Many skeptics argue that system capacity constraints will limit the viability of electric vehicles. With all the attention that the industry pays to peak demand concerns, electric vehicles must complicate matters, right? Wrong. When it comes to peak demand, electric vehicles will not have a noticeable impact for 10 or 20 years, even if charging is uncontrolled.
According to a UC Berkeley study that analyzes the effects of plug-in hybrid electric vehicle adoption on system load in California, “1 million compact car plug-in hybrid electric vehicles would not significantly affect the system peak.” We may not see 1 million electric vehicles on the road in all of the United States for 5 or 10 years, so it will be at least 10 or 20 years before we see 1 million in California alone. According to this study, even then, electric vehicles will not significantly affect the system peak. System capacity constraints will become a real issue in 20 or 30 years, but we should not worry about it too much at this point.
A concern that is not overblown is how electric vehicles impact the distribution system. There are two main issues with respect to the distribution system that electric utilities are concerned about. The first issue is obvious. The number one job of electric utilities is to keep the lights on. Electric vehicles may not have much of an impact at a system wide level, but for a given circuit or transformer, a few electric vehicles charging at the same time may lead to an increase in localized outages. Considering that adoption of electric vehicles will be concentrated in certain neighborhoods, electric utilities are concerned about maintaining the same level of reliability.
The second issue is less obvious. According to an EPRI study, concentrated charging of electric vehicles will lead to an increase in transformer degradation. The figure below plots the degradation of a given transformer as a function of the number of plug-in hybrids served. According to the study, “These transformers typically serve 5-7 households.” If a cluster of 5-7 households adopts 3 to 5 plug-in hybrids and charges them at 240 volts, transformer degradation increases precipitously. This scenario will be quite common because adoption of electric vehicles will be concentrated in certain neighborhoods and 240 volt charging will be common (in fact, 240 volt charging will be required for LEAF owners).
Stephen Lacey of RenewableEnergyWorld.com recently interviewed Matt Nielsen, a senior researcher with GE. He emphasizes many of the same points:
“Most people agree that we have enough generation capacity to meet the first wave [of electric vehicles]… One of the key stress points that folks are concerned about are the local points of connection for these vehicles, especially the local area transformers… They are designed such that they may be overloaded during the day, but then they have a cooling period at night so they can decrease their temperature and that doesn’t impact their overall lifetime. So what does that mean? Does that mean we then continue to add load to these transformers when we thought they were going to be cooling down, but now they are not cooling down? Or do we charge them during the day and we add to the peak load that they see, stressing them even more? A lot of the utilities are very proactive. They are already trying to look at where they believe adoption will occur for the electric vehicles and then try to identify those stress points in their distribution system and proactively identify a plan to correct that. So I think, on the utility side, one of the main challenges in on the distribution side.”
Like us humans, transformers need time to cool down. If they are forced to work 24/7, their overall lifetime decreases. This issue, as well as localized outages, are the key challenges that utilities will face as a result of electric vehicles. Concerns about peak demand are overblown. Concerns about the distribution system are real and need to be dealt with proactively.
To read more on how utilities can deal with these challenges, check out this white paper. It highlights some of the electric vehicle work we did at Freeman, Sullivan & Co.
Vlad,
I agree to agree…
Thoughts on your question: Who needs a grid if you have wind and solar at home? The question amounts to: can I afford to go off grid? It is a question of life style and cost. The lower energy life style makes for a cheaper set of generations and storage equipment. Being on-grid as two big potential benefits which lowers cost:
-grid can supply power above the immediate generator output or above beyond battery capacity (possibly eliminating the need for batteries).
-in many locations a utility will pay for electricity sent to the grid.
Note that an off grid house has to be smart to balance generation, keep batteries charged, and use other storage appliances appropriately.
I am strongly in favor of distributed generation, both residential and commercial. However there is plenty of place for large wind and solar farms. A big benefit for distributing wind and sun power is that the average output across a large region should vary less quickly than at a dense solar or wind development making it easier to manage. Relatively localized generation reduces transmission and distribution costs. Another benefit is that home owners are connected to the source of their power, getting more people directly involve in energy issues.
So, I see a big role for utilities going forward, balancing supply with demand, moving power from windy/sunny regions to calm/cloudy areas, and to a house with a high demand (running electric dryer) from a house down the street with excess generation, and investing in big projects such as pumped hydro or compressed air storage. Balancing demand with generation will be more challenging than it is today. The grid will remain.
Paul
I like your question of “who needs the grid”? Not because I think the need for the grid is going anywhere because of smart grid, but the dynamic surely changes.
