3 ms·

Why Vehicle To Grid is a horrible idea

4 points by billpg 17y ago

tptacek 17y ago
I have no strong opinion about using PEVs to feed the grid, but I have to respond to the first point the article makes, about PEVs making the grid too complex to manage. I was at a NIST/EPRI summit a few months back (we do some security work for smart grid devices), and small/home generation was a top of mind issue --- not from PEVs, but from renewables and small generators. The grid is eventually going to have to deal with millions of small generators one way or another. PEVs actually complicate the grid in worse ways than this article discusses. For instance, grid operators seem to believe that people will plug their cars in at friends houses, and that people will want a true accounting of their own power usage and not just that of a house.
wglb 17y ago
While I partially agree with the conclusion, some of the explanation doesn't really cut it: The “vehicle to grid” concept would mean that grid operators would need to know the status and availability of thousands or millions of tiny generators. Obviously there is a requirement that any household participating in this has some method of communication to the local utility point about the availability of power from here. They’d need to know where they are located and whether the local transmission lines had the capacity to transmit electricity to the areas where it is needed. Unless you are micro-routing, the utility is likely aggregating power from one "zone" to and external need. If there was capacity to charge the car, there is by definition the capacity to use the power from the car. They would need to be able to communicate to each of these vehicles and throttle them up and down, calculating, in real time, which ones were suitable for use. Well, the calculation would be done at the car--nobody else knows how much power is left. So that computation is local to the car, and yes, it is done in real time. Power would be flowing every which way. Residential areas would be generating some of their own power while transmitting some extra to commercial areas, which would be simultaneously getting some of their power from centralized power plants. Every which way. From the point of view of the residence, power is either going in or out to the line in the alley. Which is connected to the neighbor as well. The grid is not any more complex that it is today--there is the same connectivity. Some residential areas would be drawing power and then five minutes later, they would have surplus power. Industrial customers would be drawing power from numerous sources at once. Industrial customers draw power from the line (ok, some have TWO lines) that go into their building. They don't know, don't care, can't tell that it is from lots of places. The complexity of frequency and phase matching to the grid would need to be taken care of at the residential end, not centrally. The best point made here is that of an overnight load that leaves customer's batteries drained when they come out in the morning. I suspect that the plan was to have the vehicles plugged in by the time the evening meal/entertainment cycle pulls from the grid, and once all go to sleep to be able to charge them from the now less-loaded utility. This is an interesting enough problem that a more disciplined article would be appreciated. There isn't really enough here to form a nuanced opinion, and I don't agree that it is in-depth as TFA claims.

4 points by billpg 17y ago

I have no strong opinion about using PEVs to feed the grid, but I have to respond to the first point the article makes, about PEVs making the grid too complex to manage. I was at a NIST/EPRI summit a few months back (we do some security work for smart grid devices), and small/home generation was a top of mind issue --- not from PEVs, but from renewables and small generators. The grid is eventually going to have to deal with millions of small generators one way or another. PEVs actually complicate the grid in worse ways than this article discusses. For instance, grid operators seem to believe that people will plug their cars in at friends houses, and that people will want a true accounting of their own power usage and not just that of a house.

wglb 17y ago

While I partially agree with the conclusion, some of the explanation doesn't really cut it: The “vehicle to grid” concept would mean that grid operators would need to know the status and availability of thousands or millions of tiny generators. Obviously there is a requirement that any household participating in this has some method of communication to the local utility point about the availability of power from here. They’d need to know where they are located and whether the local transmission lines had the capacity to transmit electricity to the areas where it is needed. Unless you are micro-routing, the utility is likely aggregating power from one "zone" to and external need. If there was capacity to charge the car, there is by definition the capacity to use the power from the car. They would need to be able to communicate to each of these vehicles and throttle them up and down, calculating, in real time, which ones were suitable for use. Well, the calculation would be done at the car--nobody else knows how much power is left. So that computation is local to the car, and yes, it is done in real time. Power would be flowing every which way. Residential areas would be generating some of their own power while transmitting some extra to commercial areas, which would be simultaneously getting some of their power from centralized power plants. Every which way. From the point of view of the residence, power is either going in or out to the line in the alley. Which is connected to the neighbor as well. The grid is not any more complex that it is today--there is the same connectivity. Some residential areas would be drawing power and then five minutes later, they would have surplus power. Industrial customers would be drawing power from numerous sources at once. Industrial customers draw power from the line (ok, some have TWO lines) that go into their building. They don't know, don't care, can't tell that it is from lots of places. The complexity of frequency and phase matching to the grid would need to be taken care of at the residential end, not centrally. The best point made here is that of an overnight load that leaves customer's batteries drained when they come out in the morning. I suspect that the plan was to have the vehicles plugged in by the time the evening meal/entertainment cycle pulls from the grid, and once all go to sleep to be able to charge them from the now less-loaded utility. This is an interesting enough problem that a more disciplined article would be appreciated. There isn't really enough here to form a nuanced opinion, and I don't agree that it is in-depth as TFA claims.