Next Article in Journal
Use of CO2 in Pressurized, Fluidized Bed Gasification of Waste Biomasses
Next Article in Special Issue
Detection and Diagnosis of Dependent Faults That Trigger False Symptoms of Heating and Mechanical Ventilation Systems Using Combined Machine Learning and Rule-Based Techniques
Previous Article in Journal
Virtual In Situ Calibration for Operational Backup Virtual Sensors in Building Energy Systems
Previous Article in Special Issue
Natural Gas Consumption Forecasting Based on the Variability of External Meteorological Factors Using Machine Learning Algorithms
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Energies
Volume 15
Issue 4
10.3390/en15041393
energies-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Indirect Impact of the COVID-19 Pandemic on Natural Gas Consumption by Commercial Consumers in a Selected City in Poland

by
Tomasz Cieślik
1,2,*,
Piotr Narloch
2,3,
Adam Szurlej
2,* and
Krzysztof Kogut
4,*
1
Institute of Nuclear Physics PAN, Radzikowski St. 152, 31342 Kraków, Poland
2
Faculty of Drilling, Oil and Gas, AGH University of Science and Technology, Mickiewicz Ave. 30, 30059 Kraków, Poland
3
Polish Gas Company, Bandrowskiego St. 16, 33100 Tarnów, Poland
4
Faculty of Energy and Fuels, AGH University of Science and Technology, Mickiewicz Ave. 30, 30059 Kraków, Poland
*
Authors to whom correspondence should be addressed.
Energies 2022, 15(4), 1393; https://0-doi-org.brum.beds.ac.uk/10.3390/en15041393
Submission received: 29 December 2021 / Revised: 6 February 2022 / Accepted: 9 February 2022 / Published: 14 February 2022
(This article belongs to the Special Issue Energy and Artificial Intelligence)
Browse Figures
Versions Notes

Abstract

:
In March 2020, a lockdown was imposed due to a global pandemic, which contributed to changes in the structure of the consumption of natural gas. Consumption in the industry and the power sector decreased while household consumption increased. There was also a noticeable decrease in natural gas consumption by commercial consumers. Based on collected data, such as temperature, wind strength, duration of weather events, and information about weather conditions on preceding days, models for forecasting gas consumption by commercial consumers (hotels, restaurants, and businesses) were designed, and the best model for determining the impact of the lockdown on gas consumption by the above-mentioned consumers was determined using the MAPE (mean absolute percentage error). The best model of artificial neural networks (ANN) gave a 2.17% MAPE error. The study found a significant decrease in gas consumption by commercial customers during the first lockdown period.

1. Introduction

Natural gas is an important raw material in the sustainable energy policy due to its growing importance in the global economy [1,2]. The global economy has observed a continuous growth of natural gas consumption since the 1960s, with a growth rate of 3.5% over the last five years. In 2019, 3.92 trillion m3 of natural gas were consumed worldwide (Figure 1). Poland also shows an upward trend in the consumption of natural gas. The largest decrease was recorded in the period of political transition from 1987–1992 (Figure 2). Natural gas accounts for 36.2% of overall household energy consumption in the EU, ranking first. The share of natural gas consumption in Poland, however, is 18.4%, just behind coal consumption, which amounts to 32.3% [3]. Construction of new natural gas facilities is being considered. These would play a transitional role in the process of energy transformation towards the widest possible use of RES, which is referred to in the drafted EU energy policy and national energy development scenarios. The extent to which natural gas is used in the power industry will depend, among others, on the price of this fuel and on the prices of carbon dioxide emission allowances [4,5,6]. In the case of Europe, the risk of stagnation is observed. This is caused by a drop in consumption and restrictions regarding investments in gas-fuelled power plants and gas infrastructure development since infrastructure is not operating to its maximum capacity [7].
In light of the above information, natural gas consumption forecasts will have a major impact on public utilities and energy traders. Forecasting consumption will allow the determination of service charges for gas consumed in the future. Determining natural gas consumption in the near future enables planning decisions regarding the purchase of natural gas by national and local companies, its storage, scheduling network loads, and undertaking upgrading and investment tasks related to the day-to-day operation of the gas infrastructure by ensuring the security of the gas supply and its proper operation [8,9,10,11,12,13,14,15].

2. The COVID-19 Pandemic

A global pandemic broke out in early 2020, resulting in a reduction of production and consumption of energy resources worldwide. The disease, caused by a highly contagious and new type of coronavirus, produces a number of complications that are dangerous to human life and health. The most serious of these include cardiac [16], otolaryngological [17], dermatological [18], thrombotic [19], neurological [20], and nephrological [21] complications. The disease has a high perioperative mortality rate [22], can cause liver damage [23] and is particularly dangerous for diabetics [24]. It can also produce serious complications in patients with pneumonia [25].
To limit the spread of the epidemic, governments in many countries have decided to introduce states of emergency, which imposed many restrictions on movement, public life, and the operation of industrial plants and services. The first country was France, which introduced a state of emergency on 17 March, followed by the United States and Poland on 20 March, Germany on 22 March, and the United Kingdom on 23 March. It is estimated that from January to June 2020, the global economy reduced natural gas consumption by about 4% [26,27,28]. The restrictions triggered a crisis in the availability of consumer goods, which resulted in the loss of jobs in many areas of the economy, but they also demonstrated the importance of energy independence [29].
National restrictions and restrictions of movement (lockdown) impacted many areas of life. In effect, the lockdown has had a positive impact on air quality [30,31]. This is relevant in that pollution levels affect the progression and spread of the disease [32,33,34]. There has been a reduction in linear (road) emissions associated with reduced traffic volumes. A change also occurred in travel preferences. People chose private transport rather than public transport more often [35].
The lockdown had an impact on reducing electricity consumption by economies in countries where it was introduced [36,37,38,39]. The UK reduced electricity consumption by 6%, Italy by 11%, France by 15%, and India by 26% [40]. The United States saw more than a 10% decrease in electricity consumption from 26 March to 6 June [41]. The magnitude of the reduction in electricity consumption in the US varied by area. For the NYISO (New York Independent System Operation), it amounted to 2% in the peak hour and as much as 12% for the MISO (Midcontinent Independent System Operator) [42]. Exceptions include Switzerland, Sweden, and Norway, where no strict restrictions were introduced. Countries whose power industry is based on coal (Poland and the Czech Republic) have increased their energy imports while Italy halved its import of energy from abroad [43]. Countries that use renewable energy sources (Germany) increased their share by about 8% as compared to the same period in 2019, amounting to 55% of total consumption. The largest share was produced by wind turbines (19%), solar panels (17%), and biomass (10%) [44]. It must be pointed out that the present energy crisis, the depletion of fossil fuels, and climate change increase the interest in renewable energy sources; the studies indicate that the changes are initiated by developed countries [45].
The decline in electricity consumption intensified with the prolonged crisis caused by production cuts [46]. In China, a relationship was observed between the daily number of cases and the future drop in electricity demand, which is extremely important in planning activities during a pandemic [47].
On the other hand, household consumption of electricity and water increased, resulting in higher utility charges [48,49,50,51,52]. During the pandemic, governments in many countries took measures to protect individual consumers, including disconnection bans or cancellation of bills to help those in a difficult financial situation [40].
Along with the pandemic, there was a change in consumer habits, resulting in a shift of peak hours in utility consumption [53,54]. The hours of maximum hot water draw of residents in Seoul (South Korea) can be cited as an example. The draw was lower from 04:00 a.m. to 10:00 a.m. compared to the time before the pandemic, but there was a definite increase in water consumption from 11:00 a.m. to 8:00 p.m. A dependency was also observed between the average number of actively infected persons in a given month and changes in hot water consumption [55].
Consumption of crude oil in the United States reached its lowest level in many years, with the price at around 30 USD per barrel. The low price was dictated not only by the global crisis caused by the pandemic but also due to the lack of agreement between the OPEC countries and the US [56,57,58]. The low price of oil, with no possibility of selling it, led to an increased demand for oil storage in chartered tankers. This situation resulted in a spike in tanker rental prices [59].
Natural gas consumption in households increased while consumption in the industry during this period decreased. The consequences of changes in the consumption of energy resources in Europe depended on the region, on the dependency of the economy on natural gas, and on the severity of the pandemic during the period in question. In the European Union, there was a 7% decrease from January to May relative to the previous year. This decrease amounted to 11% from March to May and 16% from April to May. The most notable change in natural gas consumption occurred in the power sector in Italy, where, from 10 March to 25 March, there was a decrease in gas consumption from 40 to 25 million m3/day while in the industry, consumption dropped from 85 million (prior to 4 March 2020) to 35 million m3/day (25 March). A sharp and sustained decline in LNG imports to Europe was also noted in mid-March. China, which experienced a decline in LNG imports in January, saw imports increase from February to June 2020 [60,61,62,63]. Forecasts indicate that if the shutdown of the economies continues, the decline in global natural gas consumption could amount to as much as 300 billion m3 and 36 million tons of LNG. A long-term increase in natural gas consumption will depend primarily on launched relief programs and the systematic recovery of economies after the COVID-19 pandemic [64].

