Saturday, August 05, 2023

Power Augmentation in combined cycle power plants

"There is nothing so costly as bargains." - Margaret Oliphant

I've been avoiding this post for quite a long time now.  The reason I've been avoiding it is that sometimes it's difficult to explain concepts to passers-by, who may not be familiar with power plant operations, engineering, economics, and basic physics concepts.

I'll probably have to edit this post several times before I'm somewhat happy with it anyway, but here goes: 

Gas turbines have a fundamental issue, and that is a significant loss of power as ambient temperature increases.  This is true of all gas turbines, whether installed in aircraft, power plants equipped with aeroderivative turbines, and heavy industrial turbine power plants.  

Below is a chart of Electrical Output Power vs Ambient Inlet Temperature of two of the more common gas turbine power plant types.


The GE combined cycle power drops off from 323 Megawatts to 260 Megawatts - a 20% loss of capacity - as the inlet air temperature rises from 50F to 104F (10C to 40C).  The reason that power falls off so much as air temperature rises is because air density decreases.  Fewer oxygen molecules per volume of air means that less fuel can be burned with it.  Because these are mass flow engines, the more mass you put through them, the more power they make - and there is less mass in hot air than in cold air. 

This loss of generating capacity at higher temperatures is unfortunate, because power demand increases greatly during periods of high temperatures.  This power demand is mostly due to society's increased reliance on air conditioning.  Air conditioning is a heavy demand placed on the electrical grid, right when gas turbine power plants are least able to generate optimum amounts of power.

Below is the average electrical demand for each season, throughout a 24 hour time span.  As you can see, the electrical demand is far higher in the hottest part of the day during summer months - coinciding with temperatures where gas turbine power plants may be reduced to just 80% of rated capacity. 

Natural gas turbine power plants provide about 40% of the generation across the US, so this temperature-related loss of generation can be significant for grid operations.  We've now reached the point where we will discuss several ways to mitigate this 20% loss of capacity.

 Turbine inlet cooling: (There is a decent Wikipedia page covering the topic here)

Among the easiest, most cost-effective, and trivially simple ways to reduce the loss of power caused by high ambient air temperatures is evaporative inlet cooling.  The gas turbine has to draw in tremendous quantities of air, so why not run that intake air through wet evaporative media and cool it down?   


 How it works in practice:  At the bottom right of the below image, a pump sits in a tank of water and pumps water up to a header, which distributes the water evenly across the top of the evaporative media (brown).  The evaporative media breaks up the water and provides a large surface area for the air and water to contact each other - increasing evaporation and air cooling.  A mist eliminator sits behind the evaporative media to prevent water ingress into the engine.  Behind the mist eliminator is the air filter.  A float at the lower right makes up water as necessary - like in a toilet tank.

Image courtesy Munters.

Evaporative cooling comes with some minor issues, but anything that can add 10 MW for the low cost of 100 gallons/hour of raw water and a 10 horsepower pump is pretty cheap added power.  The issues are the same as home evaporative cooling systems.  Eventually the leading edge of the evaporative media scales up - because that's where the dry air first contacts the mineral-laden raw water.  This scaling  causes airflow blockage, and at that point the media must be replaced.  

There are other ways to cool the inlet air of a gas turbine.  The second method is called "Inlet Fogging", and it's about what you would expect.  This technique is used in regions with consistent high humidity, where an evaporative cooling system would not be very effective.

Inlet Fogging is also known as "Wet Compression".  Fogging is also form of evaporative cooling, but it differs significantly from media-based evaporative cooling, discussed above.  It also presents a few advantages and challenges.

Fogging systems require the use of expensive high-pressure piston pumps, similar to those used for pressure washing, although these pumps are much larger.  One advantage of fogging systems is that they can over-saturate the air, causing it to become cooler even when the air humidity is extremely high.

Below:  An inlet fogging skid with two (blue) high pressure pumps and a suction-side cartridge filter (silver canister) to remove any foreign particulates from the water.

Below:  The fogging system at my current power plant.  As you can see, there are a variety of pump/motor combinations available to use, depending on humidity and ambient temperature.



Due to the quantity of water being sprayed, and the size of the water droplets, the fogging grid and nozzles have to be installed downstream of the inlet filter.  This much water would soak an inlet filter, making it impossible for air to pass through.  Because the fogger grid is downstream from the filter, there is potential for "FOD".  "FOD" is the acronym for Foreign Object Damage.  It's quite possible for one of the fogging nozzles to come loose and be ingested into the gas turbine - potentially ruining it.  For this reason, all parts of the fogging system are lock-wired in place.

Even if a nozzle simply wears out or breaks, it can still squirt a solid stream of water into the airstream.  Over time, larger droplets will erode the rotating blades of the compressor's first stage, potentially causing them to experience buffeting and eventually release during operation.

Another issue with inlet fogging is that 100% of the water being injected into the turbine's inlet air stream will be evaporated.  This means that the inlet fogging water must be very high purity demineralized water.  Any trace contaminants in the water will be deposited within the gas turbine compressor, once the water has evaporated during the compression process.  

Below:  Salt deposits on compressor blades


The make-up needs of the demineralized water tank for the fogging system are considerable.  The plant's water treatment system must be sized appropriately - not just for boiler water make-up, but also for spraying large quantities of pure demineralized water into the gas turbine inlet.

There is a "poor man's fogging system" as well.  I've spoken with operators who used the compressor on-line water wash system - designed to be used for 10 minutes once per day - continuously "washing" the compressor during the hottest part of the day, to increase power slightly.

