Coal- and Gas-Fired Power Plants Have a New Best Friend: Data Centers

That law may be changed, due to no subsidies for wind and solar and batteries

However, Dominion has found a useful loophole to get around the law’s requirements — data centers.

.

.

Viriginia hosts the largest data center market in the world, and is home to at least 150 hyperscale data centers, with more being proposed.

In its recent integrated resource plan, Dominion cited projected energy demand from these data centers as a key reason to delay retiring existing power plants, including the Clover Power Station, a coal-powered peaker plant in Halifax County, a disproportionately low-income region.

.

In addition to delaying peaker retirements, Dominion has proposed building new CCGT plants, including a 1-GW peaker plant in Chesterfield, a community that already shoulders an undue environmental burden from existing natural gas- and coal-fired generation.

.

Similar stories have played out across the country as data centers become more and more ubiquitous, particularly in the Southeast.

Utilities in Virginia, Georgia, North Carolina and South Carolina have proposed building 20,000 MW of new gas power plants by 2040.

Data centers driving the projected load growth are being used to justify this buildout.

In Virginia, Georgia and South Carolina, data centers are responsible for at least 65% of projected load growth.

.

Data centers are also delaying the retirement of fossil fuel power plants nationwide, with at least 17 fossil fuel generators originally scheduled for closure now delaying retirement.

.

The gas bonanza is especially concerning because the projected demand from data centers could be significantly overblown.

.

Many prospective customers will submit speculative interconnection requests, sometimes across multiple states, inflating demand numbers and making accurate projections difficult.

.A study published last year by Lawrence Berkley National Lab highlighted these discrepancies across future demand projections.

The report’s low-end predictions estimate data centers will represent 6.7% of all U.S. energy use by 2028, while the high-end estimate is 12% — a difference of 255 terawatt-hours of energy, equivalent to the energy consumption of over 24 million households.

.

Despite the known issue of site speculation, utilities have an incentive to use exaggerated demand claims to justify building new infrastructure, such as Dominion’s massive Chesterfield peaker plant.

.

Regulated utilities like Dominion are guaranteed a 9% rate of return for building power plants and can pass the cost of building the power plants on to customers through rate increases.

In many cases, ratepayers are also footing the bill for the necessary transmission upgrades to service these large load customers.

West Virginia ratepayers, for example, will end up paying more than $440 million for two transmission lines destined for data centers in northern Virginia.

.

Keeping aging peaker plants online, which provide some of the most expensive energy on the grid (not to mention the most polluting), will also be borne by ratepayers.

Utility bills in several states are expected to increase by $40 to $50 a month due to data center-driven grid investments.

.

Utilities in some states have begun exploring ways to mitigate these costs.

As a way to buffer ratepayers from risk, Dominion proposed a new rate class for high energy users such as data centers.

The rate class requires a 14-year contract commitment to pay for requested power, even if the customer ends up using less than requested.

.

Ohio utility American Electric Power enacted a similar tariff, requiring large-scale data centers to pay for up to 85% of their projected energy demand each month, as well as requiring data centers to provide proof they are financially viable and able to meet those requirements.

.

Possibly in a bid to cut down on speculative interconnection requests, customers will also have to pay an exit fee if the project is canceled.

.

These types of regulatory measures are hindered by the nature of data centers themselves, which require 24/7 power and have little appetite for reducing demand during peak periods.

Large tech companies with existing decarbonization pledges have turned to procuring their own carbon-free power to address this.

Microsoft, for example, signed a power purchase agreement last year to restart the nuclear facility at Three Mile Island, a move which prompted backlash from residents.

.

Data center pursuit of nuclear power also prompts larger concerns about the impact on the grid.

Nuclear plants delivered almost half of the carbon-free power produced in the U.S. in 2023.

.

ADDITION

.

COAL ELECTRICITY LESS COSTLY, AVAILABLE NOW, NOT PIE IN THE SKY, LIKE EXPENSIVE FUSION AND SMAL MODULAR NUCLEAR 

https://www.windtaskforce.org/profiles/blogs/coal-electricity-less-...