3. Literature Review

Literature is abundant with literature reviews [65,66,67,68]. One can select core areas where gas consumption forecasts are drafted.

3.1. National Level

Literature provides information on various forecasting models in particular countries. For Poland, a nationwide gas consumption analysis has been performed using the Hubbert’s Model for a long-term forecast [69]. The Stochastic Gompertz Innovation Diffusion Process, as a stochastic growth model, has been used to determine the volume of gas consumed in Spain [70]. For Ireland, the national gas consumption forecast uses the Network Degree Day (NDDCA) model adapted to the climate [71]. For England, literature presents a relation whereby a daily air temperature drop by 1 centigrade causes an increase in electricity consumption by 1% and in gas consumption by 3% to 4% [72]. For two countries, the USA and Canada, gas consumption is forecasted using the Hubbert’s Model [73]. For Turkey (currently Türkiye), gas consumption has been forecasted using econometric models, genetic models, artificial neural networks frequently used in the forecasting area, and time series [74,75,76]. China’s gas consumption was analyzed using the PCMACP (Polynomial Curve and Moving Average Combination Projection) model [77] and Gray’s model [78,79].
Artificial neural networks forecasted gas consumption in the economies of Belgium [80], USA [81], and Poland [82].

3.2. Cities Level

For Türkiye, genetic and autoregressive algorithms have been used [1]; the MARS (Multivariate Adaptive Regression Splines) and CMASR (Conic Multivariate Adaptive Regression Splines) methods were used for the city of Ankara [83] as well as artificial neural networks [84,85], SVR [84], and the Network Degree Day method [86]. For cities on the territory of China, the authors have used the following for gas consumption forecasting: the Structure-Calibrated Support Vector Regression model (SC-SVR) [14]; the least squares method [87]; the optimized genetic algorithm improved by ANN with backpropagation [88]; and a hybrid model based on a simulated annealing algorithm, cross-correlation coefficient, and auxiliary vectors [89]. Hybrid models achieve better results than traditional models; hence, they are more frequently chosen to forecast gas consumption in cities [90,91]. Gas consumption in Poland has been forecasted using ANN [92,93] while in the case of cities in Croatia, the FTW (Fermat-Torricelli-Weber) model was applied plus the function dependent on gas consumption and temperature [94]. Daily gas demand for a selected city in Greece was forecasted using a hybrid model, comprising the continuous wavelet transform (CWT), genetic algorithm, adaptive neuro-fuzzy inference system (ANFIS), and feed-forward neural network (FFNN) [83].
In Iran, for a case study involving the city of Karaj, artificial neural networks were applied (machine learning plus genetic programming) [95] while the city of Yosuju used a hybrid model [96]. A descriptive analysis was performed for a selected city in Italy [11] whereas for a city in Slovenia, gas consumption was forecasted using machine learning, linear regression, and a recurrent neural network [97].

4. Case Study

Forecasts were made for commercial consumers (such as hotels, restaurants, and bars) that were subject to restrictions imposed by the government in March 2020 due to the coronavirus pandemic. Natural gas consumption is correlated with ambient temperature: when temperature rises, there is a decrease in gas consumption; when it drops the demand for natural gas increases (Figure 3).
Periods with the highest consumption are the winter months, such as January, February, and March as well as November and December. However, when considering gas consumption on a weekly basis, a slight drop can be observed on weekends (Figure 4 and Figure 5).

5. Neural Network

A neural network consists of an input layer, a hidden layer or layers, and an output layer. Each of these layers is made up of neurons, the basic building blocks of a neural network. An artificial neuron of the upstream layer is connected to all neurons of the downstream layer. There is no proper method for selecting the number of neurons in an ANN (artificial neural network). An excessive number of neurons can cause overfitting or the loss of the ability to generalize. Each neuron has a specific number of inputs whose significance is determined by weight values (wi). Artificial neurons can be compared to their biological models, and detailed descriptions can be found in the previously published literature [99,100]:
  • u1: inputs on dendrites (incoming signals passing through inputs)
  • w1: weights (correspond to synapses)
  • Σ: summation function (corresponds to the nucleus)
  • ϕ: activation function (corresponds to the axon hillock)
  • y: output (corresponds to the axon)
An artificial neuron picks up signals and multiplies them by weights. The signals are then routed to a summation function, which is responsible for stimulating the neuron. Stimulated signals are directed to the nonlinear activation function where the signal is generated [101]. Currently, multilayer networks are most commonly used. The input layer does not accept the signal without changing it. Neurons are then activated in the first hidden layer, where most of the calculations are performed. The hidden layer processes the signal and creates a model closely related to the analyzed process. After analyzing the signals in the hidden layer, intermediate signals are generated to all neurons of the output layer. The output layer generates the output signal [102,103,104]. There are two neural network learning methods. In supervised learning, the teacher chooses weights through the model signals at input to produce the best representation of the signal at the output. In other words, the network selects weights based on the results to achieve the best representation. In unsupervised learning, the network receives an input signal and generates an output signal without providing output data. If input/output signals cannot be applied, the gain is used. The gain delivered to the system is interpreted as a negative or positive signal [103,105,106,107,108,109].
Due to their advantages, neural networks have found a wide application: in power engineering, for controlling boilers [110], in forecasting heat energy for house heating [111] and electricity consumption [112], or in forecasting gas consumption [66,80,86,113,114].

