Continental Airlines, Boeing and GE Aviation Plan Biofuel Flight Test in 2009
Coca-Cola Bottling Company and Duke Energy Partner on PHEVs

ORNL Study Explores PHEVs’ Impact on Power Generation Requirements

Sum for all 13 regions of projected 2030 generating capacity (top left), base generation (top right), and new generation dispatched (bottom) to meet demand for each PHEV recharging scenario. Click to enlarge.

A recent Oak Ridge National Laboratory (ORNL) study examines how an expected increase in ownership of plug-in hybrid electric cars and trucks could affect regional power generation requirements depending on what time of day or night the vehicles are charged. The researchers concluded that supporting a 25% market share of light-duty (cars and SUVs) PHEVs in 2030 could require either major new power generation resources or no new resources at all, depending on when people recharge.

The ORNL study also factors in the impact of the different options for connecting vehicles to the grid. As Dr. Mark Duvall at EPRI has outlined, at 120 volts AC, a 15 amp circuit would be about a 1.4 kW load, while a 20 amp circuit would be about 2 kW. Using a 208/240 volt and 30 amp circuit instead, the load could be as much as 6 kW.

In the worst-case scenario—if all PHEV owners charged their vehicles at 5 p.m., at 6 kW of power—up to 160 large power plants would be needed nationwide to supply the extra electricity, and the demand would reduce the reserve power margins for a particular region’s system.

The best-case scenario occurs when vehicles are plugged in after 10 p.m., when the electric load on the system is at a minimum and the wholesale price for energy is least expensive. Depending on the power demand per household, charging vehicles after 10 p.m. would require, at lower demand levels, no additional power generation or, in higher-demand projections, just eight additional power plants nationwide.

In aggregate, the model predicts an increase in demand, generation, electricity prices, and emissions from the utilities created by the introduction of PHEVs. It also suggests that by 2030 almost all regions (10 out of 13) will need to add capacity to provide for charging PHEVs, mostly in the scenario where PHEVs are charged at 6 kW in the evenings. In all likelihood, to avoid these problems the utilities in the regions would expand their capacity, increase their imports, or establish demand response programs beyond the level that NEMS had calculated, but these factors were not modeled in the scenarios.

—“Potential Impacts of Plug-in Hybrid Electric Vehicles on Regional Power Generation”

Some assessments of the impact of electric vehicles assume owners will charge them only at night, said Stan Hadley of ORNL’s Cooling, Heating and Power Technologies Program.

That assumption doesn’t necessarily take into account human nature. Consumers’ inclination will be to plug in when convenient, rather than when utilities would prefer. Utilities will need to create incentives to encourage people to wait. There are also technologies such as smart chargers that know the price of power, the demands on the system and the time when the car will be needed next to optimize charging for both the owner and the utility that can help too.

—Stan Hadley

The paper also compares the fuel use, emissions, and cost of using a PHEV versus a hybrid electric vehicle (HEV) for each region, with both vehicles going 20 miles per day and the efficient HEV getting 40 miles per gallon, in the different scenarios.

In both of the 2020 scenarios and the night scenario in 2030, CO2 emissions are higher with PHEVs than with efficient HEVs. Coal and oil generation is sufficient to raise CO2 emissions higher than if the vehicles had used gasoline. However, in the 2030 2 kW evening scenario, CO2 emissions are lower, because higher-efficiency combined cycle plants and gas-fired turbines with relatively low emissions are used to meet the added demand.

Even though NOx emissions at first calculation are much higher for much of the country, these will be offset by reductions elsewhere. Even though new power plants are very clean, the plants that are on the margin and are run because of the extra demand may actually be older plants with higher emissions. SO2 emissions are likewise positive, but will be offset by reductions elsewhere in the electric system because of the legal caps on total emissions.

Costs are much lower with the PHEVs, from 22% to 42% of the gasoline cost, depending on the scenario. Electricity is likely to be much cheaper than gasoline, though this does not take into account the added initial cost of the plug-in capability for the vehicle.

—“Potential Impacts of Plug-in Hybrid Electric Vehicles on Regional Power Generation”

The researchers analyzed the potential impacts of PHEVs on electricity demand, supply, generation structure, prices, and associated emission levels in 2020 and 2030 in the 13 regions specified by the North American Electric Reliability Corporation (NERC) and the US Department of Energy’s (DOE’s) Energy Information Administration (EIA), and on which the data and analysis in EIA’s Annual Energy Outlook 2007 are based.

