WattCarbon 2023 Recap and 2024 Look Ahead - Part 2
The First EAC Marketplace
When we started WattCarbon, one of the assumptions we had was that we’d be able to facilitate transactions in existing voluntary carbon markets on behalf of building owners. Looking back, this assumption was a bit naive. Even though the nature-based offsets regime was already on the defensive at the time, we figured that providing them with an offering that was more verifiable, like eliminating emissions from buildings, would be met with open arms. Instead, we received pushback and general skepticism regarding energy-based emissions solutions. Removals are the name of the game in the offset world.
In hindsight, this was probably for the best. Sometimes these setbacks are blessings in disguise. In the past year, we’ve seen the voluntary carbon market crumble under a plague of scandal and a general recognition that the challenge facing us is not one of mitigation, but of fully transitioning our economies away from fossil fuels.
What this means is that we’re going to have to build new environmental commodity markets that are specifically designed to achieve decarbonization goals. In other words, it’s time for EAC-based measurement and verification of carbon free energy deployment, backed by actual data.
There are glimmers of hope that existing energy markets are going to start becoming more accommodating of a decarbonization focus, but today these markets reflect the constraints (both conceptual and technological) of the past. State energy offices and renewable energy bookkeepers (e.g., REC registries) have talked about expanding their coverage to include EACs, and some preliminary efforts are underway to provide hourly production data for RECs, but not with the granularity and fungibility required for EAC markets. Outside of batteries, even cutting edge standards efforts like EnergyTag have been unwilling to entertain the notion that EACs might not always be based on renewable energy generation.
For these and other reasons, we decided early in 2023 to create a new EAC-based marketplace and launch it with the goal of deploying 1,000 MWh of new residential solar, 1,000 MWh of energy savings from demand response, and 1,000 tons of CO2 emission reductions from heat pump electrification projects in a first demonstration phase. We launched the marketplace at the end of June, closed the first set of portfolios by the end of September, and issued the first payments at the end of December.
Below we’ll spend some time reviewing the decisions that we made to get the pilot launched and lessons learned in the first iteration that will impact how the next phase gets rolled out in 2024. We are excited to be able to open up the marketplace to any entity that is able to generate EACs starting in early 2024. Combined with our open-source EAC registry, we hope to enable billions of dollars to be invested in the deployment of clean distributed energy resources over the next few decades.
Marketplace Fundamentals
Developing a first-of-its-kind marketplace, we had to make some big deployment decisions, starting with how to aggregate buyers and sellers. To simplify a lot of potential market complexity, we decided to create fixed sized portfolios of similar projects that would be sold at a fixed price. A buyer could purchase a share of the portfolio in the form of a specific amount of energy (or carbon). So you could order 10 MWh confident that you would receive 10 MWh from the pool, even if you wouldn’t know exactly which 10 MWh you would receive. All of our buyers were aggregated together and the portfolio was capped. The structure ended up looking like a mashup between a Kickstarter and a Mutual Fund.
We wanted multiple suppliers in each portfolio, but we didn’t know ahead of time which ones would actually be able to deliver projects and how much savings there would be in each project. So each supplier was allocated a share of the portfolio to supply, with a minimum amount reserved for each one, but where collectively they would be competing to fill the limited amount available in the pool.
To ensure maximum impact, we decided to make advance market commitments to project developers. We wanted these projects to meet additionality thresholds. So we had to make firm commitments from WattCarbon’s own treasury that we would guarantee payments for the energy savings from their projects, even if we didn’t have sufficient buyers for these savings. After all, we didn’t know what the market would support and we didn’t know if the price point we picked would be too high or too low. But because we offered an advance market commitment, our suppliers could start to bake the cash flow from our market into the value proposition offered to their customers.
While we could have gone further to constrain how the funds were spent, we decided not to add too many other strings to the supplier agreement, in order to learn as much as possible from the pilot marketplace deployment. Generally, the idea was to give them as much flexibility as they needed to figure out how these EAC payments could accelerate the deployment of their technologies.
Another key decision was to constrain our portfolios to one of three different types of resources:
Renewable energy from solar production in the PJM grid;
Load shifting from demand response in CAISO;
Electrification from projects across the United States.
The next few sections review these portfolios in more depth, highlighting the different complexities posed by each type of resource.
Solar EACs
We partnered with Solar Holler and Atmos Financial to procure EARs (Energy Attribute Rights) from their customers for solar installations completed in 2023 and for energy generated between July 1 and December 31. For these partners, we guaranteed 1,000 MWh of purchases at a fixed price of $40/MWh. The RECS associated with these projects were first minted within the PJM GATS tracking system, then transferred to WattCarbon. We created EACs from hourly inverter production data from end customers and matched them to the corresponding RECs from GATS so that we could assign an individual serial number to each watt-hour generated by the panels and cross reference to the original REC that had been recorded and retired on GATS.
