Preface. This is a book review of Angwin’s 2020 “Shorting the Grid. The Hidden Fragility of Our Electric Grid”. It is a good primer on how the grid works, especially why Volt-Ampere Reactives (VARs) are important and why renewables don’t provide them, yet another reason renewables can’t replace fossil fuel generated electricity (which is two-thirds of electricity in the U.S.). VARs are essential for keeping the grid stable and not coming down, which Angwin describes as a bit like “riding a bicycle. The energy you put into the pedals will move the bike forward,
but you also have to put some energy into maintaining your balance, or you’ll fall over and won’t be able to move forward at all. If you are a good bicyclist on a smooth road, the “maintaining your balance energy” will be small. If you are a poor bicyclist who swerves around a lot, or if you’re on a bad road, the “maintaining your balance” energy will be larger. In either case, the “maintaining your balance” energy is necessary. That energy is also a parasitical drain on your energy effort: it doesn’t move the bike forward. A well-run grid is like a good bicyclist on a smooth road. Rotating electric machinery puts VARs on the grid, and if the entire grid was thermal (nuclear, gas, coal) and hydro units, there would rarely be a problem with VARs. These systems all run with rotating electric machinery. But wind turbines and solar make direct current that needs to be changed into alternating current, and that process does not put VARs on the grid in the same fashion. (Some older and bigger wind turbines do put VARs on the grid.) Messing up the VARs can also mess up the grid, so this is another place where the BA must be aware of what is happening on the grid.”
I’ve written two books now on why the electric grid will come down for good when fossil fuels are scarce, especially natural gas declines since other forms of dispatchable power like hydropower and batteries can’t be scaled up enough to provide storage. Natural gas is the main electricity storage today. Meanwhile, this is a great book to understand how the grid works NOW, and why a distributed grid may not be in your future. A few excerpts:
- The amount of lithium needed for backing up one large 1000 MW modern coal plant for 100 hours would require 32,000 tons of lithium. In 2018, the global production of lithium was 62,000 tons. So it would take more than half a year’s worldwide production of lithium to back up a single large coal plant.
- No agency is in charge of ensuring that there are enough power plants and power lines to keep the grid operating. In Regional Transmission Organization (RTO) areas, the grid is becoming more fragile and more expensive. In the near future, “rolling blackouts” may become common in many RTO areas. This book is about why this will happen.
- The only seasonal backup for renewables is fossil fuels (and whatever hydro is available). Similarly, there is more wind at the change of seasons (spring and fall). There are more windless and very hot days in summer, and more windless and very cold days in winter. Again, it is unreasonable to consider that charging a battery in windy March will provide power to run air conditioners in August.
- In the RTO areas, the grid is being moved inexorably to a strong reliance on intermittent renewables, coupled with an equally strong reliance on just-in-time natural-gas delivery as backup. Several different scenarios could cause this system to collapse. Natural gas, the cure-all fuel for the grid and for houses, could become scarce or expensive. A long, cold windless period or a long, hot, humid period could overwhelm the natural-gas supply for electric generation. Wind generation is often hundreds of miles from the load center, adding another level of vulnerability to the transmission system.
Quite frankly, I’m amazed the system isn’t crashed by the complex, inadequate and sometimes incompetent bureaucracy that tries to keep the grid balanced…
Alice Friedemann www.energyskeptic.com Author of Life After Fossil Fuels: A Reality Check on Alternative Energy; When Trucks Stop Running: Energy and the Future of Transportation”, Barriers to Making Algal Biofuels, & “Crunch! Whole Grain Artisan Chips and Crackers”. Women in ecology Podcasts: WGBH, Jore, Planet: Critical, Crazy Town, Collapse Chronicles, Derrick Jensen, Practical Prepping, Kunstler 253 &278, Peak Prosperity, Index of best energyskeptic posts
***
Angwin M (2020) Shorting the Grid. The Hidden Fragility of Our Electric Grid. Carnot Communications.
In many areas of the country, but especially in RTO areas, power installations that can operate only intermittently, such as solar and wind installations, are the sure bet for becoming wealthy. In the mortgage situation, the intrinsic value of the mortgage didn’t matter. In the RTO area, the value of the power produced doesn’t matter. As a matter of fact, less-valuable power is more profitable. Trouble is sure to come, and it is on its way. In these areas, we are on our way to an expensive and fragile grid.
IN THE RTO AREAS, no group or agency has the responsibility for grid reliability. This agency can do a little of this, and that agency can do a little of that, but no agency is charged with ensuring reliable power. No agency is in charge of ensuring that there are enough power plants and power lines to keep the grid operating.
In RTO areas, the grid is becoming more fragile and more expensive. Fragility is the most dangerous problem. In the near future, “rolling blackouts” may become common in many RTO areas. This book is about why this will happen and what we can do to prevent those blackouts. What about the “free market,” which could conceivably use its invisible hand to bring reliable electricity to the customers? There is no free market. There are false markets, ruled by insider decisions.
In my opinion, a grid meltdown is coming. Reliable power will become part of the Good Old Days that parents tell their children about. Unlike the heroes of The Big Short, I am not in a position to place some sort of bet that will make me rich.
THE ELECTRICITY MARKETS in the RTO areas serve nobody well. First of all, they aren’t markets. Most types of plants are constantly on the search for their “missing money”: the RTO regulations do not allow them to recapture even their costs. Incentives of various types lead to fragile grids, and nobody is the wiser. Huge decisions are made in closed rooms with only insiders (called “stakeholders”) present, and the press is often not allowed.
The grid in RTO areas has a closed-door governance. In many cases, you can’t find out what is going on and what decisions are being made. Why should I write a book saying: “This is how you are going to get beaten up. Get used to it”? I’ll start with a description of the energy auctions in RTO areas. We have to understand the way kilowatt-hours are bought and sold on an RTO grid before we can understand anything else.
Vertically integrated utilities did have some important advantages over the RTO system. The main advantage of the vertically integrated systems is that those systems had clear lines of responsibility. To move power long distances, with relatively low line loss, you need high voltages. That is why you can see really tall transmission towers crossing the countryside. The height of the tower is a rough indication of the voltage of the lines. Taller towers mean higher voltages and, therefore, less line loss. Nobody would build such high towers unless they were moving a lot of power.
What is line loss? As you move power, some of it is wasted as it moves. In general, power is wasted two ways: resistance and electromagnetic radiation. Resistance heats up the power lines. Electromagnetic radiation (radio waves) are the “static” you hear when your car drives under a power line. Both these phenomena are well understood by electrical engineers and cause very few problems for the grid. However, line loss does make moving power more expensive, since not all the power will arrive at its destination. As a rough rule, the longer the line, the more power will be lost.
The Balancing Authority (BA) makes sure that the supply of power on the grid is exactly matched with the requirement for power. Always. The grid’s requirements for power varies during the day. Power plants must be called online as the requirement rises or asked to stop producing electricity as the requirements fall.
Another problem is that different technologies have different abilities. A steam-fired power plant can put out more or less power but must change power levels relatively slowly. It runs wonderfully when it runs steadily, but it takes some time to turn around. By design, a steam plant is a steady, solid producer. Of course, a steam plant can change the output.
Even the internal combustion plants are not “instant” the way you need to have electricity be “instant.” So there’s this whole other thing called “ancillary services” on the grid: this basically consists of paying plants to be on various types of hot standby, often with turbines spinning (but no load), ready to send their power to the grid very fast, when called on. Of course, plants can’t just keep their turbines spinning out of the goodness of their hearts: they will have to pay for fuel. The Balancing Authority has to arrange for spinning reserve and pay for it. Spinning reserve, which is fast to get on the grid, will cost more than slightly slower-to-the-grid spinning reserve.
Renewables often make the BA’s job harder. They depend on real-time conditions (sun, wind), and the BA cannot order them online to match requirements. The BA can ask them to get offline (called “curtailing”), however. So the BA’s options are more limited for solar and for wind.
Wind is also famously spiky: while grid demand changes slowly, the wind starts up and dies down with comparative suddenness. Poor transmission connections plus spikiness mean that the wind is frequently curtailed.
The neck of the duck curve is becoming an issue in areas with a lot of solar. Solar input to the grid tends to be highest during summer afternoons. However, when the sun goes down, the solar goes offline rapidly. The BA then orders the dispatchable plants (thermal plants, hydro plants) to ramp up, and they often have to ramp up faster than the solar is ramping down. Faster because people have a tendency to turn on a light as the sun sets, or come inside and begin cooking dinner, and so on.
There’s a rule of thumb on the grid that no plant should be so big that it is more than 10% of the average demand on the grid. One plant going offline should not take 20% of the grid’s power with it. People look at solar as a distributed system: my rooftop, your rooftop, and a solar array down by the Interstate. No huge power plants here! However, in fact, solar often acts like a single megaplant, which switches off in the early evening.
