Electricity Cosumption Analysis in Germany

Introduction

This report summarizes the analysis of Germany’s daily electricity consumption, focusing on trends, seasonal patterns, and forecasting.

Data Overview

The data-set contains daily electricity consumption, wind, and solar production data from 2006 to 2018.

Load data

First 6 Rows of the Dataset

Date	Consumption	Wind	Solar	Wind.Solar
2006-01-01	1069.184	NA	NA	NA
2006-01-02	1380.521	NA	NA	NA
2006-01-03	1442.533	NA	NA	NA
2006-01-04	1457.217	NA	NA	NA
2006-01-05	1477.131	NA	NA	NA
2006-01-06	1403.427	NA	NA	NA

Summary

Summarize Data Frame to understand better way.

Summary Statistics

	Date	Consumption	Wind	Solar	Wind.Solar
	Min. :2006-01-01	Min. : 842.4	Min. : 5.757	Min. : 1.968	Min. : 21.48
	1st Qu.:2008-12-31	1st Qu.:1217.9	1st Qu.: 62.353	1st Qu.: 35.179	1st Qu.:172.19
	Median :2012-01-01	Median :1367.1	Median :119.098	Median : 86.407	Median :240.99
	Mean :2012-01-01	Mean :1338.7	Mean :164.814	Mean : 89.259	Mean :272.66
	3rd Qu.:2014-12-31	3rd Qu.:1457.8	3rd Qu.:217.900	3rd Qu.:135.071	3rd Qu.:338.99
	Max. :2017-12-31	Max. :1709.6	Max. :826.278	Max. :241.580	Max. :851.56
	NA	NA	NA’s :1463	NA’s :2195	NA’s :2196

Data preprocessing

## [1] "2006-01-01" "2006-01-02" "2006-01-03" "2006-01-04" "2006-01-05"
## [6] "2006-01-06"

##         Date Consumption Wind Solar Wind.Solar year month day
## 1 2006-01-01    1069.184   NA    NA         NA 2006     1   1
## 2 2006-01-02    1380.521   NA    NA         NA 2006     1   2
## 3 2006-01-03    1442.533   NA    NA         NA 2006     1   3
## 4 2006-01-04    1457.217   NA    NA         NA 2006     1   4
## 5 2006-01-05    1477.131   NA    NA         NA 2006     1   5
## 6 2006-01-06    1403.427   NA    NA         NA 2006     1   6

Trends in Electricity Consumption

Yearly Trends

The plot below shows the yearly average electricity consumption in Germany. goal : identify trends in electricity consumption over the year

Observation

## [1] "Year with minimum daily consumption: 2009 Value: 842.395"

## [1] "Year with maximum daily consumption: 2010 Value: 1709.568"

Correlation

By calculating Correlation yealy between two parameter(consumption,year) show trends like in this we got Positive value that means by over the year cause increase consumption.

## [1] "Positive Correlation that means Positive trend"

Monthly Trends

Observation

The plot below shows the monthly average electricity consumption.

Correlation

Similarly we see negative trend here that’s conclude over the month we see decreasing order in electricity consumption.

## [1] -0.07394412

## [1] "Correlation value is -ve that means -ve trend low corelation"

Daily Trends

Observation

weekends	Consumption
Wednesday	maximum consumption
Sunday	minimum consumption

Corelation

As we see the output is negative that conclude the trend is negative , over the days the consumption is become low

## [1] -0.02135832

Hypothesis Testing

ANOVA:Yearly Consumption Differences

We tested if electricity consumption differs significantly across years.

##                Df    Sum Sq Mean Sq F value Pr(>F)    
## factor(year)   11   8303614  754874   29.43 <2e-16 ***
## Residuals    4371 112120696   25651                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Conclusion

The p-value is less than 0.05, indicating significant differances in consumption across years.

ANOVA: Seasonal Consumption Differences

We also tested if consumption differs across seasons.

##               Df    Sum Sq Mean Sq F value Pr(>F)    
## season         3  15227607 5075869   211.3 <2e-16 ***
## Residuals   4379 105196703   24023                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Conclusion

The p-value is less than 0.05, indicating significant differences in consumption across seasons.

Correlation Analysis

Correlation Matrix

The matrix below shows the correlation between consumption, wind, and solar production.

Forecasting

ARIMA Model

We fitted an ARIMA model to forecast future electricity consumption.

Accuracy

ARIMA Model Accuracy

	ME	RMSE	MAE	MPE	MAPE	MASE	ACF1
Training set	-0.0354426	83.39549	59.83591	-0.3915875	4.674948	0.4853672	-0.0344969

ETS Model

We also fitted an ETS model for comparison.

Accuracy

ETS Model Accuracy

	ME	RMSE	MAE	MPE	MAPE	MASE	ACF1
Training set	-0.0416266	57.8157	32.27235	-0.1462378	2.586819	0.6242534	0.0860811

Conclusion

Electricity consumption in Germany shows clear yearly, monthly, and daily trends.

There are significant differences in consumption across years and seasons.

The ETS model provided a better fit for forecasting compared to ARIMA.