Analyzing Power Consumption Trends And Drivers In Solar Cities Dataset

The Solar Cities Project

The concerned research report has been accomplished by the University of Ballarat based on the data of that of the “Solar City Project” and for the purpose of analysis the data has been collected from the regions- Grampians and Loddon Mallee, from both households as well as business sectors. The main objective of the concerned research is to observed the changes and dynamics arising out of energy usage (Zakir, Seymour & Berg, 2015). The study emphasises on the different variables which influence the usage of power by the households in various Australian suburbs and for the purpose of the same, “IBM Watson Analytic” tool has been used (High, 2012). The different factors studied have been categorised into different sets of features and the analysis and visualisation of the data of solar energy technology are categorised into the groups namely- “Solar energy technology adoption”, “Geographical characteristics” and “Dwellings’ physical characteristics”.

On the other hand, the physical characteristics of the dwellings consists of the aspects like:

• Bedroom numbers
• Dwelling size
• Dwelling stories
• Age of the house
• Roof colour
• Material used for construction
• Suburbs
• Number and types of light
• Housing insulation and others

The research aims to explore the drivers of power consumption in these dwellings, keeping into consideration the huge amount of carbon dioxide emission which takes place due to the usage of the energy produced from the conventional energy resources like petroleum, coal and others, which in turn has hugely negative impacts on the environment (Chen, Chiang and Storey, 2012).

Answer 1: Contribution of usage of power over year by the colour of the roof 

 

Figure 1: Power usage by roof colour

As is evident from the above figure, the light-coloured roofs consume much lesser power compared to intermediate coloured ones, which again use less power than dark coloured ones, for 2012, 2013as well as 2015. However, in 2014, intermediate coloured roofs consume more colour than dark coloured ones.

 

Figure 2: Power usage by PV_Capacity over a year

As can be seen from the above figure, higher energy can be seen to be consumed by lower Photo Voltaic Capacity (0-960) while comparatively lower energy can be seen to be consumed by higher Photo Voltaic Capacity (>960). As in the case of 2014, it can be seen that highest power usage is done by Photo Voltaic Capacity of 0-960.

 

Figure 3: Power usage contribution by Insulation and PV_Capacity

In 2014, the highest usage of power can be seen in the case of lowest Photo Voltaic Capacity (0-960) and insulation equal to 1. In both 2013 and 2015, with Photo Voltaic Capacity 0-960 and insulation 1, the power usage can be seen to be following highest.

Data Set Features

 

Figure 4: Usage of power by estimated age

The usage of power can be seen to e maximum for those dwellings with estimated age of 60 years and above and is the lowest for the dwellings with estimated age of 0 years to 4 years.

 

Figure 5: Highest power usage over months

For all the years in the interval 2012 to 2015, the highest usage of power can be seen in the month of October. Highest power usage can be seen to be used in the month of July with usage almost 160 K.

 

Figure 6: Minimum power used by months

The month of November can be seen to be the one with least power usage when four respective years are taken into account.

 

Figure 7: Top drivers contributing to the power usage

The factors, “Suburb” and “Wall Construction” types seen to be the top drivers for power usage, explaining 28% of the total usage. Of these, factors the “Suburb” type is the most important single driver of energy consumption. It is to be noted that in this case, the newly generated variable “Light _Count” is ignored (Kambatla et al., 2014).

 

Figure 8: Suburb having the most number of PV_Capacity inclusive dwellings

In this case, the maximum number of houses having Photo Voltaic Capacity can be seen to be found in the “Heywood” suburb.

 

Figure 9: Age of houses more likely of having PV_Capacity

When the Photo Voltaic Capacity is 4800, the estimated age of the house most likely to have the Photo Voltaic Capacity is that of the housings of 20 years to 29 years.

 

Figure 10: House types consuming less power

As is evident from the above figure, the “Owned” dwellings can be seen to be consuming more power than that of the “Rented” ones.

 

Figure 11: Power usage by suburbs

The suburb named, “Portland” situated in Victoria can be seen to be consuming or using maximum power which is followed by that of the suburb named “Heywood”.

 

Figure 12: Power usage by square meterage and number of storeys of dwellings

The single storeyed dwellings can be seen to be using more power than the double storeyed ones and for both of these types of dwellings the power usage can be seen to be greater for rooms with area of 199 square meters, as is evident from the above figure.

Data Analysis Tasks

 

Figure 13: Power consumption by types of lights in the dwellings

Among the different types of lights which are used in the houses, the highest power consumption can be seen to be by the “Halogen” type of light which can also be seen to be accompanied by “Incandescent” types of lights in the dwellings.

 

Figure 14: More lights of any type and relation with more power

More number of lights does not imply more usage of power for any kind of light. Highest power usage can be seen for the dwellings with 17 lights (usage being nearly 140K) whereas houses having 43 lights can be seen to be using much lower (<40K) of power consuming.

 

Figure 15: Age of houses and type of construction of wall

The highest estimated age of the dwellings can be found as “Sixty to Sixty-nine”. The estimated house age can be seen to be maximum in case of those whose wall is made of brick.

 

Figure 16: Power usage by house age, area and the number of bedrooms

Power usage can be found to be highest in Portland, in those with number of bedrooms is 3 and especially for those dwellings with age ranging from 50-59 years (Najafabadi et al., 2015).