6. Calculations

When conducting neural network calculations to find the best model, the following activation functions were checked for the input layer: linear, logistic, tangentoid, and exponential. The same functions were used for the output layer neurons. The Broyden−Fletcher−Goldfarb−Shanno (BFGS) algorithm was used for the calculations. First, the effect of data quantity on the model quality was calculated. The models used learning data from three, six, nine, and twelve months. The larger the learning data set, the smaller the MAPE error (Figure 6).
Three types of temperatures were used separately to perform the calculations: 24-h average temperature, maximum and minimum temperature, or temperature at ground level. Additional data was also used for the calculations: wind speed, humidity, and additional meteorological data, including the duration of selected atmospheric phenomena (water vapour pressure, cloud cover, station level pressure, daily precipitation, day−night precipitation, amount of snowfall, water equivalent in snow, insolation, duration of rainfall, duration of the snowfall, duration of rain with snow, duration of fog, duration of dew, duration of frost, and duration of a thunderstorm).
Models were also created using historical data (from one and two days back). Backward data was used for all atmospheric data. The rule of using backward data is that in addition to information about temperature and wind speed that may have affected natural gas consumption on a given day of natural gas consumption, temperature and wind speed values that occurred in the past were used.
Based on this data, the new neural network learned to find the best model (Table 1, Table 2 and Table 3). Detailed information about the configuration of the individual networks is provided in the “Supporting Information” file (Table S1).

7. Results

Results of the MAPE calculations for each estimated model will be presented in subsequent sections of this paper. For the 24-h average temperature, two models using data from the previous day and three models using data from the two previous days were improved (accounting for 20% of the models that used the 24-h average temperature). For the minimum and maximum temperature, one model improved after using data from one day back and two models improved after using data from two days back. The MAPE index value decreased when atmospheric event durations were added. Models that used the temperature at ground level for data from one day back improved in eight instances (accounting for more than 50% of all models), and nine models improved after using data from two days back (accounting for 60% of all models using data from previous days) (Figure 7, Figure 8, Figure 9, Figure 10 and Figure 11).
The best model turned out to be the one using the multi-l perceptron (MLP) 26-13-1 BFGS 12 average temperature. This means that there were 26 neurons in the input layer and 13 neurons in the hidden layers. The following quantities, among others, were used as independent parameters: temperature, wind speed, and months as “artificial data”. The days of the week (Monday−Sunday) and months (January–December), as artificial data, are recorded as (0, 1). If consumption occurred on a specific day of the month (Monday, January), then Monday and January receive signal (1) as the input. Other days and months receive no signal (0). (Figure S1). The network was trained in the 12th learning cycle. It has the following activation functions: linear in the input layer and exponential in the output layer. Figure 12 shows the comparison of the predicted and actual natural gas consumption performed using the selected artificial neural network model for the month selected for the neural network validation process. There is a good fit between the two curves, resulting in a low MAPE error value (2.17%). In addition, Figure 13 shows the detailed fit of the forecast results. There is a good fit in the range of the “average” values; larger differences are seen only for the extremes of low and high natural gas consumption. This is due to a smaller quantity of learning data at extreme values of the variability interval in the available learning data. The forecast made by the model was compared to actual natural gas consumption in March 2020 (Figure 14).
Figure 14 shows the comparison of the consumption forecast obtained using the selected artificial neural network with the smallest MAPE error to the actual consumption in March 2020. The consumption forecast was made for the available data without taking into account the impact of lockdown. It can be observed that the shape of both curves (forecasted and actual consumption) is highly consistent. The decrease in natural gas consumption caused by the COVID-19 pandemic was calculated from these parameters. The average decrease in natural gas consumption for the whole month was over 30%, increasing to nearly 45% when only the second half of the month was taken into account and reaching a maximum value of just over 54% on 19 March.

8. Conclusions

The COVID-19 pandemic that began in 2019 has had a strong impact on public health, the economy, and all aspects of society. There have been disruptions in supply and interruption of manufacturing and commercial activities. The pattern of electricity generation changed, with an increased share of renewable energy sources, while air quality improved. This infectious disease also had a major impact on the operations of the natural gas industry. There have been declines in natural gas consumption, reflecting the severity of the pandemic and introduction of restrictions. The most pronounced changes in natural gas consumption due to lockdown can be seen in commercial consumers. The impact of the pandemic on natural gas consumption of commercial consumers was analyzed, and evidence was presented that the lockdown resulted in a reduction of natural gas consumption for this group of consumers.
Based on the developed and validated artificial neural network model, a decrease was shown in the actual natural gas consumption as compared to predicted consumption in a normal operating situation. During the analysis for the month of March 2020, it can be seen that a more pronounced decrease in natural gas consumption started after 8 March, to reach a maximum on 19 March, at a level of 54%. After this period, the decrease in natural gas consumption oscillated between 40% and 45%.
Current observations show that the COVID-19 pandemic does not pose a high risk to natural gas distribution systems. However, the development of successive stages of the pandemic and the impact of reduced natural gas consumption on the financial standing of suppliers and consumers must be taken into account.
Further research should focus on the analysis of natural gas consumption in households and temporary changes that occurred in this group of consumers and whether consumption of the fuel has returned to levels from before the pandemic.
The current situation related to the COVID-19 pandemic should encourage the governments to undertake greater efforts oriented at energy transformation.
Renewable energy sources, principally including photovoltaics and wind farms, are experiencing dynamic growth. Energy transformation adopted by the European Parliament as well as a significant increase in fossil fuel prices have encouraged people to look for new alternative energy sources. The application of a new energy source can be perceived in an overly broad perspective as a supplementation or targeted replacement of fossil fuels used in industry and distribution to municipal clients. This is a result of the growing interest and demand for fuels originating from renewable sources, in relation to the implemented EU programs based on RES.
The use of biogas can serve as an example. Unfortunately, current legal regulations do not correspond to the needs for biomethane production: despite the fact that, theoretically, it has been possible for several years to input treated biogas in the gas networks, no biogas production plant has been connected to a distribution gas network.
Restrictive requirements as to the quality of the biogas input in the network cause technical problems and increase the production costs on the manufacturer’s part.
To maximize the share of Polish businesses in the supply chain for construction and operation of biogas and biomethane plants and to stimulate the universally understood biogas and biomethane market in Poland, the “Memorandum for Biogas and Biomethane Development” was signed. This is a sector of major importance, and it is indispensable for performing the energy transformation process to achieve autonomy and self-reliance.

Supplementary Materials

The following supporting information can be downloaded at: https://0-www-mdpi-com.brum.beds.ac.uk/article/10.3390/en15041393/s1. Table S1. Information about all artificial neural networks models for average temperature. Table S2. Information about all artificial neural networks models for average temperature with 1 day ago. Table S3. Information about all artificial neural networks models for average temperature with 2 days ago. Table S4. Information about all artificial neural networks models for min/max temperature. Table S5. Information about all artificial neural networks models for min/max temperature with 1 day ago. Table S6. Information about all artificial neural networks models for min./max. temperature with 2 days ago. Table S7. Information about all artificial neural networks models at a ground-level temperature. Table S8. Information about all artificial neural networks models at a ground-level temperature with 1 day backdate. Table S9. Information about all artificial neural networks models at a ground-level temperature with 2 days ago. Figure S1. Topology of the best neural network MLP 26-13-1.

