Episode description

Last week, I wrote an introduction to the hot new trend in energy: 24/7 carbon-free energy (CFE), i.e., matching a company or city’s power consumption with production of clean electricity throughout the day, every hour of every day. If you haven’t read it yet, you’ll want to check it out before reading this post.

Today, I want to talk about a big debate around 24/7 CFE, regarding whether it’s the right goal for companies and cities to adopt at all. Exploring that debate will help us get our heads around what 24/7 CFE can and can’t accomplish.

But first, a quick refresher.

Here’s the idea: right now, in addition to generating electricity, renewable energy projects generate renewable energy certificates (RECs), one for each megawatt-hour. They can sell the RECs to any entity looking to buy renewable energy. For instance, a company or city that wants to go “100 percent renewable” can simply buy enough RECs to cover its yearly electricity consumption.

At least two changes would be required to make 24/7 CFE possible. First, “renewable energy” would expand to “carbon-free energy.” Any generator putting out electrons without carbon emissions, including nuclear or natural gas with carbon capture and sequestration (CCS), would qualify.

And second, RECs, rather than coming in month- or year-long chunks, would be issued in time-stamped increments of an hour, so that buyers could target procurement at the particular hours of the day when they need CFE. Eventually, each hourly REC would contain information about avoided carbon emissions, so buyers could tally up the carbon impact of their purchases.

That’s the vision.

In this post, I’m going to discuss an objection to 24/7 and some counter-arguments to the objection. Then, in my next post (yes, this is turning into 24/7 Month), I’ll look at some new modeling of the impact of 24/7 procurement and try to draw some conclusions. We’re going to have a good time.

Measuring carbon is mostly doable

An intrinsic part of the 24/7 CFE vision is that each hourly REC will be tagged with a certain amount of avoided carbon. This will allow buyers to make procurement decisions that take emissions into account.

There are some issues and controversies around calculating avoided carbon, though they’re not the ones I’m going to focus on today. Some carbon counters have proprietary formulas (like WattTime) and some are trying to develop open-source methods (like EnergyTag).

The numbers they produce are not radically different, but they do differ. They vary in how they calculate the marginal (most expensive) energy source on the grid at a given moment — the marginal generator is the one that will spin down to make room when the CFE is produced. They differ in how to draw the geographic boundary of analysis, which can affect results. And other stuff like that.

“To go from the generation data to the local carbon emissions data is not trivial,” says Toby Ferenczi, founder of EnergyTag, “because you're trying to model the flow of electrons. Until you can track a single electron through the system, there will always be different types of approximations.”

There’s also the question of how distributed energy resources (DERs) are treated. Right now, grid operators tend to have little visibility into or control over DERs; energy generated locally, on a distribution grid, is viewed by grid operators as reduced demand on that grid. Bringing DERs more fully into the picture as deployable resources is an important long-term challenge.

There are data issues too. If you look at electricityMap, which seeks to track the carbon intensity of every grid in the world, at every hour, you will see that there are still big holes, areas where utilities have not made the data public. New regulations and laws requiring grid operators to make these numbers available is another priority.

Anyway, I’m not going to dig into these technical issues. I have faith that, if an hourly REC market gets going, these kinds of questions will be ironed out. The general sentiment is that it is more important to have a common set of numbers than it is for those numbers to be accurate down to the decimal.

Instead, let’s turn to the more fundamental challenge to 24/7 CFE.

24/7 vs. emissionality

Unlike air pollution, which concentrates where it is emitted, carbon dioxide diffuses completely into the atmosphere. It doesn’t matter where it is emitted; all tons are the same, from a climate perspective. One company or city’s emissions are no different than any others. There’s nothing about your hourly emissions that make them special.

It follows that, if you’re a company that wants to reduce carbon emissions, the thing to do is buy clean energy on the dirtiest grid possible, wherever it will displace the most carbon-intensive energy and thus prevent the most emissions. If you take an international perspective, that will probably be somewhere overseas, in Asia or Africa; if you take a US perspective, it will be in states like West Virginia, Wyoming, and Kentucky.

The best way to do this is with bundled RECS — RECs purchased together with the energy that produced them, through long-term power purchase agreements (PPAs) — because that’s the approach most likely to actually lead to new clean energy projects being built. But “most organizations are not in a position to sign long term PPAs,” says Ferenczi. “All they know is: I want to buy good electricity, not bad electricity.” For them, unbundled RECs are the only option.

