There is good news for investors on the carbon impact front as trumpeted earlier this month by Bloomberg New Energy Finance (BNEF) at its Future of Energy Summit: 2015 should prove to be a watershed year for the “de-carbonization” of the US power sector, with record volumes of coal-fired capacity to be shuttered, renewables capacity to be built, and natural gas to be consumed. The result: CO2 emissions from the power sector should drop to their lowest level since 1994.1 So how might we persuade investors to capitalize on opportunities in clean energy technology and accelerate the market for green bonds?

CarbonCount™ is a new metric that evaluates bond investments in US energy-efficiency and renewable-energy projects based on the expected reduction in carbon dioxide (CO2) emissions resulting from each $1,000 of investment. CarbonCount™ was developed by the Alliance to Save Energy, a leading US nonprofit organization that promotes energy efficiency. And at the above-named BNEF conference on April 15th in New York City, we took home one of four top prizes in the coveted Finance for Resilience (FiRe) competition where 60 projects vied to demonstrate their “tangible” and “actionable” ability to unlock at least $1 billion of clean energy investment capital over three years. Here’s how it all works.

Quantifying “Green”

The Alliance understands that investors will not properly value carbon impacts until they are confident that those impacts have been estimated impartially and consistently. CarbonCount™ combines forward-looking project data already used for credit ratings, sophisticated emissions modeling software, and clearly documented assumptions to produce a quantitative score tailored for finance professionals. The global market for self-labeled green bonds has burgeoned in recent years with annual issuances rising from $11 billion globally in 2013 to $36.6 billion in 2014, but nearly 40% of these bonds lacked any independent review of their climate impacts.2 Third-party verification systems are available, at no small expense to the issuers, but their findings are not readily comparable between purveyors. CarbonCount™ was designed to provide a quantitative, transparent, easily comparable metric, and to do so at minimal cost.

Independent Engineers’ Estimates and Investment Grade Audits

In the case of renewable-energy projects, predicted monthly energy generation output is customarily included as part of the standard financial underwriting package. CarbonCount™ uses highly conservative production values, allocated using models and region-specific distribution tables developed by the National Renewable Energy Laboratory (NREL). When these data are unavailable, we conservatively estimate on the basis of nameplate generating capacity using the state- and technology-specific capacity factors prepared by NREL.3 In the case of energy-efficiency projects, energy service companies (ESCOs) usually give customers detailed analyses of estimated energy savings in the form of an Investment Grade Audit (IGA). Since ESCOs guarantee savings and thus assume a financial risk, a conservative bias is assumed. Unless hourly load impacts are explicitly identified, savings are allocated evenly across the year.

Quantifying Electricity-Sector CO2 Impacts

This information is combined with another publicly available analytical tool. EPA’s AVoided Emissions and geneRation Tool (AVERT), released in February 2014, is a model intended primarily to help state air and energy officials evaluate the impact of proposed energy-efficiency and renewable-energy policy initiatives.4 The Alliance has leveraged the model, combined with the information referenced above, to measure electricity-sector impacts. AVERT collects operational and emissions data from every fossil-fueled electric generating unit (EGU) in the lower 48 United States with over 25 megawatts (MW) of capacity. It analyzes historical usage patterns of fossil-fueled EGUs, recorded by the hour and grouped into ten regions, to predict future EGU behavior and, thus, emissions. AVERT estimates how each EGU will operate under future regional load scenarios, with hourly granularity. Finally, on the basis of user-defined load reductions for a particular year, achieved either through efficiency or non-fossil generation, AVERT estimates avoided metric tons of CO2 for the modeled year. The AVERT model cannot directly capture the CO2 impacts of fuel-oil or natural-gas usage offset by onsite energy-efficiency improvements, but these are easily calculated using factors developed by EPA.5 These savings can then be added to the results from the AVERT model for a total CO2 impact.

Finalizing the Metric

The final step in generating CarbonCount™ involves apportioning CO2 impacts for the fraction of project capital provided by bonds and then scaling to a common investment size – metric tons of CO2 (CO2e) offset per $1,000 of bond value. Given that AVERT forecasts on the basis of the existing generation mix, we use only one year of CO2 savings as a metric when assessing the bond, even though the capital basis covers the lifetime of the project. Whether the market will share our preference for a robust one-year metric over softer life-time projections remains to be seen, but we note that the German development bank KfW recently employed a similarly conservative approach when it scored its own US dollar-denominated green bonds.6 CarbonCount™ has been tested on five instruments to date:

  • Continental Wind LLC Senior Secured Bond (utility-scale wind);
  • Southern California Public Power Authority’s Milford Phase One Revenue Bond (utility-scale wind);
  • SolarCity Series I LMC 2013-1 Bond
  • Topaz Solar Farms LLC Series A Senior Secured Bond (utility-scale solar); and
  • Hannon Armstrong Sustainable Yield Bond, a representative of governmental ESPC projects.

The results in metric tons of CO2 offset annually per $1,000 bond are as follows:

CarbonCount table

Detailed reports on these bonds are available via the Alliance to Save Energy’s information center. Anyone using CarbonCount™ is advised to examine these reports and carefully consider the primary drivers of the metric (project cost and regionally specific emissions situations).


CarbonCount™ is an intentionally simplified solution to a complex problem, but we believe it provides a low-cost, consistent and quantitative measure of CO2 emissions reductions/offsets per unit of investment that US bond buyers can use today to evaluate the carbon impacts of specific projects. To help investors value the impact of CarbonCount, the Alliance pledges (resources permitting) to evaluate any bond brought to us by issuers in 2015 and certify that projected CO2 savings have been modeled consistently, utilizing comparable energy generation and savings forecasts.


Ken Locklin is an Affiliated Expert with the Alliance to Save Energy. He also serves as a Director of Impax Asset Management (US) LLC.
David Posner is the Financial & Economic Policy Program Manager at the Alliance to Save Energy.
2 “Bonds and Climate Change: The State of the Market in 2014,” Climate Bonds Initiative (July 2014), 6.
3 “US Renewable Energy Technical Potentials: A GIS-based Analysis,” National Renewable Energy Laboratory (July 2012), available at
4 “AVoided Emissions and geneRation Tool (AVERT) User Manual Version 1.2,” US Environmental Protection Agency (October 2014).
5 Emissions from fuel oil and natural gas combusted onsite do not vary by location or time of use; see
6 “Green Bonds – Made by KfW,” (September 2014); see