 

Figure 17: Energy consumption by roof colour and material

As is evident from the above figure, the consumption of energy of the dwellings can vary across the types of roof colour and roof material. For light coloured roof the consumption is found to be minimum compared to those with intermediate roof colours and brick walls. Maximum power usage can be seen to be used by those with dark coloured roof and concrete brick constructions.

 

Figure 18: Power consumption by dwellings with double glazed windows with window coverings

Dwellings with double glazed windows with window coverings (Blinds and curtains) can be seen to be using less power than those with single glazed or tented ones with coverings like curtains or blinds. The tented windows with curtains use most power.

The two concerned research questions can be analysed with the help of the following Dashboard Representation:

Research Question 1 

Which are the combinations of the features which highlight the areas where the efficiencies could be found in the aspect of reduction of power usage:

 

Figure 19: Dashboard Presentation 1

From the above dashboard visualisation, it can be seen that the single storeyed dwellings consume low power than the double storeyed ones, however, the usage of power can be seen to be highest for the rooms of 199 square metre in both the types of houses. Again, for the houses with dark coloured roofs and brick walls the consumption of power can be seen to be seen to be consuming highest power. followed by those with brick walls and intermediate coloured roofs. The least usage of power can be seen by the ones with light coloured roofs. Dwellings with double glazed and covered windows can also be seen to be consuming less amount of power as compared to the ones with single glazed or tainted windows with coverings (Turban et al., 2013). When measured across the different Australian suburbs, Portland can be seen to be consuming highest amount of energy, and the power consumption can also be seen to be highest for those dwellings with bedroom number 3 and those with the estimated age of thirty to thirty-nine.

BI Reporting and Dashboards

Inclusion of components in a predictive model that can explain the demand of future energy use and emission of carbon dioxide:

 

Figure 20: Dashboard Presentation 2

As can be seen from the above figure, the aged dwellings (with age 60 and above) consume much more power than those with age 0 to 4 years, which in turn consume less power. The dwellings with less Photo Voltaic Capacity can also be seen to emit more carbon dioxide and vice-versa. The double drivers which significantly influence the level of power consumption are found to be the suburb type and the wall material types, with the suburb type being the single most significant driver of the same (Minelli, Chambers & Dhiraj, 2012). Most energy consumption and carbon emission take place in Portland and Heywood suburbs. Top emission of Carbon Dioxide takes place depending upon wall material and suburb types.  

Research Report 

As can be seen from the above section of the concerned report, the Dashboards act as one of the most important and effective tools for any kind of business management, which in turn makes it more preferable and a successful real time pointer which helps the businesses in the aspects of rational and knowledgeable decisions, thereby helping in the achievement of long term objectives. In this method, the collection, organization, grouping and visualisation of the data and its visualisation becomes easy and comprehensive (West, 2012).

In this context, the first research question can be seen to be analysed with the help of Dashboard Representation 1, which in turn indicates towards the fact that construction of single storeyed buildings and those with light coloured roofs can help in reducing the consumption of power. Also installing double glazed and covered windows in the dwellings can be seen to be helping in reduction of power consumption. As power consumption can be seen to be highest for those with 3 bedrooms, avoiding the same can help in the reduction of the same.

On the other hand, with the help of dashboard visualisation, the second research aspect shows that the it is better to dwell in newly made dwellings for reducing power consumption. Presence of high PV_Capacity can also reduce energy consumption. Choosing non-brick materials and settling in suburbs other than Portland and Heywood can also help in reduction in power consumption, thereby reducing carbon dioxide emission.

From the above discussion, the following recommendations can be made for reducing energy consumption and CO2 emissions:

• Usage of light coloured roofs
• Installation of higher PV_Capacity
• Reconstruction of new dwellings from old ones
• Less use of single glazed windows
• Less usage of non-brick materials
• Avoiding to dwell in Portland and Heywood

Each of the members of the team and their individual contribution can be seen to facilitate the efficient interpretation of the real scenario exiting in the concerned aspect. The usage of “Dashboard Presentation” can be seen to be effective in depicting the causes of higher emission of CO2 and also in the aspects of finding out the feasible solutions to combat the same. The energy consumption in the year can also be seen to be forecasted in the concerned study along with the probable steps which can be taken by the managers of the project as well as the stakeholders involved in the concerned project.

References 

Chen, H., Chiang, R.H. and Storey, V.C., 2012. Business intelligence and analytics: from big data to big impact. MIS quarterly, pp.1165-1188.

High, R. (2012). The era of cognitive systems: An inside look at IBM Watson and how it works. IBM Corporation, Redbooks.

Kambatla, K., Kollias, G., Kumar, V., & Grama, A. (2014). Trends in big data analytics. Journal of Parallel and Distributed Computing, 74(7), 2561-2573.

Minelli, M., Chambers, M., & Dhiraj, A. (2012). Big data, big analytics: emerging business intelligence and analytic trends for today’s businesses. John Wiley & Sons.

Najafabadi, M. M., Villanustre, F., Khoshgoftaar, T. M., Seliya, N., Wald, R., & Muharemagic, E. (2015). Deep learning applications and challenges in big data analytics. Journal of Big Data, 2(1), 1.

Turban, E., King, D., Sharda, R., & Delen, D. (2013). Business intelligence: a managerial perspective on analytics. Prentice Hall, New York.

West, D. M. (2012). Big data for education: Data mining, data analytics, and web dashboards. Governance Studies at Brookings, 4, 1-0.

Zakir, J., Seymour, T., & Berg, K. (2015). BIG DATA ANALYTICS. Issues in Information Systems, 16(2).