Either way, if you want to reduce emissions with your CFE procurement, it must be guided not only by what’s most likely to lead to new projects (“additionality”), but also by what will reduce the most emissions (“emissionality”).

This word emissionality is a terrible neologism — the latest of many out of the energy world — but I’m living with it, because it’s a helpful way to refer to the quantification of carbon emission reductions.

Now, take note: optimizing your clean-energy procurement for emissionality is different from optimizing it for your own 24/7 consumption.

The former strategy maximizes emission reductions. The latter does not. In some cases, optimizing around your own consumption could fail to reduce emissions or even increase net emissions, despite increasing your share of CFE.

One simple example: imagine one company has signed a bunch of solar PPAs and thus has more hourly RECs during the day than it needs to cover its consumption, but it has a shortage at night; another company has signed a bunch of wind PPAs and thus has excess hourly RECs at night, but a shortage during the day. The companies can simply trade hourly RECs. Each has increased its CFE score, but no new clean energy was built and no carbon emissions were reduced.

Another example is how companies choose to deploy batteries. Mark Dyson, an energy analyst at RMI, explained it to me this way:

A battery optimized for 24/7 would charge when a buyer has procured “excess” renewable energy in a particular hour, but in most grids, for the foreseeable future, a fossil generator will usually be the marginal unit at the system level — so charging storage increases carbon emissions in that hour.

Discharging the battery later would offset generation from another fossil generator and reduce emissions, but there’s no guarantee the difference in efficiency of those power plants is greater than the round-trip-efficiency penalty of using the battery, and thus total emissions can actually increase.

In other words, optimizing battery deployment to cover 24/7 consumption will be different from, in some cases contrary to, deploying them to optimize emission reductions.

Nobody has yet modeled exactly how much these two strategies would diverge, or how frequent cases like the ones above might be, but no one disputes that they would diverge. A strategy built around emissionality would, by definition, reduce more emissions than any alternative strategy built around any other goal.

And this is the critique of 24/7 CFE: emissions are emissions. Reducing any one company’s emissions is of no particular benefit to the climate. Just reduce emissions wherever you can — that’s the climate imperative.

This same debate expresses itself in several different forms. One way to think of the distinction is between “attributional” and “consequential” carbon accounting. Critics (see, e.g., this paper from WattTime) say attributional accounting — purchasing energy with a REC attached — is fine for statutory or voluntary clean-energy requirements. But when it comes to reducing carbon emissions, companies should use consequential accounting, i.e., purchasing energy that has the most short-term emission-reduction impact.

The same debate crops up again between “hourly average” and “marginal” carbon measurement. One can either assess a unit of CFE based on its effect on the hourly average emissions on the grid in the hour it is produced or based on the carbon intensity of the marginal generator it displaces. Hourly averages are, for a variety of reasons, easier to determine, and can be used to boost your own CFE score, but a marginal approach (measuring “nodal marginal emissions”) will tell you which energy purchase will maximize short-term emission reductions.

All these debates are forms of the same question: why not focus on carbon emissions? As Henry Richardson of WattTime put it to me, “measure emissions, not megawatt hours.”

The emissionality critique — that emissions, not any company’s particular emissions, are the proper target for procurement strategies — is worth taking seriously. Everyone in the space has wrestled with it. Let’s run through a few possible responses and counter-arguments.

Volts is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Industrial policy vs. carbon policy

When I talked to Princeton energy modeler Jesse Jenkins — who contributed to the modeling of 24/7 CFE we’ll look at in the next post — he suggested a helpful analogy to the debate between emissionality and 24/7: the debate between a carbon tax and more sector-specific standards and investments, i.e., industrial policy.

A carbon tax is the most economically efficient way to reduce emissions — it will go after the cheapest emissions first. But by doing so, it will leave untouched many sectors of the economy that we will eventually need to decarbonize to get to 100 percent net-zero.

If we leave them untouched for too long, we’ll run into a wall. “We need to be thinking about the total solution,” says Melissa Lott, research director at the Center on Global Energy Policy. “Otherwise we're going to get halfway down the road, have to take a hard left, and it's going to be painful and expensive.”

The emissionality vs. 24/7 debate takes the same form. An emissionality approach would reduce emissions at a lower per-ton cost — it would go after the cheapest reductions first, usually by adding wind and solar to dirty grids.

