is then applied to classify the load profiles of PV and non-PV households. A post-clustering quantitative compositional analysis is developed to characterize the energy use variability and compositional changes for each household. The proposed PV household identification method allows network owners to identify the existence of PV installations in a particular area for voltage control and ancillary service provision. In the longer term, effective load pattern categorization and post-clustering characterization can help better understand demand characteristics of different cohorts of customers, network utilization, and network-level changes in load growth.

MODELLING THE DEMAND BEHAVIOUR OF ACTIVE CUSTOMERS AND THE IMPACT OF PRICE INCENTIVE-BASED CONTROL STRATEGIES

Current developments such as the electrification of the mobility and heat sector, an increasing spread of photovoltaics and energy storage systems as well as new pricing models are changing the load profiles of distribution grid customers. This paper presents an optimization model to derive the load profiles of future active households and compares different incentive-based control strategies to meet the challenges of these developments from a grid perspective. For this purpose, a three-phase load flow analysis is performed, which also considers power quality measures such as unbalance and static voltage limits.

Optimal Sequential Distribution Planning for Low-Voltage Network With Electric Vehicle Loads

There has been a growing presence of electric vehicles in many countries including Thailand, where many forms of incentives have been provided to build integrated infrastructure, and to encourage drivers to switch to electric vehicles (EVs). Because the immediate entry of EVs unavoidably can alter household load profiles, reinforcement on the existing system based on traditional planning may not be sufficient and can introduce over or under capital and operating expenditure over the time horizon. Therefore, if distribution systems are unreadily prepared for such an uptake, three obvious problems can be expected: 1) voltage regulation, 2) overloads of the distribution feeders and the distribution transformers, and 3) high energy loss. In this paper, an activity-based, time-sequential Monte Carlo Simulation algorithm was comprehensively developed for uncontrollable and smart charging, given annually updated information of EV locations and number of EVs, their energy consumption, hourly average vehicle speed, number of daily trips, travel distance per trip, size of EV batteries, time to arrive home and time to leave home. Minimizing the annual sum of investment and operating costs over a planning period could then be sequentially solved by a Particle Swarm Optimization (PSO) algorithm. The results from a practical 122-bus, 24 kV/400 V distribution system with different scenarios of uncontrollable and smart charging show that the sequential optimization embedded with deterministic decision can help improve customer voltage profile, keep feeder and transformer loading within acceptable operating limits and offer significant cost savings from energy loss. As far as a large number of low-voltage networks, and the associated large sum of cost savings are concerned, the proposed planning framework is practical to be applied and expected to be served as a new guideline for future implementation in Thailand.

Electrical vehicle grid integration for demand response in distribution networks using reinforcement learning

Most utilities across the world already have demand response (DR) programs in place to incentivise consumers to reduce or shift their electricity consumption from peak periods to off-peak hours usually in response to financial incentives. With the increasing electrification of vehicles, emerging technologies such as vehicle-to-grid (V2G) and vehicle-to-home (V2H) have the potential to offer a broad range of benefits and services to achieve more effective management of electricity demand. In this way, electric vehicles (EV) become distributed energy storage resources and can conceivably, in conjunction with other electricity storage solutions, contribute to DR and provide additional capacity to the grid when needed. Here, an effective DR approach for V2G and V2H energy management using Reinforcement Learning (RL) is proposed. Q-learning, an RL strategy based on a reward mechanism, is used to make optimal decisions to charge or delay the charging of the EV battery pack and/or dispatch the stored electricity back to the grid without compromising the driving needs. Simulations are presented to demonstrate how the proposed DR strategy can effectively manage the charging/discharging schedule of the EV battery and how V2H and V2G can contribute to smooth the household load profile, minimise electricity bills and maximise revenue.

Analysis and Prediction of Electromobility and Energy Supply by the Example of Stuttgart

This paper seeks to identify bottlenecks in the energy grid supply regarding different market penetration of battery electric vehicles in Stuttgart, Germany. First, medium-term forecasts of electric and hybrid vehicles and the corresponding charging infrastructure are issued from 2017 to 2030, resulting in a share of 27% electric vehicles by 2030 in the Stuttgart region. Next, interactions between electric vehicles and the local energy system in Stuttgart were examined, comparing different development scenarios in the mobility sector. Further, a travel demand model was used to generate charging profiles of electric vehicles under consideration of mobility patterns. The charging demand was combined with standard household load profiles and a load flow analysis of the peak hour was carried out for a quarter comprising 349 households. The simulation shows that a higher charging capacity can lead to a lower transformer utilization, as charging and household peak load may fall temporally apart. Finally, it was examined whether the existing infrastructure is suitable to meet future demand focusing on the transformer reserve capacity. Overall, the need for action is limited; only 10% of the approximately 560 sub-grids were identified as potential weak points.

