research paper on electricity

• Open access
• Published: 09 September 2020

Electricity load forecasting: a systematic review

• Isaac Kofi Nti   ORCID: orcid.org/0000-0001-9257-4295 1 ,
• Moses Teimeh 2 ,
• Owusu Nyarko-Boateng   ORCID: orcid.org/0000-0003-0300-2469 3 &
• Adebayo Felix Adekoya   ORCID: orcid.org/0000-0002-5029-2393 3  

Journal of Electrical Systems and Information Technology volume  7 , Article number:  13 ( 2020 ) Cite this article

24k Accesses

99 Citations

Metrics details

The economic growth of every nation is highly related to its electricity infrastructure, network, and availability since electricity has become the central part of everyday life in this modern world. Hence, the global demand for electricity for residential and commercial purposes has seen an incredible increase. On the other side, electricity prices keep fluctuating over the past years and not mentioning the inadequacy in electricity generation to meet global demand. As a solution to this, numerous studies aimed at estimating future electrical energy demand for residential and commercial purposes to enable electricity generators, distributors, and suppliers to plan effectively ahead and promote energy conservation among the users. Notwithstanding, load forecasting is one of the major problems facing the power industry since the inception of electric power. The current study tried to undertake a systematic and critical review of about seventy-seven (77) relevant previous works reported in academic journals over nine years (2010–2020) in electricity demand forecasting. Specifically, attention was given to the following themes: (i) The forecasting algorithms used and their fitting ability in this field, (ii) the theories and factors affecting electricity consumption and the origin of research work, (iii) the relevant accuracy and error metrics applied in electricity load forecasting, and (iv) the forecasting period. The results revealed that 90% out of the top nine models used in electricity forecasting was artificial intelligence based, with artificial neural network (ANN) representing 28%. In this scope, ANN models were primarily used for short-term electricity forecasting where electrical energy consumption patterns are complicated. Concerning the accuracy metrics used, it was observed that root-mean-square error (RMSE) (38%) was the most used error metric among electricity forecasters, followed by mean absolute percentage error MAPE (35%). The study further revealed that 50% of electricity demand forecasting was based on weather and economic parameters, 8.33% on household lifestyle, 38.33% on historical energy consumption, and 3.33% on stock indices. Finally, we recap the challenges and opportunities for further research in electricity load forecasting locally and globally.

Electricity is the pivot in upholding highly technologically advanced industrialisation in every economy [ 1 , 2 , 3 ]. Almost every activity done in this modern era hinges on electricity. The demand and usage of electric energy increase globally as the years past [ 4 ]; however, the process of generating, transmitting, and distributing electrical energy remains complicated and costly. Hence, effective grid management is an essential role in reducing the cost of energy production and increased in generating the capacity to meet the growing demand in electric energy [ 5 ].

Accordingly, effective grid management involves proper load demand planning, adequate maintenance schedule for generating, transmission and distribution lines, and efficient load distribution through the supply lines. Therefore, an accurate load forecasting will go a long way to maximise the efficiency of the planning process in the power generation industries [ 5 , 6 ]. As a means to improve the accuracy of Electrical Energy Demand (EED) forecasting, several computational and statistical techniques have been applied to enhance forecast models [ 7 ].

EED forecasting techniques can be clustered into three (3), namely correlation, extrapolation, and a combination of both. The Extrapolation techniques (Trend analysis) involve fitting trend curves to primary historical data of electrical energy demand in a way to mirror the growth trend itself [ 7 , 8 ]. Here, the future value of electricity demand is obtained from estimating the trend curve function at the preferred future point. Despite its simplicity, its results are very realistic in some instances [ 8 ].

On the other hand, correlation techniques (End-use and Economic models) involve relating the system load to several economic and demographic factors [ 7 , 8 ]. Thus, the techniques ensure that the analysts capture the association existing between load increase patterns and other measurable factors. However, the disadvantage lies in the forecasting of economic and demographic factors, which is more complicated than the load forecast itself [ 7 , 8 ]. Usually, economic and demographic factors such as population, building permits, heating, employment, ventilation, air conditioning system information, weather data, building structure, and business are used in correlation techniques [ 7 , 8 , 9 ]. Nevertheless, some researchers group EED forecasting models into two, viz. data-driven (artificial intelligence) methods (same as the extrapolation techniques) and engineering methods (same as correlation the techniques) [ 9 ]. All the same, no single method is accepted scientifically superior in all situations.

Also, proper planning and useful applications of electric load forecasting require particular “forecasting intervals,” also referred to as “lead time”. Based on the lead time, load forecasting can be grouped into four (4), namely: very short-term load forecasting (VSTLF), short-term load forecasting (STLF), medium-term load forecasting (MTLF) and long-term load forecasting (LTLF) [ 6 , 7 , 10 ]. The VSTLF is applicable in real-time control, and its predicting period is within minutes to 1 h ahead. The STLF is for making forecasting within 1 h to 7 days or month ahead [ 11 ]. It is usually used for the day-to-day operations of the utility industry, such as scheduling the generation and transmission of electric energy. The MTLF is used for forecasting of fuel purchase, maintenance, utility assessments. Its forecasting period ranges from 1 week to 1 year. While the LTLF is for making forecasting beyond a year to 20 years ahead, it is suitable for forecasting the construction of new generations, strategic planning, and changes in the electric energy supply and delivery system [ 10 ].

Notwithstanding the above-mentioned techniques and approaches available, EED forecasting is seen to be complicated and cannot easily be solved with simple mathematical formulas [ 2 ]. Also, Hong and Fan [ 12 ] pointed out that electric load forecasting has been a primary problem for the electric power industries, since the inception of the electric power. Regardless of the difficulty in electric load forecasting, the optimal and proficient economic set-up of electric power systems has continually occupied a vital position in the electric power industries [ 13 ]. This exercise permits the utility industries to examine the dynamic growth in load demand patterns to facilitate continuity planning for a better and accurate power system expansion. Consequently, inaccurate prediction leads to power shortage, which can lead to “dumsor” and unneeded development in the power system leading to unwanted expenditure [ 7 , 14 ]. Besides, a robust EED forecasting is essential in developing countries having a low rate of electrification to facilitate a way for supporting the active development of the power systems [ 15 ].

Based on the sensitive nature of electricity demand forecasting in the power industries, there is a need for researchers and professionals to identify the challenges and opportunities in this area. Besides, as argued by Moher et al. [ 16 ], systematic reviews are the established reference for generating evidence in any research field for further studies. Our partial search of literature resulted in the following [ 10 , 12 , 17 , 18 , 19 , 20 , 21 ] papers that focused on comprehensive systematic review concerning the methods, models, and several methodologies used in electric load forecasting. Hammad et al. [ 10 ] compared forty-five (45) academic papers on electric load forecasting based on inputs, outputs, time frame, the scale of the project, and value. They revealed that despite the simplicity of regression models, they are mostly useful for long-term load forecasting compared with AI-based models such as ANN, Fuzzy logic, and SVM, which are appropriate for short-term forecasting.

Similarly, Hong and Fan [ 12 ] carried out a tutorial review of probabilistic EED forecasting. The paper focused on EED forecasting methodologies, special techniques, common misunderstandings and evaluation methods. Wang et al. [ 19 ] presented a comprehensive review of factors that affects EED forecasting, such as forecast model, evaluation metric, and input parameters. The paper reported that the commonly used evaluation metrics were the mean absolute error, MAPE, and RMSE. Likewise, Kuster et al. [ 22 ] presented a systematic review of 113 studies in electricity forecasting. The paper examined the timeframe, inputs, outputs, data sample size, scale, error type as criteria for comparing models aimed at identifying which model best suited for a case or scenario.

Also, Zhou et al. [ 17 ] presented a review of electric load classification in the smart-grid environment. The paper focused on the commonly used clustering techniques and well-known evaluation methods for EED forecasting. Another study in [ 21 ] presented a review of short-term EED forecasting based on artificial intelligence. Mele [ 20 ] presented an overview of the primary machine learning techniques used for furcating short-term EED. Gonzalez-Briones et al. [ 18 ] examined the critical machine learning models for EED forecasting using a 1-year dataset of a shoe store. Panda et al. [ 23 ] presented a comprehensive review of LTLF studies examining the various techniques and approaches adopted in LTLF.

The above-discussed works of literature show that two studies [ 20 , 21 ] address a comprehensive review on STLF, [ 23 ] addresses forecasting models based on LTLF. The study in [ 24 ] was entirely dedicated to STLF. Only a fraction (10%) of above systematic review studies included STLF, MTLF and LTLF papers in their review; however, as argued in [ 10 ], the lead time (forecasting interval) is a factor that positively influences the performance of a chosen model for EED forecasting studies. Again, a high percentage of these studies [ 10 , 12 , 17 , 18 , 19 , 20 , 21 , 22 , 24 ] concentrated on the methods (models), input parameter, and timeframe. Nevertheless, Wang et al. [ 19 ] revealed that the primary factors that influence EED forecasting models are property (characteristic) parameters of the building and weather parameters include. Besides, these parameters are territorial dependant and cultural bond. Thus, the weather pattern is not the same world-wide neither do we use the same building architecture and materials globally.

Notwithstanding, a higher percentage of previous systematic review studies overlooked the origin of studies and dataset of EED forecasting paper. Also, only a few studies [ 12 , 17 , 19 ] that examined the evaluation metrics used in EED forecasting models. However, as pointed out in [ 17 ], there is no single validity index that can correctly deal with any dataset and offer better performance always.

Despite all these review studies [ 10 , 12 , 17 , 18 , 19 , 20 , 21 , 22 , 24 ] on electricity load forecasting, a comprehensive systematic review of electricity load forecasting that takes into account all possible factors, such as the forecasting load (commercial, residential and combined), the forecast model (conventional and AI), model evaluation metrics and forecasting type (STLF, MTLF, and LTLF) that influences EED forecast models is still an open gate for research. Hence, to fill in the gap, this study presents an extensive systematic review of state-of-the-art literature based on electrical energy demand forecasting. The current review is classified according to the forecasting load (commercial, residential, and combined), the forecast model (conventional, AI and hybrids), model evaluation metrics, and forecasting type (STLF, MTLF, and LTLF). The Preferred-Reporting Items for Systematic-Review and Meta-Analysis (PRISMA) flow diagram was adopted for this study based on its ability to advance the value and quality of the systematic review as compared with other guidelines [ 16 , 25 , 26 ]. The current study contributes to knowledge as follows:

A comprehensive and detailed assessment of previous state-of-the-art studies on electricity demand forecasting; based on used methods, timeframe, the train and test split of data, error, and accuracy metrics applied to forecast.

We present a concise summary of the useful characteristics of compared techniques in electric load forecasting.

We identified the challenges and opportunities for further studies in electric load forecasting.

The remaining sections of the current paper are structured as follows. “ Methodology ” section presents the methods and materials used in the current study. “ Data collection ” section presents the results and a detailed discussion of the outcomes, and “ Study framework ” section presents the summary of findings and direction for future studies.

Methodology

The current study presents a systematic review of pertinent literature on electrical energy forecasting.

Data collection

A total of eighty-one (81) state-of-the-art research works published in journals, conferences, and magazines, and student’s thesis relevant to the scope of the current study were downloaded from the internet, thus using keywords and terms which included AI, Electricity Prediction (EP), Energy Forecasting (EF), Machine Learning (ML), and combination of AI and EP, AI and EF, ML and EP, ML and EF. Each downloaded literature was then carefully studied and categorised into the two methods of electrical load forecasting data-driven (artificial intelligence) methods and engineering methods.

Study framework

According to Moher et al. [ 16 ], the quality of every systematic review is based on building protocol, which outlines the justification, hypothesis, and planned methods of the investigation. However, only a few systematic review study reports of their framework. A detailed, well-described structure for systematic reviews facilitates the understanding and evaluation of the methods adopted. Hence, the PRISMA model [ 26 ] was adopted in this study (Fig.  1 ). As shown in Fig.  1 , the PRISMA presents the flow of information from one stage to another in a systematic review of the literature and gives the total number of the research identified, excluded, and included and the reasons for inclusion and exclusions.

Source : Moher et al. [ 26 ]

The adopted PRISMA flow diagram.

The PRISMA flow diagram involved five (5) phases, as shown in Fig.  1 . Phase 1 consists of outlining the review scope, developing questions, and inclusion or exclusion. Phase 2 searches the literature with keywords to identify potential studies. Phase 3 includes determining the addition of a paper by screening its abstracts if it meets inclusion criteria. While phase 4 includes characterisation of paper for mapping by keywords. This review aimed to document an overview of research in the field of electric load forecasting to make way for future studies. As a result, a fifth (5) step offers an in-depth quantitative synthesis (meta-analysis) of studies included in the review.

Our search of literature retrieved one hundred and one (109) papers from online journals and eleven (11) from under sources, making one hundred and twelve (120) records in all (see Fig.  1 ). Of the 120 records, 21 were duplicates, hence, removed leaving ninety-one (99) record shortlisted for the screening stage. At the screening stage, fifteen (15) records were removed; thus, studies that were not related to electrical energy, and those that the primary publication language was not in English. Leaving eighty-four (84) records, of the 84 records, we further remove five (5) more records that were published before the year 2010, two (2) record omitted due to overlapping, this reduced eligible papers for analysis to seventy-seven (77). The 77 papers were used for the qualitative analysis. Ten (10) records that presented a review of electric load forecasting were also removed, and the remaining (67) were used for the quantitative analysis, as shown in Fig.  1 .

Results and discussion

Electrical energy consumption can be classified as residential (domestic), commercial (non-residential), or industrial. Residential or domestic refers to the home or a dwelling where people globally live from day-to-day. At the same time, commercial consumers are the business and industries that require massive supply than residential users for their businesses [ 27 ]. Selected literature was on electric load forecasting was classified into two main categories AI methods and engineering methods. However, each category was further grouped into residential and commercial or combined (residential and commercial), and the outcomes are presented. A total of seventy-seven (77) papers were eligible for the qualitative analysis, while sixty-seven (67) were included in the quantitative analysis, as already discussed above in this study.

AI methods used in electrical energy demand forecasting

This section presents the studies that were based on AI techniques.

Combined (commercial and residential)

Most works on EED forecasting sought to forecast the total load (residential and commercial) demand on the supplying authority. This section presents the selected studies that fell in this category of electric load forecasting.

Hybrid models

In a way to harness the strength in different machine learning techniques, some researchers sought to hybrid two or more ML techniques to improve the forecasting accuracy of their models.

A short-term (next-day) EED forecasting model based on the historical meteorological parameter to forecast the future load on the Greek Electric Network Grid using Support Vector Machine (SVM), ensemble XGBoost, Random Forest (RF), k-Nearest Neighbours (KNN), Neural Networks (NN) and Decision Trees (DT) was proposed by [ 28 ]. The mean absolute percentage error (MAPE) was used as a performance metric for comparison among selected models by the author. The study achieved a reduction in prediction error or + 4.74% compared with the Operator of the Electricity Market in Greece predictions. In other studies [ 29 ], long-term (10 years) EED forecasting using NN and Autoregressive Integrated Moving Average (ARIMA) was proposed to forecast the EED of Kuwait. Weather temperature and humidity, average salary, gross domestic, oil price, population, residence, passengers, currency earning rate, and economic factors like (total import and export in USD) were used as independent variables. The study concluded that NN outperformed ARIMA and weather parameters were found to be more significant than average salary, gross domestic and oil price.

A Particle Swarm Optimisation (PSO) and Differential Evolution (DE) for forecasting the Andhra Pradesh Grid using weather parameters were presented [ 6 ]. The reported concluded that better prediction accuracy was achieved with PSO and DE than the conventional time series forecast model. In another study, a Curve Fitting Algorithm (CFA) was proposed for forecasting electricity power demand for an hour/day/week/month [ 30 ]. Their study shows that future electricity demand can be effectively forecasted based on past demand. In a process to increase forecasting accuracy, historical electricity data combined with Twitter data was used as an input variable to hybrid ANN and SVM forecast model to forecast the electricity consumption in Dutch [ 14 ]. The authors compared the performance of ANN to SVM and concluded that the ANN outperformed the SVM. On the other side, the SVM performance in accuracy increases in long-term forecasting. The authors again admit that inclusion of weather data as input could not increase model performance. Similarly, in Ghana, an attempt to predict the 30-day ahead EED demand of the Bono region using a hybrid ML (MLP, SVM, and DT) based on historical weather and electricity demand was made [ 7 ]. The authors achieved 95% prediction accuracy; however, it affirms that the inclusion of household lifestyle as an input variable will improve prediction accuracy.

The hybrid of ML (SVM and RF) and time-series models Generalized Linear Model (GLM) and ARIMA forecast model was proposed for predict electricity consumption in South Africa based on the historical electricity price, load demand, and weather parameter [ 31 ]. The outcome of the study showed that the ARIMA-GLM combination performs better for long-term forecasting. Similarly, a combination of quantum search with SVM (quantum computing and the chaotic mechanics) for forecasting yearly EED in Taiwan was presented [ 32 ]. The empirical analysis revealed that the proposed model exhibits considerably enhanced forecasting performance than other SVM-based forecasting models. A hybrid of mode decomposition (EMD), PSO, and SVM model was present for forecasting short-term EED demand of the Australian electricity market [ 33 ]. Sulandari et al. [ 34 ] proposed a hybrid of ANN and Fuzzy algorithm and a recurrent formula (LRF) to predict electricity demand in Indonesian. The study results showed that the hybrid model performed well with low values of RMSE. Likewise, in [ 35 ], a hybrid of clustering technique (K-means) and ARIMA forecasting model was presented to forecast university buildings electricity demand. Paper revealed that the hybrid model outperformed the ARIMA model alone as a forecast model.

An artificial neural network (ANN)-based forecast model for short-term forecasting of Chhattisgarh State electricity demand was proposed [ 36 ], using historical weather data as input variables. The results conclude that ANN can efficiently forecast electricity demand. Likewise, an ANN model was applied to forecast the short-term electricity demand of the Iraqi National Grid [ 37 ]. The authors achieved high accuracy and a reasonable error margin. In a similar study, a DT algorithm forecast model was proposed for forecasting future EED on the Yola/Jimeta power transmission company using weather parameters. Again, a short to medium term EED forecasting using deep machine learning (ML) algorithm long short-term memory (LSTM)-based neural network enhanced with genetic algorithm (GA) for feature selection was proposed [ 9 ], to forecast France metropolitan’s electricity consumption. The mean absolute error (MAE) and root-mean-square error (RMSE) were used as the performance metrics, and the weather parameter was used in the independent variable. Their results affirm that weather parament is very useful in forecasting future electricity demand.

An enhanced Convolutional Neural Network (CNN) and enhanced SVM-based forecast model was presented for forecasting electricity price and load forecasting using [ 38 ]. Despite the enhancement made by authors, they recommended additional enhancement of classifiers to improve prediction. In a similar study, an ANN model to forecast consumer demand in North Cyprus was proposed [ 39 ]. Their results affirm the ability for ANN effective automatic modelling of electricity; however, the study concluded this could be achieved when the training and testing datasets are meaningful. A forecast model using a deep belief network using historical EED of Macedonian (2008–2014) was proposed to forecast a short-term (1 day) EED. The outcome of the forecast shows a reduction in MAPE by 8.6% by the proposed model compared with traditional techniques [ 2 ]. In Young et al. [ 40 ], an ANN with Bayesian regularisation algorithm-based model for short-term load forecasting of commercial building electricity usage was carried out. An ANN and hybrid methods to forecast electricity consumption of Turkey were proposed Aydogdu et al. [ 41 ]. The proposed model gave an average absolute prediction error of 2.25%.

Khwaja et al. [ 11 ] presented an ensemble ANN predictive model to enhance short-term electricity load forecasting. Different from existing studies, the authors combined both bagging and boosting techniques to train bagged-boosted ANNs. The study results showed that the proposed ensemble technique offered a reduction of both variance and bias compared to a bagged ANN, single ANN, and boosted ANN. Also, Ahmad et al. [ 42 ] combined Extreme Learning Machine and an enhanced Support Vector Machine to forecast short-term electricity demand. The study outcome showed that the proposed hybrid model outperformed other state-of-the-art predictive models in terms of performance and accuracy. Atef and Eltawil [ 43 ] proposed a deep-stacked LSTM forecasting model to forecast electricity demand. The paper reported that bidirectional (Bi-LSTM) networks outperformed unidirectional (Uni-LSTM) in terms of forecasting accuracy. Also, a generalised regression Neural Network (GRNN) predictive model was proposed in [ 44 ] to predict short-term electricity demand. The study results showed that a GRNN of 30 neurons offered better prediction accuracy in terms of MAPE and MAE than a GRNN of 10 neurons.

Fuzzy logic models

A Fuzzy Logic (FL)-based forecasting model for the next-day electricity demand in Albania was presented [ 45 ]. The time, the historical and forecasting value of the temperature and the previous day load (L) served as the independent variables for the forecast of the next-day consumption. The outcome of the study yielded accurate forecasting by the FL model. Motepe et al. [ 46 ] proposed an adaptive neuro-fuzzy inference system (ANFIS) model for forecasting South African electricity demand. The author concluded that adding temperature as an input parameter to the proposed model did not enhance forecast accuracy, as typically expected.

A new economy (stock indices) reflecting the STLF model for electricity demand forecasting was proposed [ 47 ]. The authors attempted to forecast the future demand for electricity based on the Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX) data. The study revealed a significant relationship between stock market function and energy demand, which was helpful to financial analysis wanting to do reverse engineering.

Residential load forecasting

A survey in 2019 shows that EED forecasting, especially short-term load forecasting for individual (residential) electricity customers, plays an increasingly essential role in the future grid planning and operation Kong et al. [ 51 ]. Similarly, the outcome of Leahy and Lyons [ 48 ] affirms that water heating styled used by a household is even more essential than the number of electrical appliances when explaining domestic electrical energy usage. Furthermore, a study in Portugal shows that residential electricity consumption since 1990 has been increasing more rapidly than the Gross Domestic Product (GDP) per capita [ 49 ]. Therefore, studies that focused on forecasting residential EED is of necessity. However, the results revealed that eight (8) out of the sixty-seven (67) representing (11.94%) reviewed works focused on residential electricity forecasting, and a section of these studies is presented.

An hourly prediction of residential energy consumption using RF and SVM was proposed by Hedén [ 50 ], using 187 households in Austin. Error metrics are mean bias (MB), coefficient of variance (CV), and MAPE. The study achieved better forecasting accuracy compared with traditional time series models. Again, an LTSM for short-term residential load forecasting based on residential meter reading [ 51 ] and resident behaviour learning [ 52 ] was studied. The outcomes of these works showed the effectiveness of the LTSM in electricity demand forecasting.

In the same way, linear regression, RF, and SVM predictive model (data-driven) for estimation of city-scale energy use in buildings were proposed by Kontokosta and Tull [ 53 ]. The outcome of the study revealed that adequate electricity consumption in a building could be predicted using actual data from a moderately small subset of buildings. Likewise, unsupervised ML algorithms such as Self-Organising Maps (SOM), k-means, and k-medoid were used to cluster residential electricity based on their trend of electricity use within the day [ 54 ]. The study found that households and how they use electricity in the home can be categorised based on specific customer characteristics. Mcloughlin et al. [ 55 ] examined residential electricity consumption patterns in Irish based on occupant socio-economic and dwelling variables. The variable examined includes the number of bedrooms, dwelling type, age of household head, social class, household composition, and water heating.

The study outcome showed a positive association between maximum demand periods and several household appliances, especially dishwashers, electric cookers tumble dryers, with electric cook topping. The study further found that the time of use of electrical appliances was dependent on occupant characteristics, and younger occupants of a household tend to use more electricity than the older. Alike, an ARIMA, NN, and exponential smoothening forecast model were proposed for forecasting household electricity demand [ 56 ]. The authors concluded that forecasting accuracy varies considerably depending on the choice of forecasting techniques/tactic and configuration/selection of input parameters. Again, a neural network model for predicting residential building energy consumption was proposed by Biswas et al. [ 57 ]. The outcome of the study showed that models based on OWO-Newton algorithms and Levenberg–Marquardt outperformed previous literature.

Engineering and traditional electrical energy demand forecasting

This section presents the electrical energy demand forecasting study based on the traditional time-series and engineering models.

A predictive model for the prediction of medium-term (1-year) electricity consumption of general households based on the lifestyle of the household using Lasso and Group Lasso was proposed [ 58 ]. Their results showed that household lifestyle such as family composition, age, and house-type is good predictors of electricity consumption in a home. Similarly, in Nigeria, an attempt was made to estimate the electricity demand of residential users to support energy transitions using the engineering approach, such as calculating the total power consumed in a household based on the power rating of appliance and their duration of use [ 59 ]. The study concludes that the proposed system can serve policymaking in Nigeria to improve the financial sustainability of the energy supply structure. An extended Autoregressive Distributed Lag (ARDL) model was proposed to estimate residential electricity consumption per capita demand function, which depends on the GDP per capita in Algeria [ 60 ]. The study concluded that promoting financial growth among citizens of Algeria would reduce electricity consumption, since wealthier people (higher income earners) mighty use of more efficient appliances.

In Bogomolov et al. [ 61 ], authors used general public dynamics derived data from cellular network and energy consumption dataset to predict the next-day energy demand. The study could serve a model to enhance the energy meter to promote energy conservation. Likewise, the historical data of on–off times of residential appliances were used to predict the next-day electricity demand using Bayesian inference [ 62 ]. The study concluded that historical electricity consumption data only is not adequate for a decent eminence hourly forecast.

Also, an ARIMA and Holt-Winter model was proposed for forecasting the national electricity consumption of Pakistan from 1980 to 2011 [ 3 ]. The study revealed that the demand in household energy consumption would higher as compare with all other sectors. Jain et al. [ 63 ] proposed an ARIMA forecaster for forecasting electricity consumption. The proposed model achieved a MAPE of 6.63%. The authors concluded that the ARIMA model has the potential of computing in EED forecasting with other forecast techniques. Integration of three (3) forecasting model, long-range energy alternative planning (LEAP), ARIMA, and Holt-Winter, was proposed for forecasting long-term energy demand in Pakistan [ 64 ]. The study would be valid for energy supplies for accurate estimation of users’ demand in the future. The combination of Extreme Learning Machine and Multiple Regression for forecasting China’s electricity demand was proposed [ 65 ]. A quantile regression (QR) model for long-term electricity demand forecasting in South Africa within 2012 and 2015 was presented by Mokilane et al. [ 66 ]. The model was helpful to power distribution industries in the country.

An attempt was made to forecast the electricity consumption of Ghana by 2030 using ARIMA-based model. The study outcome projected that Ghana’s electricity consumption would grow from 8.5210 billion kWh in 2012 to 9.5597 billion kWh in 2030 [ 67 ]. However, in 2017, a report by the energy commission revealed that electricity consumption was 14,247 GWh [ 68 ]. A generalised additive model was adopted for forecasting medium-term electric energy demand in a South African power supply system [ 69 ]. The outcome of the study revealed a useful application of the proposed model in the power generation and distribution industries in the country. An investigation between the association of causal nexus and (environmental pollution, energy use, GDP per capita, and urbanisation) in an attempt to forecast Nigeria’s energy use by 2030 was carried out using the ARIMA and ETS models [ 70 ]. The study outcome showed better forecasting accuracy by both models, and a high rise in energy demand was observed.

Multiple regression analysis approaches for forecasting the yearly electricity demand in commercial sectors and electricity access rates in rural and urban households in some selected West African countries, which included Ghana, were carried out by Adeoye and Spataru [ 71 ]. The study showed that there is a very high variation in hourly electricity demand in the dry seasons. Their results affirm Nti et al. [ 7 ] report that the demand pattern of electricity in Ghana is highly dependent on the month of the year. In a way, one can say there is a partial agreement in these two studies. A time-series regression model for forecasting South African’s peak load demand was presented [ 72 ]. Experimental fallouts indicated that when the temperature is included as an input parameter, improvement in accuracy by the forecast model was realised. ARIMA-based predictive model for predicting both sectoral and total electrical energy consumption of Turkey for the next 15 years was proposed [ 73 ]. The study points out that the demand for electrical energy in agriculture sectors, transport, public service, residential, and commercial will keep increasing.

Similarly, partially linear additive quantile regression models for forecasting short-term electricity demand during the peak-demand periods (i.e., from 6:00 to 8:00 pm) were carried out in South African [ 74 ]. The authors found out that the use of the proposed system in power utility industries for the planning, scheduling, and dispatching of electricity activities will result in a minimal cost principally during the peak-period hours. Caro et al. [ 75 ] predicted the Spanish electricity demand using the ARIMA model. The study achieved an improvement in the short-term predictions of electricity demand with less computational time. A short-term electricity load forecasting model based on dynamic mode decomposition was proposed in [ 76 ], the proposed model showed better stability and accuracy compared with other predictive models.

Quantitative analysis of findings

The descriptive statistics of the study outcome is presented under this section with tables and charts.

Algorithms used for forecast models

As part of the aim of the review, we sought to find out the most used algorithms in electricity forecasting models. The most top nine (9) used algorithms found in the sixty-seven (67) article are presented (Fig.  2 ); this includes only algorithms that were used in more than a single paper. The study outcome revealed that 90% out of the top nine algorithms were AI-based, with ANN representing 28% of AI models used in electricity forecasting. Besides, the ANN models were primarily used for STLF where electrical energy consumption patterns are more intricate than LTLF. The traditional AFRIMA recorded 17.5% due to its efficiency in LTLF, where load fluctuations and periodicity are less critical.

Top nine (9) most used algorithms for electricity forecasting

Additionally, a high percentage of regression models is used for LTLF prediction. The study outcome shows how AI is applied in various sectors of the economy to improve efficiency and profitability. Also, we observed that the SVM, PSO and Fuzzy are gaining more popularity in recent studies, a sign of increasing attention of researchers on these algorithms for EED forecasting.

Study origin

Table  1 shows the studies and their origin (countries). The origin of surveyed studies was examined, in order to find the linkage between the power crisis in the continent and studies on electrical energy demand forecasting. Concerning geographical coverage, it was found that a high number of studies (31.34%) were carried out in Europe, 17.91% in Africa and 19.40% in Asia. Making Africa the third-highest, however, interestingly, most of the studies in Africa were carried in South Africa (five representing 41.67%), three (3) representing 25% in Ghana, with the rest from Nigeria and other African countries. The energy crisis that hit Europe in 2008, reported by [ 77 ], can be attributed to the numerous studies in electricity forecast, as shown in Table  1 . It was further observed that 2 out of 3 studies in Ghana were based on national or regional electricity demand forecasting (one long-term, one short-term (1 month)). While the third study aimed at identifying the relationship between electricity demand and economic growth. Thus, it suggests that the limited number of studies in electricity demand forecasting by both academicians and professionals in this field might have partially contributed to the power shortages facing the nation, which in 2015 resulted in “Domsur”.

Used evaluation metrics

The performance of every forecasting model is examined based on the difference in error between the actual value \(\left( y \right)\) and the predicted value ( \(\hat{y}\) ). Several of these metrics were identified in the literature. However, we examined the most used in electrical energy demand forecasting to enable new beginners in the field of electricity demand forecasting to have a firm grip on which to apply in their study. Figure  3 shows the top four (4) most used error metrics in two or more studies. The results revealed that root-mean-square error RMSE (38%) was the most used error metric among EED forecasters, followed by mean absolute percentage error MAPE (35%). Due to the effectiveness in measure predictive model performance and their usefulness for short-term prediction, it was also observed that the MAPE is a standard primary metric as it is easy to both calculate and understand. The results affirm the findings in [ 12 , 19 , 22 ] that MAE, MAPE, and RMSE are the most commonly used evaluation metrics in EED forecasting model.

Most used error metrics in electricity load forecasting

Forecast type

Based on short-term load forecasting (STLF), medium-term load forecasting (MTLF), and long-term load forecasting (LTLF), it was observed that 80%, 15%, and 5% of the electrical energy demand forecasting were STLF, MTLF, and LTLF, respectively. The massive number of studies is based on STLF as compared with LTLF and MTLF call for further studies into the challenges associated with LTLF and MTLF electricity load forecasting. Again, 80% STLF forecasting affirms the 38% use of RMSE error metric, since it is for STLF forecasting. The results affirm the report in [ 22 ] that 43.6% of electricity forecasting are short-term prediction.