Volt-Ampere Reactives (VARs)
There is another property of electric current, a property with which most of us are not familiar: the VAR. The BA must keep voltage within a narrow range and balance demand on the grid (amps). The BA must also make sure the VARs are in balance. Alternating current has electromagnetic properties that have to be kept in balance. When a person attempts to explain VARs, they often end up giving incomprehensible explanations,
Hittinger makes an analogy to riding a bicycle. The energy you put into the pedals will move the bike forward, but you also have to put some energy into maintaining your balance, or you’ll fall over and won’t be able to move forward at all. If you are a good bicyclist on a smooth road, the “maintaining your balance energy” will be small. If you are a poor bicyclist who swerves around a lot, or if you’re on a bad road, the “maintaining your balance” energy will be larger. In either case, the “maintaining your balance” energy is necessary. That energy is also a parasitical drain on your energy effort: it doesn’t move the bike forward. A well-run grid is like a good bicyclist on a smooth road.
Rotating electric machinery puts VARs on the grid, and if the entire grid was thermal (nuclear, gas, coal) and hydro units, there would rarely be a problem with VARs. These systems all run with rotating electric machinery.
But wind turbines and solar make direct current that needs to be changed into alternating current, and that process does not put VARs on the grid in the same fashion. (Some older and bigger wind turbines do put VARs on the grid.) Messing up the VARs can also mess up the grid, so this is another place where the BA must be aware of what is happening on the grid.
Vermont had an interesting illustration of this issue. The Kingdom Community Wind Project connected to the grid without installing synchronous condensers. The developer said that it would install such condensers, but the wind project connected to the grid without them. Without the condensers, the grid operator had to curtail (not accept) the wind power quite often, mostly due to VAR mismatch.
After the owners invested $10 million in synchronous condensers, the VAR problem was mostly solved, and the wind turbines were not curtailed as often after the condensers were installed.
RTO payment systems
The physical operation of the grid faces more challenges than I have listed here, including gas-fired plants not getting fuel supplies when home heating is using lots of gas on cold winter days, and so forth. However, in general, physical issues are well understood by engineers and operators. The BAs know how to keep the lights on, and they generally do it. The moment-by-moment successful balance on the grid makes every functioning grid into one of the wonders of the modern world. We must now move to the more depressing area of payment and policy, which is quite capable of wrecking what the BAs and engineers have achieved.
Understanding the RTO payment system requires insider knowledge, willingness to use acronyms without examining them too carefully, and a cheerful willingness to spread high prices to low-income parts of society. Some RTOs are called Independent System Operators (ISOs). In California, CAISO.
Before 1999, all utilities were “vertically integrated.” In much of the country, this is still the case. The utility had requirements to keep the lights on: it reported to a local board of some kind, and it would be judged and possibly fined if there were too many outages. The utility also got paid a rate-of-return on its expenditures. For example, if the utility had to build a power plant, it was allowed to charge enough money to build it and to make a rate-of-return profit on building the plant and running it. However, in many jurisdictions, a utility’s rate-of-return could be lowered if its customers’ lights went out too often.
Before RTOs, it was in the utility’s best interest to spend “whatever it takes” to keep all the power plants and transmission systems in top shape and to have crews ready to fix problems very quickly. In other words, the higher the reliability, the better the chance of a high rate of return on the utility’s investment.
The second way for a utility to make more money would be to invest more money, resulting in a bigger “rate bas”. This was basically a perverse system. The way for a vertically integrated utility to make more money was … to spend more money. Unlike Joe’s deli down the street, which was always trying to be more efficient with money, in order to have larger profits, a vertically integrated utility wanted to be less efficient with money, so they could have a bigger rate base and larger profits.
One of the ways that utilities made more money was to overbuild power plants, that is, build too many power plants. Presenting a somewhat inflated projection of electricity needs in the future could justify a new power plant, which would be approved by the local regulators. The cost of building the plant would go into the utility’s rate base. In this sense, a new power plant generated utility profits, whether or not the electricity was needed. The utility couldn’t just build things and spend money. A state regulatory agency had to approve the expenditure. Depending on the state, this agency was usually called the Public Utilities Commission (PUC) or the Public Service Board (PSB).
At the state level, utilities tended to get a little cozy with their regulators (“regulatory capture”). Overbuilding and overcharging customers was often the result. On the other hand, the utility had incentives for building a very reliable grid. LOOKING AT THE PERVERSE incentives for integrated utilities, many people had the idea that a market would fix these high prices—if the utilities could just be forced to compete in a market.
One would think that deregulating the utility markets would mean the consumers are no longer ratepayers. Those ratepayers, captive to their local utility, would actually become customers—customers who made choices. Customer choice is one of the most important ways that markets work. That is how deregulation of airlines worked: choose your flight, and airlines can now compete on price.
Deregulation did not give consumers a choice.
Only 18 states have electricity choices. In most of these states, only some consumers are allowed to have a choice. The consumers who have a choice may be large industrial or commercial consumers,
And in California, some big customers get some choices.
Indiana lies in MISO territory, but consumers have a choice of only natural gas providers, not electricity providers.
It wasn’t about turning ratepayers into customers who made choices. So why would we expect it to be good for the customers? Spoiler alert: Utility “deregulation” hasn’t been good for the customers. As a matter of fact, it isn’t deregulation.
The regulations for RTO areas are longer and more complex than those for areas that didn’t choose to form an RTO. Also the RTO areas have higher prices for electricity than non-RTO areas. Generation utilities own facilities that generate electricity. Distribution utilities own the distribution systems that distribute power. Distribution utilities are usually “regulated” in the same way as vertically integrated utilities have been “regulated”: with a rate of return. In contrast, generation utilities are supposed to compete in energy markets and auctions. Independent LSEs provide customer choice, but they are available in only some parts of some states.
The central features of RTO areas are merchant generators, which are separate from distribution utilities, and generators that compete through auctions.
The distribution system causes the most frequent types of power outages. Wind, ice, and tree branches disrupt utility lines. In the RTO areas, the distribution utilities are regulated, and they have the same incentives for reliability that they had before. They get a rate of return on their expenses. As expected, most distribution utilities continue to do a good job. In some areas, distribution utilities also own power plants, which compete on price as merchant generators. In these areas, the distribution system can suffer under deregulation. If the generator portion of the company is losing money, distribution is sometimes an easy place to cut expenses. Distribution utilities buy power from generation utilities. The distribution utilities are still regulated monopolies, with every incentive to keep their power lines and substations in good repair. The distribution utilities buy the power at auctions run by the RTO. In these auctions, the power plants compete with each other in auctions that run at five-minute intervals.
The “queue” (dispatched plants) is the set of plants that bid the lowest price per kWh, and the plants in the queue provide enough power for the current real-time electricity requirements on the grid. If the load requirements are low and a plant’s electricity is expensive, the plant won’t make it into the queue at that time. The power plants want to make it into the queue. They want their power to be dispatched, and they want to be paid.
The distribution utilities buy kWh from the generators, and the RTOs run the auctions. With the auctions, power plants face a new situation, and the plants have very different incentives than they had within vertically integrated utilities. The auctions between power plants removed each power plant’s incentives for reliability.
The grid itself can become less reliable. In my area, studies of the future of the grid show the strong possibility of rolling blackouts in winter weather, due to insufficient generation to meet the high demand. Without electricity, thermostats won’t work, and fans and pumps won’t work (traditional heat sources such as forced air and hot water heating). More modern methods won’t work, either. Pellet stoves and cold-weather heat pumps need electricity. A traditional wood stove will work though.
Generators
I have friends who have their own generators for backup. However, asking everyone to buy a diesel generator is not a civilized way to run a grid. Generators are expensive, and they need to be kept in good trim. They are not suitable for everyone. They are completely impractical for people who live in apartment buildings. A generator is only a short-term fix. When the grid is down, there will be no electric pump working at the gas station to refill your diesel. Even with a home generator, after you have used the fuel you store at home, you are still dependent on the grid.
For the grid operator, this dependence on natural gas is a potential and actual problem. Gas is delivered through pipelines: it is just-in-time delivery. There is very little storage for gas at a power plant, though there is storage to feed the pipelines. But no matter how much storage can feed the pipelines, the pipelines can carry only a certain amount of gas at a given time. Since electricity is made and used instantaneously, and natural gas for power plants is delivered just in time for use, sometimes the two “just in times” don’t mesh,
The pipelines cannot deliver enough gas to the power plants in winter, because many homes are heated with gas. Homes have first priority for natural-gas supplies. On a cold day, the pipelines cannot meet both sets of needs: they cannot deliver enough gas to homes for heating needs and simultaneously deliver enough gas to power plants to make electricity. In this situation, gas-fired power plants may not be able to get fuel.
