What Happens to Unused Electricity in Distribution Transformers?

[Venkata Satish]

Thanks for your reply. Now i agree that no extra energy is generated.

But recently i went to a conference. They put these statement like "Achieved peak savings of 7.2 MW load during DR (Demand Response) Event"

What do they mean by peak savings of 7.2 MW mean? If Produced energy is same as consumed energy where is the saving? Is is commercial savings they are talking about? Meaning they committed to purchase more power but now they reduced it to 7.2 MW because of Demand Response?

Sorry one more question. If utility orders Energy say tomorrow we want 1000MW, are they obligated to buy even thought they have supplied less because of consumption?

 

[sophiecentaur]

What do they mean by peak savings of 7.2 MW mean? If Produced energy is same as consumed energy where is the saving?

I read that as meaning that the 7.2 MW was handled by lower cost generators. High cost, Peak Generation capacity was not needed. Presumably this could be achieved by stoking up coal boilers in good time to catch the peak demand. You can do this with software if you have a sluggish system (predictive techniques) or with human skill and experience. Pedalling fast downhill (forward planning) can help get you up the next hill.

 

[anorlunda]

OP your questions were all answered in post #7.

Beware sales pitches. To my knowledge, payments for demand response are so far limited to large blocks of power, not homeowners. 7,2 MW is about 1000 times more than a house. There is lots of talk about consumer level demand response, but I have not heard of it being implemented yet.

Did you go to the conference as an individual? Representing a company? Representing a city?

 

[jim hardy]

Look at the bullet just above the one you yellowed..
I would guess they remotely switched off 7.2 megawatts of water heaters, airconditioners and clothes dryers.
http://energy.gov/oe/technology-development/smart-grid/demand-response

It also includes direct load control programs which provide the ability for power companies to cycle air conditioners and water heaters on and off during periods of peak demand in exchange for a financial incentive and lower electric bills.

Wikipedia shows this picture of a clothes dryer using a demand response switch to reduce peak demand

 

[Venkata Satish]

i represented the company. My company does software.
I understood the usage of smart meters like to time of usage , if there ant defects in the meters, auto disconnect etc

But still i am not understanding how smart meters can help utilies save money at the distribution place.

Satish

 

[Venkata Satish]

Look at the bullet just above the one you yellowed..
I would guess they remotely switched off 7.2 megawatts of water heaters, airconditioners and clothes dryers.
http://energy.gov/oe/technology-development/smart-grid/demand-response

Wikipedia shows this picture of a clothes dryer using a demand response switch to reduce peak demand

But this as good as not switching the washing machine. So is it addressing financial problem or technical problem.

 

[jim hardy]

But this as good as not switching the washing machine. So is it addressing financial problem or technical problem.

I'm a power plant maintenance guy. So to me it's a technical problem.
At my utility there were days when customers use was pushing our ability to generate so close to the limit that we turned off lights on the boilers and cut out feedwater heaters just for a few extra kilowatts. We sure could have used some gas turbine peaking units but this was 1960's...

To a power company system dispatcher it's a financial problem. If he doesn't have to start one of those horridly inefficient gas turbine peaking units he can save thousands of dollars. 7 megawatts at ten cents per kilowatt for jet engine fuel is S700 an hour. At one time we had a hundred megawatts of those peaking turbines..

 

[Jeff Rosenbury]

I think the key is to understand that different energy sources have wildly different costs. Some hydroelectric plants cost about ½ ¢/kWh while gas peaking plants may run 9 ¢/kWh. Different plants also have different response times. A gas plant can change its power output quickly, but a coal plant takes a few minutes. Nuclear plants can take hours (weeks to shut down, as Fukishima showed us). So power companies want to run the cheap plants full out and use the expensive peaking plants only to balance things when unexpected loads come.

By controlling even a little of the load, the use of expensive peaking power sources is reduced. This saves costs -- a lot of costs.

BTW, lower natural gas costs likely lower these problems. But an increasing reliance on unstable wind and solar power dramatically increases the problem.

If you are interested in the financial side, get the annual report from a few power companies (you might want to buy some stock). They go into more detail about costs and how they are reduced. Public utilities are mostly transparent for those willing to do the work of understanding.