By Willem Post

.

Coal gets very little direct subsidies in the US.

Here is an example of the lifetime cost of a coal plant.

The key is running steadily at 90% output for 50 years, on average

.

Assume mine-mouth coal plant in Wyoming; 1800 MW (three x 600 MW); turnkey-cost $10 b; life 50 y; CF 0.9; no direct subsidies.

Payments to bank, $5 b at 6% for 50 y; $316 m/y x 50 = $15.8 b

Payments to Owner, $5 b at 10% for 50 y; $504 m/y x 50 = $21.2 b

Lifetime production, base-loaded, 1800 x 8766 x 0.9 x 50 = 710,046,000 MWh

Ignored cost: 1) O&M escalates at 4%; 2) insurance escalates at 4%; 3) taxes; 4) periodic overhauls.

.

For lower electricity cost/kWh, borrow more money, say 70%

Traditional Nuclear has similar economics; life 60 to 80 y; CF 0.9

.

Wyoming coal, at mine-mouth $15/US ton, 8600 Btu/lb, plant efficiency 40%, Btu/ton = 2000 x 8600 = 17.2 million

Lifetime coal use = 710,046,000,000 kWh/y x (3412 Btu/kWh/0.4)/17,200,000 Btu/US ton = 353 million US ton 

Lifetime coal cost = $5.3 billion

.

Electricity cost = (15.8 + 21.2 + 5.3) x 1,000,000,000/710,046,000,000 = 6 c/kWh; this cost will be higher, because some costs were ignored.

The Owner can deduct interest on borrowed money, and can depreciate the plant over 50 y, or less, which helps him achieve his 10% return on investment; that is a general government subsidy.

For perspective, China used 2204.62/2000 x 4300 = 4740 million US ton in 2024

.

THE DYSFUNCTIONAL STATE OF MASSACHUSETTS WITH GIANT BATTERIES

https://www.windtaskforce.org/profiles/blogs/the-dysfunctional-stat...

By Willem Post

.

A recent announcement is for a statewide, 4-h battery system, installed capacity 5000 MW/20,000 MWh.

Tesla recommends not charging to more than 80% full and not discharging to less than 20% full, to achieve normal life of 15 years and normal aging at 1.5%/y.
The delivered capacity would be 20,000 MWh x 0.6, Tesla factor x aging factor x 0.9, outage factor = 10,800 MWh

The batteries would 1) absorb midday solar peaks and deliver the electricity during peak hours of late afternoon/early evening, and 2) stabilize the grid, due to varying W/S output, 24/7/365

The turnkey cost would be about $600/installed kWh, delivered as AC at battery outlet, 2024 pricing, or $600/kWh x 20 million kWh = $12.0 billion, about every 15 years.

There will be annually increasing insurance costs for risky W/S/B projects.

If 50% were borrowed from banks, the cost of amortizing $6 billion at 6% over 15 years = $608 million/y

If 50% were from Owners, the cost of amortizing $6 billion at 10% over 15 years = $774 million/y

The two items total $1,382 million/y; another hell-of-a-big subsidy for W/S systems

There are many more cost items

Less 50% subsidies (tax credits, 5-y depreciation, loan interest deduction, etc.)

Subsidies shift costs from project Owners to ratepayers, taxpayers, government debt

https://www.windtaskforce.org/profiles/blogs/battery-system-capital...

.

No banks will finance W/S/B projects at acceptable interest rates and no insurance companies will insure them at acceptable premiums, no matter what the leftist, woke bureaucrats are announcing.
The sooner the U-turn, the better for New England, the US and Europe

.

NOTE: Trump has declared a National Energy Emergency. A new gas line from Pennsylvania to New England and new gas/oil storage systems near each CCGT power plant are needed, because most of the “planned” W/S/B systems will never be built, especially after the application of tariffs.

.