Model input parameters

The efficiency of every predictive model is believed to be partially dependent on the independent (input) variable [ 7 ]. At this level, the focus was to examine the different input variables used for electricity load forecasting. The type of independent variables (input features) used by electricity load forecasters was also examined. The current study observed that sixty (60) out of the sixty-seven (67) papers made known the input parameter to the proposed model. Table  2 presents the type of input variables and the percentage of studies that utilised it. It was observed that a high percentage (50%) of electricity demand forecasting was based on weather parameters; next was the historical electricity consumption pattern. The outcome exposed that little attention is given to household lifestyle in electricity demand forecasting. However, Nishida et al. [ 58 ] argue that residential (domestic) energy consumption differs depending on the lifestyle of the family. Family lifestyle, according to [ 56 ], cannot be undermined in electrical load forecasting. According to these studies, these factors include the life stage family composition, house type, age, home appliances possessed and their usage, family income, cultural background, social life, and lifestyle habits, which include how long to stay at home and how to spend holidays. The observations open the opportunity for further studies on the association between EED and household lifestyle.

Conclusions

The current study sought to reviewed state-of-the-art literature on electricity load forecasting to identify the challenges and opportunities for future studies. The outcome of the study revealed that electricity load forecasting is seen to be complicated for both engineers and academician and is still an ongoing area of research. The key findings are summarised as follows.

Several studies (90%) have applied AI in electrical energy demand forecasting as compared with traditional engineering and statistical method (10%) to address energy prediction problems; however, there are not enough studies benchmarking the performance of these methods.

There are few studies on EED forecasting in Africa countries (12 out of 67). Though the continent has progressive achievement in the creation of Regional Power Pools (PPP) over the last two decades, the continent still suffers from a lousy power network in most of its countries, leaving millions of people in Africa without electricity.

Temperature and rainfall as an input parameter to the EED forecasting model are seen to have a divergent view. At the same time, some sections of research recorded an improvement in accuracy and reported no improvement in accuracy when introduced and input. However, the current study attributes this to the difference in automorphic temperature globally and the different economic status among countries. An additional investigation will bring more clarity to the literature.

This study revealed that EED forecasting in the residential sector had seen little attention. On the other hand, Guo et al. [ 78 ] argue that the basic unit of electricity consumption is home.

It was observed that there had been a global increase in residential electricity demand, this according to the report in [ 49 ] can be attributed to the growing rate of buying electrical equipment and appliances of low quality due to higher living standards. However, a further probe into Soares et al. [ 49 ] assertion will bring clarity to literature because of the discrepancy in opinions in literature.

Lastly, the study revealed that there is a limited number of studies on load forecasting studies in Ghana. We, therefore, recommend rigorous researchers in this field in the country to enhance the economic growth of the country.

Our future study will focus on identifying the relationship between household lifestyle factors and electricity consumption in Ghana and predict load consumption based on identified factors since it is an area that has seen little or no attention in Ghana.

Availability of data and materials

All data generated or analysed during this study are included in this published article.

Abbreviations

• Electrical energy demand
• Artificial intelligence

Artificial neural network

Root-mean-square error

Mean absolute percentage error

Very short-term load forecasting

Short-term load forecasting

Medium-term load forecasting

Long-term load forecasting

Preferred-reporting items for systematic-review and meta-analysis

Electricity prediction

Energy forecasting

• Machine learning

Support vector machine

Random forest

k-nearest neighbours

Neural networks

Decision trees

Autoregressive integrated moving average

Particle swarm optimisation

Differential evolution

Curve fitting algorithm

Generalised linear model

Long short-term memory

Genetic algorithm

Mean absolute error

Convolutional neural network

Fuzzy logic

Adaptive neuro-fuzzy inference system

Generalised regression neural network

Taiwan stock exchange capitalisation-weighted stock index

Gross domestic product

Coefficient of variance

Self-organising maps

Autoregressive distributed lag

Long-range energy alternative planning

Quantile regression

Aung SS (2015) Electric power is the main driving force for industrialization. http://www.globalnewlightofmyanmar.com/electric-power-is-the-main-driving-force-for-industrialization/ . Accessed 2 Apr 2018

Dedinec A, Filiposka S, Dedinec A, Kocarev L (2016) Deep belief network based electricity load forecasting: an analysis of Macedonian case. Energy 115:1688–1700. https://doi.org/10.1016/j.energy.2016.07.090

Article   Google Scholar  

Hussain A, Rahman M, Memon JA (2016) Forecasting electricity consumption in Pakistan: the way forward. Energy Policy 90:73–80. https://doi.org/10.1016/j.enpol.2015.11.028

Jevgenijs S, Joeri deW, Kochnakyan A, Vivien F (2017) Forecasting electricity demand: an aid for practitioners. http://www.worldbank.org/energy/livewire . Accessed 15 Jun 2019

Zaman MU, Islam A, Sultana N (2018) Short term load forecasting based on internet of things (IoT). BRAC University, Dhaka

Google Scholar  

Kumar CHJ, Veerakumari M (2012) Load forecasting of Andhra Pradesh grid using PSO, DE algorithms. Int J Adv Res Comput Eng Technol 1:179–184

Nti IK, Asafo-Adjei S, Agyemang M (2019) Predicting monthly electricity demand using soft-computing technique. Int Res J Eng Technol 06:1967–1973

Eeeguide.com (2014) Forecasting methodology. http://www.eeeguide.com/forecasting-methodology/ . Accessed 4 Jan 2019

Bouktif S, Fiaz A, Ouni A, Serhani MA (2018) Optimal deep learning LSTM model for electric load forecasting using feature selection and genetic algorithm: comparison with machine learning approaches †. Energies (MDPI) 11:1–20. https://doi.org/10.3390/en11071636

Hammad MA, Jereb B, Rosi B, Dragan D (2020) Methods and models for electric load forecasting: a comprehensive review. Logist Sustain Transp 11:51–76. https://doi.org/10.2478/jlst-2020-0004

Khwaja AS, Anpalagan A, Naeem M, Venkatesh B (2020) Joint bagged-boosted artificial neural networks: using ensemble machine learning to improve short-term electricity load forecasting. Electr Power Syst Res 179:106080. https://doi.org/10.1016/j.epsr.2019.106080

Hong T, Fan S (2016) Probabilistic electric load forecasting: a tutorial review. Int J Forecast 32:914–938. https://doi.org/10.1016/j.ijforecast.2015.11.011

Selvaraj KR, Sundararaj S, Ravi T (2013) Artificial neutral network based load forecasting and economic dispatch with particle swarm optimization. Int J Sci Eng Res 4:139–145

Stavast P (2014) Prediction of energy consumption using historical data and twitter. The University of Groningen, Groningen

Ouedraogo NS (2017) Modeling sustainable long-term electricity supply-demand in Africa. Appl Energy 190:1047–1067. https://doi.org/10.1016/j.apenergy.2016.12.162

Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, Stewart LA, Group P-P (2015) Preferred reporting items for systematic review and and explanation meta-analysis protocols (PRISMA-P) 2015: elaboration. Syst Rev 4:1–25

Zhou K, Yang S, Shen C (2013) A review of electric load classification in smart grid environment. Renew Sustain Energy Rev 24:103–110. https://doi.org/10.1016/j.rser.2013.03.023

Gonzalez-Briones A, Hernandez G, Corchado JM, Omatu S, Mohamad MS (2019) Machine learning models for electricity consumption forecasting: a review. In: 2019 2nd international conference on computer applications and information security, IEEE, pp 1–6. https://doi.org/10.1109/cais.2019.8769508

Wang Z, Li J, Zhu S, Zhao J, Deng S, Zhong S, Yin H, Li H, Qi Y, Gan Z (2019) A review of load forecasting of the distributed energy system. IOP Conf Ser Earth Environ Sci 237:042019. https://doi.org/10.1088/1755-1315/237/4/042019

Mele E (2019) A review of machine learning algorithms used for load forecasting at microgrid level. In: Sinteza 2019-International scientific conference on information technology and data related research, Singidunum University, pp 452–458. https://doi.org/10.15308/sinteza-2019-452-458

Shah RB (2019) A technological literature review on load forecasting in power system using artificial intelligence. Paripex-Indian J Res 8:14–15

Kuster C, Rezgui Y, Mourshed M (2017) Electrical load forecasting models: a critical systematic review. Sustain Cities Soc 35:257–270. https://doi.org/10.1016/j.scs.2017.08.009

Panda SK, Mohanty SN, Jagadev AK (2017) Long term electrical load forecasting: an empirical study across techniques and domains. Indian J Sci Technol 10:1–16

Jacob M, Neves C, Vukadinović-Greetham D (2020) Short term load forecasting. Forecast Assess Risk Individ Electr Peaks Math Planet Earth. https://doi.org/10.1007/978-3-030-28669-9_2

Article   MATH   Google Scholar  

Appiahene P, Ussiph N, Missah YM (2018) Information technology impact on productivity. In: International conference on applied science and technology, Kumasi Technical University, Kumasi, Ghana, pp 122–133

Moher D, Liberati A, Tetzlaff J, Altman DG (2009) PRISMA Group, preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 151:264–269. https://doi.org/10.7326/0003-4819-151-4-200908180-00135

Electricchoice.com (2016) Residential electricity for houses, apartments, condos, trailers, etc. https://www.electricchoice.com/residential-electricity . Accessed 10 Aug 2019

Kouroupetroglou P-N (2017) Machine learning techniques for short-term electric load forecasting. Aristotle University of Thessaloniki, Thessaloniki

Zakarya S, Abbas H, Belal M (2017) Long-term deep learning load forecasting based on social and economic factors in the Kuwait region. J Theor Appl Inf Technol 95:1524–1535

Soni A, Sharma AK (2013) Electricity load forecast for power system planning. Int Ref J Eng Sci 2:52–57

Angamuthu CR, Mukherjee A, Campion M, Salehfar H, Hansen T, Lin J, Ranganathan P (2018) A multi-stage price forecasting model for day-ahead electricity markets. Forecasting 1:3. https://doi.org/10.3390/forecast1010003

Lee CW, Lin BY (2016) Application of hybrid quantum tabu search with support vector regression (SVR) for load forecasting. Energies 9:1–16. https://doi.org/10.3390/en9110873

Hong WC, Fan GF (2019) Hybrid empirical mode decomposition with support vector regression model for short term load forecasting. Energies 12:1–16. https://doi.org/10.3390/en12061093

Sulandari W, Subanar MH, Lee PC (2020) Rodrigues, Indonesian electricity load forecasting using singular spectrum analysis, fuzzy systems and neural networks. Energy 190:116408. https://doi.org/10.1016/j.energy.2019.116408

Nepal B, Yamaha M, Yokoe A, Yamaji T (2020) Electricity load forecasting using clustering and ARIMA model for energy management in buildings. Jpn Archit Rev 3:62–76. https://doi.org/10.1002/2475-8876.12135

Ghore S, Goswami A (2015) Short term load forecasting of Chhattisgarh grid using artificial neural network. Int J Eng Dev Res 3:391–397

Kuhba H, Al-tamemi HAH (2016) Power system short-term load forecasting using artificial neural networks. Int J Eng Dev Res 4:78–87

Zahid M, Ahmed F, Javaid N, Abbasi R, Zainab KH, Javaid A, Bilal M, Akbar M, Ilahi M (2019) Electricity price and load forecasting using enhanced convolutional neural network and enhanced support vector regression in smart grids. Electronics 8:122. https://doi.org/10.3390/electronics8020122

Nwulu NI, Agboola OP (2012) Modelling and predicting electricity consumption using artificial neural networks. In: 2012 11th international conference on environmental and electrical engineering, pp 1059–1063

Young TC, Raya H, Youngdeok H, Young ML, (2016) Artificial neural network model for forecasting sub-hourly electricity usage in commercial buildings. Energy Buildings 111:184–194.  https://doi.org/10.1016/j.enbuild.2015.11.045

Aydodu G, Yildiz O (2017) Forecasting the annual electricity consumption of Turkey using a hybrid model. In: 2017 25th signal processing and communication application conference, Antalya, Turkey, pp 1–4. https://doi.org/10.1109/SIU.2017.7960283

Ahmad W, Ayub N, Ali T, Irfan M, Awais M, Shiraz M, Glowacz A (2020) Towards short term electricity load forecasting using improved support vector machine and extreme learning machine. Energies 13:1–17. https://doi.org/10.3390/en13112907

Atef S, Eltawil AB (2020) Assessment of stacked unidirectional and bidirectional long short-term memory networks for electricity load forecasting. Electr Power Syst Res 187:106489. https://doi.org/10.1016/j.epsr.2020.106489

Azar AT, Khamis A, Kamal NA, Galli B (2020) Short term electricity load forecasting through machine learning. In: Hassanien A, Azar A, Gaber T, Oliva D, Tolba F (eds) International conference on artificial intelligence computer vision (AICV2020). AICV 2020. Advances in intelligent systems and computing, vol 1153, Springer, Cham. https://link.springer.com/chapter/10.1007/978-3-030-44289-7_40 . Accessed 5 July 2020

Konica JA, Hanelli L (2016) Forecasting next-day the electricity demand based on fuzzy logic method case for. J Multidiscip Eng Sci Technol 3:6172–6180

Motepe S, Hassan AN, Stopforth R (2019) South African distribution networks load forecasting using ANFIS. In: Proceedings of 2018 IEEE international conference power electronic drives energy system PEDES 2018, IEEE, Chennai, India, pp 1–6. https://doi.org/10.1109/pedes.2018.8707876

Lin CT, Der Chou L (2013) A novel economy reflecting short-term load forecasting approach. Energy Convers Manag 65:331–342. https://doi.org/10.1016/j.enconman.2012.08.001

Leahy E, Lyons S (2010) Energy use and appliance ownership in Ireland. Energy Policy 38:4265–4279. https://doi.org/10.1016/j.enpol.2010.03.056

Soares A, Gomes Á, Antunes CH (2014) Categorisation of residential electricity consumption as a basis for the assessment of the impacts of demand response actions. Renew Sustain Energy Rev 30:490–503. https://doi.org/10.1016/j.rser.2013.10.019

Hedén W (2016) Predicting hourly residential energy consumption using random forest and support vector regression an analysis of the impact of household clustering on the performance accuracy. KTH Royal Institute of Technology, Stockholm

Kong W, Dong ZY, Jia Y, Hill DJ, Xu Y, Zhang Y (2019) Short-term residential load forecasting based on LSTM recurrent neural network. IEEE Trans Smart Grid 10:841–851. https://doi.org/10.1109/TSG.2017.2753802

Kong W, Dong ZY, Hill DJ, Luo F, Xu Y (2018) Short-term residential load forecasting based on resident behaviour learning. IEEE Trans Power Syst 33:2016–2017. https://doi.org/10.1109/TPWRS.2017.2688178

Kontokosta CE, Tull C (2017) A data-driven predictive model of city-scale energy use in buildings. Appl Energy 197:303–317. https://doi.org/10.1016/j.apenergy.2017.04.005

McLoughlin F, Duffy A, Conlon M (2015) A clustering approach to domestic electricity load profile characterisation using smart metering data. Appl Energy 141:190–199. https://doi.org/10.1016/j.apenergy.2014.12.039

Mcloughlin F, Duffy A, Conlon M (2012) Characterising domestic electricity consumption patterns by dwelling and occupant socio-economic variables: an Irish case study. Energy Build 48:240–248. https://doi.org/10.1016/j.enbuild.2012.01.037

Veit A, Goebel C, Tidke R, Doblander C, Jacobsen H (2014) Household electricity demand forecasting-benchmarking state-of-the-art methods. arXiv:1603.00751 , pp 1–10

Biswas MAR, Robinson MD, Fumo N (2016) Prediction of residential building energy consumption: a neural network approach. Energy 117:84–92. https://doi.org/10.1016/j.energy.2016.10.066

Nishida K, Takeda A, Iwata S, Kiho M, Nakayama I (2017) Household energy consumption prediction by feature selection of lifestyle data. In: International conference on smart grid communication, Dresden, Germany, pp 235–240

Olaniyan K, Mclellan BC (2018) Estimating residential electricity consumption in Nigeria to support energy transitions. Sustain 10:1–22. https://doi.org/10.3390/su10051440

Bouznit M, Pablo-Romero MP (2018) Residential electricity consumption and economic growth in Algeria. Energies (MDPI). https://doi.org/10.3390/en11071656

Bogomolov A, Lepri B, Larcher R, Antonelli F, Pianesi F, Pentland A (2016) Energy consumption prediction using people dynamics derived from cellular network data. EPJ Data Sci. https://doi.org/10.1140/epjds/s13688-016-0075-3

Huber P, Gerber M, Rumsch A, Paice A (2018) Prediction of domestic appliances usage based on electrical consumption. Energy Inform 1:265–428

Jain PK, Quamer W, Pamula R (2018) Electricity consumption forecasting using time series analysis. In: Singh M, Gupta P, Tyagi V, Flusser J, Ören T (eds), Advanced computer data science ICACDS 2018. Communications in computer and information science, Springer, Singapore, Singapore. https://doi.org/10.1007/978-981-13-1813-9_33

Rehman USA, Cai Y, Fazal R, Das Walasai G, Mirjat NH (2017) An integrated modeling approach for forecasting long-term energy demand in Pakistan. Energies 10:1–23. https://doi.org/10.3390/en10111868

Liang Y, Niu D, Cao Y, Hong WC (2016) Analysis and modeling for China’s electricity demand forecasting using a hybrid method based on multiple regression and extreme learning machine: a view from carbon emission. Energies 9:1–22. https://doi.org/10.3390/en9110941

Mokilane P, Galpin J, Yadavalli VSS, Debba P, Koen R, Sibiya S (2018) Density forecasting for long-term electricity demand in South Africa using quantile regression. S Afr J Econ Manag Sci 21:1–14. https://doi.org/10.4102/sajems.v21i1.1757

Sarkodie SA (2017) Estimating Ghana’s electricity consumption by 2030: an ARIMA forecast, energy sources. Part B Econ Plan Policy 12:936–944. https://doi.org/10.1080/15567249.2017.1327993

Energy Commission, Energy Commission, Ghana 2018 Energy (Supply and Demand) Outlook for Ghana, Accra, 2018. http://www.energycom.gov.gh/planning/data-center/energy-outlook-for-ghana . Accessed 1 Feb 2020

Sigauke C (2017) Forecasting medium-term electricity demand in a South African electric power supply system. J Energy S Afr. https://doi.org/10.17159/2413-3051/2017/v28i4a2428

Asumadu-Sarkodie S, Owusu PA (2016) Forecasting Nigeria’s energy use by 2030, an econometric approach, Energy Sources. Part B Econ Plan Policy 11:990–997. https://doi.org/10.1080/15567249.2016.1217287

Adeoye O, Spataru C (2019) Modelling and forecasting hourly electricity demand in West African countries. Appl Energy 242:311–333. https://doi.org/10.1016/j.apenergy.2019.03.057

Sigauke C, Chikobvu D (2016) Peak electricity demand forecasting using time series regression models: an application to South African data. J Stat Manag Syst 19:567–586. https://doi.org/10.1080/09720510.2015.1086146

Ozturk S, Ozturk F (2018) Prediction of energy consumption of Turkey on sectoral bases by Arima model. Energy Econ Lett 5:23–30. https://doi.org/10.18488/journal.82.2018.51.23.30

Lebotsa ME, Sigauke C, Bere A, Fildes R, Boylan JE (2018) Short term electricity demand forecasting using partially linear additive quantile regression with an application to the unit commitment problem. Appl Energy 222:104–118. https://doi.org/10.1016/j.apenergy.2018.03.155

Caro E, Juan J, Cara J (2020) Periodically correlated models for short-term electricity load forecasting. Appl Math Comput 364:124642. https://doi.org/10.1016/j.amc.2019.124642

Article   MathSciNet   MATH   Google Scholar  

Kong X, Li C, Wang C, Zhang Y, Zhang J (2020) Short-term electrical load forecasting based on error correction using dynamic mode decomposition. Appl Energy 261:114368. https://doi.org/10.1016/j.apenergy.2019.114368

Auverlot D, Beeker É, Hossie G, Oriol L, Rigard-Cerison A (2014) The crisis of the european electricity system: diagnosis and possible ways forward. https://www.strategie.gouv.fr/sites/strategie.gouv.fr/files/archives/CGSP_Report_European_Electricity_System_030220141.pdf . Accessed 18 Apr 2019

Guo Z, Zhou K, Zhang C, Lu X, Chen W, Yang S (2018) Residential electricity consumption behavior: influencing factors, related theories and intervention strategies. Renew Sustain Energy Rev 81:399–412. https://doi.org/10.1016/j.rser.2017.07.046

Download references

Acknowledgements

Not applicable.

The authors declare that they did not receive any funding for this research.

Author information

Authors and affiliations.

Department of Computer Science, Sunyani Technical University, Box 206, Sunyani, Ghana

Isaac Kofi Nti

Department of Electrical and Electronic Engineering, Sunyani Technical University, Sunyani, Ghana

Moses Teimeh

Department of Computer Science and Informatics, University of Energy and Natural Resources, Sunyani, Ghana

Owusu Nyarko-Boateng & Adebayo Felix Adekoya

You can also search for this author in PubMed   Google Scholar

Contributions

We, at this moment, declare that IKN and MT contributed to the design and implementation of the research. MT, IKN ON-B and AFA contributed to the analysis of the results and the writing of the manuscript. The manuscript has been read and approved by all named authors, and there are no other persons who satisfied the criteria for authorship but are not listed. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Isaac Kofi Nti .

Ethics declarations

Competing interests.

The authors of the current study declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and Permissions

About this article

Cite this article.

Nti, I.K., Teimeh, M., Nyarko-Boateng, O. et al. Electricity load forecasting: a systematic review. Journal of Electrical Systems and Inf Technol 7 , 13 (2020). https://doi.org/10.1186/s43067-020-00021-8

Download citation

Received : 06 April 2020

Accepted : 30 August 2020

Published : 09 September 2020

DOI : https://doi.org/10.1186/s43067-020-00021-8

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Electricity load forecasting

Accessibility Links

• Skip to content
• Skip to search IOPscience
• Skip to Journals list
• Accessibility help
• Accessibility Help

Click here to close this panel.

The rise of electric vehicles—2020 status and future expectations

Matteo Muratori 13,1 , Marcus Alexander 2 , Doug Arent 1 , Morgan Bazilian 3 , Pierpaolo Cazzola 4 , Ercan M Dede 5 , John Farrell 1 , Chris Gearhart 1 , David Greene 6 , Alan Jenn 7 , Matthew Keyser 1 , Timothy Lipman 8 , Sreekant Narumanchi 1 , Ahmad Pesaran 1 , Ramteen Sioshansi 9 , Emilia Suomalainen 10 , Gil Tal 7 , Kevin Walkowicz 11 and Jacob Ward 12

Published 25 March 2021 • © 2021 IOP Publishing Ltd Progress in Energy , Volume 3 , Number 2 Focus on Transport Electrification Citation Matteo Muratori et al 2021 Prog. Energy 3 022002 DOI 10.1088/2516-1083/abe0ad

Article metrics

57430 Total downloads

Permissions

Get permission to re-use this article

Share this article

Author e-mails.

[email protected]

Author affiliations

1 National Renewable Energy Laboratory, Golden, CO, United States of America

2 Electric Power Research Institute, Palo Alto, CA, United States of America

3 Colorado School of Mines, Golden, CO, United States of America

4 International Transport Forum in Paris, France

5 Toyota Research Institute of North America, Ann Arbour, MI, United States of America

6 University of Tennessee, Knoxville, TN, United States of America

7 University of California, Davis, CA, United States of America

8 University of California, Berkeley, CA, United States of America

9 The Ohio State University, Columbus, OH, United States of America

10 Institut VEDECOM, Versailles, France

11 Calstart, Pasadena, CA, United States of America

12 Carnegie Mellon University, Pittsburgh, PA, United States of America

Author notes

13 Author to whom any correspondence should be addressed.

Matteo Muratori https://orcid.org/0000-0003-1688-6742

Doug Arent https://orcid.org/0000-0002-4219-3950

Morgan Bazilian https://orcid.org/0000-0003-1650-8071

Ahmad Pesaran https://orcid.org/0000-0003-0666-1021

Emilia Suomalainen https://orcid.org/0000-0002-6339-2932

Gil Tal https://orcid.org/0000-0001-7843-3664

Jacob Ward https://orcid.org/0000-0002-8278-8940

• Received 3 August 2020
• Accepted 27 January 2021
• Published 25 March 2021

Peer review information

Method : Single-anonymous Revisions: 3 Screened for originality? Yes

Buy this article in print

Electric vehicles (EVs) are experiencing a rise in popularity over the past few years as the technology has matured and costs have declined, and support for clean transportation has promoted awareness, increased charging opportunities, and facilitated EV adoption. Suitably, a vast body of literature has been produced exploring various facets of EVs and their role in transportation and energy systems. This paper provides a timely and comprehensive review of scientific studies looking at various aspects of EVs, including: (a) an overview of the status of the light-duty-EV market and current projections for future adoption; (b) insights on market opportunities beyond light-duty EVs; (c) a review of cost and performance evolution for batteries, power electronics, and electric machines that are key components of EV success; (d) charging-infrastructure status with a focus on modeling and studies that are used to project charging-infrastructure requirements and the economics of public charging; (e) an overview of the impact of EV charging on power systems at multiple scales, ranging from bulk power systems to distribution networks; (f) insights into life-cycle cost and emissions studies focusing on EVs; and (g) future expectations and synergies between EVs and other emerging trends and technologies. The goal of this paper is to provide readers with a snapshot of the current state of the art and help navigate this vast literature by comparing studies critically and comprehensively and synthesizing general insights. This detailed review paints a positive picture for the future of EVs for on-road transportation, and the authors remain hopeful that remaining technology, regulatory, societal, behavioral, and business-model barriers can be addressed over time to support a transition toward cleaner, more efficient, and affordable transportation solutions for all.

Export citation and abstract BibTeX RIS

This article was updated on 29 April 2021 to add the name of the fifth author and to correct the name of the eighth author.

1. Introduction

First introduced at the end of the 1800s, electric vehicles (EVs) 12 have been experiencing a rise in popularity over the past few years as the technology has matured and costs (especially of batteries) have declined substantially. Worldwide support for clean transportation options (i.e. low emissions of greenhouse gasses [GHG] to mitigate climate change and criteria pollutants) has promoted awareness, increased charging opportunities, and facilitated adoption of EVs. EVs present numerous advantages compared to fossil-fueled internal-combustion-engine vehicles (ICEVs), inter alia: zero tailpipe emissions, no reliance on petroleum, improved fuel economy, lower maintenance, and improved driving experience (e.g. acceleration, noise reduction, and convenient home and opportunity recharging). Further, when charged with clean electricity, EVs provide a viable pathway to reduce overall GHG emissions and decarbonize on-road transportation. This decarbonization potential is important, given limited alternative options to liquid fossil fuels. The ability of EVs to reduce GHG emissions is dependent, however, upon clean electricity. Therefore, EV success is intertwined closely with the prospect of abundant and affordable renewable electricity (in particular solar and wind electricity) that is poised to transform power systems (Jacobson et al 2015 , Kroposki et al 2017 , Gielen et al 2019 , IEA 2020b ). Coordinated actions can produce beneficial synergies between EVs and power systems and support renewable-energy integration to optimize energy systems of the future to benefit users and offer decarbonization across sectors (CEM 2020 ). A cross-sectoral approach across the entire energy system is required to realise clean future transformation pathways (Hansen et al 2019 ). EVs are expected to play a critical role in the power system of the future (Muratori and Mai).

EV success is increasing rapidly since the mid-2010s. EV sales are breaking previous records every year, especially for light-duty vehicles (LDVs), buses, and smaller vehicles such as three-wheelers, mopeds, kick-scooters, and e-bikes (IEA 2017 , 2018a , 2019 , 2020 ). To date, global automakers are committing more than $140 billion to transportation electrification, and 50 light-duty EV models are available commercially in the U.S. market (Moore and Bullard 2020 ). Approximately 130 EV models are anticipated by 2023 (AFDC 2020 , Moore and Bullard 2020 ). Future projections of the role of EVs in LDV markets vary widely, with estimates ranging from limited success (∼10% of sales in 2050) to full market dominance, with EVs accounting for 100% of LDV sales well before 2050. Many studies project that EVs will become economically competitive with ICEVs in the near future or that they are already cost-competitive for some applications (Weldon et al 2018 , Sioshansi and Webb 2019 , Yale E360 2019 , Kapustin and Grushevenko 2020 ). However, widespread adoption requires more than economic competitiveness, especially for personally owned vehicles. Behavioral and non-financial preferences of individuals on different technologies and mobility options are also important (Lavieri et al 2017 , Li et al 2017 , McCollum et al 2018 , Ramea et al 2018 ). EV adoption beyond LDVs has been focused on buses, with significant adoption in several regions (especially China). Electric trucks also are receiving great attention, and Bloomberg New Energy Finance (BloombergNEF) projects that by 2025, alternative fuels will compete with, or outcompete, diesel in long-haul trucking applications (Moore and Bullard 2020 ). These recent successes are being driven by technological progress, especially in batteries and power electronics, greater availability of charging infrastructure, policy support driven by environmental benefits, and consumer acceptance. EV adoption is engendering a virtuous circle of technology improvements and cost reductions that is enabled and constrained by positive feedbacks arising from scale and learning by doing, research and development, charging-infrastructure coverage and utilization, and consumer experience and familiarity with EVs.

Vehicle electrification is a game-changer for the transportation sector due to major energy and environmental implications driven by high vehicle efficiency (EVs are approximately 3–4 times more efficient than comparable ICEVs), zero tailpipe emissions, and reduced petroleum dependency (great fuel diversity and flexibility exist in electricity production). Far-reaching implications for vehicle-grid integration extend to the electricity sector and to the broader energy system. A revealing example of the role of EVs in broader energy-transformation scenarios is provided by Muratori and Mai, who summarize results from 159 scenarios underpinning the special report on Global Warming of 1.5 °C (SR1.5) by Intergovernmental Panel on Climate Change (IPCC). Muratori and Mai also show that transportation represents only ∼2% of global electricity demand currently (with rail responsible for more than two-thirds of this total). They show that electricity is projected to provide 18% of all transportation-energy needs by 2050 for the median IPCC scenario, which would account for 10% of total electricity demand. Most of this electricity use is targeted toward on-road vehicle electrification. These projections are the result of modeling and simulations that are struggling to keep pace with the EV revolution and its role in energy-transformation scenarios as EV technologies and mobility are evolving rapidly (McCollum et al 2017 , Venturini et al 2019 , Muratori et al 2020 ). Recent studies explore higher transportation-electrification scenarios: for example, Mai et al ( 2018 ) report a scenario in which 75% of on-road miles are powered by electricity, and transportation represents almost a quarter of total electricity use during 2050.

Vehicle electrification is a disruptive element in energy-system evolution that radically changes the roles of different sectors, technologies, and fuels in long-term transformation scenarios. Traditionally, energy-system-transformation studies project minimal end-use changes in transportation-energy use over time (limited mode shifting and adoption of alternative fuels), and the sector is portrayed as a 'roadblock' to decarbonization. In many decarbonization scenarios, transportation is seen traditionally as one of the biggest hurdles to achieve emissions reductions (The White House 2016 ). These scenarios rely on greater changes in the energy supply to reduce emissions and petroleum dependency (e.g. large-scale use of bioenergy, often coupled to carbon capture and sequestration) rather than demand-side transformations (IPCC 2014 , Pietzcker et al 2014 , Creutzig et al 2015 , Muratori et al 2017 , Santos 2017 ). In most of these studies, electrification is limited to some transportation modes (e.g. light-duty), and EVs are not expected to replace ICEVs fully (The White House 2016 ). More recently, however, major mobility disruptions (e.g. use of ride-hailing and vehicle ride-sharing) and massive EV adoption across multiple applications are proposed (Edelenbosch et al 2017 , Van Vuuren et al 2017 , Hill et al 2019 , E3 2020 , Zhang and Fujimori 2020 ). These mobility disruptions allow for more radical changes and increase the decarbonization role of transportation and highlight the integration opportunities between transportation and energy supply, especially within the electricity sector. For example, Zhang and Fujimori ( 2020 ) highlight that for vehicle electrification to contribute to climate-change mitigation, electricity generation needs to transition to clean sources. They note that EVs can reduce mitigation costs, implying a positive impact of transport policies on the economic system (Zhang and Fujimori 2020 ). In-line with these projections, many countries are establishing increasingly stringent and ambitious targets to support transport electrification and in some cases ban conventional fossil fuel vehicles (Wentland 2016 , Dhar et al 2017 , Coren 2018 , CARB 2020 , State of California 2020 ).

