Marginal Grid Analysis Algorithms

WattTime is founded on new marginal emissions analysis technology that makes it possible to identify what times the electricity is clean or dirty.

We upgrade our software periodically to incorporate the latest research breakthroughs. As of our April 2016 upgrade, WattTime is based on a combination of the methods of Rogers et al (2013) and Callaway et al (2015).

How Better Accounting Can More Cheaply Reduce Carbon Emissions, by J. LaRiviere, G. McCormick and S. Kawano 2016

Much of corporate renewable energy carbon accounting today relies on an analytical technique that the Greenhouse Gas Protocol notes is a significant simplifying assumption. We show with a simple example that this widespread accounting practice can mismeasure carbon savings by up to 45 percent. Recent advances in estimating the emissions foregone from renewable energy generation have made significantly more accurate accounting now entirely practical. There are environmental and financial reasons why the most accurate emissions accounting would be socially valuable--it would allow for locations which offset more carbon to be identified and thus receive more investment.

Location, location, location: The variable value of renewable energy and demand-side efficiency resources, by D. Callaway, M. Fowlie and G. McCormick 2015

Greenhouse gas mitigation efforts in the electricity sector emphasize accelerated deployment of energy efficiency measures and renewable energy resources. We evaluate renewable energy (RE) and energy efficiency (EE) technologies across regional power systems in the United States in terms of carbon dioxide emissions displaced, operating costs avoided, and capacity value generated. We estimate that external, emissions-related benefits account for between one quarter and one half of the total value generated per MWh over our study period. Regional variation in these emissions benefits gives rise to economically significant, regional differences in second-best production subsidies. This variation is not reflected in the prevailing policy incentives that currently guide new investments.

Spatial and temporal heterogeneity of marginal emissions: Implications for electric cars and other electricity-shifting policies, by J.S. Graff Zivin, M.J. Kotchen and E.T. Mansur 2014

In this paper, we develop a methodology for estimating marginal emissions of electricity demand that vary by location and time of day across the United States. The approach takes account of the generation mix within interconnected electricity markets and shifting load profiles throughout the day. Using data available for 2007 through 2009, with a focus on carbon dioxide (CO2), we find substantial variation among locations and times of day. Marginal emission rates are more than three times as large in the upper Midwest compared to the western United States, and within regions, rates for some hours of the day are more than twice those for others. We apply our results to an evaluation of plug-in electric vehicles (PEVs). The CO2 emissions per mile from driving PEVs are less than those from driving a hybrid car in the western United States and Texas. In the upper Midwest, however, charging during the recommended hours at night implies that PEVs generate more emissions per mile than the average car currently on the road. Underlying many of our results is a fundamental tension between electricity load management and environmental goals: the hours when electricity is the least expensive to produce tend to be the hours with the greatest emissions. In addition to PEVs, we show how our estimates are useful for evaluating the heterogeneous effects of other policies and initiatives, such as distributed solar and real-time pricing.

Evaluation of a rapid LMP-based approach for calculating marginal unit emissions, by M.M. Rogers, Y. Wang, C. Wang, S.P. McElmurry and C.J. Miller 2013

To evaluate the sustainability of systems that draw power from electrical grids there is a need to rapidly and accurately quantify pollutant emissions associated with power generation. Air emissions resulting from electricity generation vary widely among power plants based on the types of fuel consumed, the efficiency of the plant, and the type of pollution control systems in service. To address this need, methods for estimating real-time air emissions from power generation based on locational marginal prices (LMPs) have been developed. Based on LMPs the type of the marginal generating unit can be identified and pollutant emissions are estimated. While conceptually demonstrated, this LMP approach has not been rigorously tested. The purpose of this paper is to (1) improve the LMP method for predicting pollutant emissions and (2) evaluate the reliability of this technique through power system simulations. Previous LMP methods were expanded to include marginal emissions estimates using an LMP Emissions Estimation Method (LEEM). The accuracy of emission estimates was further improved by incorporating a probability distribution function that characterize generator fuel costs and a membership function (MF) capable of accounting for multiple marginal generation units. Emission estimates were compared to those predicted from power flow simulations. The improved LEEM was found to predict the marginal generation type approximately 70% of the time based on typical system conditions (e.g. loads and fuel costs) without the use of a MF. With the addition of a MF, the LEEM was found to provide emission estimates with errors typically less than 25% for CO2, and less than 50% for SO2 and NOX. Overall, the LEEM presented provides a means of incorporating pollutant emissions into demand side decisions.

Regional variations in the health, environmental, and climate benefits of wind and solar generation, by K. Silar-Evans, I.L. Azevedo, M. Granger Morgen and J. Apt 2013

When wind or solar energy displace conventional generation, the reduction in emissions varies dramatically across the United States. Although the Southwest has the greatest solar resource, a solar panel in New Jersey displaces significantly more sulfur dioxide, nitrogen oxides, and particulate matter than a panel in Arizona, resulting in 15 times more health and environmental benefits. A wind turbine in West Virginia displaces twice as much carbon dioxide as the same turbine in California. Depending on location, we estimate that the combined health, environmental, and climate benefits from wind or solar range from $10/MWh to $100/MWh, and the sites with the highest energy output do not yield the greatest social benefits in many cases. We estimate that the social benefits from existing wind farms are roughly 60% higher than the cost of the Production Tax Credit, an important federal subsidy for wind energy. However, that same investment could achieve greater health, environmental, and climate benefits if it were differentiated by region.


The prevalent soda vending machine industry in the US could yield reductions in energy consumption by addressing operational use. A study by the National Renewable Energy Laboratory estimates that each of the 4.6 million vending machines in the US consumes between 7 and 13kWh per day. Currently, soda vending machines keep their products at a consistent temperature regardless of the time of day. Although no formal soda vending machine usage patterns have been observed, we hypothesize that usage patterns primarily follow time of day with high utilization during midday and afternoon and low utilization during the night and morning. However, soda is generally non-perishable and does not need to be refrigerated during periods of low to no soda demand. In this report, we construct a thermodynamic, state space refrigerator model and integrate a hypothetical soda demand schedule in order to optimize the operation of a soda vending machine that minimizes energy and carbon impact while maximizing the delivery of the appropriately chilled soda.


This project aimed to create a smart refrigerator for both perishables and non-perishables that is capable of using prior usage data and the user’s choice of objectives to optimize its energy usage and performance. Using prior usage data, the refrigerator can predict times at which it is most likely to be opened and its contents consumed, and (when in non-perishables mode) adjust its temperature setpoint accordingly to ensure that it reaches the desired temperature at that time. In addition, an online dashboard allows the user to view sensor data and control the refrigerator mode remotely. In either of the two modes, the refrigerator’s compressor schedule is optimized using a thermal-model-based mixed integer linear program to minimize marginal carbon emissions (using the WattTime API) and electricity cost (using a time-of-use rate schedule).