BATTERY SYSTEM CAPITAL COSTS, OPERATING COSTS, ENERGY LOSSES, AND AGING
https://www.windtaskforce.org/profiles/blogs/battery-system-capital...

by Willem Post

. 

Utility-scale, battery system pricing usually is not made public, but for this system it was.

Neoen, in western Australia, has just turned on its 219 MW/ 877 MWh Tesla Megapack battery, the largest in western Australia.

Ultimately, it will be a 560 MW/2,240 MWh battery system, $1,100,000,000/2,240,000 kWh = $491/kWh, delivered as AC, late 2024 pricing. Smaller capacity systems will cost much more than $500/kWh

.

Annual Cost of Megapack Battery Systems; 2023 pricing
Assume a system rated 45.3 MW/181.9 MWh, and an all-in turnkey cost of $104.5 million, per Example 2
Amortize bank loan for 50% of $104.5 million at 6.5%/y for 15 years, $5.484 million/y
Pay Owner return of 50% of $104.5 million at 10%/y for 15 years, $6.765 million/y (10% due to high inflation)
Lifetime (Bank + Owner) payments 15 x (5.484 + 6.765) = $183.7 million
Assume battery daily usage for 15 years at 10%, and loss factor = 1/(0.9 *0.9)
Battery lifetime output = 15 y x 365 d/y x 181.9 MWh x 0.1, usage x 1000 kWh/MWh = 99,590,250 kWh to HV grid; 122,950,926 kWh from HV grid; 233,606,676 kWh loss
(Bank + Owner) payments, $183.7 million / 99,590,250 kWh = 184.5 c/kWh
Less 50% subsidies (tax credits, 5-y depreciation, loan interest deduction) is 92.3c/kWh

Subsidies shift costs from project Owners to ratepayers, taxpayers, government debt
At 10% throughput, (Bank + Owner) cost, 92.3 c/kWh
At 40% throughput, (Bank + Owner) cost, 23.1 c/kWh
. 
Excluded costs/kWh: 1) O&M; 2) system aging, 1.5%/y, 3) 20% HV grid-to-HV grid loss, 4) grid extension/reinforcement to connect battery systems, 5) downtime of parts of the system, 6) decommissioning in year 15, i.e., disassembly, reprocessing and storing at hazardous waste sites. Excluded costs would add at least 15 c/kWh
 

COMMENTS ON CALCULATION

Almost all existing battery systems operate at less than 10%, per EIA annual reports i.e., new systems would operate at about 92.4 + 15 = 107.4 c/kWh. They are used to stabilize the grid, i.e., frequency control and counteracting up/down W/S outputs. If 40% throughput, 23.1 + 15 = 38.1 c/kWh. 

A 4-h battery system costs 38.1 c/kWh of throughput, if operated at a duty factor of 40%.

That is on top of the cost/kWh of the electricity taken from the HV grid to feed the batteries

Up to 40% could occur by absorbing midday solar peaks and discharging during late-afternoon/early-evening, which occur every day in California and other sunny states. The more solar systems, the greater the peaks.

See URL for Megapacks required for a one-day wind lull in New England

40% throughput is close to Tesla’s recommendation of 60% maximum throughput, i.e., not charge above 80% and not discharge below 20%, to perform 24/7/365 service for 15 y, with normal aging.

Owners of battery systems with fires, likely charged above 80% and discharged below 20% to maximize profits.

Tesla’s recommendation was not heeded by the Owners of the Hornsdale Power Reserve in Australia. They excessively charged/discharged the system. After a few years, they added Megapacks to offset rapid aging of the original system, and added more Megapacks to increase the rating of the expanded system.

http://www.windtaskforce.org/profiles/blogs/the-hornsdale-power-res...

Regarding any project, the bank and Owner have to be paid, no matter what. I amortized the bank loan and Owner’s investment

Divide total payments over 15 years by the throughput during 15 years, you get c/kWh, as shown.