Pooling the buyers together allowed us to allocate the EACs on a proportional basis. If a buyer purchased 1MWh of the 1,000 MWh in the pool, they would receive 0.1% of the EACs minted for each hour of the day for each system in the pool. Because we were able to fractionalize down to the individual watt-hour, we could implement an allocation system that efficiently made sure each buyer received the correct allocation of EACs within the pool and that no buyer received a materially different set of EACs than any other buyer (such as different times of day). Once each buyer receives their EACs, they are able to selectively retire or transact with another party on an individual watt-hour basis.
In general, solar is the most straightforward use case for creating EACs (other renewable generation would be similarly straightforward). The most challenging aspect is the discrepancy that can show up between inverter data and the records kept by the REC registry. It is important that the EACs match the RECs, if both are minted, but figuring out the causes of discrepancies between what should be the same set of data can be difficult. Still, we expect to greatly expand our renewables coverage in 2024, making it possible for anyone to convert an existing or new REC into an EAC so that it can be transacted at the hourly granular level.
Load Flexibility EACs
The second type of EAC that we included in the marketplace was for energy savings related to demand flexibility. Increasingly, grid managers are asking customers to turn off or shift their electrical loads to reduce consumption during particular times of the day. While renewable energy RECs and EACs are fairly familiar to the industry, demand flexibility is part of an emerging market for Virtual Power Plants (VPPs) powered by Distributed Energy Resources (DERs). The underlying variability in renewable energy production and limitations of existing infrastructure to manage periods of peak demand has created new value in active energy load management. Both the Department of Energy and the Federal Energy Regulatory Commission (FERC) have implemented policies to spur the deployment of VPPs. FERC order 2222 specifically orders RTO markets to incorporate aggregated DERs on an equal basis with traditional forms of generation.
Despite the inclusion of these carbon-free energy resources in wholesale energy markets, VPPs have not received equal treatment in environmental markets. Given how central these resources are to the expansion of renewable energy generally, and their importance in filling in the gaps where renewables are offline, we think this exclusion is wholly unwarranted.
To prove the concept of EACs for demand flexibility, we contracted with Leap, a company that aggregates large quantities of distributed energy resources and bids them into competitive energy markets. We asked Leap to bid up to 1,000 MWh of load reductions into the CAISO market between July 1st and December 31st. We paid them an extra $30 for each MWh that was successfully bid into the California energy market. The idea behind this subsidy was to demonstrate how recognizing the environmental benefits of DERs allows these resources to become more price competitive in energy markets. Currently, the main economic value of DER deployment is limited to a handful of times per year when demand is at an all-time high and grid prices soar accordingly. However, there are many days when the grid is reasonably expensive and the primary source of power is a dirty fossil fuel power plant, where an extra subsidy would make it economically feasible to bid the DER into the market so that it would displace the fossil fuel plant.
What makes demand flexibility more complicated than traditional renewable energy generation is the need to calculate the load impact. A reduction in energy use relative to a baseline requires establishing a counterfactual number of what the “expected” load would be. This is a familiar problem for those in the energy efficiency and demand response industries, but can seem confusing for those coming from the outside.
As both contributors to and members of the governing board of the OpenEEmeter open source project for calculating energy savings, we actually have more experience than most in making these sorts of calculations. While the CAISO (and all other RTOs) set their own guidelines for calculating energy savings from load flexibility resources, we also explored how we might use a more advanced set of methods that could also account for pre- and post-event impacts.
While there is some justification for using these more sophisticated methods, we believe that at this point the best way to align the calculation of the environmental benefits is to use the same energy values recognized by the grid operator. Just like the hourly production of solar panels should add up to the MWh of REC recognized by the grid operator, individual DER savings should also add up to the same energy savings value recognized by the market.
One additional complicating factor when working with DERs is that each individual participating device might only contribute a small fraction of the overall impact of the event. A grid operator will often require small resources to be pooled together or aggregated in order to qualify for market participation. However, for the purposes of avoiding double counting, it is important to know which individual units contributed to the event and which did not. For reporting purposes, WattCarbon is aggregating multiple individual units together to generate “aggregated” EACs. But each portfolio of contributing devices is tracked at the unit level as well so that there is full auditability available to ensure that the same unit is not getting credit multiple times for the same event. This auditability will be a core feature of the open source EAC Registry that we will release early in 2024.