EV charging undoubtedly will impact the electricity sector in terms of overall energy use, demand profiles, and synergies with electricity supply. Mai et al ( 2018 ) show that in a high-electrification scenario, transportation might grow from the current 0.2% to 23% of total U.S. electricity demand in 2050 and significantly impact system peak load and related capacity costs if not controlled properly. Widespread vehicle electrification will impact the electricity system across the board, including generation, transmission, and distribution. However, expected changes in U.S. electricity demand as a result of vehicle electrification are not greater than historical growth in load and peak demand. This finding suggests that bulk-generation capacity is expected to be available to support a growing EV fleet as it evolves over time, even with high EV-market growth (U.S. DRIVE 2019 ). At the same time, many studies have shown that 'smart charging' and vehicle-to-grid (V2G) services create opportunities to reduce system costs and facilitate the integration of variable renewable energy (VRE). Charging infrastructure that enables smart charging and alignment with VRE generation, as well as business models and programs to compensate EV owners for providing charging flexibility, are the most pressing required elements for successfully integrating EVs with bulk power systems. At the local level, EV charging could increase and change electricity loads significantly, which could impact distribution networks and power quality and reliability (FleetCarma 2019 ). Distribution-network impacts can be particularly critical for high-power charging and in cases in which many EVs are concentrated in a specific location, such as clusters of residential LDV charging and possibly fleet depots for commercial vehicles (Muratori 2018 ).

This paper provides a timely status of the literature on several aspects of EV markets, technologies, and future projections. The paper focuses on multiple facets that characterize technology status and the role of EVs in transportation decarbonization and broader energy-transformation pathways focusing on the U.S. context. As appropriate, global context is provided as well. Hybrid EVs (for which liquid fuel is the only source of energy) and fuel cell EVs (that power an electric powertrain with a fuel cell and on-board hydrogen storage) have some similarities with EVs and could complement them for many applications. However, these technologies are not reviewed in detail here. The remainder of this paper is structured as follows. Section 2 focuses on the status of the light-duty-EV market and provides a comparison of projections for future adoption. Section 3 provides insights on market opportunities beyond LDVs. Section 4 offers a review of cost and performance evolution for batteries, power electronics, and electric machines that are key components of EV success. Section 5 reviews charging-infrastructure status and focuses on modeling and analysis studies used to project charging-infrastructure requirements, the economics of public charging, and some considerations on cybersecurity and future technologies (e.g. wireless charging). Section 6 provides an overview of the impact of EV charging on power systems at multiple scales, ranging from bulk power systems to distribution networks. Section 7 provides insights into life-cycle cost and emissions studies focusing on EVs. Finally, section 8 touches on future expectations.

1.1. Summary of take-away points

1.1.1. ev adoption.

• The global rate of adoption of light-duty EVs (passenger cars) has increased rapidly since the mid-2010s, supported by three key pillars: improvements in battery technologies; a wide range of supportive policies to reduce emissions; and regulations and standards to promote energy efficiency and reduce petroleum consumption.
• Adoption of advanced technologies has been underestimated historically in modeling and analyses; EV adoption is projected to remain limited until 2030, and high uncertainty is shown afterward with widely different projections from different sources. However, EVs hold great promise to replace conventional LDVs affordably.
• Barriers to EV adoption to date include consumer skepticism toward new technology, high purchase prices, limited range and lack of charging infrastructure, and lack of available models and other supply constraints.
• A major challenge facing both manufacturers and end-users of medium- and heavy-duty EVs is the diverse set of operational requirements and duty cycles that the vehicles encounter in real-world operation.
• EVs appear to be well suited for short-haul trucking applications such as regional and local deliveries. The potential for battery-electric models to work well in long-haul on-road applications has yet to be established, with different studies indicating different opportunities.

1.1.2. Batteries and other EV technologies

• Over the last 10 years, the price of lithium-ion battery packs has dropped by more than 80% (from over $1000 kWh −1 to $156 kWh −1 at the end of 2019, BloombergNEF 2020 ). Further price reduction is needed to achieve EV purchase-price parity with ICEVs.
• Over the last 10 years, the specific energy of a lithium-ion battery cell has almost doubled, reaching 240 Wh kg −1 (BloombergNEF 2020 ), reducing battery weight significantly.
• Reducing or eliminating cobalt in lithium-ion batteries is an opportunity to lower costs and reduce reliance on a rare material with controversial supply chains.
• While batteries are playing a key role in the rise of EVs, power electronics and electric motors are also key components of an EV powertrain. Recent trends toward integration promise to deliver benefits in terms of increased power density, lower losses, and lower costs.

1.1.3. Charging infrastructure

• With a few million light-duty EVs on the road, currently, there is about one public charge point per ten battery electric vehicles (BEVs) in U.S. (although most vehicles have access to a residential charger).
• Given the importance of home charging (and the added convenience compared to traditional refueling at public stations), charging solutions in residential areas comprising attached or multi-unit dwellings is likely to be essential for EVs to be adopted at large scale.
• Although public charging infrastructure is clearly important to EV purchasers, how best to deploy charging infrastructure in terms of numbers, types, locations, and timing remains an active area for research.
• The economics of public charging vary with location and station configuration and depend critically on equipment and installation costs, incentives, non-fuel revenues, and retail electricity prices, which are heavily dependent on station utilization.
• The electrification of medium- and heavy-duty commercial trucks and buses might introduce unique charging and infrastructure requirements compared to those of light-duty passenger vehicles.
• Wireless charging, specifically high-power wireless charging (beyond level-2 power), could play a key role in providing an automated charging solution for tomorrow's automated vehicles.

1.1.4. Power system integration

• Accommodating EV charging at the bulk power-system level (generation and transmission) is different in each region, but there are no major known technical challenges or risks to support a growing EV fleet, especially in the near term (approximately one decade).
• At the local level, however, EV charging can increase and change electricity loads significantly, causing possible negative impacts on distribution networks, especially for high-power charging.
• The integration of EVs into power systems presents opportunities for synergistic improvement of the efficiency and economics of electromobility and electric power systems, and EVs can support grid planning and operations in several ways.
• There are still many challenges for effective EV-grid integration at large scale, linked not only to the technical aspects of vehicle-grid-integration (VGI) technology but also to societal, economic, business model, security, and regulatory aspects.
• VGI offers many opportunities that justify the efforts required to overcome these challenges. In addition to its services to the power system, VGI offers interesting perspectives for the full exploitation of synergies between EVs and VRE as both technologies promise large-scale deployment in the future.

1.1.5. Life-cycle cost and emissions

• Many factors contribute to variability in EV life-cycle emissions, mostly the carbon intensity of electricity, charging patterns, vehicle characteristics, and even local climate. Grid decarbonization is a prerequisite for EVs to provide major GHG-emissions reductions.
• Existing literature suggests that future EVs can offer 70%–90% lower GHG emissions compared to today's ICEVs, most obviously due to broad expectations for continued grid decarbonization.
• Operational costs of EVs (fuel and maintenance) are typically lower than those of ICEVs, largely because EVs are more efficient than ICEVs and have fewer moving parts.

1.1.6. Synergies with other technologies and future expectations

• Vehicle electrification fits in broader electrification and mobility macro-trends, including micro-mobility in urban areas, new mobility business models regarding ride-hailing and car-sharing, and automation that complement well with EVs.
• While EVs are a relatively new technology and automated vehicles are not readily available to the general public, the implications and potential synergies of these technologies operating in conjunction are significant.
• The coronavirus pandemic is impacting transportation markets negatively, including those for EVs, but long-term prospects remain undiminished.
• Several studies project major roles for EVs in the future, which is reflected in massive investment in vehicle development and commercialization, charging infrastructure, and further technology improvement. Consumer adoption and acceptance and technology progress form a virtuous self-reinforcing circle of technology-component improvements and cost reductions.
• EVs hold great promise to replace ICEVs affordably for a number of on-road applications, eliminating petroleum dependence, improving local air quality and enabling GHG-emissions reductions, and improving driving experiences.
• Forecasting the future, including technology adoption, remains a daunting task. However, this detailed review paints a positive picture for the future of EVs across a number of on-road applications.

2. Status of electric-LDV market and future projections

This section provides a current snapshot of the electric-LDV market in a global and U.S. context, but focuses on the latter. The global rate of adoption of electric LDVs has increased rapidly since the mid-2010s 13 . By the end of 2019, the global EV fleet reached 7.3 million units—up by more than 40% from 2018—with more than 1.25 million electric LDVs in the U.S. market alone (IEA 2020 ). EV sales totaled more than 2.2 million in 2019, exceeding the record level that was attained in 2018, despite mixed performances in different markets. Electric-LDV sales increased in Europe and stagnated or declined in other major markets, particularly in China (with a significant slowdown due to changes in Chinese subsidy policy in July 2019), Japan, and U.S. U.S. EV adoption varies greatly geographically—nine counties in California currently see EVs accounting for more than 10% of sales (8% on average for California as a whole), but national-level sales remain at less than 3% (Bowermaster 2019 ). BEV sales exceeded those of plug-in hybrid electric vehicles (PHEVs) in all regions.

The rapid increase in EV adoption is underpinned by three key pillars:

• (a)   Improvements and cost reductions in battery technologies, which were enabled initially by the large-scale application of lithium-ion batteries in consumer electronics and smaller vehicles (e.g. scooters, especially in China, IEA 2017 ). These developments offer clear and growing opportunities for EVs and HEVs to deliver a reduced total cost of ownership (TCO) in comparison with ICEVs.
• (b)   A wide range of supportive policy instruments for clean transportation solutions in major global markets (Axsen et al 2020 ), which are mirrored by private-sector investment. These developments are driven by environmental goals (IPCC 2014 ), including reduction of local air pollution. These policy instruments support charging-infrastructure deployment (Bedir et al 2018 ) and provide monetary (e.g. rebates and vehicle-registration discounts) and non-monetary (e.g. access to high-occupancy-vehicle lanes and preferred parking) incentives to support EV adoption (IEA 2018a , AFDC 2020 ).
• (c)   Regulations and standards that support high-efficiency transportation solutions and reduce petroleum consumption (e.g. fuel-economy standards, zero-emission-vehicle mandates, and low-carbon-fuel standards). These regulations are being supported by technology-push measures, consisting primarily of economic instruments (e.g. grants and research funds) that aim to stimulate technological progress (especially batteries), and market-pull measures (e.g. public-procurement programs) that aim to support the deployment of clean-mobility technologies and enable cost reductions due to technology learning, scale, and risk mitigation.

Transport electrification also has started a virtuous self-reinforcing circle. Battery-technology developments and cost reductions triggered by EV adoption provide significant economic-development opportunities for the companies and countries intercepting the battery and EV value chains. Adoption of alternative vehicles both is enabled and constrained by powerful positive feedback arising from scale and learning by doing, research and development, consumer experience and familiarity with technologies (e.g. neighborhood effect), and complementary resources, such as fueling infrastructure (Struben and Sterman 2008 ). In this context, more diversity in make and model market offerings is supporting vehicle adoption. As of April 2020, there are 50 EV models available commercially in U.S. markets (AFDC 2020 ), and ∼130 are anticipated by 2023 (Moore and Bullard 2020 ).

Measures that support transport electrification have been, and increasingly shall be, accompanied by policies that control for the unwanted consequences. Thus, the measures need to be framed in the broader energy and industry contexts.

When looking at the future, EV-adoption forecasts remain highly uncertain. Technology-adoption projections are used by a number of stakeholders to guide investments, inform policy design and requirements (Kavalec et al 2018 ), assess benefits of previous and ongoing efforts (Stephens et al 2016 ), and develop long-term multi-sectoral assessments (Popp et al 2010 , Kriegler et al 2014 ). However, projecting the future, including technology adoption, is a daunting task. Past projections often have turned out to be inaccurate. Still, progress has been made to address projection uncertainty (Morgan and Keith 2008 , Reed et al 2019 ) and contextualize scenarios to explore alternative futures in a useful way. Scenario analysis is used largely in the energy-environment community to explore the possible implications of different judgments and assumptions by considering a series of 'what if' experiments (BP 2019 ).

Adoption of advanced technologies historically has been underestimated in modeling and analysis results (e.g. Creutzig et al 2017 ), which fail to capture the rapid technological progress and its impact on sales. Historical experiences suggest that technology diffusion, while notoriously difficult to predict, can occur rapidly and with an extensive reach (Mai et al 2018 ). Projecting personally owned LDV sales is particularly challenging because decisions are made by billions of independent (not necessarily rational) decision-makers valuing different vehicle attributes based on incomplete information (e.g. misinformation and skepticism toward new technologies) and limited financial flexibility.

Many studies make projections for future EV sales (see figure 1 for a comparison of different projections). Some organizations (e.g. Energy Information Administration [EIA]) historically have been conservative in projecting EV success, mostly due to scenario constraints and assumptions. Still, U.S. EV-sales projections from EIA in recent years are much higher than in the past. Others (e.g. BloombergNEF and Electric Power Research Institute [EPRI]) consistently have been more optimistic in terms of EV sales and continue to adjust sales projections upward. Policy ambition for EV adoption is also optimistic. For example, in September 2020, California passed new legislation that requires that by 2035 all new car and passenger-truck sales be zero-emission vehicles (and that all medium- and heavy-duty vehicles be zero-emission by 2045) (California, 2020). Projected EV sales and outcomes from major energy companies vary widely, ranging from somewhat limited EV adoption (e.g. ExxonMobil) to full market success (e.g. Shell). A survey from Columbia University (Kah 2019 ) considers 17 studies and shows that 'EV share of the global passenger vehicle fleet is not projected to be substantial before 2030 given the long lead time in turning over the global automobile fleet' and that 'the range of EVs in the 2040 fleet is 10% to 70%'. The studies compared in figure 1 show an even greater variability for 2050 projections, ranging from 13% to 100% of U.S. EV adoption for LDVs.

Figure 1.  Electric LDV (BEV and PHEV) new sales projections from numerous international sources. Unless otherwise noted, data refer to new U.S. sales. AEO2015 = EIA Annual Energy Outlook 2015, Reference Scenario. AEO2017 = EIA Annual Energy Outlook 2017, Reference Scenario. AEO2020 = EIA Annual Energy Outlook 2020, Reference Scenario. AEO2020HO = EIA Annual Energy Outlook 2020, High Oil Scenario. EFS Med = National Renewable Energy Laboratory (NREL) Electrification Futures Study, Medium Scenario. EFS High = NREL Electrification Futures Study, High Scenario. EPRI Med = EPRI Plug-in Electric Vehicle Projections: Scenarios and Impacts, Medium Scenario. EPRI High = EPRI Plug-in Electric Vehicle Projections: Scenarios and Impacts, High Scenario. EPRI NEA = EPRI National Electrification Assessment, Reference Scenario. GEVO NP = IEA Global EV Outlook 2019, New Policies Scenario. GEVO CEM = IEA Global EV Outlook 2019, Clean Energy Ministerial 30@30 Campaign Scenario. BNEF = BloombergNEF EV Outlook 2020. Equinor Riv = Equinor 2019 Energy Perspectives, Rivalry Scenario. Equinor Ren = Equinor 2019 Energy Perspectives, Renewal Scenario. Shell Sky = Shell Sky Scenario. ExxonMobil = 2019 ExxonMobil Outlook for Energy. IEEJ Ref = The Institute of Energy Economics, Japan. 2019 Outlook, Reference Scenario (global sales). IEEJ Adv = The Institute of Energy Economics, Japan. 2019 Outlook, Advanced Technologies Scenario (global sales). CA ZEV Mandate = California zero-emission vehicle (ZEV) Executive Order N-79-20 (September 2020).

Download figure:

The future remains uncertain, but there is a clear trend in projections of light-duty EV sales toward more widespread adoption as the technology improves, consumers become more familiar with the technology, automakers expand their offerings, and policies continue to support the market.

A number of studies analyze the drivers of EV adoption (Vassileva and Campillo 2017 , Priessner et al 2018 ) and highlight several barriers for EVs to achieve widespread success, including consumer skepticism for new technologies (Egbue and Long 2012 ); uncertainty around environmental benefits (consumers wonder whether EVs actually are green; see section 7 for more clarity on the environmental benefits of EVs) and continued policy support; unclear battery aging/resale value; high costs (Haddadian et al 2015 , Rezvani et al 2015 , She et al 2017 ); lack of charging infrastructure (Melaina et al 2017 , Narassimhan and Johnson 2018 ); range anxiety (the fear of being unable to complete a trip) associated with shorter-range EVs; longer refueling times compared to conventional vehicles (Franke and Krems 2013 , Neubauer and Wood 2014 ; Melaina et al 2017 ); dismissive and deceptive car dealerships (De Rubens et al 2018 ); and other EV-supply considerations, such as limited model availability and limited supply chains.

A recent review of 239 articles published in top-tier journals focusing on EV adoption draws attention to 'relatively neglected topics such as dealership experience, charging infrastructure resilience, and marketing strategies as well as identifies much-studied topics such as charging infrastructure development, TCO, and purchase-based incentive policies' (Kumar and Alok 2019 ). Similar reviews published recently focus on different considerations, such as market heterogeneity (Lee et al 2019a ), incentives and policies (Hardman 2019 , Tal et al 2020 ), and TCO (Hamza et al 2020 ). Other than some limited discussions on business models and TCO, the literature is focused on one side of the story, namely demand. However, the availability (makes and models) of EVs is extremely limited compared to ICEVs (AFDC 2020 ). This is justified, in part, by new technologies requiring time to be introduced, and, in part, by the higher manufacturer revenues associated with selling and providing maintenance for ICEVs. Moreover, slow turnover in legacy industry (Morris 2020 ) and other supply constraints can be a major barrier to widespread EV uptake (Wolinetz and Axsen 2017 , De Rubens et al 2018 ). Kurani ( 2020 ) argues that in most cases, 'Results of large sample surveys and small sample workshops mutually reinforce the argument that continued growth of PEV markets faces a barrier in the form of the inattention to plug-in electric vehicles (PEVs) of the vast majority of car-owning and new-car-buying households even in a place widely regarded as a leader. Most car-owning households are not paying attention to PEVs or the idea of a transition to electric-drive.'

3. EVs beyond light-duty applications

While much of the recent focus on vehicle electrification is with LDVs and small two- or three-wheelers (primarily in China), major progress also is being made with the electrification of medium- and heavy-duty vehicles. This includes heavy-duty trucks of various types, urban transit buses, school buses, and medium-duty vocational vehicles. As of the end of 2019, there were about 700 000 medium- and heavy-duty commercial EVs in use around the world (EV Volumes 2020 , IEA 2020 ).

A major challenge facing both manufacturers and end-users of medium- and heavy-duty vehicles is the diverse set of operational requirements and duty cycles that the vehicles encounter in real-world operation. When designing powertrain configurations and on-board energy-storage needs for new technologies, it is of critical importance to represent vehicle behavior accurately for different operations, including possible changes triggered by electrification (Delgado-Neira 2012 ). Medium- and heavy-duty vehicles can require a large number of powertrain and battery configurations, control strategies, and charging solutions. These needs depend on vehicle type (covering the full U.S. gross vehicle weight ratings [GVWR] spectrum from class 3 to class 8, 10 001–80 000 lb [4536–36 287 kg]), commercial operational situations and activities, and diverse drive cycles and charging opportunities (e.g. depot-based operations vs. long-haul). An example of this potential variability and its effect on the required battery capacity across multiple vehicle vocations is shown in figure 2 (Smith et al 2019 ).

Figure 2.  Battery capacity requirements vs. weight class for medium- and heavy-duty vehicles (Smith et al 2019 ).

Another example of the highly variable use cases for medium- and heavy-duty EVs shows energy efficiencies range between 0.8 kWh mile −1 and 3.2 kWh mile −1 (0.5–2.5 kWh km −1 ) (Gao et al 2018 ). If the on-board energy-storage needs for these vehicles are considered, assuming a daily operational range of between 50 miles and 200 miles (80–322 km), this results in battery-size requirements between 40 kWh and 640 kWh (assuming that the vehicle is recharged once daily). If additional charging strategies are considered (with their variability in expected charge times and associated power ratings), the range of vehicle-hardware and charging-infrastructure possibilities increases further. When adding variability across use cases with respect to temperature effects, battery-capacity degradation, payload, and road grade, it becomes clear that medium- and heavy-duty truck manufacturers face a significant challenge in designing, developing, and manufacturing systems that are able to meet the diverse operational requirements.

There are potential synergies between components of light-duty and medium- and heavy-duty electric vehicles. However, the requirements of medium- and heavy-duty vehicles place much greater burdens on powertrain components. The power and energy needs in heavy-duty applications are much larger than in light-duty applications. Heavy-duty vehicles could demand twice the peak power, four times the torque, and can consume more than five times the energy per mile (or km) driven compared to LDVs. In addition to using more energy per mile (or km) driven, typically, commercial vehicles drive many more miles (or km) per day, requiring much larger batteries and possibly much higher-power charging. Moreover, heavy-duty-vehicle users expect their vehicles to last more than a million miles, pointing to significantly higher durability requirements for heavy-duty-vehicle components (Smith et al 2019 ). Overall, these requirements, in combination with the needs for very high durability and very high-power drivelines and charging, may cause battery chemistries of heavy-duty vehicle batteries to diverge from those that are used in LDVs, hindering economies of scale. Demands for high efficiency, high power, and lower weight will put pressure on commercial vehicles to work at higher voltages than LDVs do. While LDVs are designed typically with powertrains that operate at a few hundred volts, it may be desirable to design large EVs with kilovolt powertrains. This will have a particularly significant impact on power electronics and could drive the development of wide-bandgap power electronics.

Historically, EVs have not been considered capable alternatives to heavy-duty diesel trucks (above 33 000 lb [14 969 kg] GVWR) due to high capital costs, high energy and power requirements, and weight and range-related battery constraints. International Council on Clean Transportation (ICCT), for example, suggests that while conventional EV-charging methods may be sufficient for small urban commercial vehicles, overhead catenary or in-road charging are required for heavier vehicles (Moultak et al 2017 ). Recent studies dispute this, anticipating a much greater opportunity for EVs to replace diesel trucks in the short-term, even for long-haul applications (Mai et al 2018 , McCall and Phadke 2019 , Borlaug et al Forthcoming ), but the potential for battery-electric models to work well in long-haul applications has yet to be established (NACFE 2018 ). Studies show that a significant amount of payload capacity will be consumed by batteries, potentially up to 7 tons or 28% of capacity in a truck with a 500 mile (805 km) range with 1100 kWh battery capacity (Burke and Fulton 2019 ). Thus, batteries would reduce significantly the amount of cargo that can be carried. Other studies suggest this could be much less―on the order of 4% of lost payload capacity for 500 mile range (805 km) trucks and with overall lower TCO than diesel trucks (Phadke et al 2019 ). For short-haul applications, such as port drayage and regional or local deliveries, EVs appear well suited and battery weight may not affect the cargo or payload capacity adversely. Several heavy-duty battery-electric trucks for short- and medium-haul applications have been developed and tested in recent years by Balqon, Daimler Trucks NA, Peterbilt, TransPower, Tesla, US Hybrid, Volvo, and others (AFDC 2020 ).

Urban buses are also a major emerging market for electrification. In California, Innovative Clean Transit rules require transit agencies to transition completely to zero-emission technologies (batteries or fuel cells), with all new bus purchases being zero-emission by 2029 (CARB 2018 ). Eight of the ten largest transit agencies in California already are adopting zero-emission technologies into their fleets (CARB 2018 ). In a comparative study of urban buses running on diesel, compressed natural gas, diesel hybrid, fuel cells, and batteries, the battery buses are estimated to have the lowest CO 2 emissions in both California and Finland bus duty cycles at the time of the study (Lajunen and Lipman 2016 ). This study also shows that battery buses have only slightly higher overall costs per mile (or km) than fossil-fuel-based alternatives. Future projections out to 2030 show that electric buses have the lowest overall life-cycle costs, especially when CO 2 costs are included (Lajunen and Lipman 2016 ).

Medium-duty delivery vehicles (typically 10 000–33 000 lb [4536–14 969 kg] GVWR) are another attractive emerging area for electrification. The goods-delivery market is growing at approximately 9% per year in recent years, with a projected $343 billion global industry value in 2020 (Accenture 2015 ). The 'last mile' delivery vehicles that are needed for this market are undergoing changes and present good opportunities for electrification. Amazon, for example, has announced plans to purchase 100 000 custom-designed Rivian electric delivery vans by 2030, with 10 000 of the vehicles delivered by late 2022 (Davies 2019 ).

A significant challenge with electrifying these heavy- and medium-duty vehicles revolves around the installation of the required charging infrastructure (either at depots or along highways). While LDVs typically charge at power levels of 3 kW–10 kW, and potentially 50 kW–250 kW with DC fast chargers (DCFCs), a heavy-duty vehicle may require higher-power charging, depending on its duty cycle. Fleets of these vehicles charging in one location, such as a truck depot or travel center, may require several megawatts of power. This requires expensive charging infrastructure, potentially including costly and time-consuming distribution-grid upgrades, to provide the higher voltage and current levels that are needed. For example, a single 350 kW DCFC that may be suitable for heavy-duty applications costs almost $150 000 today (Nelder and Rogers 2019 , Nicholas 2019 ). These costs would, in turn, impact the business case for vehicle electrification. Potential costs of grid upgrades to support these new electrical loads would be additional expenses that may or may not be supported by the local utility, depending on the circumstances. To enable reliable, low-cost charging, which is crucial when considering the TCO for a fleet owner, the installation and operational costs of the charging infrastructure must be optimized, requiring engagement with power-supply stakeholders.

4. Batteries, power electronics, and electric machines

Electrification is a key aspect of modern life, and electric motors and machines are prevalent in manufacturing, consumer electronics, robotics, and EVs (Zhu and Howe 2007 ). One reason for the recent success and rise in adoption of EVs is the use of advanced lithium-ion batteries with improved performance, life, and lower cost. Improved energy and power performance, increased cycle and calendar life, and lower costs are leading to EVs with longer electric range and better acceleration at lower cost premia that are attracting consumers. This section summarizes the state-of-the-art for batteries and for power electronics, electric machines, and electric traction drives in terms of cost, performance, power and energy density, and reliability, and highlights some research challenges, pathways, and targets for the future.

4.1. Batteries

Over the last 10 years, the price of a lithium-ion battery pack has dropped by almost 90% from over $1000 kWh −1 in 2010 to $156 kWh −1 at the end of 2019 (BloombergNEF 2020 ). Meanwhile, the specific energy of a lithium-ion battery cell has almost doubled from 140 Wh kg −1 to 240 Wh kg −1 during that same window of time (BloombergNEF 2020 ). The improvement in performance and cost comes mainly from engineering improvements, use of materials with higher capacities and voltages, and development of methods to increase stability for longer life and improved safety. Improvements in cell, module, and pack design also help to improve performance and lower costs. Increases in manufacturing volume due to EV sales contribute significantly to cost reductions (Nykvist and Nilsson 2015 , Nykvist et al 2019 ). However, further reductions in battery costs, along with a reduction in the cost of electric machines and power electronics, are needed for EVs to achieve purchase-price parity with ICEVs. This parity is estimated by U.S. Department of Energy (DOE) to be achieved at battery costs of ∼$100 kWh −1 (preferably less than $80 kWh −1 ) (VTO, 2020 ). At that point, EVs should have both a purchase- and a lifetime-operating-cost benefit over ICEVs. Such cost benefits are likely to trigger drastic increases in EV sales. Figure 3 shows the observed price of lithium-ion battery packs from 2010 to 2018, as well as estimated prices through 2030. BloombergNEF projects that by 2024 the price for original equipment manufacturers (OEMs) to acquire battery packs will go below $100 kWh −1 and reach ∼$60 kWh −1 by 2030 if high levels of investments continue in the future (BloombergNEF 2020 ).

Figure 3.  Evolution of battery prices over the last 10 years and future projections (Goldie-Scot 2019 ). BloombergNEF 2019.

The typical anode material that is used in most lithium-ion EV batteries is graphite (Ahmed et al 2017 ). Research is underway to utilize silicon, in addition to graphite, due to its higher specific-energy capacity. For cathodes, there is more variety (Lee et al 2019 , Manthiram 2020 ). Consumer electronics such as mobile phones and computers almost exclusively have used lithium cobalt oxide, LiCoO 2 , due to its high specific-energy density (Keyser et al 2017 ). Most EV manufacturers (except Tesla) have avoided using LiCoO 2 in EVs due to its high cost and safety concerns. Lithium iron phosphate also has been used for electric cars and buses because of its long life and better safety and power capabilities. However, due to its low specific-energy density (110 Wh kg −1 ) when paired with a graphite anode, lithium iron phosphate is not used commonly for light-duty EVs in U.S. In recent years, battery makers and vehicle OEMs have moved to lithium nickel manganese cobalt oxides (NMC) with varying ratios of the three transition metals. Initially, OEMs used NMC111 (the numbers represent the molar fractions of nickel, manganese, and cobalt, which are equal in this case), but they have transitioned to NMC532 and utilize NMC622 now while working to stabilize the NMC811 cathode structure. The goal is eventually to reduce the amount of cobalt in the cathode to less than 5% and perhaps even eliminate the use of cobalt. The use of these cathodes with higher specific-energy density and less cobalt leads to lower battery cost per unit energy ($ kWh −1 ). Table 1 shows the specific energy and estimated (bottom-up) cost from Argonne National Laboratory's BatPaC Battery Performance and Cost model (Ahmed et al 2016 ) based on large-volume material processing, cell manufacturing, and pack manufacturing.

Table 1.  Calculated specific energy and cost of advanced lithium-ion batteries with different cathode/anode chemistries. Numbers are from BatPaC (Ahmed et al 2016 ) and are intended for relative comparison only. Final values can change depending on the components used and production volume, and costs reported do not reflect what a negotiated price could be between a battery and EV maker.

The cost of batteries is expected to decline in the future due to improved capacity of materials (such as Si anodes), increased percentage of active material components, use of lower-cost elements (no cobalt), improved packaging, and continued automation to increase yield while leading to a longer electric range. However, price increases for certain metals such as Ni and Li could prevent achieving those lower-battery-cost projections. Moreover, different battery chemistries can lead to very different costs and specific energies. For example, table 1 shows results obtained from bottom-up calculations with Argonne National Laboratory's BatPaC Battery Performance and Cost Model (Ahmed et al 2016 ), for a 100 kWh battery pack showing great variability in battery cost and performance for different chemistries.

Opportunities to improve performance and reduce costs further are being pursued in a number of major research areas. The battery community is investigating a number of materials, with the aim of reducing the cost and increasing the energy density of battery systems (Deign and Pyper 2018 ). Future work will involve utilizing silicon (Salah et al 2019 ) or lithium metal (Zhang et al 2020 ) as the anode while utilizing high-energy cathodes, such as NMC811 or lithium sulfur (Zhu et al 2019 ). Reducing the amount of critical material in lithium-ion batteries, especially cobalt, is an opportunity to lower the cost of batteries and improve supply-chain resilience. The private and public sectors are working toward developing new cathode materials along these lines (Li et al 2009 , 2017b ). Research and development (R&D) projects are underway to develop infrastructure and recycling technologies to collect batteries and recover the key battery materials economically and environmentally (Harper et al 2019 ). Reuse of end-of-life batteries from EVs would delay the need for additional battery materials, which should have positive environmental benefits (Neubauer et al 2012 ). Different battery technologies also are being explored. To increase energy density, reduce cost, and improve safety, the battery community is pursuing development of solid-state batteries with solid-state electrolytes (Randau et al 2020 ) that have ionic conductivities approaching those of today's liquid electrolyte systems. Solid-state lithium batteries enable the use of metallic lithium anodes, together with solid electrolytes and high-energy cathodes (such as high-nickel NMC or sulfur). Lithium-metal batteries based on solid electrolytes can, in principle, alleviate the safety concerns with current lithium-ion batteries with a flammable organic electrolyte. The main challenges facing lithium-metal anodes are dendritic growth, especially at low temperatures and higher current rates. Dendritic growth could lead to short circuit and thermal runaway and low Coulombic efficiency leading to poor cycle life (Xia et al 2019 ). Slow ion transport through the solid-state electrolyte leading to low power densities and manufacturing challenges, including poor mechanical integrity, pose additional challenges. Significant R&D activities are focused on developing solid-state electrolytes that prevent dendrite growth, have high ionic conductivity, good voltage-stability windows, and low impedance at the electrode–electrolyte interface. Recent cathode formulations in Li-S cells overcome the polysulfide problem, which could lead to lower efficiency and cycle life. Nevertheless, the deployment of cells with lower electrolyte-to-sulfur ratios for scale-up to large sizes is a remaining challenge. It may take another 5 to 10 years to mass-produce solid-state lithium batteries for EV applications.

