In 1993, Japan broke through  with the first commercial-scale  

Ultra-Supercritical steam turbines.

For thirty years, turbines operated  at mere supercritical temperatures.

Limited by the properties of the  steel they were produced from.

It took nearly two decades of  R&D for Japan to develop the  

technologies to bring that steel to the market.

In today's video, we explore  a coal-centric technology. The  

30-year march from supercritical to  ultra-supercritical steam turbines.

## Beginnings

Power-generating steam turbines are a  technology well over a hundred years old.

Inside a thermal power plant, an energy  center heats up water in a boiler. This  

energy center can be coal, oil, nuclear,  or geothermal. Most of the time it is coal.

The heated water turns into steam. That steam then  hits the turbine blades and violently expands,  

producing mechanical energy that spins  a generator and creates electricity.

The steam - now cooler and under less  pressure - is then returned to the  

boiler where it is condensed back into a  liquid state. This is the Rankine Cycle.

Thermal plants are not as efficient  as a hydroelectric plant - which can  

get to as high as 90% as compared to  a thermal plant's 30-60%. But thermal  

plants are cheaper, smaller,  and less location-specific.

## Getting More Efficient We define a steam turbine's efficiency as how much

of the fuel's heating input is  turned into usable electricity.

Like all heat engines, steam  turbines have a maximum efficiency:  

the Carnot heat engine efficiency,  

the difference between the energy levels of  the steam entering and leaving the turbine.

The steam’s energy level leaving  the turbine tends to be fixed,  

as it is tied to the turbine's environment.  So the most practical thing to do is raise the  

other side of the equation: The energy  of the steam going into the turbine.

So over the past 70 years, that has been the  general summary of the steam turbine's technical  

evolution: Getting the steam hotter and putting it  under higher pressure to make it more energetic.

There was an interesting efficiency  side-quest involving reheat cycles.

After the steam expands in  a high-pressure turbine,  

it is sent back to the boiler for reheating.

The reheated but lower-pressure steam  then hits a second intermediate turbine,  

and maybe even a third lower-pressure turbine.

Only after that is the steam  finally condensed back to liquid.

Reheat cycles both add thermal efficiency  and protect the turbine blades by reducing  

the steam's moisture content. Though they do add  considerable complexity to the turbine design.

Anyway, the more efficient the turbine is,  

the less coal or oil we need to burn to get  the same number of megawatts. Which is a big  

deal because fuel is the dominant  factor in the cost of electricity.

One additional percentage point of  efficiency can save millions of tons  

of coal from burning each year - reducing  carbon dioxide emission by about 2-3%.

## The Big Beautiful Boil

A major efficiency problem that  soon emerged however was boiling.

Imagine a pot of water on the stove at sea level.  You add heat to the pot until the water reaches  

its boiling temperature of 100 degrees Celsius.  And right on cue, the water starts to boil.

Now suddenly you find that adding more heat  no longer increases the temperature. More heat  

energy only produces more bubbles.  It does not give you hotter steam,  

which is what you need to further  raise the turbine's efficiency.

But if the water's pressure  and heat go beyond a certain  

critical point - 22.1 mega-pascals and 374  degrees Celsius - something weird happens.

It becomes a dense, fog-like thing that  we call supercritical fluid - bestowed  

with the properties of both gas and liquids. It  effectively becomes steam without boiling. We  

can now raise input steam temperatures with more  heat energy - achieving ever better efficiencies.

Going supercritical also made these boilers  somewhat safer. Older, subcritical boilers  

had something called a steam drum. After water is  boiled into steam, the drum separates the steam.

Going supercritical means we no longer  need the steam drum. Removing it not only  

reduces the system's weight but also improves  safety because the drums being full of hot,  

highly pressurized fluid posed explosion  risks. As many a train has discovered.

We call such designs "once-through" boilers -  first patented by the engineer Mark Benson in  

the 1920s - because water passes  through the heating system only  

once. It goes into the boiler cold  and leaves as supercritical steam.

And yes, we probably shouldn’t call  supercritical boilers “boilers” anymore,  

since they are not actually boiling.  But inertia is a powerful thing.

## Philo Unit 6

In the 1950s, the American Electric Power  company joined with turbine-maker General  

Electric and boiler-maker Babcock  & Wilcox Company to build the first  

commercial supercritical power  generation unit: Philo Unit 6.

