This "Vehicle to Grid" idea has been around for a while. The article cites some progress (a trial program in SoCal).
It's noted that using one's car to contribute to grid stability may void current battery warranties, about which the article notes:
"Innovators in the field are gradually convincing car manufacturers of the potential to create a "value proposition for the car owner" and thus boost sales, Gage said. Ultimately, however, carmakers may be put at ease by experiments being conducted by the military."
This idea sounds great ... until you need to hop in your car and drive somewhere midday and learn that the battery is mostly drained because it's a hot day and everyone's turned their A/C on full blast.
I'm sure the engineers have thought about that scenario and maybe only allow 20% of the battery to be used for the purpose of grid maintenance, but I imagine it will be difficult to sell the average car owner on the idea, because people are always thinking about worst cases.
I mean, think of it in terms of a conventional car. How would you like to never know with certainty whether you have a full tank or not when you get in your car?
I wouldn't like that, indeed. I'd want my participation real-time controllable, real-time voluntary, and real-time limited.
That said, my uncertainty about vehicle use isn't as bad upon return to home, at which time I'd be more amenable to allow discharge -- and this is right when my local leg of the grid needs a bump in supply: I'm about to switch on lights, TV, microwave, heat or A/C; things I'd leave off whilst at work.
With self-driving cars on the horizon, will people even be owning their own cars for much longer? Centralization of car ownership would change how your scenario plays out.
"On the horizon" to me implies a timespan that's a lot shorter than my own view on when self-driving cars will be commonplace. I think we're decades away at least. Though in terms of the scope of the modernization of the electrical grid that would be required to make this workable, decades is probably the best case scenario anyway.
As much as I like the idea of this, I don't see it playing out that way. We currently have a form of self driving car - taxis, and yet the majority of people prefer to own their car. Golf clubs in the trunk, child seats in the back, favorite music in the glove box, etc. and the sense of "mine" seems to be enough to drive this preference for most.
While true self-driving cars should be cheaper than taxis and also allow for longer distances, I don't think this will be sufficient to change the dynamic significantly.
> While true self-driving cars should be cheaper than taxis [...]
Different costs can lead to radically different consequences. Just imagine how smart phones or tablets would play out without, say, cheap touch screens. People wouldn't care much for them.
Or imagine if power was much cheaper, then we could run eg desalination for almost free, and no coastal area or island would ever have any problems getting fresh water.
The problem with contemporary taxies is that a large chunk of the money you pay goes towards paying the driver.
Think about it; You need to pay a driver something like $40k each year; While the car itself may cost $40k to purchase and will last for 5-10 years.
Having human drivers also modifies the way you deploy your taxies. The driver must return home at the end of each day, so intercity trips don't happen. The driver needs to be paid even if the taxi is just sitting waiting for a passenger, so you want to closely match the number of active taxies to the number passengers. This also means electric taxies are unattractive for for manned taxies because your taxi shouldn't be stopped for long enough to charge between rides.
But if you move to self driving electric taxies you get rid of idle costs. It costs almost nothing to have a taxi which doesn't have passengers, so you can overstock the taxies and give time for them to charge between rides. Since you have extra taxies, you can also afford to leave them in remote areas where you know you will have return passengers in the future.
Basically, a self driving taxi doesn't have the same cost model, so it is reasonable to expect a change in the usage patterns.
An even larger cost to running a taxi is the cost of the License plate which because of severe restrictions of supply amount in some cities to hundreds of thousands of dollars
“At the moment it costs taxi operators more than $30,000 a year to lease a licence, or more than $400,000 to buy one. As a result, about $4 of every $20 fare goes towards funding that licence.”
The car batteries are really only stop-gap measures to allow the utilities to better match their plants with demands. It's for synchronization only.
A coal power plant might take several days to reach a "set point" and to generate a certain amount of power.
A natural gas plant might take as little as 5 minutes or as long as an hour to get operational.
A diesel generator might turn on and make power in as little as 30 seconds.
A hydro plant might be able to go from 0% to 100% in just a few seconds.
