Power grid change may disrupt clocks

nsaspook · June 2011

The group that oversees the U.S. power grid is proposing an experiment would allow more frequency variation than it does now without corrections, according to a company presentation obtained by The Associated Press.

In the future, more use of renewable energy from the sun and wind will mean more variations in frequency on the grid, McClelland said. Solar and wind power can drop off the grid with momentary changes in weather. Correcting those deviations is expensive and requires instant backup power to be always at the ready, he said.

http://www.google.com/hostednews/ap/article/ALeqM5giHrMC9wYlOzOkUg9wNC2jVKugkw?docId=371623ab59694aef9f0a02fe83faca8a

Almost every cheap AC powered alarm clock for the last 30 years has an LM8560 or similar chip that uses the AC signal as the timebase with most having a crappy RC based timer when on battery backup. Somebody really needs to rethink this experiment.

niel · June 2011

Re: Power grid change may disrupt clocks

egads, i think that test will take place near me because cmu is involved.:grr

mcclelland is full of dodo as far as renewables causing problems. for 1 thing they don't allow more than 1% of re generation and for another thing everybody just does not gain or lose the sun or wind at the same time. where do they dig these people up at?:roll:

James · June 2011

Re: Power grid change may disrupt clocks

Maybe they want to allow more variations because they foresee the coming issues with electric grid infrastructure. They might just want a bit more room for variation because of the cost to maintain tight standards. Bringing renewable energy into the mix might offer an easy excuse?

solarix · June 2011

Re: Power grid change may disrupt clocks

My understanding of the grid as far as clock accuracy goes is that the moment by moment frequency accuracy is not all that good anyway. But the long term accuracy (like the cycles per hour) is very accurate which is what clocks rely upon. This is a much easier control problem and not a concern when adding "unstable" renewables to the grid. I really don't see how the freq control is anywhere near the problem that voltage stability is anyway.

niel · June 2011

Re: Power grid change may disrupt clocks

James wrote: »

Bringing renewable energy into the mix might offer an easy excuse?

that last sentence is exactly my thinking, but may not be as you meant it. switching high powered generators in and out throughout the day adds small variances. pvs and other re sources aren't that big in relation to the size of the power being switched and re sources do not switch on/off at the same time. if i have a cloud then you could still be seeing full sunshine and along that line the winds that may hit you may not reach the next guy's turbine for another few minutes even if nearby. there isn't any need to lower the frequency requirements in my mind as they already have the means to keep it within the current tolerances over 24hrs without it costing them more $.

i also should add that re sources are frequency dependent on the grid and only follow the frequencies the grid itself presents.

what re does is save them $ because they failed to upgrade their own grid system as load demands increased. they didn't forget to bill you for the upkeep and maintenance they failed to do and give themselves raises to boot.:grr

Cariboocoot · June 2011

Re: Power grid change may disrupt clocks

The amount of RE contributing to the grid now doesn't even make up for the daily fluctuations caused by people turning things on and off.

Or, "Pull the other one; it's got bells on".

Another case of exponential power growth without improved infrastructure being blamed on something that isn't part of the problem because it eats into the profits.

Marc "admittedly opinionated cynical old man" Beebe

nsaspook · June 2011

Re: Power grid change may disrupt clocks

I think part of the problem is the control loops for frequency stabilization in most base-load generation plants are very damped and slow to react to rapid load variations and that's a good thing normally with massive power 'on one shaft'. My understanding is a very small changes in frequency against the grid is the method used with massive generators to increase torque and power delivered to the grid. So there is always a slight bias to run fast with unstable loads to keep the system synced on positive side. I would think for inverter produced power from solar and wind it would be fairly easy to control the slight variations but with a large amount (above a few percent?) of Induction generator wind turbines needing to mechanically synchronize rapidly to the grid by frequency shifting several times a day or even per hour, that could create a fast bias unless corrected.