The estimates of power plant supplies and regional hourly electricity demand come from publicly available sources from EIA and the Federal Energy Regulatory Commission. Electricity requirements for PHEVs are based on analysis from the Electric Power Research Institute, with an optimistic projection of 25% market penetration established by 2020 and continuing at this percentage level through 2030, involving a mixture of sedans and sport utility vehicles. Even though the market share remains the same, the increase in the fleet size results in larger annual sales of PHEVs post-2020.

The calculations were done using the Oak Ridge Competitive Electricity Dispatch (ORCED) model, a model developed over the past 12 years to evaluate a wide variety of critical electricity sector issues. Seven scenarios were run for each region for 2020 and 2030, for a total of 182 scenarios. In addition to a base scenario of no PHEVs, the authors modeled scenarios assuming that vehicles were either plugged in starting at 5:00 p.m. (evening) or at 10:00 p.m.(night) and left until fully charged. Three charging rates were examined: 120V/15A (1.4 kW), 120V/20A (2 kW), and 220V/30A (6 kW).

The analysis, the authors note, uses simplifications in modeling electric sector supply and demand. The model applies rigid recharging schedules to all vehicles in each scenario, it does not adjust electricity supply to meet increased demand, its inventory of supply is based on results from a different model (the National Energy Modeling System, or NEMS) that simulates supply and demand somewhat differently, it does not model the transmission system, and it does not reflect all the complexities of air emissions regulations.

However, it does offer insights into the issues involved with PHEVs and the electric grid. PHEV penetration of the vehicle market will potentially create a substantial change on the electric grid. By evaluating these issues early, DOE will be able to help utilities, manufacturers, and regulators understand the issues involved, suggest ideas that will better optimize the combined system, and help avoid negative consequences.



Harvey D


I've seen a shopping centre equipped with a few similar parking places on our recent trip to Argentina. There was an extra charge equivalent to $1.10 USD for 4 hours.
It was worth that much just to keep the car out of direct sunshine when the temperature was at 32C+.

This idea should be very popular when PHEVs and BEVs sell in high volume.

Justin VP

This is another good study identifying the potential challenges of a shift in our automotive fuel sources. Identifying issues, including baseline scenarios, is necessary to ensure a smoother and lower cost transition.

Supporters of green power shouldn't see these studies as a threat, but rather an encouraging sign that the potential of mass-marketed plug-in vehicles is being taken seriously.


Southern California Edison has baseline which is usually set low, so that most homes go over it even in low use times. The next level charges more and the next level beyond that even more in the summer.

The latest offer was to put a control box on your AC, where they can allow it to turn on or not. This is what they call modern.


The big question is to what extent will we need to switch to solar electricity during this time frame? Natural gas is peaking, coal has CO2 problems, nuclear has image problems, oil is gone, wind is variable. A substantial reliance on solar would likely invert our electrical price schedule, with electricity being more expensive at night than during the day. In that case the whole analysis in this article is backward.


I agree that this and similar studies should be read as stimuli and informing discussion of the subject.
I'd like to weave a thread or two amongst this framework to show one alternative version

Another analysis of the grid could show the cost of discounted off peak, bulk commercial discounted as well as likely write off dumping as under performing sectors of the grid. That is to say put a money (or Energy $) value of the loss in the debit column rather than see it as an uncounted loss.
Some investment on the problem may be required to rationalize the current situation.
There would be winners and losers and as usual those who stand to suffer a potential economic loss will lobby.
On the other hand the Rationality of the system will produce winners to the value waste reduction. We all love numbers so lets be brave and just throw one up for starters ~ 20% ? My feeling is that this is a way low figure when it is understood that grids prefer around 10% above expected high demand.(peak times)

