Chicago, United States — photo by @alexlivi

The Water-Electricity Nexus & AI

Changing Climate Predictions With Machine Learning Correlating Water and Energy

A recent journal article from Purdue University argues that most modelling assumes water and electricity is used separately, in silos. While they do the modelling and estimation of the impact with these two factors in relation to each other.

“The model uses artificial intelligence to make predictions on climate change impact. As a learning algorithm, the model is fed years of data from a region’s utilities and weather services and then trained to predict changes in electricity and water use given certain climate-change scenarios.”

What they arrive at specifically with the regional-scale model is leveraged to make city-level projections is the indication that there will be a 10–20% (2–5%) increase in electricity (water) use across the analysed cities due to a warming climate.

Apparently water has become so costly in Chicago that some citizens cannot afford it.

Imagine thinking about turning on the tap, but realising you cannot afford it.

What do you do when you have no water?

In February 2019 NPR wrote an article detailing how water was cut off from marginalised citizens who could not afford to pay.

“Campbell is partially blind and lives on a fixed income from disability payments. She dedicates most of her time to helping her community. Her church includes a resource center that provides food and shelter for poor and homeless people. She couldn’t pay off her water debt, and in August her water was turned off. The Chicago water department offered her a payment plan but required a more than $1,700 deposit before restoring her water. She didn’t have it.”

In certain areas in the US the cost of water has been increasing.

“The APM Reports investigation found that the rising cost of water has hit poor families the hardest; the government-run water utilities in these six cities have issued at least 367,740 shut-off notices in the past decade. And an analysis of shut-off data revealed disproportionately high concentrations of water shut-offs in poorer areas and in majority black and Latino neighborhoods in every city.”

If there is already an issue with water being expensive and usage is expected to increase.

What the authors of the journal mentioned previously says is:

  1. The inclusion of additional climate variables beyond the baseline provides a significant improvement in predictive accuracy
  2. The climate-sensitive portions of summer electricity and water use are expected to increase in the region by 19% and 7%, respectively.

In an interview in Science Daily on the 5th of March one of the authors say:

“Such scenarios are fundamental for understanding the joint response of electricity and water uses to future changes in climatic conditions as to understand to what degree our current management and technological strategies need to adapt to the future changes.”

If we have more accurate models perhaps we can operate this infrastructure better at least to some degree when understanding maintenance, demand and how to deliver good services.

This is #500daysofAI and you are reading article 280. I am writing one new article about or related to artificial intelligence every day for 500 days. My current focus for 100 days 200–300 is national and international strategies for artificial intelligence. I have decided to spend the last 25 days of my AI strategy writing to focus on the climate crisis.

AI Policy and Ethics at Student at University of Copenhagen MSc in Social Data Science. All views are my own.

AI Policy and Ethics at Student at University of Copenhagen MSc in Social Data Science. All views are my own.