http://www.google.com/url?sa=t&source=web&cd=20&ved=0CEIQFjAJOAo&url=http%3A%2F%2Fciteseerx.ist.psu.edu%2Fviewdoc%2Fdownload%3Fdoi%3D10.1.1.116.7279%26rep%3Drep1%26type%3Dpdf&ei=gD8GTqTsAtPciALFzsCuDQ&usg=AFQjCNEXsyzBu3061T6RRyu_oYNy9eh0ag&sig2=b3Zru5Nm_VniMy3KI8C2Cg

RCinFLA · June 2011

Re: Power grid change may disrupt clocks

Renewable energy on grid is a nit.

Allowing more fluxuation releaves stress on the distribution grid and reduces the amount of standby reserve power that must be kept fired up to handle peaks in load that would slow down the grid, saving fuel. The eastern U.S. grid is massive. I don't see them allowing more fluxuation without allowing more frequent disconnects between the approximately ten sub-grids. It would likely be based on distribution stress points where a disconnect from the master grid would releave power flowing through particular set of distribution transformers.

It might be an improvement if the 'big one' CME hits. (solar Coronal Mass Ejection)

They could alternately allow more fluxuation in line voltage but that has more threat to the grid and end users.

You cannot just snap your fingers and get a steam boiler to pump out more power to run turbines. There is a time lag. Nuclear is even worse because they have the lag of nuclear furnace in front of steam generation. Natural gas turbines are fastest as they are jet engines with turboshaft drives. Gas turbines cogeneration take the jet exhaust to boil water to extract more energy. They can run over 50% efficient on fuel BTU content to electric generation.

nsaspook · June 2011

Re: Power grid change may disrupt clocks

I guess we will see if increasing random noise and lowering feedback to the grid control system reduces stress or only reduces operating costs at the expense of stability.

The Bonneville Power Administration solution is to just flip off the wind generation switch to maintain baseload stability.

www.crr.vutbr.cz/system/files/brozura_07_1009.pdf

TheBackRoads · June 2011

Re: Power grid change may disrupt clocks

I don't understand how allowing a looser Hz range makes the grid more stable. I read in the paper that the east coast's clocks might be off by 30 min a year and the west coast only 4 min and Texas like 2 min. I also read that "they don't know what will break" sooo that's just nifty, some day I could come home and something might just not work. I guess I just don't understand how important Hz are to AC.. novice here!

niel · June 2011

Re: Power grid change may disrupt clocks

gee hz is what makes it ac.;)

with variations of the frequency it'll sometimes hertz a little less or it hertz a little more.:p
couldn't resist.:-)

waynefromnscanada · June 2011

Re: Power grid change may disrupt clocks

niel wrote: »

gee hz is what makes it ac.;)

with variations of the frequency it'll sometimes hertz a little less or it hertz a little more.:p
couldn't resist.:-)

Good one Niel

But seriously, if we are talking about an interconnected grid here, how can things not simply explode when huge currents and voltages get out of sync, which they MUST, if one part of the grid starts operating on a different frequency than other parts? Generators etc MUST be in sync before connecting to the grid. Yes, the phase may lead or lag somewhat, but never stray to another frequency without a catastrophic failure. Electrically isolated parts of a grid may operate on it's own frequency, I've seen that as a kid during storms when our "local" 360,000KW hydro plant would be isolated from grid due to downed lines. That of course was before the advent of computer controls etc and if there was a problem at "the plant", an alarm would go off in a nearby employees home to get him out of bed and go see what the problem was. Even as a kid I was aware of the change in tone and speed of fridge compressors, water pumps etc at such times. But always, that hydro plant had to be brought back in sync with the grid before reconnection, to avoid it's destruction. Can you imagine the huge currents that would surge across any grid that had even one part, just for even an instant, slip 180 degrees out of phase? These facts have had me assuming from the very start of this thread, that it was an April Fools joke in June.

