Keywords

1 Introduction

In 2018, the electricity use in the residential sector represented 27% of total electricity consumption worldwide (IEA, 2020). Households use energy for various purposes, including space cooling, water heating, cooking, and lighting. In developing countries, households mainly use energy for cooking and heating services; solid biofuels and liquefied petroleum gas being the predominant fuels. Other services, including lighting, space cooling, and other plugging appliances, use electricity (Poznaka et al., 2015). The literature mentions various factors that influence electricity consumption behaviour in the residential sector. These factors include local weather, dwellings type, number of inhabitants, and income. In Senegal, the residential sector consumed 11.7% of the total final energy consumption of fossil fuels and 25% of the electricity produced in the country in 2018. The share of electricity demand by the residential sector is projected to increase by a factor of 3 by 2040 (IEA, 2019).

Seasons and geographic location are determinants of energy consumption. Ndiaye et al. (2017) found that electricity consumption in Dakar and Niamey is relatively low during the rainy season (June–September) compared to other periods of the year. Consumption peaks are reached during the period from March to May when temperatures are high, followed by a decrease in consumption during the rainy season, which leads to another consumption peak arising in October that is the hottest month of the year in the Sahel region, with high demand for space cooling. Conversely, space heating represents the biggest share of energy end-use, with more than 50% of domestic energy consumption in OECD countries. An analysis of electricity consumption and households’ behaviour in Swedish rented apartments showed that income is the main determinant of electricity consumption (Vassileva et al., 2012).

Besides economic conditions, the geographical location associated with households also affects their electricity demand. In China, Xinye Zheng et al. (2014) noted that urban dwellings consume more energy than dwellings located in rural areas. Daioglou et al. (2012) investigated energy consumption in both urban and rural areas of developing countries. They found that in both areas, cooking is predominant in end-use services, but other services that use electricity, such as space cooling and plugging appliances, have increasing shares.

Estiri (2014) identified four parameters that affect electricity consumption in the residential sector: weather, location, income, and plugging appliances. This study aims to investigate the weight of these parameters on electricity consumption in the residential sector, taking as examples the districts of Grand-Yoff and Grand-Dakar, which are two working-class areas located in the city of Dakar, Senegal.

2 Methodological Approach

2.1 Data Collection

The study methodology consists of an analysis of data collected within the framework of a study conducted by the Senegalese Agency in charge of energy savings (AEME). In 2016, AEME commissioned a survey in Grand-Dakar and Grand-Yoff in partnership with the German cooperation (GIZ). The survey sample was 120 dwellings located in these districts. The survey collected both quantitative and qualitative data in order to understand households’ behaviour related to electricity consumption.

The qualitative data of the survey were collected during interviews with heads of households and include information on the demographic profile of their residence, characteristics of plugging appliances in the household and the level of knowledge on best practices in improving energy efficiency.

The quantitative data of the survey were collected during campaigns to measure the electricity consumption of the three main plugging appliances: fridge, freezer, and television (TV), and to measure the household consumption profile per week. In addition, data such as the number of plugging appliances per category were collected. The survey also collected information on the electricity consumption of these dwellings recorded by the utility (Senelec) over a period of one year.

2.2 The Study Area

Grand-Yoff and Grand-Dakar are two of the 19 districts of the city of Dakar, Senegal. Figure 12.1 displays the position of the districts on the Dakar map.

Fig. 12.1
figure 1

Survey districts

Full size image

Grand-Yoff is a working-class district located in the municipality of Parcelles Assainies. According to the 2013 census, Grand-Yoff had a population of 185,503. The population was projected to reach 200,000 by 2020 (ANSD, 2016). Grand- Dakar is part of the municipality of the same name located in south-centre of Dakar. The population of Grand-Dakar was 47,012 in 2013 and was projected to reach 55,000 by 2020 (ANSD, 2016). The survey targeted dwellings in four sub-districts in Grand-Dakar and ten sub-districts in Grand-Yoff.

2.3 Data Analysis

This study focuses on the analysis of qualitative and quantitative data related to two themes: characteristics of plugging appliances and socio-economic parameters in household energy behaviour.

