OXYGEN PRODUCTION for OXYFUEL POWER PLANTS Status of Development

OXYGEN PRODUCTION for OXYFUEL POWER PLANTS Status of Development Linde Engineering Dr. Dimitri Goloubev Workshop on Oxyfuel-FBC Technology, 28.06.12...
Author: Frank Dalton
0 downloads 1 Views 1MB Size
OXYGEN PRODUCTION for OXYFUEL POWER PLANTS Status of Development

Linde Engineering

Dr. Dimitri Goloubev Workshop on Oxyfuel-FBC Technology, 28.06.12

Linde Engineering

Agenda

• Oxygen Requirements for Oxyfuel Combustion Why new development for Air Separation Unit (ASU)?

• Development History and Current Status • Heat Integration between ASU and Power Plant • ASU Load Following Capability

• Discussion

Linde AG Linde Engineering Division

Linde Engineering

Oxygen Requirements for oxyfuel combustion

• Large amount (e.g. 280 t/h for Wel = 300 MW) • Low purity (< 97%) • Low pressure (nearly equal to atmospheric pressure)

• No demand for any significant quantity of other products (nitrogen, argon or liquid products)

Linde AG Linde Engineering Division

Linde Engineering

Why new development for ASU?

• Power consumption of ASU reduces the net power output  High efficiency is required! • Large scale ASU means significant CAPEX  Capital cost should be minimized!

Linde AG Linde Engineering Division

Linde Engineering

Why new development for ASU?

Composition of atmospheric air

Pure oxygen product:  99,5% O2 and 0,5% Ar (Argon has a big impact on the rectification process) Low purity oxygen product:  95% O2, 2% N2 and 3% Ar (approx.) (almost no impact, nearly the whole Argon remains in the oxygen product)

Separation of argon from oxygen is not required for "impure" oxygen product!  Saving potential is available

Linde AG Linde Engineering Division

Linde Engineering

“Classical” ASU Process with Double Column

1.2 bar

UN2 to Evap. Cooler and for Adsorber regeneration

1

MS Adsorber Turbine Air Precooling

Condenser

G POWER

5.6 bar

Main Air Compressor

POWER

AIR

Oxygen purity: 95% Linde AG Linde Engineering Division

1

Linde Engineering

“Classical” ASU Process with Double Column McCabe-Thiele Diagram for the upper (low pressure) column

Oxygen purity: 95% Linde AG Linde Engineering Division

Linde Engineering

“Classical” ASU Process with Double Column

DISADVANTAGES: •

High power consumption  Thermodynamical losses in the low pressure column lead to additional losses at the main air compressor The process is not capable to realize the saving potential to a large extent

ADVANTAGES: •

Low capital cost  Small equipment dimensions due to high (usuall) air pressure  Reduced volume of main heat exchanger as a result of excess turbine refrigeration

Linde AG Linde Engineering Division

Linde Engineering

PGAN

GOX

Single pressure Dual Reboiler ASU process

1.2 bar

LP Column

2

UN2 to Evap. Cooler and for Adsorber regeneration 1

Condenser 1 Heat Exchanger

MS Adsorber Turbine Air Precooling

G POWER

Condenser 2

2

~ 4.85 bar

Oxygen purity: 95% Linde AG Linde Engineering Division

HP Column

AIR

1

Subcooler

Main Air Compressor

POWER

Linde Engineering

LP Column

Linde AG Linde Engineering Division

HP Column

Subcooler

Oxygen purity: 95%

Nitrogen in vapour phase

Single pressure Dual Reboiler ASU process

Linde Engineering

Dual Reboiler process

Efficiency improvement of the Dual Reboiler process: •

Dual Reboiler process with the feed air stream under two different pressures (MAC and BAC)



Introduction of a third condenser into the process  Side condenser for evaporation of product oxygen only



The use of advanced condenser types  Falling film condenser  Forced flow condenser

The power consumption of the OPEX opimised Dual Reboiler process is around 13% lower compared to the "classical" Double Column process* * compared assuming identical efficiency numbers for compressors Linde AG Linde Engineering Division

Linde Engineering

PGAN

GOX

Triple Column ASU process

Low pressure column is devided into two parts Medium pressure column (MAC pressure) High pressure column (BAC pressure)

1.2 bar UN2 to Evap. Cooler GAN for Adsorber regeneration

Booster Air Compressor ~ 4.8 bar approx. 22% of the total ASU power LP Column Part 1

LP Column Part 2

HP Column 2

G

HP Column 1

Air Precooling

Heat Exchanger

MS Adsorber

~ 3.1 bar

AIR

Oxygen purity: 95% Linde AG Linde Engineering Division

Side Condenser

Subcooler

Main Air Compressor

POWER

Linde Engineering

Triple Column ASU Process



The introduction of a third column into the Dual Reboiler process with feed air stream under two pressures (MAC and BAC) lead to futher reduction of losses in the rectification part and allows the process optimisation with quite low pressure at MAC outlet



Slight advantage in power can be additionally reached with two adsorber stations operating at different pressures as well as by precooling of the air stream to the Booster Air Compressor



The power consumption of ASU with this process cycle is around 20% lower than with "classical" double column process*

* compared assuming identical efficiency numbers for compressors Linde AG Linde Engineering Division

Linde Engineering

Multi-column ASU process?