Seldom used - due to expense of installation and operating cost - is inlet air refrigeration.  Essentially there is an industrial-scale Air Conditioning unit on the turbine inlet.  This is only feasible on the smallest gas turbine installations, due to the large quantities of air that need to be cooled - if possible down near freezing.  

I am familiar with only one turbine equipped with inlet air refrigeration cooling.  This is a small 15 Megawatt turbine, in a facility that already used anhydrous ammonia as a refrigerant for several other processes.  This made hot-rodding the gas turbine using a refrigeration cycle an easy choice.

The final choice for turbine inlet cooling is thermal storage - using chilled water that you cool down and store in a tank during off-peak demand hours.  During off-peak hours, part of the turbine's output will power a refrigeration unit.  The refrigeration unit will chill a large volume of water, which is stored in an insulated tank.  When the peak electrical demand arrives, the chiller is shut off, and the previously chilled water is then circulated through cooling coils in the turbine inlet, allowing the turbine to make greater power than it would without inlet cooling.

Supplemental duct firing:

Many gas turbines are installed in a combined-cycle configuration.  This means that the waste heat from the gas turbine exhaust is used to generate steam which drives a steam turbine, to make extra power from heat that would have otherwise gone out the stack.

Below:  Simplified diagram of a combined cycle power plant.  The gas turbine (6) at the bottom is the one that suffers from loss of power on hot days.  This also reduces steam production in the boiler (5).

The most efficient use of burning natural gas is to burn it in the gas turbine, then extract the waste heat.  However, it's possible to burn additional gas downstream of the gas turbine to increase steam production.  While this is less efficient, it will increase the output of the Steam Turbine (2)

We have now reached the topic of duct firing (red arrow and flame).

Adding extra heat to the gas turbine exhaust will cause the boiler to generate more steam, allowing the steam turbine to produce more power.  If the economics of electricity pricing justify the loss of efficiency, duct firing can (and is) used to increase overall power output.

Duct firing is accomplished by installing long natural gas lances with diffusers horizontally across the boiler in several tiers.  This spreads the heat out before it reaches the boiler tubes.

Below: A short duct burner.  The jagged sheet metal is the diffuser


Below:  Duct burner banks being installed in a brand-new smallish size boiler.

Below:  Side view of a duct burner in service, viewed through the inspection port.

There are a few issues that arise while operating a duct firing system.  Emissions increase.  As previously mentioned, efficiency decreases - you burn a lot more gas for a little bit more power - all of which comes from the steam turbine.  The additional steam flow and pressure also cause stress in the steam turbine.

The steam turbine will be designed with duct firing in mind, but it is usually not selected for the ability to deal with those sort of flows and pressures over the long-term, as that would add expense.  Over time, duct firing gradually damages the steam turbine.  Typically the steam turbine damage is "dishing" of the diaphragms due to high differential pressure.

The diaphragms are semi-circular arrays of stationary blades that are housed in the upper and lower turbine shells.  Diaphragms are held in place around the perimeter, but they are unsupported at the center. The added steam pressure and flow generated by duct firing causes axial bowing at the inner edge of these assemblies.

 Below:  A steam turbine shell with installed diaphragms.  The centers of these flex towards the viewer during duct firing operations.  Over time, the centers "dish" and are not axially aligned with the perimeter.

Below:  A steam turbine diaphragm removed for refurbishing

In addition to dishing of the diaphragms, the turbine shell will also begin to develop cracks where the diaphragms are held in place.  This is due to the greater force being applied by the additional steam on the diaphragm, and the small surface area that has to retain the diaphragm in place.

All of these factors are taken into account before the power plant is built.  The added maintenance, emission control (ammonia), and fuel expenses associated with duct firing are calculated.  At some electricity price level the threshold is reached where the cost of operating duct burners exceeds the increased expenses of using them - and duct burners are placed in service.

Steam Power Augmentation:

Steam Power Augmentation is the injection of steam into the gas turbine combustors.  This process is incredibly inefficient, but it allows the plant to make significantly more power on a very hot day than it would be able to otherwise.

Below:  The red line indicates steam from the boiler (HRSG) going into the turbine combustion chamber.

This form of power augmentation is the least efficient, because to do this, the duct burners have to already be in service.  The process requires about 150 gallons/minute (570 liters/minute) of steam, which all gets blown out the stack into the wind.  It takes a massive demineralized water treatment system to pull this off, in order to make up the lost demin water during off-peak hours. 

The gas turbine will make quite a bit of extra power - about 25 MW.  That's because that steam is cool compared to combustion heat.  The steam will expand a bit more, and there will be additional mass flowing through the gas turbine.  However :(  This additional mass flow has a maintenance wear factor of 1.3.  That means that for every 3 hours of steam injection, tack on a 4th hour towards the next maintenance outage.  

With full steam power augmentation in service, the plant will be throwing away 150 gallons per minute of demin water (per turbine), adding wear and tear to the gas turbine, burning excess gas with the duct burners, and adding wear and tear to the steam turbine.  However inefficient, output will be boosted by about 15%.  If electricity prices are high enough, the added inefficiency and maintenance costs will pay for themselves many times over.  A merchant power plant can run at a loss 11 months of the year, and recover to profitability in 2-3 weeks of extreme energy pricing.  This is the energy market when the cost of installing and operating power augmentation pays off.

Most gas turbine power plants will use one or more of these techniques to offset the loss of power caused by high ambient temperatures.  Plants in dry regions will use evaporative cooling continuously, unless there is danger of ice forming in the winter months.

I'd meant to do this a while back, but just now got motivated enough to think it through and post it.  It'll surely have typos and additions at some point, but at least it's finally out there for anyone with an interest.





No comments:

Post a Comment