Cariboocoot · June 2011

Re: Power grid change may disrupt clocks

Friends;

I'm seeking permission to post here information from my inside source with proper accreditation.
Basically he says the problem is as bad as it seems, possibly worse. There is a lot more RE power source connected to the grid than we (or at least I) imagine and it is causing some difficulties, to put it mildly. Since this fellow design the software used to control the power grids through N.A. as well as some of the rest of the world he knows what he's talking about.

When you think about it, you can see the problem: add unpredictability to production (which as we all know can't be ramped up and down at a moments notice - we're not starting Hondas here) to the wildly unpredictable demands and you have a situation where by the time the one catches up with the other it needs to change again. This is happening in the course of a single cycle and on a megawatt scale.

The frequency fluctuation allowance is because the grid is largely frequency driven. When huge loads or production come on or off it is like the momentary change of speed on a standard generator, only times a billion or so. Naturally these shifts cause the frequency to go out of spec.

I'll warn you all in advance he has some disparaging remarks about RE inclusion in the grid, and you may not like nor agree with what he has to say. Since he's retired he's not making money off it, so he doesn't have a vested interest. For years he has lectured utilities and politicians alike on the need to upgrade infrastructure to avoid blackouts - like the large ones that still happen because they don't listen.

Hopefully he will give me permission to reprint his remarks on the subject here. They are relevant and perhaps even useful to us. I for one will admit to not being aware as to just how much RE is now part of commercial power. This is especially wind turbines, which seem to be the worst for predictability. It won't surprise anyone here that the grid's biggest problem is storage, just like our little off-grid installs.

One other thing; he didn't say anything about how far the frequency shift allowance may be, but has it occurred to anyone that if it's too far that grid-tie inverters may be sporadically knocked off-line? That's probably a bigger concern than clocks. At least for many folk here. :roll:

BB. · June 2011

Re: Power grid change may disrupt clocks

In the "old'n days", electric grids were regional. But as power plants are now build where the sources of power are (rivers, coal fields, cooling water for nuclear, wind, sun, etc.)--grids are being interconnected into larger and larger mutl-state (and multi-country) systems. And they built the loads near the cheap power sources (aluminum smelters near hydroelectric power dams).

And as Marc says, they all have to be operating at the same frequency or else very bad things can happen.

The AC interconnects have to be designed to manage large surge currents (heavy load turned on in one location needs power from a remote generator site to support)--or the local grid has to disconnect and make do with local AC power sources (and probably slow the frequency down a bit until everything is back up to speed).

And, that disconnect inter-tie--now the generating sources have lost a huge power sink--so they have their own voltage/current/frequency issues to address too.

One method around the frequency lock issue has been the use of DC very high voltage (and very high energy) inter-ties between regions. This allows the local grid to manage their frequency without regard to the remote generator's frequency/phase relationships.

DC power transmission is actually more efficient than AC for long distances (even including the AC to DC to AC power conversions)--But I am sure that these have there own issues too.

I will be very interested in Marc's friends insights. System Stability (and the math behind it) is very interesting (and complex). A simple example is using the brake and accelerator peddles on a car and how new drivers will have varying speed and braking which results in a jerky/uncomfortable ride. Now imagine trying to control the power sources and loads on a national (or even international) scale.

-Bill

Cariboocoot · June 2011

Re: Power grid change may disrupt clocks

Bill;

When you hear about what's involved from someone who does this stuff, it's scary. The software actually tries to anticipate changes, but it's like trying to pick the lottery numbers - sixty times a second. What's worse is that the sheer distance involved between source and load creates a measurable lag time.

These problems are exactly what has created the "super black outs" in the past years. One source-to-circuit connection gets overloaded, trips out. It did so because it was carry too much current being closest to the peak load. Now the loads haven't changed but you've just lost one source of power. Everything it was supplying gets dumped on the other sources. Bang! Another one overloads and trips out. Then another. Then another. Cascade failure; the East Coast goes dark before operators can catch on, locate the area that's the problem, and start manually disconnecting other regions. They can chart trends in power use, but never exact numbers. One spell of unseasonably hot or cold weather will set the thing off. Or some idiot with a hair dryer in Buffalo.