With a max capacity of 125 megawatts,  Unit 6's feedwater was pressurized at  

37.9 mega-pascals. Prior to this, no steam  power plant had breached 22.1 mega-pascals.

The steam's main operating temperature reached  621 degrees Celsius - 28 degrees higher than  

what had been previously possible. At these  temperatures, the steel pipes literally glow red.

The unit's thermal efficiencies touched 40% - also  a clear cut above anything else then available.

After Philo Unit 6, a second supercritical  power generation unit fired up in 1961:  

Eddystone Unit 1. Built by the Philadelphia  Electric Company, Eddystone's steam reached  

operating temperatures of 649 degrees  Celsius and pressures of 34 mega-pascals.

The Soviets also got into the fun, launching a  prototype turbine called the SKR-100 in 1968.  

It achieved steam conditions of 30  mega-pascals and 650 degrees Celsius.

But Unit 6 led the way. A manager  for mechanical engineering at the  

successor company AEP would later say about it:

> For its day, Philo 6 was like flying to  the moon without taking the intermediate  

steps of first orbiting the Earth and  then sending up an unmanned space ship

But perhaps as the metaphor implies,  it soon became clear that Unit 6 and  

its early peers had taken a step  too far. Such intense heat and  

pressures were not sustainable. The  reason had to do with the steels.

## Going Too Far

Steam turbines share a few  common steel components.

The casings and shells are big pieces of  steel that offer structural support and  

hold in steam. Being so big, their  steels cannot be too expensive.

Bolts hold things together. These  have to be highly resistant to  

stress and might find themselves  exposed to very high temperatures.

The blades spin around. They experience the steam  gas directly, but are thin and cooled by the flow.

So it is the turbine rotor - the part that  spins the blades - that experiences the hottest  

temperatures. The rotor is thick and solid  and receives heat conducted into it from the  

blades. It is one of the biggest challenge  areas from a metallurgical perspective.

Turbines and boilers are made from a limited  range of steels to ensure good match of thermal  

properties. Supercritical units in the  day used largely two types of steel.

For most components, they used 2.25Cr-1Mo steel,  

or T22 steel. The name refers to its  components of chromium and molybdenum.

The high chromium content makes T22 part of  a class of steels known as "ferritic steels".  

They are called that because they have a  body-centered cubic crystalline structure.

T22 is a good steel that welds easily  and offers good creep strength. Creep,  

meaning a tendency for the steel to deform after  long periods of high temperatures and strong  

mechanical forces. Creep is a very serious  problem for both steel and people alike.

But above 560 degrees Celsius, T22 starts  to lose that strength. So the hottest parts  

of the turbine like the rotors were made  from what are called Austenitic steels.

These steels have a face-centered  cubic crystalline structure. 

Austenitic steels contain high amounts of nickel  and chromium and are very temperature-resistant.

They are a bit expensive, thanks  to that high nickel content but  

there were two other more serious problems.

First, these Austenitic steels expand  a lot in high heat while also poorly  

transferring that heat. So when the  turbines start up and shut down,  

their thick-walled components will have  hotter outsides but cooler insides. The  

hotter areas expand more than the cooler  inner areas thus leading to cracking.

Second was oxidation. Superheated steam  can oxidize the steel, creating oxides  

on its surface that eventually flake off. These  flakes then either build up in the boiler tubes,  

blocking their flows, or chip away at  the turbine's insides - breaking them.

## The Steel Stall These were complicated problems.

Therefore, subsequent turbines ramped down to 23.8  mega-pascals and main steam temperatures between  

541 and 566 degrees Celsius. This neighborhood  is generally classified as "supercritical".

The old neighborhood - anything  over 25 mega-pascals and 593  

degrees Celsius - formed a new category called  "Ultra-supercritical". Though I must admit that  

the borders between the categories are quite  fuzzy. Various sources have their own numbers.

Great work was needed to get to this category.  But with coal being so cheap, there was little  

economic incentive in the US for this efficiency  gain. The vast majority of America’s thermal  

plants remained subcritical. By the 1960s  and 1970s, just 15% were supercritical-class.