• The introduction of further condensers and rectification columns doesn't bringt any significant advantages and only increases the level of complexity

Linde AG Linde Engineering Division

Linde Engineering

Further Development of ASU process cycles for Oxyfuel Power Plants

Target • To minimise the power consumption and capital cost

The way • Process Improvement (cost and complexity reduction) • Improvements in "hardware" technology • Heat Integration with Power Plant

Linde AG Linde Engineering Division

Linde Engineering

Heat Integration between ASU and Oxyfuel Power Plant

• Thermodynamic losses in the air compressor of ASU are still very high  The isothermal compressor efficiency is around 75% (25% of consumed power is lost at compressor itself) • The heat flow from the air compressor can be recovered at the power plant to reduce this exergy loss • The heat recovering must be maximised with minimisation of the compressor power consumption

Linde AG Linde Engineering Division

Linde Engineering

Heat Integration between ASU and Oxyfuel Power Plant

• Optimisation is possible  adiabatic compression at lower pressures doesn't lead to significant power penalty due to saving the pressure loss in intercooler but allows to recover the heat at higher temperature level  The power penalty can be totally avoided with use of an axial compressor stage at lower pressures

Linde AG Linde Engineering Division

Linde Engineering

Heat Integration between ASU and Oxyfuel Power Plant LP Air Stream to ASU, ~ 3.0 – 3.2 bar

second compressor section Chilled H2O H2O

MP Air Stream to ASU, ~ 4.9 – 5.1 bar

AIR

first compressor section without intercooler

Cold feed water T ~ 300 K Warm feed water to power plant T > 400 K

Axial compressor with radial stage for MP Air Stream can be used

example picture Linde AG Linde Engineering Division

Linde Engineering

Heat Exchanger for "Heat Integration" The requirements for "Integration"- heat exhanger are quite hard:  Very large amount of Heat is to be transffered with small temeprature difference (MTD=10-15 K or even less)  Very small allowed pressure loss for the air stream (≤ 100 mbar)

Coil-Wound Heat Exchanger • Efficient cross-flow counter-current principle • Air flow as a shell side stream • Coiled tubes in layers for Feed-Water • Compactness • High mechanicall robustness • Linde Technology (LNG Heat Exchangers) example picture for a small coil-wound HEX Linde AG Linde Engineering Division

Linde Engineering

ASU process cycles for Oxyfuel Power Plants

Oxygen purity: 95%

Linde AG Linde Engineering Division

Oxygen pressure: 1.2 bar

Linde Engineering

Load following capability of ASU for Oxyfuel Power Plants

• The requirements for load following capability for Oxyfuel Power Plants are higher than for conventional applications • Tests with increased speed of load change were carried out to investigate the dynamic behavior of ASU: To identify the highest speed of load change without big fluctuations in products purity  To find out the critical controllers and process parameters which restrict the maximal speed of load change  To collect the practical experience with an adjustment of parameters needed for high speed load change • The tests were carried out at the existing Linde ASU with production capacity of 6000 Nm3/h of low pressure gaseous oxygen with a purity of 99.5% (double column process cycle with side condenser) Linde AG Linde Engineering Division

Linde Engineering

Load change with 4% per minute (MAXMINMAXMIN) Process Air flow, GOX Product flow and GOX Product purity

Linde AG Linde Engineering Division

Linde Engineering

Load change with 4% per minute (MAXMINMAXMIN) „Liquids-Management" in the plant

Linde AG Linde Engineering Division

Linde Engineering

Load following capability of ASU for Oxyfuel Power Plants

• ASU load changes between 75 and 100% with a speed of 4% per minute and up to three load changes in a row were performed without appreciablechanges in oxygen product purity. The liquid levels in condensers could be held at required set point values • The load changes with the speed of 8% per minute were successfully performed also without appreciable fluctuation of oxygen purity. The relevant parameters were taken over from the load change tests with 4% per minute without any further adjustments. The maximal deviation between the ramped set point and process value of oxygen product flow amounted to approx. 5% • The results of experiments allowed the adaption of simulation models for ASU dynamic behavior prediction and gain confidence in offering the ASU that enables working with increased speed of load change

Linde AG Linde Engineering Division

Linde Engineering

Acknowledgement

To my colleagues for their actual and former contribution… • Dr. Alexander Alekseev • Dr. Dirk Schwenk • Dr. Thomas Rathbone

Linde AG Linde Engineering Division

Many thanks for your attention

Linde Engineering

Suggest Documents