Author Contributions

Conceptualization, T.C. and P.N.; methodology, T.C.; validation, T.C.; formal analysis, T.C., P.N., A.S. and K.K.; investigation, T.C. and K.K.; resources, T.C., P.N., A.S. and K.K.; data curation, T.C.; writing—original draft preparation T.C., P.N. and K.K.; writing—review and editing, T.C., P.N., A.S. and K.K.; visualization, T.C., P.N., A.S. and K.K.; supervision, T.C. and K.K.; project administration, T.C. and K.K.; funding acquisition, A.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research was cofinanced by the Research Subsidy of the AGH University of Science and Technology for the Faculty of Energy and Fuels (No. 16.16.210.476) and by the Faculty of Drilling, Oil and Gas (No. 16.16.190.779).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Ervural, B.F.; Beyca, O.F.; Zaim, S. Model estimation of ARMA using genetic algorithms: A case study of forecasting natural gas consumption. Procd. Soc. Behv. 2016, 235, 537–545. [Google Scholar] [CrossRef]
  2. Voudouris, V.; Matsumoto, K.; Sedgwick, J.; Rigby, R.; Stasinopolus, D.; Jeffersin, M. Exploring the production of natural gas through the lenses of the ACEGRS model. Energy Policy 2016, 64, 124–133. [Google Scholar] [CrossRef]
  3. GUS. “Główny Urząd Statystyczny” (Statistics Poland). 2018–2020. Available online: https://stat.gov.pl/ (accessed on 3 June 2021).
  4. Szurlej, A.; Ruszel, M.; Olkuski, T. Will natural gas be competitive fuel? (in Polish: Czy gaz ziemny będzie paliwem konkurencyjnym?). Rynek Energii 2015, 5, 3–10. [Google Scholar]
  5. Ciechanowska, M. Poland’s Energy Policy until 2050 (in Polish: Polityka Energetyczna Polski do 2050 roku). Naft. Gaz 2014, 11, 839–842. [Google Scholar]
  6. Europen Commission. Energy Roadmap 2050; European Commissioner for Energy: Luxembourg, 2012. [Google Scholar]
  7. Kosowski, P.; Kosowska, K. Valuation of Energy Security for Natural Gas—European Example. Energies 2021, 14, 2678. [Google Scholar] [CrossRef]
  8. Statistical Review of World Energy. Available online: https://www.bp.com/en/global/corporate/energy-economics/statistical-review-of-world-energy.html (accessed on 3 June 2021).
  9. Ozmen, A.; Yilmaz, Y.; Weber, G.-W. Natural gas consumption forecast with MARS and CMARS models for residential users. Energy Econ. 2018, 70, 357–381. [Google Scholar] [CrossRef]
  10. Sanchez-Ubeda, E.F.; Berzosa, A. Modeling and forecasting industrial end-use natural gas consumption. Energy Econ. 2007, 29, 710–742. [Google Scholar] [CrossRef]
  11. Baldacci, L.; Golfarelli, M.; Lombardi, D.; Sami, F. Natural gas consumption forecasting for anomaly detection. Expert Syst. Appl. 2016, 62, 190–201. [Google Scholar] [CrossRef]
  12. Khan, M.A. Modelling and forecasting the demand for natural gas in Pakistan. Renew. Sust. Energ. Rev. 2015, 49, 1145–1159. [Google Scholar] [CrossRef]
  13. Bartnicki, G.; Nowak, B. Model ARIMA w prognozowaniu zużycia gazu w cyklach miesięcznych. Zesz. Nauk. Inst. Gospod. Surowcami Miner. I Energią Pol. Akad. Nauk 2018, 103, 145–158. [Google Scholar] [CrossRef]
  14. Bai, Y.; Li, C. Daily natural gas consumption forecasting based on a structure-calibrated support vector regression approach. Energ Build. 2016, 127, 571–579. [Google Scholar] [CrossRef]
  15. Brown, R.H.; Kaftan, D.J.; Smalley, J.L.; Fakoor, M.; Graupman, S.J.; Povinelli, R.J.; Corliss, G.F. Improving Daily Natural Gas Forecasting by Tracking and Combining Models. In Proceedings of the International Symposium on Forecasting, Cairns, QLD, Australia, 25–28 June 2017; Available online: https://epublications.marquette.edu/electric_fac/288 (accessed on 3 June 2021).
  16. Samidurai, A.; Das, A. Cardiovascular Complications Associated with COVID-19 and Potential Therapeutic Strategies. Int. J. Mol. Sci. 2020, 21, 6790. [Google Scholar] [CrossRef] [PubMed]
  17. Krajewska, J.; Krajewski, W.; Zub, K.; Zatoński, T. COVID 19 in otolaryngologist practice: A review of current knowledge. Eur. Arch. Oto-Rhino-Laryngol. 2020, 277, 1885–1897. [Google Scholar] [CrossRef] [PubMed]
  18. Gottlieb, M.; Long, B. Dermatologic manifestations and complications of COVID-19. Am. J. Emerg. Med. 2020, 38, 1715–1721. [Google Scholar] [CrossRef] [PubMed]
  19. Klok, F.A.; Kruip, M.J.; Van der Meer, N.J.; Arbous, M.S.; Gommers, D.A.; Kant, K.M.; Kaptein, F.H.; Van Paassen, J.; Stals, M.A.; Huisman, M.; et al. Incidence of thrombotic complications in critically ill ICU patients with COVID-19. Thromb. Res. 2020, 19, 145–147. [Google Scholar] [CrossRef] [PubMed]
  20. Bridwell, R.; Long, B.; Gottlieb, M. Neurologic complications of COVID-19. Am. J. Emerg. Med. 2020, 38, 1549.e3–1549.e7. [Google Scholar] [CrossRef]
  21. Kunutsor, S.K.; Laukkanen, J.A. Renal complications in COVID-19: A systematic review and meta-analysis. Ann. Med. 2020, 52, 345–353. [Google Scholar] [CrossRef]
  22. Aminian, A.; Safari, S.; Razeghian-Jahromi, A.; Ghorbani, M.; Delaney, C.P. COVID-19 Outbreak and Surgical Practice: Unexpected Fatality in Perioperative period. Ann. Surg. 2020, 272, e27–e29. [Google Scholar] [CrossRef]
  23. Alqahtani, S.; Schattenberg, J. Liver injury in COVID-19: The current evidence. United Eur. Gastroent. 2020, 8, 509–519. [Google Scholar] [CrossRef]
  24. Pal, R.; Bhadada, S.K. COVID-19 and diabetes mellitus: An unholy interaction of two pandemics. Diabetes Metab. Syndr. Clin. Res. Rev. 2020, 14, 513–517. [Google Scholar] [CrossRef]
  25. She, J.; Jiang, J.; Ye, L.; Hu, L.; Bai, C.; Song, Y. 2019 novel coronavirus of pneumonia in Wuhan, China: Emerging attack and management strategies. Clin. Trans. Med. 2020, 9, 1–7. [Google Scholar] [CrossRef] [PubMed]
  26. WHO (World Health Organization). Coronavirus Disease 2019 (COVID-19) Situation Report-88. 2020. Available online: https://www.who.int/ (accessed on 3 June 2021).
  27. Ministerstwo Zdrowia (Ministry of Health). Ordinance of the Minister of Health of 20 March 2020 on the declaration of an epidemic in the Republic of Poland (in Polish: Rozporządzenie Ministra Zdrowia z dnia 20 marca 2020 w sprawie ogłoszenia na obszarze Rzeczpospolitej Polskiej stanu epidemii). 2020. Available online: https://isap.sejm.gov.pl/isap.nsf/DocDetails.xsp? (accessed on 3 June 2021).