So for twenty or so years, the operating  temperatures and pressures of the world’s  

top coal-fired power plants  remained steady. Instead,  

American utilities focused on scaling turbine  size and capacity to 600-1000 megawatts.

Even this hit limits at the end of the decade.  The insane physics and tolerances on metals  

inside such a huge turbine caused unexpected  maintenance and extended downtime costs.

## Japan In the 1950s and 1960s, Japanese companies

gained technical proficiency through  technology transfers with the West.

By the 1970s, Japanese oil-fired  thermal plants operated steam of 16.6  

mega-pascals and 566 degrees Celsius.  Short of supercritical conditions,  

but good enough for 35% thermal efficiency - on  par with similar plants in the US and Europe.

Then came the Oil Crises of the  1970s. High oil costs - plus an  

international ban on new oil-fired  thermal plants - forced a transition  

to a diverse energy portfolio of  imported coal, LNG, and nuclear.

Thusly the government funded a rapid transition  to supercritical-class turbines. The first of  

which were two 500-megawatt units  in the Matsushima Power Plant.

The Japanese government then  embarked on a R&D project to make  

Ultra-supercritical power generation a  reality - seeing them as enabling coal  

diversification while also meeting  internal carbon emissions goals.

## Going Ultra-Supercritical

After feasibility studies, the research program  at the Wakamatsu institute began in 1982.

As I said, the ultimate issue was if the turbine  steels can sustain the high thermal and mechanical  

stresses over long periods of time. It would take  over ten years to develop the metals for this.

Phase 1 of the project spanned until  1994 and was split into two steps,  

which I shall call 1A and 1B.

Phase 1A studied older Ferritic  steels to achieve conditions of 31.4  

mega-pascals and 595 degrees Celsius.

Meanwhile Phase 1B looked at Austenitic  steels with the hope of achieving 34.3  

mega-pascals and 650 degrees Celsius  (with 595 degrees in the intermediate and  

lower-pressure reheat cycles). There  was great hope in this, initially.

But in the end, the Austenitic  steels in phase 1B failed to work.  

Though blends were found that can sustain  those temperatures, tests concluded that  

the thermal expansion coefficients still  caused them to eventually warp and break.

After building test turbines at Wakamatsu,  

Mitsubishi Heavy Industries leveraged the Phase  1A learnings to build the first commercial-scale  

Ultra-supercritical turbine in the  world: The 700 mega-watt Hekinan Unit 3.

Unit 3 operated at temperatures of 593 degrees  Celsius. It first fired up in April 1993,  

marking a long-awaited return to  Ultra-Supercritical conditions.

## Some Special Steel

Meanwhile, in Phase 2, the Japanese program - a  joint venture between Wakamatsu, Mitsubishi and  

the Japanese steel company Kobelco - discovered  a line of "Advanced 12Cr" Ferritic steels.

There are four of these steels, but the most  well-covered one is Mitsubishi's TMK1 steel. I  

am in awe of this steel. I don't know how many  people care, but this is some special steel.

TMK1 descends from a 12% Chromium steel  originally made in England by William  

Jessop & Sons for jet engine discs. Called,  MEL-TROL H46. Produced with a proprietary blend  

of 9 element additives from carbon to boron to  vanadium to tungsten, H46 had good creep strength.

General Electric took this steel  and in 1965 reduced the proportion  

of Niobium. This produced a line of 12%  Chromium steels sometimes used for rotors  

during the supercritical 566 degree  Celsius era. They called it 12CrMoVNb.

To create TMK1, the Japanese  simply had to do one thing:  

Tune the amount of molybdenum in relation  to how much tungsten was in the steel.

This was done thanks to experiments on H46  done by a Professor Fujita in the 1970s.  

Fujita discovered that raising  molybdenum content from 1% to 1.5%  

helped hold together the steel's internal  structure and keep it from creeping.

Too much molybdenum however would cause the  steel to create delta-ferrite structures that  

undermine that long-term creep strength.  So 1.5% precisely. No more, no less.

Fujita's steel, called TAF, was not suitable  for large items like rotors. So Mitsubishi and  

Kobelco worked together to scale up the  methods to produce larger steel items.

Producing this steel requires some insane  skills. The raw steel mix is first melted  

using electricity in a vacuum, which  removes hydrogen and nitrogen gas  

impurities. The output is then cast  into a solid intermediate product.

