NexSteppe producing dedicated biomass crop at commercial scale
Ballard providing fuel cell power modules to new customer for 8 buses in China

CMU study finds controlled EV charging can reduce generation cost, but at greater health and environmental costs depending upon the generation mix

In a study focused on the PJM portion of the US electricity grid, researchers at Carnegie Mellon University (CMU) found that although charging electric vehicles at night (when electricity is cheap and wind power is typically more plentiful) could lower electricity costs, doing so also creates more air emissions, and that the health and environmental costs from these emissions outweigh the electricity cost savings. A paper describing the work is published in the ACS journal Environmental Science & Technology.

Results from the study also suggest that with sufficient coal plant retirement and sufficient wind power, controlled charging could result in positive net benefits instead of negative. The result of the analysis depends on the details of the region, notes CMU Professor Jeremy Michalek, corresponding author—i.e., other parts of the US and the world could be different. The question of electricity costs vs. health and environmental cost is important to ask everywhere, Michalek said.

Reduction in annual generation cost and external emissions costs due to controlled charging compared to uncontrolled charging ($2010). Stacked bars show the change in generation cost combined with the median damages by pollutant assuming the 2010 social cost of carbon given by the Office of Management and Budget ($31 in $2010).

Black dots show the change in net social benefit due to controlled charging with error bars representing a 95% confidence interval. Credit: ACS, Weis et al. Click to enlarge.

Although electric vehicles have lower tailpipe emissions than gasoline powered vehicles, the changes in emissions associated with vehicle electrification on a life cycle basis will depend on the emissions associated with the operations of the power plants used to charge the battery. Power plants currently produce 71% of national SO2 emissions, 1% of primary particulate matter emissions, and 14% of NOx emissions, which cause their own set of health and environmental problems. SO2 from power plants is a particular concern, as SO2 is a precursor of particulate matter. Power generation also accounts for over 40% of GHG emissions. Electric vehicle charging may affect these trends. Finally, the additional electricity demand from charging vehicles will affect the operations of the power system and potentially affect the costs of electricity.

—Weis et al.

In the study, the CMU team evaluated the economic, environmental, and health costs and benefits of controlled electric vehicle charging in the PJM interconnection. The PJM Interconnection is a regional transmission organization (RTO) that coordinates the movement of wholesale electricity in all or parts of Delaware, Illinois, Indiana, Kentucky, Maryland, Michigan, New Jersey, North Carolina, Ohio, Pennsylvania, Tennessee, Virginia, West Virginia and the District of Columbia.

PJM the team noted, is an “interesting” power system to examine, they said, as it is the largest independent system operator in the United States by population and has a large installed coal capacity.

The team developed a unit commitment and economic dispatch model to estimate the operation costs and the air emissions externality costs attributable to new electric vehicle electricity demand under controlled vs. uncontrolled charging schemes.

Unlike earlier studies, the CMU team combined detailed modeling of the operating constraints of the electric grid with an estimate of the environmental and health damages from the additional emission due to vehicle charging, in addition to evaluating the change in operating costs.

The researchers used five different scenarios to investigate how different factors will affect emissions and the costs of charging:

  1. Base Case: In this scenario, they assumed an electric vehicle fleet based on the PHEV35 model in GREET10 (similar to the Chevy Volt) and a fleet of power plants representing the PJM system in 2010.

  2. Small Battery: For this scenario, they modified the base case so that the vehicle fleet is based on the Toyota Plug-in Prius.

  3. Large Battery: For this scenario, they modified the base case so that the vehicle fleet is based on the Tesla Model S.

  4. Future: For this scenario, they modified the base case to model a power plant fleet in 2018 by accounting for planned new power plant construction, plant retirement, and updated emissions rates and marginal generation costs.

  5. High Wind Future: In this scenario, they modified the future case to add wind plants sufficient to produce 20% of generation.

For each scenario, they evaluated uncontrolled electric vehicle charging, in which drivers plug in their vehicles immediately after the last trip of the day, and controlled charging, in which vehicle charging is optimized to minimize the cost of generating electricity.

They found that controlled electric vehicle charging can reduce associated generation costs by 23%−34% in part by shifting loads to lower-cost, higher-emitting coal plants. However, this shift results in increased externality costs of health and environmental damages from increased air pollution.

The net implication is that controlled electric vehicle charging creates negative net social benefits in the recent grid scenarios but might produce positive net social benefits in a future grid with sufficient coal retirement and wind penetration. This finding is robust to uncertainty in vehicle adoption patterns, transmission constraints, reserve requirements, fuel prices, and air emissions implications.

In general, controlled charging has potential for reducing generation costs, but its net implications depend on the characteristics of the power plant fleet. In other regions with tighter environmental regulations, more renewable generation, less coal power, and/or inexpensive natural gas plants, controlled charging could lead to lower environmental and health damages. Our results also suggest that the externality costs missing from the current power system operations based on generation cost minimization are substantial and should be considered when making policy decisions to avoid large increases in human health and environmental costs.

—Weis et al.


  • Allison Weis, Jeremy J. Michalek, Paulina Jaramillo, and Roger Lueken (2015) “Emissions and Cost Implications of Controlled Electric Vehicle Charging in the U.S. PJM Interconnection” Environmental Science & Technology doi: 10.1021/es505822f



This is nothing like a show stopper, and simply points to the need to clean up the grid.

It does however show that claims to be 'running a BEV on sunshine' are wholly disingenuous/innumerate, nightworkers apart.

One solution might be to store the energy in batteries overnight, but the loses to and from batteries are usually higher than might be hoped, and seem to normally be in the range of 10-20%, so the overall efficiency pending improvements in that is not great.