It is probably easier to put a 'no or low value' tag on a perfectly useful product than explain the 'loss' of billions of dollars worth of stock.
Efficient reclamation through utilization of this resource may come via any number of innovation incl ph/bev this can be priced differentially and controlled by mobile phone networks for 80-90% coverage globally. Or( Whatever happened to Internet over the grid?)
The possibility of price signalling could be applied as a hedge against grid stress may then be fully integrated to daytime loads.
Consumer preference and budgets can be accommodated via differential pricing with greatest savings occurring at times of surplus.
Wind solar and other renewables could be absorbed by such a system. Guaranteeing and encouraging utilization of renewables through pricing structures and integration.
This is an example of myth debunking in this xample the myths may be.
1; That renewables cannot supply baseload. Answer : Change the definition of 'baseload'.
2; The control of such a system would be too hard or expensive. Answer: (without exploring all options showing merit) The phone co's already price their services on a differential system with the customer able to choose from a variety of budget and convenience options.
The Infrastructure for wireless signalling already exists for most consumers and is probably under utilized hence wasteful. The opportunities for rorting (stealing)can be low. System robustness is high. The power cos could use roaming or other (mobile to mobile) links at small cost to maintain connectivity.
The smarts may sit on the vehicle, on the meter point in the domestic or industrial location and may discuss the billing arrangements between themselves. This would allow cashless open access.
3; that consumer are resistant to change: Answer pricing signals and convenient operating sytems will enable more consumers including those who chose independance via stand alone renewable sytems that integrate with the nationnal infrastructure.

My earlier point about miracles , the impossible etc is that All sorts of things are possible but prettywell nothing will work in an environment as constrained and pressurised by the (technically irrelevant) demands imposed by pre existing interests. So moving convention (ways of doing business) is one huge inertia that stands squarely in the path of achieving identified savings.
Of course people behaviors are another seemingly insurmountable obstacle - if it's allowed.
Now regarding the part where the miracles come in well that takes a little longer.


You can make up NG with SNG from biomass. Concentrated solar cells have reached over 35% efficiency. Wind turbines are popping up and with pumped hydro, you can use the energy any time. I think we have a lot of solutions to use, the question is whether we will use the capital to create them.


"The calculations were done using the Oak Ridge Competitive Electricity Dispatch (ORCED) model, a model developed over the past 12 years to evaluate a wide variety of critical electricity sector issues."

So they have this model you see, and they get a grant to do a study. So they use the computer model they have and publish their grand analysis. Never mind that the accuracy of any computer model depends first and foremost on the assumptions made. Assume the power meters are the same as the ones invented in 1888 by Elihu Thomson (no computers needed here; no time of day-dependent rates, etc.) and the model takes garbage-in and gives garbage-out.

Assume instead that time of use metering is used to make people pay more for electricity that costs more to produce, and voila! The problem is pretty much solved with a lot less hand wringing.


It sounds easy, but who is going to replace 5 million meters in Southern California? If each meter costs $1000 to replace with parts, labor, networking, computer systems and all other costs, that is $5 billion dollars for one utility. That is the price of 5-10 very large power plants. We can all say it can be done, but who is going to do it and who is going to pay for it? Ultimately it is the consumer that pays for it, but are they willing?

Bill Young


In their major energy white paper last year, the British government found that the single most cost effective means to reducing CO2 emissions was smart metering of business and providing feedback to that business. See graph on page 286 and discussion in section 2 of this document:



I am referring to the situation in the U.S. Privatization ran rampant in the late 90s and the electric utilities were blasted in the process. Most people know what a disaster Enron and deregulation were in California. It was open season for scam artists.

Since the utilities got so badly burned there, they are less likely to make investments that could lose money or strand capital. A smart meter system sounds good, but if it only benefits a few PHEV owners, do not expect all the tax payers to pay for it.

I would say that if you were thinking of buying a PHEV, you should talk to your electric company first. Since your hopes are that you will get a lot of your propulsion power from them, that would be a good idea. A lot of people seem to think that this is like buying an appliance, you just guy it and plug it in. It is more like converting from gas to electric cooking. The electric utility company wants to know about this.

At that time you and they could work out a plan for time of use and smart meter. You would pay up front and and amount per month for the change in meter and of course they would have to make the investment in the central computer system. This is where the rub comes in. In the U.S. the Public Utilities Commission may not authorize and expenditure for a central system that would only benefit a few but is paid for by many.

Mick Cowles

When I lived in the southern US, air conditioner compressors had the potential to bring the grid down. Under a demonstration project, we got a discount on our electric bill by allowing PEPCO to install a radio controlled switch on our a/c compressor to keep all the a/c compressors from all coming on at exactly the same time. The effect was imperceptable on the air conditioner's performance. Perhaps a similar technology might reduce the need for new generating capacity that would otherwise result from the increased demands of PHEVs and EVs.