This metal is then carefully re-melted by  turning it into an electrode and passing  

AC electric current through it. A method  known as Electroslag remelting. The metal  

melts drop-by-drop so it can be cast into  the cleanest, most uniform ingot possible.

The massive ingot is then forged into shape  while it is still hot. Then finally the metal  

is heated, cooled, and reheated four times at  temperatures of 700 to 1,100 degrees Celsius.

These heat treatments are to create and  lock in the steel crystal microstructures  

necessary for the steel to survive  insanely high temperatures, pressures,  

and mechanical forces without  fail for over 100,000 hours.

Mitsubishi Heavy Industries used  TMK1 for the rotors of the Matsuura  

Thermal Power Plant Unit 2. It began  operations in 1997. At 1,000 megawatts,  

it was the first large-scale  ultra-supercritical steam plant.

With operating steam pressures of  24.1 mega-pascals and temperatures  

of 593 degrees Celsius, its thermal  efficiency of about 42% broke new  

ground. Recall that supercritical  plants had about 34-35% efficiency.

Matsuura's success kicked off a  new turbine boom in Japan. Better  

steel recipes came out with improved heat and  creep resistance. And by 2001, there were 13  

Ultra-supercritical power generating  units online in Japan. By 2013, 25 units.

This includes some of the most efficient  coal-fired power plants in the world. A notable  

one being the Isogo Thermal Power Plant Unit  2, an 600 megawatt ultra-supercritical turbine.

Produced by Hitachi, the turbine operates at a  

wand-erful 25 mega-pascals and 600 degrees  Celsius. Notably, it adds one reheat cycle  

raising it to 620 degrees. The thermal  efficiency is a record-breaking 45%.

## Advanced Ultra-Supercritical It is worth nothing that the European community

has also spent efforts to develop  Ultra-supercritical thermal plants.

The core of the European effort  was a program called AD700.

Begun in 1998, it sought to produce a  demonstration plant with steam operating  

at 34 mega-pascals and the neighborhood of  700 degrees Celsius. Maybe even 750 degrees.

Which is basically basaltic lava flow  range. Such a system would have thermal  

efficiencies of 50%. They dubbed this  category: "Advanced Ultra-supercritical"

The limits of ferritic steels and  others are around 620 degrees Celsius,  

so the focus has shifted to  nickel-based superalloys. Their  

production requires even more demanding  melting processes and heat treatments.

However, by the 2000s the European Community had  started a shift towards renewables. The AD700  

demonstration plant was eventually built in 2005  - COMTES700 in Germany, which operated for about  

four years. But continued work on 50% efficient  coal-fired thermal plants has been de-emphasized.

These technologies have since moved over to  Asia. Japan remains a technology pioneer but  

they are increasingly adopting LNG. So  these thermal plants are most thriving  

in India and the People's Republic of China  - countries that still heavily rely on coal.

## Conclusion

Coal-fired thermal plants occupy  a funny place in the portfolio.

The steam turbine is the OG turbine, but  there are others out there. A notable one  

is the gas turbine, which directly takes  in natural gas or other refined fuels  

to produce electricity. Lots  of overlap with jet engines.

There are even versions where the heat in a  gas turbine's exhaust is captured to power  

a steam turbine a la Human Centipede:  A Combined Cycle Gas Turbine or CCGT.

Measured on point-to-point efficiency, CCGTs  beat standalone steam turbines at about 55-60%  

for the former compared to 42-45% for the  latter. And since they cook with gas directly,  

you skip a big boiler - though  you still need to produce steam.

But nothing quite beats coal's versatility,  

storability, and low sticker price.  Yeah CCGTs are more efficient,  

but dead cheap and plentiful coal beats more  expensive LNG - carbon pricing not included.

Moreover, these steam turbines are huge. Gas  turbines generate anywhere between 100 and 400  

megawatts of power. Modern steam turbines  can get to a staggering 1,500 megawatts,  

operating for weeks or months on end to  provide steady baseload power to the grid.

Despite the rise in world renewables capacity,  coal remains a dominant power source - and  

I struggle to see a path away from it  entirely despite the carbon footprint.  

So producing more electricity from less  coal should be a key goal in the future.