Electrification EACs
The final type of portfolio that we included, electrification, looks fairly different from either renewable generation or demand response. For starters, unlike renewables and demand flexibility, electrification is not bid into competitive energy markets. Clean thermal energy is considered an eligible resource in several state Renewable Portfolio Standards (RPS), but there is no consensus on how to account for the environmental benefits of an electrification project. Despite the lack of formal accreditation, electrification projects that eliminate dependence on fossil fuels for heating, cooking, water heating, and transportation represent one of the greatest opportunities for decarbonization at our disposal. While existing accounting frameworks struggle (this is a Scope 1 reduction, but also a Scope 2 increase, and it may need to be claimed as a Scope 3 offset), the fact of the matter is that electrification is vital to achieving climate goals.
We followed the same general portfolio design for electrification as we did for renewables and demand response, but some of the fundamental principles were different. To start with, electrification projects are entirely prospective and the carbon savings are highly dependent on the buildings in which the projects take place. Even the units (do we measure in therms, kWh, or tons of CO2?) are ambiguous relative to solar or demand response. The benefits of electrification are measured over the long run, but unlike solar there is no natural payback period, so the entire value of the subsidy has to be anchored to the upfront decision to replace a furnace or gas water heater with a heat pump in the first place.
Methodologically, electrification represents a significantly more complex problem than either renewables or demand flexibility. For starters, we have to create three different numbers: 1) a forecasted savings number so that our project developer partners would know what type of incentive to offer their customers; 2) a projected savings number that could be used to pay the project developer at the point of project completion; 3) a calculated savings number that could be used to create EACs once the results of the project were measurable. We also had to solve the problem of accounting for baseline fuel of one type and reporting period fuel of a different type. We would expect to see a decrease in natural gas (or fuel oil) consumption and an increase in electricity consumption. But since these end uses are not independently measured, we have to estimate them using a combination of weather-normalized energy models from actual data plus calibrated end-use load shape models from aggregated data.
For example, to address the expected increase in electricity consumption from a heat pump, we estimated a typical increase in electricity consumption and then purchased clean energy to net out the carbon emissions that would result from increased electricity consumption. But to encourage greater efficiency and a more holistic approach to the project, we also provided a rebate to the project developer based on the difference between the estimated increase in electricity consumption and the actual increase in electricity consumption. That is, first we modeled so that we could issue a payment, then we measured so that we could incentivize performance.
To work out the differences between different types of business models, we partnered with BlocPower, Elephant Energy, and QuitCarbon and offered to pay $67.50/ton for reductions in natural gas consumption net of whatever the costs of procuring clean energy to offset the increase in electricity consumption turned out to be. In cleaner grids, this cost would be lower. In dirtier grids, this cost would be higher. We also decided to pay for the full lifetime of carbon savings in the first year, mirroring the way that energy efficiency programs pay for long-term savings, and we adopted the conventions of energy program technical resource manuals to determine the number of years that would be counted in the performance lifetime (13 years in the case of water heaters, 15 years for heat pumps). To calculate the savings of the projects, we required each of our partners to connect utility data from the customer through a provider like UtilityAPI. Not all natural gas utilities (and certainly no fuel oil companies) offer online meter data, which complicates the verification process.
Despite all of this complexity, we were extremely pleased with the outcome of the electrification pilot. By the end of 2023, we had made initial estimates of carbon savings for heat pump water heater projects with QuitCarbon and issued our first payments for these savings. We were able to develop forecasting tools and a strategy for project level M&V that would prove emissions reductions using metered utility data. It will take a full year before we can put these measured savings methods to work, but we are confident that the impacts will prove real.
The major lingering question we are currently working through is whether or not to adopt the Massachusetts standard for quantifying the benefits of heat pumps. One advantage of this approach is that it standardizes the calculation and converts it into kWh for consistency with other RPS resources. But because it is based on a square footage conversion, it overlooks the benefits of energy efficiency measures that may be implemented at the same time and ignores other factors that could be picked up with a weather normalized model.
Looking ahead at 2024
For 2024, WattCarbon is planning to open up its marketplace to any building or entity that would like to buy or sell certified EACs. We’ll be starting early enrollment starting next week, so anyone interested who has not been in touch with us should feel free to reach out. We’ll be allowing any EACs to be registered that are directly metered or that have a certified third party attestation of savings (such as a utility or ISO market operator). As I’ll explain in part 3 of our annual review, we’ll also be introducing a vPPA for DERs, in which the EACs come bundled with the delivery of energy in the form of a long-term contract. More on this to come!
Great work!