Before reading further, everyone should go over to the web site http://www.acpropulsion.com and read the white paper on grid-interactive electric vehicles. Got that? Good. Consider what just happened to a large chunk of the USA: one powerplant goes out, the excess load falls onto other plants, and they go off-line too. As the load falls onto plants farther and farther away, they go off-line in turn and the blackout spreads. It's going to take upwards of a day to bring everything back on-line, according to DTE Energy. (At least three days according to the latest news.) Exactly how did this happen? If I understand the mechanism correctly, the problem is due to a lack of "spinning reserve", a term which means powerplants on-line (spinning) but not actually producing power. A grid is supposed to have enough spinning reserve to handle the failure of the largest plant currently feeding the network. Bluntly, this didn't work. Changing the subject to cars for a moment, the total power of the vehicle powerplants in the country vastly exceeds the total power of the electric powerplants. Total nameplate power output of all the electric powerplants in the USA is something around 700 million kilowatts. Figuring 250 million vehicles in the USA at 100 horsepower minimum, call it 70 kilowatts each, is 1.75 billion kilowatts, more than twice as much. Unfortunately, that capacity isn't connected to the electric grid. Yet. AC Propulsion, Inc touts the value of grid-connected electric cars for stabilizing the electrical grid. The price of not taking that course has suddenly come home to us today. If even 10% of the vehicles on the road had substantial battery packs and were connected to the power grid, they could feed enough power back to the grid to make up for the loss of several large powerplants. This could balance transient overloads until the grid operators could fix them, and would prevent the kind of cascading outages which we've seen and make us all much safer and more secure. Electric and grid-connected hybrid cars are more than just a matter for the environment. They are increasingly important for our national security. (The author is blacked out as he writes this on Thursday night, but is writing this item using a computer powered by a UPS on a 105 amp-hour, 12-volt storage battery. It is working considerably better than Grex seems to be!)39 responses total.
Russ, is that a deep cycle marine battery? What brand of UPS? I had considered doing that with my APC, but wondered whether that would screw something up due to slight voltage curve differences. One problem with the grid supplimenting idea: If there's a possible power outage to occur and I had an electric car, I'd be reluctant to release any of its charge unless compensated extremely well for it - and power companies, when one does manage to get them to buy back electricity, rarely pay full price for it. It *would* be useful for powering ones house though. (For my case, I'm going to need a wireless extension cord, as it were, to make it work for me. To wit, the same method proposed for beaming power from satellites to ground stations.)
Somebody tell me why something which is good for the environment also has to be "good for national security" as well, for it to sell? Who will care about national security when the environment is f*cked, hmm?
Politics is so screwed, man. You'll never get anyone to agree on anything.
Re #1: Yeah, me too. The last thing I'd want is for an outage to happen anyway, and end up stuck with no transportation because my car's battery was discharged trying to prop up the grid.
Incidentally, this outage made me aware of something I wasn't before -- that nuclear plants can't run safely without outside power available. When the grid started to fail all the nuclear power plants shut down. This suggests to me that building new nuclear power plants in an attempt to improve power generation capacity will make the grid more unstable, not less, since at the first sign of trouble they'll shut down and make the problem worse.
They don't need outside power. They shut down generation for the grid but had power for their own use. Power plants don't really "shut down": they disconnect from the grid because the grid will no longer accept their power and they then power down. The grid is inherently unstable. Realize that every power plant has to be generating power in phase (locally): if phase slips by a small fraction of 1/60-th of a second, things tend to go BOOM. A sudden enormous new load on a power plant slows the generators, slips phase, causes an enormous power load as phases mismatch, and causes disconnection. This is the one enormous disadvantage for AC powwer, compared to DC power.
Actually, they *do* need outside power. They pull outside power in to run cooling systems and security cameras and so forth. If outside power fails, they have to shut down so that the rods don't overheat. From a New York Times story this weekend: The Nuclear Regulatory Commission said the blackout today caused seven nuclear power plants in New York and New Jersey and two in the Midwest to shut down automatically. The power plants feed electricity out into the wider grid of transmission wires, but they run their systems on power pulled back in from that grid. When there is a disruption in that incoming power, the plants automatically shut down their reactors and switch to diesel-fueled backup generators. The nuclear plants cannot be turned back on until the rest of the grid has power produced by other plants running on coal, gas or other fuels, said Scott Burnell, a spokesman for the regulatory commission in its headquarters in Rockville, Md. Such shutdowns are a normal response to a power disruption and do not raise risks of radiation releases or other problems, federal nuclear regulatory officials said. The backup generators support vital systems like coolant pumps and security cameras. They must have sufficient fuel to run for seven days, said Edwin Lyman, a senior scientist at the Union of Concerned Scientists, a private group that studies nuclear safety issues. The generators are tested frequently because if one were to fail, leading to a complete blackout at a nuclear plant, the nuclear fuel could overheat dangerously. [http://www.nytimes.com/2003/08/15/nyregion/15GRID.html?pagewanted=3], FRR.