We "outsiders" tend to think of power distribution as being point-to-point: the generator, through the extension cord, to the power saw. The idea that the extension cord loops back to another generator and then to another and so on across an entire continent is a bit difficult to think of. And all long the way people have plugged in their power saws and they all flick the switch at the same time.

There are no pull ropes on power plants.

BB. · June 2011

Re: Power grid change may disrupt clocks

AC power systems are a very different animal. Probably close to a hundred years ago, when they where first learning how to send power long distances--In Germany (story from Elec. Eng. class instructor)--They had a breaker trip on the destination of a ~500 mile long power line. Problem was that this line was something like the 1/4 wave length of the 50 Hz (or whatever they were using at the time).

When you "open" a wire with AC--it creates a reflection back to the source--In this case, the reflection was roughly 180 degrees out of phase with the source--To the generator, it looked worse than a dead short and caused massive over current failures at the power plant.

AC power distribution on large scale is really a strange beast. Through in consumers (large and small) that want to turn power on/off without thinking about it. And add politicians that, through legislation want to add "uncontrolled renewable energy that has no storage/throttling element behind it", will create havoc.

Even a simple thing like a home GT inverter, without central control is an issue.

Besides the issue of a large housing track all with solar GT will generate more peak wattage than the development will consume peak/on average (i.e., to "break even"), you use power 24 hours per day but can only generate power ~6 hours per day--so you are looking at generating 5-6x or so as much power as you consume on average.

There would be simple things that can help... A GT Inverter (from what little I have seen) is designed to output current in phase with the voltage... I.e., Power Factor = 1.0 or Cosine of 0.0 degrees = 1.0 ...

However, it is not unusual for summer power factors "seen by the power company" to be on the order of 0.6 to 0.8 -- I.e., the power company generators need still to operate to supply the "imaginary" current/power component of the power to electric motors (Air-conditioning, pumping, etc.).

The current is "real", and the power plants need to supply this current--even though, for home owners (however, many commercial customers are charged for out of phase current) never pay for the off-phase current.

There are many "little issues" like this with conservation/government legislation that add up to be big head aches for power companies (lots of non-PFC computers, cfl and other high efficiency lighting, etc.).

And, sometimes that creates even more legislation (like requiring PFC for larger computers and other equipment).

It is not simple.

-Bill

niel · June 2011

Re: Power grid change may disrupt clocks

sure, i'd like to hear it. he should go into detail like explaining this to somebody that does not know all aspects of it or may just be overlooking some of those aspects. take as much time and space as you would need.

solar_dave · June 2011

Re: Power grid change may disrupt clocks

Cariboocoot wrote: »

Friends;

One other thing; he didn't say anything about how far the frequency shift allowance may be, but has it occurred to anyone that if it's too far that grid-tie inverters may be sporadically knocked off-line? That's probably a bigger concern than clocks. At least for many folk here. :roll:

I think I occasionally (rarely) see this now, where the inverter pulls the plug for 5 minutes and gets re-synced. I don't think it has been voltage related as I see very little variation in the voltages here. I am pretty sure the utility here drops in large capacity generation quite often to handle the peak production times, causing the inverter to go wonky. I suspect frequency sync is the issue. You can occasionally hear it in running motors. if it was voltage related I would think you would see it in lighting variations.

Anyone know what the inverter algorithms for out of phase look like? How many cycles of out of phase before it pulls the plug?

RCinFLA · June 2011

Re: Power grid change may disrupt clocks

solar_dave wrote: »

I think I occasionally (rarely) see this now, where the inverter pulls the plug for 5 minutes and gets re-synced. I don't think it has been voltage related as I see very little variation in the voltages here. I am pretty sure the utility here drops in large capacity generation quite often to handle the peak production times, causing the inverter to go wonky. I suspect frequency sync is the issue. You can occasionally hear it in running motors. if it was voltage related I would think you would see it in lighting variations.