In many ways, having so high a percentage of natural gas on the grid was “cruising for a bruising.” Electricity must be produced at the exact moment it is needed. Oil, coal, and nuclear plants are prepared for demand variations because they store fuel on-site. Many gas-fired plants can also burn oil if they have it, and the bulk of reliability-program money goes toward paying such gas-fired generators to keep oil on-
If someone was in charge, in the RTO areas, the simple solution would be to order a certain number of gas plants to retrofit themselves to burn an alternate fuel and then to order such plants to stockpile that fuel. Their payments for the fuel stockpiles would go into their regulated rate base, and everyone would be happy. Adding to the rate base in order to increase reliability was (and is) a common strategy in areas of the country where the utilities are vertically integrated. Oops, there isn’t a grownup in charge in the RTO areas. There is nobody who can promise a power plant a better rate of return for being more reliable.
It was uphill work for ISO-NE to set up the Winter Reliability Program, and FERC required them to dismantle it after three years. The RTOs set up many auctions, including new auctions to correct the results of older auctions. If there is a problem on the grid, an RTO will usually try to set up an auction to solve it. To satisfy FERC, ISO-NE had to come up with a plan to encourage fuel storage without actually mentioning fuel storage. Well, RTOs are not heavily staffed with lawyers for nothing! ISO-NE came up with a complex, clumsy scheme for Pay for Performance, in which plants that don’t go online when called to do so have to pay part of their capacity payments to plants that come online. The scheme worked out so that the penalty payments would encourage plants to (let me whisper the words here) “keep fuel on-site” to avoid the penalties.
A demand-response customer would be paid for not using electricity, just as a power plant would be paid for keeping oil or LNG on-site. Unfortunately for ISO-NE and the grid, there usually isn’t a big response to demand response. In 2013, ISO-NE bought 1.9 million Megawatt-hours (MWh) worth of bids for the Winter Reliability Program. Only 4,000 MWh of demand-response customers bid into the auction. The rest of the bids were for oil on-site.
Very few groups want to go without power in winter. In fact, demand-response bidding functioned as greenwashing for the oil-storage program. The idea of Pay for Performance is that plants will be penalized if they don’t go online when the grid is stressed. Hopefully, the fear of a penalty will encourage plants to keep oil on-site, without the RTO actually mentioning “oil.
ISO-NE has attempted to set up rewards and penalties that will work as well (they hope) as simply paying plants to keep fuel on-site. The complex set of payments and penalties is meant to force plants to go online in bad weather. These payments and penalties, carefully defined by the RTO, are supposedly an important aspect of a Free Market for electricity. As a matter of fact, the entire RTO system is a complex set of payments and penalties that are supposedly in the service of a Free Market.
Ma Bell, a monopoly that mostly ended in 1982. Long-distance calls were madly expensive. Not content with owning the phone lines, the phone company also owned your phone and charged you more if you had an extension phone in your house.
RTO retail electricity prices have stayed higher than non-RTO retail prices for 25years.
What went wrong with utility deregulation?
No transparency: The stakeholders (insiders) meet in secret or semi-secret. They all have very good economic reasons to use their full power to influence which power plants are on the grid and how those plants get paid. Mere ratepayers will never learn what “stakeholders” share with each other.
No accountability: Nobody is accountable for the grid. A power-plant owner can run his power plant well or badly: it’s not his plant’s fault if the grid can’t get enough power. Just-in-time natural-gas plants can take over a grid, and just-in-time deliveries can fail in cold weather. It’s not the RTO’s problem to keep a mix of plants on the grid, for resilience. In RTO areas, generation utilities own power plants, and distribution utilities own power lines and substations. Distribution utilities still have regulated rates of returns and have to keep their lines and substations in good shape. Generators do not have that level of responsibility. The RTOs that manage the whole system do not have a high level of responsibility. Wherever you are in an RTO area, unless you are a distribution utility maintaining local lines, the buck never stops with you.
IN CALIFORNIA, AS ELSEWHERE, the distribution utilities were responsible for keeping their lines and substations running. The lines and substations ran fine. The RTO (CAISO) was responsible for dispatching the plants that were available for dispatch. If too many of them were offline for “maintenance” and there had to be “load shedding” (rolling blackouts), well, CAISO is not responsible for plant maintenance, just for dispatch.
THE SYSTEM WAS SET to be gamed. Opportunities arose. Enron, Shell, and NRG owned several plants that fed into the California grid. They closed plants for “maintenance” in order to create scarcity and drive up the wholesale electricity price. Their other plants received the new high prices. Lack of generation led to rolling blackouts in California.
Meanwhile, the state of California knew the markets were being manipulated. They complained to the RTO but got no help from that source. The state then complained to FERC, the federal regulator. FERC also offered no help whatsoever.
In New England, there was nobody who could order a plant to keep oil on-site. In California, there was nobody who could order a plant to go online if it claimed it needed maintenance. In 2015, a federal appeals court affirmed that the California market had been manipulated in 2002 and that nobody (including FERC) had stopped it. The Sacramento Bee carried an opinion piece about this ruling, including the unanswerable question: What took everybody such a long time to do something about this? Market manipulation was successful in California because there was nobody responsible for making the system work. In an RTO area, the buck never stops anywhere. Not even today.
The RTO areas are more heavily regulated than the non-RTO areas. They are not markets as we know markets. They are complex systems, with new regulations constantly tweaking and trying to improve existing regulations. They are a bureaucratic thicket, not a market. It’s Orwellian. RTOs are “deregulated” only if “deregulated” actually means “lots more regulation. “War is peace.” “Deregulation” is “lots more regulation.” Orwell would be amused.
LET’S EXAMINE THE INCENTIVES for a power-plant owner in an RTO area. The owner has to decide: Will this power plant run often enough to generate enough revenue for its upkeep? Will it be a hard winter (my power plant will run a great deal, and I will make lots of money) or a mild winter (maybe not so good for me)? How much will my fuel cost? What will be the price on the grid? Each power-plant owner had to make a decision about paying for maintenance of the power plant. If an owner has more than one plant, the owner must decide separately
All these decisions looked dangerous for reliability. Companies could just shut down their power plants. Too many power-plant owners could decide that their power plant wasn’t going to be paid enough, and then there wouldn’t be enough power plants on the grid. Or power-plant owners could game the system by shutting down power plants (as they did in the early days in California).
In many RTO areas, plants sell “available capacity” whether or not they are making power at the moment. Power plants get capacity payments, and they can use these payments to maintain their plants. ERCOT, the Texas RTO, and CAISO, the California RTO, did not implement such capacity payments. As I write this, ERCOT is facing very low reserve margins this summer. California expects to have sufficient margins, partially due to abundant hydro energy from a generous snowpack.
The least-expensive plants are dispatched first. If the grid needs 1000 MW of power, and a low-priced plant can meet that demand, the low-priced plant will be dispatched to meet it. If the grid needs more power, and the only plants remaining are higher-priced, then they will be dispatched next. All other things being equal, the dispatcher will choose to dispatch plants that are near the load so there won’t be excessive line loss. The dispatcher must also keep track of how much electricity a line can carry, and not overload any of the lines.
An RTO needs 500 MW for the next five minutes. Offers: Plant A: 200 MW at 15 cents per kWh. Plant B 100 MW at 20 cents per kWh. Plant C 300 MW at 30 cents per kWh. The RTO might reply: “Okay, plant A and plant B, I’ll take all your output. Now I have 300 MW. Plant C, I’ll buy 200 MW of your output, but I don’t need all 300 MW. I’ve got my 500 MW now. “All you plants, you get 30 cents per kWh for your output. Plant C has set the clearing price for this round.” At this point, the people in the class begin to shake their heads: “So plant A bid in at 15 cents and is getting 30 cents?” The answer is: “Yes, indeed.” That is how the auctions go.
The RTO controls the bids. They know your fuel costs. They know your heat rate, which is how efficient your plant is at turning fuel into electricity. They can calculate the cost of your next kWh. It doesn’t matter, really, what you want to bid. The RTO knows the marginal cost of the next kWh your plant makes, and that is what you are going to bid. They are checking.
Capital costs are not allowed in your calculation of how much you bid. However, outside sources of income are allowed and welcome. If you are receiving payments for Renewable Energy Certificates, you can take those payments into account and bid at a lower cost for your kWh.
This system doesn’t actually help steam plants that much, because when renewables are forcing the price on the grid down to one cent, zero cents, or minus one cent per kWh (we’ll pay you to take our power), steam plants often choose to keep running and take the price, far below their operating costs, for several hours a day.
The supposed fear is that “Pay as Bid” would lead to higher prices. In fairness, that could be an issue for power plants, where there are just not that many bidders (unlike the hundreds of farmers that could supply zucchini). There are relatively few big generators on any grid, but they don’t have monopoly power. Grids have a comparatively small number of large plants that run all the time. Most grids have a great many power plants that don’t run all the time but come on the grid to follow load.
