Utilities are under real pressure right now. Demand is climbing fast, with data centers and other new large loads driving visible — and at times controversial — demand, right as worsening extreme weather causes further strain.
But building new avenues for generation or transmission takes years and comes with high costs, permitting challenges, and community pushback. So in some contexts, grid decision-makers under strain are turning to what’s already there. That means bringing old assets back online.
That is where the return of so-called “zombie” power plants comes in.
Facilities that were slated to be retired years ago are being restarted to close the gaps, and in the case of coal plants especially, that comes with an emissions and economic toll. Last fall, at least 33 fossil fuel generating units across 15 power plants had their planned retirements postponed due to reliability concerns associated with energy demand from data centers. These decisions typically do not come from one party alone. They often involve the asset owner — whether an investor-owned utility or independent power producer — working with a grid operator and regulators to determine whether keeping or bringing a plant back is necessary.
At the same time, though, research suggests that the grid is underutilized, and may have capacity to spare. This raises a bigger question: If we are restarting aging infrastructure while parts of the grid sit underused, what are we missing about how the system can optimally operate?
Restarting plants doesn’t fix reliability
Bringing an old plant back online can help cover a generation shortfall, but comes with tradeoffs.
These assets were often retired because they were inefficient, dirty, or difficult to maintain. When they return, those same issues follow. Maintenance cycles become more frequent, parts are harder to source, and documentation is often outdated, leaving generating reliability teams to bring assets back into the portfolio from a time well before distributed energy resources were as widespread.
When the information is inaccurate or incomplete, even slightly, it starts to show. Maintenance gets missed, response times slip, and avoidable outages occur — exactly what reviving the “zombie” plants was designed to avoid.
The problem of aging assets
Demand today does not behave the way it did even a decade ago. Heatwaves push load far beyond what a region has historically planned for — think 106 degrees Fahrenheit in Phoenix, Arizona, in March — or else a new data center shifts consumption patterns.
At the same time, power is coming from more places across the network, and it doesn’t behave the same way every day. A hot, sunny day increases demand and could reduce capacity available in transmission throughput — but it also increases what may be generated through solar sources. While demand can swing, so can production in certain areas; transmission has become a balancing act for utility providers.
But a lot of our older infrastructure wasn’t built for that kind of dynamic balancing. When legacy plants come back online, they complicate things further.
Peaker plants are designed to fire up to meet cyclical demand during a day, usually paired with predictable temperature and human rhythm shifts. But today, certain places are using them to help keep generating costs steady — by, for instance, operating a small coal plant located on the side of a transmission bottleneck where higher demand has nudged rates higher.
Meanwhile, some utilities are getting more out of their corner of the grid by understanding it better with real time condition metrics. Utilities using data on things like ambient temperature, wind, and line temperature can better decide how much energy a line can safely carry; those without those insights, however, must operate at a lower throughput, and potentially leave capacity on the table.
Underutilization with limited visibility
Research from The Brattle Group says the U.S. grid is significantly underutilized, operating at roughly half capacity at any given time and that better utilization could save customers between $110 billion and $170 billion over the next decade. This is a tremendous opportunity, especially at a moment when new capacity is being added at significant cost.
However, it’s not just a matter of flipping a switch and using the grid more heavily. Underutilization does not happen completely on purpose. It reflects how difficult it is to see and manage the system in real time.
Without clear insight into how assets are performing and how demand is evolving, utilities cannot safely push systems closer to full utilization. Improving that visibility changes the equation. When facility reliability teams trust the data and understand it, they plan differently. They begin targeting maintenance most strategically, assets stay online longer and returns for inputs increase. This reflects what is needed to meet the demand of the future: an “all of the above” approach that includes maximizing existing assets while expanding generating capacity.
This approach doesn’t obviate the need for new infrastructure, but it will take pressure off the system. The potential returns from improved utilization represent about 10% of the $1.3 trillion that utilities have requested to spend through 2030 — a clear and absolute win for everyone involved.
Looking ahead
There isn’t one fix for the demand the industry is seeing. Meeting it successfully will come from a mix of new generation, more transmission capacity, and better use of renewables. Restarting older assets will likely continue to play a role, especially in the near term, but it can’t be a long-term strategy.
The best plans start with the clearest picture of the energy system as it is. That means better visibility into assets, as well as better flow of information across teams. Those who grapple with this challenging moment most effectively will be the ones that use what is already in place most efficiently
Jason Kraus is a principal solution consultant at Accruent’s asset solutions group, where he focuses on improving facilities, equipment, and processes. He has worked in the enterprise IT space for 15 years. The opinions represented in this contributed article are solely those of the author, and do not reflect the views of Latitude Media or any of its staff.