This response has been erased.
What I use is beside the point, Drew (but see response in item 162). The whole power outage was caused by a transient condition which cascaded like dominos across 1000 miles in 9 seconds. This should not have happened; our power networks should not be so brittle, but it's obvious that they are. The question is how to fix them. If you had read the paper at acpropulsion.com, you'd know that a grid-interactive car can offer considerable value to the grid operators. If you're offering a high-value service, you can often get a fair fraction of that value back in your pocket. According to the paper the grid suffers local imbalances on a time-scale of minutes, and these imbalances are a major headache for the operators. The buffering capacity of a large number of grid-connected electric or hybrid vehicles would be perfect for their purposes, and they'd be willing to pay the suppliers of this service. Such a buffer would have been able to prevent this blackout entirely. The antidote to brittleness is toughness, the ability to absorb stress by yielding without damage. Even 100,000 cars on the grid at 20 kilowatts apiece would have been able to replace (or absorb) the output of 3 typical large powerplants. That is a lot of stress; it is far more than the amount of power going over the two 345 KV power lines which are suspected of being the trigger for this outage. If the operators had even 60 seconds to shed load and adjust generators, the powerplants would have remained on-line, all the dominos would have remained standing, and we wouldn't be talking about the Great Blackout of 2003.
Re #8: I'm not convinced yet that it was a Blaster worm issue, but I guess time will tell. This sort of thing has happened on at least three other occasions, in other parts of the grid, without any Internet worm being involved. I suspect the trigger event may turn out to be something as random as a geomagnetic storm, or a circuit breaker failure like the one that triggered the 1977 blackout.
Knowing almost nothing about the power grid, but having done a fair amount of work on other complex networks, I strongly suspect that anybody who claims to know what caused this, or what would have prevented it, is wrong. If the last failure this massive was almost 38 years ago, it is a fairly safe bet that almost nobody still working on the power grid would ever have seen a failure like this before. That said, I do find it somewhat hard to imagine that a bunch of people hooking up their hybrid cars to the grid and starting them would have prevented this. My pretty limited understanding is that the grid is supposed to cut off failing pieces within milliseconds to save the rest of the system, and in this case the entire grid was somehow shut off. Ignoring for the moment the likelyhood of people reliably plugging in their cars whenever they park to guard against a once every 40 year event, I would expect that determing which cars needed to be turned on to stabilize things, and then starting them all, would be a far more complex operation than whatever the grid failed to do this time. Moreover, I suspect that the act of all these people regularly connecting and disconnecting their cars from the power grid would cause probably of its own, probably more often than every 40 years. There's a saying in my industry, that "the number one cause of network outages is network engineers." That is while people do need to work on systems to maintain them or make them grow, most failures happen because somebody changed something and the results weren't what was expected.
Re #7: I think you may be mistaken. The plants do not *need* outside power to restart but they are *required* to have it. In fact, they are required to have multiple sources of power. For an example of this see http://www.energy-northwest.com/Information/03.14.03.html In addition, it seems that nuclear power plants are not *allowed* to start up a grid, because they are so finicky and the consequences of a problem can be more serious than for a non-nuclear plant. There they are *reauired* to only feed power into a grid that is already up.