In my opinion, these part-time plants could undercut a large generator that was charging too much and acting like a monopoly. In real markets, after a while, everyone learns how to bid properly to make a living. Also, real markets tend to be a little cutthroat, which tends to benefit the consumer.
I have concluded that the reason people don’t know much about the grid in the RTO areas is because people can’t find out what is happening on the grid or who is controlling it. The RTO areas have extraordinarily complex rules. Only insiders can follow the complexity. In contrast, in vertically integrated utility areas, an ordinary person can go to a state PUC hearing and figure out what is going on. The RTO areas often actively discourage public participation. My own power pool (New England Power Pool) bans reporters from its important meetings. Other RTOs don’t explicitly ban reporters, but their rules are so complex and their meetings are so lengthy that only industry journals (RTO Insider, Utility Dive) report on their activities. Only “insiders” go to the meetings.
A person can learn about NEPOOL by reading the bylaws and the sector rosters and the minutes of previous meetings. In other words, you can learn about NEPOOL by reading what NEPOOL chooses to publish. The NEPOOL meetings themselves are closed to the public. Among other things, this means that no reporters can attend the meetings as reporters. One can hope that Boards of Directors can provide some outsider perspective on what the staff of the organization is doing. NEPOOL does not have a board of directors. NEPOOL officers are chosen from within the Participants Committee.
Most people agree that plants should be ready to run on a cold day. Yet each individual plant is owned by a different entity (some of the entities owning more than one plant, of course), and nobody is in charge of the grid. This makes the situation virtually impossible to manage. Nobody can just say, “These twelve plants are our designated cold-weather plants, and we will pay them to keep fuel on-site.” In an RTO, that would not be fair to all the other plants. In an RTO, there clearly needed to be a new type of auction or something. The RTO must strive for “fairness” and “to keep the grid functioning.” Somehow, for “fairness,” being “fuel-neutral” has become a major obligation of the RTOs. How to get plants to keep fuel on-site without actually saying, “Keep fuel on site”? It is a difficult problem but provides much employment for grid experts and lawyers.
FERC is supposed to be fuel-neutral. Any special support for low-emissions plants is supposed to come from the states, from other federal regulators (such as the EPA) and so forth. And yet, somehow, when it comes right down to it, the NEPOOL plan does not reward nuclear, coal, or hydro for being available when the grid is stressed, the ruling was another de facto vote in favor of gas-fired plants.
The RTO grids are moving toward fragility and rolling blackouts. Personally, I don’t think that the RTOs are capable of solving this problem.
When there’s not enough supply of electricity to meet demand, a grid operator cuts power to one section of the grid to keep the rest of the grid from failing. After a while, the operator restores the power to the blacked-out area and moves the blackout to another section. That is a “rolling blackout.
“Scheduled” is too strong a word for the status of that power line. Conservation Law Foundation supports the line, and the line has many of its permits, but my own experience is that such lines are problematic to build. For one thing, they run through one of the two north-south states (New Hampshire and Vermont), but they don’t deliver any power to those states. The power goes to Massachusetts. This means that people in the north-south states have very little reason to welcome the lines. Consequently, people in Vermont and New Hampshire are likely to protest the lines, lie down in front of the bulldozers, and so forth.
In assuming we can get more electricity from Canada (or even the same amount as we obtain now), the Synapse Report doesn’t look at important evidence: the history and the contracts of imports from Canada in the winter. ISO-NE was modeling what happens in cold weather, because that is when New England is most likely to have power shortfalls. Canada has really cold weather at the same time New England does. Canadians are aware of this. Therefore, there have been and will be shortfalls in cold-weather power inputs from Canada. Even ISO-NE assumes we get twice the amount of energy that we are likely to get from Canada in cold weather. It takes days to order up new LNG, and there is competition for this gas going into northern pipelines (here and in Canada). There is also worldwide competition for LNG.
WE ARE MAXED OUT IN New England. For once, ISO-NE, Synapse, and I all agree: no major new natural-gas pipelines have been built in New England in many years, and none are likely to be built in the near future. In several areas of New England, natural-gas companies are refusing to hook up any more homes. Synapse assumes a very slow rate of increase in home use of natural gas.
As the offshore fields get depleted, we can expect Canada to supply less natural gas to New England than it supplies now, especially in cold weather, when both Canada and New England have high demands for gas. That doesn’t sound good for fuel security in New England.
RTO areas are often more controlled by insiders than is good for governance. Insiders have strong financial reasons to see certain conclusions finalized. For them, it is not just wishful thinking. It’s about money. The power pool groups are generally not accountable to anyone. They make up the power pools (such as NEPOOL). The power pool groups are generally not accountable to anyone. They don’t even have a board of directors. Sometimes they are closed to the press. And they have strong standing before the regulator.
In short, if ISO-NE paid for oil for winter reliability, the power plants were willing to buy and store oil. Without ISO-NE funding, they didn’t buy oil. This is perfectly in line with the perverse incentives on the grid. The power plants do best when the grid is doing worst. When many power plants have uncertain fuel supplies, and the grid is stressed, then electricity prices rise. At that point, many power plants make excellent profits. They don’t even have to manipulate the markets to do this: the incentives are perverse enough on their own.
RTOs, however, I will focus on the issues of transmission lines that are required for reliability. These are the lines for which the grid operator does assessments and assigns costs and so forth. The socialized lines. How do you socialize a line? “Socialized” means that the various distribution utilities pay for transmission in proportion to their share of the system load. For example, let’s look at my own little state of Vermont. Vermont is, in general, only about 4% of the power usage in the ISO-NE area. Therefore, Vermont pays only 4% of the cost of transmission-structure upkeep. The costs are “socialized” according to each state’s use of the infrastructure. Massachusetts pays much more for the transmission infrastructure than Vermont pays, because Massachusetts has more people and industries. Note that this payment is completely unrelated to where the lines are located: Vermont may host an important long line, but that doesn’t mean that Vermont will be charged more for the upkeep of that line.
If the grid operator feels that there needs to be a new line in New Hampshire for reliability, New Hampshire will not be charged extra money to pay for that line. Share-of-load pricing will still continue.
If your New England state wants a transmission line to bring power from distant wind turbines to its city center, in the pre-FERC 1000 days, that would be something the ratepayers of that state would pay for. After FERC 1000, one state may decide on a policy, but all states will pay for it.
When all states have to pay transmission costs for one state’s policies, FERC 1000 will set up a “tragedy of the commons” for transmission. The classic “tragedy of the commons” is when a shared resource is overused and therefore depleted, because each individual user does not have to pay for the effects of his own overuse of the system.
Why shouldn’t every state decide to buy some new transmission and meet their renewable goals—at the expense of all the other states? Why should state A pay for state B’s goals? And if state A is going to have to pay for state B’s goals, State A will try to arrange that State B pays for state A’s goals, too. This is the tragedy of the commons: every state expands its own share of the socialized payments, since all the other states are expanding their share of the payments.
More and more money will be required, and the commons will be overgrazed. In this case, “the commons” are the ratepayers. They will find their power bills increasing, and there will be nothing they can do about it.
The Northeastern grid is heavily dependent on natural gas, which gets delivered by pipeline. When homes and businesses are using natural gas for heating, power plants sometimes cannot get gas. The power plants often use oil instead, but sometimes the winter weather is so bad that it is difficult to deliver the oil. I have several chapters on the problems of just-in-time gas delivery and questionable oil delivery. Winter is tough in New England.
The utilities are urging conservation in summer because they are playing the Game of Peaks. It’s a utility game about money. If they play the Game of Peaks well, they can shift some costs from themselves over to neighboring utilities. CUTTING BACK ON ELECTRICITY use on the hottest day of the summer is not a moral imperative. It is merely part of The Game of Peaks. The percentage of power a utility uses during the peak hour is the percentage of transmission costs that the utility has to pay. If a utility can lower its electricity use for that one peak hour, it will save a lot of money by paying lower transmission costs for the grid.
Somebody is still paying that $200k for transmission: the overall cost of grid transmission hasn’t changed. Some other utility is paying that cost. As Green Mountain Power describes it: “… when utilities can lower demand during that key hour, they can create savings for customers.
Their 5 MWh of storage will not make much difference to the expense of transmission on the grid.
The New England grid runs about 10,000 MW at night and up to around 25,000 MW during a hot day. 5 MWh storage is pretty small.
Being thrifty and not using excess power is always a good thing. Still, especially if you live in New England, it helps the environment more if you are thrifty with electric usage in winter (with all that oil and coal-burning). It helps your local utility’s bottom line more if you are thrifty with electric use in summer.
Utilities are masters of greenwashing: The official announcement encourages you to help the environment. If you could hear the little man behind the curtain, he would be saying, “Do this to help my company’s bottom line.
PG&E operates more like a vertically integrated utility (it owns both power plants and transmission lines) than a distribution utility or merchant generator in an RTO area. I did need to mention the California situation, murky as it is. The resolution of the California claims will influence transmission requirements everywhere. Nobody can say how they will be influenced, but they will be influenced.