As is discussed in section 5 , a network of fast chargers and batteries that can handle high charging-power rates is needed to address any potential barriers to widespread EV adoption. Research is focusing on developing batteries that can be charged very quickly (e.g. 80% of capacity in less than 15 min). A number of challenges to high-power charging, such as lithium plating, thermal management, and poor cycle life, need to be addressed (Ahmed et al 2017 ; DOE 2017 , Michelbacher et al 2017 ). Significant efforts also have focused on developing electrochemical and thermal modeling of batteries for EV applications (Kim et al 2011 , Chen et al 2016 , Keyser et al 2017 , Zhang et al 2017 ) to improve battery lifetime and efficiency in real-world applications. These efforts include lifetime-estimation and degradation modeling under different real-world climate and driving conditions (Hoke et al 2014 , Neubauer and Wood 2014 , Liu et al 2017b , Harlow et al 2019 , Li et al 2019 ); simplified models for control and diagnostics (e.g. state-of-charge estimation) (Muratori et al 2010 , Fan et al 2013 , Cordoba-Arenas et al 2015 , Bartlett et al 2016 ); and developing effective thermal management and control strategies (Pesaran 2001 , Serrao et al 2011 ).

Besides EV applications, batteries can offer energy-storage solutions for hybrid- or distributed-energy systems. These solutions include the use of batteries in integrated configurations with wind or solar photovoltaic (PV) systems or with EV fast-charging stations (Bernal-Agustín and Dufo-Lopez 2009 , Badwawi et al 2015 , Muratori et al 2019a ). Batteries also can provide stabilization and flexibility and can improve resilience and efficiency for power systems in general, especially for critical services or when a high share of variable power generation (e.g. from solar or wind) is expected (Divya and Østergaard 2009 , Denholm et al 2013 ; De Sisternes et al 2016 ). Lithium-ion batteries that have been developed for EV applications have found their way into stationary applications (Pellow et al 2020 ) because of their lower cost and modularity compared to other energy-storage technologies (Chen et al 2020 ). Moreover, EV batteries can be reused or repurposed at the end of their 'vehicle life' (usually considered when energy storage capacity drops below 70%–80% of the original nominal value, (Podias et al 2018 )) for stationary applications, improving their economic and environmental performance (Assuncao et al 2016 , Ahmadi et al 2017 , Martinez-Laserna et al 2018 , Olsson et al 2018 , Kamath et al 2020 ).

4.2. Power electronics, electric machines, and electric-traction-drive systems

While batteries are playing a key role in the rise of EVs, power electronics and electric motors and machines are also key components of an EV powertrain. Traditionally, the motor and power electronics drive were separate components in an EV. However, recent trends toward integration promise to deliver benefits in terms of increased power density, lower losses, and lower costs compared with separate motor and motor-drive solutions (Reimers et al 2019 ). Figure 4 shows the 2020 power density for power electronics, electric machines, and electric-traction-drive system from some example commercial vehicles as well as the 2025 DOE and U.S. DRIVE Partnership targets for near-term improvements (U.S. DRIVE 2017 , Chowdhury et al 2019 ). Commercially available vehicles exceed the 2020 power-density target. However, the 2025 target is at least a factor of six to eight higher than current commercial baselines. U.S. DRIVE Partnership also proposes electric-traction-drive-cost targets for 2020 and 2025: $8 kW −1 and $6 kW −1 , respectively, both of which are challenging targets (U.S. DRIVE 2017 , Chowdhury et al 2019 ). The authors are not aware of commercial systems meeting the 2020 target, and the 2025 target represents a further 33% reduction.

Figure 4.  Integrated electric-drive system and inverter power density for several commercial light-duty vehicles and DOE targets (data from U.S. DRIVE 2017 , Chowdhury et al 2019 ).

Improvements in compact power electronics and electric machines are applicable to novel emerging wheel-integrated solutions as well (Iizuka and Akatsu 2017 , Fukuda and Akatsu 2019 ). The development of advanced electric traction drive with improved efficiency is a strategy for increasing the range of electric-drive vehicles. In addition to this, chassis light-weighting is another strategy that is being pursued by the industry and the research community for increasing EV driving ranges. There are several technical challenges in meeting the DOE power-density targets (shown in figure 4 ). Challenges in meeting related DOE cost targets remain as well. A range of integration approaches are proposed in the literature, including surface mounting the power electronics on the motor housing (Nakada, Ishikawa, and Oki 2014 ), mounting on the motor stator iron (Wheeler et al 2005 ), and piecewise integration. Piecewise integration involves modularizing both power modules and machine stators into smaller units (Brown et al 2007 ). In all cases, the close physical positioning of the power electronics relative to the machine and the associated harsh thermal environment necessitate new concepts related to the active cooling of both components. A first strategy may be to isolate the power electronics from the machine thermally using parallel cooling mechanisms (Wheeler et al 2005 ). Another approach may be to use a fully integrated, series-connected, active-cooling loop (Tenconi et al 2008 , Gurpinar et al 2018 ). In either case, cost benefits may be realized through the possible elimination or combination of cooling loops. Significant research also has been focused on reducing rare-earth and heavy-rare-earth materials within the electric machines because that is an additional important pathway to reduce costs (U.S. DRIVE 2017 ).

Higher levels of integration go hand-in-hand with the utilization of wide-bandgap (WBG) semiconductor devices, which may be used at higher operational temperatures (e.g. >200 °C versus 150 °C for silicon) with reduced switching loss (Millán et al 2014 ). However, the adoption of WBG devices requires new packaging technologies to support the end goals of high temperature, high frequency, higher voltages, and more compact footprints. High-performance electrical interconnects (Cheng et al 2013 ), die-attach (Liu et al 2020 ), encapsulation (Cao et al 2010 ), and power-module-substrate technologies (Stockmeier et al 2011 ), along with thermal management and reliability of these technologies (Moreno et al 2014 , Paret et al 2016 , 2019 ), are critical aspects to consider. The new materials, devices, and components must be cost-effective and high-temperature-capable to be compatible with WBG devices. The downsizing of passive electrical components is another added benefit of adopting WBG devices and a further necessity for integrated machine-drive packaging solutions. Fortunately, the higher switching frequencies that are supported by WBG devices enable the downsizing of both the inductors and capacitors found in a traditional power-control unit (Hamada et al 2015 ). The development of economically viable and high-temperature-capable passives, capacitors in particular (Caliari et al 2013 ), is an area of great interest.

Besides EV applications, power electronics and electric machines with low cost, high performance, and high reliability are important for numerous energy-efficiency and renewable-energy applications, such as solar inverters, generators and electric drives for wind, grid-tied medium-voltage power electronics, and sensors and electronics for high-temperature geothermal applications (PowerAmerica 2020 ).

5. Charging infrastructure

Infrastructure planning and deploying an ecosystem of cost-effective and convenient public and private chargers is central to supporting EV adoption (CEM 2020 ). The lack of a sufficient refueling infrastructure has hampered many past efforts to promote alternatives to petroleum fuels (McNutt and Rodgers 2004 ). Extensive research is being done to address the diverse challenges that are posed by a transition from fossil-fuelled ICEVs to EVs and the special role of charging infrastructure in this transition (Muratori et al 2020b ).

At the end of 2019, there were an estimated 7.3 million EV chargers (or plugs) worldwide, of which almost 0.9 million were public, including approximately 264 000 public DCFCs (81% in China) (IEA 2020 ). Significant government support and private investments are helping to expand the network of public charging stations worldwide. With about 7.2 million light-duty BEVs on the road, there is about one public charger per 10 light-duty BEVs, and most vehicles have access to a residential charger. However, the number of public chargers per BEV varies widely among the 10 countries with the most BEVs (figure 5 ) because of different strategies for deploying fast versus slow public chargers. In addition to these LDV chargers, IEA estimated there are 184 000 fast chargers dedicated to electric buses (95% in China).

Figure 5.  Public charging availability by country in 2019, measured as Level-1 and Level-2 chargers per BEV and DCFC per 10 BEVs (Data from IEA 2020 ).

Studies show consistently that today's EVs do the majority (50%–80%) of their charging at home, followed by at work (15%–25% when workers use their vehicles to commute), and using public chargers (only about 5% of charging) (Hardman et al 2018 ). PHEVs conduct more charging at home than BEVs do, and they rely more on level-1 charging (Tal et al 2019 ). While single-household detached residences readily can accommodate level-1 or -2 charging, multi-unit dwellings require curbside public charging or installations in shared parking facilities (Hall and Lutsey 2017 ). Historical data on the charging behavior of California BEV owners reveals that 11% of their charging sessions were at level 1, 72% were at level 2, and 17% used DCFCs (Tal et al 2019 ). Use of DCFCs is lowest for BEVs with less than 100 miles (161 km) of range, highest for medium-range BEVs, and lower again for BEVs with ranges of 300 miles (483 km) or more.

5.1. Charging-siting modeling

Public charging infrastructure is clearly important to EV purchasers and supports EV sales by adding value (Narassimhan and Johnson 2018 , Greene et al 2020 ). However, how best to deploy charging infrastructure, in terms of numbers, types, locations, and timing remains an active area for research (Ko et al 2017 , Funke et al 2019 provide reviews). The literature includes many examples of geographically and temporally detailed models to optimize the location, number, and types of charging stations (e.g. Wood et al 2017 , Wu and Sioshansi 2017 , Zhao et al 2019 ). Geographically and temporally detailed data recording the movements of PEVs and their charging behavior are scarce. With few exceptions (e.g. Gnann et al 2018 ), simulation analyses rely on conventional ICEV databases (e.g. Dong et al 2014 , Wood et al 2015 , 2018 ), which do not reflect the changes PEV owners will make to maximize the utility of PEVs.

Given the importance of home charging, access to chargers for on-street parking in residential areas comprising attached or multi-unit dwellings is likely to be essential for PEVs to be adopted at large scale. Grote et al ( 2019 ) employ heuristic methods with geographical-information systems to locate curbside chargers in urban areas using a combination of census and parking data. The works of Nie and Ghamami ( 2013 ), Ghamami et al ( 2016 ), and Wang et al ( 2019 ) are examples of the variety of optimization methods that are applied to design DCFC networks to support intercity travel. Despite these examples, applied research is hindered by the scarcity of data on long-distance vehicle travel by PEVs (Eisenmann and Plötz 2019 ). Jochem et al ( 2019 ) estimate that 314 DCFC stations could provide minimum coverage of EU intercity routes with approximately 0.7 charging points per 1000 BEVs. Using a database of simulated U.S. intercity travel, He et al ( 2019 ) employ a mixed-integer model to optimize the location and number of DCFCs. They conclude that 250 stations could serve 98% of the long-distance miles of BEVs with ranges of 150 miles (241 km) or greater but only 73% of the long-distance miles of 100 mile range (161 km range) BEVs. Similarly, Wood et al ( 2017 ) estimate that 400 DCFC stations are required to cover the U.S. interstate-highway network with a 40 mile (64 km) spacing between stations. Others consider the optimal location of dynamic, wireless charging in combination with stationary charging (Liu and Wang 2017 ).

Optimization models for locating chargers to support commercial PEV fleets also appear in the literature (Jung et al 2014 , Shahraki et al 2015 ). In the future, if vehicle sharing becomes much more common, the downtime for charging could be an important disadvantage for PEVs. Using an integer model to optimize station allocation and PEV assignment, Roni et al ( 2019 ) find that charging time represents 72%–75% of vehicle downtime but that charging time could be reduced by almost 50% by optimal deployment of charging stations.

5.2. Beyond LDV charging

The electrification of medium- and heavy-duty commercial trucks and buses introduces unique charging and infrastructure requirements compared to those of LDVs. These requirements stem from the significantly higher battery capacities required on-board the vehicles, potentially shorter charging-dwell times (due to the in-service time requirements of the vehicles), and the potential of large facility charging loads (due to multiple trucks or buses charging in one location). One challenge is to understand the costs associated with the multitude of charging scenarios for commercial vehicles for current operations as well as future operations. It is expected that on-road freight vehicle miles (or km) traveled will increase by 75% from 2012 to 2045 (McCall and Phadke 2019 ). This increase may bring about new business models and potentially new charging-infrastructure approaches to meet this demand with electrified trucks. California's Innovative Clean Transit regulation, which will require California transit agencies to adopt zero-emission buses by 2040, is likely to drive large charging-infrastructure investments for buses (CARB 2018 ).

Today's commercial diesel-powered trucks in small fleets typically are fueled at publicly available on-road fueling stations, while nearly half of trucks in fleets of 10+ vehicles use company-owned facilities (Davis and Boundy 2020 ). Likewise, commercial EVs are charged primarily in fleet-owned facilities as their daily schedule allows (most often overnight). This depot-charging approach, which enables seamless integration of EVs into fleet logistics, might limit the electrification of some vehicle segments in the long term due to the battery capacity that is needed to satisfy their daily-range requirements (the need to complete their full-day function) and return to the facility to recharge fully 14 . Some studies suggest that long-haul battery-electric trucks are technically feasible and economically compelling (Phadke et al 2019 ) while others are more skeptical (Held et al 2018 ). Publicly available, high-power charging or en-route charging infrastructure for commercial vehicles could enable electrification for longer-distance vehicles (by enabling smaller on-board battery-capacity needs), but this scenario has cost challenges. En-route, high-power charging of over 1 MW might be needed to enable 500 miles (805 km) or more of daily driving. Installation of a 20 MW truck-charging station in California (capable of multiple 1.5 MW charge events for heavy-duty freight vehicles) is estimated to cost as much as 15 million USD. McCall and Phadke ( 2019 ) estimate that as many as 750 of these stations are needed to electrify the fleet of California Class-8 combination trucks. Charging commercial vehicles at depots requires additional infrastructure costs to install lower-power EV-supply equipment networks (e.g. 50 kW–100 kW) capable of charging multiple vehicles at these lower rates. These depot charging systems also will challenge existing facility electrical systems by adding a significant load that was not planned previously at the facility (Borlaug et al Forthcoming ).

5.3. Economics of public charging

PEV-charging economics vary with location and station configuration and depend critically on equipment and installation costs and retail electricity prices, which are dependent on utilization (Muratori et al 2019b , Borlaug et al 2020 ). In the early stages of market development, when there are relatively few vehicles, future demand is uncertain, and most charging is done at an EV's home base (Nigro and Frades 2015 , Madina et al 2016 ). Public charging stations tend to be lightly used during these initial stages (e.g. INL 2015 ), which poses a difficult challenge for private investment. Understanding and quantifying the value of public charging is hindered by lack of experience with PEVs on the part of consumers (Ito et al 2013 , Greene et al 2020 , Miele et al 2020 ) and the complexity of network effects in the evolution of alternative-fuel-vehicle markets (Li et al 2017a ). Nevertheless, it is likely that DCFCs will be profitable with sufficient demand. Considering vehicle ranges of between 100 km and 300 km and charging-power levels of between 50 kW and 150 kW, Gnann et al ( 2018 ) conclude that charger-usage fees could be between 0.05 € kWh −1 and 0.15 € kWh −1 in addition to the cost of electricity. The estimates were based on simulations with average daily occupancy of charging points of 10%–25% and peak-hour utilization of 20%–70%. In their simulations, utilization rates increase with increasing charger power and decrease with increased EV range. For intercity travel along European Union highways, Jochem et al ( 2019 ) estimate that a surcharge of 0.05 € kWh −1 of DCFC would make a minimal coverage of 314 stations (with 20 charge plugs each) profitable, even for station capital costs of one million EUR. He et al ( 2019 ) optimize DCFC locations along U.S. intercity routes and conclude that providing an adequate nationwide charging network for long-distance travel by 100 mile (161 km) range BEVs is more economical than increasing vehicle range and reducing the number of charging stations. Muratori et al ( 2019a ) consider a set of charging scenarios from real-world data and thousands of U.S. electricity retail rates. They conclude that batteries can be highly effective at mitigating electricity costs associated with demand charges and low station utilization, thereby reducing overall DCFC costs.

Early estimates show that the cost of public DCFC in U.S. can vary widely based on the station characteristics and level of use (Muratori et al 2019a ). Numerous new technology options are being explored to provide lower-cost electricity for light-duty passenger and medium- and heavy-duty commercial BEVs. Increasing the range of EVs through higher-power public charging stations as well as accommodating new potential BEV business models, such as transportation-network companies or automated vehicles, are driving new charging-technology solutions. Managed charging solutions that are available today can provide increased value to the BEV owner (lower electricity costs), charging station owner (lower operating costs), or grid operator (lower infrastructure-investment costs). For example, a managed-charging solution has been adopted and is currently in operation at a Santa Clara Valley Transportation Authority depot to charge a fleet of Proterra electric buses optimally to ensure minimal stress on the grid (Ross 2018 ).

5.4. Emerging charging technologies

Wireless charging, specifically high-power wireless charging (beyond level-2 power levels), could play a key role in providing an automated charging solution for tomorrow's automated vehicles (Lukic and Pantic 2013 , Qiu et al 2013 , Miller et al 2015 , Feng et al 2020 ). Wireless charging also can enable significant electric range for BEVs by providing en-route opportunity charging (static or dynamic charging opportunities). If a network of wireless charging options is available to provide convenient and fast en-route charging, it could help reduce the amount of battery that is needed on-board a vehicle and reduce the cost of ownership for a BEV owner. Wireless charging is being developed for power levels of up to 300 kW for LDVs, 500 kW for medium-duty vehicles, and 1000 kW for heavy-duty vehicles. Bidirectional functionality, improved efficiency, interoperability of different systems, improved cybersecurity, and increased human-safety factors continue to be developed (Ozpineci et al 2019 ).

Connectivity and communication advances will enable new BEV-charging infrastructure and managed charging solutions. However, emerging cybersecurity threats also are being identified and should be addressed. There are concerns associated with data exchange, communications network, infrastructure, and firmware/software elements of the EV infrastructure (Chaudhry and Bohn 2012 ), and new charging-system security requirements and protocols are being developed to address these concerns (ElaadNL 2017 ). New emulation and simulation platforms also are being developed to address these threats and help understand the consequences and value of mitigating cyberattacks that could affect BEVs, electric-vehicle-supply equipment, or the electric grid (Sanghvi et al 2020 ).

6. Vehicle-grid integration (VGI)

Connecting millions of EVs to the power system, as may occur in the coming decades in major cities, regions, and countries around the world, introduces two fundamental themes: (a) challenges to meet reliably overall energy and power requirements, considering temporal load variations, and (b) VGI opportunities that leverage flexible vehicle charging ('smart charging') or V2G services to provide power-system services from connected vehicles. Multiple studies, which are reviewed in detail below, investigate the potential load growth, impact on load shapes, and infrastructure implications of increased EV adoption. These works focus especially on impact on distribution systems and opportunities for flexible charging to reshape aggregate power loads. Mai et al ( 2018 ), for example, shows that in a high-electrification scenario, transportation might grow from the current 0.2% to 23% of total U.S. electricity demand by 2050. This growth would impact system peak load and related capacity costs significantly if not controlled properly. In-depth analytics indicate a complex decision framework that requires critical understanding of potential future mobility demands and business models (e.g. ride-hailing, vehicle sharing, and mobility as a service), technology evolution, electricity-market and retail-tariff design, infrastructure planning (including charging), and policy and regulatory design (Codani et al 2016 , Eid et al 2016 , Knezovic et al 2017 , Borne et al 2018 , Hoarau and Perez 2019 , Gomes et al 2020 , Muratori and Mai 2020 , Thompson and Perez 2020 ).

While accommodating EV charging at the bulk-power (generation and transmission) level will be different in each region, no major technical challenges or risks have been identified to support a growing EV fleet, especially in the near term (FleetCarma 2019 , U.S. DRIVE 2019 , Doluweera et al 2020 ). At the same time, many studies show that smart charging and V2G create opportunities to reduce system costs and facilitate VRE integration (Sioshansi and Denholm 2010 , Weiller and Sioshansi 2014 , IRENA 2019 , Zhang et al 2019 ). Therefore, charging infrastructure that enables smart charging (e.g. widespread residential and workplace charging) and alignment with VRE generation and business models and programs to compensate EV owners for providing charging flexibility are critical elements for successful integration of EVs with bulk power systems.

6.1. Impact of EV loads on distribution systems

At the local level, EV charging can increase and change electricity loads significantly, having possible negative impacts on distribution networks (e.g. cables and distribution transformers) and power quality or reliability (Khalid et al 2019 ). Residential EV charging represents a significant increase in household electricity consumption that can require upgrades of the household electrical system which, unless managed properly, may exceed the maximum power that can be supported by distribution systems, especially for legacy infrastructure and during times of high electricity utilization (e.g. peak hours and extreme days) (IEA 2018b ). The impact of EVs on distribution systems also is influenced by the simultaneous adoption of other distributed energy resources, e.g. rooftop PV panels. While this interdependency complicates assessing the impact of EV charging, Fachrizal et al ( 2020 ) show that the two technologies support one other. Similarly, Vopava et al ( 2020 ) show that line overloads caused by rooftop PV panels can be reduced (but not avoided) by increasing EV adoption and vice versa.

The impact of EV charging on distribution systems is particularly critical for high-power charging and in cases in which many EVs are concentrated in specific locations, such as clusters of residential LDV charging and possibly fleet depots for commercial vehicles (Saarenpää et al 2013 , Liu et al 2017a , Muratori 2018 ). Smart charging, by which EV charging is timed based on signals from the grid and electricity prices that vary over time, or other forms of control, can help to minimize the impact of EV charging on distribution networks. However, smart charging requires both appropriate business models and signals (with related communication and distributed-control challenges). The market for distribution-system operators to provide such services is not mature yet (Everoze 2018 , Crozier, Morstyn, and McCulloch 2020 ). Time-varying pricing schemes, which are effective at influencing the timing of EV charging (PG&E 2017 ), typically do not include any distribution-level considerations. Thus, while consumers are responsive to such signals, the business models to include distribution-level metrics still are lacking. Moreover, price signals are offered usually to a large consumer base with the intent of reshaping the overall system load. At the local level, however, multiple consumers responding to the same signal might cause 'rebound peaks' (Li et al 2012 , Muratori and Rizzoni 2016 ) that can overstress distribution systems, calling for coordination among consumers connected to the same distribution network (e.g. direct EV-charging control from an intermediate aggregator).

Charging of larger commercial vehicles and highway fast-charging stations typically involves higher power levels: DCFC is typically at 50 kW/plug today, but power levels are increasing rapidly. Commercial charging locations with multiple plugs co-located at a specific location may lead to possible MW-level loads, which is roughly equivalent to the peak load of a large hotel. Commercial DCFC may require costly upgrades to distribution systems that can impact the cost-effectiveness of public fast charging heavily, especially if stations experience low utilization (Garrett and Nelder 2016 , Muratori et al 2019b ). While charging timing and speed at commercial stations is less flexible (consumers want to charge and leave or commercial fleets must meet business requirements), business models are often already in place to incentivize curbing maximum peak power from commercial installations. For example, demand charges (a fixed monthly payment that is proportional to the peak power that is drawn during a given month) are fairly common in U.S. retail tariffs and provide a reason to limit peak power. Furthermore, Muratori et al ( 2019a ) show that distributed batteries can be effective at mitigating the cost associated with demand charges by up to 50%, especially for 'peaky' or low-utilization EV-charging loads. Batteries also can facilitate coupling EV-charging stations with local solar electricity production or can provide grid services (Megel, Mathieu, and Andersson 2015 ), generating additional revenue.

6.2. Value of managed ('smart') EV charging for power systems

The integration of EVs into power systems presents several opportunities for synergistic improvement of the efficiency and economics of electromobility and electric power systems. These synergies stem from two inherent characteristics of EVs and power systems. Demand response and other forms of demand-side flexibility can be of value for power-system planning and operations (Albadi and El-Saadany 2007 , 2008 , Su and Kirschen 2009 , Muratori et al 2014 ). Contemporaneously, most personal-vehicle driving patterns entail vehicle-use for mobility purposes a relatively small proportion of the time (Kempton and Letendre 1997 ). If EVs are grid-connected for extended periods of time, they can provide demand-side flexibility in the form of smart charging or V2G services. Such use of an EV can improve its economics by leveraging cheaper electricity at little incremental cost (e.g. the costs of monitoring, communication, and control equipment that are needed to manage smart charging). EVs can support grid planning and operations in a number of ways. Figure 6 summarizes the key support services that EVs can provide. These services include reducing peak load and generation-, transmission-, and distribution-capacity requirements, deferring system upgrades, providing load response, supporting power-system dispatch (including VRE integration and real-time energy and operating reserves), providing energy arbitrage, and supporting power quality and end retail consumers.

Figure 6.  Summary of opportunities for EVs to provide demand-side flexibility to support power system planning and operations across multiple timescales.

Habib et al ( 2015 ) and Thompson and Perez ( 2020 ) provide detailed surveys of different potential uses of EVs for smart charging and V2G services. This includes active- and reactive-power services, load balancing, power-quality-related services (e.g. managing flicker and harmonics), retail-bill management, resource adequacy, and network deferral. In addition, Habib et al ( 2015 ) discuss different standards and technology needs relating to V2G services.

Kempton and Letendre ( 1997 ) provide the first description of the concept of EVs providing grid services, either in the form of smart charging or bidirectional V2G services (which can involve discharging EV batteries). Denholm and Short ( 2006 ) study the benefits of controlled overnight charging of PHEVs for valley-filling purposes. They demonstrate that with proper control of vehicle charging, up to 50% of the vehicle fleet could be electrified without needing new generation capacity to be built and at substantial savings compared to using liquid fuels for transportation. They show also that under conservative utility-planning practices, PHEVs could replace a significant portion of low-capacity-factor generating capacity by providing peaking V2G services. Tomić and Kempton ( 2007 ) examine the economics of using EVs for the provision of frequency reserves and demonstrate that such services can yield substantive revenues to vehicle owners in a variety of wholesale markets. Thompson and Perez ( 2020 ) conduct a meta-analysis of V2G services and value streams and find that power-focused services are of greater value than energy-focused services. They distinguish the two types of services based on the extent to which EV batteries must be discharged and degraded. Sioshansi and Denholm ( 2010 ) come to a similar conclusion in comparing the value of using PHEV batteries for energy arbitrage and operating reserves.

Another important synergy between EVs and power systems is using the flexibility of EV charging to manage the integration of VRE into power systems (Mwasilu et al 2014 , Weiller and Sioshansi 2014 ). Hoarau and Perez ( 2018 ) develop a framework for examining the synergies between EV charging and the integration of photovoltaic-solar resources into power systems. They find that the spatial footprint across which solar resources and EVs are deployed and the regulatory, policy, and market barriers to cooperation between solar resources and EVs are critical to realizing these synergies. Szinai et al ( 2020 ) find that controlled EV charging in California under its 2025 renewable-portfolio standards can reduce operational costs and renewable curtailment compared to unmanaged charging. They find that properly designed time-of-use retail tariffs can achieve some, but not all, of the benefits of controlled EV charging. They show also that these two approaches to managing EV charging (controlling EV charging directly and time-of-use tariffs) reduce the cost of infrastructure that is necessary to accommodate EV charging relative to a case of uncontrolled EV charging. Chandrashekar et al ( 2017 ) conduct an analysis of the Texas power system and find similar benefits of controlled EV charging in reducing wind-integration costs. Coignard et al ( 2018 ) show that under California's 2020 renewable-portfolio standards, controlled EV charging can deliver the same renewable-integration benefits that California's energy-storage mandate does but at substantially lower costs. They show that bidirectional V2G services deliver up to triple the value of controlled EV charging. Kempton and Tomić ( 2005 ) show that high penetrations of wind energy in U.S. could be accommodated at relatively low costs if 3% of the vehicle fleet provides frequency reserves and 8%–38% of the fleet provides operating reserves and energy-storage services to avoid wind curtailment. Loisel et al ( 2014 ) and Zhang et al ( 2019 ) conduct more forward-looking analyses of the synergies between EVs and renewables. The former examines German systems, and the latter examines California systems under potential renewable-deployment scenarios in the year 2030.

An important assumption underlying these works is that EV owners (or aggregations of EVs) are exposed to prices that signal the value of these services and that there are regulatory and business models that allow such services to be exploited (i.e. consumer are willing to engage in these programs and are compensated properly for providing flexible charging). Several pilot studies suggest that EV owners have interest in participating in utility-run controlled-charging programs and that a set of different compensation strategies beyond time-varying electricity pricing might maximize engagement (Geske and Schumann 2018 , Hanvey 2019 , Küfeoğlu et al 2019 , Delmonte et al 2020 ).

Niesten and Alkemade ( 2016 ) survey the literature on these topics and numerous European and U.S. pilot programs in terms of value generation for V2G services. They find that the ability of an aggregator to scale is related to its ability to develop a financially viable business model for V2G services. Another important consideration is the availability of control and communication technologies to manage EV charging based on power-system conditions. Key considerations in the design of control strategies are robustness in the face of uncertainty (e.g. renewable availability, EV-arrival times and charge levels upon arrivals, and EV-departure times), data privacy, and robustness to communication or other failures. Le Floch et al ( 2015a ), Le Floch et al ( 2015b ), ( 2016 ), develop a variety of distributed and partial-differential-equation-based algorithms for controlling EV charging. Rotering and Ilic ( 2011 ) develop a dynamic-optimization-based approach to control EV charging and bidirectional V2G services (with a focus on the provision of ancillary services). Donadee and Ilic ( 2014 ) develop a Markov decision process to optimize the offering behavior of EVs that participate in wholesale electricity markets to provide frequency reserves.

6.3. Remaining challenges for effective vehicle-grid integration at scale

There are still many challenges to tackle before smart charging and V2G can be deployed effectively at large scale. These challenges are linked to the technical aspects of VGI technology but also to societal, economic, security, resilience, and regulatory questions (Noel et al 2019a ). With regard to the technical challenges of VGI, existing barriers notably include battery degradation, charger availability and efficiency, communication standards, cybersecurity, and aggregation issues (Eiza and Ni 2017 , Sovacool et al 2017 , Noel et al 2019a ).

While the technical aspects of VGI are studied widely, this is much less the case for its key societal aspects. Societal issues include the environmental performance of VGI, its impact on natural resources, consumer acceptance and awareness, financing and business models, and social justice and equity (Sovacool et al 2018 ). There are also various regulatory and political challenges linked to clarifying the regulatory frameworks applicable to VGI as well as market-design issues, such as the proper valuation of VGI services and double taxation (Noel et al 2019a ) and the trade-offs between bulk power and distribution-system needs. Regulatory changes may be required to enable distribution-network operators and EV owners (or aggregators) to take a more active role in electricity markets. The Parker project, an experimental project on balancing services from an EV fleet, underlines some of the barriers to providing ancillary services, such as metering requirements (Andersen et al 2019 ). It is argued that insufficient regulatory action might keep us from attaining the full economic and environmental benefits of V2G (Thompson and Perez 2020 ) and that regulations are lagging behind technological developments (Freitas Gomes et al 2020 ). The lack of defined business models is seen by many experts as a key impediment (Noel et al 2019b ).

Major challenges that are linked to data-related aspects of VGI, including who has the right to access data from EVs (e.g. the state of EV batteries and charging) and how these data can be exploited, remain. Privacy concerns are one of the major obstacles to user acceptance (as is fear of loss of control over charging) (Bailey and Axsen 2015 ). In addition, there are also questions linked to cybersecurity (Noel et al 2019a ).

Nevertheless, VGI offers many opportunities that justify the efforts required to overcome these challenges. In addition to its services to the power system, VGI offers interesting perspectives for the full exploitation of synergies between EVs and renewable energy sources as both technologies promise large-scale deployment in the future (Kempton and Tomić 2005 , Lund and Kempton 2008 ). Exploiting EV batteries for VGI also is appealing from a life-cycle perspective, as the manufacturing of EV batteries has a non-trivial environmental footprint (Hall and Lutsey 2018 ). However, there are a few future developments that might compromise the potential of VGI, most notably cheaper batteries (including second-life EV batteries) that might compete with EVs for many potential services (Noel et al 2019b ). In addition, the impacts of new mobility business models, such as the rise of vehicle- and ride-sharing, on grid services remain unclear. Although smart charging will come first in the path toward grid integration, V2G services have the potential to provide additional value (Thingvad et al 2016 ).