2.3.1 Characteristics of Plugging Appliances

The processing of data provided information on the number of appliances in surveyed households, their rated power, and usage time per day and per period. The first part of the study consists in analyzing the energy behaviour associated with each service considering the attributes of appliances. The electricity consumed in lighting is calculated based on the number and type of light bulbs in the dwelling. The same is done for each of the other services that include space cooling (fans and air conditioners), and other services provided by plugging appliances (TVs and fridges). The objective of the analysis is to design relevant measures that improve efficiency on energy behaviour, which could target either appliances’ rated power or usage per day. A similar approach has been applied in the analysis of levers of efficiency in the electricity consumption of the residential sector in Sweden (Vassileva et al., 2012 and in China (Xinye Zheng et al., 2014).

2.3.2 Socio-Economic Parameters of Energy Behaviour

The parameters classified as social and/or economic in the questionnaire collected information on the determinants of the household’s energy behaviour. For instance, the level of income in the household may explain the number and rated power of appliances used in supplying the services studied. The data was first compiled in Excel, and the analysis used statistical tests to identify the distribution pattern, whether it is normal or not. Regression methods were used to study relations between the total electricity consumption per year and each of the parameters hypothesized as determinants of electricity consumption. Similar method has been used to analyze determinants of energy consumption in the United States (Estiri, 2014), and in China (Murata et al., 2008).

3 Results and Discussion

3.1 Characterization of the Plugging Appliances

3.1.1 Cold Appliances

Cold appliances include fridges that represent an average of 60% of the household electricity consumption, followed by freezers (37%), and minibars (3%). The majority of households in both districts are equipped with one cold appliance.

The preferred location for fridges and freezers is the kitchen; but we observed that in 20% of households, these appliances are placed in open areas such as hallways or balconies, which can affect their energy performance. Twenty-two per cent (22%) of cold appliances have an energy-saving Label, which is usually the energy performance rate based on the European rating system from where these appliances are exported. Thirty-six per cent (36%) of appliances have a plate indicating the climate class; out of this, 45% are of the tropical class that includes tropical T-class, subtropical ST class, and extended subtropical SN-T class. The remaining 55% are class N and SN (Fig. 12.2).

Fig. 12.2
figure 2

Number of cold equipment per dwelling

Full size image

3.1.2 TV Sets

Television (TV) appliances are mainly placed in closed areas such as the living room and the bedroom. Fifty-six per cent (56%) of TVs are of the type Small and medium-sized Cathode Tube (32 inches), and 21% are of the type Plasma large screen size. For Cathode tube TV, the power measured in stabilized operating mode is 73 W; in sleep mode, this power is 7 W, making it the most energy-intensive TV on standby mode. Next is the Plasma technology TV with 4 W in standby mode and 96 W in stabilized operating mode. The LED TVs represent 14% of TV appliances inventoried during the survey, and LCD TVs represent 9% of these appliances. The LCD TV has a power measured in a stabilized operating mode equivalent to 109 W. This relatively high value is due to the brightness adjustment system of this type of screen. The majority of households surveyed have one TV set (Fig. 12.3).

Fig. 12.3
figure 3

Number of TV sets per dwelling

Full size image

3.1.3 Space Cooling Appliances

The measurement campaign took place from December to February, when temperatures are below 25 degrees Celsius in Dakar. Among the 120 dwellings surveyed, only one had a window-type air conditioner of class CV2. Fans and air conditioners were not used in any of surveyed dwellings, which prevented reliable measurements in normal operating mode. The majority of dwellings surveyed in Grand-Dakar have one fan: The situation is more contrasted in Grand-Yoff as visible in Fig. 12.4.

Fig. 12.4
figure 4

Number of fans per dwelling

Full size image

3.1.4 Lighting Appliances

The majority of dwellings surveyed have four or five lamps. Since January 2018, only energy-saving lamps that include compact fluorescent lights (CFL), halogen, and light-emitting diode (LED) lamps are authorized for sale in Senegal. The pre-legislation campaign conducted by the utility should have normally removed incandescent lamps by the time of the survey. However, 60 W and 40 W incandescent lamps were found in households, and they represent 12% of the lamps inventoried in Grand-Yoff and 25% of those inventoried in Grand-Dakar. Incandescent bulbs consume more electricity and have a relatively shorter lifespan but cost less.