RTOs are supposed to be fuel neutral, but state policies have no such constraints. Around here, biomass is often about 40% of our renewable energy on the grid, and waste-to-heat is another 40%. When the wind is blowing strongly, wind can be 40% or even 50% of renewable energy. In that case, the percentage of renewables on the grid increases with the extra wind energy, and the percentage of biomass and waste heat shrink as a percentage of renewables. Overall, however, biomass and waste-to-heat are the steady performers.
How many times have you seen a report extolling the renewable future, and it features an illustration of a waste-to-energy plant? Never, I suspect. It’s always a solar panel or a wind turbine on the cover.
In general, a standard hydro plant cannot be baseload because the turbines cannot run all the time. To operate, the hydro power plant requires a certain water level in the pond behind it. The level of water in the pond behind the dam sinks as water goes through the turbines. When the turbines stop running, the level of water rises as the streams refill the pond. Hydro is generally used for load following, or for filling in when other renewables (such as wind turbines) go offline. Power
Nuclear plants’ capacity factors average above 93%, the highest capacity factor of any type of power plant. The average capacity factor for a hydro plant in America is about 40%, and this capacity factor mostly depends on the availability of water. In areas with many dams, hydro can provide all or most of the electricity required, including base load. While a single hydro plant is limited by its capacity factor, using several hydro plants sequentially can provide baseload power.
There are also run-of-the-river dams that do not have significant ponds and are designed to make power whenever the river is flowing. Many large run-of-the-river dams take advantage of a separate upstream dam to provide an even flow through their hydro plant. For example, the Chief Joseph Dam on the Columbia River is a run-of-the-river dam, with only a small pond. In order to operate steadily, it depends on a steady supply of water from the upstream Grand Coulee Dam.This type of run-of-the-river hydro is basically another way to use several dams sequentially to provide baseload.
their rebuttal paper pointed out that the Jacobson paper describes hydro power as providing 700 to 1300 GW. However, existing installed hydro capacity is 87 or 145 GW, depending on whether pumped hydro is included, and the most useful sites have already been exploited. Many people, including myself, feel that the fact that Jacobson even brought a lawsuit has had a chilling effect on the whole renewable-energy debate. If scientists can’t debate each other in peer-reviewed journals without fear of lawsuits, science will not be able to move forward very well.
How will a grid operator deal with this spikiness? Basically, the operator will arrange for fast-startup plants to be available for backup. However, simply having such a plant on the grid does not necessarily provide fast backup for the renewable’s spikiness. The plant must also be ready to begin operation very quickly. This often means keeping the plant running at a low level, or keeping the turbine spinning without a load, or various other ways to be sure that the plant can come up to speed quickly. Keeping the backup plant ready to start quickly may require that the plant burn some fuel. The cost of such fuel will be part of “ancillary services” paid for by the grid operator, since the backup plant’s readiness is part of grid reliability.
The NBER worldwide macro-study described above looked at how renewables affect the grid. This study showed that the grid needs slightly more fast-reacting fossil available than it has intermittent renewables installed. (1.0 MW fossil installation is needed for every 0.88 renewable installation.) This macro-study looked at many countries over two decades.
Matt Nesvisky A role for fossil fuels in Renewable Energy Diffusion NBR working paper 22454, National bureau of economic research. www.nber.org/digest/oct16/w22454.html
Hallquist showed a chart of her system’s demand on a partly cloudy day. This showed the effect of net metering solar on her system. In a typical net-metering solar installation, a homeowner has a solar installation on her roof, and it sells electricity to the grid while the sun is shining. When the sun is not shining, the homeowner buys electricity from the grid. The term “net metering” refers to the fact that the homeowner’s electric bill (or her payment from the utility) will be the net of how much electricity she sold to the grid and how much she bought from the grid.
The peak demand that Vermont Electric Cooperative needs to supply did not change due to net metering and installed solar. The system has to have the same availability of power to meet demand—as usual. The power requirements for the system were no longer steady. They were spiky as solar net-metering cut in and out of the power supply, as clouds went by. She had to have close to the same amount of nonsolar power as she had needed before the solar was installed. This micro-example illustrates the same issues as the NBER study addressed. A grid, large or small, needs as much quick-reacting fossil capacity as it has intermittent-renewable capacity.
Stop-and-go driving. This is very hard to quantify but needs to be said. Running a gas turbine in an on-and-off backup mode is like running your car in stop-and-go city driving. You don’t get the gas mileage in city driving that you get on the highway. Running an engine steadily is most efficient. So, using a gas turbine for renewable backup requires more gas per kWh than running the turbine steadily.
For a gas turbine backing up renewables, the amount of “stop and go” depends on circumstances. However, there is no question that some of the renewable advantage of “clean energy” is offset by extra gas burned inefficiently as backup.
One of the most efficient power plants on earth is a Combined Cycle Gas Turbine (CCGT). When you run such a machine steadily in baseload operation, it can turn about 60% of the energy in the gas into electric energy. In contrast, a simple gas turbine will turn about 40% of the energy in the fuel into electricity.
A combined cycle plant consists of two plants that operate together. The first is a gas turbine; the exhaust from that plant is hot enough to heat water for a steam-powered cycle. The combination of the gas turbine with a steam generator leads to huge thermodynamic efficiency. You can get 50% more electricity from a unit of fuel with a combined-cycle plant than you can obtain with a simple gas-turbine cycle.
However, the second part of the cycle is usually a steam cycle, and it does not respond quickly. To back up renewables, the fast-acting part of the plant will be utilized, and the steam cycle may or may not be used. In other words, the plant will be operating at 40% efficiency instead of 60% (its efficiency with the steam cycle). Using a combined-cycle plant in a mode that is optimized for flexibility will use considerably more fuel for the same output of kWh.
Renewables are supposed to “clean up” a grid. It is certainly true that a photovoltaic installation will make no emissions as it makes electricity, and it may displace the need for a gas-fired plant to operate. In this simple example, emissions decrease with the introduction of renewables. However, the real world is not always that simple. Emissions can also increase with the increased use of renewables. Backing up renewables can cause inefficient operation of fossil plants, leading to an increase of emissions on the grid. In other words, as renewables increase, emissions can also increase. Two studies show this effect, though both are a bit ambiguous. More renewables can mean lower emissions … or higher emissions, as the case may be.
First, look at carbon dioxide and Ireland: A study of wind (using 2014-2015 data and published in 2016) in Ireland shows that, when the fleet of CCGT (combined-cycle gas turbines) in Ireland run steadily at about 55% fuel efficiency, the fleet produces 335kg CO2/MWh. However, when backing up wind turbines, with more starts and stops and a lower fuel efficiency, the fleet produces 500–600kg CO2 per MWh.
In my opinion, I think it quite probable that ramping combined-cycle gas turbines produces more NOx than running them steadily. I worked on controlling NOx, and controlling NOx emissions is probably the most delicate balancing act in the science of pollution control. Control of NOx during the combustion cycle depends heavily on the balance between oxygen and fuel and on the temperature of combustion. Control includes controlling the air entry into the combustion process and may include shooting steam into a gas turbine to lower the combustion temperature. When a plant is changing power rapidly, these balancing acts get far more difficult and far less effective.
The ramping “neck” of the duck curve is energy inefficient. Think of how much gas your car will burn if you speed away from a stoplight as if you were in a race, compared to how much it will burn if you are cruising down the highway at a steady clip. It’s the same with a gas turbine.
Advocates for “renewables can do it all” are often most scornful of baseload plants. If coal or a nuclear plant shuts down, they celebrate. It doesn’t matter that the coal plant may have been polluting and the nuclear plant was not polluting. The sin that both these types of plants have committed is that they are not “flexible,” and we need “flexible” plants for the new grid.
I have never heard of an all-renewable advocate complaining about a high-efficiency gas combined-cycle plant needing to be run in less-efficient single-cycle mode. As a combined-cycle unit, it is also “inflexible,” which is considered to be a major problem.
All-renewable advocates are advocating for a kludge system, not a well-designed system. In their future, all plants are optimized for flexibility, because only flexible plants that can back up intermittent renewables are allowed on the grid.
Any plant optimized for steady operation, whether it is a combined-cycle gas plant, a coal plant, or a nuclear plant, is considered “your grandfather’s grid.” The emissions don’t matter. Only flexibility matters.
Vermont, sometimes solar panels are covered in snow. There is no reasonable type of battery that can be charged up on the summer solstice and provide power for the winter solstice.
The only seasonal backup for renewables is fossil fuels (and whatever hydro is available). Similarly, there is more wind at the change of seasons (spring and fall). There are more windless and very hot days in summer, and more windless and very cold days in winter. Again, it is unreasonable to consider that charging a battery in windy March will provide power to run air conditioners in August.