Re #2: Not every argument carries weight with every person. Someone might discount the need to cut back on oil consumption for the sake of the environment, but take terrorism only too seriously. A measure which can be justified for several different reasons (appealing to different people) is far more likely to attract support. Re #4: Go to the site and read the paper. Bluntly, that wouldn't happen. If you had a hybrid car, you'd have to be running out of fuel for it to make a difference. For an electric, you'd reserve enough energy to get home and let the utility pay you for the right to use the rest. For an event like Thursday's, you're talking very little energy. If your car back-fed the grid with 20 KW for 60 seconds while a situation was resolved, that is a whole 1/3 of a kilowatt-hour. That amount of energy would probably drive your car less than 2 miles. Besides, the idea is that the cars would balance the grid so it stays up and you can charge. Re #6: Actually, nuclear powerplants do NOT run on their own power; they have diesel generators for powering pumps and such when the grid is down. The same appears to be true of coal-fired plants; when the lines tripped, several DTE plants took condenser damage when the boilers dumped their steam as the plants shut down. Big plants need big loads, and the little stuff around the plant just won't cut it. (No, I don't know why they don't use dummy loads.) AC networks are not inherently unstable. Power is transferred by voltage differences in DC systems, but by phase differences in AC systems; leading phase transfers power to lagging phase. If I understand it correctly, there are few problems with interconnecting things so long as the voltages are synced properly (frequency and phase) before the switch closes. In modern systems this is done automagically, but long ago it used to be done with a set of 3 lamps, one across each phase of the 3-phase switch, and the alternerator would be brought up to speed and then tweaked into phase by manual adjustment until all of the lamps went out. Once the switches were closed, torque could be applied to the alternator and it would start supplying power to the line. Of course, any system can be pushed past its limits. Power transfer over a line peaks at 90 degrees phase difference. Below that, the system is stable (at least in theory), but if you try pushing more power across the line without boosting the voltage (phase and voltage are independent variables) the two ends fall out of sync, the stable coupling is lost, and large overcurrent events can occur. This Is Bad. Transmission lines are designed to operate well below the limit. Having more and bigger buffers for power on either end of a transmission line would help keep excursions like Thursday's from happening. Putting these buffers in our cars seems to be the smart thing to do.
re#0: Current opinion is not power plant failure but multiple line failures in ohio with one or more previously known broken alarm systems. ( see ap wire service stories on the issue) Lets see if I understand, DC batteries in individual cars are connected using an infrastructure that duplicates the existing distribution network at a cost far greater than merely upgrading capacity to distribute generator capacity that currently exheeds distribution network ability to carry load. In a nutshell, upgrading current distribution network is far cheaper and better play.
Re #11: You didn't read the paper either, did you? (Can't *anyone* be
bothered to take in the background information before trying to discuss
it? Is there any point in misconceptions born of complete ignorance?)
For your information, Thursday's outage happened quickly enough that by
the time anyone could have gone to their car to plug it in, the event
would have been over. People would plug cars in every day because they'd
get charged up, and possibly paid in the bargain. Cars would not be
started on demand (in the case of an electric, what's to start?), but
their batteries would be part of the ebb and flow of the grid.
Stabilizing things is easier than it sounds; holding frequency and
voltage steady at the connection will automatically react to surplus
or deficit elsewhere on the grid. (How do you think utilities do it?
That's why "spinning reserve" does what it does.)
You may be right that lots of vehicles going on and off the grid would
create some management issues, but you ignore two things:
1.) Aggregate behavior of large groups of people is very predictable.
2.) If as few as 30,000 vehicles can buffer the gain or loss of a
600 megawatt powerplant, only a small fraction of the total
vehicle fleet would have to be plugged in at once to keep the
grid stable as a table.
It sounds like, indirectly, deregulation is to blame for the outage. From a Detroit News article: "The transmission system in the United States was designed largely for individual, stand-alone utilities with some interconnections with their neigboring utilities for short emergencies," said Anthony F. Earley, chairman and CEO of DTE Energy Co. "It wasn't originally designed to transport large amounts of bulk power over long distances." "As deregulation of the wholesale markets occurred, the economic driver was to go find the cheapest power you could wherever you could and transport it to where it was needed. So, we've seen over the last five years a significant increase in the amount of power transmitted across the system." ...Energy experts say power plants now are being built too far from reliable transmission lines. The generating companies building these plants are concerned mainly with the cost of transporting fuel to the power plant, not about upgrarding transmission lines to get the new power safely into the grid. And transmission companies have little incentive to invest in new lines because they are no longer guaranteed that they can raise rates to recover the costs.
Re #14: Let's test that understanding. Fill in the blanks, Beady:
1.) The lines (3 of them, according to the latest) went out.
What are the lines for?
(Answer: For getting power from a place where it's in
surplus to where it's in deficit.)
2.) What can storage do for you?
(Answer: Absorb a surplus, or fill a deficit.)