Aside from the fuel cell car alternative, whose real emissions relative to BEVs since the time of day is not important are clearly reduced if the coal burn in the grid are taken into account when BEVs are really charged, then a proper costing should give a boost to covered solar parking at work, which really would provide low emission charging for cars, in addition to reducing load anyway by providing shade and cutting transmission losses.


There are several scenarios here:

a: You have a lot of solar (Germany / California)

b: You lave a lot of wind (Denmark, Ireland etc.)

c: You don't have much in the way of intermittent renewables.

In the cases a; and b: you need weather forecasting and hence renewables forecasting. With this, you can predict when you will have the lowest emissions electricity over the next 48 hours. In some cases, you will be able to schedule your charging to match this.

In all other cases, you may as well just go for lowest cost charging.

In reality, this will mean charging when you get home in the evening, early in the night or late in the night, unless you have charging facilities at work.

Then, you have the matter of cost: Lets say a standard unit is 20 cents and a night rate one is 10 cents - what rate do they charge for a "low Co2" unit ?
20 ?, 11, 9 ?

If they charge 11 or 9, you might as well always charge for lowest CO2, if they charge the full price, you might have to think a little.

However, the cost of electricity is so low (compared to gas) that many people would charge for lowest emissions anyway, especially if they were given stickers for every 25 nights they charged for lowest co2.

You should be able to set up a smart charger to charge for lowest cost, lowest emissions or a blend of the two as long as it has access to emissions and price forecasting.

European or international standards would be nice in this case.


For wind the big problem is intermittency, so that much of the power is delivered in gales etc.

So Spain for instance with a nominal ~10% of the grid supplied by wind, on occasion has MORE than 100%, which gets wasted.

Since it also of course has long slack periods, the so-called 'back up' runs a heck of a lot of the time, as it does in Germany.

Storing it as hydrogen is one solution.


Replacing ICEVs with FCEVs and BEVs will not be enough to reduce GHG and pollution if the old CPPs are not progressively closed.

Since NPPs (as the ideal clean base load energy producers) are way too costly ($0.16+/kWh) and no longer accepted by the population; Hydro, Geothermal, Wind and Solar will have to be further developed to replace (1) CPPs and latter (2) NGPPs.

Lower cost large energy storage systems will be required. The H2 avenue seems to be inevitable until other technologies are developed.

Nikita Sidorov

The answer is simple - emissions need to be taxed. Want to burn coal - pay a tax. Want to burn gasoline - pay a tax. And then let the market optimize for the lowest cost that includes the cost of emissions.

To make it fair, the government should take all such collected taxes and split it equally among residents. That way those who are polluting less are rewarded.

By the way, today is exactly one year that my province of Ontario is coal free:

Nick Lyons

Nuclear baseload replacing coal is the obvious, rational solution. Other countries (China, India) will show the way forward, since we are still impaired by irrational anti-nucear hysteria and misinformation.


Fast neutron, burn the waste.


US stats seem to indicate that CPPs are not generally capable of being quickly [large scale] ramped down, then back up again.
Apparently higher quality coal helps with this - as does having part (pot/s?) of the generating plant already operational.

Having extremely dirty brown coal electricity generation in the Victorian state of Australia I had presumed that with continuously burning our 'peat' we had already produced the CO2, & other byproducts.
(We do condense the steam generated to drive the turbines, however, in order to save water :-)

If, in fact, the coal is still burned, then might one argue that only a marginal cost is involved?
For 'spare' generation capacity, perhaps no additional health hazard?

In Victoria any spare [overnight] electricity generation (at the cost of further water usage) might 'fuel' EVs?

As a newbie, I would welcome comments on this.

Henry Gibson

Coal can be combusted without any particles produced. The main irrational bias against coal is that more CO2 is produced per unit energy, but people do not drive automobiles acting as though they know that more than twice the CO2 is produced if the speed they drive between towns on motorways is double; they also do not know all of the CO2 produced during refining and all transportation of crude and products as well as the CO2 produced in flaring of gas at the well and the CO2 produced from the decomposition of spills in the production areas. People will drive as they will but they will irrationally demand that other people not have affordable electricity for their homes from fossil fuels. Air with CO2 is not dirty, it comes from all plants and animals including humans. Nothing that the UK or the US or the EU or Japan can do now can have any noticible effect on the reduction of CO2 in the air by reducing the consumption of fossil fuels because of the vast consumption of the rest of the world, but they all can erect new nuclear electrical facilities not to reduce CO2 but as an example to China and India and other industrializing countries. A very safe nuclear energy source that also does not reduce CO2 is being demonstrated on Mars in the large rover which is Mar's first fully nuclear powered automobile. The prior rovers were heated not powered by nuclear energy because and they had to idle during the cold Mars winter with not enough sun to heat or run. The 80 year half life isotope weight 238 gives off about half a watt of heat per gram. Three times as much electricity could have been had from the 238 on Mars if Stirling generators had been used rather than thermoelectric elements as in the past. Both could have been used with heat pipes to move the heat efficiently. Nuclear radiation from isotope 238 is nearly 100 percent absorbed inside and a thin coating of tungsten can absorb the rest but the metal is very hot and would make a wonderful source of heat for a stirling generator. Isotope 238 can be produced in all nuclear reactors with slight modifications. CANDU heavy water reactors would be very good. They can also burn used light water fuel if diluted with thorium. Isotope 238 cannot be used as nuclear explosive, and would be a good way to produce constant energy when cheaply produced. All of the US and russian owned 238 should be in alaska or antartica or siberia. ..HG..

The comments to this entry are closed.