Negative pricing is caused by subsidized renewables, and this policy choice drives stable plants to retire. Such retirements make the grid more vulnerable to disruption. Solar, wind, and batteries depend on inverters to change the direct current that they produce into the alternating current that the grid uses. The grid runs on alternating current (AC): 60 cycles per second in the US,
The direct current from solar, wind, and batteries have to be converted to AC before it can be added to the grid. A wind turbine is rotating electric machinery, so it actually starts out by making AC. However, the speed of the rotation is determined by the wind speed. The current cannot be added directly to lines, because the number of cycles per second depends on the wind speed, not on the requirements of the grid. Therefore, wind-turbine AC is customarily converted to direct current (DC), and then the DC is re-converted to AC at the proper number of cycles per second for the grid.
Batteries and photovoltaics make DC naturally, without conversion from AC to DC. Their power must be converted to AC to be put on the grid, but they require only that one conversion cycle.
Energy from an inverter can be matched to the grid current, but it takes some action. Harmonics (overtones of the cycles per second) need to be matched, and VARs must be matched. These are pretty much solvable problems (more solvable than spikiness), but sometimes renewables are connected without the investment that is needed to match the grid constraints effectively.
Ground faults: As more power is added to the grid, transmission ground fault overvoltage (TGFOV) can occur. Under fault conditions, such as a tree falling on a power line, TGFOV can lead to serious damage to the grid, including damage to substations and transformers. When large power stations are added to the grid, the power plant developer must pay for upgrades to the substations. When homeowners add solar, the question becomes “Who should pay for the upgrade?” The homeowner whose proposed solar installation would “tip” the substation into a vulnerable state? Should that homeowner pay the $75,000 for the substation upgrade? But what about all the homeowner’s friends, who put their solar panels in earlier and didn’t have to pay the costs? Or should everyone on the grid pay for the upgrade, though only the customers with solar installations will make money by connecting to the grid? (That is why power plants have traditionally paid for the substation upgrades when they connect to the grid.) Some of the issues are described in this article in VTDigger: “Proposed solar fee raises questions about who pays for grid upgrades.”
When the costs are spread over the whole grid, paid by all the customers, the cost of the electricity will go higher, but it will not be an easy calculation to ascribe the higher costs to the presence of renewables.
Daily spikiness, seasonal reliability, and the effects of inverters (needed for wind, solar, and batteries) make it difficult to integrate renewables on the grid. The more renewables, the more difficult. When more power is spiky and making harmonics, when VARs become harder to manage, then managing the grid becomes more and more difficult. The reliability of our grid would be in danger … except for the fact that such a renewable grid would require 100% backup by quick-acting fossil plants, which could step into the breach when things get too bad.
Despite the reality of laws of nature—the sun doesn’t always shine, the wind doesn’t always blow, inverters don’t make VARs for the grid—legislators make other laws saying their state grid must be 100% renewable. The laws of nature are not repealed by these renewable-mandate laws, and yet the laws are passed. Renewable-mandate laws have unrealistic plans for renewables (to put it mildly). They will not succeed in building grids that are 100% renewable. However, such laws will succeed in making the grid more fragile and more expensive.
We have to notice that 100 MWh is quite a small amount when we are thinking at grid scale. Consider that Vermont Yankee nuclear plant was too small to be cost-effective, and it made 620 MW every hour, for months at a time. The Tesla battery can supply only 100 MW for one hour.
If only variable renewables and storage were available, generation and storage-installed capacity would have to be five to eight times the peak-systems demand. Such a system would need reserve margins of 400% to 700% of peak demand. In contrast, on our current national grid, reserve margins of around 15% of peak demand are common. This would be an unbelievably wasteful way to run a grid.
After overbuilding by five to eight times, we would waste around 100% of a year’s supply of electricity by curtailment (“Turn off your wind turbine—we can’t use the energy now”).
In a blog post, Malhotra calculated the amount of lithium needed for backing up a significant portion of the grid. He started his calculation by looking at what would be needed to back up one large modern coal plant. Malhotra calculates that providing 100 hours of backup for a single massive (1000 MW) coal plant would require 32,000 tons of lithium. In 2018, the global production of lithium was 62,000 tons. So it would take more than half a year’s worldwide production of lithium to back up a single large coal plant for 100 hours.
Nickel-iron batteries use nickel and iron, which are not in tight supply. However, nickel-iron batteries need to be charged with about a third more electricity than they are able to deliver at discharge. In other words, while these batteries are not as resource intensive for the battery itself, they are resource intensive in terms of the power that they require. They lose charge at the rate of about 20% per month, whereas lithium batteries lose charge at 2% per month. While the material to produce nickel-iron batteries is abundant, the number of batteries required would be much greater than the number of lithium batteries needed.
The problem is not with renewables or with batteries. The problem is that people aren’t planning for their use or how they might be most useful. Renewables and batteries are overhyped and are beginning to be overbuilt. Both can be helpful to the grid. Even together, they cannot be the grid.
On an RTO grid, a power plant can’t make a living by just selling kWh. The power plants need subsidies to compete with other power plants that also receive subsidies. It’s an endless battle for subsidy payments. Providing kWh to the grid is close to irrelevant. First, we will review the role of subsidies. Then we will watch FERC attempt to keep the lights on by making a rule for some generators and some competitions: set-your-price-as-if-you-didn’t-receive-a subsidy. The Minimum Offer Price Rule (MOPR) was designed to prevent heavily subsidized plants from completely dominating the market. Then we will watch ISO-NE follow up with its own rule, Competitive Auctions with Sponsored Policy Resources (CASPR), which tries to help plants that might have been hurt by MOPR. The CASPR rule was cleverly designed to help some plants in some competitions but hopefully not affect other aspects of related competitions. (“We will add more regulations until we get it right.”)
The history of the federal tax credits is a zombie history. The tax credits are always supposed to die. A recent (July 2019) article in Reuters illustrates how the tax credits are revived. So far, these tax credits have always been renewed. A state can enact a renewable portfolio standard, and then the utilities in the state must buy a certain percentage of their electricity from renewable sources. A state can enact a zero-emissions credit, and then utilities in the state will be required to buy zero-emission electricity, which generally includes renewables and nuclear. The utilities will usually be mandated to pay some extra fee to the zero-emission suppliers.
There is a balance of power between the RTOs and the states. In terms of renewables, the states have the most freedom of action. The RTOs just follow along, trying to keep the lights on at what they hope will be low costs. State mandates might require too many renewables too fast: this could raise grid prices while lowering grid reliability.
In many cases, renewables don’t have to make any money by actually selling their energy to the grid. They make money by selling RECs (Renewable Energy Certificates) and by receiving production tax credits. Renewables can pay the grid to take their power (negative pricing) and still come out ahead financially.
Zero-cost energy bids will lower the grid prices for selling kWh. This is widely trumpeted by renewable advocates as “Due to renewables, prices are going down.” But there is a catch. They mean “wholesale prices on the grid kWh auctions are going down.” Prices to the consumer are going up.
THE GRID PRICE per kWh is a fallacious way of accounting for renewable costs because the grid price for kWh does not show the entire picture of what the customer pays.
Let’s look at a generator that is selling renewable kWh on the kWh auction. Say that this is a wind farm and is bidding into the auction at zero cents per kWh. Due to the auction method, the price for kWh on the grid will be lower, due to the presence of the wind farm. Yes, the clearing price at the auction will probably be lower. However, the wind farm also expects to sell RECs (Renewable Energy Certificates) as well as kWh. Some utility will have to buy those RECs to meet a renewable portfolio standard. The RECs will then be part of that other utility’s overhead, and, therefore, a ratepayer will pay for the RECs. One ratepayer is paying the grid clearing price for the wind kWh, and another ratepayer is paying for a wind REC through his distribution utility.
EVERY NOW AND AGAIN, we will see an announcement that some company is doing its part for the environment by using “100% renewable” electricity. This is supposed to make us imagine a big industrial facility surrounded by wind turbines and solar panels. Well, no. That’s not the right image. Better to think about an accountant.
Usually, RECs are sold to a distribution utility in a state that requires utilities to use a certain percentage of renewable power. With a fistful of RECs, the utility can claim to have bought the correct amount of renewables. The RECs are the renewables, from an accounting point of view. A utility in Connecticut can buy a REC from a wind farm in Maine, and the Connecticut utility can claim to be using renewable electricity. Meanwhile, almost all the electricity on the local grid comes from high-emissions power plants burning gas, and from low-emissions sources such as nuclear and large hydro plants. In most states, the nuclear and hydro plants are not allowed to make or sell RECs.
RECs don’t even have to be on the same grid as the REC buyer. A Connecticut utility buying RECs from Maine is bad enough. The power from Maine is being used in Maine, or maybe in New Hampshire. Nobody can track an electron, but it is unlikely that the power made in Maine is finding its end user hundreds of miles away in Connecticut. But at least the two states are on the same grid system.