Given that the utility companies can't KNOW for sure EXACTLY how much power people will demand in the future they've always got excess capacity ready and waiting. But these idling plants aren't free and that drives the price of electricity up.
The idea behind V2G is that the utility companies can get some of their demand power from the cars whilst they fire up diesel generators or natural gas plants or whatever, thus saving them from having to keep those plants idling until they're definitely needed.
EDIT: The idea isn't that you constantly charge/discharge car batteries it's that you charge them in the morning after the drive to work so they're full by noon. Then as the day heats up you can pull small amounts of power from large amounts of cars until the demand is high enough that you fire up a peaker plant. Use that to charge the cars back up and provide the afternoon A/C electricity and shut it down as everyone starts to go home.
Instead of having the peaker plant be at ~20% capacity from 12-2, ~60% capacity from 2-5 and ~30% capacity from 5-7 you get to run 0% from 12-2, 80-100% from 2-5 and 30% from 5-7. Two hours of runtime saved per day is 600 hours per year. Turbine rebuilds aren't cheap; a buddy of mine is a private jet pilot and they have to put away between $500 and $2500 per hour that engines are running for overhaul at either 1000 or 2000 hour intervals. And that's for the kind of small engines in a 6-12 seater business jet. I'd wager that 20MW (~20,000 HP) natural gas turbines are substantially more.
I think the electric car as dispatchable load holds the most promise. I don't know the physics of the batteries but I imagine that reducing the rate of charge of many car batteries would temporarily reduce demand without causing a charge / discharge cycle of the battery and would benefit the stability of the system in the short term if required
Edit:
Also using car batteries as dispatchable load and not dispatchable generation simplifies the problem for utilities. They still have to deal with increased load but not drastic changes in power flows.
You wrote a very long post about the benefits. Thank you, but I already understand that. This really doesn't address my post in the slightest! Unless I'm just thick as a brick here.
I think msandford is thinking that you think that this system would charge-cycle a car more than once per day, and if you do think this, he's addressing it by pointing out that the car would be charge-cycled only once per day, or less given that the system wouldn't need to utilize every car on a given day. So this extra <= 1 charge cycle per day shouldn't "murder" the battery longevity (or at least murder is way too strong of a word) compared to normal daily driving.
Probably more like ~20-30 charge-cycles per year, as it would only be used on really hot days where AC usage is high and peaker plants are needed to meet that demand.
Yes I was getting to that in a roundabout way. It wasn't very clear. You're going to have two charge cycles per day driving to and from work probably, and one or less for the V2G scenario. So it doesn't seem terribly murderful on battery life to me. Should have made that a lot more clear.
As a counterexample, while it's not likely, I can see needing to drive down to my mother's house on very short notice, and that would be around 180 miles or so.
As a counterexample, while it's not likely, I can see needing to drive down to my mother's house on very short notice, and that would be around 180 miles or so.
The primary issue of grid stabilization isn't supplying large amounts of energy to the grid over time, it's getting enough power to it right now. Highly responsive storage is a tremendous asset.
I've got my doubts over the viability of EV-based grid storage, and a lot of the viability depends on technologies. If we see a period of rapid charging carbon nanotube or graphene batteries which can literally be fully charged in seconds to minutes, and have a tremendous number of charge cycles, it starts looking more viable. Even a Tesla Supercharger situation is pretty good in terms of time -- few commuters will be inconvenienced if their vehicle only receives a net charge for the last 30 minutes of its idle cycle, which during a daytime commute scenario typically lasts 8-9 hours.
Presently the alternatives are ... few. Pumped hydro works but is very limited by available siting. There simply aren't that many places where you can run a pump/turbine between two large bodies of water with a significant grade separation, and far fewer in Kansas (or Illinois, or Florida for that matter). One of the more reactive alternatives is kinetic storage through flywheels, which offer tremendous charge/discharge rates, but are also very pricey: about $10,000 per kWh capacity
The biggest obstacle, in my opinion, to vehicle-to-grid (V2G) applications is distribution utilities. All utilities have comprehensive procedures for studying and approving connection of distribution generation resources to their networks (rooftop PV, small CCGT, etc.). These take in account the maximum possible output of the generation asset as well as how they will behave under fault conditions.
