California must do more and faster to prevent future blackouts as it transforms its energy generation system to combat climate change, the state’s governor wrote this week in a letter to energy regulators, Reuters reported.
Governor Gavin Newsom requested a post-event analysis and re-examination of reliability assumptions after a heatwave and record electricity consumption forced grid operators to introduce rolling blackouts to manage demand.
Electric customers expect the service to be reliable and affordable as well as avoid air pollution and climate change.
But recent events show California’s power grid failed the reliability test, the top priority for suppliers and grid operators given electricity’s importance to the modern economy.
Reliability is often taken for granted until supplies fail, and rolling blackouts, even if brief and rare events, signal the grid is under stress and that the reserve margins needed for reliable operation are no longer adequate.
Consider them the “canary in the mine”, which should always prompt a thorough re-evaluation of operating practices and policies before reliability deteriorates further.
Peak load
Reliable electric systems must have sufficient generating capacity to meet wide variations in load over seconds, minutes, hours, days, seasons and even years.
In the very short term, reliability is maintained by fast-responding generating units connected to the grid, while users can be paid to curtail consumption under demand-response programmes.
Reliability can also be boosted by linking neighbouring grids to form a super-grid. California’s is linked to neighbouring states through the critical Western Interconnection, allowing imbalances between generation and load to be evened out over a much larger area.
Over longer periods, whether days or years, reliability is maintained by having spare capacity capable of meeting any conceivable level of peak load.
Since the turn of the century, California’s highest loads have all occurred between 3.00 pm and 5.30 pm, between July and September, when air-conditioning demand is most intense.
During the highest annual peaks, in July 2006 and September 2017, the grid served over 50 Gigawatts of load, more than double the year-round average of 24 GW.
Spare capacity must be able to cover load even if a major generation asset, such as a large power station or a high-capacity transmission line, suddenly fails.
Load curve
Maintaining reliability at different times of year and under a mix of weather conditions presents a formidable challenge.
Grid controllers forecast demand on a daily, seasonal and annual basis, carrying out dynamic scenario-planning to ensure there is adequate generation available, even if they lose the largest grid-connected asset.
A load duration curve shows the number of hours each year in which load exceeds a given level.
In 2019, California’s Independent System Operator (CAISO), which operates the network across most of the state, had to supply hourly loads ranging from just 17.0 GW to as much as 44.1 GW.
Load-served was below 24.2 GW for 50% of the time, and below 40.5 GW 99% of the time, according to data from CAISO’s Energy Management System.
But for 12 hours last year, CAISO had to supply more than 43.0 GW, and at one point more than 44.0 GW.
Even that understates the pressure variable load imposes on the system, because the summer of 2019 was relatively cool and peak load was therefore relatively low compared with other years.
Since 2000, annual loads have peaked around 45.8 GW on average, and in two years have climbed to more than 50.0 GW.
Flexibility
Like other grid operators, CAISO needs a lot of flexibility.
Spare generation capacity, grid-connected batteries, imports from neighbouring states and voluntary curtailment can all help manage daily and seasonal peaks.
Gas-fired power plants are ideally suited for peak-management because they can ramp up and ramp down their output rapidly over a time-scale ranging from a few minutes to a couple of hours.
Wind and solar generators are more problematic because their output is determined by external factors (time of day, season and weather conditions) and they are not responsive to the needs of the grid.
The increasing proportion of wind and solar generation has therefore made the grid less flexible and reduced reliability, especially in the summer when it is already stretched by peak air-conditioning loads.
Transition
As wind and solar account for a greater share of in-state generation, CAISO and other state regulators have sought to build in other sources of flexibility.
California’s reliability plan, heavily dependent on importing power from neighbouring states when needed, implicitly assumes those states have excess capacity of their own.
In recent days, however, the high pressure weather system that produced a heatwave, cut wind power and drove up air-conditioning loads in California created the same conditions in adjacent states, limiting the ability to import electricity.
In the long term, there are hopes electric vehicles and more widespread time-of-use pricing can make electricity demand from residential and small business customers more responsive to the grid.
The ultimate aim is to flatten the load curve, increasing consumption at times of the day and year when loads are currently low, and reducing consumption when loads are currently high.
Flattening the curve would make it much easier and cheaper for generators to supply the necessary power without carrying so much spare capacity, which is expensive.
But California is still very far from that point. Flexibility has been reduced by wind and solar faster than it has been replaced by other measures, resulting in a predictable deterioration in reliability.
Recent blackouts are a warning that the state’s current mix of electricity policies is not reliable and an opportunity to adjust them before worse outages occur in future.
