Concepts for efficient hydrogen liquefaction

David Berstad, Geir Skaugen and Øivind WilhelmsenThe Role of Large-Scale Hydrogen, Hyper closing seminarBrussels, 11-12 December 2019

Overview of the presentation

• A look at the present and beyond-present state-of-the-art for LH2 generation.

• Dissecting the loss of efficiency – where does it come from?

• How do we realize the potential of LH2generation? – the next steps

0

2

4

6

8

10

12

14

16

20% 30% 40% 50% 60% 70% 80%

Pow

er re

quire

men

t [kW

h/kg

]

Efficiency of hydrogen liquefier

Power requirement vs exergy efficiency of liquefier

State of the art (5–10 t/d blocks)

Long-term identified potential

In perspective:Full-scale LNG plants: 10 000–20 000 tLNG/dRational (exergy) efficiency: Up to 48 %

Minimum liquefaction work

• Main influence:• Feed pressure

• Key influences:• Final liquid storage pressure

• 1.3 bar (blue dotted)• 1.5 bar (black)• 1.7 bar (red dashed)

• Target para-hydrogen concentration

• Ambient temperature

Conversion of ortho to para-hydrogen

para-hydrogen ortho-hydrogen

Process overview

HX-7

Ejector

Boil-off from storage

LH2 to storage

HX-6

HX-1

HX-2

LPC-1LPC-2HPC-1HPC-2MRC-1 MRC-2

H2 feed

HX-4

HX-3

HX-5

Mixed refrigerant pre-cooling cycle

Hydrogen Claude cycle

Regenerative adsorption and adiabatic ortho-para conversion

• Assumed capacity: 125 t/d• Additionally 7.1 t/d BOG from storage

• Mixed-refrigerant "PRICO" precooling of feed hydrogen to 114 K

• Hydrogen Claude cooling to 30 K• Continuous ortho-para conversion in heat

exchangers between 114 K and 30 K • Expansion and BOG recompression to

1.85 bar with ejector• Condensing and subcooling before

transfer to final LH2 storage• Storage pressure: 1.50 bar

Overall liquefier efficiency

• Two independent ways to calculate the overall efficiency (and to verify the calculations):

• Overall, top-down. Based on exergy flows in and out of the system boundaries

• Detailed, bottom-up. Calculating the exergy destruction in each single process component

• The former method is sufficient for calculating the overall efficiency of the plant

• The latter method gives a full breakdown of all losses and gives directions as to where the improvement potentials lie

HX-7

Ejector

Boil-off from storage

LH2 to storage

HX-6

HX-1

HX-2

LPC-1LPC-2HPC-1HPC-2MRC-1 MRC-2

H2 feed

HX-4

HX-3

HX-5

Mixed refrigerant pre-cooling cycle

Hydrogen Claude cycle

Regenerative adsorption and adiabatic ortho-para conversion

Overall, top-down approach• Only considers exergy streams crossing

the system boundaries• Hydrogen feed flow exergy• BOG flow exergy• LH2 product flow exergy• Compression power• (Turbine power, assumed dissipated)• (Intercooler heat, assumed dissipated)

• Exergy efficiency:= (ELH2 prod – (EH2 feed + EBOG)) / Wcompr

= (14 361 kW / 36 729 kW)= 39.1 %

Boil-off from storage

LH2 to storage

Hydrogen feed

Compressor and pump power

Turbine power

WTURB

WCOMPR

Pre-cooling cycle to 114 K• Single, 5-component refrigerant mixture• Estimated stand-alone exergy efficiency for pre-

cooling to 114 K: 50.75 %• Further efficiency improvements can be made:

• Replacing throttling valve with a liquid expander, or expander in series with throttling valve

• Reduces entropy in expansion process and leads to savings other places in the process

• Dual-mixed refrigerant cycle:• Tailored refrigerant mixtures for two temperature

ranges instead of a single one• Allows even deeper precooling, possibly down to 80 K