They're already quite stingy about approving customer-owned DGs (because there is little or no financial benefit to the utility) - I can only imagine how they'll feel about people connecting a generator they don't control, at locations and times they can't predict, with the ability to disconnect them arbitrarily and drive away.
In order for V2G to become a reality, utilities will need far more advanced software and metering infrastructure to perform real-time load-flow analysis of their systems and determine if/how to dispatch connected EVs. At present few utilities even have the metering infrastructure to enable that kind of intelligent control, let alone the supervisory control software.
>[eigenvector]... utilities will need far more advanced software and metering infrastructure
...and as I take it from other discussion, re-design, -engineering and replacement of some distribution station equipment and software to allow net power flow upstream when a substation's downstream DG exceeds downstream load, presently taken to be a fault condition.
A DG participation fee payable to the utility for infrastructure development and upgrades seems reasonable to me in light of these considerations.
I would imagine that the best batteries would be reservoirs. Take two spaces that hold water. One at high ground and one at low ground. Pump the water from low ground to high ground when energy is plentiful and let it run through generators when you need energy. Yeah, there is going to be energy loss, but this approach should scale to the quantity of energy necessary to provide the same amount of energy as a power plant for 12-24 hours.
For a consumer though, a single KW (3.6MJ) of power would be generated by moving 10m^3 (i.e. 10KL) of water 36m (approximating gravity at 10ms^-2). You can tweak those numbers but you end up needing two sizable water tanks quite far apart. Great if you have loads of land but not much use for consumer storage.
Feel free to correct any flaws in the calculations, I'd be much happier if I were off by an order of 10 or more.
There are several large pumped-storage plants in the USA too. They can provide about 2 GW for one day, enough to completely power a city of 1M people, or to smooth out supply fluctuations for a much larger region.
There's a second benefit to this approach in that you could use potable water tanks and fully sealed piping. This would provide many people with water to use in the event of an emergency. Obviously they would need to balance the needs of water with its utility as a way to store energy, but so long as you conserve both in the event of a disaster, you should be able to replenish what you've used via rainwater if the system is set up to collect rainwater and filter it (except in dry areas or during a drought of course).
How long would 1 KW last for the average household, and, if 1 KW is not enough, how many cubic meters of water would the average household have to store?
How long 1 kilowatt hour would last is a difficult question - depends on how efficient the home is and what tasks need to use that energy (for example, AC is very expensive but tends to be needed most when solar energy is available). Heating and cooling are most expensive.
Our energy usage in a not particularly well insulated home but in a subtropical Spring averaged about 9kWh/day - so 1kWh (3.6MJ) wouldn't get us far - it probably barely powers the fridge! However, you'd only need the storage overnight, and use solar energy the rest of the time. In sunny climes, at least.
Again, how much water depends on how much vertical drop, but it would likely be a lot of water (if you sat one cubic tank with 4.4m sides on top of another then emptying the top into the bottom would theoretically generate 1kWh)
It'll be interesting how the energy landscape gets defined in the near future. Of note, there's a funded kickstarter that stores energy in residential flywheels that appears to be another promising solution: http://www.kickstarter.com/projects/1340066560/velkess-energ...
Recently there have been quite a few debates on the subject of solar/alternative-energy users feeding power back onto the grid .. and the latest story seems to be that power companies are attempting to fight off the impending doom of self-powered individuals by stating that 'it costs too much to update the grid to allow everyone to push to it, so these alternative-energy users should pay a hefty fee to upgrade the grid'.
Well, this story seems to point to the solution: put batteries in all the things! If its going to cost more to feed power back to the grid, then why not just make everything in your home store its own local power supply ..
To some extent, I'm already doing this - pretty much every electronic device in my house uses rechargeable batteries. So maybe the campaign by the utilities against feeding power to the grid is going to have the effect of pushing everyone to localise their power needs even further, and instead of pushing excess to an ageing, antiquated monopoly, us alternative-energy freaks are instead going to foster the creation of a heftier market for battery-protected devices. (I'd be quite happy to run my TV on battery power, if it only had one in it, like my laptops and cell phones and toothbrushes do ..)