  28. Deloitte Touche Tohmatsu India LLP. Member of Deloitte Touche Tohmatsu Limited. 2020. Available online: https://www2.deloitte.com/content/dam/Deloitte/in/Documents/finance/in-fa-impact-of-covid-19-on-o-and-g-industry-noexp.pdf (accessed on 3 June 2021).
  29. Brosemer, K.; Schelly, C.; Gagnon, V.; Arola, K.L.; Pearce, J.M.; Bessette, D.; Olabisi, L.S. The energy crises revealed by COVID: Intersections of Indigeneity, inequity, and health. Energy Res. Soc. Sci. 2020, 68, 101661. [Google Scholar] [CrossRef] [PubMed]
  30. Mostafa, M.K.; Gamal, G.; Wafig, A. The impact of COVID 19 on air pollution levels and other environmental indicators—A case study of Egypt. J. Environ. Manag. 2021, 277, 111496. [Google Scholar] [CrossRef] [PubMed]
  31. Filonchyk, M.; Hurynovich, V.; Yan, H. Impact of COVID-19 lockdown on air quality in the Poland, Eastern Europe. Environ. Res. 2020, 198, 110454. [Google Scholar] [CrossRef] [PubMed]
  32. Karan, A.; Kabeer, A.; Teelucksingh, S.; Sakhamuri, S. The impact of air pollution on the incidence and mortality of COVID-19. Glob. Health Res. Policy 2020, 5, 1–3. [Google Scholar] [CrossRef]
  33. Nurshad, A.; Farjana, I. The Effects of Air Pollution on COVID-19 Infection and Mortality—A Review on Recent Evidence. Front. Public Health 2020, 8, 580057. [Google Scholar] [CrossRef]
  34. Comunian, S.; Dongo, D.; Milani, C.; Palestini, P. Air pollution and COVID-19: The role of particulate matter in the spread and increase of COVID-19’s morbidity and mortality. Int. J. Environ. Res. Public Health 2020, 17, 1–20. [Google Scholar] [CrossRef]
  35. COVID-19 pandemic impacts on traffic system delay, fuel consumption and emissions. Int. J. Transp. Sci. Technol. 2020, 10, 184–196. [CrossRef]
  36. Prol, J.L.; Sungmin, O. Impact of COVID-19 Measures on Short-Term Electricity Consumption in the Most Affected EU Countries and USA States. iScience 2020, 23, 101639. [Google Scholar] [CrossRef]
  37. Department for Business, Energy & Industrial Strategy. Energy Trends. 2021. Available online: https://www.gov.uk/government/statistics/electricity-section-5-energy-trends (accessed on 3 June 2021).
  38. Mahajan, M. Estimating, U.S. Energy Demand and Emissions Impacts of COVID-19 with the Energy Policy Simulator. Energy Innovation: San Francisco, CA, USA, 2020. Available online: https://energyinnovation.org/wp-content/uploads/2020/05/Modeling-COVID-Impacts-On-US-Emissions.pdf (accessed on 3 June 2021).
  39. Bulut, M. Analysis of the COVID-19 impact on electricity consumption and production. Sak. Univ. J. Comput. Inf. Sci. (SAUCIS) 2020, 3, 283–295. [Google Scholar] [CrossRef]
  40. Elavarasan, R.M.; Shafiullah, G.M.; Raju, K.; Mudgal, V.; Arif, M.T.; Jamal, T.; Subramanian, S.; Balaguru, V.S.S.; Reddy, K.S.; Subramaniam, U. COVID-19: Impact analysis and recommendations for power sector operation. Appl. Energy 2020, 279, 115739. [Google Scholar] [CrossRef] [PubMed]
  41. Ruan, G.; Wu, J.; Zhong, H.; Xia, Q.; Xie, L. Quantitative assessment of U.S. bulk power systems and market operations during the COVID-19 pandemic. Appl. Energy 2021, 286, 116354. [Google Scholar] [CrossRef] [PubMed]
  42. Eryilmaz, D.; Partia, M.; Heilbrum, C. Assessment of the COVID-19 pandemic effect on regional electricity generation mix in NYISO, MISO, and PJM markets. Electr. J. 2020, 33, 106829. [Google Scholar] [CrossRef]
  43. Werth, A.; Gravino, P.; Prevedello, G. Impact analysis of COVID-19 responses on energy grid dynamics in Europe. Appl. Energy 2021, 281, 116045. [Google Scholar] [CrossRef] [PubMed]
  44. Halbrügge, S.; Schott, P.; Weibelzahl, M.; Buhl, H.U.; Fridgen, G.; Schöpf, M. How did the German and other European electricity systems react to the COVID-19 pandemic? Appl. Energy 2021, 285, 116370. [Google Scholar] [CrossRef]
  45. Abadie, L.M. Energy Market Prices in Times of COVID-19: The Case of Electricity and Natural Gas in Spain. Energies 2021, 14, 1632. [Google Scholar] [CrossRef]
  46. Ghiani, E.; Galici, M.; Mureddu, M.; Pilo, F. Impact on electricity consumption and market Pricing of energy and ancillary services during pandemic of COVID-19 in Italy. Energies 2020, 13, 3357. [Google Scholar] [CrossRef]
  47. Kalbusch, A.; Henning, E.; Brikalski, M.P.; Vieira de Luca, F.; Konrath, A.C. Impact of coronavirus (COVID-19) spread-prevention actions on urban water consumption. Resour. Conserv. Recycl. 2020, 163, 105098. [Google Scholar] [CrossRef]
  48. Huang, L.; Liao, Q.; Qiu, R.; Liang, Y.; Long, Y. Prediction-based analysis on power consumption gap under long-term emergency: A case in China under COVID-19. Appl. Energy 2021, 283, 116339. [Google Scholar] [CrossRef]
  49. Eastman, L.; Smull, E.; Patterson, L.; Doyle, M. COVID-19 Impacts on Water. Utility Consumption and Revenues. Preliminary results. Available online: www.raftelis.com/covid-19-resources (accessed on 3 June 2021).
  50. Ong, A.; Nielsen, E. Economic Impacts of COVID-19 on the Water Sector. 2020. Available online: https://www.water.org.uk/wp-content/uploads/2020/12/Impact-of-COVID-19-on-the-water-sector_FINAL-REPORT-STC-141220.pdf (accessed on 3 June 2021).
  51. Cheshmehzangi, A. COVID-19 and household energy implications: What are the main impacts on energy use. Heliyon 2020, 6, e05202. [Google Scholar] [CrossRef]
  52. Nemati, M. COVID-19 and Urban Water Consumption. Giannini Foundation of Agricultural Economics. ARE Update 2020, 24, 9–11. Available online: https://s.giannini.ucop.edu/uploads/giannini_public/6c/45/6c45e9f6-7f90-4c99-9d14-63905593539a/v24n1_3.pdf (accessed on 3 June 2021).
  53. Balacco, G.; Totaro, V.; Iacobellis, V.; Manni, A.; Spagnoletta, M.; Piccinni, A.F. Influence of COVID-19 spread on water drinking demand: The case of Puglia Region (Southern Italy). Sustainability 2020, 12, 5919. [Google Scholar] [CrossRef]
  54. Mastropierto, P.; Rodilla, P.; Batlle, C. Emergency measures to protect energy consumers during the COVID-19 pandemic: A global review and critical analysis. Energy Res. Soc. Sci. 2020, 68, 101678. [Google Scholar] [CrossRef]