Today’s grid is definitely not cheap, just in New England we are spending billions on new tramsmission and looking at increasing capacity charges for generation. Why? Because the disparity between the base and peak continues to widen and widen dramatically. In NE we maintain a supply of generating capacity that is twice our typically daily demand i.e. when its neither very hot or very cold. Unfortunately the end use customer is indifferent about their use because their electric rate does not reflect the cost to deliver peak ekectricity.
Now we introduce AMI and real time prices and customers will now be able to see the harm of more expensive times and the benefit or opportunity of less expensive times. Coming back to the grid as a whole, my long term vision (getting through the learning curve for the massess or for masses friendly solutions may take a while) entails consumers being managers of their total energy pie. Their opportunity to save is not just a few bucks on a $60 monthly electric bill, but instead how much does it cost to heat their home with oil or put gasoline in their car per month? Off peak electricity may be less expensive. Meanwhile it may be advantageous for them to sell their on site solar electricity back to the grid during peak periods thus contributing to their net energy cost.
So the grid remains but no longer simplistically as a set of poles and wires to deliver central station power to passive consumers, but instead a flexible and dynamic system that rewards smart consumers and balances all energy sources against each other based on their availability and cost.
I would argue that the national and global passion for renewables as the pathway to addressing climate change and energy independence is all dependent on a smart grid to free up the economics at all levels.
Good news: there is a lot of statistics available on line – the numbers below illustrate the “big picture” of electricity being less efficient than fossil fuels (the balance appears to be fragile …)
Bad news: it appears so that I lost a 1K multiplier in my previous post…
Constants ——————————————-
Energy equivalent 3,412 BTU/kWh
Gasoline energy 114,000 BTU/gal (Wikipedia)
= 33.4 kWh/gal
Diesel fuel energy 138,700 BTU/gal (Wikipedia)
= 40.7 kWh/gal
2008/09 US statistics —————————————–
Electrical utilities:
Total generation 4,119 B kWh (EIA 2008)
T&D losses 10.3 % (tuned to the avg retail)
Utilities’ revenues 363.7 B $ (EIA 2008)
Avg electricity value 0.097 $/kWh (EIA 2008)
Transportation energy 27,800 T BTU (EIA 2008)
= 8,148 B kWh (3,412 BTU per kWh)
Avg cost of gasoline 2.35 $/gal (EIA 2009)
Gasoline energy value 0.070 $/kWh
Avg cost of diesel fuel 2.46 $/gal (EIA 2009)
Diesel energy value 0.061 $/kWh
Vlad,
Comparing transport efficiency of gasoline with electric is more complicated than converting to BTU or KWH. You can probably google some comparisons done people far more expert than I am.
Chevy claims the Volt claims 40 miles per day cost $1.50 where a 40 mpg vehicle would cost about 2.75 in Denver today. Another google hit reports the volt at 230 mpg. Nissan claims the Leaf is 367 mpg. Electric vehicles with regenerative braking have a big advantage over ICE vehicles which turn motion in to hot brake pads. My brother says gets about 40mpg overall in his camry, EPA says the 2008 camry 6cylinder get 19/28, 4 cyl 21/31, hybrid 33/34 city/highway. None of these get to your point, electricity is inefficient at the generation, gas autos are inefficient, but gas water heaters are far more efficient than a typical electric water heater (new heat pump water heaters cost about the same as gas to operation per their EPA info). Not simple stuff.
Also, how do we think about efficiency of “renewable electrons”? Since there is no fuel cost, efficiency seems to be 100% or 0%, 100% when the electrons are used, 0% if they are not used. Thoughts?
Paul, thank you for giving a thought to my poorly defined argument.
It sure does take an expert to consider all things involved.
From the other hand, if the ends do not meet at an abstract level, why should experts bother to argue about the details? I bet you would not talk to an inventor of yet another “Perpetuum Mobile”…
Here, just cannot help throwing in another illustration: during 8 hours of the out-of-peak night time the total output of a 25kVA pole-mounted distribution transformer (serves 5 houses in my area) equals to 6 (six) gallons of gasoline.
Under 50% load (since a substantial 20…30% demand still exists), it will take 4 (four) hours for such a transformer to charge a single Chevrolet Volt for a single 40-mile trip (no energy conversion or storage efficiency calculations were used).
My intent is to demonstrate that the present electrical GTD system as we know it after one hundred, or more, years of optimization around the prevailing energy consumption patterns is not able to deal with electrical transportation, even on a small scale.
Thus, significantly higher cost of electrical transportation will not benefit anyhow the pressing environmental issues…
If we are truly concerned about the environment, we have to either reduce fundamentally our consumption of energy (a defensive approach) or rebuild the entire electrical GTD system (seems to be a more constructive approach).
The latter implies introduction of abundant distributed renewable sources.
How does it comply with the centralized/monopolized GTD structure?