A Cross-Country Model for End-Use Specific Aggregated Household Load Profiles

End-use specific residential electricity load profiles are of interest for energy system modelling that requires future load curves or demand-side management. We present a model that is applicable across countries to predict consumption on a regional and national scale, using openly available data. The model uses neural networks (NNs) to correlate measured consumption from one country (United Kingdom) with weather data and daily profiles of a mix of human activity and device specific power profiles. We then use region-specific weather data and time-use surveys as input for the trained NNs to predict unscaled electric load profiles. The total power profile consists of the end-use household load profiles scaled with real consumption. We compare the model’s results with measured and independently simulated profiles of various European countries. The NNs achieve a mean absolute error compared with the average load of 6.5 to 33% for the test set. For Germany, the standard deviation between the simulation, the standard load profile H0, and measurements from the University of Applied Sciences Berlin is 26.5%. Our approach reduces the amount of input data required compared with existing models for modelling region-specific electricity load profiles considering end-uses and seasonality based on weather parameters. Hourly load profiles for 29 European countries based on four historical weather years are distributed under an open license.

Solar panels and smart thermostats: The power duo of the residential sector?

Rising peak demand is a major cause for high emissions from the electricity sector. In this study, we investigate how different combinations of distributed energy technologies affect peak grid load, energy consumption from the grid, and emissions in the residential sector under time-varying prices. To do so, we develop an optimization framework in which households with varied amalgamations of distributed energy technologies minimize electricity costs, amortized capital, and operational costs over a year, with marginally increasing penalties for deviating from room temperature set-points. The four technologies we consider are: solar panels, lithium-ion batteries, ice cold thermal energy storage, and smart thermostats. The study also incorporates a one-parameter thermal model of the home, so that the discomfort penalties can apply to the room temperature rather than the total appliance load. Based on empirical energy consumption profiles and solar generation data from 25 homes in Austin, we find that residential customers would keep overall annual expenditure and environmental footprint low by investing in solar panels and smart thermostats. The capital costs of both storage systems are still too high to make them economically profitable investments for typical residential customers. Additionally, the Value of Solar policy disincentivizes solar customer investment in storage systems. The study also shows that, while the energetic effect of the two storage systems can be favorable or detrimental depending upon the pricing structure and the household load profile, lithium-ion batteries are the main instruments to avoid high demand charges. Thus, we recommend that, as an effective peak load control mechanism, electric utilities should offer significant rebates to encourage residential customer investment in storage systems in addition to subjecting them to demand charges.

Deep Learning-Assisted Short-Term Load Forecasting for Sustainable Management of Energy in Microgrid

Nowadays, supplying demand load and maintaining sustainable energy are important issues that have created many challenges in power systems. In these types of problems, short-term load forecasting has been proposed as one of the management and energy supply modes in power systems. In this paper, after reviewing various load forecasting techniques, a deep learning method called bidirectional long short-term memory (Bi-LSTM) is presented for short-term load forecasting in a microgrid. By collecting relevant features available in the input data at the training stage, it is shown that the proposed procedure enjoys important properties, such as its great ability to process time series data. A microgrid in rural Sub-Saharan Africa, including household and commercial loads, was selected as the case study. The parameters affecting the formation of household and commercial load profiles are considered as input variables, and the total household and commercial load profiles of the microgrid are considered as the target. The Bi-LSTM network is trained by input variables to forecast the microgrid load on an hourly basis by recognizing the consumption pattern. Various performance evaluation indicators such as the correlation coefficient (R), mean squared error (MSE), and root mean squared error (RMSE) are utilized to analyze the forecast results. In addition, in a comparative approach, the performance of the proposed method is compared and evaluated with other methods used in similar studies. The results presented for the training phase show an accuracy of R = 99.81% for the Bi-LSTM network. The test and load forecasting stage are performed by the Bi-STLM network, with an accuracy of R = 99.34% and forecasting errors of MSE = 0.1042 and RMSE = 0.3243. The results confirm the high performance of the proposed Bi-LSTM technique, with a high correlation coefficient when compared to other methods used for short-term load forecasting.

Assessing the temporal load resolution effect on the photovoltaic energy flows and self-consumption

The electrical load temporal resolution on the simulated photovoltaic system electrical energy flows and self-consumption are investigated in order to determine the analysis uncertainty. The renewable energy system under investigation includes a grid-connected photovoltaic system with a capacity of 0.55, 1.1, 1.65, 2.2 and 2.75 kWp. Results show that the load temporal resolution influences the energy flow in the modelled system even for a single household load and a simple photovoltaic system without batteries. It has been found that the household load profile characteristic impact the required data temporal resolution. For high oscillating power, the temporal resolution for the energy consumption of 1 min should be considered, however, if power fluctuations are not high and rapid, a sampling rate of 15 min or even 60 min in most cases is appropriate. The most sensitive parameter for analysis is self-consumption which decreases at high load resolution and increases at low load resolution and may affect instantaneous analysis uncertainty substantially. On average the yearly estimated absolute and relative errors for energy self-consumption in the analysed configuration is 67.33 kWh and 13% at 1.65 kWp photovoltaic system and they decrease with system capacity increase. Finally, two mathematical models are developed and validated for estimation of the relative and absolute errors generated due to the load resolution.