 

[anorlunda]

Sigh, we are getting deeper into the woods than I planned.

Electric energy is a fungible commodity, just like shares in a company stock. So all people who buy/sell IBM stock today at 1019 EDT may see a clearing price of $110.27. Some people buy. Some sell. The profit/or loss of each party is not considered. Everybody trades at the same price. If you want a minimum of $120 for your stock, it won't get sold at this moment.

It is the same with generators in the electric markets. If we are forced to start a high cost gas turbine, the clearing price for all generators goes up. Even the low cost generators will be paid the high price until a new price is calculated a short time in the future. In theory, an electric market running along steady at $30/MWH, could spike to $1000/MWH for 15 minutes. In real life (after the year 2000 when California learned the hard way to do it correctly) such extreme spikes almost never happen. A single market clearing price is a simple matter of economic fairness and even handedness. However, it makes it even more expensive to meet a peak demand.

On the wholesale markets, $50/MWH is a typical price paid for an increment of generation. $500/MWH is the price paid for demand response DR. However, those payments are not available to small players, because there is a significant engineering & legal overhead to set up the DR contracts, communications, certification and annual audit.

The law plays a role too. In New York State, the law says that utilities may not consider price when purchasing enough generation to meet the demand. That prevents them from deliberate and involuntary area blackouts as a tactic to avoid price spikes. DR can be considered as a voluntary black out.

(See why I say that these power topics are multidisciplinary? We go from Ohms Law, to Economics 101, to the law almost in the same paragraph.)

My info on DR may be dated. Does anyone know of a homeowner who received a check in the mail for turning off their HVAC under DR control? Time-based rates are not DR. DR is when the grid sends a command DROP YOUR LOAD RIGHT NOW (or there will be a stiff penalty for non-compliance.)

 

[johnbbahm]

I was thinking about what problems a large number of grid attached solar customers will be.
The supply could have much broader swings in a local area.
The power plants can adjust their supply, but only down to zero,
some months in spring and fall, may not have much local load, but a large
surplus of supply.
I am thinking there will need to be some place to dump the surplus, least the heat
harm the grid.

 

[BillyT]

Post 10 has correct answer for a grid, with several generators feeding it.
On a single transformer POV, if all of the customer it supplies disconnect, then the current the transformer takes from the gird drop drastically - only the hysteresis loses in the iron core are being met as the magnetic field alternates. The current in the secondary is zero.

 

[anorlunda]

I was thinking about what problems a large number of grid attached solar customers will be.
The supply could have much broader swings in a local area.
The power plants can adjust their supply, but only down to zero,
some months in spring and fall, may not have much local load, but a large
surplus of supply.
I am thinking there will need to be some place to dump the surplus, least the heat
harm the grid.

Everything you said may become true 20-30 years from now, but for now wind and solar are a tiny fraction of total capacity. 0.6% solar and 4.7% wind in 2015, https://www.eia.gov/tools/faqs/faq.cfm?id=427&t=3
But they are growing rapidly. The northeast USA grids did a study. They found that wind / solar can grow to 20% of total capacity before any drastic changes are needed in the grid operations.

 

[jim hardy]

My info on DR may be dated. Does anyone know of a homeowner who received a check in the mail for turning off their HVAC under DR control?

My utility gives a small discount, around five bucks a month, on the bill for customers who let them install a DR relay on airconditioner and water heater.

My house had one when i moved in but i didnt know what it was. When it failed to "disconnected" state in dead of winter and Fair Anne got a cold shower, i took it out.
That was under threat of "Wrath of the Valkyries" - she thought i'd turned it off on purpose.
I need to call electric company and get one put back in - it's good neighbor policy.

old jim

 

[anorlunda]

utility gives a small discount, around five bucks a month, on the bill for customers who let them install a DR relay on airconditioner and water heater.

Thanks Jim. I am out of date. So now the discussion has to be broadened to wholesale and retail DR. The $500/MWH number I quoted was wholesale. The Supreme Court just decided a case about wholesale DR.

I've been living off the grid on a sailboat for 11 years. I haven't seen a power bill (or a water heater or an AC) that whole time.