There is about a 20% round-trip loss, from HV grid to 1) step-down transformer, 2) front-end power electronics, 3) into battery, 4) out of battery, 5) back-end power electronics, 6) step-up transformer, to HV grid, i.e., you draw about 50 units from the HV grid to deliver about 40 units to the HV grid, because of A-to-Z system losses. That gets worse with aging.

A lot of people do not like these c/kWh numbers, because they have been repeatedly told by self-serving folks, battery Nirvana is just around the corner.

.

NOTE: Aerial photos of large-scale battery systems with many Megapacks, show many items of equipment, other than the Tesla supply, such as step-down/step-up transformers, switchgear, connections to the grid, land, access roads, fencing, security, site lighting, i.e., the cost of the Tesla supply is only one part of the battery system cost at a site.

.

NOTE: Battery system turnkey capital costs and electricity storage costs likely will be much higher in 2023 and future years, than in 2021 and earlier years, due to: 1) increased inflation rates, 2) increased interest rates, 3) supply chain disruptions, which delay projects and increase costs, 4) increased energy prices, such as of oil, gas, coal, electricity, etc., 5) increased materials prices, such as of tungsten, cobalt, lithium, copper, manganese, etc., 6) increased labor rates.

.

HIGH COST/kWh OF W/S SYSTEMS FOISTED ONTO A BRAINWASHED PUBLIC 

https://www.windtaskforce.org/profiles/blogs/high-cost-kwh-of-w-s-s...

By Willem Post

.

What is generally not known, the more weather-dependent W/S systems, the less efficient the traditional generators, as they inefficiently counteract the increasingly larger ups and downs of W/S output. See URL

https://www.windtaskforce.org/profiles/blogs/fuel-and-co2-reduction...

.

W/S systems add great cost to the overall delivery of electricity to users; the more W/S systems, the higher the cost/kWh, as proven by the UK and Germany, with the highest electricity rates in Europe, and near-zero, real-growth GDP.

.

At about 30% W/S, the entire system hits an increasingly thicker concrete wall, operationally and cost wise.

The UK and Germany are hitting the wall, more and more hours each day.
The cost of electricity delivered to users increased with each additional W/S/B system

.

Nuclear, gas, coal and reservoir hydro plants are the only rational way forward.

Ignore CO2, because greater CO2 ppm in atmosphere is essential for: 1) increased green flora to increase fauna all over the world, and 2) increased crop yields to better feed 8 billion people.

.
Net-zero by 2050 to-reduce CO2 is a super-expensive suicide pact, to increase command/control by governments, and enable the moneyed elites to get richer, at the expense of all others, by using the foghorn of the government-subsidized/controlled Corporate Media to spread scare-mongering slogans and brainwash people.

.

Subsidies shift costs from project Owners to ratepayers, taxpayers, government debt:

1) Federal and state tax credits, up to 50% (Community tax credit of 10 percent – Federal tax credit of 30 percent - State tax credit and other incentives of up to 10%);

2) 5-y Accelerated Depreciation write off of the entire project;

3) Loan interest deduction

.

Utilities pay 15 c/kWh, wholesale, after 50% subsidies, for electricity from fixedoffshore wind systems

Utilities pay 18 c/kWh, wholesale, after 50% subsidies, for electricity from floating offshore wind

Utilities pay 12 c/kWh, wholesale, after 50% subsidies, for electricity from larger solar systems

.

Excluded costs, at a future 30% W/S annual penetration on the grid, based on UK and German experience: 

- Onshore grid expansion/reinforcement to connect distributed W/S systems, about 2 c/kWh

- A fleet of traditional power plants to quickly counteract W/S variable output, on a less than minute-by-minute basis, 24/7/365, which leads to more Btu/kWh, more CO2/kWh, more cost of about 2 c/kWh

- A fleet of traditional power plants to provide electricity during 1) low-wind periods, 2) high-wind periods, when rotors are locked in place, and 3) low solar periods during mornings, evenings, at night, snow/ice on panels, which leads to more Btu/kWh, more CO2/kWh, more cost of about 2 c/kWh

- Pay W/S system Owners for electricity they could have produced, if not curtailed, about 1 c/kWh

- Importing electricity at high prices, when W/S output is low, 1 c/kWh

- Exporting electricity at low prices, when W/S output is high, 1 c/kWh

- Disassembly on land and at sea, reprocessing and storing at hazardous waste sites, about 2 c/kWh

Some of these values exponentially increase as more W/S systems are added to the grid
.
The economic/financial insanity and environmental damage of it all is off the charts.
No wonder Europe’s near-zero, real-growth GDP is in de-growth mode.