Mis-stateth John Taylor:

Line voltage drops with increasing load, and increases with increased power generation.
Actually, in an AC system the frequency drops with increasing load; voltage is determined by reactance, not power.  Counterintuitive but true.



You are right, heavy surge current devices can be controlled that way and have been for quite a while in commercial buildings. There is no technical reason why similar things can not be done in residential.

If I were a local electrical power provider, just to increase the life of the transformers in a local neighborhood grid, I would want to implement such a program. It makes economic sense to do this.

However, that works with ACs that turn on and off a lot. With BEV, PHEVs the timer might be set to come on at midnight and turn off at 6 am, or when charged. This represents an almost continuous load (charge is heavy at the start and tapers at the end) and you could imagine 10,000 4 kw loads all switching on in a city at midnight.

If that is the case, I would want to know how many there are, where they are located and make sure that they do not all turn on at the same time. Technical people will tell you that this can be done, but I have not heard the plans to do this by the car makers nor the electric utility providers.


Until there is a substantial number of PHEVs in the market consumers will trickle charge at night at one of two voltages (110 or 220V 3ph). Aftermarket sales of "smart" chargers can utilize Arnold's price signaling concept provided the utility is made to offer variable rates and energy sources.

The consumer signs up to a rate scheme that benefits both the util (more off peak charging) and the consumer (lower rates, daytime charge, renewables only). More variables can be added with time such as charging with only renewable energy, charging with renewable energy from a local/municipal solar/wind farm, etc.

There is danger in resting too much authority in monopolies. Utilities need to closely watchdogged and forced to buy energy from independent producers and resell that energy to consumers. Non-centralized grids are our best defense against system failure and security concerns.

Andrey Levin

From EP comment:

“Actually, in an AC system the frequency drops with increasing load; voltage is determined by reactance, not power. Counterintuitive but true.”

From Wiki:

“A power outage may take one of three forms:
2) Brownout
where the voltage level is below the normal minimum level specified for the system. Systems supplied with three-phase electric power also suffer brownouts if one or more phases are absent, at reduced voltage, or incorrectly phased. Such malfunctions are particularly damaging to electric motors. Some brownouts, called voltage reductions, are made intentionally to prevent a full power outage.”

Reduction of residential grid voltage down to 190 Volt (from standard 220) was regular occurrence in last years of former USSR, take my word for it. Grid frequency very rarely dropped for less than 49.8 Hz from standard 50.


Distributed generation does have its benefits, but I have not seen any grand plans for our brave new future. If everyone just does their own thing, that can lead to problems.

Utilities know that they are in a new market environment and that they have to adapt, but after many decades of the same ways, it is difficult. It seems like a good item for a true national energy policy to get all interested parties together for some planning soon.


Yes, Andrey.  Now ask yourself how the voltage was lowered.  Hint:  it wasn't an automatic effect of increased load, it was probably done by changing transformer taps.  The decrease in grid frequency was a direct effect of load; parts of the grid which draw power from others have their phase lag behind.


This study seems to confirm the obvious. How much were they paid?

I assume vehicles will be fitted with GPRS devices and be able to negotiate with the utilities a convenient download time (or up load if needed).

Kit P

In the North America, grid frequency is maintained constant at 60 hz. It is a mechanical thing. Large Steam turbines are designed to avoid resonance frequencies and would be subject to fatigue if operated for extended times at other than design rpm. Loss of load or turbine trips can result frequency deviations which are corrected quickly with control circuits.

Voltage is also maintained constant at the generator using voltage regulating controls. Again, upsets on the grid can have very short effects while the control system is responding. If the response is not fast enough the lights go out to prevent destroying equipment.

The Russian grid is very different from North America. The grid in North America is essentially infinite while Russia has several smaller systems and low reserve capacity. As Andrey suggested, Russia has a sever shortage of electricity and it is projected to get worse. Many of Russia's fossil plants are very inefficient (e.g., 17%) because of huge reserves of natural gas. Russia has a very aggressive nuclear construction program so the can sell the natural gas to the Germans who want to shut down nuclear. Go figure.


I wrote a long rebuttal to Kit P's erroneous assertion, but it's stuck in moderation since yesterday.

The comments to this entry are closed.