Anyone know what the inverter algorithms for out of phase look like? How many cycles of out of phase before it pulls the plug?

It is more likely a voltage surge or dip causes a grid tie inverter to release then it is frequency variation. The grid drifts very slowly.

A hybrid grid tie inverter on a generator will have more possiblity of releasing due to frequency drift. A grid tie inverter has a phase lock loop to continually track the AC input. It has a narrow loop bandwidth and if the generator slews frequency too quickly (like when a heavy startup surge hits causing engine rpm dip) the inverter cannot track fast enough and releases from generator.

On a hybrid grid tie inverter, even though the frequency tracking range spec may be +/- 5 Hz, if the rate of change is faster then the tracking bandwidth, it will release even though the frequency is within the tracking range.

BB. · June 2011

Re: Power grid change may disrupt clocks

I don't recall exactly--but I thought I remembered reading somewhere the inverter had to disconnect in 5 cycles if the voltage/frequency was out of specifications.

5 cycles * 1/60 Hz = 0.083 seconds

Solar Guppy or XW Guy would probably know.

-Bill

solarix · June 2011

Re: Power grid change may disrupt clocks

While the grid is humongous and complex, barring abnormal events, it is this hugeness that helps the grid deal with unpredictability. You get enough loads all doing whatever and it generally averages out, becomes smoother and needs less response time. If you are turning off your load, likely someone else is turning their's on. Human activity is pretty predicable on a macro scale and what this expert's software does is characterize the patterns and try to anticipate the cumulative effects. On the other hand, if the something unusual happens bigger and faster than the response time of the grid - whamo - the house of cards will crumble, and you may want to have your own battery backed system.

As the song says - "do you really know what time it is?"

"And another angel came out of the temple, crying with a loud voice to him that sat on the cloud, Thrust in thy sickle, and reap: for the time is come for thee to reap; for the harvest of the earth is ripe."

solar_dave · June 2011

Re: Power grid change may disrupt clocks

RCinFLA wrote: »

It is more likely a voltage surge or dip causes a grid tie inverter to release then it is frequency variation. The grid drifts very slowly.

A hybrid grid tie inverter on a generator will have more possiblity of releasing due to frequency drift. A grid tie inverter has a phase lock loop to continually track the AC input. It has a narrow loop bandwidth and if the generator slews frequency too quickly (like when a heavy startup surge hits causing engine rpm dip) the inverter cannot track fast enough and releases from generator.

On a hybrid grid tie inverter, even though the frequency tracking range spec may be +/- 5 Hz, if the rate of change is faster then the tracking bandwidth, it will release even though the frequency is within the tracking range.

I thought the freq track was more like +/-0.5% to ensure the disconnect. I know without a full time scope on the line it is hard to tell the root cause, it is more like I feel freq is the issue than I know it. I suppose tossing more generation on the line or switching substations cause the voltage to surge/sag.

RCinFLA · June 2011

Re: Power grid change may disrupt clocks

BB. wrote: »

DC power transmission is actually more efficient than AC for long distances (even including the AC to DC to AC power conversions)--But I am sure that these have there own issues too.

Power bridges between the three continental U.S. major grids do use D.C. for the transfer between non-syncronous grids.

The main issue with HVDC transmission is cost and reliablity. Most HVDC systems today are using stacked thyristor commutating converters which are the weak link for reliability. They must be very well protected from lightning strikes. Other issue is high voltage DC circuit breakers are difficult to build, generally including a separate scheme to momentarily take the current away from breaker during opening and closure.

HVDC transmission best application is for underground or underwater transmission lines where the outside protective shielding of the cable over the distance creates too much capacitance to drive with A.C. (like the D.C. power line under the English Channel).