P.S. Just cannot believe that such events as Three-Mile Island, Chernobyl, Hubble, Kursk, Toyota and BP was not enough to disable the nuclear alternative…
Vlad,
I understand your points, I agree with them all, but I see the grid issues as tweaks rather than obstacles. The utilities will have additional constraints (EV/PHEV, renewables, smart stuff) and more revenue potential (EV/PHEV). The distribution networks are continually maintained and upgraded and smart grid is probably a bigger challenge than the electrical load at final transformers as smart grid is a huge change to philosophy vs the engineering problem of transformers.
If I follow your example, the transformer can support 2 EV per night for a full change. The Prius as an example which may apply to EV but not Hummers, sold about 815K vehicles in the US total since 2001 (about 100K in 2009, but fewer in the early years). Wild guess of 1 in 100 EV will overlap transformers, or for 100K EV sold 1K transformers affected per year (for 1 in 10 it is still only 10K/year). I’d guess far more than 10K transformers are already replaced annually. I repeat my point that extensive wind/solar may create new opportunities for off peak rates, particularly from wind as storm fronts pass, extending the cheap charging hours. We can use these when smart metering is in place to better use the renewable energy as it is created, rather than the overnight off peak created because old gas/coal or nuke plants can not be throttled down/up efficiently.
I think we will work both efficiency and renewable generation. New vehicle mileages standards (including big trucks), new building codes, and public opinion are evidence on the efficiency side. Renewable generation is growing and presents its own challenges, including pricing.
Your list of disasters are the normal type, humans design a system which can work, but accidents and human errors, including over confidence, conspire to create a catastrophe. I’m not confident that humans will learn to take the simpler but longer path when there is a shiny new high tech short cut, or a way which can make them tons of money.
check out http://en.wikipedia.org/wiki/Toyota_Prius
The technical improvements from the original to the current model are impressive. Newest is rated at EPA 50mpg overall and has some cool features.
Paul,
Just wanted to share a couple of quotations applicable to the case
Pro — “Progress has little to do with speed, but much to do with direction” (anonymous)
Con — “You can never solve a problem on the level on which it was created” (Enstein)
One of the integration issues with wind generation is that it’s production often happens at night during light load conditions resulting in over generation conditions on the grid. The addition of a large amount of EV or PHEV charging at night will assist in the increased integration of wind resources on the grid.
Warren Frost
Considering the US in 2006, the amount of energy consumed by the means of transportation powered up by gasoline and diesel fuel was 26 trillion BTU, which equals to 7.7 billion kWh. In the same year, the amount of generated electricity was 3.8 billion kWh (minus 8…12% of T&D losses).
Taking into account the limited efficiency of batteries as energy storage and the unideal efficiency of energy conversion from 240Vac to 12Vdc (or 36…42Vdc), the all electric-powered transportation in the US would demand 4 times more electricity than it was produced.
Vlad,
Very interesting big picture comment.
Cars and trucks represent a huge new load on the electric system so we should expect to install more generation capacity to support them. I suspect we will need less than the 4X additional that your calculations suggest due to the following, but won’t guess at the level.
-improved efficiency (hybrids use about 1/2 the gas of regular cars) -new standards are being developed for big trucks
-higher efficiency electrical systems in cars and distribution
As I mention in my previous posts I hope that much of the generation will be from wind, solar, and other variable renewable sources and that vehicles will enable a high percentage of renewable electricity on the grid.
Well, we seem to be in agreement that transportation consumes the lion’s share of energy. Also, let’s agree that renewable sources of energy imply distributed generation. Then, if solar and wind generators in every backyard will produce more energy than delivered now by the centralized GTD system, who will need the grid then? Either smart or dumb…
The addition of electric cars on the load side and wind + solar will have profound affects on the grid. There will be times when wind is strong and will make electricity cheap during the day. Similarly, picture a clear cold day in Phoenix (say in February) cheapest electricity may be at noon.
The grid will be different in the future, but it will take years to build the equipment: cars, wind farms, solar gardens, smart grid /control equipment, pricing models, utility commissions, public awareness, political will.
The benefits to countries which adapt quickly will be big:
-some immunity to oil price shocks, the 2008 oil price spike were a contributor to the current financial ‘situation’.
-chance to export the solution to the rest of the world
-less air pollution and related health and climate
Josh,
Thanks for this perspective on plug in vehicles and their possible effect on the distribution system. I view the Smart Grid as a complete economic and operational transition for the grid. Today the grid is inherently inefficient due to the wide gap between base demand and peak. As more and more homes even in Northern New England go to central A/C, the disparity between base and peak widens. Yet there is no price consequence to the consumer. Since 1980 the annual usage of new generation has steadily declined as each unit is fired less of a given year because much of the capacity has been added to meet peak demand.