 

[AgentCachat]

OK, dummer question still, perhaps off topic. What is base load? Is this related to what the OP asked?

 

[anorlunda]

Base load is that which is there 24x7. However, the definition can be much looser than that.

Think of base load. Peak load and intermediate. Fit your generation source to the closest of those three choices.

 

[sophiecentaur]

As a matter of interest - and someone is bound to have an idea about this - what is the sort of time involved between increasing the rate of coal feed to an increase in generator output? Is it 2 minutes, 10 minutes, 30 minutes? The same figure for a nuclear system would also be interesting (also the reaction time to a drop in fuel input).

 

[Jeff Rosenbury]

I was thinking about what problems a large number of grid attached solar customers will be.
The supply could have much broader swings in a local area.
The power plants can adjust their supply, but only down to zero,
some months in spring and fall, may not have much local load, but a large
surplus of supply.
I am thinking there will need to be some place to dump the surplus, least the heat
harm the grid.

There are some economic activities that thrive on cheap power. Making aluminum is an example. I could easily foresee industrial plants designed to take up the slack.

Another option is energy storage. There are several competing technologies being developed, plus some old fashioned hydo projects. Pump the water uphill when power is cheap and generate peaking power when it's expensive. Still, none of these solutions are inexpensive or likely to get that way soon.

 

[Jeff Rosenbury]

As a matter of interest - and someone is bound to have an idea about this - what is the sort of time involved between increasing the rate of coal feed to an increase in generator output? Is it 2 minutes, 10 minutes, 30 minutes? The same figure for a nuclear system would also be interesting (also the reaction time to a drop in fuel input).

I'm pretty sure power can be cut to a nuclear plant quickly, within a few minutes.

There are three steps. First the fast (prompt) neutrons used in chain reactions are quenched. This takes less than 1ms. That's pretty fast. Next the decay neutrons need to be poisoned. That takes longer. I don't know how long. Finally, about 7% of full power remains due to decay heat. (As opposed to decay neutrons which cause more reactions, but below the breakeven for chain reactions.) This 7% drops off over months, years, decades, centuries, etc.

I seem to recall that Fukishima should have had a month of forced cooling before the fuel rods could boil water on their own without danger.

 

[jim hardy]

As a matter of interest - and someone is bound to have an idea about this - what is the sort of time involved between increasing the rate of coal feed to an increase in generator output? Is it 2 minutes, 10 minutes, 30 minutes? The same figure for a nuclear system would also be interesting (also the reaction time to a drop in fuel input).

Westinghouse plants are designed to "load follow" but they're base loaded because of economics.
i'm having trouble remembering numbers as to procedural limits on rate of load increase/decrease.
I have seen my reactor returned to full power within an hour of a trip. So your "thirty minute" estimate of 3.3% per minute sounds about right for graceful operation.
Smaller changes can go faster. A 10% step change is quite manageable. I've seen step decreases of 50% which, when everything worked as designed our plant handled.

Coal plants - i never worked in one. I'd guess they're similar, 3 to 5 % a minute because that's what the grid needs.

There are three steps. First the fast (prompt) neutrons used in chain reactions are quenched. This takes less than 1ms. That's pretty fast.

It takes a little over a half second for the rods to fall the twelve feet from top of core to bottom. The "prompt" neutrons disappear as the rods fall, and indeed our instruments show flux decreasing in the top half of core first .
That top-to-bottom flux difference propagates through the Reactor Protection System briefly actuating relays whose purpose is to protect the reactor from flux imbalance (for you PWR guys: fΔI term in OPSP and OTSP equations ).
When reviewing post-trip data it's at first confusing why you got those alarms .

A "trip" closes the steam valves as fast as they can move, ~1/10 second and steam bleeds out of the turbine over the next few seconds. So a "trip" puts a near immediate stop to generation.

I hope that helps. Somebody who's more current than i will have better answers.

old jim

 

[Venkata Satish]

Guy i am so sorry. i tough the initial question is very simple and a Dumb question. It turns out there is lot deeper that I thought

My only question is how Smart Meters are providing profits to utilities. I say this again and again because this is our company has this statement and we sell smart metering solution. All along i have been thinking that we predict load at the consumption point and hence lesser load the utility orders to generation company and Utility save money. But you guys all convinced me that no extra energy is created at all in the first place. So what are the advantages of smart meters for Utilties (Apart from auto disconnect, Usage patterns, Meter statuses etc)

My question when Utility commit(purchase) commit for some Energy say 100MW , Is Utility suppose to pay even thought the 100MW is NOT consumed? Is that we are talking ?