But a 24/7 approach will direct investment toward technologies that fill the gaps left by wind and solar. “And there are gaps,” says Lott. “These gaps aren't eight or even 100 hours, which can be solved with different battery technologies. They're eight to 14 days.”

To cover those gaps will require “clean firm” generation, many sources of which are still in nascent forms of development. The pursuit of 24/7 CFE will stimulate innovation and growth in the entire suite of technologies needed to smooth out variable renewables — sources all grids will eventually need and many already do. (California power providers are already putting out solicitations for clean-firm projects.)

In fact, says Brian Janous, Microsoft’s director of energy and renewables, even the early talk of 24/7 CFE has gotten people thinking about solutions. “We're seeing more and more utilities, and more and more energy service providers, come to us and say, hey, we think we can solve this problem for you,” he says.

Companies pursuing 24/7 CFE are undertaking voluntary industrial policy, channeling attention and investment to gaps in current clean-energy technology, bringing down the costs so that other companies can use them more easily. That could have impacts well beyond their own emissions.

Here’s how Jenkins put it to me:

The heart of 24/7 carbon-free procurement is the pursuit of transformative impact on electricity systems via accelerated innovation. Think about the indirect emissions impacts from helping accelerate the time to maturity (or enable in the first place) one or more clean firm technologies or long-duration energy storage technologies that can go on to widespread adoption and make reaching 100 percent carbon-free electricity easier for the world.

Leadership isn't just about doing one's part. It is about making it easier for others to follow. For a company, even one as large as Google, this impact is likely to far outpace any direct emissions reductions they achieve via procurement.

24/7 CFE needs to be seen in its full context

None of the entities pursuing 24/7 CFE today see their own achievement of 24/7 CFE as the ultimate end goal. The goal is grid decarbonization.

“We break it into three pillars,” says Michael Terrell, Google’s director of energy. “First is transacting,” i.e., contracting with developers to ensure Google’s own 24/7 supply of CFE. “Second is advancing technology, both on the demand side and the supply side,” i.e., the industrial-policy piece. “Lastly is policy and grid decarbonization,” i.e., advocating for clean-energy policies before state public utility commissions (PUCs) and legislatures, to hasten decarbonization of the grids in which it operates.

“For us, it's not a win if the only way we get to 24/7 in each place is by transacting,” he says. “We want to get the grids moving in that direction, too.”

When it comes to the standard way of getting to “100 percent clean energy,” companies can just buy cheap RECs from distant grids. They don’t need to get involved beyond that. “That was a concern of ours,” Terrell says. “Companies were getting to 100 percent without having to consider the future of the grids where they were operating or do any policy.”

In contrast, if a company is trying to cobble together a 24/7 supply of CFE on its local grid, it becomes much more invested in the state of that grid. The more CFE is on the grid, the higher the baseline from which it begins transacting for its own CFE. That will get companies involved in pushing utilities to make clean-energy commitments, pushing PUCs to clear away anachronistic regulations, and pushing legislatures to pass clean-energy policy.

“We are trying to drive massive system change well beyond Google,” says Terrell. “The idea behind 24/7 is, you want corporates to have a stake in every grid where they operate. You want them to be banging the table, driving system change on these grids, getting these grids to carbon-free as fast as possible.”

Janous says that Microsoft also wrestled with the 24/7 vs. emissionality debate as it determined its next steps. “Ultimately, we determined that local influence is still important,” he says. “Our ability to influence PUCs and local utilities, and do that worldwide across dozens and dozens of different markets, was more important than taking a pure-play emissionality approach in one market.”

Time will tell how strong that local influence proves to be. What happens if progress on local grids is slow? Lott thinks the pursuit of 24/7 will move forward some tough calls. Entities pursuing 24/7 “are going to have to make a decision here in the next few years,” she says. “Do we keep our data center in this location where we don't see a clear path to [24/7], or do we move it? Do we shift investment somewhere else? This is going to be an interesting tension that will play out around 2025, ‘26.”

This is an aspect I think critics of 24/7 CFE tend to miss: the social dynamics. If it becomes the new standard for climate-conscious companies and cities, there will be dozens, maybe hundreds of them doing it, spread out across all of America’s many balkanized grids. Each will have reason to serve as a local clean-energy emissary. Each will be invested in the others’ success — one company’s PPA boosts the grid mix for every other company on that grid.

Companies will be incentivized to pool their resources for greater impact, as many are already doing through the Clean Energy Buyers Association (CEBA), which organizes and accelerates voluntary clean energy procurement. (A telling tidbit: until quite recently, CEBA was REBA, but “renewables” have given way to “clean.”)