  55. Kim, D.; Yim, T.; Lee, J.Y. Analytical study on changes in domestic hot water use caused by COVID-19 pandemic. Energy 2021, 231, 120915. [Google Scholar] [CrossRef]
  56. Collette, M.W.; Baffes, J.; Kabundi, A.; Kindberg-Hanlon, G.; Nagle, P.S.; Ohnsorge, F.L. Adding fuel to the fire. Cheap oil during the COVID-19 Pandemic. In Policy Research Working Paper World Bank Group; The World Bank: Washington, DC, USA, 2020; Available online: http://hdl.handle.net/10986/34129 (accessed on 3 June 2021).
  57. Aloui, D.; Goutte, S.; Guesmi, K.; Hchaichi, R. COVID-19’s impact on crude oil and natural gas S&P GS Indexes. HAL Sci. Hum. Et Soc. 2020, 1–17. [Google Scholar] [CrossRef]
  58. Prawiraatmadja, W. COVID-19 Pandemic: Impact on the Oil and Gas Industry; Institute for Essential Services Reform (IESR): Jakarta, Indonesia, 2020; Available online: http://iesr.or.id/wp-content/uploads/2020/05/Covid19-Pandemic-Impact-on-the-Oil-and-Gas-Industry-IESR.pdf (accessed on 3 June 2021).
  59. Ghosh, S. Marine Insight. 2021. Available online: https://www.marineinsight.com/know-more/oil-tanker-business-boomed-during-covid-19-pandemic/ (accessed on 3 June 2021).
  60. IEA. Countries and Regions. 2021. Available online: https://www.iea.org/countries (accessed on 3 June 2021).
  61. Amara, R.; Belaifa, M. COVID-19 and Its Implications on the Italian Natural Gas Market. 2020. Available online: https://www.gecf.org/_resources/files/events/gecf-expert-commentary---covid-19-and-its-implications-on-the-italian-natural-gas-market/covid-19-and-its-implication-in-the-italian-gas-market.pdf (accessed on 3 June 2021).
  62. Honore, A. Natural Gas Demand in Europe: The Impacts of COVID-19 and other Influences in 2020; The Oxford Institute for Energy Studies: Oxford, UK, 2021; Available online: https://www.oxfordenergy.org/wpcms/wp-content/uploads/2020/06/Natural-gas-demand-in-Europe-the-impacts-of-COVID-19-and-other-influences-in-2020.pdf (accessed on 3 June 2021).
  63. IGU. Global Gas Report 2020. Available online: https://www.igu.org/resources/global-gas-report-2020/ (accessed on 3 June 2021).
  64. Koyama, K.; Suehiro, S. COVID-19 and the Outlook for Oil, Natural Gas, and LNG Demand in 2021; The Institute Of Energy Economics, Japan: Tokyo, Japan, 2020; pp. 1–5. Available online: https://eneken.ieej.or.jp/data/8933.pdf (accessed on 3 June 2021).
  65. Soldo, B. Forecasting Natural Gas Consumption. Appl. Energy 2012, 92, 26–37. [Google Scholar] [CrossRef]
  66. Musilek, P.; Palikan, T.; Brabec, M.; Simunek, M. Recurrent Neural Network Based Gating for Natural Gas Load Prediction System. In Proceedings of the 2006 IEEE International Joint Conference on Neural Network Proceedings, Vancouver, BC, Canada, 16–20 July 2006; pp. 3736–3741. [Google Scholar] [CrossRef]
  67. Wang, X.; Luo, D.; Liu, J.; Wang, W.; Jie, G. Prediction of natural Gas Consumption in Different regions of China using a hybrid MVO-NNGBM. Math. Probl. Eng. 2017, 2017, 6045708. [Google Scholar] [CrossRef]
  68. Jinynuan, L.; Nan, W.; Shouxi, W.; Xi, C.; Hanyui, X.; Wang, J. Natural gas consumption forecasting: A discussion on forecasting history and future challenges. J. Nat. Gas. Sci. Eng. 2021, 90, 103930. [Google Scholar] [CrossRef]
  69. Siemek, J.; Nagy, S.; Rychlicki, S. Estimation of natural-gas consumption in Poland based on the logistic-curve interpretation. Appl. Energy 2003, 75, 1–7. [Google Scholar] [CrossRef]
  70. Gutierrez, R.; Nafidi, A.; Gutierrez Sanchez, R. Forecasting total natural gas consumption in Spain by using the stochastic Gompertz innovation diffusion model. Appl. Energy 2005, 80, 115–124. [Google Scholar] [CrossRef]
  71. Oliver, R.; Duffy, A.; Enright, B.; O’Connor, R. Forecasting peak-day consumption for year-ahead management of natural gas networks. Util Policy 2017, 44, 1–11. [Google Scholar] [CrossRef]
  72. Thornton, H.E.; Hoskins, B.J.; Scaife, A.A. The role of temperature in the variability and extremes of electricity and gas demand in Great Britain. Environ. Res. Lett 2016, 11, 114015. [Google Scholar] [CrossRef]
  73. Reynolds, D.; Kolodziej, M. North America Natural Gas Supply Forecast: The Hubbert Method including the Effects of Institutions. Energies 2009, 2, 269–306. [Google Scholar] [CrossRef]
  74. Taspinar, F.; Celebi, N.; Tutkun, N. Forecasting of daily natural gas consumption on regional basis in Turkey using various computational methods. Energy Build. 2013, 56, 23–31. [Google Scholar] [CrossRef]
  75. Melikoglu, M. Vision 2023: Forecasting Turkey’s natural gas demand between 2013 and 2030. Renew. Sustain. Energy Rev. 2013, 22, 393–400. [Google Scholar] [CrossRef]
  76. Aras, N. Forecasting Residential Consumption of natural Gas Using Genetic Algorithms. Energy Explor. Exploit 2008, 26, 241–266. [Google Scholar] [CrossRef]
  77. Xu, G.; Wang, W. Forecasting China’s natural gas consumption based on combination model. J. Nat. Gas. Chem. 2010, 19, 493–496. [Google Scholar] [CrossRef]
  78. Shaikh, F.; Ji, Q.; Shaikh, P.H.; Mirjat, N.H.; Uqaili, M.A. Forecasting China’s natural gas demand based on optimized nonlinear grey models. Energy 2017, 140, 941–951. [Google Scholar] [CrossRef]
  79. Yifei, M.; Yanli, L. Analysis of the supply-demand status of China’s natural gas to 2020. Pet. Sci. 2010, 7, 132–135. [Google Scholar] [CrossRef]
  80. Suykens, J.A.K.; Lemmerling, P.; Favoreel, W.; De Moor, B.L.R.; Crepel, M.; Briol, P. Modelling the Belgian gas consumption using neural networks. Neural. Process. Lett. 1996, 4, 157–166. [Google Scholar] [CrossRef]
  81. Brown, R.H.; Kaftan, D.J.; Feng, X.; Piessens, L.P.; Nestor, D. Development of feed-forward network models to predict gas consumption. In Proceedings of the 1994 IEEE International Conference on Neural Networks (ICNN’94), Orlando, FL, USA, 28 June–2 July 1994; Volume 2, pp. 802–805. [Google Scholar] [CrossRef]
  82. Paliński, A. Forecasting gas demand using artificial intelligence methods (In Polish: Prognozowanie zapotrzebowania na gaz metodami sztucznej inteligencji). Naft. Gaz 2019, 2, 111–117. [Google Scholar] [CrossRef]
