We need a better way of purchasing and accounting for renewable electricity

Kumar Venkat
6 min readNov 27, 2022


Over 73% of global GHG emissions come from energy used in industry, transport, buildings, and agriculture. While certain other emissions might be unavoidable and may need to be offset or captured, energy-related emissions can be largely eliminated by electrifying just about everything and then supplying that electricity from zero-emission sources. Spurred by the Inflation Reduction Act, new investment in renewable energy in the US is expected to reach $1.2 trillion through 2035.

While we can expect renewables to eventually become the dominant sources powering the US electric grid, purchasing and correctly accounting for renewable electricity today is a daunting task for commercial and industrial users who are attempting to reduce their Scope 2 emissions. There are two major problems that need to be addressed.

The problem of additionality

The latest data from the US Department of Energy shows that unbundled renewable energy certificates (RECs) accounted for nearly 44% of all voluntary renewable electricity purchases, followed by power purchase agreements (PPAs) at 30%. Both categories are growing rapidly with increasing demand. While buyers receive a REC for each MWh of electricity purchased in both cases, the underlying transactions are very different.

Source: US DOE - Status and Trends in the U.S. Voluntary Green Power Market (2021 Data)

Unbundled RECs are simply the environmental attributes of renewable electricity separated from the power itself and sold separately, typically from existing power generation projects that are already operational. PPAs on other hand are long-term contracts to purchase electricity from a specific provider where purchasers receive RECs but may or may not receive physical electricity delivery. The PPA purchaser is taking on some of the project risk with a long-term commitment to physically or virtually purchase the power produced.

From the perspective of accounting standards such as the GHG Protocol Corporate Standard (under market-based accounting) or emission reduction protocols like the Science Based Targets initiative, what matters is that the purchasers receive RECs in both cases and can therefore claim zero emissions for each MWh of power (from whatever source) that is matched by a REC purchase. However, purchase of unbundled RECs from existing projects does absolutely nothing for climate change mitigation since these projects are already viable and operating with or without the sale of RECs.

The price of RECs in the voluntary market has been highly variable lately but is not significant or reliable enough relative to the price of power and other financial inputs. Revenue from RECs is unlikely to incentivize additional renewable capacity in the future. Unbundled RECs are therefore not additional.

PPAs on the other hand have a reasonable claim to additionality because those projects would not have happened in most cases without the financial commitments of the purchasers. This distinction hasn’t stopped large companies like Intel, Microsoft and others from purchasing unbundled RECs to “reduce” emissions and meet climate targets using market-based accounting. Bloomberg reports that S&P 500 companies bought 32.7 million MWhs of unbundled RECs in 2020, nearly 40% of the RECs sold that year.

Source: EPA — Introduction to VPPAs

The problem of location and timing

The second problem with any renewable electricity purchase is that the location and timing of the power generation may be very different from the location and timing of the power consumption. Matching a MWh of generation by way of a REC (unbundled or obtained through a PPA) to a MWh of consumption — as companies are doing today using market-based accounting — says nothing about the net reduction in emissions as a result of the purchase.

The graphic above from the Energy Information Administration shows a recent hourly mix of energy sources in the US Northwest as an example of how energy sources and therefore emissions could vary by location (which could be a large grid region or a smaller balancing authority), time of day and time of year. Given different electricity generation patterns at the generation and consumption sites, the emission calculation would have to consider the avoided emissions when and where the renewable energy was added to the local grid, and the induced emissions when and where the energy was consumed. The difference between these two emissions would give us the net emissions from the purchase of a MWh of renewable energy (this is the so-called impact accounting method proposed by WattTime and illustrated below).

Source: WattTime — Accounting for Impact

Google, for example, has been matching 100% of its global electricity use with renewable energy purchases since 2017 but recognizes that 100% renewable does not mean zero carbon. The company’s next step is to run on 24x7 carbon-free energy by 2030 by matching hourly electricity consumption with clean energy. The Energy Information Administration publishes hourly data on energy sources used in each grid region or balancing authority, net generation, demand and other factors that can be used to calculate fine-grained emission factors for a given location and hour.

The path forward

Of the two problems identified here, the location/timing issue is a technical problem that can be solved using the available grid data. There are already solutions emerging that could robustly address this, and standards could be updated to support a more fine-grained and real-time emissions model.

Additionality, however, is a more serious issue where the only viable solution is to move away from unbundled RECs derived from existing power generation projects and require that new renewable energy be added to the grid before any Scope 2 emission reductions are claimed.

PPAs are the main voluntary instrument through which new renewable energy is added to the grid today. PPAs, however, can be complex and require significant expertise to put together. Companies will also need to be able to absorb various risks associated with them, including power price risk and counterparty credit risk. This largely rules out small and medium-sized companies from participating in renewable energy unless they team up with other companies to create the critical mass needed.

One way to move forward in the next few years is to create emission reduction pools. Companies could work collaboratively in larger groups or consortiums than in silos of their own in order to increase their negotiating power and credit worthiness. Larger companies could also lend their size and capacity to help move smaller companies along in the emission-reduction journey.

One recent example of this is Walmart’s renewable energy accelerator known as Gigaton PPA which puts together aggregated PPAs that Walmart’s suppliers can participate in. Suppliers will eventually be able to claim Scope 2 emission reductions as the PPA projects are implemented while Walmart will benefit from the aggregated Scope 3 emission reductions as a result.

Longer term, if renewables become the dominant energy source on the grid, then RECs (even those from PPAs) might gradually become less valuable and could eventually become obsolete. But as long as renewable electricity is something that consumers seek out and purchase, RECs will be needed to provide assurance that the MWhs purchased came from a renewable source.

The gaps in the process today — establishing the additionality of the RECs and then accurately calculating the net impact of the energy purchase and consumption — need to be filled in and standardized before anyone can legitimately claim emissions reductions from renewable electricity purchases.



Kumar Venkat

Technologist. Scientific modeler. Climate analyst. Founder/CEO at Model Paths. Previously: CTO @ Planet FWD, CEO @ CleanMetrics. https://www.modelpaths.ai