Organizations and individuals can buy RECs from any supplier, which means that a company in Vermont can buy RECs generated in South Dakota.
Two types of entities purchase RECs: Utilities that have to meet their renewable-portfolio standards for renewable purchases. This is called the “compliance market.” These RECs tend to be more expensive, because utilities need to buy a certain amount. Businesses (mostly) that want to wave a green banner in front of their customers. This is called the “voluntary” REC market, and the RECs tend to be less expensive.
The user is not depending on intermittent renewables for its electricity. The user can claim to be “using renewable energy,” but that is about accounting, not about the energy they are actually using.
RECS ARE ONLY PART OF the costs of renewables. Estimating the costs of renewables, nationwide, is practically an impossible task. There are so many types of costs, and they vary by jurisdiction. I will list a few here. Subsidy costs—net metering: Also, net-metered solar usually means paying solar owners retail prices for wholesale power. These net-metering payments raise the retail price for everyone else.
Subsidy costs—curtailment: In some areas, wind turbines get paid if they cannot get online at the times that they are available. Only renewables get this type of payment. As wind turbines became a bigger presence in the Bonneville region, their owners complained to FERC about not being dispatched when the rivers were running high. When a wind turbine is able to run (the wind is blowing) but it is not dispatched, it loses the money it could have made in production tax credits and selling RECs. FERC agreed with the wind turbines, and now Bonneville Power has to compensate the wind farms for lost revenue if it does not dispatch them when the wind is blowing. The wind turbines must be paid when they are curtailed. This ruling is only for the Pacific Northwest: wind turbines in other areas are paid or not paid when curtailed, depending on the jurisdiction.
As much backup capacity on the system with renewables as she needed without renewables. The NBER study, a worldwide review, determined that a grid needs 1.14 MW of installed fossil capacity for each MW of intermittent renewable capacity.
Redundancy Costs… Transmission costs: The costs of transmission are steadily going up. Some fraction of the increased costs of transmission are due to the expense of connecting far-flung renewables to the grid. However, it is almost impossible to say whether this is a large or small portion of the increased costs. Renewables are rarely, if ever, cheaper than traditional generation.
A recent study at the University of Chicago took a different approach: Instead of trying to track all these separate costs, Greenstone and Nath looked at the effect that renewable portfolio standards had on consumer bills. For the states that have instituted standards, they compared the consumer prices in that state before and after implementation. They found that renewable portfolio standards raise the consumer cost of power.
The estimates indicate that, 7 years after passage of an RPS program, the required renewable share of generation is 1.8 percentage points higher and that average retail electricity prices are 1.3 cents per kWh, or 11% higher; the comparable figures for 12 years after adoption are a 4.2 percentage-point increase in renewables’ share and a price increase of 2.0 cents per kWh, or 17%.
The authors estimated that consumers in the states with renewable portfolio standards had paid a total of $125 billion more for electricity than they would have paid without the policies.
THE SUN MAY SHINE only during daylight hours, but a business may decide to buy solar RECs that add up to 100% of the power it uses and then advertise that it runs on 100% solar. Since the electricity and the REC are disconnected, you can use a solar REC at midnight, which is very misleading. Being “100% renewable” by buying RECs gives a business the best of both worlds: bragging rights and virtue signaling for its customers and reliable grid power for its operation.
Vermont buys RECs from Hydro-Québec at a low price, because nobody else wants them, and then sells its own RECs to southern New England states, where they want RECs to meet their renewable portfolio standards. This REC arbitrage cuts my personal power bill here in Vermont. Other states pay part of my costs. Why should I be against it? I’m against it because, with complex rules and state-by-state complexities, what could have been a good idea has, in my opinion, devolved into a game, based on arbitrage and heavily driven by public-relations opportunities.
When the policies get too complicated, the results of such policies become hard to predict, even by the people who wrote the policies. Looking at the whole complex system, I think the games-with-RECs are going to backfire on the renewable industry. At some point, the shell games will be exposed. The bad news isn’t even that the RECs might backfire. The bad news is that complex policies and games-with-RECs do not lead to a reliable grid.
By bidding in at a low “capacity value,” a wind turbine will receive some capacity payments, but they won’t get hit too badly with fines if the wind doesn’t blow. That is the wind turbine’s way of looking at the capacity value, perhaps. From the grid’s point of view, capacity value estimates how valuable the wind turbine is to the grid. For example, if a traditional power plant bids into a capacity auction, it generally bids at its nameplate capacity, or close to it. If a wind turbine bids in at 10% of its nameplate capacity, it is pretty much admitting that it is only 1/10 as valuable to the grid as a standard dispatchable power plant.
PJM looked at hourly load shapes as well as the wind-availability shapes. According to a PJM report in September 2018,206 the mean effective load-carrying capability for wind was 11.5% of nameplate capacity. That number is a type of estimation of wind’s capacity value. In contrast, wind capacity factors range between 22% and 45%
When a company decides to build a plant, it will bid into the capacity market. If its bid is accepted (its bid isn’t too high), it can bank on capacity payments when it comes online. As mentioned before, capacity payments are the major way that many gas-fired plants are funded, so a guarantee of such payments can encourage bank loans for building the power plant. The capacity payments also encourage profitability once the plant has been built.
An IRP basically consists of a load forecast and plans for how the load will be met. The plans include power plants built, retired, and upgraded; transmission lines built, retired, and upgraded; expected demand-side management possibilities; and so on. In other words, for the next twenty years, how are you (the utility) going to meet the demand for electricity, what will have to be built or upgraded, and how much will it cost?
33 states still require utilities to file Integrated Resource Plans with their state PUCs. In many areas, the RTO is also theoretically responsible for system planning. However, since the RTO has to be fuel neutral, and the auctions are supposed to be the “market” that the RTO is operating, the RTO actually has very little power to implement plans. It would be close to impossible for an RTO to plan and implement a resilient mix of power plants and fuels. A single-state Public Utility Board could insist on a strong and resilient IRP plan: an RTO can’t.
But the RTO auctions do provide cover for state mandates. From the point of view of a state legislator, why not feel good and vote for a law that says the state electricity will be X percent renewables by year Y? It doesn’t seem to hurt anyone, and it may get you some votes in some areas. Voting for a “high percentage of renewables” is voting for the modern version of “motherhood and apple pie.
When solar panels are added, how many more gas turbines need to be built? In the RTO areas, without any real oversight of requirements, renewable resources are overbuilt, and they lock the grid into gas-turbine backup.
Complying with state renewable standards is expensive, everywhere. When it is said that “renewables are the cheapest power on the grid”— that usually means the lowest bid in an RTO auction. In terms of renewables, states within RTOs seem to be more willing to encourage high percentages of renewables by a renewable portfolio standard, because their state PUCs do not have responsibility for the cost of power (the RTO auctions supposedly take care of that) or the reliability of power (the RTO auctions supposedly take care of that, too).
The difference between the wholesale (grid price) and the retail (consumer price) is not a matter of the utility laughing all the way to the bank. For that price differential, the utility has to pay for distribution line maintenance, storm repairs, billing, help desks, providing electricity for low-income people who fall behind on their bills, paying the transmission authority for transmission costs. The utility must also pay the balancing authority and/or the RTO for balancing the grid and running the auctions. If the utility is meeting a renewable-energy requirement, such as a renewable portfolio standard, the utility must pay for RECs bought from other utilities. All these costs add up to the differential between the cost of wholesale cost of “raw” electricity in the kWh auctions and the cost of electricity as delivered to the customer.
In general, with net metering, the customer with a solar panel gets paid the retail price for the electricity he sells, instead of the wholesale price. This is a very good deal for the homeowner. Too good a deal. In most states, not all customers are allowed to take part in net metering. There is usually a “cap,” and net metering is “opened up for new sign-ups” and then closed again. The reason is clear. If all kWhs of electricity were bought at the same price that it is sold to the customer—where is the utility going to get money for line maintenance and so forth? Sometimes utilities get permission to ask homeowners to pay a connection fee for net metering. People fight this sort of fee, tooth and nail. As you can imagine, since early net-metering customers didn’t have to pay such a fee, if a new customer has to pay it, that person may well say, “Why me? Why are you picking on me? My neighbor has no such fee.” If fees are applied across the board, the neighbor, who never had to pay such a fee in the past, can well say, “Boy, bait and switch!
If net metering were allowed for everyone, without a cap, we would have a “tragedy of the commons.” Prices would rise for everyone.
California also found itself with excessive costs due to net metering. It did not attempt to roll back the original payment plans, as Nevada attempted to do, though it did make the policy for new connections (NEM 2.0) stricter and more expensive. A customer enrolled in NEM 2.0 is also automatically enrolled in “Time of Use” pricing. This means that solar panels which face southwest will be paid better than solar panels that face southeast. Also, under NEM 2.0, net-metering customers will have to pay non-bypassable charges on any electricity delivered by the utility. These charges fund energy-efficiency programs and low-income-support programs.