3.2 Analysis of Electricity Consumption

3.2.1 Total Electricity Consumption Per Annum

The survey also recorded measurements of daily energy consumption from existing metres. The data collected were extrapolated in models to calculate the theoretical electricity consumption per year using the formula below:

Equation 12.1: Theoretical electricity consumption of dwellings

$$ \frac{Cons\ \mathrm{weekdays}\times 5+ Cons\ week\mathrm{end}}{7}\times 30\times 12 $$

The theoretical electricity consumption calculated refine the data obtained from direct measurements in dwellings and project consumptions per month and per annum in the dwellings. The electricity consumption data from SENELEC invoices are labelled as the real electricity consumption of the dwelling during the same period. As shown in Fig. 12.5, the difference between real electricity consumption and theoretical estimation of electricity consumption is statistically significant.

Fig. 12.5
figure 5

Real and theoretical electricity consumption per annum

Full size image

For instance, the invoice of the dwelling coded as number 52 in the survey documents shows a real electricity consumption equivalent to 5524.13 kWh per annum, which is the highest of surveyed dwellings, but the estimation is based on direct measurements return a total of 3960 kWh per annum. Considering all dwellings surveyed, the real electricity consumption represented on average 65% of the theoretical consumption. This demonstrates that planning electricity supply systems based on extrapolation of direct measurements can cause biases and results on sub-optimal systems with excess capacity, which is frequent in developing countries (Urban et al., 2007).

3.2.2 Electricity Consumption Per Capita

Per capita electricity consumption in surveyed dwellings varies between 33 kWh and 706 kWh per annum. Figure 12.6 shows variations in electricity consumption per capita in surveyed dwellings. Different parameters can explain this large difference in the electricity consumption of individuals targeted by the survey. The number of persons living in a dwelling is positively correlated with the per capita consumption (R = 0.34). However, the correlation is weak, suggesting at most an influence on per capita electricity consumption.

Fig. 12.6
figure 6

Electricity consumption per capita

Full size image

3.2.3 Impact of Social Parameters in Electricity Consumption

In this section, we look at the impact of four parameters on the real electricity consumption of surveyed dwellings. The number of residents in the dwellings, the number of residents in active employment, the level of income, and the number of appliances in the dwelling. Figure 12.7 displays the shape of plotted data.

Fig. 12.7
figure 7

Correlation between electricity consumption and social parameters

Full size image

The correlation coefficient of the parameters number of residents per dwelling and annual electricity consumption is R = 0.34. The correlation is weak, suggesting that the number of residents per dwelling does not significantly affect the annual electricity consumption.

The correlation coefficient of the parameters number of residents in active employment and annual electricity consumption is R = 0.25. The correlation is weak, suggesting that active employment of the dwelling’s residents, and therefore absence during the daytime, does not significantly affect the annual electricity consumption.

More surprisingly, the correlation coefficient of the parameters number of plugging appliances in the dwelling and annual electricity consumption is R = 0.25. A higher correlation factor could have been hypothesized, with the annual electricity consumption per dwelling increasing with the number of plugging appliances in the dwelling. However, several other parameters affect the electricity consumption of these appliances, including their rated power and usage time per day.

The correlation coefficient of the parameters income of the dwelling and annual electricity consumption is the lowest of all factors at R = 0.01. This weak correlation suggests that electricity consumption in surveyed districts has more to do with social behaviours not identified in this study than economic purchasing power.

4 Conclusion

This chapter completed an in-depth analysis of the determinants of electricity consumption in two working-class districts of Dakar. The findings on statistical relationships between electricity consumption in these districts, with parameters that include the number of residents and the number of plugging appliances per dwelling, provide necessary information for electricity planning models. Understanding the way households consume electricity supports the design of optimal energy solutions in cities.

Findings from the study demonstrate that information on the number, rated power and usage time of plugging appliances is not enough to project the electricity demand of the dwelling. The study did not identify the parameters that have a statistically significant impact on dwellings’ electricity consumption. Still, the weak correlation between this consumption and the number of plugging appliances, as well as the difference between real electricity consumption recorded by the utility and estimates based on extrapolation of periodic measurements, suggest other parameters play out in this consumption. Thus, it would be necessary to conduct advanced studies. The study provided results that are important to consider in planning future energy systems for working-class districts of cities in developing countries, which expand rapidly, driven by accelerated urbanization.

Firstly, planning energy systems at the level of city districts such as community grids through the extrapolation of data collected from a periodic survey may cause biases that would affect the size of the system.

Secondly, past invoices are not enough to project the future electricity demand of a district. Therefore, it is necessary to improve planning models with parameters different from those hypothesized in this study.