The Sense And Nonsense Of Virtual Power Plants

Over the past decades power grids have undergone a transformation towards smaller and more intermittent generators – primarily in the form of wind and solar generators – as well as smaller grid-connected batteries. This poses a significant problem when it comes to grid management, as this relies on careful management of supply and demand. Quite recently the term Virtual Power Plant (VPP) was coined to describe these aggregations of disparate resources into something that at least superficially can be treated more or less as a regular dispatchable power plant, capable of increasing and reducing output as required.

Although not actual singular power plants, by purportedly making a VPP act like one, the claim is that this provides the benefits of large plants such as gas-fired turbines at a fraction of a cost, and with significant more redundancy as the failure of a singular generator or battery is easily compensated for within the system.

The question is thus whether this premise truly holds up, or whether there are hidden costs that the marketing glosses over.

Reactive Power

The power triangle, showing the relationship between real, apparent and reactive power. (Source: Wikimedia)
The power triangle, showing the relationship between real, apparent and reactive power. (Source: Wikimedia)

The alternating current (AC) based electrical grid is a delicate system that requires constant and very careful balancing to ensure that its current current and voltage don’t go too far out of phase, lest grid frequency and voltage start following it well beyond tolerances. The consequence of getting this wrong has been readily demonstrated over the decades through large-scale blackouts, not the least of which being the 2025 Iberian Peninsula blackout event that plummeted all of Spain and Portugal into darkness. This occurred after attempts to reduce the amount of reactive power in the system failed and safeties began to kick in throughout these national grids.

This is also the point where the idea of a VPP runs into a bit of a reality check, as the recommendation by the grid operators (transmission system operators, or TSOs) is that all significant generators on the grid should be capable of grid-forming. What this means is that unlike the average invertor on a wind- or PV solar installation that just follows the local grid frequency and voltage, it should instead be able to both absorb and produce reactive power.

Effectively this involves adding local energy storage, which is where the idea seems to be that you can sort of fudge this with distributed dumb inverters and grid-connected batteries in the form of people’s whole house batteries and whatever Vehicle-to-Grid (V2G) capable BEV is currently plugged in within that subsection of the grid.

Theoretically with enough of these scattered generators and storage elements around, along with a few grid-forming converters and remotely controlled loads like EV chargers and airconditioning units, you could simulate the effect of a regular thermal- or hydropower plant. The question is whether you can make it work well enough, and as a logical follow-up question, there are those who would like to know who is really footing the bill in the end.

Battery Rental

Electricity generation by type, 2001-2024. (Credit: California Energy Commission)
Electricity generation by type, 2001-2024. (Credit: California Energy Commission)

An example of such a VPP in action can be found in California, where PG&E and others have recently been running tests. A big focus here is on these home batteries, which are also used for peak-shaving in these tests, with the battery owner compensated for withdrawn power. In a report sponsored by Sunrun and Tesla Energy, the Brattle Group describes this system in which the Demand Side Grid Support (DSGS) program aspect is hailed as a major revolution.

Fire at the Moss Landing Power Plant. (Credit: Guy Churchward)
Fire at the Moss Landing Power Plant. (Credit: Guy Churchward)

The idea here is that regular grid-connected consumers install batteries which the grid operator can then tap into, which can compensate for California’s increasing amount of non-dispatchable, non-grid forming generation sources. Of note here is that grid-scale energy storage can never provide enough capacity to bridge significant spans of time, ergo the proposal here is primarily to provide an alternative to expensive peaker plants, of which California already has a significant number.

With a predicted 4 GW of home battery capacity by 2040, this could then save the grid operators a lot of cash if they can use these batteries instead of running special peaker plants, or installing more large batteries as at the (PG&E-operated) Moss Landing battery storage facility.

Incidentally, said Moss Landing battery storage facility has repeatedly caught fire, which highlights another potentially major savings for grid operators, as the fallout of such events are instead borne by the operator of the battery, which for the DSGS would be the home owner. So far, remote adjustment of air-conditioning doesn’t seem to be a big part of the discussion yet, but this would seem to be only a matter of time, considering the significant power savings that way, even if it relies just on paid volunteers like with the DSGS.

Signs Of Market Failure

Although it can seem tempting to imagine making money off that expensive home battery or electric car by letting the local grid operator tap into it, the same general issues apply as with the much older V2G discussion. Not only is there the question of battery wear, but as mentioned there are also insurance considerations, and the problem that home batteries and BEVs tend to be sited far from where they are likely needed. While a site like Moss Landing is directly plugged into the big transmission lines, home batteries are stuck on some local distribution grid, making dispatching their power a bit of a nightmare.

This is also the impression one gets when reading certain articles on VPPs over at the US Department of Energy, with a VPP plan in Illinois targeting larger commercial and community solar generators rather than residential, giving them a rebate if they want to foot the bill for installing a grid-following converter, which presumably would involve some level of on-site storage. A major problem with distributed resources is their distributed nature, which precludes any planning or siting considerations that directly address demand in the form of building a power plant or pumped hydro plant with a direct transmission line to where it’s needed.

Projected electricity generation pathways by 2040. (Credit: S&P Global Inc.)
Projected electricity generation pathways by 2040. (Credit: S&P Global Inc.)