3.) Storage capacities of hybrid vehicles are in the neighborhood
of 1 KWH; typical electric vehicles carry 10-20 KWH. Power
output ranges from 30 to 150 KW. At 20 KW apiece, how many
vehicles would it take to absorb or supply the power going
through the lines which went down, and for how long could
they do it?
(Answer: If you're so smart, you tell me.)
On other points:
>... DC batteries in individual cars are connected using an infrastructure
>that duplicates the existing distribution network
Um, no. The distribution network has next to no energy storage; the
various inductances, capacitances and kinetic energy of rotating
machinery hold energy for less than a second of demand. Storage is
so valuable that systems such as the Ludington pumped-storage facility
are highly coveted by network managers.
>at a cost far greater than merely upgrading capacity to distribute
>generator capacity
Far greater costs? Really? Even when the batteries can supply energy
to the car for less than the price of gasoline [1], and their use for
supplying ancillary services to the grid is effectively free?
If upgrading distribution is so cheap, why has "demand-side management"
been a buzzword of the electric industry for the past decade-plus?
Hint: Network costs scale with peak capacity, while revenue scales
with energy shipped. Flattening the demand curve pays dividends.
>that currently exheeds distribution network ability to carry load.
You're including all the peaking generators which cost $.25/KWH
and upwards to run, yes? You're also including full nameplate
capacity of all plants when some of them would usually be down
for maintenance or running at less than 100% for various reasons,
like not having enough water behind the dam?
>In a nutshell, upgrading current distribution network is far cheaper
>and better play.
Yeah, right. Nobody in the industry has believed that for ages.
[1] Commuter Cars Corp. claims that their car, the Tango, can be run
for about half the cost of gasoline. See http://www.commutercars.com
for their supporting data. AC Propulsion (http://www.acpropulsion.com)
makes similar claims.
Ok, if its so good and cheap why isn't it being done, buttboy?
Sounds like partial deregulation creates perverse incentives. Isn't anyone in authority smart enough to see these things coming? Incidentally, some of Beady's assertions are specifically denied (one might even say refuted) by Patrick Lincoln's first piece in the latest RISKS digest (/a/r/u/russ/risks/risks-22.86). Elsewhere in that issue, Edward Reid states that the Niagara area saw a flow change of 3 gigawatts in under one second. (3 GW is 150,000 cars at 20 KW each. How many cars do you think there are split between Buffalo and the Toronto/Hamilton area?) A final quote from Reid's piece: Conservation is far more cost-effective than new construction at ensuring continuous availability of electricity. But this is not a market-savvy investment, so until we accept that we need non-market investments in conservation, we will continue to waste our most effective resource.
Yep, that definately proves your point, buttboy.
I'm not sure transmission line deregulation would solve anything. Think about it. For a stable grid, for every major transmission path you need capacity that's never used except in an emergency. Power engineers call this "reserve." A deregulated free market calls it "waste" and tries to eliminate it.
The problem with the "free market solves all problems" mindset is that it assumes that the people running the companies are planning long term. These days pretty much most big companies seem to be run on the "how fast can I pump up my bonuses and cash out" philosophy.
This response has been erased.
"Buttboy"? Look, another deserter from the battle of intellect!
People with further interest in the various issues should see this:
http://www.spectrum.ieee.org/WEBONLY/publicfeature/aug03/wind.html
This backs my point that storage is *important*. Some quotes:
The giant shock absorber that Stahlkopf envisions for the Big
Island would mediate between the power grid and the turbines,
and it would combine power electronics with an advanced energy
storage device, such as an ultracapacitor or a battery.
When the turbines are going full bore, Stahlkopf explains, the
power electronics will divert some power into the storage system,
drawing it out again when the wind dips. If the line voltage
drops in a fault, the power electronics will dig deeper into the
storage reservoir to generate reactive power and prop up the line.
For the project, Hawaii Electric, the state, and the U.S.
Department of Energy are now considering various storage
technologies and capacities, and Stahlkopf reports that a
shock-absorbing system could be on-line by the end of next year.
...
The flow batteries in the system Apollo envisions could back up
a 20-MW wind farm for several minutes. It wouldn't be cheap;
building and operating the battery system could cost $1.8 million
over its projected 17-year life span. But by keeping the wind farm
going during the 200-plus hours each year when erratic winds would
otherwise force operators to shut down some turbines, ZBB and its
partners insist that the battery storage system would more than
pay for itself -- to the tune of $5.4 million over its lifetime.
(Or weigh it against the cost of the grid going down for a day.)