As customers begin to react to price signals we can shave the peak thus enhancing capacity utilization and avoiding peak driven transmission investment. This more efficient grid will in my estimation save money across the Board and that those savings are offset by perhaps the the sort of effect that electric vehicles may have on the distribution system, especially in avery localized way.
Very interesting comment, Commissioner O’Brien. Utilities in the north are slowly becoming summer peaking, which presents a number of problems. As you point out, customer price responsiveness is a key part of the solution.
California utilities have been trying to implement price responsiveness for their residential and small and medium C&I customers since the energy crisis in 2001. Although there have been some growing pains, the effort has been quite successful. California utilities are in the process of defaulting all small and medium C&I customers onto critical peak pricing (CPP). San Diego Gas & Electric was the first California utility to make this transition in 2008, and only 30% of small and medium C&I customers opted out after being defaulted onto CPP.
In Pacific Gas & Electric service territory, a voluntary CPP rate was made available to residential customers. In 2008, this program achieved a 16.6% load reduction for the average activation day. Although similar programs would have a different experience in Vermont, it is encouraging to know that residential and small and medium C&I customers are open to participating in these alternative rates, and that they provide substantial load reductions.
Thanks Josh. I am going to have our team here look at the CA experience. Our understanding generally is that widespread use of dynamic pricing has yet taken hold in any utility territory. CPP is a bit different, but it is heartening to see that only 30% of the C&I customer base opted out after implementation of CPP.
Josh,
Good article.
I assume a slightly different transformer design can handle more power (or specific transformers will have shorter life times). Its a cost of doing business (more kWh delivered, more revenue, more expenses). As you say, knowing which circuits are affected and replacing and upgrading equipment will be necessary but it is a well understood problem.
I’m particularly interested in plug in cars as a demand/generation matching tool. The car user and the utility both have needs and so both have inputs to the car charging transaction.
Wind and solar vary, car charging can be prioritized and varied to increase or decrease demand. Variation is biggest for the smallest area. A neighborhood circuit is affected by one passing cloud over the roof top PV arrays. A region is covered by entire storm front, although it has a bigger total effect there is a less rapid regional change.
A few possible charging algorithms, select by car owners. Utility manages demand with pricing signals. Care has to be taken to develop stable management algorithms.
-battery low 1/4, need to be 3/4 full at specific time, pay what it takes to get to 3/4 then only use cheap electrons.
-battery already 3/4, allow charging only at lowest rates
-battery full, allow utility to by electrons at peak price (two way flow requires more expensive equipment, but may get to be the lowest cost of ownership)
-This is similar to filling a car’s tank now. On a long highway trip use the station closest to the exit, pay what they ask. Close to home, use the favorite (cheapest or convenient) station.
Transformers are not like humans, there is no need for cool down. Rather changing temperatures will do more harm then good. I think the transformer are designed to deliver for a short moment an amount of power, without the need for a complicated cooling system to keep thing at a specific regulate temperature. A transformer for a 200 kW might have a cooling system good enough for a continues load of 100 kW, continues loading at 200 kW will drive the temperature to high, giving rise to brownouts/blackouts.
As is usual, lines and capacity are often 2 to3 times higher as stated. So the first couple of EV wont be a problem: the system can handle it. If the loaders are smart, than it is no problem at all, as long as the number of EV is well below the rated power. I think thanks to the continues load, transformers will have a longer life instead of a shorter life.
If transformer would be humans, then they need a cool down and auto repair themselfs. That would be cool. Technique has not become that smart that it will fix itself during a pause. Maybe later.
Josh,
Another good article, I agree with the position in the article. However, I would think from a base loading prospective the producers could see a major change in their normal operations. I am far from being an expert, however looking into the future would not the typical evening to nighttime base load see an elevated output? I guess I am asking how would this effect the normal operation of the base loading facilities…does typical O&M have to change? Would this mean some load would have to (temporally) shift to peaker in the load stack for the O&M to be performed at the base loading facility? I know this might be extreme questions, but unfortunately this could mean higher pricing?
As I mentioned previously, I am not an expert just curious if this could be an issue.
Thanks,
Joe
As I understand the process, in order to install a 240 V charging station, a homeowner will be required to get a permit which will notify the utility of the planned upgrade and they will be able to take appropriate action to ensure that the distribution system is adequate to handle the additional load. Also, the charging stations all have built in demand response capabilities to allow the utilities to shift demand out of peak periods
Households who want to start charging should request an upgrade of their individual connections, feeding information that the local utility will use to plan corresponding distribution investments. I’m sorry I don’t know enough to talk about transformer degradation.
The utilities will stage upgrades to targeted sub-distribution transformers to alleviate or mitigate the loading and loss-of-life issues. Upgrades to distribution feeders will be more difficult to arrange and also expensive.
Warren