Richard Boehnke is a consultant.
Our electricity grid is a vast balancing act between supply and demand, with a tightly-guarded and centralized structure created in the early 20th century to manage electricity production, transmission and delivery. This system is dominant not through persistent innovation or transparency, but by maintaining legislated monopoly control, with guaranteed revenue from ratepayers, and perpetuating the vision that excess generation capacity must be built in order for utilities to fund capital investments.
We are using electricity in ever more diverse and unpredictable ways, and our utility monopoly model is increasingly expensive and unreliable as it attempts to adapt to the new realities of electricity demand. Instead of another 100 years of our current system, what if the utilities of the 21st century were more like internet service providers? The utility network is the system’s foundation, like the internet — a platform that creates an ecosystem for innovation. Such a reimagining would create significant opportunities, far exceeding any logistical challenges.
To understand what a 21st-century innovation could look like, we need to cover a few basics about our current system. First, the grid is not a battery. It does not store energy and deliver it to your device when you turn it on or plug it in. It is one massive teeter-totter, near-instantaneously balancing supply and demand (think microseconds). Practically, this means when you turn on your toaster, your local power plant responds.
Second, demand fluctuates significantly each day. According to the Midcontinent Independent System Operator’s Operations Displays, on June 17, the minimum electricity demand was roughly 68 GW with a peak at roughly 102 GW. This means that in MISO, 34 GWs of power plants turn on and off on a regular summer day.
Beyond these daily fluctuations, heat waves and other events can cause spikes in demand. The average length of load curtailment to accommodate these spikes lasts 1.7 to 2.5 hours, and these events generally account for less than 1% of the total hours in a year.
Why is this crucial? Because utilities are required to build infrastructure to accommodate this peak demand. Imagine footing the bill of a $2.4 billion gas peaker plant for it to lie dormant more than 8,700 hours each year, waiting for demand to spike. This is not a hypothetical scenario but, rather a well-worn plan that utilities have and will continue to execute unless we reimagine our energy system.
Gas peaker plants are built to provide a resource that can be called on to meet demand, so we can turn on our lights, cool our homes, query our data centers and manufacture our goods. If we recognize that the goal is not just to push more electrons but to actually accomplish these tasks, we can begin to rethink what our system looks like and explore ways of meeting our needs without buying an expensive asset that lies dormant 99% of the year.
One way to do this is to take advantage of existing battery storage that already interacts with the grid. In Michigan, there are roughly 100,000 electric vehicles and plug-in hybrids on the road. Assuming the average battery capacity of these vehicles is 65 kWh, this means that the storage infrastructure owned and available in Michigan is 6.5 GWh.
When vehicles are plugged into the grid using bi-directional chargers, they can both draw power from or deliver power to the grid. With half of Michigan’s vehicles plugged into bi-directional chargers (assuming 12 hours of charging or connection time) and drawing only 50% of each plugged-in vehicle’s capacity, between 575 MW and 1.6 GWh of capacity could be available to the grid today.
This is a massive resource — a virtual power plant (VPP) — greater than the entire battery capacity cleared in the 2025 MISO auction. This is a resource that already exists in Michigan. A resource that does not require investing billions of dollars in infrastructure and a decade before it can operate. A resource that could be used today.
So, instead of spending $2.4 billion on 1 GW of peaker plants that will remain idle 99% of the year, let’s invest in $4,000 bi-directional EV chargers for the existing 100,000 EVs. The maximum capacity of these EV chargers is 1.15 GW, providing 575 MW of capacity at any given time. And, all of this at a cost of $400 million, a sixth of the cost of a utility proposal.
Or, what if we built a power plant out of new electric vehicles? The Chevy Bolt with the federal tax credit costs roughly $20,000 and a bi-directional charger and installation costs roughly $5,000 and is supported by a $1,000 federal tax credit. This means that for the same cost of a 1-GW gas-peaker plant, we could buy roughly 100,000 EVs, providing an additional 6.5 GWh (assuming 65 kWh each) of total capacity.
These vehicles and chargers could be purchased and installed for customers to use at no cost, provided each customer committed to participating in 20 to 80 hours of peak events each year. This VPP could provide the same required capacity on peak days as a power plant and act as an additional grid asset for a variety of grid services as systems develop. Further, this VPP would provide clean, reliable transportation to Michigan residents year-round.
This type of VPP should be one part of the solution to our growing demands. It provides mechanisms to better utilize existing generation capacity without buying expensive stranded assets with high operating costs, ranging from $14/kW-hour to $63/kW-year, or $14 million to $63 million, in fixed costs each year. Similar services can be delivered now, with comparatively minimal investments, by aggregating existing distributed energy resources.
The energy transition is happening. Ratepayers are buying assets that can be aggregated into VPPs in the form of EVs, solar panels, heat pumps, energy efficiency upgrades and a variety of other resources to deliver the grid services we desire. Why should we also be forced to pay for and maintain additional power plants for utilities?