  83. Panapakidis, I.P.; Dagoumas, A.S. Day-ahead natural gas demand forecasting based on the combination of wavelet transform and ANFIS/genetic algorithm/neural network model. Energy 2017, 118, 231–245. [Google Scholar] [CrossRef]
  84. Beyca, O.F.; Ervural, B.C.; Tatoglu, E.; Ozuyar, P.G.; Zaim, S. Using machine learning tools for forecasting natural gas consumption in the province of Istanbul. Energy Econ. 2019, 80, 937–949. [Google Scholar] [CrossRef]
  85. Kizilaslan, R.; Karlik, B. Comparison neural networks models for short term forecasting of natural gas consumption in Istanbul, 2008. Proceeding of the First International Conference on the Applications of Digital Information and Web Technologies (ICADIWT), Ostrava, Czech Republik, 4–6 August 2008; pp. 448–453. [Google Scholar] [CrossRef]
  86. Sarak, H.; Satman, A. The degree-day method to estimate the residential heating natural gas consumption in Turkey: A case study. Energy 2003, 28, 929–939. [Google Scholar] [CrossRef]
  87. Yu, Y.; Zheng, X.; Han, Y. On the demand for natural gas in urban China. Energy Policy 2014, 70, 57–63. [Google Scholar] [CrossRef]
  88. Yu, F.; Xu, X. A short-term load forecasting model of natural gas based on optimized genetic algorithm and improved BP neural network. Appl. Energy 2014, 134, 102–113. [Google Scholar] [CrossRef]
  89. Lu, H.; Azimi, M.; Iseley, T. Short-term load forecasting of urban gas using a hybrid model based on improved fruit fly optimization algorithm and support vector machine. Energy Rep. 2019, 5, 666–677. [Google Scholar] [CrossRef]
  90. Wei, N.; Li, C.; Peng, X.; Li, Y.; Zeng, F. Daily natural gas consumption forecasting via the application of a novel hybrid model. Appl. Energy 2019, 250, 358–368. [Google Scholar] [CrossRef]
  91. Wei, N.; Li, C.; Duan, J.; Liu, J.; Zeng, F. Daily Natural Gas Load Forecasting Based on a Hybrid deep Learning Model. Energies 2019, 12, 218. [Google Scholar] [CrossRef]
  92. Szoplik, J. Forecasting of natural gas consumption with artificial neural networks. Energy 2015, 85, 208–220. [Google Scholar] [CrossRef]
  93. Cieślik, T.; Metelska, K. Modeling of gas consumption in the city. AGH Drill. Oil Gas 2017, 34, 439–453. [Google Scholar] [CrossRef]
  94. Sabo, K.; Scitovski, R.; Vazler, I.; Zekic-Susac, M. Mathematical models of natural gas consumption. Energ Convers. Manag. 2011, 52, 1721–1727. [Google Scholar] [CrossRef]
  95. Izadyar, N.; Ong, H.C.; Shamshirband, S.; Ghadamian, H.; Tong, C.W. Intelligent forecasting of residential heating demand for the District Heating System based on the monthly overall natural gas consumption. Energy Build. 2015, 104, 208–214. [Google Scholar] [CrossRef]
  96. Karimi, H.; Dastranj, J. Artificial neural network-based genetic algorithm to predict natural gas consumption. Energy Syst. 2014, 5, 571–581. [Google Scholar] [CrossRef]
  97. Hribar, R.; Potocnik, P.; Silc, J.; Papa, G. A comparison of model for forecasting the residential natural gas demand of an urban area. Energy 2019, 167, 511–522. [Google Scholar] [CrossRef]
  98. Meteomodel. 2021. Available online: https://meteomodel.pl/ (accessed on 3 June 2021).
  99. Mohamed, Z.E. Using the artificial neural networks for prediction and validating solar radiation. J. Egypt. Math. Soc. 2019, 27, 47. [Google Scholar] [CrossRef]
  100. Cieślik, T.; Kogut, K. Forecasting the work of gas network by means of artificial neural network (in Polish: Prognozowanie pracy sieci gazowej za pomocą sztucznych sieci neuronowych). Naft. Gaz 2016, 6, 443–450. [Google Scholar] [CrossRef]
  101. Żurada, J. Artificial Neural Networks (In Polish: Sztuczne Sieci Neuronowe); PWN: Warszawa, Poland, 1996. [Google Scholar]
  102. Tadeusiewicz, R.; Gaciarz, T.; Borowik, B.; Leper, B. Exploring the Properties of Neural Networks (in Polish: Odkrywanie Własności Sieci Neuronowych); Polska Akademia Umiejętności: Kraków, Poland, 2007. [Google Scholar]
  103. Silva, N.I. Artificial Neural Network Architectures and Training Processes. In Artificial Neural Networks; Springer International Publishing: Cham, Switzerland, 2017. [Google Scholar] [CrossRef]
  104. Wang, S.-C. Interdisciplinary Computing in Java Programming; Springer Science+Business Media: New York, NY, USA, 2003. [Google Scholar] [CrossRef]
  105. Gallo, C. Artificial neural network: Tutorial. In Encyclopedia of Information Science and Technology, 3rd ed.; Khosrow-Pour, M., Ed.; IGI Global: Hershey, PA, USA, 2015; pp. 6369–6378. [Google Scholar]
  106. Wójcik, M. A Gas Network Model Based on Artificial Neural Networks (In Polish: Model Sieci Gazowniczej Oparty o Sztuczne Sieci Neuronowe). Master of Thesis, AGH University Science and Technology, Kraków, Poland, 2005. [Google Scholar]
  107. Maciejasz, M. Application of Neural Networks for the Analysis of Transmission Network Operation (In Polish: Zastosowanie Sieci Neuronowych do Analizy Pracy Sieci Przesyłowych). Master of Thesis, AGH University Science and Technology, Kraków, Poland, 2005. [Google Scholar]
  108. Kogut, K. Analysis of Natural Gas Transmission Network Modeling Capabilities (In Polish: Analiza Możliwości Modelowania Sieci Przesyłowej Gazu Ziemnego). Ph.D. Thesis, AGH University Science and Technology, Kraków, Poland, 2007. [Google Scholar]
  109. Stefanowski, J.; Lectures on Neural Networks (In Polish: Wykłady z Sieci Neuronowych). Machine Learning. Available online: http://www.cs.put.poznan.pl/jstefanowski/mlteaching.html (accessed on 3 June 2021).
  110. Lichota, J. Neural Control of Thermal Energy Objects (in Polish: Neuronowe Sterowanie Obiektami Termoenergetycznymi); Oficyna Wydawnicza Politechniki Wrocławskiej: Wrocław, Polonia, 2013. [Google Scholar]
  111. Dombayci, A.O. The prediction of heating energy consumption in a model house by using artificial neural networks in Denizli–Turkey. Adv. Eng. Softw. 2010, 41, 141–147. [Google Scholar] [CrossRef]
  112. Cieślik, T. Change in the structure of electricity generation in the USA, China, Japan and the EU, and a forecast of electricity consumption. AGH Drill. Oil Gas 2017, 34, 291–302. [Google Scholar] [CrossRef]