In short, the movement to 24/7 has the potential to drive social and political change in a way that traditional REC markets never could and arguably a pure emissionality approach couldn’t either.

Emissionality can inform a hybrid approach

The choice between 24/7 and emissionality does not have to be a stark either/or. It is possible to use both perspectives for different jobs, or to blend them.

Take the hourly-average vs. marginal debate. “All of these accounting systems have their advantages,” says Terrell. If you’re thinking about offsetting the consumption of a large facility, “you want to be looking long term, at the 10-to-15-year roadmap of that grid, and average emissions is fine for that,” he says.

On the other hand, if you’re thinking about offsetting the emissions of product supply chains and product use — which are spread out across the country — there are no local consumption concentrations to target, so you might take a marginal approach to seek the cheapest, fastest emission reductions.

Put another way: you can use hourly averages to offset your scope 2 emissions and marginal to offset your scope 3 emissions.

Microsoft, Janous says, takes emissionality very seriously:

The way I think about it is, there are three stages of impact. The first one is attribution — it's just buying RECS. Then, five years in, we moved to we would call additionality — it's not good enough for me to have this attribute, I need to have an attribute tied to something that I actually did.

We're now getting into this third era, which is what I would deem consequential — not only do I need to say that I caused it, I need to be able to demonstrate that I'm materially changing the makeup of carbon on the grid.

With that in mind, he says, Microsoft has developed a “hybrid form.”

By pursuing 24/7, “we are going to look at each grid where we operate and we're going solve for that,” he says. “If we can achieve 100 percent decarbonization of our energy supply across our portfolio, we've demonstrated that you can do it just about anywhere.”

“By virtue of taking that grid-level approach, we are not going to have perfect optimization for emissionality,” he says, “but we're going to apply [emissionality] within that grid context.” In other words, within the grids containing Microsoft’s local loads, Microsoft will purchase the CFE that reduces the maximum emissions.

In this hybrid approach, neither 24/7 CFE nor emissionality is perfectly optimized, but both are pushed forward together.

“We feel like the principle [of emissionality] is still extraordinarily valuable,” Janous says, “even when you do it in this hybrid way.”

Other values could supplement 24/7 as well

Janous, Terrell, and pretty much everyone else I spoke to emphasized that there are other values beyond emission reduction that are important to integrate into procurement decisions as well, most notably environmental justice.

Last year, Salesforce released a white paper, “More Than a Megawatt,” explaining its evolving view on large-scale corporate procurement. It wants to go beyond emissionality to assess potential clean-energy projects based on a whole range of criteria, from equity to land use to impacts on wildlife.

There are always trade-offs among these metrics, so Salesforce has developed a “procurement matrix” that will help weigh all these factors and determine which projects best optimize for multiple values. (A similar whitepaper was recently released by LevelTen Energy, a renewable energy procurement platform.)

Notice that the more of these values enter your procurement matrix, the farther you are from a pure 24/7 play. Instead of optimizing for any single value, you are — as in life generally — trying to balance multiple competing values under time and resource constraints.

That can be complicated. It will help if big players like Salesforce create some standardized tools and metrics that make it easier for mid-sized companies to follow suit. Individual companies and cities can decide for themselves how much weight they want to give to 24/7 CFE relative to other values.

Anyway! This post, like the last one, has gotten way too long. If you’ve stayed with me this long, you are definitely a Volts reader and should purchase a subscription!

In my next post, we’ll have a close look at some new modeling of 24/7 procurement from Princeton’s Zero Lab and then see, based on that and all that has come before, whether we can draw some provisional conclusions about 24/7 CFE.

Notice that the more of these values enter your procurement matrix, the farther you are from a pure 24/7 play. Instead of optimizing for any single value, you are — as in life generally — trying to balance multiple competing values under time and resource constraints.

That can be complicated. It will help if big players like Salesforce create some standardized tools and metrics that make it easier for mid-sized companies to follow suit. Individual companies and cities can decide for themselves how much weight they want to give to 24/7 CFE relative to other values.

Anyway! This post, like the last one, has gotten way too long. If you’ve stayed with me this long, you are definitely a Volts reader and should purchase a subscription!

In my next post, we’ll have a close look at some new modeling of 24/7 procurement from Princeton’s Zero Lab and then see, based on that and all that has come before, whether we can draw some provisional conclusions about 24/7 CFE.