7. Life-cycle cost and emissions

EVs differ from conventional ICEVs on an emissions basis. While the operation of gasoline- or diesel-powered ICEVs produces GHG and pollutant emissions that are discharged from the vehicle tailpipe, EVs have no tailpipe emissions. In a broader context, EVs still can be associated with so-called 'upstream' emissions from the processes that generate, transmit, and distribute the electricity that is used for their charging. Fueling an ICEV also involves upstream 'fuel-cycle' emissions from the raw-material extraction and transportation, refining, and final-product-delivery processes that make gasoline or diesel fuel available at a retail pump. These fuel-cycle emissions give rise to the colloquial jargon 'well-to-pump' emissions. Accordingly, a 'well-to-wheels' (WTW) life-cycle analysis (LCA) is an appropriate framework for comparing EV and ICEV emissions. WTW considers both upstream emissions from the fuel cycle ('well-to-pump') and direct emissions from vehicle operation ('pump-to-wheels') for a standardized functional unit and temporal period. WTW studies have a history of over three decades of use to evaluate direct and indirect emissions related to fuel production and vehicle operations (Wang 1996 ). WTW emissions are expressed typically on a per-mile or per-kilometer basis over a vehicle's assumed lifetime.

WTW analyses typically focus only on fuel production and vehicle operation. Some studies consider broader system boundaries that include vehicle production and decommissioning (i.e. recycling and scrappage) in an LCA framework. This broader system boundary considers what is commonly called the 'vehicle cycle' and provides a so-called 'cradle-to-grave' (or 'C2G') analysis. Vehicle-cycle emissions typically account for 5%–20% of today's ICEV C2G emissions and can be as low as 15% or as high as 80% of today's BEV emissions, depending on the underlying electricity-generation mix. Lower-carbon mixes result in vehicle-cycle emissions accounting for a greater portion of total emissions. As an extreme illustrative example, the case of zero-carbon electricity implies that vehicle-cycle emissions account for 100% of C2G emissions. In general, BEV vehicle-cycle emissions are 25% to 100% higher than their ICEV counterpart (Samaras and Meisterling 2008 , Ambrose and Kendall 2016 , Elgowainy et al 2016 , Hall and Lutsey 2018 , Ricardo 2020 ). As this section explores, higher initial BEV vehicle-cycle emissions almost always are counterbalanced by lower emissions during vehicle operation (with notable exceptions in cases in which BEVs are charged from especially high-emissions electricity).

Even including upstream emissions, EVs are championed as a critical technology for decarbonizing transportation (in line with anticipated widespread grid decarbonization). National Research Council ( 2013 ) identifies EVs as one of several technologies that could put U.S. on a path to reducing transportation-sector GHG emissions to 80% below 2005 levels in 2050. Furthermore, National Research Council ( 2013 ) estimates that BEVs would reduce emissions by 53%–72% compared to ICEVs in 2030. IEA ( 2019 ) contends, similarly, that EVs can reduce WTW GHG emissions by half versus equivalent ICEVs in 2030. Recently published literature also agrees, even on a C2G basis, estimating that future EV pathways offer 70%–90% lower GHG emissions compared to today's ICEVs (Elgowainy et al 2018 ). As such, the broad view across national, international, and academic-research perspectives is that EVs offer the potential to reduce transportation-related GHG emissions by 53% to 90% in the future.

Several studies find that EVs already reduce WTW GHG emissions today by as little as 10% or as much as 41% on average versus comparable ICEVs based on current electricity-production mixes. Samaras and Meisterling ( 2008 ), who are among the first to relate a range of potential electricity carbon intensities to associated EV-lifecycle emissions explicitly, estimate a 38%–41% GHG emissions benefit for EVs powered by the average 2008 U.S. grid. Hawkins et al ( 2012a ), informed by a meta-study of 51 previous LCAs, highlight great variations based on different electricity generation assumptions and vehicle lifetime. Hawkins et al ( 2012b ) estimate a decline of 10%–24% global warming potential (a measure proportional to GHG emissions) for EVs powered by the average 2012 European electricity mix. Elgowainy et al 2016 , 2018 ) estimate that EVs emit 20%–35% fewer GHG emissions when operating on the average 2014 U.S. grid mix.

Many factors contribute to variability in EV WTW emissions and estimated reduction opportunities compared to ICEVs—electricity-carbon intensity, charging patterns, vehicle characteristics, and even local climate (Noshadravan et al 2015 , Requia et al 2018 ). To illustrate these variabilities, figure 7 compares WTW GHG emissions of EVs versus comparable ICEVs. Relative emissions reductions are generally larger for larger vehicles. Woo et al ( 2017 ) find that electrifying SUVs reduces emissions more than electrifying sub-compact vehicles on a WTW basis versus comparable ICEVs (30%–45% and 10%–20%, respectively, assuming median national grid mixes). Ellingsen et al ( 2016 ) find that large EVs emit proportionally less than small EVs compared to comparable ICEVs on a C2G basis (27% and 19%, respectively).

Figure 7.  WTW GHG emissions for EVs versus comparable ICEVs on average and with illustrative variability by market segment, electricity generation pathway, grid mix, and ambient temperature.

Low-carbon electricity can lead to greater reductions in EV emissions. Electricity that is produced from coal, which has a high carbon intensity, can increase EV emissions by as much as 40% or decrease EV emissions by as much as 5% compared to an ICEV (depending on other assumptions). Conversely, electricity from hydropower, nuclear, solar, or wind, all of which offer near-zero carbon intensities, can decrease EV emissions by more than 95% compared to an ICEV (Woo et al 2017 ). Such variability in electricity-generation pathways affects the relative benefits of real-world grid mixes. For example, while EVs offer 30%–65% lower emissions versus comparable ICEVs on average in Europe (Woo et al 2017 , Moro and Lonza 2018 ), in individual countries relative emissions can range from as much as 95% lower to 60% higher (Orsi et al 2016 , Moro and Lonza 2018 ). Typically, U.S. EVs provide emissions reductions, but in some regions EV emissions are higher compared to an efficient ICEV (Reichmuth 2020 ). Changes in regional climate and daily weather add further variability: EV emissions can vary between 40% and 50% lower than a comparable ICEV even when charged from the same grid mix (Yuksel et al 2016 ). While outside the scope of a typical WTW comparison, the additional consideration of refueling infrastructure (i.e. gasoline stations for ICEVs and recharging equipment for EVs) is estimated to increase EV emissions by 4%–8% compared to a more modest 0.3%–0.7% increase for ICEV emissions (Lucas et al 2012 ).

When assessing EV emissions, average or marginal grid-emission factors are considered (Anair and Mahmassani 2012 , Traut et al 2013 , EPRI 2015 , Nealer and Hendrickson 2015 , Nealer et al 2015 , Elgowainy et al 2018 ), leading to significantly different results. Average emissions factors consider all electricity loads as equivalent, while marginal emission factors consider EVs as an additional load on top of existing electricity demands and estimate the associated incremental generation emissions. Marginal emissions could be higher or lower than average, depending on the relative emissions of marginal plants compared to the average in different regions. Different questions lead to using average or marginal metrics. Proper assessment of indirect EV emissions associated with electricity generation is complicated by numerous factors, including timescale (short or long term, aggregate or temporally explicit), system boundaries, impact of EV loads on power-system-expansion and -operation decisions, and non-trivial supply-demand synergies and allocation complexities. Yang ( 2013 ) reviews different grid-emissions-allocation methods concluding that there is no ideal approach to the allocation of emissions to specific end-use and stressing how different assumptions make it difficult to determine EV emissions and compare them to other alternatives and across studies. Nealer and Hendrickson ( 2015 ) discuss whether it is more appropriate to use marginal or average grid-emission factors to estimate EV emissions, concluding that 'average emissions may be the most accessible for long-term comparisons given the assumptions that must be made about the future of the electricity grid.'

Just as EVs offer typically a WTW-emissions reduction compared to ICEVs while shifting those emissions from the tailpipe to upstream, EVs shift costs as well. Operational (fuel and maintenance) costs of EVs are typically lower than those of ICEVs, largely because EVs are more efficient than ICEVs and have fewer moving parts. While data are still scarce, a recent Consumer Reports study estimates that maintenance and repair costs for EVs are about half over the life of the vehicle and that a typical EV owner who does most fueling at home can expect to save an average of $800 to $1000 a year on fuel costs over an equivalent ICEV (Harto 2020 ). Insideevs ( 2018 ) estimates a saving of 23% in servicing costs over the first 3 years and 60 000 miles (96 561 km). Borlaug et al ( 2020 ) estimate fuel savings between $3000 and $10 500 compared with gasoline vehicles (over a 15 year time horizon). However, vehicle capital costs for EVs are higher (principally due to the relatively high cost of EV batteries). In general, studies use a TCO metric to combine and compare initial capital costs with operational costs over a vehicle's lifetime. While some studies find that EVs are typically cost-competitive with ICEVs (Weldon et al 2018 ), others find that EVs are still more costly, even on a TCO basis (Breetz and Salon 2018 , Elgowainy et al 2018 ), or that the relative cost depends on other contextual factors, such as vehicle lifetime and use, economic assumptions, and projected fuel prices. Longer travel distance and smaller vehicle sizes favor relatively lower EV TCO (Wu et al 2015 ), as do lower relative electricity-versus-gasoline price differentials (Lévay et al 2017 ). Despite these differences regarding TCO conclusions across studies, there is general agreement that future EV costs will decline (Dumortier et al 2015 , Wu et al 2015 , Elgowainy et al 2018 ).

The existing literature suggests future EV emissions will decline, in large part due to expectations for continued grid decarbonization (Elgowainy et al 2016 , 2018 , Woo et al 2017 , Cox et al 2018 ). For example, Ambrose et al ( 2020 ) anticipate that evolution in vehicle types and designs could accelerate future decreases for EV GHG emissions. Several studies also posit repurposing used EV batteries for stationary applications could accrue additional GHG benefits (Ahmadi et al 2014 , 2017 , Olsson et al 2018 , Kamath et al 2020 ). Cox et al ( 2018 ) suggest future connectivity and automation technologies will enable energy-optimized EV-recharging behavior and associated lower carbon emissions. Similarly, future EV costs also are expected to decline as battery costs continue to decline (cf section 4 ), and new mobility modes such as ride-hailing lead to higher vehicle use that favors the business case for highly efficient EVs compared to ICEVs.

8. Synergies with other technologies, macro trends, and future expectations

Vehicle electrification fits within broader electrification trends, including power-system decarbonization and mobility changes. The latter include micro-mobility in urban areas, new mobility business models revolving around 'shared' services as opposed to vehicle ownership (e.g. ride-hailing and car-sharing), ride pooling, and automation. These trends are driven partially by the larger availability of efficient and cost-effective electrified technologies (Mai et al 2018 ) and the prospect of abundant and affordable renewable electricity and by other technological and behavioral changes (e.g. real-time communication). Abundant and affordable renewable electricity is a conditio sine qua non for EVs to provide a pathway to decarbonize road transportation. Direct use of PV on-board vehicles (i.e. PV-powered or solar vehicles) also is being considered. However, this concept still faces many challenges (Rizzo 2010 , Aghaei et al 2020 ). Yamaguchi et al ( 2020 ) show potential synergies for integration but also highlight that for this technology to be successful, the development of high-efficiency (>30%), low‐cost, and flexible PV modules is essential.

Urban micro-mobility is emerging recently as an alternative to traditional mobility modes providing consumers in most metropolitan areas worldwide with convenient options for last-mile transportation (Clewlow 2019 , Zarif et al 2019 , Tuncer and Brown 2020 ). Virtually all micro-mobility solutions use all-electric powertrains. Shared electric scooters and bikes (often dockless), e.g. those pioneered by Lime and Bird in the U.S., are experiencing rapid success and are 'the fastest-ever U.S. companies to reach billion-USD valuations, with each achieving this milestone within a year of inception' (Ajao 2019 ). Future expectations for micro-mobility remain uncertain due to issues related to sidewalk congestion, safety, and vandalism (heavily impacting the business case for these technologies). However, the nexus with EVs has not been questioned.

Similarly, ride-hailing—matching drivers with passengers at short notice for one-off rides through a smartphone application, which date back to Uber's introducing the concept in 2009—is an attractive alternative to traditional transportation solutions. These mobility-as-a-service solutions cater to the consumer's need for quick, convenient, and cost-effective transportation and may lead to drops in car-ownership and driver-licensure rates (Garikapati et al 2016 , Clewlow and Mishra 2017 , Movmi 2018 , Walmsley 2018 , Henao and Marshall 2019 , Arevalo 2020 ). After just over 10 years, ride-hailing is widely available and extremely successful, with hundreds of millions of consumers worldwide and 36% of U.S. consumers having used ride-hailing services (Mazareanu 2019 ). While most ride-hailing vehicles today are ICEVs (in line with the existing LDV stock), many ride-hailing companies are exploring electrification opportunities (Slowik et al 2019 ). EVs offer a number of potential advantages as high vehicle usage promotes a more favorable business model for recovering the higher EV purchase price by leveraging cheaper fuel costs (Borlaug et al 2020 ). At the same time, long-range vehicles and effective charging solutions are required for ride-hailing companies to transition to EVs (Tu et al 2019 ). Moreover, EVs can mitigate additional fuel use and emissions related to increased travel, mostly due to deadheading, which is estimated to be ∼85% (Henao and Marshall 2019 ). EVs also provide access to restricted areas in some cities (driving some regional goals for ride-hailing electrification). For example, Uber aims for half of its London fleet to be electric by 2021 and 100% electric by 2025 (Slowik et al 2019 ).

Automation trends are also poised to have the potential to disrupt transportation as we know it. The combination of electric and connected automated vehicles (CAVs) is hypothesized to offer natural synergies, including easier integration with CAV sensors and a greater affinity for cheaper fuels aligning with greater travel (Sperling 2018 ). The chief counterargument relates to high power requirements for a heavily instrumented CAV, which would deplete EV batteries quickly and may be accommodated better with PHEV powertrains. Wireless EV charging, both stationary and dynamic, increases the potential synergies enabling autonomous recharging. Also, CAVs may be required to maximize the efficiency of dynamic wireless charging. In fact, without the alignment accuracy enabled by CAVs, in-road dynamic charging may have limited efficacy. The literature on these synergies is relatively sparse, though some studies are beginning to investigate the implications of combining EV and CAV technologies.

Even though the technology is not widely available commercially, several studies are beginning to examine how consumer preferences may be influenced by the combination of connected, automated, and electric vehicles 15 . Thiel et al ( 2020 ), for example, highlight how full EV success may emerge as automated shared vehicles become predominant in a world where the border between public and private transport will cease to exist. Tsouros and Polydoropoulou ( 2020 ) develop a survey combining traditional attributes (e.g. car type and vehicle style) alongside future technology attributes (e.g. fuel type and degree of automation) and estimate preferences using a latent-class structural-regression approach. They find a specific class of consumers, described as technology-savvy, who have a high proclivity for both alternative-fuel vehicle technologies and higher degrees of automation. While the proportion of the population that can be classified as technology-savvy is unclear, Tsouros and Polydoropoulou ( 2020 ) provide early compelling evidence that consumers see explicit value in the combination of EVs and automation. Hardman et al ( 2019 ) provide a complementary perspective of early adopters of automated vehicles based on a survey of existing U.S. EV owners. Similar to the work of Tsouros and Polydoropoulou ( 2020 ), Hardman et al ( 2019 ) find that the type of consumers who would pursue automated vehicles have similar lifestyles, attitudes, and socio-demographic profiles as EV adopters. These include high-income consumers, with high levels of knowledge about technology features, who have positive perceptions of CAV attributes and technology in general, provided that safety concerns are resolved.

Another benefit of the combined technologies is the potential to integrate charging events better with the needs of the electricity grid. Several studies assess the combination of these technologies with new mobility services such as car-sharing systems to optimize VGI. Iacobucci et al ( 2018 ) consider a case study in Tokyo of the ability of connected, automated EVs to be dispatched to respond to both transportation demand and charging to meet demands and constraints of the electricity system. The authors observe the vehicles can take advantage of a variety of different time-of-day pricing structures—leading to a tradeoff between wait times and cost benefits from lower fuel prices. They find that the vehicles in Tokyo can supply on the order of 3.5 MW of charging flexibility per 1000 vehicles, even during times of high mobility demand. Miao et al ( 2019 ) conduct a similar study in a generic region. The authors develop an algorithm that simulates operational behavior of the connected, automated EV technology that includes trip demand and vehicle usage, vehicle relocation, and vehicle charging. Their results indicate that charging behavior is highly sensitive to different levels of charging due to the length of charging—which can affect service provision of trip demand.

The final topic of study considering synergistic opportunities between connected, automated vehicles and EVs focused on emissions benefits. Taiebat et al 2018 explore the environmental impacts of automated vehicles showing net positive environmental impacts at the local vehicle-urban levels due to improved efficiency, but acknowledge that greater vehicle utilization and shifts in travel patterns might to offset some of these benefits. Of course, EVs provide the significant benefit of eliminating tailpipe criteria-pollutant emissions, yielding significant human-health benefits. Regarding GHG emissions, two of the earliest studies on this topic examine the net effect of automation on reducing transport GHG emissions (Brown et al 2014 , Wadud et al 2016 ). Greenblatt and Saxena ( 2015 ) conduct a case-study application of connected and automated vehicles in taxi fleets and find large emissions benefits associated with electrification. They find a decrease of GHG emissions intensities ranging from 87% to 94% below comparable ICEVs in 2014 and 63% to 82% below hybrid electric vehicles (HEVs) in 2030. The total emissions benefit is augmented relative to privately owned vehicles due to the higher travel intensity of taxi vehicles. Following these earlier works, additional case studies examine the hypothetical application of automated and electric fleets. These include two studies in Austin, Texas. Loeb and Kockelman ( 2019 ) examine a variety of scenarios to simulate the operation of different vehicle fleets replacing current-day transportation network companies and taxis. The primary goal of their work is to estimate costs associated with operation. They find that automated EVs are the most profitable and provide the best service among the vehicle-technology options that they examine. Gawron et al ( 2019 ) also perform a case study in Austin, Texas, but focus on the emissions benefits of electrifying an automated taxi fleet. They find that nearly 60% of emissions and energy in a base case CAV fleet can be reduced by electrifying powertrains. These improvements can be pushed up to 87% when coupled with grid decarbonization, dynamic ride-sharing, and various system- and technology-efficiency improvements. These results are consistent with a more generalized study by Stogios et al ( 2019 ), who, in a similar approach simulating fleet behavior, find that emissions from CAVs are most dramatically improved via electrification.

While EVs are a relatively new technology and automated vehicles are not widely available commercially, the implications and potential synergies of electrification and automation operating in conjunction are significant. The studies mentioned in this section are investigating a broad set of impacts when CAVs are coupled with EVs. Future research is necessary to generalize and refine many of these results. However, the potential for transformative changes to transportation emissions is clear.

8.1. Expectations for the future

EVs hold great promise to replace ICEVs for a number of on-road applications. EVs can provide a number of benefits, including addressing reliance on petroleum, improving local air quality, reducing GHG emissions, and improving driving experience. Vehicle electrification aligns with broader electrification and decarbonization trends and integrates synergistically with mobility changes, including urban micro-mobility, automation, and mobility-as-a-service solutions. The effective integration of EVs into power systems presents numerous opportunities for synergistic improvement of the efficiency and economics of electromobility and electric power systems, with EVs capable of supporting power-system planning and operations in several ways. Full exploitation of the synergies between EVs and VRE sources offers a path toward affordable and clean energy and mobility for all, as both technologies promise large-scale deployment in the future. To enable such a future continued technology progress, investments in charging infrastructure (and related building codes), consumer education, effective and secure VGI programs, and regulatory and business models supporting all aspects of vehicle electrification are all critical elements.

The coronavirus pandemic is impacting LDV sales in most countries negatively, and 2020 EV sales are expected to be lower than 2019, marking the first decline in a decade (BloombergNEF 2020 ). However, sales of ICEVs are set to drop even faster and, despite the crisis, EV sales could reach a record share of the overall LDV market in 2020 (Gul et al 2020 ). Despite these short-term setbacks, long-term prospects for EVs remain undiminished (BloombergNEF 2020 ).

Several studies project major roles for EVs in the future, which is reflected in massive investment in vehicle development and commercialization, charging infrastructure, and further technology improvement, especially in batteries and their supply chains. Consumer adoption and acceptance and technology progress form a virtuous self-reinforcing circle of technology-component improvements and cost reductions that can enable widespread adoption. Forecasting the future, including technology adoption, remains a daunting task. Nevertheless, this detailed review paints a positive picture for the future of EVs for on-road transport. The authors remain hopeful that technology, regulatory, societal, behavioral, and business-model barriers can be addressed over time to support a transition toward cleaner, more efficient, and affordable mobility solutions for all.

Acknowledgments

The authors thank Paul Denholm, Elaine Hale, Trieu Mai, Caitlin, Murphy, Bryan Palmintier, and Dan Steinberg for valuable comments on figure 6 , as well as two anonymous reviewers for helpful comments on the paper. This work was co-authored by National Renewable Energy Laboratory (NREL), which is operated by Alliance for Sustainable Energy, LLC, for U.S. Department of Energy (DOE) under Contract No. DE-AC36-08GO28308. No funding was received to support this work. The views expressed in this article do not necessarily represent the views of DOE or the U.S. Government. The findings and conclusions in this publication are those of the authors alone and should not be construed to represent any official U.S. Government determination or policy, or the views of any of the institutions associated with this study's authors.

 EVs are defined as vehicles that are powered with an on-board battery that can be charged from an external source of electricity. This definition includes plug-in hybrid electric vehicles (PHEVs) and battery electric vehicles (BEVs). EVs often are referred to as plug-in electric vehicles (PEVs).

 Transport electrification is confined not only to electric LDVs. Transport electrification includes a wide range of other vehicles, spanning from small vehicles that are used for urban mobility, such as three-wheelers, mopeds, kick-scooters, and e-bikes, to large urban buses and delivery vehicles. In 2019, the number of electric two-wheelers on the road exceeded 300 million and buses approached 0.6 million (IEA 2019 , Business Wire 2020 ), with new deliveries in 2019 close to 100 thousand units (EV Volumes 2020 ).

 Just over 10% of the U.S. heavy-duty truck (Class 7–8) population requires an operating range of 500 miles (805 km) or more, while nearly 80% operate within a 200 mile (322 km) range and around 70% within 100 miles (161 km). Only ∼25% of heavy truck VMT require an operating range of over 500 miles (805 km) (Borlaug et al Forthcoming ).

 As a counterargument, Tesla states that 'all new Tesla cars come standard with advanced hardware capable of providing Autopilot features today, and full self-driving capabilities in the future—through software updates designed to improve functionality over time'.

Towards Sustainable Energy: A Systematic Review of Renewable Energy Sources, Technologies, and Public Opinions

Ieee account.

• Change Username/Password
• Update Address

Purchase Details

• Payment Options
• Order History
• View Purchased Documents

Profile Information

• Communications Preferences
• Profession and Education
• Technical Interests
• US & Canada: +1 800 678 4333
• Worldwide: +1 732 981 0060
• Contact & Support
• About IEEE Xplore
• Accessibility
• Terms of Use
• Nondiscrimination Policy
• Privacy & Opting Out of Cookies

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity. © Copyright 2023 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.

• Doing Business and related research

Research on Getting Electricity

Doing Business considers the following list of papers as relevant for research on getting electricity. Some papers—denoted with an asterisk (*)—use Doing Business data for their empirical analysis. If we've missed any important research, please let us know .

Addressing the electricity access gap

Connecting businesses to the electrical grid in 183 economies*, connection charges and electricity access in sub-saharan africa, electricity cost and firm performance: evidence from india, managing an electricity shortfall: a guide for policy makers, power tariffs: caught between cost recovery and affordibility, moving beyond lcoe: impact of various financing methods on pv profitability for sids, implementing structural reforms in abenomics: how to reduce the cost of doing business in japan, "getting to gender equality in energy infrastructure. lessons from electricity generation, transmission, and distribution projects ", electricity sector constraints for firms across economies : a comparative analysis, "when do firms generate evidence on in-house electricity supply in africa", "how do electricity shortages affect productivity evidence from india.", power outages and firm performance in sub-saharan africa, power outages and economic growth in africa, power infrastructure quality and manufacturing productivity in africa: a firm level analysis, papers on doing business topics.

• Starting a Business
• Dealing with Construction Permits
• Getting Electricity
• Registering Property
• Getting Credit
• Protecting Minority Investors
• Paying Taxes
• Trading across Borders
• Enforcing Contracts
• Resolving Insolvency
• Download full list of research papers related to Doing Business topics (Excel)

Feedback has been successfully submitted.

Research Paper: A New Electricity Era

In a decarbonized power system, extended electrification in the transport and buildings sectors has the potential to reduce fossil fuel use by at least 10-20% by 2040, leading to a reduction in CO2 emissions of 2-4GT per annum. Electrification of some industrial processes could increase this potential even further.

This is the main conclusion of a joint study carried out by the Climate Policy Initiative and Copenhagen Economics for the Energy Transition Commission.

Currently, 78% of energy used by the end-use sectors (transport, buildings and industry) comes from fossil fuels, of which 17% comes from the use of electricity generated from fossil fuels.

Electrification has two direct benefits:

1- Energy productivity benefits: Electricity delivers an equivalent energy service with less energy input, because it avoids conversion losses associated with burning fossil fuels.

2- Decarbonization benefits: Greater electrification of several energy end uses, when coupled with decarbonization of the power supply, could accelerate decarbonization by extending its benefits to a broader range of economic activities.

This study relies on the assumption that the decarbonization of the power system is possible at low-cost, building on another piece of research carried out by Climate Policy Initiative for the Energy Transitions Commission, which is available here.

In the transport sector, electrification could reduce fossil fuel use by at least 10 to 30% if the right investments in electric transport infrastructure are deployed.

This assumes a 50% electrification of light road vehicles by 2040 and very limited electrification of heavy duty vehicles, although innovation creating new opportunities in freight transport cannot be ruled out.

In the buildings sector, expanding the electrification of water heating, space heating and cooking could replace 35% of fossil fuel use by 2040.

Although buildings are already electrified to a relatively large degree, especially in terms of appliances and lighting, further progress is possible. Technologies are widely available, but key barriers lie in the high capital cost of some of these technologies, the lack of incentives to encourage tenants to make the necessary investments and the low turnover rate of large equipment.

In the industry sector, this study has conservatively assumed no further electrification by 2040.

Currently, the share of electricity in energy consumption varies considerably across sectors, from 54% in aluminum production to 10% in chemicals and petrochemicals. Significant R&D efforts are still required to develop and deploy electrified processes, such as electro-thermal technologies, in a range of industrial sectors. Alternative decarbonization routes are also likely to play a role, including bioenergy, hydrogen and carbon capture.

About the Energy Transitions Commission:

The Energy Transition Commission (ETC) is a uniquely diverse group of individuals and organizations from across the energy landscape (investors, incumbent energy companiesindustry disruptors, equipment suppliers, non-profit organizations, and research institutes), and from the developed and developing worlds. We aim to accelerate change towards low-carbon energy systems that enable robust economic development and limit the rise in global temperature to well below 2°C. ,

Download paper

Related posts

New report: the etc calls for a rapid phase-down of fossil fuel demand and supply, new report: scale-up of critical materials and resources required for energy transition, new insights briefing: clean energy technology supply chains must get better, faster, cleaner, get in touch.

[email protected]

Join our newsletter

Sign up to our mailing list to receive regular updates

© 2020 Energy Transitions Commission. Privacy policy | Cookies | Website by Regency

• Our ambition
• Energy programmes
• Keeping 1.5°C Alive
• Regional programmes
• Sector decarbonisation
• Publications
• News & events

Privacy Overview

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 14 November 2023

Evidence of an upper ionospheric electric field perturbation correlated with a gamma ray burst

• Mirko Piersanti   ORCID: orcid.org/0000-0001-5207-2944 1 , 2   na1 ,
• Pietro Ubertini   ORCID: orcid.org/0000-0003-0601-0261 2   na1 ,
• Roberto Battiston   ORCID: orcid.org/0000-0002-5808-7239 3 , 4   na1 ,
• Angela Bazzano 2   na1 ,
• Giulia D’Angelo   ORCID: orcid.org/0000-0002-9214-2051 2   na1 ,
• James G. Rodi 2   na1 ,
• Piero Diego 2   na1 ,
• Zhima Zeren 5 ,
• Roberto Ammendola   ORCID: orcid.org/0000-0003-4501-3289 6 ,
• Davide Badoni   ORCID: orcid.org/0000-0002-9027-2039 6 ,
• Simona Bartocci 6 ,
• Stefania Beolè 7 ,
• Igor Bertello 2 ,
• William J. Burger 4 ,
• Donatella Campana 8 ,
• Antonio Cicone 2 , 9 ,
• Piero Cipollone 6 ,
• Silvia Coli 7 ,
• Livio Conti   ORCID: orcid.org/0000-0003-2966-2000 6 , 10 ,
• Andrea Contin 11 , 12 ,
• Marco Cristoforetti   ORCID: orcid.org/0000-0002-0127-1342 4 , 13 ,
• Fabrizio De Angelis 2 ,
• Cinzia De Donato 6 ,
• Cristian De Santis   ORCID: orcid.org/0000-0002-7280-2446 6 ,
• Andrea Di Luca   ORCID: orcid.org/0000-0002-9074-2133 3 , 4 ,
• Emiliano Fiorenza 2 ,
• Francesco Maria Follega   ORCID: orcid.org/0000-0003-2317-9560 3 , 4 ,
• Giuseppe Gebbia   ORCID: orcid.org/0000-0001-7252-7416 3 , 4 ,
• Roberto Iuppa 3 , 4 ,
• Alessandro Lega 3 , 4 ,
• Marco Lolli 12 ,
• Bruno Martino 2 , 14 ,
• Matteo Martucci 6 ,
• Giuseppe Masciantonio   ORCID: orcid.org/0000-0002-8911-1561 6 ,
• Matteo Mergè 6 , 15 ,
• Marco Mese 8 , 16 ,
• Alfredo Morbidini 2 ,
• Coralie Neubüser   ORCID: orcid.org/0000-0002-2008-8404 4 ,
• Francesco Nozzoli   ORCID: orcid.org/0000-0002-4355-7947 4 ,
• Fabrizio Nuccilli 2 ,
• Alberto Oliva   ORCID: orcid.org/0000-0002-6612-6170 11 , 12 ,
• Giuseppe Osteria   ORCID: orcid.org/0000-0002-9871-8103 8 ,
• Francesco Palma 6 ,
• Federico Palmonari 11 , 12 ,
• Beatrice Panico 8 , 16 ,
• Emanuele Papini 2 ,
• Alexandra Parmentier 2 , 6 ,
• Stefania Perciballi   ORCID: orcid.org/0000-0003-2868-2819 7 ,
• Francesco Perfetto 8 ,
• Alessio Perinelli   ORCID: orcid.org/0000-0001-5603-3950 3 , 4 ,
• Piergiorgio Picozza 6 , 17 ,
• Michele Pozzato 12 ,
• Gianmaria Rebustini 6 ,
• Dario Recchiuti   ORCID: orcid.org/0000-0002-9530-6779 2 , 3 ,
• Ester Ricci   ORCID: orcid.org/0000-0002-4222-9976 3 , 4 ,
• Marco Ricci 18 ,
• Sergio B. Ricciarini 19 ,
• Andrea Russi   ORCID: orcid.org/0000-0001-7884-2310 2 ,
• Zuleika Sahnoun 12 ,
• Umberto Savino 7 ,
• Valentina Scotti   ORCID: orcid.org/0000-0001-8868-3990 8 , 16 ,
• Xuhui Shen 20 ,
• Alessandro Sotgiu   ORCID: orcid.org/0000-0001-8835-2796 6 ,
• Roberta Sparvoli 6 , 17 ,
• Silvia Tofani 2 ,
• Nello Vertolli 2 ,
• Veronica Vilona   ORCID: orcid.org/0000-0001-9893-9419 3 ,
• Vincenzo Vitale   ORCID: orcid.org/0000-0001-8040-7852 6 ,
• Ugo Zannoni 2 ,
• Simona Zoffoli   ORCID: orcid.org/0000-0003-3573-9051 15 &
• Paolo Zuccon   ORCID: orcid.org/0000-0001-6132-754X 3 , 4  

Nature Communications volume  14 , Article number:  7013 ( 2023 ) Cite this article

7073 Accesses

1128 Altmetric

Metrics details

• Atmospheric dynamics
• Natural hazards

Earth’s atmosphere, whose ionization stability plays a fundamental role for the evolution and endurance of life, is exposed to the effect of cosmic explosions producing high energy Gamma-ray-bursts. Being able to abruptly increase the atmospheric ionization, they might deplete stratospheric ozone on a global scale. During the last decades, an average of more than one Gamma-ray-burst per day were recorded. Nevertheless, measurable effects on the ionosphere were rarely observed, in any case on its bottom-side (from about 60 km up to about 350 km of altitude). Here, we report evidence of an intense top-side (about 500 km) ionospheric perturbation induced by significant sudden ionospheric disturbance, and a large variation of the ionospheric electric field at 500 km, which are both correlated with the October 9, 2022 Gamma-ray-burst (GRB221009A).

