NTNU T. Gundersen Slide no. 1 Dept. of Chemical and Biomolecular Engineering National University of Singapore Seminar 10 March 2008 “A new Process Synthesis Methodology utilizing Pressure based Exergy in Subambient Processes” by Truls Gundersen Department of Energy and Process Engineering Norwegian University of Science and Technology Trondheim, Norway

NTNU T. Gundersen Slide no. 2 People: NTNU4.300 SINTEF2.000 Students Budgets: - NTNU3,5 bill NOK - SINTEF1,6 bill NOK Trondheim in Summer Time NTNU/SINTEF is the Norwegian Center of Gravity for Science & Technology and Research & Development

NTNU T. Gundersen Slide no. 3 Trondheim in Winter Time We sometimes get a lot of Snow....

NTNU T. Gundersen Slide no. 4 Trondheim in Winter Time Then we need proper Equipment....

NTNU T. Gundersen Slide no. 5 Trondheim in Winter Time But Snow is not all that bad....

NTNU T. Gundersen Slide no. 6 Norway - an Energy Nation ……. 3 Generations of Energy Development: Hydro Power, Petroleum, Renewables

NTNU T. Gundersen Slide no. 7 Brief Outline Motivation and Background Limitations of existing Methodologies Subambient Process Design The ”ExPAnD” Methodology How to play with Pressure? Attainable Region for Composite Curve Contributions from individual Streams Small Example to illustrate the Procedure Industrial Example to demonstrate the Power Concluding Remarks

NTNU T. Gundersen Slide no. 8 Motivation and Background Stream Pressure is an important Design Variable in above Ambient Heat Recovery Systems  Pressure Levels in Distillation & Evaporation affect the Temperature of important (large Duties) Heat Sinks & Sources Pressure is even more important below Ambient  Phase changes link Temperature to Pressure  Boiling & Condensation  Pressure changes link Temperature to Power  Expansion & Compression Why do we ”go” Subambient?  To liquefy volatile Components (LNG, LH 2, LCO 2 )  To separate Mixtures of volatile Components (Air) Subambient Cooling is provided by Compression  Yet another important Link to Pressure

NTNU T. Gundersen Slide no. 9 The Onion Diagram revisited R SHU The “traditional” Onion Smith and Linnhoff, 1988 R S C&EC&E H The “forgotten” Onion The User Guide, 1982 R SH The “subambient” Onion Aspelund et al., 2006 U C&EC&E

NTNU T. Gundersen Slide no. 10 Limitations of Existing Methodologies Pinch Analysis is heavily used in Industry  Only Temperature is used as a Quality Parameter  Exergy Considerations are made through the Carnot Factor  Pressure and Composition are not Considered Exergy Analysis and 2 nd Law of Thermodynamics  Considers Pressure, Composition and Temperature  Focus on Equipment Units not Flowsheet (Systems) Level  No strong Link between Exergy Losses and Cost  Often a Conflict between Exergy and Economy ExPAnD Methodology is under Development  ”Extended Pinch Analysis and Design”  Combines Pinch Analysis, Exergy Analysis and (soon) Optimization (Math Programming and/or Stochastic Opt.)

NTNU T. Gundersen Slide no. 11 The ExPAnD Methodology Currently focusing on Subambient Processes A new Problem Definition has been introduced:  ”Given a Set of Process Streams with a Supply and Target State (Temperature, Pressure and the resulting Phase), as well as Utilities for Heating and Cooling  Design a System of Heat Exchangers, Expanders and Compressors in such a way that the Irreversibilities (or later: TAC) are minimized” Limitations of the Methodology (at present)  Relies Heavily on a Set of (10) Heuristics, 6 different Criteria (Guidelines) and suffers from a rather qualitative approach  Strong need for Graphical and/or Numerical Tools to replace/assist Heuristic Rules and Design Procedures  Using the Concept of Attainable Region is a small Contribution towards a more quantitative ExPAnD Methodology A. Aspelund, D.O. Berstad and T. Gundersen, ”An Extended Pinch Analysis and Design Procedure utilizing Pressure based Exergy for Subambient Cooling”, accepted for Applied Thermal Engineering, April 2007.

NTNU T. Gundersen Slide no. 12 Classification of Exergy Thermomechanical Exergy can be decomposed into Temperature based and Pressure based Exergy e (tm) = (h – h o ) – T o (s – s 0 ) = e (T) + e (p)

NTNU T. Gundersen Slide no. 13 Exergy Balance in (ideal) Expansion Exp ambient Exp

NTNU T. Gundersen Slide no. 14 Temperature/Enthalpy (TQ) ”Route” from Supply to Target State is not fixed Supply State Target State The Route/Path from Supply to Target State is formed by Expansion & Heating as well as Compression & Cooling a)Hot Streams may temporarily act as Cold Streams and vice versa b)A (Cold) Process Stream may temporarily act as a Utility Stream c)The Target State is often a Soft Specification (both T and P) d)The Phase of a Stream can be changed by manipulating Pressure The Problem is vastly more complex than traditional HENS

NTNU T. Gundersen Slide no. 15 General Process Synthesis revisited Glasser, Hildebrand, Crowe (1987) Attainable Region Applied to identify all possible chemical compositions one can get from a given feed composition in a network of CSTR and PFR reactors as well as mixers Hauan & Lien (1998) Phenomena Vectors Applied to design reactive distillation systems by using composition vectors for the participating phenomena reaction, separation & mixing We would like to “ride” on a “Pressure Vector” in an Attainable Composite Curve Region for Design of Subambient Processes

NTNU T. Gundersen Slide no. 16 How can we Play with Pressure? Given a ”Cold” Stream with T s = - 120ºC, T t = 0ºC, p s = 5 bar, p t = 1 bar Basic PA and the 2 ”extreme” Cases are given below: Heating only Heating before Expansion before Heating ºC ºC

NTNU T. Gundersen Slide no. 17 How can we Play with Pressure? Given a ”Cold” Stream with T s = - 120ºC, T t = 0ºC, p s = 5 bar, p s = 1 bar Preheating before Expansion increases  (mCp):

NTNU T. Gundersen Slide no. 18 How can we Play with Pressure? Given a ”Cold” Stream with T s = - 120ºC, T t = 0ºC, p s = 5 bar, p s = 1 bar Heating beyond Target Temperature before Expansion:

NTNU T. Gundersen Slide no. 19 How can we Play with Pressure? Given a ”Cold” Stream with T s = - 120ºC, T t = 0ºC, p s = 5 bar, p s = 1 bar Attainable Region with One Expander:

NTNU T. Gundersen Slide no. 20 How can we Play with Pressure? Given a ”Cold” Stream with T s = - 120ºC, T t = 0ºC, p s = 5 bar, p s = 1 bar Attainable Region with Two Expanders:

NTNU T. Gundersen Slide no. 21 Attainable Region for infinite # Expanders

NTNU T. Gundersen Slide no. 22 The simplest possible Example H1: T s = -10  C T t = -85  C mCp = 3 kW/K Q H1 = 225 kW P s = 1 bar P t = 1 bar C1: T s = -55  C T t = 10  C mCp = 2 kW/K Q C1 = 130 kW P s = 4 bar P t = 1 bar CC GrCC Insufficient Cooling Duty at insufficient (too high) Temperature, but we have cold Exergy stored as Pressure Exergy !!

NTNU T. Gundersen Slide no. 23 Targeting by Exergy Analysis (EA) EA with simplified Formulas and assuming Ideal Gas (k = 1.4) gives: H1:  E X T = 65 kW  E X P = 0 kW  E X tm = 65 kW Inevitable Losses due to Heat Transfer (  T min = 10  C):  E X Loss = 14 kW C1:  E X T = -20 kW  E X P = -228 kW  E X tm = -248 kW Exergy Surplus is then:  E X Surplus = 248 – ( ) = 169 kW Required Exergy Efficiency for this Process:  X = 79/248 = 31.9 % It should be possible to design a Process that does not require external Cooling First attempt: Expand the Cold Stream from 4 bar to 1 bar prior to Heat Exchange

NTNU T. Gundersen Slide no. 24 After pre-expansion of the Cold Stream CC GrCC Modified Composite and Grand Composite Curves Evaluation: New Targets are: Q H,min = 60 kW (unchanged) and Q C,min = 12.5 kW (down from 155 kW) Power produced: W = kW (ideal expansion) Notice: The Cold Stream is now much colder than required (-126  C vs. -85  C -  T min )

NTNU T. Gundersen Slide no. 25 Pre-heating before Expansion of C1 CC GrCC Modified Composite and Grand Composite Curves Evaluation: New Targets are: Q H,min = 60 kW (unchanged), Q C,min = 0 kW (eliminated) Power produced: W = 155 kW (ideal expansion) Notice: The Cold Stream was preheated from -55  C to  C Temperature after Expansion is increased from -126  C to -115  C

NTNU T. Gundersen Slide no. 26 Expanding the Cold Stream in 2 Stages to make Composite Curves more parallel Evaluation: New Targets are: Q H,min = 64 kW (increased), Q C,min = 0 kW (unchanged) Power produced: W = 159 kW (ideal expansion) Reduced Driving Forces improve Exergy Performance at the Cost of Area CC GrCC This was an economic Overkill

NTNU T. Gundersen Slide no. 27 An Industrial Application - the Liquefied Energy Chain Oxyfuel Power Plant CO 2 Liquefaction Natural Gas Liquefaction Air Separation ASU NG Air LINLNG W CO 2 NG LCO 2 O2O2 LNG H2OH2O Power Production from ”stranded” Natural Gas with CO 2 Capture and Offshore Storage (for EOR) This Presentation

NTNU T. Gundersen Slide no. 28 The Base Case - using basic Pinch Analysis Heat Recovery first, Pressure Adjustments subsequently

NTNU T. Gundersen Slide no. 29 Base Case Composite Curves External Cooling required for Feasibility External Heating is ”free” (Seawater) Seawater CO 2 N2N2 NG LNG

NTNU T. Gundersen Slide no. 30 After a number of Manipulations The Composite Curves have been ”massaged” by the use of Expansion and Compression A. Aspelund, D.O. Berstad and T. Gundersen, ”An Extended Pinch Analysis and Design Procedure utilizing Pressure based Exergy for Subambient Cooling”, accepted for Applied Thermal Engineering, April 2007.

NTNU T. Gundersen Slide no. 31 A novel Offshore LNG Process Self-supported w.r.t. Power & no flammable Refrigerants

NTNU T. Gundersen Slide no. 32 The Natural Gas ”Path”

NTNU T. Gundersen Slide no. 33 The CO 2 ”Path”

NTNU T. Gundersen Slide no. 34 The Nitrogen ”Path”

NTNU T. Gundersen Slide no. 35 Concluding Remarks Current Methodologies fall short to properly consider important options related to Pressure in the Design of Subambient Processes The Problem studied here is considerably more complex than traditional HENS  TQ behavior of Process Streams are not fixed  Vague distinction between Streams and Utilities  HEN is expanded with Compressors & Expanders The Attainable Composite Curve Region is an important new Graphical Representation  Provides Insight into (subambient) Design Options  Quantitative Tool in the ExPAnD Methodology  Small Contribution to the area of Process Synthesis