HX-1

LPHPC-1HPC-2MRC-1 MRC-2

H2 feed

Mixed refrigerant pre-cooling cycle

Regenerative adsorption and adiabatic ortho-para conversion

Useful exergy output

Irreversibilities, HX-1

Irreversibilities, Compressor stage 1

Irreversibilities, Stage 1 intercoolerIrreversibilities, Throttling valve

Irreversibilities, Stage 2 intercoolerIrreversibilities, Compressor stage 2

Irreversibilities, Gas-Liquid Mixing

Irreversibilities, Pump

Compression power

0

500

1000

1500

2000

2500

3000

3500

4000

Exergy conversion Exergy input

kW

Detailed, bottom-up approach Calculating the irreversibility rate (exergy destruction rate) in each component (40+)

Cumulative

0,00

0,50

1,00

1,50

2,00

2,50

3,00

3,50

4,00

4,50

5,00

Hydrogenturbines anddissipative

brakes

Cryogenic heatexchangers

Hydrogenintercoolers

Hydrogencompressors

Hydrogenejector

MR intercoolers MRcompressors

Hydrogen/JTliquid turbineand throttling

valve

MR throttling Mixers Adiabatic ortho-para conversion

Throttling tostorage

kWh/

kg

Irreversibility [kWh/kg H2 feed]

Verification of top-down and bottom-up approaches

Exergy added to reliquefied BOG

Exergy added to hydrogen feed

Hydrogen turbines and dissipative brakes

Cryogenic heat exchangers

Hydrogen intercoolers

Hydrogen compressors

Hydrogen ejectorMR intercoolersMR compressorsHydrogen/JT liquid turbine and throttling valve

MR throttlingMixers

Adiabatic ortho-para conversion

Throttling to storage

Hydrogen compressors power

MR Compressors power

0

5000

10000

15000

20000

25000

30000

35000

40000

Exergy conversion and destruction Exergy input

kW

Improved efficiency from recovering turbine power • Efficiency of the liquefaction process without power recovery:

• Exergy efficiency: 39.1 %• 7.05 kWh/kg scaled with hydrogen feed rate (125 ton per day)• 6.67 kWh/kg scaled with hydrogen feed rate + boiloff gas reliquefaction rate

• For these numbers, 2769 kW of hydrogen turbine shaft power is assumed to be dissipated• If, e.g. 80 % of this power is recovered with electric generators the improved efficiency

figures will be:• Exergy efficiency: 41.6 %• 6.62 kWh/kg scaled with hydrogen feed rate (125 ton per day)• 6.27 kWh/kg scaled with hydrogen feed rate + boiloff gas reliquefaction rate

0

2

4

6

8

10

12

14

16

20% 30% 40% 50% 60% 70% 80%

Pow

er re

quire

men

t [kW

h/kg

]

Efficiency of hydrogen liquefier

Power requirement vs exergy efficiency of liquefier

State of the art (5–10 t/d blocks)

Long-term identified potential

Investigated process

Further improvements• High capacity shifts the cost weighting more towards OPEX, hereunder energy cost,

while the impact of CAPEX decreases. This motivates for, and necessitates, more advanced and integrated process designs

• Precooling: Dual mixed refrigerant to also extend the range from 114 down towards 80 K. Will shift power consumption over from hydrogen to MR compression

• Feed expansion from 30 K: Possibility for a liquid expander upstream of the ejector• Subcooling process: Can be two-stage instead of single-stage to improve efficiency• H2 feed pressure: Autothermal reforming as well as some electrolysers can supply

hydrogen at e.g. 30–40 bar. There are several means for taking advantage of this for improving the efficiency

• Turbocompressors and novel cryogenic refrigerant: More easily scalable, boiling at low temperatures.

For more: See D. Berstad, Ø. Wilhelmsen, K. Banasiak. Hydrogen liquefaction. Influence of ejector and expander configuration. Hyper Technical Seminar. Trondheim, 15 May 2019

15

Acknowledgements

This publication is based on results from the research project Hyper, performed under the ENERGIX programme. The authors acknowledge the following parties for financial support: Equinor, Shell, Kawasaki Heavy Industries, Linde Kryotechnik, Mitsubishi Corporation, Nel Hydrogen, Gassco and the Research Council of Norway (255107/E20).

Internationally outstanding