Introduction

Evidence of ionospheric disturbance induced by a gamma-ray burst (GRB) was first reported in 1988 by Fishman and Inan 1 as due to the GRB occurred on 1st August 1983, the strongest ever observed at that time, with a total fluence exceeding 10 −3 e r g s / c m 2 / s . The measured bulk effect on the ionosphere was the amplitude change of Very Low Frequency (VLF) radio signals, proof of the perturbation induced in the lower part of the ionosphere by that very energetic extrasolar event.

During cosmic GRB (and solar flare too), the intense high energy photon flux can abnormally ionize the lower ionosphere 2 by producing a large increase of free electron density 3 . As a consequence, the electron density grows giving rise to a variation of the ionospheric conductivity leading to a pronounced alteration in both VLF and ELF (Very Low and Extremely Low Frequency) electric field behaviour, respectively. Using ground VLF emitters, Inan et al. 4 showed that, if the burst is sufficiently severe (total fluence exceeding 10 −3 e r g s / c m 2 ) and long-lasting, the ionospheric perturbation caused by a GRB can be observed in the bottom-side ionosphere (from about 60 km up to about 350 km of altitude). Although dedicated satellites recorded an average of more than one GRB per day in the last decade, intensive ionospheric reactions were seldom observed. In fact, only a handful of papers have reported the detection of ionospheric perturbations due to GRBs events 3 , 4 , 5 , 6 , 7 , 8 , though always in the bottom-side ionosphere.

In addition, both Sentman et al. 9 , and Price and Mushtak 10 have investigated GRB effects on Earth’s ionosphere finding no significant variation on the ELF electromagnetic wave data. Nonetheless, Tanaka et al. 11 reported a clear detection of transient ELF signal caused by the December 27, 2004, event, a very intense cosmic gamma-ray flare, inducing a clear variation in the ionospheric Schumann resonance 12 detected by electromagnetic ground stations.

In this work we present the evidence of variation of the ionospheric electric field at about 500 km induced by the strong GRB occurred on October 9th, 2022. Using both satellite observations and a new ad hoc developed analytical model, we prove that the GRB221009A deeply impacted on the Earth’s ionospheric conductivity, causing a strong perturbation not only in the bottom-side ionosphere 13 , 14 , but also in the top-side ionosphere (at around 500 km).

On October 9th, 2022, at 13:21 UT, a highly bright and long-lasting GRB (hereafter GRB221009A), triggered many of the X and Gamma-ray space observatories, in particular Swift 15 , 16 , Fermi 17 , 18 , MAXI 19 , AGILE 20 , 21 and INTEGRAL 22 , 23 . The GRB follow-up was observed by most operative telescopes in space and on-ground. The INTEGRAL (see Integral satellite data section for more details) gamma-ray observatory 24 detected the GRB both using the SPI spectrometer (SPectrometer of Integral) and the IBIS imager (Imager on-Board the INTEGRAL Satellite) as a complex, impulsive, very strong photon signal followed by a very intense gamma-ray afterglow 25 . The GRB221009A zenith was located over India and the GRB photon flux was illuminating Europe, Africa, Asia and part of Australia (Fig.  1 ).

A map of the Earth with the CSES satellite orbit trace shown in blue. The green-colored part along the orbit marks the time of the electric field variation triggered by the GRB and detected by EFD. The gray shaded area shows the estimated illumination area of GRB221009A impinged at a latitude of 19.8 ∘ and a longitude of 71 ∘ (red circle).

The light curves from the SPI detector and the IBIS imager (Fig.  2 ) show a multi-peaked structure with a moderately intense precursor, starting at 13:16:58 UTC, followed by a very strong prompt GRB emission, peaking at 13:21 UTC and, a long-lasting, sustained, soft gamma afterglow detected by both instruments in the energy range 75–1000 keV (SPI) and 0.250–2.6 MeV (IBIS), respectively. Optical follow-up with the OSIRIS (Optical System for Imaging and low-Intermediate-Resolution Integrated Spectroscopy) at the 10.4m GTC (Gran Telescopio CANARIAS) telescope confirmed the presence of a strong optical afterglow in the range 3700-10000Å with features suggesting a supernova progenitor 26 .

Time profile of G R B 221009 A detected by INTEGRAL, scaled for background. The red curve shows the SPI/ACS count rate on 1s time-bin plotted in e r g / c m 2 / s in the energy range 75-1000 keV; the black curve shows the IBIS/PICsIT data in the energy range 0.25–2.60 MeV. The differences between the two light curves are due to: i) difference in computing the two energy bands, ii) statistical fluctuations (IBIS/PICSiT is less sensitive in this case because of the partial shield absorption to low energy photons), iii) instrument saturation and/or telemetry loss due to the exceptionally strong photon flux from GRB221009A.

The fluence of the prompt emission (i.e. the total time-integrated energy per unit area), lasting about 800s, was 0.013 e r g / c m 2 in the 75–1000 keV energy range 23 . This value is a lower limit estimate, due to the partial saturation and pile-up caused by the intense GRB photon flux. As far as we know, this GRB is among the largest ever detected. Assuming both a measured distance corresponding to a red-shift z=0.151 27 and an isotropic emission (E-Iso) only in the high energy band, the energy emitted during the prompt GRB was about 8  ⋅  10 53 ergs. It should be noted that both fluence and E-Iso values are lower limits, due to the partial saturation of instruments (SPI) and telemetry data transmission (IBIS). The prompt emission was followed by an unusual strong soft gamma-ray long tail 25 decaying with a power index around -2, and lasting at least 40 minutes before crossing the detection threshold of both SPI and IBIS detectors (Fig.  2 ).

GRB221009A strongly perturbed the D-region 13 (about 60−100 km of altitude) and, for the first time, its effect was observed also in the top-side ionosphere (507 km) by the Electric Field Detector (EFD) 28 onboard the Low Earth Orbit (LEO) Chinese Seismo Electromagnetic Satellite (CSES - see CSES satellite electric field data section for more details) 29 , which was orbiting from North to South over the European sector (blue line in Fig.  1 ). Figure  3 shows the comparison between the SPI/ACS gamma-ray flux (panel a) and the ionospheric electric field measured by EFD (panel b). At 13:17:01 UT, EFD was switched on just before entering the Auroral Oval (AO). In this region (blue-shaded region), the electric field variations are strongly dominated by the auroral electrojet (a large horizontal current flowing mainly in the E region of the ionosphere, located at an altitude of about 100 − 150 km), generated by complex solar wind-magnetosphere interaction processes 30 , 31 , 32 . This effect results in the impossibility to correlate the evolution of GRB221009A peaks from 13:17:07 to 13:20:44 UT. The position of the AO boundaries were determined by using Ding et al. algorithm 33 . At 13:25:03, about 1.5 min (Δ t G I ) after the beginning of the third and final peak of GRB221009A, the EFD observed a strong peak in the ionospheric electric value of about 54 m V / m . We hypothesize that such an electric field variation in the top-side ionosphere can be driven by the GRB221009A occurrence.

Comparison between the SPI gamma-ray flux (panel a ) and the ionospheric electric field observed by the CSES satellite (panel b ), after the subtraction of the v s  ×  B induced electric field ( v s and B are the spacecraft speed and the local magnetic field, respectively). The blue-shaded region corresponds to the CSES flight over the Auroral Region. The CSES electric field observations were measured at an altitude of 507 km.

In fact, Δ t G I might be related to a characteristic feature of the ionosphere in response to ionizing flux 34 , 35 , which in general, depends on the balance between the electron production rate (dominated by photo-ionization) and the electron losses (resulting from recombination) 34 , 36 , 37 . The physical effect caused by the electron loss process is to delay the response of the changes in electron density ρ e to changes induced by the photo-ionization process. As a consequence Δ t G I should represents the time taken for the ionospheric photo-ionization recombination processes to recover the equilibrium after an increase of irradiance. The higher is the ionospheric density, the larger is the delay time 35 , 37 , 38 .

Figure  4 shows the CSES electric field observations during the GRB221009A occurrence for the three geographical components E x (panel a), E y (panel b) and E z (panel c), where x is directed northward, y westward, and z along the (negative) radial direction. It can be easily seen that the EFD variation (black curve) is superimposed to a low-frequency modulation induced by v s  ×  B effect 28 .

CSES electric field waveform as function of time during the GRB221009A occurrence for the three components E x (panel a ), E y (panel b ) and E z (panel c ) with (black curve) and without (blue curve) the v s  ×  B induced electric field component. The blue-shaded region corresponds to the CSES flight over the Auroral Region. The CSES electric field observations were measured at an altitude of 507 km.

At 13:25:03 UT a large peak in the ionospheric electric field is visible along the three components, whose amplitudes are: Δ E x  = 32.6 m V / m ; Δ E y  = − 39.5 m V / m ; Δ E z  = 27.9 m V / m .

These observations are consistent with an anomalous high ionization in the ionosphere. In general, such a ionospheric perturbations are caused by solar flares and/or solar particle events leading to sudden radio wave absorption (in both the medium frequency - MF - and high frequency - HF - ranges) 39 . These effects are detected in the D-region and are called Sudden Ionospheric Disturbances (SID) 40 . In the present case, the very strong and long-lasting photon flux due to GRB221009A triggered an unprecedented level of ionization in the ionosphere producing both a significant SID in the bottom-side ionosphere and a strong electric field variation in the top-side ionosphere.

We hypothesize that the strong variations of the ionospheric electric field measured by CSES at an altitude of 507 km can only originate from a strong variation in the ionospheric parallel conductivity ( σ 0 ) 41 , 42 , which is directly dependent on the plasma density (see equation ( 5 ) in Analytical model for top-side Ionospheric Electric field variation induced by a GRB section). To confirm such a scenario, on the one hand we investigated the distribution of the ionospheric Total Electron Content (TEC) over Europe, as measured by Global Navigation Satellite System (GNSS) receivers (see GNSS Total Electron Content Data section). As from Fig.  5 , GNSS receivers located in the Mediterranean area recorded a significant TEC increase on October 9th (panel b) between 13:00 and 14:00 UT compared to the day before (panel a) and after (panel c) at the same time, thus confirming the ionizing effect of the intense GRB 1 , 13 , 43 .

Map of the vertical total electron content (TEC) around the CSES satellite position one day before (panel a ), at the moment of (panel b ) and one day after (panel c ) the GRB occurrence. All the maps have been averaged over 1 hour, between 13:00 UT and 14:00 UT. Colors are representative of the TEC value.

On the other hand, we developed an analytical model (see Analytical model for top-side Ionospheric Electric field variation induced by a GRB for more details) able to give a first rough quantitative evaluation of the top-side ionospheric electric field variation driven by an impulsive photon flux (e.g., the impinging of a GRB). As can be seen from Fig.  6 , an impulsive photon source can generate a variation in the top-side ionospheric electric field of about 30 m V / m only if the ratio R α β between ion production ( α ) and absorption ( β ) rates are greater than 5. In addition, if R α β is lower than 2, the effect of the ionization seems not to be able to produce significant variation in the electric field. Such a result is in agreement with the previous experimental observations related to GRBs impinging the ionosphere 1 , 4 , 43 , 44 .

Model results of top-side ionospheric electric field time variation induced by a impulsive photon source. Colours are representative of different photon production/absorption rate ratio.

In addition, our model predicts a time delay (Δ t t h ) between the peak of the GRB and the peak of the ionospheric electric field variation of 1.22 minutes for R α β  = 5. This Δ t t h is in agreement with the Δ t G I observed.

As previously said, being the observations in the bottom-side ionosphere analogous to the effects induced by solar flares (Solar Flare Effect - SFE) 45 , 46 , we investigated the possibility of a sudden intensification of the Solar quiet (Sq) ionospheric current system 47 , 48 and of the ionospheric Equatorial Electrojet (EEJ) 49 , 50 induced by the GRB221009A 51 . The Sq ionospheric electric currents are located in the E-region and are responsible of the diurnal variation in the geomagnetic field observed at ground 52 . Figure  7 b shows the comparison between the equatorial electrojet, estimated in terms of the variation of the North-South component of the geomagnetic field (H - see Equatorial electrojet evaluation section for more details), calculated for a solar quiet day (October 12 th , 2022, black line) and for the day of the GRB occurrence (October 9 th , 2022, red line). It can be seen that the occurrence of the GRB221009A (vertical black dashed line) generated a perturbation of the EEJ. Indeed, superimposed to the long-term variation, featured in both days and characterized by a minimum around both dawn and dusk, and by a maximum around the noon, at about 13:21 UT a low frequency (0.35 mHz) fluctuation appears. Such a variation is more clear in the original magnetometer data used for the EEJ evaluation (panels a, b, c and d) in Fig.  7 ). In fact, looking at Tatuoka data (panels b and d) which is located inside the EEJ, we can see that during quiet conditions (panel b) the geomagnetic field reaches its maximum values around the local noon remaining almost stable for about 2.5 hours before decreasing down as the station approaches the local dusk 49 . Differently, on October 9th (panel d), before the GRB occurrence, as expected the H field reaches its maximum value, but, around 13:21 UT, in coincidence with the occurrence of the first peak of the GRB (vertical black dashed line), instead of remaining stable, starts to fluctuate with a low frequency of 0.35 mHz. Interestingly, such alteration lasted up to 19:00 UT, possibly sustained by the hard GRB221009A long tail (Fig.  2 ), containing more than 10% of the total energy of the prompt emission.

Estimation of the Equatorial Electrojet for the a solar quiet day of October 2022 (black line) and for the day of the GRB occurrence (red line): panel a ) and c ) show original observations of the H component of the geomagnetic field from San Juan magnetometer station during quiet and GRB day, respectively; panel b ) and d ) show original observations of the H component of the geomagnetic field from Tatuoka magnetometer station during quiet and GRB day, respectively; panel e ) shows the EEJ results in terms of Δ H . Black dashed lines represent the time occurrence of the first peak of the GRB.

In conclusion, the unprecedented photon-flux associated to the GRB221009A deeply impacted on the Earth’s ionospheric conductivity, causing a strong perturbation not only in the bottom side ionosphere 13 , 14 , where it is typically observed using ground VLF antennas 53 , but also in the top-side ionosphere (at around 500 km). In fact, a huge variation of the ionospheric electric field, induced by the strong ionospheric conductivity change was detected in the top side ionosphere (507 km) as a consequence of a GRB impact, which increased the ionospheric plasma density by the huge photo-ionization (even in the dayside), as depicted in Fig.  5 . The analytical model described in this work supports the observations and confirms the hypothesis that the interaction between GRB and top-side ionosphere is a threshold process 1 , 4 , 44 . Our model suggests that such a threshold strictly depends on both the production-to-loss-rate ratio of ions and the time duration of the ionization process.

As a closing remark, we want to highlight that, differently to previous similar studies 13 focused on the impact of GRB on both D- and F- regions by using TEC data 6 , 43 and/or VLF ground electromagnetic transmitters 1 , 4 , 14 , our work represents, at our knowledge, the first-ever top-side ionospheric (507km) measurement of electric field variation triggered by impulsive cosmic photons.

This section contains the description of the datasets used in this study and the analytical description of the model developed for the explanation of the experimental results.

INTEGRAL satellite data

INTEGRAL, an ESA lead space observatory for observations in the energy range from a few keV up to 10 MeV, was launched in 2002 and is still fully operative. In this study data from the Imager IBIS 54 and the SPectrometer SPI 55 have been used. In particular, IBIS observes from 25 keV to 10 MeV, with an angular resolution of 12 arcmin, enabling a bright source to be located to better than 1 arcmin. SPI observes radiation between 20 keV and 8 MeV with an high energy resolution of 2 keV at 1 MeV, capable to resolve candidate gamma-ray lines 56 . The INTEGRAL instruments were pointed to a sky direction at about 60 degree offset respect to the GRB arrival direction and the signal were detected by the omni-directional SPI/ACS shield and by the IBIS/PICsIT detector through the telescope shield (see annex material for detailed telescope response 57 ). The INTEGRAL data are transmitted continuously in real-time to ground, and distributed in almost real-time via GCN web network and also through the Interplanetary Network (IPN).

CSES satellite electric field data

CSES-01 (Chinese Seismo-Electromagnetic Satellite) is a LEO satellite orbiting sun-synchronously at about 507 km  since February 2018 29 , 58 . CSES-01 has nine instruments on board for the electromagnetic field, waves and charged particle observations in the upper ionosphere. For this analysis we used electric field data from the Electric Field Detector (EFD) 59 . EFD is able to measure the electric field in four frequency bands: ULF (DC -16 Hz) with a sampling frequency of 125 Hz; ELF (6 Hz–2.2 kHz) with a sampling frequency of 5 kHz; VLF (1.8 kHz–20 kHz) with a sampling frequency of 50 kHz; and HF (High-frequency, 18 kHz–3.5 MHz) with a sampling frequency of 50 kHz. Due to the limitation of telemetry capability, the waveform data are only available for both the ULF and ELF bands, and for a few minutes, over the global seismic belts, for both VLF and HF bands. During the remaining part of the orbit the VLF and HF data are transmitted as Fast Fourier Transform (FFT) 28 .

To eliminate the v s  ×  B effect ( v s and B are the spacecraft speed and the local magnetic field, respectively), induced by the motion of the satellite inside the geomagnetic field, from the E field components, we applied the technique described in Diego et al. 28 .

GNSS total electron content data

To investigate the ionospheric scenario leading to the observed impulsive variation of the current generated in the ionosphere, we collected and processed standard daily RINEX files provided by the permanent stations, located in Europe, of the University NAVSTAR Consortium and of the Rete Integrata Nazionale GNSS 60 managed by the Istituto Nazionale di Geofisica e Vulcanologia (INGV). In particular, to calibrate vertical total electron content (vTEC) data, we processed GNSS measurements as described in D’Angelo et al. 61 by using the technique by Ciraolo et al. 62 and Cesaroni et al. 63 . Specifically, to generate maps over a specific world zone, we performed an average of one hour vTEC observations over 1 ∘  × 1 ∘ of geographic latitude and longitude bin using data recorded by all satellites in view of each selected GNSS ground receiver.

Equatorial electrojet evaluation

The equatorial electrojet (EEJ) was obtained using the method described in Soares et al. 64 . We considered the H (North-South) component the geomagnetic field at ground alone, being directly related to the east-west flow of the EEJ 49 . We used two pairs of ground stations consisting of one magnetometer close to the magnetic equator and one out at almost the same meridian. This assumption allows to have only one observatory under the influence of the EEJ. To estimate the EEJ, we evaluated the difference between the H component measured by the two pair stations after the subtraction of the nighttime baseline. Finally the EEJ signal at the longitude of the equatorial stations is obtained referred to as Δ H . The ground stations information used for the EEJ estimation are reported in Table  1 .

Magnetometer data were obtained from INTERMAGNET magnetometer array network. INTERMAGNET is a consortium of observatories and operating institutes that guarantees a common standard of data released to the scientific community, allowing the possibility to compare the measurements carried out at different observation points.

Analytical model for top-side Ionospheric Electric field variation induced by a GRB

In order to develop a model able to represent the effect of GRB impinging the top-side ionosphere, we started from the ionospheric Ohm’s law 65 :

where E is the electric field, B is the ambient magnetic field, σ is the conductivity tensor with σ p , σ H , and σ 0 being respectively the Pedersen, Hall, and parallel conductivity. The formation of the electric current in the ionized layer is caused by the difference between the velocities of ions (typically NO + , O \({}_{2}^{+}\) , O + , H + , H \({}_{e}^{+}\) and N + ) and electrons. In ionosphere, the temporal variability of the electrodynamics processes is slow enough that one can ignore the displacement current in Maxwell’s equations (i.e., the term ∂E/∂ t ) 41 , therefore Ampère-Maxwell law reduces to

where μ 0 is vacuum magnetic permeability. By combining equations ( 1 ) and ( 2 ), we obtain:

At about 500 km (i.e. CSES orbiting altitude) both σ H and σ p are negligible with respect to σ 0 (see Fig.  7 in Denisenko et al. 66 ). As a consequence, equation ( 3 ) simplifies to

Once σ 0 is known, equation ( 4 ) can be numerically solved to obtain the E field behaviour. Equation for the parallel conductivity in the ionosphere as given by Maeda 42 reads

where n e is the electron density, ν e , n is electron-neutral collision frequency, ν e , i electron-ion collision frequency, q e is the unsigned electric charge (i.e. 1.602  ⋅  10 −19 C ), and m e is the electron mass (i.e. 9.109  ⋅  10 −31 k g ). Following the results of Aggarwal et al. 67 , we can estimate the electron collision frequency at about 500 km as ν e  = 10 2 sec −1 .

Being σ 0 directly dependent on the electron density, it is straightforward that any variation of n e causes a changes in E. In general, the rate of change of the electron density is expressed by a continuity equation 68 :

where A is the production coefficient and L the loss coefficient by recombination/losses. Naturally, the recombination coefficient depends of what ion species are present, and hence on the ionospheric altitude. At high altitudes (>200km, i.e. top-side ionosphere) where O + is the dominant ion species, L becomes proportional to the electron density 68 . So, equation ( 7 ) becomes:

where β is the loss rate. Equation ( 8 ) is valid only at altitudes higher than 200 km (and hence at the altitude of our electric field observations), being L , at lower altitudes (bottom-side ionosphere), proportional to the square of the electron density 68 . To simulate the production rate induced by a GRB, we used a Gaussian impulsive function of the form \(\alpha {e}^{-{(\frac{t-{t}_{0}}{{s}_{0}})}^{2}}\) , so that equation ( 8 ) can be written as:

where α is the production rate induced by the GRB that depends on its photon flux, t 0 is the time of the maximum production rate, and s 0 is the width of the pulse.

Putting together equations ( 4 ), ( 5 ), and ( 9 ), and assuming at 500 km both an average electron density of 1.2  ⋅  10 11 cm −3   69 and an average loss rate coefficient of 0.6  ⋅  10 −6 sec −1   70 , 71 , 72 , we can model the electric field variation induced by a GRB as a function of the ionospheric plasma density variation at 500 km of altitude. Figure  6 shows the results of our model for different ratios between production ( α ) and loss ( β ) rate. It can be easily seen that the effect of a GRB is negligible if α / β  < 3. To obtain results similar to what was observed on October 9th, 2022, our model requires a production-to-loss ratio greater than 5.

The usage of a formalism directly related to the ratio between α and β allows the model to being independent (for the present analysis) of the calculation of a realistic photon production rate caused by a GRB, whose evaluation needs a Montecarlo approach and the estimation of the real top-side ionospheric ion cross-section, which is out of the scope of the present work but a more accurate modelling of the effect of a GRB on the top-side ionospheric electric field is in progress.

Anyway, despite being very simplified, our model can be used to give a first quantitative explanation of the effect induced in the top-side ionosphere by GRB221009.

Data availability

We cannot supply our source data in any public depository since they are property of: European Space Agency (INTEGRAL satellite data); Italian Space Agency (CSES satellite data); International Real-time Magnetic Observatory Network (ground magnetometer data); University NAVSTAR Consortium (GNSS satellite data). Anyway all of them can be freely downloaded from the relative website after registration. CSES satellite data are freely available at the LEOS repository ( www.leos.ac.cn/#/home , accessed on 08/09/2023) after registration; GNSS data are freely available at University NAVSTAR Consortium ( https://www.unavco.org/accessedon08/09/2023 ) after registrtion. INTEGRAL SPI data are freely available at the ISDC ( https://www.isdc.unige.ch/integral/ , accessed on 08/09/2023) repository. INTEGRAL PiCsIt/IBIS data are proprietary data of authors of the paper without any restriction. Ground magnetometer data are freely available at INTERMAGNET website ( https://imag-data.bgs.ac.uk/GIN_V1/GINForms2 , accessed on 08/09/2023). The datasets generated during and/or analysed during the current study are available from the corresponding author on request.

Code availability

Codes used to produce results and figures were obtained using Matlab software package. They are not public but can be made available upon request to the corresponding author.

Fishman, G. & Inan, U. Observation of an ionospheric disturbance caused by a gamma-ray burst. Nature 331 , 418–420 (1988).

Article   ADS   Google Scholar  

Hargreaves, J. K. The upper atmosphere and solar-terrestrial relations - An introduction to the aerospace environment (Cambridge University Press, 1979).

Tanaka, Y. et al. First very low frequency detection of short repeated bursts from magnetar sgr j1550- 5418. Astrophys. J. Lett. 721 , L24 (2010).

Inan, U. S. et al. Massive disturbance of the daytime lower ionosphere by the giant γ -ray flare from magnetar sgr 1806-20. Geophys. Res. Lett . 34 (2007).

Maeda, K. et al. Ionospheric effects of the cosmic gamma ray burst of 29 march 2003. Geophys. Res. Lett . 32, https://doi.org/10.1029/2005GL023525 (2005).

Huang, W.-G., Gu, S.-F. & Shen, H. Response of the total electron content of terrestrial ionosphere to grb041227. Chin. Astron. Astrophys. 32 , 65–72 (2008).

Hudec, R. et al. Detection of grbs and ots by all-sky optical and sid monitors. Adv. Astron . (2010).

Slosiar, R., Hudec, R., Kocka, M., Marko, R. and Zatko, M. Indirect detections and analyses of grbs by ionospheric response: toward a sid-monitor network. In AIP Conference Proceedings, vol. 1358, 393–396 (American Institute of Physics, 2011).

Sentman, D., Heavner, M., Baker, D., Cayton, T. and Fraser, B. Effects of solar storms on the schumann resonances in late 1989. In 10th Annual Conference on Atmospheric Electricity. Soc. of Atmos. Electr. of Japan, Osaka, Japan (1996).

Price, C. & Mushtak, V. The impact of the august 27, 1998, γ -ray burst on the schumann resonances. J. Atmos. Sol.-Terrest. Phys. 63 , 1043–1047 (2001).

Tanaka, Y. T. et al. Detection of transient elf emission caused by the extremely intense cosmic gamma-ray flare of 27 december 2004. Geophys. Res. Lett . 38, https://doi.org/10.1029/2011GL047008 (2011).

Schumann, W. O. Über die strahlungslosen eigenschwingungen einer leitenden kugel, die von einer luftschicht und einer ionosphärenhülle umgeben ist. Z. f.ür. Naturforsch. A 7 , 149–154 (1952).

Article   ADS   MATH   Google Scholar  

Hayes, L. A. & Gallagher, P. T. A significant sudden ionospheric disturbance associated with gamma-ray burst grb 221009a. Res. Notes AAS 6 , 222 (2022).

Pal, S. et al. First detection of global ionospheric disturbances associated with the most powerful gamma ray burst grb221009a. Atmosphere 14 , 217 (2023).

Dichiara, S. et al. Swift j1913.1+1946 a new bright hard x-ray and optical transient. GRB Coord. Netw. 32632 , 1 (2022).

Google Scholar  

Kennea, J. A. & Williams, M. Grb 221009a: Swift detected transient may be grb. GRB Coord. Netw. 32635 , 1 (2022).

Veres, P., Burns, E., Bissaldi, E., Lesage, S. & Roberts, O. Grb 221009a: Fermi gbm detection of an extraordinarily bright grb. GRB Coord. Netw. 32636 , 1 (2022).

Bissaldi, E., Omodei, N. & Kerr, M. Grb 221009a or swift j1913.1+1946: Fermi-lat detection. GRB Coord. Netw. 32637 , 1 (2022).

Negoro, H. et al. MAXI/GSC detection of the new X-ray transient Swift J1913.1+1946. Astron. Telegr. 15651 , 1 (2022).

ADS   Google Scholar  

Ursi, A. et al. Grb 221009a (swift j1913.1+1946): Agile/mcal detection. GCN Circ . 32650 (2022).

Piano, G. et al. Grb 221009a (swift j1913.1+1946): Agile/grid detection. GCN Circ . 32657 (2022).

Svinkin, D. et al. Ipn triangulation of extremely bright grb 221009a. GCN Circ . 32641 (2022).

Gotz, D., Mereghetti, S., Savchenko, V., Ferrigno, C. & Bozzo, E. Grb221009a/swift j1913.1+1946: Integral spi/acs observations. GRB Coord. Netw. 32691 , 1 (2022).

Winkler, C. et al. The integral mission. Astron. Astrophys. 411 , L1–L6 (2003).

Article   ADS   CAS   Google Scholar  

Savchenko, V. et al. Grb221009a: Integral detection of hard x-ray emission up to 38 hours after trigger. GRB Coord. Netw. 32691 , 1 (2022).

de Ugarte Postigo, A. et al. Grb 221009a: Spectroscopic detection of emerging sn features. GRB Coord. Netw. 32800 , 1 (2022).

de Ugarte Postigo, A. et al. Grb 221009a: Redshift from x-shooter/vlt. GRB Coord. Netw. 32648 , 1 (2022).

Diego, P. et al. The electric field detector on board the china seismo electromagnetic satellite-in-orbit results and validation. Instruments 5 , https://doi.org/10.3390/instruments5010001 (2021).

Shen, X., Zong, Q.-G. & Zhang, X. Introduction to special section on the china seismo-electromagnetic satellite and initial results. Earth Planet. Phys. 2 , 439 (2018).

Smith, A. R. A., Beggan, C. D., Macmillan, S. & Whaler, K. A. Climatology of the auroral electrojets derived from the along-track gradient of magnetic field intensity measured by pogo, magsat, champ, and swarm. Space Weather. 15 , 1257–1269 (2017).

Materassi, M. et al. Stepping into the equatorward boundary of the auroral oval: preliminary results of multi scale statistical analysis. Ann. Geophys. 62 , GM455–GM455 (2019).

Consolini, G. et al. On turbulent features of e × b plasma motion in the auroral topside ionosphere: Some results from cses-01 satellite. Remote. Sens . 14, https://doi.org/10.3390/rs14081936 (2022).

Ding, G.-X., He, F., Zhang, X.-X. & Chen, B. A new auroral boundary determination algorithm based on observations from timed/guvi and dmsp/ssusi. J. Geophys. Res. Space Phys. 122 , 2162–2173 (2017).

Mitra, A. P. Ionospheric effects of solar flares, vol. 46 (Springer Nature, 1974).

Appleton, E. V. A note on the “sluggishness" of the ionosphere. J. Atmos. Terr. Phys. 3 , 282–284 (1953).

Nina, A., Cadež, V., Sreckovic, V. & Šulic, D. Altitude distribution of electron concentration in ionospheric d-region in presence of time-varying solar radiation flux. Nucl. Instrum. Methods Phys. Res. Sect. B: Beam Interact. Mater. At. 279 , 110–113 (2012).

Žigman, V., Grubor, D. & Šulic, D. D-region electron density evaluated from vlf amplitude time delay during x-ray solar flares. J. Atmos. Sol.-terrest. Phys. 69 , 775–792 (2007).

Basak, T. & Chakrabarti, S. K. Effective recombination coefficient and solar zenith angle effects on low-latitude d-region ionosphere evaluated from vlf signal amplitude and its time delay during x-ray solar flares. Astrophys. Space Sci. 348 , 315–326 (2013).

Hegde, S., Bobra, M. G. & Scherrer, P. H. Classifying signatures of sudden ionospheric disturbances. Res. Notes AAS 2 , 162 (2018).

Dellinger, J. H. Sudden ionospheric disturbances. Terr. Magn. Atmos. Electr. 42 , 49–53 (1937).

Article   Google Scholar  

Kelley, M. C. The Earth’s ionosphere: Plasma physics and electrodynamics (Academic press, 2009).

Maeda, K.-I. Conductivity and drifts in the ionosphere. J. Atmos. Terr. Phys. 39 , 1041–1053 (1977).

Mahrous, A. Ionospheric response to magnetar flare: signature of sgr j1550-5418 on coherent ionospheric doppler radar. Ann. Geophys. 35 , 345–351 (2017).

Mondal, S. K. & Chakrabarti, S. K. Earth’s Ionosphere as a Gigantic Detector of Extra-terrestrial Energetic Phenomena: A eview. AIP Conf. Proc. 1286 , 311–330 (2010).