36 comments
[ 3.5 ms ] story [ 81.5 ms ] threadIt's noted that using one's car to contribute to grid stability may void current battery warranties, about which the article notes:
"Innovators in the field are gradually convincing car manufacturers of the potential to create a "value proposition for the car owner" and thus boost sales, Gage said. Ultimately, however, carmakers may be put at ease by experiments being conducted by the military."
This idea sounds great ... until you need to hop in your car and drive somewhere midday and learn that the battery is mostly drained because it's a hot day and everyone's turned their A/C on full blast.
I'm sure the engineers have thought about that scenario and maybe only allow 20% of the battery to be used for the purpose of grid maintenance, but I imagine it will be difficult to sell the average car owner on the idea, because people are always thinking about worst cases.
I mean, think of it in terms of a conventional car. How would you like to never know with certainty whether you have a full tank or not when you get in your car?
That said, my uncertainty about vehicle use isn't as bad upon return to home, at which time I'd be more amenable to allow discharge -- and this is right when my local leg of the grid needs a bump in supply: I'm about to switch on lights, TV, microwave, heat or A/C; things I'd leave off whilst at work.
While true self-driving cars should be cheaper than taxis and also allow for longer distances, I don't think this will be sufficient to change the dynamic significantly.
Different costs can lead to radically different consequences. Just imagine how smart phones or tablets would play out without, say, cheap touch screens. People wouldn't care much for them.
Or imagine if power was much cheaper, then we could run eg desalination for almost free, and no coastal area or island would ever have any problems getting fresh water.
Think about it; You need to pay a driver something like $40k each year; While the car itself may cost $40k to purchase and will last for 5-10 years.
Having human drivers also modifies the way you deploy your taxies. The driver must return home at the end of each day, so intercity trips don't happen. The driver needs to be paid even if the taxi is just sitting waiting for a passenger, so you want to closely match the number of active taxies to the number passengers. This also means electric taxies are unattractive for for manned taxies because your taxi shouldn't be stopped for long enough to charge between rides.
But if you move to self driving electric taxies you get rid of idle costs. It costs almost nothing to have a taxi which doesn't have passengers, so you can overstock the taxies and give time for them to charge between rides. Since you have extra taxies, you can also afford to leave them in remote areas where you know you will have return passengers in the future.
Basically, a self driving taxi doesn't have the same cost model, so it is reasonable to expect a change in the usage patterns.
Read more: http://www.smh.com.au/nsw/more-taxi-licences-needed-to-lower...
A well-planned system of autonomous cars can anticipate your needs to a high degree of accuracy, either as an individual or as a group.
A coal power plant might take several days to reach a "set point" and to generate a certain amount of power.
A natural gas plant might take as little as 5 minutes or as long as an hour to get operational.
A diesel generator might turn on and make power in as little as 30 seconds.
A hydro plant might be able to go from 0% to 100% in just a few seconds.
Given that the utility companies can't KNOW for sure EXACTLY how much power people will demand in the future they've always got excess capacity ready and waiting. But these idling plants aren't free and that drives the price of electricity up.
The idea behind V2G is that the utility companies can get some of their demand power from the cars whilst they fire up diesel generators or natural gas plants or whatever, thus saving them from having to keep those plants idling until they're definitely needed.
EDIT: The idea isn't that you constantly charge/discharge car batteries it's that you charge them in the morning after the drive to work so they're full by noon. Then as the day heats up you can pull small amounts of power from large amounts of cars until the demand is high enough that you fire up a peaker plant. Use that to charge the cars back up and provide the afternoon A/C electricity and shut it down as everyone starts to go home.
Instead of having the peaker plant be at ~20% capacity from 12-2, ~60% capacity from 2-5 and ~30% capacity from 5-7 you get to run 0% from 12-2, 80-100% from 2-5 and 30% from 5-7. Two hours of runtime saved per day is 600 hours per year. Turbine rebuilds aren't cheap; a buddy of mine is a private jet pilot and they have to put away between $500 and $2500 per hour that engines are running for overhaul at either 1000 or 2000 hour intervals. And that's for the kind of small engines in a 6-12 seater business jet. I'd wager that 20MW (~20,000 HP) natural gas turbines are substantially more.