Read the link http://www.powerwise.gov.ae/en/research/programmes-projects/demand-side-management.html

http://etap.com/substation-automation/substation-automation-images/demand-side-management-software-large.jpg

 

[Jeff Rosenbury]

My question when Utility commit(purchase) commit for some Energy say 100MW , Is Utility suppose to pay even thought the 100MW is NOT consumed? Is that we are talking ?

That's a contractual question I don't know the answer though I suspect it depends on which market they buy it from. I'm sure there are penalties involved. They could also recoup some losses by selling some excess on the real time market.

 

[anorlunda]

My question when Utility commit(purchase) commit for some Energy say 100MW , Is Utility suppose to pay even thought the 100MW is NOT consumed? Is that we are talking ?

Now you are wasting our time. You had a direct answer to that question in #7. I'll repeat it.

Actually, that is a very good question. In your first post, you asked about earthing the excess power. Excess power is never generated.

But now you make it clear that you are asking about payment. Yes, if the power plant contracts 100MW in the day-ahead market but only 95MW is needed in real time, he is still paid for 100. But consumers pay only for the energy they use.

Where does the money come from four that 5MW difference? All power plants and all utilities contribute to a fund to pay for someone to run that DAM market. The money comes from that fund.

Suppose the power plant sold 90MW in the DAM market but the next day 95 is needed. The extra 5MW are purchased on the real time RT market where prices may be different than DAM prices. The RT market is calculated every 15 minutes.

The markets are motivated to be as accurate as possible in how much they buy in advance.

Edit: demand response can be thought of as a negative way to balance load and generation. If an extra 10MW is needed, the market can pay a power plant to make 10 more, or it can pay demand response to demand 10 less. Demand response is not normally sold in the DAM but only in the RT market. The prices paid for demand response are much higher than power plant purchases. Therefore demand response is not used unless an emergency is near.

 

[jim hardy]

what is the sort of time involved between increasing the rate of coal feed to an increase in generator output?

ReReading that question i may have mis-answered.

In our plants the heat source, boiler or reactor, follows the turbine. An increase or decrease in generation causes the fuel pumps to inject more oil and/or natural gas , or causes control rods to move in the appropriate direction.

Generation is sensed by pressure drop across turbine , from a sensor just downstream of the throttle valves.
That signal gives feed-forward to combustion controls in the fossil and to rod control in the nuke .
It also gives feed-forward to boiler(called steam generator in nuke) water level controls.
Final control trim of heat generation is by pressure in the fossil or by temperature in the nuke.

So the delay you mention is actually the other way - change in generation precedes change in heat production but only by the time for signals to propagate through the automatic controls and move the final actuators - seconds at most.
Boiler follows turbine.
Remember - we follow the customer not the other way round...

BWR's might be different, and we have some good BWR guys here..

old jim

 

[sophiecentaur]

So the delay you mention is actually the other way - change in generation precedes change in heat production.
Boiler follows turbine.

I understand what you are saying.
It's a bit of a chicken and egg situation but in normal feedback ('loop' is the relevant term) systems, the 'slew rate' of the Amplifier (or equivalent) is the limiting factor to the possible reaction time. In this case, it is the lag in getting fuel to burn, produce steam and to vary the power from the turbine that must be the critical factor. Your feedback signal is the departure of the turbine output from what's required and that can be detected in a few hundred milliseconds but there is a limit to how fast that signal can be used to vary the energy supply to the generator etc.. I guess what I wanted to know is how fast the response is to a step change in fuel input when there is no feedback - the Open Loop performance - which is what determines the possible closed loop performance. (That's just the simple feedback theory I was taught)
Your reply is more to do with the system with the loop closed but the 3-5% per minute is meaningful to me. You would need to dump say ten minute's worth of power if the demand suddenly dried up (a broken single line feeder, for instance).
But you mentioned feed forward, which is the way to deal with high values of lag.