Rastogi, R. Electromagnetic induction due to solar flares at equatorial stations. J. Atmos. Sol.-Terrest. Phys. 63 , 599–604 (2001).

Rastogi, R. Effect of solar disturbances on the geomagnetic h, y, and z fields in american equatorial electrojet stations. solar flare effects. J. Indian Geophys. Union 7 , 43–51 (2003).

Matsushita, S. & Xu, W.-Y. Equivalent ionospheric current systems representing solar daily variations of the polar geomagnetic field. J. Geophys. Res. Space Phys. 87 , 8241–8254 (1982).

Campbell, W. The regular geomagnetic-field variations during quiet solar conditions. Geomatik 3 , 385–460 (1989).

Chapman, S. & Rao, K. R. The h and z variations along and near the equatorial electrojet in india, africa and the pacific. J. Atmos. Terrest. Phys. 27 , 559–581 (1965).

Lühr, H. & Manoj, C. The complete spectrum of the equatorial electrojet related to solar tides: Champ observations. Ann. Geophys. 31 , 1315–1331 (2013).

Nogueira, P. A. B. et al. Modeling the equatorial and low-latitude ionospheric response to an intense x-class solar flare. J. Geophys. Res. Space Phys. 120 , 3021–3032 (2015).

Campbell, W. H. An introduction to quiet daily geomagnetic fields. Quiet Dly. Geomagn. Fields 315–331 (1989).

Scherrer, D. et al. Distributing space weather monitoring instruments and educational materials worldwide for ihy 2007: The awesome and sid project. Adv. Space Res. 42 , 1777–1785 (2008).

Ubertini, P. et al. Ibis: The imager on-board integral. Astron. Astrophys. 411 , L131–L139 (2003).

Vedrenne, G. et al. Spi: The spectrometer aboard integral. Astron. Astrophys. 411 , L63–L70 (2003).

Kuulkers, E. et al. Integral reloaded: Spacecraft, instruments and ground system. N. Astron. Rev. 93 , 101629 (2021).

Savchenko, V. et al. Integral ibis, spi, and jem-x observations of lvt151012. Astron. Astrophys. 603 , A46 (2017).

Shen, X. et al. The state of the art of the china seismo-electromagnetic satellite mission. Sci. China Technol. Sci. 61 , 634 (2018).

Huang, J. et al. The electric field detector (efd) onboard the zh-1 satellite and first observational results. Earth Planet. Phys. 2 , 469 (2018).

Selvaggi, G. et al. Rete integrata nazionale gps. Istituto Nazionale di Geofisica e Vulcanologia (INGV): Rome, Italy (2016).

D’Angelo, G. et al. Investigation of the physical processes involved in gnss amplitude scintillations at high latitude: A case study. Remote. Sens. 13 , 2493 (2021).

Ciraolo, L., Azpilicueta, F., Brunini, C., Meza, A. & Radicella, S. M. Calibration errors on experimental slant total electron content (tec) determined with gps. J. Geod. 81 , 111–120 (2007).

Cesaroni, C. et al. L-band scintillations and calibrated total electron content gradients over brazil during the last solar maximum. J. Space Weather. Space Clim. 5 , A36 (2015).

Soares, G. et al. Equatorial counter electrojet longitudinal and seasonal variability in the american sector. J. Geophys. Res. Space Phys. 123 , 9906–9920 (2018).

Vasyliunas, V. M. The physical basis of ionospheric electrodynamics. Ann. Geophys. 30 , 357–369 (2012).

Denisenko, V., Rycroft, M. & Harrison, R. Mathematical simulation of the ionospheric electric field as a part of the global electric circuit. Surv. Geophys. 40 , 1–35 (2019).

Aggarwal, K., Nath, N. & Setty, C. Collision frequency and transport properties of electrons in the ionosphere. Planet. Space Sci. 27 , 753–768 (1979).

Briggs, B. H. & Rishbeth, H. An analogue solution of the continuity equation of the ionospheric f region. Proc. Phys. Soc. 78 , 409 (1961).

Pignalberi, A. et al. Inter-calibration and statistical validation of topside ionosphere electron density observations made by cses-01 mission. Remote. Sens. 14 , 4679 (2022).

Ratcliffe, J. A., Schmerling, E. & Setty, C. The rates of production and loss of electrons in the f region of the ionosphere. Philos. Trans. R. Soc. Lond. Ser. A, Math. Phys. Sci. 248 , 621–642 (1956).

Taylor, G. Integrated electron production and loss rates in the ionosphere. Planet. Space Sci. 13 , 507–520 (1965).

smith III, F. Electron production and loss rates in the f region. J. Geophys. Res. 73 , 7385–7398 (1968).

Download references

Acknowledgements

The authors thank the Italian Space Agency for the financial support under the contract ASI “LIMADOU Scienza+” n ∘ 2020-31-HH.0, and the financial support under the “ INTEGRAL ASI-INAF” agreement n ∘ 2019-35-HH.0. M.P., and G.d.A. thank the ISSI-BJ project “The electromagnetic data validation and scientific application research based on CSES satellite” and Dragon 5 cooperation 2020-2024 (ID. 59236). Part of the research leading to the result has receiving founding support from the EuropeanUnion’s Horizon 2020 Programme under the AHEAD2020 project (grant agreement n. 871158). This material is based on services provided by the GAGE Facility, operated by EarthScope Consortium, with support from the National Science Foundation, the National Aeronautics and Space Administration, and the U.S. Geological Survey under NSF Cooperative Agreement EAR-1724794. The authors thank the INGV group for providing the RING data. We thank the national institutes that support INTERMAGNET for promoting high standards of the magnetic observatory practice ( www.intermagnet.org ) used in this paper.

Author information

These authors contributed equally: Mirko Piersanti, Pietro Ubertini, Roberto Battiston, Angela Bazzano, Giulia D’Angelo, James G. Rodi, Piero Diego.

Authors and Affiliations

Department of Physical and Chemical Sciences, University of L’Aquila, 67100, L’Aquila, Italy

Mirko Piersanti

National Institute of Astrophysics, IAPS, Rome, 00133, Italy

Mirko Piersanti, Pietro Ubertini, Angela Bazzano, Giulia D’Angelo, James G. Rodi, Piero Diego, Igor Bertello, Antonio Cicone, Fabrizio De Angelis, Emiliano Fiorenza, Bruno Martino, Alfredo Morbidini, Fabrizio Nuccilli, Emanuele Papini, Alexandra Parmentier, Dario Recchiuti, Andrea Russi, Silvia Tofani, Nello Vertolli & Ugo Zannoni

Department of Physics, University of Trento, Povo, Italy

Roberto Battiston, Andrea Di Luca, Francesco Maria Follega, Giuseppe Gebbia, Roberto Iuppa, Alessandro Lega, Alessio Perinelli, Dario Recchiuti, Ester Ricci, Veronica Vilona & Paolo Zuccon

TIFPA-INFN, Povo, 38123, Trento, Italy

Roberto Battiston, William J. Burger, Marco Cristoforetti, Andrea Di Luca, Francesco Maria Follega, Giuseppe Gebbia, Roberto Iuppa, Alessandro Lega, Coralie Neubüser, Francesco Nozzoli, Alessio Perinelli, Ester Ricci & Paolo Zuccon

National Institute of Natural Hazards, Ministry of Emergency Management of China, Beijing, 100085, People’s Republic of China

Zhima Zeren

INFN, University of Rome Tor Vergata, Rome, 00133, Italy

Roberto Ammendola, Davide Badoni, Simona Bartocci, Piero Cipollone, Livio Conti, Cinzia De Donato, Cristian De Santis, Matteo Martucci, Giuseppe Masciantonio, Matteo Mergè, Francesco Palma, Alexandra Parmentier, Piergiorgio Picozza, Gianmaria Rebustini, Alessandro Sotgiu, Roberta Sparvoli & Vincenzo Vitale

INFN - Sezione di Torino, 10125, Torino, Italy

Stefania Beolè, Silvia Coli, Stefania Perciballi & Umberto Savino

INFN-Sezione di Napoli, Naples, 80126, Italy

Donatella Campana, Marco Mese, Giuseppe Osteria, Beatrice Panico, Francesco Perfetto & Valentina Scotti

Dipartimento di Ingegneria e Scienze dell’Informazione e Matematica, University of L’Aquila, 67100, L’Aquila, Italy

Antonio Cicone

Uninettuno University, 00186, Rome, Italy

Livio Conti

University of Bologna, Bologna, 40127, Italy

Andrea Contin, Alberto Oliva & Federico Palmonari

INFN - Sezione di Bologna, 40127, Bologna, Italy

Andrea Contin, Marco Lolli, Alberto Oliva, Federico Palmonari, Michele Pozzato & Zuleika Sahnoun

Fondazione Bruno Kessler, 38123, Povo, TN, Italy

Marco Cristoforetti

CNR, V. Fosso del Cavaliere 100, 00133, Rome, Italy

Bruno Martino

Agenzia Spaziale Italia, Rome, 00133, Italy

Matteo Mergè & Simona Zoffoli

Università degli Studi di Napoli Federico II, 80126, Naples, Italy

Marco Mese, Beatrice Panico & Valentina Scotti

Department of Physics, University of Rome Tor Vergata, Rome, 00133, Italy

Piergiorgio Picozza & Roberta Sparvoli

INFN-LNF, Frascati, Rome, 00100, Italy

Marco Ricci

IFAC-CNR, Sesto Fiorentino, Florence, 50019, Italy

Sergio B. Ricciarini

National Space Science Center, Chinese Academy of Sciences, Beijing, 100190, People’s Republic of China

You can also search for this author in PubMed   Google Scholar

Contributions

M.P. writing–original draft, formal analysis, methodology and supervision; P.U. writing–revision, editing and methodology; R.B. writing–revision, formal analysis and validation; A.B. writing–revision; G.d.A. writing–revision, formal analysis; J.C.R. writing–revision, formal analysis; P.D. data validation, funding, Z.Z data validation. R.A., D.B., S.B., S.Be., I.B., W.J.B., D.C., A.C., P.C., S.C., L.C., A.Co., M.C., F.d.A., C.d.D., C.d.S., A.d.L., E.F., F.M.F., G.G., R.I., A.L., M.L., B.M., G.M., M.M., M.Me., M.Mes, A.M., C.N, F.N., F.Nu, A.O., G.O., F.P., F.Pa., B.P., E.P., A.P., S.P., F.P., A.Pe., P.P., M.Po., G.R., D.R., E.R., M.R., S.B.R., A.R., X.S., Z.S., U.S., V.S., A.S., R.S., S.T., N.V., V.V., V.Vi, U.Z., S.Z., P.Z., are part of the CSES-Limadou Collaboration whose significant contribution made satellite observations possible.

Corresponding author

Correspondence to Mirko Piersanti .

Ethics declarations

Competing interests.

This research received no external funding. The authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Peer review file, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and Permissions

About this article

Cite this article.

Piersanti, M., Ubertini, P., Battiston, R. et al. Evidence of an upper ionospheric electric field perturbation correlated with a gamma ray burst. Nat Commun 14 , 7013 (2023). https://doi.org/10.1038/s41467-023-42551-5

Download citation

Received : 04 February 2023

Accepted : 13 October 2023

Published : 14 November 2023

DOI : https://doi.org/10.1038/s41467-023-42551-5

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

We use cookies to enhance our website for you. Proceed if you agree to this policy or learn more about it.

• Essay Database >
• Essays Samples >
• Essay Types >
• Research Paper Example

Electricity Research Papers Samples For Students

355 samples of this type

WowEssays.com paper writer service proudly presents to you an open-access catalog of Electricity Research Papers aimed to help struggling students tackle their writing challenges. In a practical sense, each Electricity Research Paper sample presented here may be a pilot that walks you through the critical phases of the writing process and showcases how to compose an academic work that hits the mark. Besides, if you need more visionary help, these examples could give you a nudge toward a fresh Electricity Research Paper topic or inspire a novice approach to a threadbare subject.

In case this is not enough to satisfy the thirst for effective writing help, you can request customized assistance in the form of a model Research Paper on Electricity crafted by an expert from scratch and tailored to your particular requirements. Be it a simple 2-page paper or an in-depth, lengthy piece, our writers specialized in Electricity and related topics will deliver it within the pre-set timeframe. Buy cheap essays or research papers now!

Free Research Paper On Assuming The Diode Is Forward Biased, Using Kvl We Have

Research paper on heated hose pipes, wind energy research papers examples, introduction.

Don't waste your time searching for a sample.

Get your research paper done by professional writers!

Just from $10/page

Emerson Electric Company Research Paper Example

Various operational challenges facing emerson electric company, good example of research paper on lincoln electric strategy and itw strategy, industry analysis.

Industry Structure Welding is an important process for different types of structures such as roads, bridges, railroads, oil facilities and general infrastructure. Arc welding is the most popular type of welding in the welding industry. The potential for growth in the industry remains high because infrastructural structures are always coming up in different places in the world. One of the most critical success factors is being able to adapt competitively to an increasingly innovative industry. Having skilled employees who are committed to the company objectives and customer-focused is an invaluable asset to any company in this industry.

Industry Rivalry

Research paper on potential causes of co2 pollution and feasible solutions, example of research paper on governments plan of action to secure electricity supply, free research paper about alternatively fueled vehicles policies, free the development of alternate current and voltage through time from its beginning research paper sample, electrical engineering: a journey through the sub-disciplines research paper example, tesla the electric car research paper samples, good research paper about what keeps people from buying electrical cars, free research paper about tesla motors case analysis, please type your name here, environmental and consumer advantages and disadvantages of hybrid vehicles research papers examples.

Alternative engines in vehicles have been produced by the automotive industry to address the issue of the diminishing natural oil reserves and the continual degradation of the environment. Eco-friendly vehicles like electric and hybrid vehicles have become the most popular non-pollutant vehicles to arrive in the market.

Environmental and consumer advantages of hybrid vehicles

Good research paper about south and east asia, free research paper about chuck berry- passed off due to exhaustion on-stage, example of hybrid cars are better for the environment research paper, example of environmental impacts research paper, order# 209828430, fname lname research paper sample, technology 101, good a paper presented to research paper example.

(Institution Name)

BIOGRAPHY: BOB DYLAN

(Name of the Professor)

In Requirement of fulfillment of the Course

History of wireless power transmission research paper, tesla inc research paper examples, introduction:.

TESLA is an American Car Manufacturing company that designs, manufactures and sells electric cars and related components. The company trades on NASDAQ under the ticker symbol of ‘TSLA’. The company has been operating under losses since 10 years with Net loss for 2012 amounting to $396213.

Analyzing Accounts Payables:

Nuclear power research paper examples, geothermal energy research paper, uk energy policy research papers example, example of motor research paper, alternative energy resource research paper sample, geothermal heating research paper examples, advantages of hybrid cars research paper, hybrid cars, research paper on solar energy, efficiency of renewable energy research papers example, good research paper on the future of electric car, good research paper on technology.

Meeting the demands of this generation requires embrace to new technology and a greater sense to the future. The world at this juncture suffers from the financial crisis that need to be solved; there is energy crisis and a cry for the greener environment. This is what Chevy Volt kept in perspective and developed one of the automotive that has satisfied most of its buyer. The vehicle is pocket friendly, effective, safer and modern. This has reduced environmental degradation, and the life span of the battery is long lasting ("The car of the future | Drive For Innovation – AVNETExpress").

Major Innovation of Chevy volts

Renewable energy in kenya research paper, what were the effects of electricity to labor; ultimately destroying the american research papers examples, research paper: electric power through 1932, research paper on thomas edison dc verses tesla ac voltage.

Electricity can flow in two different ways; either in direct current (DC) or in alternating current (AC). Electricity is a moving electron through or along a conductor, for instance a wire, which have been connected for energy (Daly, Peckham & AudioGO, 2013). Therefore, the variation between DC and AC is the direction of electron flow. In direct current, the electrons flow in one direction (forward). In alternating current, electrons keep changing directions, sometimes forward and at times backwards (Daly, 2013).

Origins of DC and AC Current

Good research paper about history of fire alarms, free science milestones and their impact on society research paper example, on the origin of species, research paper on energy efficiency, product overview: refi-back pack research papers examples, research paper on chernobyl disaster.

<Student’s name> <Professor’s name>

Free Research Paper On Driving National And EU Energy Policy With Electric Vehicles

Free research paper on fukushima daiichi nuclear plant, free research paper on smart grid computing, example of research paper on domestic energy security, commentary one, renewable energy research paper, research paper on ge (general electric), transportation of energy research paper, commentary 1, transportation of energy research paper examples, research paper on general electric, car buying research paper sample, tesla coil research paper example, example of energy sector research paper, energy sector research paper examples, construction of kariba dam research paper examples, example of structure of the atom research paper, engineering research paper examples, solar energy project in flagstaff research paper sample, future changes research paper examples, research paper on examining your communitys source of energy.

Password recovery email has been sent to [email protected]

Use your new password to log in

You are not register!

By clicking Register, you agree to our Terms of Service and that you have read our Privacy Policy .

Now you can download documents directly to your device!

Check your email! An email with your password has already been sent to you! Now you can download documents directly to your device.

or Use the QR code to Save this Paper to Your Phone

The sample is NOT original!

Short on a deadline?

Don't waste time. Get help with 11% off using code - GETWOWED

No, thanks! I'm fine with missing my deadline

• Reference Manager
• Simple TEXT file

People also looked at

Original research article, leveraging sanitized data for probabilistic electricity market prediction: a singapore case study.

• 1 Energy Research Institute @NTU, Nanyang Technological University, Singapore, Singapore
• 2 The Industrial Training Centre, Shenzhen Polytechnic University, Shenzhen, China
• 3 College of Physics and Optoelectronic Engineering, Shenzhen University, Shenzhen, China
• 4 Electric Power Research Institute, State Grid Liaoning Electric Power Company Ltd., Shenyang, China
• 5 PacificLight Power Pte Ltd., Singapore, Singapore

In deregulated electricity markets, predicting price and load is a common practice. However, market participants and shareholders often seek deeper insights into other system statuses associated with price prediction, such as power flow and market share of generation companies (GenCos). These insights are challenging to obtain using purely data-driven methods. This paper proposes a physics-based solution for the probabilistic prediction of market-clearing outcomes, using real sanitized offer data from the National Electricity Market of Singapore (NEMS). Our approach begins with approximating the generator offers that have been historically cleared. Using this pool of offer data, we propose a probabilistic market-clearing process. This process allows for the probabilistic prediction of market prices. By considering the power system network and its constraints, we also naturally obtain probabilistic predictions of power flow and market shares. We validate our approach using actual NEMS data. Our findings show that while the overall performance of price prediction is comparable to existing methods, our proposed method can also provide probabilistic predictions of other associated system operating conditions. Furthermore, our method enables scenario studies, such as the impact of demand-side participation and the penetration of rooftop photovoltaic (PV) systems on the Uniform Singapore Energy Price (USEP).

1 Introduction

The past few decades have witnessed the liberalization of electricity markets all over the world ( Bunn et al., 2021 ; Yang et al., 2019 ), including the National Electricity Market of Singapore (NEMS) ( Services, 2021 ; Energy Market Authority of Singapore, 2018 ). As a fundamental aspect of these markets, electricity price prediction has been the subject of extensive research ( Zhao et al., 2008 ; Abedinia et al., 2017 ; Wan et al., 2017 ; Monte et al., 2018 ; Nowotarski and Weron, 2018 ; Brusaferri et al., 2019 ; Chai et al., 2019 ; Afrasiabi et al., 2020 ; He et al., 2020 ; Hong et al., 2020 ; Li et al., 2020 ; Fraunholz et al., 2021 ; Taylor, 2021 ; Uniejewski and Weron, 2021 ; Meng et al., 2022 ; Heidarpanah et al., 2023 ). However, most existing tools prioritize price prediction, often overlooking the comprehensive insights that additional system parameters, such as power flow and market share of generation companies (GenCos), can offer. The repercussions of such an oversight can be manifold. Without these nuanced insights, prediction models could inadvertently misjudge critical supply–demand imbalances at specific nodes. This, in turn, can lead to potential inaccuracies in price forecasts. GenCos, if left uninformed about these pivotal parameters, might grapple with challenges in streamlining their offer-making decisions, inadvertently paving the way for market inefficiencies. Furthermore, the lack of these comprehensive insights might foster an environment ripe for market volatility, with speculative bidding amplifying price fluctuations.

Over the past 2 decades, prediction techniques have significantly advanced ( Zhao et al., 2008 ; Abedinia et al., 2017 ; Wan et al., 2017 ; Monte et al., 2018 ; Nowotarski and Weron, 2018 ; Brusaferri et al., 2019 ; Chai et al., 2019 ; Afrasiabi et al., 2020 ; He et al., 2020 ; Hong et al., 2020 ; Li et al., 2020 ; Fraunholz et al., 2021 ; Taylor, 2021 ; Uniejewski and Weron, 2021 ; Meng et al., 2022 ; Zhou et al., 2022 ; Heidarpanah et al., 2023 ). Point forecasting techniques, for instance, have been widely adopted in this field ( Abedinia et al., 2017 ; Fraunholz et al., 2021 ; Hong et al., 2020 ; Heidarpanah et al., 2023 ). In the NEMS, the market operator publishes point forecasts for demand and the Uniform Singapore Energy Price (USEP) ( Energy Market Company, 2023 ). Probabilistic forecasting , however, goes a step further. It reflects the probabilistic and heteroscedastic nature of electricity price and load, providing predictions in the form of intervals ( Zhao et al., 2008 ; Wan et al., 2017 ; Taylor, 2021 ; He et al., 2020 ), quantiles ( Uniejewski and Weron, 2021 ), and densities ( Chai et al., 2019 ; Brusaferri et al., 2019 ; Li et al., 2020 ; Afrasiabi et al., 2020 ). Zhao et al. (2008) and Wan et al. (2017) construct optimal prediction intervals. Taylor (2021) proposes expectile-bounded intervals. He et al. (2020) introduce a method based on a deep neural network model. To address the vulnerability of quantile regression averaging (QRA) to low-quality predictors, a regularized variant of QRA is proposed ( Uniejewski and Weron, 2021 ). Density forecasting , which provides comprehensive uncertainty information, has its own advantages. Chai et al. (2019) propose a reliable strategy for constructing predictive densities oriented toward continuous ranked probability scores. A Bayesian deep learning-based technique is developed by Brusaferri et al. (2019) . Li et al. (2020) combine density probabilistic load forecasts to enhance the performance of the final probabilistic forecasts. Afrasiabi et al. (2020) construct a deep neural network (DNN) model from historical data to directly predict the probability density function (PDF) of residential loads based on past time series.

Many AI-based models inherently focus on identifying patterns in historical data, often sidelining real-world grid considerations, such as transmission capacities, voltage limits, and other network-related constraints. This data-driven approach can lead to predictions that, while statistically accurate, are operationally infeasible, particularly in complex scenarios or during events that stress the grid. Furthermore, these models typically struggle to derive additional system parameters, such as local marginal price (LMP), power flow, and the market share of GenCos, which are crucial for a comprehensive understanding of market dynamics. The “black-box” nature of these models also presents challenges in interpretability, making it difficult to understand the rationale behind certain predictions.

In contrast, simulation-based prediction methods, which account for the intricacies of the power system network, offer a way to obtain these additional system parameters. However, there exists limited literature on this topic. Ji et al. (2017) provide a forecasting method from the vantage point of a system operator who has access to system operating conditions. In a short-term forecasting algorithm proposed by Zhou et al. (2011) , supply-offer behaviors are assumed static. Quadratic cost function and lossless power flow are also assumed. Bo and Li (2009) investigate the impact of load uncertainty on LMP forecasting. In these studies, GenCos’ offers are often assumed or assigned as fuel-cost-based functions due to the sanitization of generator offer data in a typical electricity market.

One approach to this challenge is to approximate offers from accessible historical data. In Durvasulu and Hansen (2018) , generator types are identified and clustered using only publicly available data in the PJM market. Market-based cost functions are then fitted and used in optimal power flow (OPF) calculations to obtain the marginal cost. However, this approach becomes more complicated due to the variability of available data across different markets. For instance, in markets such as the NEMS, the available dataset is more limited. Not only is the generator ID sanitized, but so is the offer quantity. LMP is also not available, as detailed in Figure 3 , Section 2 .

This paper presents a comprehensive solution for probabilistic market prediction that addresses several key aspects in a single package:

1. It takes into account the issue of data availability and bases the modeling on sanitized data.

2. It incorporates network constraints, enabling the prediction of other system parameters, such as power flow and generator output, which are associated with price prediction.

3. It offers probabilistic prediction on price and other market-clearing outcomes.

4. It serves as a platform for studying future scenarios, including the impact of demand-side participation and increased penetration of renewable energy.

The proposed methodology is validated using real NEMS data. To the best of our knowledge, this work is the first to provide a probabilistic prediction for price, market share, and power flow using real-world sanitized data.

The framework for typical prediction methods is shown in Figure 1A . Figure 1B shows how we address the four points mentioned previously. To tackle the data availability issue (point 1), the first step is to estimate historically cleared offers based on accessible data, for which we develop an optimization method. In the second step, we sort the offer data estimated from the first step using a clustering technique. We derive clusters of functions, representing possible offers using regression, and then assign them to generators, each with associated conditional probabilities. The obtained offer curves enable the calculation of optimal power flow (OPF), addressing point 2. In the final step, we propose a conditional probabilistic market-clearing process. With load forecast as the input and using Monte Carlo simulation, we can obtain probabilistic estimates of price along with other market-clearing outcomes, addressing point 3. Since our methodology is physics-based, we can conduct scenario studies with assumed system parameters, addressing point 4.

FIGURE 1 . Comparison of market prediction methods: (A) typical method for market price prediction; (B) proposed method for comprehensive market prediction.

The remainder of this paper is structured as follows: Sections 2–4 detail the proposed methodology against the backdrop of the NEMS. A solution for the historical offer estimation with sanitized information is proposed in Section 2 . The formulation of conditional offer functions is presented in Section 3 . A probabilistic clearing process for market prediction is proposed in Section 4 . In Section 5 , actual NEMS data are used to validate the methodology. Its performance is benchmarked against both classical and advanced prediction methods. In Section 6 , future scenarios are studied. Section 7 concludes the paper.

2 Historical offer estimation

2.1 market basics.

In the NEMS, the following data are available either publicly or through subscription ( Energy Market Company, 2023 ; Energy Market Authority, 2023 ): historical USEP and load for each area, registered information of generating units, overhaul schedule (i.e., generator’s availability) and its marginal price (i.e., offer price cleared historically), daily market share of GenCos, and load forecasts. The 230 kV and 400 kV transmission network lines of the Singapore power system are available in Limited (2016) , Joseph et al. (2021) , and Global Energy Network Institute, 2023 .

Figure 2 shows forms of offers and pricing throughout the market-clearing process, as well as their data availability in the NEMS. Like many modern electricity markets, the NEMS uses nodal pricing. This means that, subject to the physical properties and constraints, electricity price varies across transmission nodes. Unique to the NEMS, while generators are paid the nodal price (local marginal price or LMP), consumers are charged a uniform price to prevent locational disadvantages. This is referred to as the Uniform Singapore Energy Price or USEP .

FIGURE 2 . Offers and pricing throughout the market-clearing process and data availability in the NEMS.

Definition (Uniform Singapore Energy Price): USEP is the weighted average of LMP over all the nodes that withdraw energy:

where p i , t L and q i , t D are LMP of node i and demand at period t , respectively. N B is the total number of buses in the system.

2.2 Approximating historically cleared offers

An offer in the context of electricity markets consists of a price and its corresponding quantity. As illustrated in Figure 2 , in the NEMS, only cleared offer price and historical USEP are available. As outlined in Figure 1 , our initial step is to approximate the offers that have been cleared historically. To achieve this, we formulate an optimization problem. The objective of this problem is to approximate the historical quantity (as shown in the second block of Figure 2 ) in such a way that the resulting USEP estimation closely matches the recorded values (2). In this paper, the terms “offer” and “price–quantity pair” are used interchangeably.

where p t U and p ^ t U are the actual and estimated USEP at the t th period, respectively.

In the setup of this problem, q i , t D and N B in (1) are known, and the LMP p i , t L is determined by the marginal price and quantity of the generators. Therefore, (1) becomes

In this paper, p represents the price ($/MWh), while q is the quantity (MW). The LMP p i , t L at bus i is defined as the price to serve the next MW of load at that location as expressed in (4):

where C . is the total production cost of all generators. The estimated LMP p ^ i , t L also depends on the generators’ accepted offers:

where p t M and q ^ t M denote vectors of accepted prices and estimated quantities cleared of all generators at t as expressed in (6) and (7), respectively:

where p g , j , t M and q ^ g , j , t M represent the price and estimated quantity cleared historically, respectively, for the j th generator from the g th GenCo at period t . N G represents the total number of GenCos in the system, and N g is the number of generators in the g th GenCo.

In addition to the typical constraints for DC power flow calculations, such as power limits and balance, and branch limits, the function C . is also subjected to an additional constraint—GenCos’ market shares. In the NEMS, the historical data of each GenCo’s daily market share are known, as represented by (8) :

where q g is the total daily production of GenCo g . It is worth noting that only the output of GenCos, rather than individual generators, is given. A GenCo typically owns multiple generators, hence the summation in (8) . Furthermore, if generators under the same GenCo have the same type and capacity, and their cleared prices are the same, we assume that they share the same quantity, as expressed in (9) :

As can be seen from the aforementioned formulation, the optimization aims to estimate historically accepted offer quantity q ^ t M so that the resultant USEP p t M for each period closely matches the actual USEP, while observing all transmission limits and additional market constraints. Given the typically large number of generators and their capacities, there could be many solutions to q ^ t M . (8) and (9), along with network constraints, help narrow down the solution space. If a solution cannot be found, (9) is relaxed.

2.3 Solution

The solution to the aforementioned problem is illustrated in Figure 3 . In the present study, we have chosen to employ the self-adaptive differential evolution (SaDE) algorithm, as detailed in Qin et al. (2009) , given its demonstrated efficiency in optimization tasks. One of the primary advantages of SaDE over traditional evolutionary algorithms (EAs) is its intrinsic ability to converge more rapidly. This accelerated convergence is achieved through SaDE’s unique capability to adaptively select learning strategies and fine-tune control parameters throughout the evolutionary process ( Qin et al., 2009 ; Qin and Suganthan, 2005 ). Unlike many other algorithms that require manual parameter tuning, SaDE’s self-adaptive mechanisms render it more user-friendly and often more robust across a variety of optimization challenges. By dynamically balancing exploration and exploitation phases, SaDE reduces the risk of premature convergence, often a challenge in traditional EAs. Furthermore, the algorithm’s adaptability means that it is less sensitive to initial parameter settings, thereby offering a level of flexibility seldom found in conventional differential evolution approaches. This dynamic and adaptive nature of SaDE ensures a holistic and efficient optimization process, making it a compelling choice for our research.

FIGURE 3 . Flowchart for calculating cleared offers.

For corroboration, basic EA options, such as the genetic algorithm (GA), are also employed. The power flow can be solved using MATPOWER ( Zimmerman et al., 2011 ). In this problem, the offer quantity is the variable, while the price is given. Therefore, the cost/offer functions for generators are constant functions within each iteration.

3 Conditional probabilistic offer functions

Once the estimated offer quantity q ^ t M for each period is determined, we form the historical data pool of offers. Assuming that a generator would not significantly change its offer strategy, each data point ( p g , j , t M , q ^ g , j , t M ) represents a possible offer the generator might make in the future. However, those data points cannot be directly used in market-clearing or OPF calculation, where offer functions are required.

Figure 4 plots the historical price–quantity pairs of a typical generator (the 8th) in the NEMS. It can be observed that most of the time, the cleared offer price is under 300 Singapore dollars per MWh. However, there is also a possibility that the price may exceed 1,000 $/MWh. In this paper, “$” denotes the Singapore dollar.

FIGURE 4 . Historical price–quantity pairs of a generator in the NEMS (from 1 January 2019 to 30 September 2019).

Second, generators choose their offer submissions based on various market conditions. As a result, these multiple choices of offers need to be modeled. In this part of our model, we form offer clusters and regress offer functions. Each cluster/function is associated with its respective probability. The framework for this step is summarized in Figure 5 , with the details provided in the following subsections.

FIGURE 5 . Flowchart for obtaining conditional probabilistic offer curves.

3.1 Clustering

To reflect the decision-making process of GenCos, the offers obtained in the first step ( Section 2 ) are classified into clusters. Each cluster represents a group of price–quantity pairs from which a generator might choose its offer. For instance, as depicted in Figure 4 , the offers made by generator 8 can be clustered into four groups.