RCinFLA · June 2011

Re: Power grid change may disrupt clocks

solar_dave wrote: »

I thought the freq track was more like +/-0.5% to ensure the disconnect. I know without a full time scope on the line it is hard to tell the root cause, it is more like I feel freq is the issue than I know it. I suppose tossing more generation on the line or switching substations cause the voltage to surge/sag.

The UL1741 spec for grid tie tracking is -0.7 Hz to +0.5 Hz. That is -1.2% to +0.8%.

The tracking rate is left to the manufacturer but the actual grid rate of change of frequency is very very slow.

You have think about all the alternators rotating syncronized to the grid. A sudden phase shift would practically rip them from their mounts.

techntrek · July 2011

Re: Power grid change may disrupt clocks

Cariboocoot - any word from your friend?

Cariboocoot · July 2011

Re: Power grid change may disrupt clocks

techntrek wrote: »

Cariboocoot - any word from your friend?

Coincidentally, yes just today. :cool:
The trouble is that he wishes to remain anonymous, which I feel removes some credibility from his comments. (My fellow moderators know his identity.)

I am ambivalent about including his remarks here now, as they are comments made to me and not a scientific analysis of the situation as a whole.

Is it worth posting someone else's opinion without divulging his identity and thus adding credence to same? I would warn in advance that people would not like what he has to say and it would prompt further discussion (which he's not willing to continue for several reasons) and might rapidly deteriorate into argument.

Opinions?

drees · July 2011

Re: Power grid change may disrupt clocks

RCinFLA wrote: »

Gas turbines cogeneration take the jet exhaust to boil water to extract more energy. They can run over 50% efficient on fuel BTU content to electric generation.

The best ones are over 60% efficient now...

chevenstein · July 2011

Re: Power grid change may disrupt clocks

Cariboocoot wrote: »

Coincidentally, yes just today. :cool:
Is it worth posting someone else's opinion without divulging his identity and thus adding credence to same?
Opinions?

In my opinion, yes. I think those of us who know a few things about the topic will be able value the piece appropriately based on its content, so provenance isn't such a big issue. Also, this sort o thing can be a very interesting read (to me, at least).

techntrek · July 2011

Re: Power grid change may disrupt clocks

Since others know his identity that takes care of his credibility, so fire away.

BB. · July 2011

Re: Power grid change may disrupt clocks

From what Marc has said--The person is OK with posting the commentary publicly.

It may be interesting for folks to read about some of the engineering/network management issues of large power systems--But everyone will need to take a deep breath. None of people here are knowledgeable to understand the trade-offs made (which may vary by equipment/region/company/country regulations).

I have seen lots of clunky code (and hardware) that "works"... And somebody tries to fix a piece and it all collapses.

I have also seen somebody take a thousand pages of code and rewrite it into 100 pages of code--just because the old stuff was not understandable--and the guy preferred to "adjust" the new code because he knew it better (they eventually killed the whole software project because it was so kludgy).

If Marc chooses to post some of the information--just take it as snapshot in time of something that is interesting to many of us here. And be kind to the source--he is probably not here to discuss the details, or to defend.

-Bill

niel · July 2011

Re: Power grid change may disrupt clocks

to go easy on his commentary because he is not here doesn't wash as then we aren't free to fire our opinions if they oppose his. if i miss my guess i believe he may be citing large re installations like solar or wind farms where they could be in megawatts and these can have large impacts switching on/off at one time, but scattered small installations will not have the same impact due to lower power and different times for the on/off cycles. most times it is a utility that operates those large installations and is a different scenario than that presented by small scattered re installs. even at that the megawatt installs are still far below the 1% total in most areas, but like i said the re installations in europe have not posed huge problems with their much higher percentages than we experience here. they need to stop saying it's our fault for their bad grid and shaky equipment imho and i'd welcome the opportunity for him to sway me otherwise.

Power grid change may disrupt clocks

Comments

Categories