Meanwhile a recent study commissioned by the American Clean Power Association (ACP) concludes that in the US alone, electricity demand by 2040 is likely to surge 35-40% compared to today, requiring an extremely fast buildout of additional generating resources involving mostly the same kind of power mix as today. At a projected 5.5 – 6 TWh by 2024 compared to about 4 TWh today with a significant boost in non-dispatchable generators, it seems fair to question how far home batteries and a handful of V2G-enabled EV cars can support this effort in some kind of national VPP system.

Asking The Basic Questions

Although it’s often said that ‘distributed electricity generation’ is the future, it’s rarely quantified why exactly this would be the case. Simply looking at how AC power grids work, along with the tracing of the kilometers of required transmission lines across a map in order to connect all disparate generators should give one plenty of pause. It seems obvious enough that an abundance of distributed, non-dispatchable, non-grid-forming generators on a grid would also prove to be problematic, especially in the wake of the Iberian blackout this year.

Patching around this by making end-users foot the bill for battery storage and grid-forming converters and calling it VPPs then feels disingenuous. Here a more reasonable model – that has also been repeatedly suggested and occasionally implemented – involves homes and businesses equipped with local storage that only serves to reduce demand on the grid. These batteries can be charged from the grid when the ¢/kWh rate is very low, providing a balancing influence on the grid without remote control by TSOs or similar levels of complexity.

Ultimately it would seem that the European TSOs (ENTSO-E) with their focus on eradicating dumb converters and requiring grid-forming ones are on the right track. After all, if every wind and solar generator installation acts for all intents and purposes as a dispatchable generator with the ability to absorb and generate reactive power, then the whole VPP debate and much of the grid-storage debate is instantly irrelevant. It just means that the investors for these variable generators will have to spend significantly more instead of palming these costs off on end-users as some kind of grand deal.

15 thoughts on “The Sense And Nonsense Of Virtual Power Plants

  1. This article uses a LOT of very specific “grid operator” language like “forming dumb coverters” and lots of TLA’s. What did I just read?! Where’s the “Hackaday” angle? Where is the homeowner who built their own VTG setup? The em dashes right at the beginning give me pause.

    1. Everything is cited and referenced in the article for your background information pleasure needs. There are also no em-dashes in the text, so I’m not sure what ChatGPT-regurgitated alternative reality version you think you just read.

  2. Overall, VPPs seem like an effort to avoid (or even merely delay) infrastructure investments by power companies. It seems quite inevitable that grid-scale battery facilities will be constructed and installed across the planet. The part missing from the equation is a producer of dirt-cheap sodium-ion batteries.

    1. When they introduced huge numbers of VRE companies on the system, that didn’t have part or stake in power transmission at all, under subsidies and right-of-way on the grid rules, it simply created a situation where the utilities profit from the fact that the supply doesn’t meet demand.

      The market has to transmit more power over longer distances to make ends meet. More transmission, more revenue, more profit. This is similar to the situation in California couple decades back, when the utilities faked transmission capacity bottlenecks in order to hike prices, pretending that power wasn’t available. This time they don’t have to fake it, because the system is actually broken.

      And they don’t want to pay anything to fix it. Why would they?

      The grid isn’t a grid but a tree for the most part, so placing the batteries near the trunk in central locations would reduce average transmission distance, and result in less revenue for the transmission company and fewer opportunities to hike prices. Installing the batteries at peoples’ homes would keep the transmission distances longer, and have someone else pay for it.

  3. One thing which isn’t really explored here is Distribution System Operators, whose primary role is to do just this. Managing energy flows within the distribution network to optimise it’s capacity / capability, but at the same time working with the TSO(s) to help manage regional flows.

  4. “It just means that the investors for these variable generators will have to spend significantly more instead of palming these costs off on end-users as some kind of grand deal.”

    Ah like work at home then. Make Home Offices Great Again.

    1. In a way, but even worse, as at least with a home office you can get a lot of expenses covered and even some insurance and tax benefits if you have it as a proper separate home office from your domicile.

      This whole approach doesn’t seem to give you even such benefits, based on the programs I have looked at so far.

  5. Honestly power plants should just implement load-shedding until the average consumer gets the picture and buys solar. Trying to wrangle a million battery banks into a semi-functional power grid is a complete boondoggle. Just give up, we aren’t going to fix it. The grid is to electricity what mainframes were to computing; it’s time to let it go.

    1. consumer gets the picture and buys solar. Trying to wrangle a million battery banks into a semi-functional power grid is a complete boondoggle.

      That’s the exact same thing. The homes need even more batteries if they can’t rely on the grid, and hundreds of millions of homes switching to grid power when they run empty is just as unmanageable.

  6. Gen IV nuclear fission reactors need to be built and taken seriously.

    If you want wind and solar to be of ANY use, especially where load managment is concerned, they should be used to add power to the off peak storage equation. Gravity or Hydro storage for best use. That way the intermittent availability of those forms of energy would be offset by the fact that they don’t contribute directly to what we call power generation.

    As the information age enters what will be the next big leap where data centers are going to gobble up gigawatts of energy we need to get serious about what we can do. We should use the waste heat the data centers generate to heat water as needed for any number of things instead of dumping it to the atmosphere or waterways.

    But please yes, let’s just put on those rose colored glasses and think that the “green energy” that we’re being sold is the answer to everything. I am sorry to say it is not. A part of a solution perhaps, but even then I’m not convinced.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.