 

[anorlunda]

Response rates aren't just a simple time constant. In many cases, they are restricted to a ramp rate, X %/minute.

Here's a short list of response rates versus generation type. 1 most responsive - 6 least.

1. Batteries/flywheels
2. Hydro
3. Gas Turbines
4. Combined Cycle
5. Gas fired boilers
6. Nukes & Coal
7. Solar and Wind

I put solar & wind last because they don't respond to grid commands at all (Unless you trip their breakers). In fact there is a 50% chance that the gust of wind or passing cloud might increase or decrease their output in opposition to the grid's instantaneous need for incremental changes. They are bad actors regarding grid frequency control.

In theory, some wind generators could change the blade pitch dynamically to be able to respond, but I never heard of them doing that. They like to run at 100% anyhow, leaving zero room to respond to a + change.

 

[jim hardy]

I guess what I wanted to know is how fast the response is to a step change in fuel input when there is no feedback - the Open Loop performance - which is what determines the possible closed loop performance. (That's just the simple feedback theory I was taught)

I'm not quite sure how to approach that. No feedback = no valve motion ,
so i'll venture
A step change in heat production would raise temperature per heat capacity of the boiler system ,
raising pressure hence density hence mass flow rate of steam
hence raising rate of heat removal,
so you'd approach a new equilibrium pressure, exponentially, as heat removal rate approaches new heat input rate.
That'd take probably twenty minutes to quit moving visibly. If that's five time constants then there's a 4-minute-ish time constant.

Point being there's natural feedbacks inherent in the machine. Thermal capacity is high enough it's inherently slow so it needs feedforward as you said...
That's why in our fossil boilers we control fueling rate by pressure and give it feedforward from that turbine first stage pressure signal, which is rate of heat removal.
Bumping the governor so throttle valves go a teeny bit further open causes an almost immediate pressure dip, an upswing in the boiler level as steam bubbles expand, and a near simultaneous increase in firing due to feedforward.
To best of my ancient memories pressure returns in about a minute, with little overshoot as in any well tuned control system.
For the size of the machinery and given its considerable thermal storage it's surprisingly nimble .

The nuke has internal natural feedbacks too
opening throttle valve cools the boiler which cools the water returning to reactor which makes reactor power increase... that takes just a very few seconds.. but the temperature coefficient of reactivity is designed to retain a lot of margin to stability and you'll still have to move rods to settle at the new power.

@anorlunda have you ever done analytical work on control systems that might help Sophie ? I only kept them running , have more of a hands on understanding of them than a theoretical one.
Our fossil gear was pneumatic analog Bailey ... nuke was analog Westinghouse 7100.

Sophie - best i can do is say i think you're looking at several minutes for natural "slew rate"
reduced to somewhat less than a minute by feedforward in the control systems.

old jim

 

[sophiecentaur]

Sophie - best i can do is say i think you're looking at several minutes for natural "slew rate"
reduced to somewhat less than a minute by feedforward in the control systems.

Thanks. That is ball park enough to make me think I have a better grasp than I started with. That could account for a fair few MWh of 'slop'.

 

[jim hardy]

I should know the heat capacity of the plant
but it's one of those numbers that just slipped away with twenty years of non-use.

That could account for a fair few MWh of 'slop'.

Slop ?

A MWh being ~3.4 million BTU... the heat stored by 3.4 degreeF in just 120,000 gallons of water...
Sounds reasonable enough !

 

[sophiecentaur]

I should know the heat capacity of the plant
but it's one of those numbers that just slipped away with twenty years of non-use.
A MWh being ~3.4 million BTU... the heat stored by 3.4 degreeF in just 120,000 gallons of water...
Sounds reasonable enough !

I guess my appreciation of the significance of 'Quantities' in Power Engineering is adrift somewhat! That amount would certainly help me with my Electricity Bill - but I am just an individual.
A point I made, earlier on, about the resulting change in Volts for the other consumers when a load comes or goes, wasn't picked up but that really must be relevant. The system has a vast 'smoothing capacity' hanging on it, in the form of all the other consumers.
My concern was about a single power station problems. But that is not relevant to running a whole network.