Will Occasional Blackouts Become the Norm?
An analysis of California’s most recent rolling blackouts and how to prevent them
When California passed the Global Warming Solutions Act (AB 32) in 2006, it was one of the most ambitious pieces of climate change legislation ever created. Ten years later, California further solidified its leadership role with a commitment to reducing greenhouse gas emissions to pre-1990 levels by 2030. At a time when it already boasted half of the country’s total installed solar PV capacity, the state offered new incentives to encourage even more residential and commercial investment in renewable technologies.
California was always supposed to provide a shining example for other states to follow when designing and executing their own climate change policies. However, recent events suggest that this supposedly winning template for “greening the grid” is deeply flawed. In mid-August 2020, California experienced record-setting heat waves that led to a series of rolling blackouts across the state.
But how could rapid decarbonization have sparked outages like this? After all, renewables like solar and wind were supposed to make the grid more reliable – not less.
The Inherent Limitations of Greening the Grid
In October 2020, the California Independent System Operator (CAISO), California Public Utilities Commission (CPUC), and California Energy Commission (CAE) issued a detailed analysis explaining why the rotating outages had occurred – and what steps the state could do to address this issue moving forward.
As one would expect, extreme weather conditions certainly played a role, with the utilities experiencing unprecedented demand as AC usage skyrocketed across the state. Another major factor was grid operators’ inability to accurately forecast supply and demand when dealing with intermittent energy sources like solar and wind. In addition, those same operators used outdated guidelines originally created for America’s legacy grid. More specifically, California utilities are required to initiate rolling blackouts if reserve generation ever falls below 6% of current demand. This threshold might have made sense when dealing with centralized power generation. But it’s obsolete when managing tens of thousands of distributed PV systems and wind turbines – many of which are outside the utility’s direct control.
Even safeguards of last resort weren’t enough to prevent these rolling blackouts. For example, the report outlines how California was unable to buy electricity from neighboring states – many of which were dealing with their own supply shortages due to inaccurate forecasting during the heatwave. Throughout the entire Western portion of the country, in fact, demand for electricity far outstripped what utility operators had anticipated – leading to regional rolling blackouts.
Why These Grid Outages Really Happened
CAISO’s 121-page report paints an accurate picture of the factors at play. However, the underlying causes of these outages run much deeper than outdated rules or excessive air conditioning usage.
It comes down to outdated thinking.
Human decision-makers can manage limited data sets. But when those data sets grow exponentially, our brains just aren’t up to the task. And given the sheer number of variables that operators and utilities must now factor into their decision, managing today’s increasingly diverse energy grid is beyond the scope of human actors.
Peak demand is a prime example. It is normally midday – precisely when temperatures were highest and businesses were the most active. But as the record-setting heat wave continued into the “off-peak” evening hours, PV panels couldn’t keep up after the sun went down. Unprecedented demand for air-conditioning in the evening created shortfalls in electricity distribution. And this wasn’t something utility operators saw coming.
As the report points out,
In transitioning to a reliable, clean, and affordable resource mix, resource planning targets have not kept pace to lead to sufficient resources that can be relied upon to meet demand in the early evening hours. This makes balancing demand and supply more challenging. These challenges were amplified by the extreme heat storm.
Sometimes, supply disruptions can stem from something as simple as a passing cloud obscuring some panels. Other times, mismatches between supply and demand result from poor forecasting based on inaccurate market signals. CAISO’s analysis explains that, “Some practices in the day-ahead energy market exacerbated the supply challenges under highly stressed conditions.”
Whatever the cause, the end result is the same – i.e. unreliable electricity delivery that leaves hundreds of thousands of residents literally in the dark.
But what if it didn’t have to be this way?
Smart Grid Optimization – Using Real-Time, Predictive Artificial Intelligence
With more data to process, making accurate decisions in the present moment is difficult. And future predictions are impossible – at least when using human processing power.
However, managing big data is precisely where AI excels. Veritone’s energy solutions, for example, use artificial intelligence (AI), machine learning, and Hamiltonian models to accurately predict where electricity demand, supply and pricing will be weeks, days and just minutes into the future to determine optimal economic energy generation, storage, and dispatch to the grid.. More specifically, this suite of AI-powered solutions can:
- Forecast where demand and rates are heading by analyzing historical and current weather conditions, production levels, industry regulations, and utility rates
- Use this forecast data to intelligently control hardware devices and deliver a highly-optimized supply mix across all the distributed energy resources within that utility network
- Synchronize edge devices to increase grid reliability, smooth out solar supply, and stabilize the voltage. This control exists across the entire energy mix of renewables, fossil fuel, and batteries. And it extends down to autonomous control at the device level.
Our AI energy solutions do all this in real-time without any direct intervention. And through machine learning, it continuously incorporates new data and environmental parameters into its models to create even more accurate forecasting and optimization moving forward.
Making Sustainability More Sustainable
The challenges outlined above aren’t new. Just 2 summers ago, for example, PG&E filed for bankruptcy in the wake of record-setting heatwaves, wildfires, and rolling blackouts. However, these challenges are becoming more complex as rapid decarbonization and extreme weather conditions continue to shape America’s energy landscape. That the past 5 years were also the hottest in recorded history suggests this trend will only accelerate moving forward.
Using AI to intelligently synchronize and optimize the grid can help reverse this trend and move us closer to a truly sustainable energy ecosystem that delivers clean and reliable electricity to everyone.