There are many clustering techniques available. In this paper, we use k -means clustering ( Arthur and Vassilvitskii, 2006 ). The probability of selecting an offer from cluster k is proportional to the number of data points in that cluster, as shown in (10) .

where Pr k is the probability of cluster k and N k is the number of data points in the cluster.

After clusters are formed, an offer function is used as a representation of each cluster.

3.2 Conditional offer-making

Variables that influence electricity prices are listed in Chai et al. (2019) . Unlike the Nordic system ( Chai et al., 2019 ), the Singapore power system primarily relies on domestic generation, with no presence of hydropower and nuclear power. Among the accessible data in the NEMS, conditions affecting generators’ offer-making include demand, generation, time of the day, and day of the week. Under these conditions, (10) becomes

where Pr k . is the probability for an offer to be chosen from the k th cluster, Q G is the total capacity of generation, Q D is the load demand, t is the period of the day, and N k . is the number of data points in cluster k under conditions. With this conditional filter applied, the data points are refined and better represent offers made/cleared under conditions. Figure 6 shows offer data filtered with conditions.

FIGURE 6 . Example of offer data and fitted offer functions, under conditions of capacity margin, time of the day, and day of the week. [Offers made between (A) 13:00 and 13:30 and (B) 6:30 and 7:00 of the typical working day from 1 January 2019 to 30 September 2019.].

Within each filtered cluster of the dataset, conditional offer functions can be obtained using either linear or polynomial regression. In this paper, linear regression is used, as illustrated in Figure 6 . For this combination of conditions ( Figure 6A ), the probabilities associated with the four offer functions (from bottom to top) are 94.51%, 4.38%, 1.09%, and 0.36%, respectively. In addition, the set of offer curves varies with the conditions. For example, the offers made by the same generator during an early period of the day could only exist in the lowest cluster, as shown in Figure 6B .

When applying conditions/filters to the offer data, we consider the following:

1) While results could benefit from an increased number of clusters and effective conditions in (11) , data points with matching conditions might become sparse as each additional condition may filter out some data points. In this regard, more data are required to ensure that the offer functions obtained from the regression are good representatives of respective clusters. This is especially true for high-price-point clusters since their occurrence is far less frequent.

2) Conditions used are subjected to data availability. Both historical and projected conditions are required. Datasets with matching conditions are selected to perform prediction. In the NEMS, both scheduled generator status and load forecast are available. They are the input of the model ( Figure 7 ). Although price variations, especially price spikes, are more likely to be related to emergencies, the true causes and relevant records are not publicly available in the NEMS. Neither do forecasts on emergencies exist.

3) For conditions such as the total generation capacity Q G , historical records that exactly match a given value could be limited. In that case, offer points corresponding to adjacent generation capacities are used to facilitate regression. The k -nearest neighbors technique fulfills this task ( Altman, 1992 ; Friedman et al., 1977 ).

FIGURE 7 . Flowchart of probabilistic market-clearing for comprehensive market prediction.

4 Probabilistic market-clearing

The concept of probabilistic market-clearing is integral to our approach. Instead of relying on deterministic values, we employ a probabilistic method that accounts for uncertainties inherent in the electricity market. The crux of our probabilistic market-clearing method lies in the utilization of Monte Carlo simulation. This simulation leverages the conditional offer functions and their respective probabilities, which we previously derived. The process involves multiple iterations, each time selecting offer functions based on their conditional probabilities and subjecting them to random sampling.

For each iteration, the selected offer functions play a pivotal role in market-clearing. Here, we calculate the LMP and USEP using Eq. 3 . This iterative method ensures that the resulting solution is stable and reflective of market dynamics. To ensure convergence and computational efficiency, we have set specific criteria: the simulation halts when the change in variance across 1,000 consecutive iterations is less than 0.1%. An upper limit of 1,000,000 iterations is set to prevent excessive computation.

The primary objective of market-clearing in the NEMS, in the absence of demand-side bidding, is to minimize the total cost, represented as C p t M , q ^ t M . For the detailed computation of power flow and network losses, we employ MATPOWER ( Zimmerman et al., 2011 ), a widely acknowledged tool in the domain. For an in-depth understanding of the market-clearing process, especially its formulation in the NEMS, readers are directed to refer to Zhou et al. (2016) and Gao et al. (2017) .

5 Numerical study

Data from the NEMS spanning from 1 January 2019 to 30 September 2019 are used for verification. During this period, there were 46 registered generators in the NEMS, with their capacities ranging from 4.8 to 431 MW. The total capacity was 9.51 GW, and the peak load was 7.24 GW. There were 48 trading/dispatch periods in a day. The proposed methodology is used for day-ahead market prediction. A 3-week dataset is selected for testing, while the remaining datasets are used for modeling.

5.1 Historical offer estimation

The proposed methodology for historical offer estimation ( Section 2 ) is implemented in the MATLAB environment. Offer quantities q ^ g , j , t M for 12,096 periods (269 days) are estimated. Given the computationally intensive calculation involved, the high-performance computing platform of Nanyang Technological University ( Nanyang Technological University, 2021 ) is used. A total of 400 CPU (AMD EPYC™ 7702 2.0 GHz) cores/workers are deployed in parallel. Using the SaDE algorithm ( Qin et al., 2009 ), it takes approximately 40 h to converge. It should be noted that this step of offer approximation is carried out offline, as the estimated offers serve as the basis for the following steps and would not change once calculated.

Figure 8 shows the histogram of the values of the objective function (1). The average of the USEPs for the 9-month period is 98.28 $/MWh. The mean error between the calculated and actual USEPs ((1)) is 0.084 $/MW (i.e., 0.084% of the average USEP); 99.52% of the cases result in errors less than 1%.

FIGURE 8 . Performance of the offer estimation—histogram of USEP errors.

5.2 Comprehensive market prediction results

With the estimated historical price–quantity pairs, we can determine offer functions with conditional probabilities ( Section 3 ), followed by USEP forecasting using probabilistic market-clearing ( Section 4 ). The number of clusters k is set to 7. The prediction results for a typical day and a day with price spikes are shown in Figures 9A, B , respectively.

FIGURE 9 . USEP prediction results vs. observations: (A) 102nd day (12 April) and (B) 76th day (17 March) in 2019.

For a typical day in the NEMS, the USEP is approximately 100 $/MWh in 2019. However, there are possibilities to witness price spikes, especially during the periods from 10:30 to 12:00 and 13:00 to 14:00. The distribution of the prediction is more converged during the early hours. This is because, historically, the prices during early hours are most likely to settle approximately 100 $/MWh.

Generally speaking, the prediction errors mainly stem from errors in 1) the offer estimation ( Section 2 ) and 2) the clustering, filtering, and linear regression ( Section 3 ). The benchmark for USEP prediction and its comparison with other methods are discussed in the following subsection.

The prediction results for daily market share are shown in Figure 10 . The predicted results for typical normal days are close to the mean values ( Figure 10A ), whereas the prediction is more scattered for a day with price spikes ( Figure 10B ). This is because generators’ offers are more likely to settle in lower price tiers on a normal day, leading to more converged results, as shown in Figure 10A . This is consistent with the observations in price prediction, as shown in Figure 9 . It is important to remember that in the NEMS, only daily market share is available (according to (8)). The calculated market shares for each period and each generator ( Section 2 ) could differ from the actual market shares.

FIGURE 10 . Prediction results for daily market share vs. its observation: (A) 102nd day (12 April) and (B) 76th day (17 March) in 2019.

5.3 Performance benchmarking

The performance of USEP prediction is evaluated against both classical and advanced prediction models. These include a basic empirical unconditional density forecast model (EU), a classical time-series model—generalized autoregressive conditional heteroskedastic model (GARCH) with the Gaussian error distribution ( Jónsson et al., 2014 ), and three AI models—an extreme learning machine logistic continuous ranked probability score-based ensemble model output statistics (ELC-EMOS) ( Chai et al., 2019 ), a regularized quantile regression averaging method which utilizes the least absolute shrinkage and selection operator (LQRA) ( Uniejewski and Weron, 2021 ), and a combined model (Combine) ( Li et al., 2020 ).

The continuous ranked probability score (CRPS) is used for benchmarking. The CRPS measures the closeness of forecast distribution to the corresponding observation and is one of the popular indices for measuring the performance of probabilistic prediction ( Chai et al., 2019 ; Li et al., 2020 ; Afrasiabi et al., 2020 ; Matheson and Winkler, 1976 ; Hersbach, 2000 ; Gneiting et al., 2007 ; Gneiting and Raftery, 2007 ; Zhang et al., 2020 ). The same NEMS data are used for training and testing, respectively. However, network constraints are not considered in the benchmark models. For quantile forecast, quantiles 0.01 to 0.99 are estimated in steps of 0.01.

Table 1 presents a comparative analysis of daily average CRPS for selected testing cases, bifurcated into normal days and days with price spikes. Among the selected 21 testing cases, two groups—normal days and days with spikes—are formed and benchmarked. Among them, 11 are normal days in which the daily average USEP varies from 87.30 to 107.61 $/MWh, and 10 days with price spikes in which the daily average varies from 112.01 to 520.58 $/MWh. The smaller the CRPS score, the better the performance.

TABLE 1 . Daily average CRPS (S$/MWh) benchmarks.

5.3.1 Classical methods

Among the classical methods, the EU stands out for its simplicity, relying on a constant empirical distribution. Information such as Q G , Q D , and t in Section 3.2 is not applied in the EU. However, its performance lags behind, as evidenced by the highest CRPS scores. The GARCH model shows improvement, attributing its better performance to its heteroscedastic modeling ( Jónsson et al., 2014 ).

5.3.2 AI-based methods

Turning to AI-based methods, they collectively outperform their classical counterparts. The combined method slightly edges out others in both normal days and days with spikes. This indicates the potential of AI in capturing intricate patterns and nuances in the data.

5.3.3 Proposed method

Our proposed method outperforms, especially during days with price spikes. While it competes closely with AI-based methods on normal days, it takes a clear lead during volatile days. When comparing the proposed method, for normal days, our proposed method is on par with AI-based methods, while AI methods have slight edges. The average CRPSs for ELC-EMOS, LQRA, Combine, and the proposed method are 7.89, 8.15, 7.51, and 8.16 $/MWh, respectively. As discussed in Section 3.1 and Section 3.2 , increasing the number of clusters can be one solution to improve the results. The offer information is distilled and represented by a limited number of linear functions. For a given number of explanatory information (i.e., conditional filter Q G , Q D , and t ), an increased cluster number means less loss of information due to regression. However, more historical data are also required accordingly.

For days with spikes, on the other hand, the proposed method takes a clear lead in terms of CRPS. The average CRPSs for ELC-EMOS, LQRA, Combine, and the proposed method are 34.80, 33.95, 32.63, and 30.22 $/MWh, respectively. This can be mainly attributed to the following reasons: first, USEP in Singapore context is a form of local marginal pricing that considers the grid’s network property, which conventional AI-based prediction models do not take as the input. Compared with normal days in which offers are cleared at similar prices, the network’s effect on LMP becomes pronounced when more generators’ offers are cleared at high price points. As a result, the LMP on the bus increases. As the weighted average of LMPs (1), USEP also increases.

The advantage of our method can be ascribed to two key factors:

Network influence on LMP: The inherent network properties of the Singapore grid play a pivotal role in shaping USEP, especially during high-clearance offers. Traditional AI models ( Wan et al., 2017 ; Chai et al., 2019 ; Wan et al., 2014 ) often overlook this constraint. Our method, however, integrates these network effects, leading to enhanced prediction accuracy.

Data constraints on AI models: While AI-based methods typically thrive on large datasets, our study was constrained to 9 months of data. This limitation could pose challenges in tuning AI models, potentially leading to suboptimal performance.

The results underscore the importance of considering network properties and constraints in electricity market predictions. While AI offers promise, its efficacy is often contingent upon data availability. Our proposed method, which seamlessly blends these considerations, emerges as a robust solution, especially during market volatilities.

6 Scenario study

6.1 impact of demand-side participation.

The proposed method also has the advantage of being able to study future scenarios with varied system parameters. For instance, the NEMS is actively considering consumers’ participation in the price discovery process, such as demand-side bidding ( Energy Market Authority, 2013 ). To study the impact, fixed-rate pricing schemes for demand-side bidding are assumed. Apart from the pricing, another influencing factor is the penetration level of the dispatchable load, defined here as the proportion in relation to the annual peak load. In this case, 1% of the penetration level is equivalent to 72.37 MW. The dispatchable load is assumed to be distributed across the whole network in proportion to demand on each node. In our model, it is treated as generators with negative output. A price floor of 300 $/MWh for demand-side bidding in the NEMS is designed to address the potential gaming issue ( Energy Market Authority, 2013 ).

By varying the price and the penetration level of the dispatchable load in the system, the changes in the average USEP can be obtained, as shown in Figure 11 . It can be seen from the figure that USEP drops with the demand-side’s participation. For a given amount of dispatchable load, the reduction in USEP decreases with the increase in the bidding price. For instance, with 1% penetration and 300 $/MWh bidding price, the USEP reduction is 1.30 $/MWh (1.32% of the average USEP). Increasing the price up to 500 $/MWh, the USEP reduction remains similar. The impact of demand-side bidding dwindles when its price is between 600 and 1,000 $/MWh. Beyond 1,000 $/MWh, its effect on USEP reduction diminishes. The higher the bidding price, the smaller the chance it can be cleared. This observation becomes more noticeable with higher penetration levels.

FIGURE 11 . Impact of dispatchable load on the USEP, with fixed-rate pricing.

The second observation concerns the penetration level. Given the bidding prices, the USEP continuously reduces as the capacity of dispatchable load increases, but not always in a linear way. When the price is low, for example, 300 $/MWh, the USEP reduction is almost in a linear relationship with the penetration level. As the price increases, especially approximately from 600 to 900 $/MWh, the USEP reduction does not grow in proportion with the penetration level. The explanation is that at lower prices, bids from the demand-side compare favorably with generators. However, at higher price levels, offers from traditional generators start to compete despite increased participation of flexible load. From the aforementioned results, the flexible load can hardly be cleared when the bidding price is higher than 1,000 $/MWh.

6.2 Impact of rooftop PV penetration

Singapore’s access to renewable energy sources is limited, with solar photovoltaics (PVs) being one of the few options. Given the country’s land constraints, rooftop PV installations have become a popular solution. We obtained power measurement data from a typical commercial building equipped with solar panels. This building consumes approximately 60 MWh of energy daily, and its rooftop PV system has a capacity of 1 MWp. Commercial loads account for 35% of the Singapore’s total demand, with the majority located in the central area.

We studied the impact of rooftop PV systems on USEP by assuming that similar PV systems are installed on other commercial buildings. The penetration level of rooftop PV, defined as the installed capacity, ranges from 0 to 1 MWp per 60 MWh commercial load. Offer strategies from GenCos are assumed to remain the same. Table 2 shows the average daily USEP reduction under different rooftop PV penetration levels.

TABLE 2 . Impact on USEP of rooftop PV penetration on commercial buildings in Singapore.

The table reveals that an increase in PV capacity indeed lowers the electricity price by reducing demand. The price drop is approximately proportional to the PV penetration level. With the highest coverage (1.0 MWp per 60 MWh of demand), the PV system can support up to 2.46% of the total daily demand on a sunny day (solar radiation intensity ranging from 800 to 1000 W/m 2 around solar noon, a typical intensity for clear days in Singapore’s equatorial climate ( The National Environment Agency, 2023 )). In our experiments, this translates into a modest reduction (0.93%) in the daily USEP.

The limited impact can be attributed to three contributing factors: 1). PV can only operate during daylight hours (approximately 10 h), leaving the USEP unchanged for the remaining periods. Additionally, sunlight availability is not always guaranteed; 2). the penetration of PV is still relatively limited when installed solely on commercial buildings’ rooftops; and 3). the USEP is the weighted sum of LMP across all nodes, which means the overall impact is diluted on average.

7 Conclusion

This paper introduces an innovative analytical method tailored for electricity market prediction, especially in scenarios with incomplete market offer data. Our research underscores the following key findings and contributions:

1. Comprehensive predictions : Beyond price forecasts, our approach furnishes crucial insights into GenCos’ market shares and load flows, adding depth to market predictions.

2. Performance benchmarking : Using real NEMS data, our method displayed superior performance, particularly during price spike events.

3. Scenario analyses : Our studies highlighted the potential of demand-side bidding to mitigate the USEP, with effects varying based on bidding price and dispatchable load capacity. Moreover, rooftop PV implementations on commercial infrastructures were found to exert a modest downward effect on prices due to solar energy limitations and installation capacities.

4. Methodological advantages : The approach stands out for its adaptability to other deregulated electricity markets, even without complete historical offer data. Its provision of added predictive parameters aids market stakeholders in gauging potential market outcomes. Furthermore, its physics-based nature paves the way for diverse scenario studies.

In essence, this work offers a robust and adaptable toolkit for electricity market prediction, promising expansive applications and avenues for future exploration.

8 Future work

While our methodology provides a comprehensive approach to predicting GenCo’s offer-making decisions, it is worth noting that we have based our predictions on a set of factors including demand, generation, time of the day, and day of the week. In real-world scenarios, GenCos often base their decisions on a myriad of factors, especially in situations leading to price hikes. For instance, external factors, such as weather patterns, can significantly influence energy demand, especially with the increasing penetration of renewable energy sources. Similarly, broader economic indicators can hint at potential changes in energy consumption patterns.

To enhance the predictive accuracy of our model, future research endeavors should delve deeper into integrating these additional factors. Incorporating weather data, for example, can provide insights into potential changes in demand due to temperature variations. Similarly, studying economic indicators can give a clearer picture of how broader economic trends might influence energy consumption and, consequently, GenCo offers. While our study has been tailored to the unique intricacies of the Singaporean electricity market, it would be invaluable to apply this methodology to other electricity markets. This would not only test the model’s universal applicability but also highlight potential adjustments required to cater to different market dynamics and structures.

Data availability statement

The raw data supporting the conclusion of this article will be made available by the authors, without undue reservation.

Author contributions

NX: writing–original draft. XG: data curation, formal analysis, and writing–review and editing. SC: writing–review and editing. MN: writing–review and editing.

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work is sponsored by the Scientific Research Startup Fund for Shenzhen High-Caliber Personnel of SZPT (No. 6022310042k).

Conflict of interest

Author MN was employed by State Grid Liaoning Electric Power Company Ltd. Author JY was employed by PacificLight Power Pte Ltd.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors, and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Abedinia, O., Amjady, N., and Zareipour, H. (2017). A New feature selection technique for load and price forecast of electrical power systems. IEEE Trans. Power Syst. 32 (1), 62–74. doi:10.1109/tpwrs.2016.2556620

CrossRef Full Text | Google Scholar

Afrasiabi, M., Mohammadi, M., Rastegar, M., Stankovic, L., Afrasiabi, S., and Khazaei, M. (2020). Deep-based conditional probability density function forecasting of residential loads. IEEE Trans. Smart Grid 11 (4), 3646–3657. doi:10.1109/tsg.2020.2972513

Altman, N. S. (1992). An introduction to kernel and nearest-neighbor nonparametric regression. Am. Stat. 46 (3), 175–185. doi:10.2307/2685209

Arthur, D., and Vassilvitskii, S. (2006). k-means++: the advantages of careful seeding . Stanford .

Google Scholar

Bo, R., and Li, F. (2009). Probabilistic LMP forecasting considering load uncertainty. IEEE Trans. Power Syst. 24 (3), 1279–1289. doi:10.1109/tpwrs.2009.2023268

Brusaferri, A., Matteucci, M., Portolani, P., and Vitali, A. (2019). Bayesian deep learning based method for probabilistic forecast of day-ahead electricity prices. Appl. Energy 250, 1158–1175. doi:10.1016/j.apenergy.2019.05.068

Bunn, D. W., Inekwe, J. N., and Macgeehan, D. (2021). Analysis of the fundamental predictability of prices in the British balancing market. IEEE Trans. Power Syst. 36 (2), 1309–1316. doi:10.1109/tpwrs.2020.3015871

Chai, S., Xu, Z., and Jia, Y. (2019). Conditional density forecast of electricity price based on ensemble ELM and logistic EMOS. IEEE Trans. Smart Grid 10 (3), 3031–3043. doi:10.1109/tsg.2018.2817284

Durvasulu, V., and Hansen, T. M. (2018). Market-based generator cost functions for power system test cases. IET Cyber-Phys. Syst. Theory Appl. 3 (4), 194–205. doi:10.1049/iet-cps.2018.5046

Energy Market Authority (2023). Energy market authority. [Online]. Available at: https://www.ema.gov.sg/ .

Energy Market Authority (2013). Implementing demand response in the national electricity market of Singapore - final determination paper .

Energy Market Authority of Singapore (2018). Intermittency pricing mechanism for intermittent generation sources in the national electricity market of Singapore: final determination paper .

Energy Market Company (2023). Energy market Company . [Online]. Available at: https://www.emcsg.com/ .

Fraunholz, C., Kraft, E., Keles, D., and Fichtner, W. (2021). Advanced price forecasting in agent-based electricity market simulation. Appl. Energy 290, 116688. doi:10.1016/j.apenergy.2021.116688

Friedman, J. H., Bentley, J. L., and Finkel, R. A. (1977). An algorithm for finding best matches in logarithmic expected time. ACM Trans. Math. Softw. 3 (3), 209–226. doi:10.1145/355744.355745

Gao, Y., Zhou, S., Shu, Z., and Beng, G. H. (2017). “Market clearing model for Singapore electricity market incorporating transmission loss,” in 2016 Asian Conference on Energy, Power and Transportation Electrification (ACEPT 2016) , Singapore , 25-27 October 2016 ( IEEE ).

Global Energy Network Institute (2023). Global energy network Institute. [Online]. Available at: http://www.geni.org/ .

Gneiting, T., Balabdaoui, F., and Raftery, A. E. (2007). Probabilistic forecasts, calibration and sharpness. J. R. Stat. Soc. B Stat. Methodol. 69 (2), 243–268. doi:10.1111/j.1467-9868.2007.00587.x

Gneiting, T., and Raftery, A. E. (2007). Strictly proper scoring rules, prediction, and estimation. J. Am. Stat. Assoc. 102 (477), 359–378. doi:10.1198/016214506000001437

He, H., Lu, N., Jiang, Y., Chen, B., and Jiao, R. (2020). End-to-end probabilistic forecasting of electricity price via convolutional neural network and label distribution learning. Energy Rep. 6, 1176–1183. doi:10.1016/j.egyr.2020.11.057

Heidarpanah, M., Hooshyaripor, F., and Fazeli, M. (2023). Daily electricity price forecasting using artificial intelligence models in the Iranian electricity market. Energy 263, 126011. doi:10.1016/j.energy.2022.126011

Hersbach, H. (2000). Decomposition of the continuous ranked probability score for ensemble prediction systems. Weather Forecast 15 (5), 559–570. doi:10.1175/1520-0434(2000)015<0559:dotcrp>2.0.co;2

Hong, Y. Y., Taylar, J. v., and Fajardo, A. C. (2020). Locational marginal price forecasting using deep learning network optimized by mapping-based genetic algorithm. IEEE Access 8, 91975–91988. doi:10.1109/access.2020.2994444

Ji, Y., Thomas, R. J., and Tong, L. (2017). Probabilistic forecasting of real-time LMP and network congestion. IEEE Trans. Power Syst. 32 (2), 831–841. doi:10.1109/tpwrs.2016.2592380

Jónsson, T., Pinson, P., Madsen, H., and Nielsen, H. A. (2014). Predictive densities for day-ahead electricity prices using time-adaptive quantile regression. Energies (Basel) 7 (9), 5523–5547. doi:10.3390/en7095523

Joseph, A., Gao, Z., Kong, X., Xu, N. Z., and Ngin, H. T. (2021). Dynamic modelling and simulation of Singapore transmission network (230/400kV) . Technology Disclosure, (NTU Ref: 2021-305).

Li, T., Wang, Y., and Zhang, N. (2020). Combining probability density forecasts for power electrical loads. IEEE Trans. Smart Grid 11 (2), 1679–1690. doi:10.1109/tsg.2019.2942024

Limited, S. P. P. A. (2016). Offering circular . Singapore .

Matheson, J. E., and Winkler, R. L. (1976). Scoring rules for continuous probability distributions. Manage Sci. 22 (10), 1087–1096. doi:10.1287/mnsc.22.10.1087

Meng, A., Wang, P., Zhai, G., Zeng, C., Chen, S., Yang, X., et al. (2022). Electricity price forecasting with high penetration of renewable energy using attention-based LSTM network trained by crisscross optimization. Energy 254, 124212. doi:10.1016/j.energy.2022.124212

Monteiro, C., Ramirez-Rosado, I. J., and Fernandez-Jimenez, L. A. (2018). Probabilistic electricity price forecasting models by aggregation of competitive predictors. Energies (Basel) 11 (5), 1074. doi:10.3390/en11051074

Nanyang Technological University (2021). High performance computing centre (HPCC) . Singapore: Nanyang Technological University . [Online]. Available at: https://ntuhpc.org/ .

Nowotarski, J., and Weron, R. (2018). Recent advances in electricity price forecasting: a review of probabilistic forecasting. Renew. Sustain. Energy Rev. 81 (2016), 1548–1568. doi:10.1016/j.rser.2017.05.234

Qin, A. K., Huang, V. L., and Suganthan, P. N. (2009). Differential evolution algorithm with strategy adaptation for global numerical optimization. IEEE Trans. Evol. Comput. 13 (2), 398–417. doi:10.1109/tevc.2008.927706

Qin, A. K., and Suganthan, P. N. (2005). “Self-adaptive differential evolution algorithm for numerical optimization,” in 2005 IEEE Congress on Evolutionary Computation Vol. 2 , Edinburgh, UK , 2-4 September 2005 , 1785–1791.

Services, S. P. (2021). Open electricity market. [Online]. Available at: https://www.openelectricitymarket.sg .

Taylor, J. W. (2021). Evaluating quantile-bounded and expectile-bounded interval forecasts. Int. J. Forecast. 37 (2), 800–811. doi:10.1016/j.ijforecast.2020.09.007

The National Environment Agency (2023). Meteorological service Singapore. [Online]. Available: http://www.weather.gov.sg .

Uniejewski, B., and Weron, R. (2021). Regularized quantile regression averaging for probabilistic electricity price forecasting. Energy Econ. 95, 105121. doi:10.1016/j.eneco.2021.105121

Wan, C., Niu, M., Song, Y., and Xu, Z. (2017). Pareto optimal prediction intervals of electricity price. IEEE Trans. Power Syst. 32 (1), 817–819. doi:10.1109/tpwrs.2016.2550867

Wan, C., Xu, Z., Wang, Y., Dong, Z. Y., and Wong, K. P. (2014). A hybrid approach for probabilistic forecasting of electricity price. IEEE Trans. Power Syst. 5 (1), 463–470. doi:10.1109/tsg.2013.2274465

Yang, J., Zhao, J., Wen, F., and Dong, Z. (2019). A model of customizing electricity retail prices based on load profile clustering analysis. IEEE Trans. Smart Grid 10 (3), 3374–3386. doi:10.1109/tsg.2018.2825335

Zhang, H., Liu, Y., Yan, J., Han, S., Li, L., and Long, Q. (2020). Improved deep mixture density network for regional wind power probabilistic forecasting. IEEE Trans. Power Syst. 35 (4), 2549–2560. doi:10.1109/tpwrs.2020.2971607

Zhao, J. H., Dong, Z. Y., Xu, Z., and Wong, K. P. (2008). A statistical approach for interval forecasting of the electricity price. IEEE Trans. Power Syst. 23 (2), 267–276. doi:10.1109/tpwrs.2008.919309

Zhou, F., Huang, Z., and Zhang, C. (2022). Carbon price forecasting based on CEEMDAN and LSTM. Appl. Energy 311, 118601. doi:10.1016/j.apenergy.2022.118601

Zhou, Q., Tesfatsion, L., and Liu, C. C. (2011). Short-term congestion forecasting in wholesale power markets. IEEE Trans. Power Syst. 26 (4), 2185–2196. doi:10.1109/tpwrs.2011.2123118

Zhou, S., Shu, Z., Tan, K., Gooi, H. B., Chen, S., and Gao, Y. (2016). “Study of market clearing model for Singapore’s wholesale real-time electricity market,” in 2016 IEEE International Conference on Power System Technology (POWERCON 2016) , Wollongong, Australia , September 2016 ( IEEE ).

Zimmerman, R. D., Murillo-Sánchez, C. E., and Thomas, R. J. (2011). MATPOWER: steady-state operations, planning, and analysis tools for power systems research and education. IEEE Trans. Power Syst. 26 (1), 12–19. doi:10.1109/tpwrs.2010.2051168

Nomenclature

Keywords: electricity market, probabilistic forecast, market clearing, market share, Monte Carlo simulation, National Electricity Market of Singapore, self-adaptive differential evolution, k -means clustering

Citation: Xu NZ, Gao X, Chai S, Niu M and Yang JX (2023) Leveraging sanitized data for probabilistic electricity market prediction: a Singapore case study. Front. Energy Res. 11:1296957. doi: 10.3389/fenrg.2023.1296957

Received: 19 September 2023; Accepted: 30 October 2023; Published: 15 November 2023.

Reviewed by:

Copyright © 2023 Xu, Gao, Chai, Niu and Yang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Xiang Gao, [email protected]

This article is part of the Research Topic

Advanced Technologies for Planning and Operation of Prosumer Energy Systems, volume III

An official website of the United States government

Here's how you know

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Real-Life Benefits of Exercise and Physical Activity

On this page:

Why is physical activity important?

Emotional benefits of exercise.

Exercise and physical activity are good for just about everyone, including older adults. No matter your health and physical abilities, you can gain a lot by staying active. In fact, studies show that “taking it easy” is risky. Often, inactivity is more to blame than age when older people lose the ability to do things on their own. Lack of physical activity also can lead to more visits to the doctor, more hospitalizations, and more use of medicines for a variety of illnesses.

Including all 4 types of exercise can benefit a wide range of areas of your life. Staying active can help you:

• Keep and improve your strength so you can stay independent
• Have more energy to do the things you want to do and reduce fatigue
• Improve your balance and lower risk of falls and injuries from falls
• Manage and prevent some diseases like arthritis, heart disease, stroke, type 2 diabetes, osteoporosis, and 8 types of cancer, including breast and colon cancer
• Sleep better at home
• Reduce levels of stress and anxiety
• Reach or maintain a healthy weight and reduce risk of excessive weight gain
• Control your blood pressure
• Possibly improve or maintain some aspects of cognitive function , such as your ability to shift quickly between tasks or plan an activity
• Perk up your mood and reduce feelings of depression

Research has shown that exercise is not only good for your physical health, it also supports emotional and mental health. You can exercise with a friend and get the added benefit of emotional support. So, next time you’re feeling down, anxious, or stressed, try to get up and start moving!

Physical activity can help:

• Reduce feelings of depression and stress, while improving your mood and overall emotional well-being
• Increase your energy level
• Improve sleep
• Empower you to feel more in control

In addition, exercise and physical activity may possibly improve or maintain some aspects of cognitive function , such as your ability to shift quickly between tasks, plan an activity, and ignore irrelevant information.

Here are some exercise ideas to help you lift your mood:

• Walking, bicycling, or dancing. Endurance activities increase your breathing, get your heart pumping, and boost chemicals in your body that may improve mood.
• Yoga. This mind and body practice typically combines physical postures, breathing exercises, and relaxation.
• Tai Chi. This "moving meditation" involves shifting the body slowly, gently, and precisely, while breathing deeply.
• Activities you enjoy. Whether it’s gardening, playing tennis, kicking around a soccer ball with your grandchildren, or something else, choose an activity you want to do, not one you have to do.

You may also be interested in

• Finding tips to help stay motivated to exercise
• Exploring safety tips for exercising outdoors
• Reading about the four types of exercise

Sign up for e-alerts about healthy aging

For more information.

YMCA 800-872-9622 [email protected] www.ymca.net

Centers for Disease Control and Prevention (CDC) 800-232-4636 888-232-6348 (TTY) [email protected] www.cdc.gov

MedlinePlus National Library of Medicine       www.medlineplus.gov

This content is provided by the NIH National Institute on Aging (NIA). NIA scientists and other experts review this content to ensure it is accurate and up to date.

Content reviewed: April 03, 2020

nia.nih.gov

An official website of the National Institutes of Health