That economy has been tied into knots by inane people.

YOUR tax dollars are building these projects so YOU will have much higher electric bills.

Remove YOUR tax dollars using your vote, and none of these projects would be built, and YOUR electric bills would be lower.

.

NUCLEAR PLANTS TOO EXPENSIVE?

https://www.windtaskforce.org/profiles/blogs/nuclear-plants-too-exp...

By Willem Post

. 

In France, the turnkey cost of the 1,600 MW Flamanville plant was $13.7 billion, or $8,563/installed MW

Plants built by Russia, China and South Korea are about $5,500/installed MW

Expensive nuclear plant building is strictly a "rules-based" Western thing.

.

Nuclear Plants by Russia

According to the IAEA, during the first half of 2023, a total of 407 nuclear reactors are in operation at power plants across the world, with a total capacity at about 370,000 MW

Nuclear was 2546 TWh, or 9.2%, of world electricity production in 2022

https://www.windtaskforce.org/profiles/blogs/batteries-in-new-england

Rosatom, a Russian Company, is building more nuclear reactors than any other country in the world, according to data from the Power Reactor Information System of the International Atomic Energy Agency, IAEA.

The data show, a total of 58 large-scale nuclear power reactors are currently under construction worldwide, of which 23 are being built by Russia.

.

In Egypt, 4 reactors, each 1,200 MW = 4,800 MW for $28.75 billion, or about $5,990/kW, 

As per a bilateral agreement, signed in 2015, approximately 85% of it is financed by Russia, and to be paid for by Egypt under a 22-year loan with an interest rate of 3%.
That cost is at least 40% less than US/UK/EU

.

In Turkey, 4 reactors, each 1,200 MW = 4,800 MW for $20 billion, or about $4,200/kW, entirely financed by Russia. The plant will be owned and operated by Rosatom

.

In India, 6 VVER-1000 reactors, each 1,000 MW = 6,000 MW at the Kudankulam Nuclear Power Plant.

Capital cost about $15 billion. Units 1, 2, 3 and 4 are in operation, units 5 and 6 are being constructed

.

In Iran, Rosatom started site preparation for a nuclear power plant at the Bushehr site.

Phase 1: Unit 1 went on line in 2012.

Phase 2: 2 VVER-1000 units, each 1050 MW. Construction started March 2017. Units 2 and 3 to be completed in 2024 and 2026.

.

In Bangladesh: 2 VVER-1200 reactors = 2400 MW at the Rooppur Power Station

Capital cost $12.65 billion is 90% funded by a loan from the Russian government. The two units generating 2400 MW are planned to be operational in 2024 and 2025. Rosatom will operate the units for the first year before handing over to Bangladeshi operators. Russia will supply the nuclear fuel and take back and reprocess spent nuclear fuel.

https://en.wikipedia.org/wiki/Rooppur_Nuclear_Power_Plant

.

Russia is the only country with nuclear powered ice breakers.

The biggest ones steadily go through up to 7 METERS of ice.

.

Rosatom, created in 2007 by combining several Russian companies, usually provides full service during the entire project life, such as training, new fuel bundles, refueling, waste processing and waste storage in Russia, etc., because the various countries likely do not have the required systems and infrastructures

. 

Remember, these nuclear plants reliably produce steady electricity, at reasonable cost/kWh, and have near-zero CO2 emissions

In the US, they have about 0.90 capacity factors, and last 60 to 80 years

Nuclear does not need counteracting plants. They can be designed as load-following, as some are in France

.