DISTRIBUTED GENERATION
Most proponents will usually explain that the grid has to be modernized to be a “smart grid” before the transition can take place. The reason the grid needs to be modernized is that electricity production and consumption will be a two-way street. An electricity consumer will also be an electricity producer, and so communications between the grid and the “prosumer” (producer and consumer) will be important.
A small turbine in a non-windy area (where most people place their houses) will not make much power and will be a maintenance hassle. Few homeowners would choose to own a wind turbine, especially when compared to the relative ease of owning some solar panels.
Are there other types of distributed generation that people can install, in order to be prosumers? Dairy farmers can have methane digesters and small diesels attached to the digester. They would use the manure from their barns as fuel. So dairy farmers can be prosumers. What if I own a house and a woodlot? Maybe I could be a biomass prosumer? Not likely. A prosumer doesn’t just use biomass at home in a wood-burning stove. A prosumer interacts with the grid, supplying the grid with electricity and taking electricity from the grid. I would have to build a wood-fired boiler, raise steam, spin a turbine, attach a generator, and connect the whole thing to the grid. No, biomass electricity is not suitable for home use, and it’s not likely to turn people into prosumers.
A college campus might be a prosumer with wood-fired electricity, but an individual will not be. We could go down the list of other renewable-energy sources (waste burning, small hydro), and we will realize that some institutions may be able to host Distributed Energy Resources for the New Smart Grid. Most residential consumers will, at best, be able to have a solar panel on their house. In other words, for the individual consumer, Distributed Generation will look a great deal like Solar Net Metering. Solar installations are far more common in suburbia and rural areas than they are in cities.
I can imagine net-metering projects coming to a crashing halt. They are already restricted in most states. This would leave many prosumers in the lurch. The end of net metering would leave some prosumers with financially useless solar panels and no source of power after the sun goes down.
In general, people choose net metering because it is more flexible and cost-effective than investing in batteries.
GENERATORS
If big utilities go out of business, someone in the neighborhood may purchase a set of noisy and polluting diesel generators. The generator owner can sell the power to their neighbors, through jury-rigged and dangerous wires.
In various parts of the world where the electricity supply is not reliable, people do sell to their neighbors. This is true in some parts of the Middle East. The generator owner is a power person in the area. A recent Wired article on “Beirut’s Electricity Brokers” describes the situation in areas where electricity is not provided by big companies but rather by freelance generator owners.
Robert Bryce visited Beirut and spoke with people there. They referred to the electricity “brokers” as the “electricity Mafia.” They paid two electricity bills each month: one for about $35 to the state-owned power company, for the power they could provide, which was available about six hours a day. Then they pay around $100 a month to their local “mafia” generator. Bryce asked one man why he didn’t just buy his own generator, since he was paying his neighbor a significant amount of money. The answer was that, if he broke away from the local “mafia” generator, he might be killed. At the very least, the wire to his generator would be cut. Bryce reports how a clash between two generator-owners left two people dead and required the Lebanese army to end the violence.
YOU MIGHT CONSIDER the situation in Beirut to be a sort of microgrid. People do not rely on a central power station, and the electricity-generation owners are not huge corporations. Instead, they are your neighbors. When microgrids are introduced into areas that do not have a connection to the larger grid, they have mixed responses. Almost every medium-size actual island is basically a microgrid, and usually it runs with diesel generators.
It’s easy to say, “Baseload is an outmoded concept.” It’s not quite as easy to live in an area where solar panels don’t provide reliable lighting at night. In India, solar panels and battery backup didn’t do the job.
Around 2015, Greenpeace decided to put their money where their mouth was and provided a poor village in the Bihar region of India with solar panels and battery backup. The first morning after the system was installed, the batteries were drained overnight. One young man had hoped to study in the early morning before he had to go to work in the fields. He discovered that his lights would not go on in the early morning. “We want real electricity!
When you have enough electricity to run a washing machine, women become empowered. When you can run a washing machine, you have “real electricity.” And you are probably connected to the grid.
Net metering raises the price of electricity for those who do not have it, which is why many states have put caps on the amount of net metering allowed.
David JC MacKay’s book, Sustainable Energy—without the hot air. This classic book, published in 2009, is dedicated “to those who will not have the benefit of two billion years’ accumulated energy reserves.” MacKay does the calculations: if we captured every drop of water that fell in the English highlands and got it to run through a hydro plant, how much electricity could we make? He concludes that “if every river were dammed and every drop [of water] perfectly exploited,” Britain could make only 1.5 kWh of electricity per person per day from hydro power.
One sixty-watt light bulb, running the entire day, uses 1.5 kWh of electricity.
Not all Energy Star appliances have connected functionality, but Energy Star has set criteria for clothes washers, clothes dryers, dishwashers, lighting, refrigerator-freezers, room air conditioning, pool pumps, and thermostats. A list of appliances and Energy Star-connected functionality specifications dates can be found on a University of California-Irvine webpage.258 Connected-functionality appliances can share information with the local utility, and they can be entered into a utility’s demand-response program.
not only can the utility learn what appliances you are using and when you are using them, but if you enter those appliances into a demand-response program, the utility can turn them off.
When you go to ISO-NE headquarters, or you look at the websites of the other RTOs and the POOLs and FERC, or you go to a CLG meeting, in which lawyers and consultants are among the speakers, you can’t help but realize that all the consequences are discussed, modeled, and discussed again, from the economic and legal points of view. In any jurisdiction, there are tens of people engaged full time in such analysis. If there is a consequence to a change in a rule or tariff, it has been considered, lobbied for, lobbied against, and decided upon. Most of the people discussing the policy change are either “stakeholders” (insiders) or consultants to the “stakeholders.” In my opinion, all these people think through the consequences and try to influence them. These pressure groups have skin in the game, and they are watching and influencing the game. And yes, a lot of this happens behind closed doors. There are no sunshine laws on the grid.
In general, in the RTO areas, single-cycle gas-fired plants (not combined cycle) plus renewables will ultimately be used for everything: covering baseload, following load, covering peak load, and providing reserve. As noted above, gas is just-in-time, and renewables are not dispatchable. The RTO areas are inexorably sliding toward complete dependence on gas deliveries for the reliability of their power.
if the gas prices rise again, consumer bills would soar. Consumer bills are now loaded with much higher transmission, distribution, and renewable-certificate payments than they were in 2008. If gas prices go up, could those payments continue?
If and when natural-gas prices go up, prices will rise, rather suddenly, in the RTO areas.
Robert Bryce describes how electricity improves the life of women. Without electricity, women must pump water, heat water over a fire or on a stove, use the heated the water for washing or cooking, and constantly tend the fire or the stove. They would have a way to keep cooked food overnight in the refrigerator. A woman’s life would not be a constant round of drudgery.
In the RTO areas, the grid is being moved inexorably to a strong reliance on intermittent renewables, coupled with an equally strong reliance on just-in-time natural-gas delivery as backup. Several different scenarios could cause this system to collapse. Natural gas, the cure-all fuel for the grid and for houses, could become scarce or expensive. A long, cold windless period or a long, hot, humid period could overwhelm the natural-gas supply for electric generation. Wind generation is often hundreds of miles from the load center, adding another level of vulnerability to the transmission system. And so forth.
To some extent, the California RTO is a poster child for how not to run a grid. California is closing down zero-emission nuclear plants, setting high requirements for widespread use of renewables, depending heavily on natural gas (no surprise there) and on imported electricity. California rates are far higher than they should be for a state with significant hydro power and in-state natural-gas supplies. But the California ISO is running out of California money.
Or rather, they are running out of California grid. California has provided such extensive supports for intermittent renewables that their grid is often overloaded. In that case, the renewables are often “curtailed.” That is, they are not allowed to put power on the lines because the lines cannot accept so much power at once. Figure 19 shows the curtailment in California in the years 2018 and 2019. One solution that California tries to propose is a regional grid, an RTO that covers more of the West. New transmission lines (probably) would send California renewables all over the West, within a SuperISO made up of California and neighboring states that are now vertically integrated. As you can see in figure 19, in most months, CAISO has to curtail tens of thousands of MWh of renewable power, because there is too much power for their system in the middle of the day.
Some people might think it was foolish and wasteful to build such an oversupply in the first place, but California’s preferred remedy is to “regionalize” their RTO to the neighboring states. (A vertically integrated state system would have held a PUC hearing where citizens could have objected to the cost of the oversupply.)
States that are not in RTO areas are usually quite happy with their vertically integrated utilities, local (state-level) regulatory oversight, and low rates. They’re not in much of a rush to join RTOs.
Behind every group pushing for impossibly hard to meet renewable standards for a state or region, there’s another well-funded group eager to sell even more natural gas.