Estimating hourly lighting load profiles of rural households in East Africa applying a data-driven characterization of occupant behavior and lighting devices ownership

To design energy access solutions for rural households in developing countries it is important to have an accurate estimation of what their electricity consumption is. Studies reveal that they mainly use electricity to meet their lighting needs, as they cannot afford high power-consuming appliances. However, the scarce data availability and modeling complexity are a challenge to compute correctly the load profiles without collecting data on-site. This paper presents a methodology that computes the hourly lighting load profiles of rural households in East Africa requiring a small amount of publicly available input data. Combining data from household surveys, climate, and satellite imagery, the methodology applies machine learning for determining occupant behavior patterns, and lamps ownership for indoor and outdoor usage. For this, an average prediction accuracy of 80% is reached. After applying lighting requirement functions, load profiles are generated and then validated using measured data from 13 households in Kenya. Results show that the methodology is able to compute the load profiles with an average normalized root mean squared error of 0.7%, which is less compared to existing simulation approaches using on-site data. To demonstrate a broad application, the monthly lighting consumption is computed and projected geospatially for households in Kenya.

Two-Stage Energy Management Strategy of EV and PV Integrated Smart Home to Minimize Electricity Cost and Flatten Power Load Profile

The efficient use of the incorporation of photovoltaic generation (PV) and an electric vehicle (EV) with the home energy management system (HEMS) can play a significant role in improving grid stability in the residential area and bringing economic benefit to the homeowner. Therefore, this paper presents an energy management strategy in a smart home that integrates an electric vehicle with/without PV generation. The proposed strategy seeks to reduce the household electricity costs and flatten the load curve based on time-of-use pricing, time-varying household power demand, PV generation profile, and EV parameters (arrival and departure times, minimum and maximum limit of the state-of-charge, and initial state-of-charge). The proposed control strategy is divided into two stages: Stage A, which operates in three operating modes according to the unavailability of PV power generation, and Stage B, which operates in five operating modes according to the availability of PV generation. In this study, the proposed strategy enables controlling the amount of energy absorbed by the EV from the grid and/or PV and the amount of energy injected from the EV to the load to ensure that the household electricity costs are minimized, and the household power load profile is flattened. The findings show that both household electricity costs reduction and flattening of the power load profile are achieved. Moreover, the corresponding simulation results exhibit that the proposed strategy for the smart home with EV and PV provides better results than the smart home with EV and without PV in terms of electricity costs reduction and power load profile flattening.

Italian Household Load Profiles: A Monitoring Campaign

The increasing share of renewable energy sources on the supply side, as well as the so-called electrification pathways on the demand side, has led to peculiar challenges for electrical systems: Indeed, the increasing load demand has to be balanced from the supply-side viewpoint. In particular, the residential sector contributes to nearly 26% of the final energy consumption in Europe, suggesting that a further understanding of households’ consumptions and load profiles is needed to support an energy transition. In this context, this paper contributes to the existing discussion by proposing a sociodemographic analysis of Italian households’ load profiles using a smart metering experimental study, while also considering the households in energy poverty conditions. For the sake of generality, results are presented based on a previously proposed household segmentation of the Italian residential sector. The outcomes point out three prominent peaks on load profiles for all the identified clusters, with a notable distinction in intensity. Where children are present, a higher load profile is noted, reaching a maximum value of 600 W of absorbed power between 19:30 and 22:30. Conversely, households in an energy poverty condition show a relatively regular load profile, ranging from a minimum of 110 W of absorbed power in baseload conditions at night time to a maximum of 280 W in the evening hours. The findings in this paper are in agreement with existing research in the field, and accordingly the study proposes a better focus on domestic appliances and sociodemographic parameters.

Influence of Data Sampling Frequency on Household Consumption Load Profile Features: A Case Study in Spain.

Smart meter (SM) deployment in the residential context provides a vast amount of data of high granularity at the individual household level. In this context, the choice of temporal resolution for describing household load profile features has a crucial impact on the results of any action or assessment. This study presents a methodology that makes two new contributions. Firstly, it proposes periodograms along with autocorrelation and partial autocorrelation analyses and an empirical distribution-based statistical analysis, which are able to describe household consumption profile features with greater accuracy. Secondly, it proposes a framework for data collection in households at a high sampling frequency. This methodology is able to analyze the influence of data granularity on the description of household consumption profile features. Its effectiveness was confirmed in a case study of four households in Spain. The results indicate that high-resolution data should be used to consider the full range of consumption load fluctuations. Nonetheless, the accuracy of these features was found to largely depend on the load profile analyzed. Indeed, in some households, accurate descriptions were obtained with coarse-grained data. In any case, an intermediate data-resolution of 5 s showed feature characterization closer to those of 0.5 s.