Virtual Power Plants and Why You Should Care
In a decarbonized world, transportation and buildings have shifted away from fossil fuels and transitioned towards electric devices; such as heat pumps, electric water heaters, and of course EVs. These electric devices would be supplied by a decarbonized grid and charged at economically efficient electricity rates. Almost 50 years ago, Professor Schweppe painted a picture set in 2000: where a sophisticated control system manages all of the devices at the building-level, balancing supply and demand in real time, and optimizing against localized, real-time prices. Driven by his vision, I learned how challenging it would be to bring this picture into fruition while interning in Santiago de Chile in 2018. My internship project was to help propose a pricing structure for a community of homes with solar and storage devices such that the neighbors could transact electricity among themselves. Not only was it impossible to tackle this PhD-worthy question in 10 weeks, but even moreso to describe my takeaways in Spanish. I’ve been working towards making this idea of a “decarbonized, electrified world,” into a reality for the last five years, and with the recent buzz around Virtual Power Plants (VPPs), I wanted to explain why they are more important today than ever before.
Historically, electricity was managed in a top-down, centralized way. Now, with more Distributed Energy Resources (DERs) such as smart thermostats, rooftop solar, and electric vehicles, they introduce two important shifts in customers’ relationships with the grid: 1) End customers are consuming and also generating energy– with rooftop solar or bidirectional EV chargers – and sending surplus energy back into the grid. 2) There is more flexibility on when you use your electricity. Traditionally, when you flip on the light switch, you expect power right away. But now, there are more devices that introduce flexibility on when they need the power, as long as they are switched on within a certain time period (e.g. Apple’s new “Clean Energy Charging” option).
There have been a lot of debates on how DERs will play a role in the energy transition. Some utilities and academics saw DERs as a threat to the existing paradigm of the top-down utility business model. While some economists may argue about the cost effectiveness of rooftop solar or home batteries, Electric Vehicles introduce an undeniable wave of change which will force us to rethink how we value DERs moving forward. This begs the question of how to (and how to not) value DER exports, as we explored in the last article. One answer to this question: Virtual Power Plants.
What is a Virtual Power Plant?
Let’s picture all of the DERs on your street – smart thermostats, electric vehicles, rooftop solar, and other flexible energy devices. Just looking at one street, there may not be enough to make a substantial impact on your local grid. Now let’s zoom out – if we consider all of the flexible energy devices in Silicon Valley, the amount of DERs could be on the order of magnitude of a few MWs. Last Labor Day, there was a multi-day heatwave across California that placed so much strain our electrical grid to the point where everyone received a text from Governor Newsom asking to reduce their electricity consumption. Sending a text message that asks people to run their dishwasher later at night is not the answer. In 2023, where I passed multiple self-driving cars this morning, I think we can do a little better than that. Instead of relying on people to physically delay their laundry loads, we can instead send signals to a group of devices that make up a Virtual Power Plant (VPP).
What does that look like in implementation? We can shift your EV charging to later in the evening, we can pre-cool your home while there is more solar, pre-heat your water before you go for a shower. With more demand flexibility, we can shift energy consumption to cheaper, higher-renewable hours devices – all while maintaining customer comfort. Even by dispatching the VPP a few times a year, they will help reduce emissions, prevent blackouts, and save both the utility and customers money. By avoiding the use of fossil fuel plants, the utility may be able to retire dirty gas plants and not replace them. They could also avoid transmission & distribution upgrades, or delay the investments for a few years. When the utility saves money on such an investment, the savings get passed onto customers; win-win. We need to explain how engaging with energy flexibility within your community is an example of collective action, and allows for your community to help fight climate change together.
Why is this important, you may ask?
Our energy system is changing in so many ways. We are building more renewables and driving towards aggressive decarbonization goals, which introduces intermittency challenges (e.g. passing clouds over a solar farm). We are switching to electric vehicles and electric heating, which will increase electricity consumption and rising peak demands. We are also facing more frequent extreme weather events due to climate change (Winter Storm Yuri in Texas, 2021, California’s Labor Day heat wave, 2022, and the winter storm on the East Coast during Christmas Eve, 2022). With these shifting trends, resiliency and reliability of the grid have been on the top of minds for most energy nerds. Virtual Power Plants can play a very valuable role in providing capacity and reducing peak demand during these increasingly frequent reliability issues, ultimately preventing blackouts.
As customers have more flexibility on when they use their power, we can monetize that “demand flexibility” to avoid or delay Transmission & Distribution upgrades – which will become more and more relevant as everyone plugs in their EV in the coming years.
From a supply chain perspective, we need to rethink how and where we are getting our critical minerals (lithium, cobalt, nickel) for batteries and EVs, in order to meet our adoption goals. We can’t afford to put a battery in every car and to not use it in its most effective and efficient way.
Currently, we use “peaker plants” during peak energy demand events, which are only used ~3% of the year on average. The majority of these plants have a disproportionate amount of low-income residents within a three-mile radius. Many of these peaker plants are aging, inefficient, and up for retirement. Instead of replacing them with oil and gas power plants, virtual power plants could fulfill the same need. In turn, this will help save money (for both the utilities and customers), reduce emissions and benefit disadvantaged communities.
Why is this hard?
It is very computationally challenging to optimize thousands of diverse distributed assets. Each device has its own technical constraints (e.g. battery size, minimum state-of-charge), customer constraints (e.g. I need my car charged by 8 am, or my contract says I only get dispatched 4 times per summer), and economic constraints (e.g. some devices are cheaper to dispatch than others). The system operator who manages our bulk grid could not account for each energy device at the distribution level. With more distributed energy assets generating energy and shifting energy consumption, we essentially need an independent system operator at the distribution level (a DSO). However, shifting towards a DSO model would be expensive and the market design has not always proven successful. VPPs fit into current market structures, while providing similar visibility and benefits that a DSO would.
Ideally, an energy customer would go on living their life without sacrificing comfort or productivity. In order to remove the decision burden from the consumer, VPPs can control and dispatch devices directly. This is not easy, as each device may use their own way of communicating, or the device manufacturer may not contractually agree to allowing VPPs to connect to their device.
Example of an Electrified Home, from Rewiring America
Barriers for VPPs:
High upfront capital expenditures; Electric vehicles, electric water heaters, heat pumps are quite expensive. The early adopters (affluent customers) may not be interested in a $25 monthly incentive to dispatch their new Ford F-150, which makes it difficult to get this business model off the ground until there are enough DER deployments and we can take advantage of network effects – across all customer segments and demographics. We don’t want to replicate the same mistakes leading to inequity carried out by rooftop solar, but instead learn from those mistakes and perhaps introduce alternative financing mechanisms to lower the barrier to entry for middle-income customers.
Open Protocols; for example, if Tesla’s EVs can only enroll in Tesla’s VPPs, it makes it challenging to have a vendor-agnostic, technology-agnostic VPP. If I want to sign up for a VPP with my ecobee thermostat, and my Rheem water heater, but cannot enroll my Model S into the VPP because Tesla is not playing nice, this is a big hurdle for monetizing my entire portfolio of demand flexibility.
Ultimately, these distributed energy resources will be adopted aggressively across the US, especially with EV goals acting as a further impetus. We need to manage and optimize these assets and monetize this higher level of flexibility in the most efficient and intelligent way. It doesn’t make sense to have millions of batteries on wheels and to not use them to their fullest potential – neither our pockets nor the grid could afford it.