  113. Khotanzad, A.; Elragal, H.; Lu, T.L. Combination of artificial neural-network forecasters for prediction of natural gas consumption. IEEE Trans. Neural. Netw. 2000, 11, 464–473. [Google Scholar] [CrossRef]
  114. Panek, W.; Wlodek, T. Natural Gas Consumption Forecasting Based on the Variability of External Meteorological Factors Using Machine Learning Algorithms. Energies 2022, 15, 348. [Google Scholar] [CrossRef]
Energies 15 01393 g001 550
Figure 1. Changes in natural gas consumption worldwide [8].
Figure 1. Changes in natural gas consumption worldwide [8].
Energies 15 01393 g001
Energies 15 01393 g002 550
Figure 2. Changes in natural gas consumption in Poland [8].
Figure 2. Changes in natural gas consumption in Poland [8].
Energies 15 01393 g002
Energies 15 01393 g003 550
Figure 3. Natural gas consumption in 2019, depending on temperature [98].
Figure 3. Natural gas consumption in 2019, depending on temperature [98].
Energies 15 01393 g003
Energies 15 01393 g004 550
Figure 4. Average gas consumption in individual months (x-axis: natural gas consumption in m3/day (24 h)).
Figure 4. Average gas consumption in individual months (x-axis: natural gas consumption in m3/day (24 h)).
Energies 15 01393 g004
Energies 15 01393 g005 550
Figure 5. Average gas consumption on individual days of the week (x-axis: natural gas consumption in m3/day (24 h)).
Figure 5. Average gas consumption on individual days of the week (x-axis: natural gas consumption in m3/day (24 h)).
Energies 15 01393 g005
Energies 15 01393 g006 550
Figure 6. Impact of the learning set size on model quality.
Figure 6. Impact of the learning set size on model quality.
Energies 15 01393 g006
Energies 15 01393 g007 550
Figure 7. MAPE error results for the models using the 24-h average temperature.
Figure 7. MAPE error results for the models using the 24-h average temperature.
Energies 15 01393 g007
Energies 15 01393 g008 550
Figure 8. MAPE error results for the models using the min/max temperature.
Figure 8. MAPE error results for the models using the min/max temperature.
Energies 15 01393 g008
Energies 15 01393 g009 550
Figure 9. MAPE error results for the models using the temperature at ground level.
Figure 9. MAPE error results for the models using the temperature at ground level.
Energies 15 01393 g009
Energies 15 01393 g010 550
Figure 10. Ratio of models that have improved by adding information from the previous day.
Figure 10. Ratio of models that have improved by adding information from the previous day.
Energies 15 01393 g010
Energies 15 01393 g011 550
Figure 11. The ratio of models that have improved by adding information from two days back.
Figure 11. The ratio of models that have improved by adding information from two days back.
Energies 15 01393 g011
Energies 15 01393 g012 550
Figure 12. Comparison of natural gas consumption forecast to the actual measurement.
Figure 12. Comparison of natural gas consumption forecast to the actual measurement.
Energies 15 01393 g012
Energies 15 01393 g013 550
Figure 13. Forecast value and actual natural gas consumption fit on individual days.
Figure 13. Forecast value and actual natural gas consumption fit on individual days.
Energies 15 01393 g013
Energies 15 01393 g014 550
Figure 14. Forecast consumption vs. actual consumption in March 2020 after lockdown.
Figure 14. Forecast consumption vs. actual consumption in March 2020 after lockdown.
Energies 15 01393 g014
Table
Table 1. Data ranges used to calculate the best model. Temperature: daily average, min/max, or at ground level.
Table 1. Data ranges used to calculate the best model. Temperature: daily average, min/max, or at ground level.
Model No.Independent Parameter
Atmospheric DataArtificial Data
TemperatureWind SpeedHumidityAdditional Data with Duration of Atmospheric EventDays of the WeekMonths
1×
2××
3×× ×
4×× ××
5×× ×
6× ×
7×××
8××× ×
9××× ××
10××× ×
11× ××
12××××
13×××× ×
14××××××
15×××××
Table
Table 2. Data ranges used to calculate the best model. Temperature: daily average, min/max, or at ground level with data from 1 day back.
Table 2. Data ranges used to calculate the best model. Temperature: daily average, min/max, or at ground level with data from 1 day back.
Model No.Independent Parameter
Atmospheric DataArtificial Data
TemperatureWind SpeedHumidityAdditional Data with Duration of Atmospheric EventData from One Day BackDays of the WeekMonths
1× ×
2×× ×
3×× × ×
4×× ×××
5×× ××
6× × ×
7××× ×
8××× × ×
9××× ×××
10××× ××
11× ×××
12×××××
13××××× ×
14×××××××
15××××××
Table
Table 3. Data ranges used to calculate the best model. Temperature: daily average, min/max, or at ground level with data from 2 days back.
Table 3. Data ranges used to calculate the best model. Temperature: daily average, min/max, or at ground level with data from 2 days back.
Model No.Independent Parameter
Atmospheric DataArtificial Data
TemperatureWind SpeedHumidityAdditional Data with Duration of Atmospheric EventData from One Day BackData from Two Days BackDays of the WeekMonths
1× ××
2×× ××
3×× ×× ×
4×× ××××
5×× ×××
6× × ××
7××× ××
8××× ×× ×
9××× ××××
10××× ×××
11× ××××
12××××××
13×××××× ×
14××××××××
15×××××××
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Cieślik, T.; Narloch, P.; Szurlej, A.; Kogut, K. Indirect Impact of the COVID-19 Pandemic on Natural Gas Consumption by Commercial Consumers in a Selected City in Poland. Energies 2022, 15, 1393. https://0-doi-org.brum.beds.ac.uk/10.3390/en15041393

AMA Style

Cieślik T, Narloch P, Szurlej A, Kogut K. Indirect Impact of the COVID-19 Pandemic on Natural Gas Consumption by Commercial Consumers in a Selected City in Poland. Energies. 2022; 15(4):1393. https://0-doi-org.brum.beds.ac.uk/10.3390/en15041393

Chicago/Turabian Style

Cieślik, Tomasz, Piotr Narloch, Adam Szurlej, and Krzysztof Kogut. 2022. "Indirect Impact of the COVID-19 Pandemic on Natural Gas Consumption by Commercial Consumers in a Selected City in Poland" Energies 15, no. 4: 1393. https://0-doi-org.brum.beds.ac.uk/10.3390/en15041393

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop