Techno-Economic Modeling of S U S TAINABILI Algal P r o cesses TwY i th Aspen Some Lessons Learne d S Eric H. Dunlop C E A. Kimi Coaldrake I N R E T R E G M Pan Pacific Technologies Pty Ltd E O Adelaide T I N C Australia S O CE e d u n lo p @ p a n p a ci fi c . c o m .a u ABO Algal Biomass Summit, San Diego, October 2014 1 Lesson Learned 1: M&EB • A fully converged mass and energy balance for the total SYSTEM is essential. • Without it, a model has limited use. • It can be started on a spreadsheet, but cannot finish there. ABO Algal Biomass Summit, San Diego, October 2014 2 Lesson Learned 2: Choice of Simulator • Process simulation is essential. • Using a full process simulator such as Aspen Plus is a good investment in money and time, but it is not enough. • A range of simulators is available (see next slide) but we found that AspenPlus and Comsol worked best for us. ABO Algal Biomass Summit, San Diego, October 2014 3 Lesson Learned 2: Choice of Simulator ABO Algal Biomass Summit, San Diego, October 2014 4 Lesson Learned 3: Reactive vs Proactive Simulation Excellent when the process has been decided. as needs a lot of information to be supplied. Great for a mature process. Unable to make predictions about an immature process. REACTOR CONFIGURATION DEFINED INPUT STREAMS DEFINED OUTPUT STREAMS PREDICTED Provide: Pond areas Oil Production Cell density at harvesting Reaction kinetics etc etc “NORMAL” ASPEN+ OPERATION GOOD FOR DEALING WITH THE PAST REACTOR CONFIGURATION PREDICTED INPUT STREAMS DEFINED Pond Areas predicted Productivities predicted Oil content of cell predicted OUTPUT STREAMS DEFINED Identifies oportunites for improvement(cid:31)(cid:31) Takes a lot of setting up Slow to converge “REVERSE” ASPEN+ OPERATION i.e. tell Aspen what you want and let it tell you how to get it. Typically solves 50,000 simulataneous equations (and their derivatives) so if what you ask for is unreasonable it will never converge. GOOD FOR DEALING WITH THE FUTURE Long convergence times can be a problem (3-4 hours) ABO Algal Biomass Summit, San Diego, October 2014 5 What actually happens when Aspen runs? • What happens? Each unit operation , flow stream, physical properties of each reactant, product and intermediate is run sequentially until convergence. Each possible ion species is solved for, including precipitation etc. • How many equations? A separate insight is available in the “equation oriented mode” where it can be seen that Aspen is solving 20,000 to 50,000 equations simultaneously. • What is the result? If the problem is “well-posed” i.e. the number of degrees of freedom is deliberately kept small, then the result is a rigorous imposition of a carefully designed mass and energy balance. ABO Algal Biomass Summit, San Diego, October 2014 6 Selected Results from Proactive Use of Aspen Selected Simulation Results Full recycle, no lipid loss Process operation 7,884 hour/year (90 % ) CO removal 1,000,000 tonne/year CO 2 2 Insolation 4.2 kWh/m2/day (Corpus Christi, TX) Evaporation 1.27 m3/hr/hc (Corpus Christi, TX) Carbon conversion to biodiesel (9 Methyl Esters) 95 wt % Biodiesel produced (as 9 Methyl Esters ) 46.4 tonne/hr 2,630,450 barrel oil equivalent /yr Total Area 20,033 hc Reaction Operation 1 Area for algae generation 484 hc Reaction Operation 2 Area for oil generation 19,549 hc Oil productivity (based on total area) 5.6 g/m2/day Effective oil content 57% wt% Genetic stability requirements -Number of generations/year (binary fision) 43.6 ABO Algal Biomass Summit, San Diego, October 2014 7 Outputs from Aspen Model The eighteen areas are: 1. Process Summary 2. Mass Balances 3. Energy Balance (1st Law) 4. Exergy (2nd Law) 5. Energy Analysis of Reactor 1 6. Energy Analysis of Reactor 2 7. Effluent/Bleed, Salinity, TDS 8. Water and Recycle 9. The Role of Evaporation 10. Control of Hazardous Precipitation of Inorganic Solids 11. Summary of Input and Output Streams 12. Photosynthetic Efficiency 13. Preliminary Check of Process Costs 14. Data Assembly for Flow Diagrams 15. Flow Diagrams : (i) Carbon (ii) Water (iii) Mass (iv) Enthalpy (1st Law of Thermodynamics) (v) Exergy (2nd Law of Thermodynamics) (vi) Exergy and Lost Work (2nd Law of Thermodynamics) 16. Establishing the Value Chain 17. The Value of Adding Extra Glycerol 18. Full Financial Analysis ABO Algal Biomass Summit, San Diego, October 2014 8 Carbon Flow (kg/hr) A-100 S-100 R-100 R-200 S-101 S-102 CO Absorber Flow Splitter Reaction Operation/ Reaction Operation/ Flow Splitter Flow Splitter 2 CO distribution Pond 1 Pond 2 Algae Purge Primary Cell Separation 2 Dissolved Air Flotation 17,979 Purge 681 Methanol 90 18,570 2,249 52,577 48,108 73,268 73,183 54,613 54,348 54,348 36,031 35,941 38,190 36,166 Product 34,598 00 00 4,469 00 23,136 00 01 02 03 03 04 05 04 06 1 1 1 2 1 1 1 1 1 1 1 1 A- S- R- R- S- S- S- R- S- S- R- S- Flue Gas In 18,667 85 265 18,317 new 2,024 Glycerol recycle S-103 R-103 S-104 S-105 R-104 S-106 Flow Splitter Cell Lysis Flow Splitter Flow Splitter Transesterification Flow Splitter Secondary Cell Separation Ball Mill Debris Clarifier Oil-Water Separation Reactor Glycerol Separation Decanter Centrifuge Carbon flowrate (kg/hr) ABO Algal Biomass Summit, San Diego, October 2014 9 Lesson Learned 4: This is an Energy Project -funded by the Department of Energy • Biofuels projects rely heavily on the science of energy aka thermodynamics. • Aspen and other data bases have been vastly expanded with new versions but it is never enough. • It is often necessary to calculate (because no-one has measured) ΔH formation and ΔG to establish heats of reaction. formation ΔH =ΔH -ΔH reaction formation. products formation. reactants) • If running Aspen forwards ( reactive) you mainly only need the thermodynamics of water. • If running Aspen in reverse ( proactive) then you need good data. ABO Algal Biomass Summit, San Diego, October 2014 10