Edit: Also using car batteries as dispatchable load and not dispatchable generation simplifies the problem for utilities. They still have to deal with increased load but not drastic changes in power flows.
More concerning is battery wear (in my mind).
I've got my doubts over the viability of EV-based grid storage, and a lot of the viability depends on technologies. If we see a period of rapid charging carbon nanotube or graphene batteries which can literally be fully charged in seconds to minutes, and have a tremendous number of charge cycles, it starts looking more viable. Even a Tesla Supercharger situation is pretty good in terms of time -- few commuters will be inconvenienced if their vehicle only receives a net charge for the last 30 minutes of its idle cycle, which during a daytime commute scenario typically lasts 8-9 hours.
Presently the alternatives are ... few. Pumped hydro works but is very limited by available siting. There simply aren't that many places where you can run a pump/turbine between two large bodies of water with a significant grade separation, and far fewer in Kansas (or Illinois, or Florida for that matter). One of the more reactive alternatives is kinetic storage through flywheels, which offer tremendous charge/discharge rates, but are also very pricey: about $10,000 per kWh capacity
http://physics.ucsd.edu/do-the-math/2011/09/got-storage-how-...
I suspect idle EVs will be at best a small part of the total solution, but even a few-percent solution could be meaningful.
They're already quite stingy about approving customer-owned DGs (because there is little or no financial benefit to the utility) - I can only imagine how they'll feel about people connecting a generator they don't control, at locations and times they can't predict, with the ability to disconnect them arbitrarily and drive away.
In order for V2G to become a reality, utilities will need far more advanced software and metering infrastructure to perform real-time load-flow analysis of their systems and determine if/how to dispatch connected EVs. At present few utilities even have the metering infrastructure to enable that kind of intelligent control, let alone the supervisory control software.
...and as I take it from other discussion, re-design, -engineering and replacement of some distribution station equipment and software to allow net power flow upstream when a substation's downstream DG exceeds downstream load, presently taken to be a fault condition.
A DG participation fee payable to the utility for infrastructure development and upgrades seems reasonable to me in light of these considerations.
For a consumer though, a single KW (3.6MJ) of power would be generated by moving 10m^3 (i.e. 10KL) of water 36m (approximating gravity at 10ms^-2). You can tweak those numbers but you end up needing two sizable water tanks quite far apart. Great if you have loads of land but not much use for consumer storage.
Feel free to correct any flaws in the calculations, I'd be much happier if I were off by an order of 10 or more.
https://en.wikipedia.org/wiki/Raccoon_Mountain_Pumped-Storag...
https://en.wikipedia.org/wiki/Bath_County_Pumped_Storage_Sta...
https://en.wikipedia.org/wiki/Ludington_Pumped_Storage_Power...
How long would 1 KW last for the average household, and, if 1 KW is not enough, how many cubic meters of water would the average household have to store?
Our energy usage in a not particularly well insulated home but in a subtropical Spring averaged about 9kWh/day - so 1kWh (3.6MJ) wouldn't get us far - it probably barely powers the fridge! However, you'd only need the storage overnight, and use solar energy the rest of the time. In sunny climes, at least.
Again, how much water depends on how much vertical drop, but it would likely be a lot of water (if you sat one cubic tank with 4.4m sides on top of another then emptying the top into the bottom would theoretically generate 1kWh)
Well, this story seems to point to the solution: put batteries in all the things! If its going to cost more to feed power back to the grid, then why not just make everything in your home store its own local power supply ..
To some extent, I'm already doing this - pretty much every electronic device in my house uses rechargeable batteries. So maybe the campaign by the utilities against feeding power to the grid is going to have the effect of pushing everyone to localise their power needs even further, and instead of pushing excess to an ageing, antiquated monopoly, us alternative-energy freaks are instead going to foster the creation of a heftier market for battery-protected devices. (I'd be quite happy to run my TV on battery power, if it only had one in it, like my laptops and cell phones and toothbrushes do ..)