(doubtless I'll be asked to RTFA after this question, but...) Russ, weren't you dissing the usefulness of some electric vehicle a while back, based on the issue of charge/discharge cycles wearing down the battery? I vaguely recall somebody talking about a hilly commute to work...
Re #25: >... weren't you dissing the usefulness of some electric vehicle a while >back, based on the issue of charge/discharge cycles wearing down the battery? Possibly. Not every vehicle is good for everything, and it's in my character to say that something isn't appropriate for a given purpose. HOWEVER, battery-discharge issues are largely mooted by hybrid vehicles, and IIRC the small variations in charge used by the grid stabilization scheme don't appear to decrease battery life much even for lead-acid batteries. If you could get paid more than the wear on the battery costs, you'd come out ahead. Just in case I'm getting it wrong, see my source for this: http://www.acpropulsion.com/Veh_Grid_Power/V2G%20Final%20Report%20R5.pdf Incidentally, the problem of battery damage by deep discharges goes down as the battery capacity goes up, and it appears that a relatively small and light lithium-ion battery would hold more than enough energy to power a vehicle for most people's daily driving.
I've had a chance to re-read the final report (finally!), and I'd like
to quote from it to give everyone an idea of what's possible:
2.3 Results
Results showed that wireless data transmission times were within
ISO system requirements, and that the daily energy throughput
through the battery pack while performing regulation is of the
same magnitude as that from typical daily driving. Battery
heating during the test periods was negligible. The long term
effects on battery life were beyond the scope of the study;
however, it was noted that battery energy capacity increased by
about 10 percent during the course of the testing. The monetary
value created by these services varies with the market price for
grid ancillary services. A sample vehicle usage profile was
developed, with on average 22.6 hours of grid connect time every
day. The value of grid regulation was determined for this usage
profile based on regulation pricing from sample weeks in
December 201, April 202, and July 2002. The annualized gross
value created with 80-Amp grid connections available at home and
at work ranged from a low of $3,038 for December 2001 to a high
of $5,038 for July 2002. The results support the ZEV program
goals by demonstrating the feasibility of vehicle-based grid
services or V2G.
Note well the various benefits:
1.) Negligible battery heating (implies very high efficiency).
2.) Battery capacity increased (implies long life).
3.) Gross value created was up to $420/month, and never below
$250/month. (That's based on market prices for regulation
services, if you read the rest of the paper.)
When you consider that a complete lead-acid battery pack for an electric
car is only about $2500, getting that much extra value out of the
batteries ought to get everyone's attention. The car could go a long
way toward paying for itself.
People who have any interest in the possibilities of electric cars should take a look at this article: http://www.nytimes.com/2003/09/19/automobiles/19ELEC.html
I like the progress so far - and it would go really well with an
Amick-style wind foil. Could have virtually infinite range that way, since
it regenerative-brakes.
But must these things be so expensive? The quoted price would buy 44000
gallons of gasoline at $5/gallon. Is there a way to get such capability a
couple of orders of magnitude cheaper?
I suppose you could make the car cheaper if you didn't build it by hand. I hear that this tends to drive up the price a bit. More seriously, I'm told that Li-ion cells are now available for roughly $1/amp-hour. At a nominal voltage of 3.6 volts, a string of 75 cells at 200 AH would hold 54 KWH and cost $15,000. That would be enough to go a couple hundred miles in an efficient car. The price is bound to keep coming down with time.
Do NYT user IDs expire if you don't use them or something? I don't go there much, but every time I do none of the ones I thought I'd created work. I'm getting sick of creating a new one every time someone posts an article.
We believe they do.
I've given up on the NYT website because of this. They are just losing an audience if others also find this onerous.
re #31: I've got an infrequently used one that I've been using occasionally for nearly five years now and another one that I use for daily browsing. Despite going months between uses the infrequently used one has never expired.
Probably it's the NYT login cookie(s) getting expired on in/by your browser.
I keep my cookie file locked (thanks, rcurl) with the nyt cookie and one or two others in it. All others disappear as soon as the browser quits. I've never had to re-enter my nyt login.
(IIRC, one of the others is a Doubleclick opt-out cookie.)
There must be something different about the settings of those having problems. I'm using my original NYT account opened at least a couple years ago and it's never burped on my login. I use Safari and have cookies enabled but only for sites I navigate to and accept.
Re #33: I doubt that bothers them. It's not like they make any money off the website.
You have several choices: