HTTPS SSH

Urban Heat Challenge

How might we better use this data to communicate the urban heat challenges of Parramatta, and how might we improve on this in Parramatta LGA during summer?

Eligibility: Use the Parramatta Temperature Dataset

Our visualisations predict the amount of power required by air conditioners in the local goverment area of Parramatta.

We used the Parramatta temperature data set to determine how the maximum temperature changed everyday and combined this with the energy cost of running the air conditioner at 35C to calculate the energy cost when running at the maximum temperature per day. We then used the census data to determine how many households there are in each of suburbs where the data was reported from to calculate total energy expenditure by air conditioners per hour for all suburbs.

Another dataset that was used showed the amount of hours spent by each household per year either cooling or heating the house. Since the data we received was from January, we only looked at the cooling hours to determine how long a household would spend cooling their house per day. This led to much better energy prediction results.

One use case of this data would be to determine how much extra power would need to be inserted into the grid on days where the temperature is very high.

To determine how much extra power would be required to be inserted, the energy company could calculate the amount of power used on a day where air conditioners would not be used as much (on a relatively cool day) and then use our data to predict how much power exactly they would need to supply on a hot day. Then they would be able to predict how much more power they would need to inject in the grid on hot days.

Another usecase would be for power companies to reduce the price per unit of power supplied when it is in demand which would convince more people to leave the air conditioning on for longer. This could potentially lead to better revenues.

If we were to develop this product in the future, we would look at data with more specific daily results such as average amount of hours spent running the air conditioner per day rather than per year. This would allow us to cater our results to different parts of the year where temperatures can vary a lot. We would also combine data from the Bureau of Meteorology to expand our predictions to various locations around the country, especially rural areas as they would be unlikely to have air conditioners.