Wind: Offshore wind systems produce variable, unreliable power, at very high cost/kWh, are far from CO2-free, on a mine-to-hazardous landfill basis.
They have lifetime capacity factors, on average, of about 0.40; about 0.45 in very windy places

They last about 15 to 20 years in a salt water environment 
They require:

1) A fleet of quick-reacting power plants to counteract the up/down wind outputs, on a less-than-minute-by-minute basis, 24/7/365,

2) Major expansion/reinforcement of electric grids to connect the wind systems to load centers,

3) A lot of land and sea area,

4) Curtailment payments, i.e., pay owners for what they could have produced

. 

Major Competitors: Rosatom’s direct competitors, according to PRIS data, are three Chinese companies: CNNC, CSPI and CGN.
They are building 22 reactors, but it should be noted, they are being built primarily inside China, and the Chinese partners are building five of them together with Rosatom.

American and European companies are lagging behind Rosatom, by a wide margin,” Alexander Uvarov, a director at the Atom-info Center and editor-in-chief at the atominfo.ru website, told TASS.

.

SMALL MODULAR REACTORS

https://www.windtaskforce.org/profiles/blogs/small-modular-reactors

By Willem Post

.

SMRs sounds good, but the electricity cost/kWh would be at least 2 times gas fired CCGT plants.

Such plants are up to 60% efficient, have very low CO2/kWh.

. 

It would take at least 5 to 8 years to build SMRs at a rate of say 50 units per year, because the US no longer has the thousands of educated and trained nuclear engineering professionals capable of designing any nuclear plants. 
The US lost that capability after Three Mile Island in March, 1979, more than 45 years ago.

.

Also, the US has not enough working-age people who 1) know how to do more complicated stuff, 2) care enough to do it, 3) have the work ethic and mental discipline, or 4) are otherwise inspired to make themselves useful.

Factories have 400,000 unfilled jobs, but there are few skilled, ambitious people to take them. 
People have weird expectations; they want to make big bucks doing nothing.

.

The US has a total lack of Science/Technology/Engineering/Mathematics (STEM) professionals who are in high places to call the shots. 

The US has been filling the shortfall with Chinese, Indian, etc., STEM folks.

The vacuum at the top was filled by lawyer/liberal arts/enviro functionaries who know next to nothing, except obstruction; Hochul, Newsom, etc., are demagogue-style examples.

.

At present, no country is set up to produce, say 50 SMRs per year, at 200 MW each.

China, Russia, South Korea, and the US, with large command/control economies, would be the only countries able set up the required A-to-Z infrastructures.

.

A 500 MW (2 units at 250 MW each) CCGT power plant can be built in two years, at a turnkey cost of $2000/kW.

New York State has finally agreed to allow the building of the gas pipeline from Pennsylvania to New England.

.

If four countries were building 50 SMRs/y each, it would require:

. 

Increased uranium mining,

Processing the uranium into fuel bundles,

Constructing factories to produce components and subassemblies,

Constructing factories for assembling the final units near harbors.

Shipping the assembled unis to the site, likely by ship or barge,

Selection and preparation of the site near harbors,

Adding the remaining balance of plant systems,

Plant test operation of each subsystem,

Connecting the plant to the grid, with switchyard,

Test operation of the entire plant,

Commissioning the plant to produce electricity at design output

. 

AI systems require lots of steady electricity 

Each major AI system should be required to have its own power plant

.

Any SMRs shipped to Africa and other such areas, would be turnkey-built in Europe, the US, Russia, Korea, China, and then shipped by special barges to Africa, etc. The SMRs would stay on the barges and send power to shore. No fuss, no muss

.

By definition, highly subsidized, very expensive, environment-uglifying, bird/bat/sea fauna/tourism/fishery/viewshed-destroying, weather-dependent, variable/intermittent, grid-disturbing, expensive-electricity-producing wind/solar/battery systems do not qualify.

https://www.windtaskforce.org/profiles/blogs/high-cost-kwh-of-w-s-s...