# $Id: hen-10sp1.in 2 2004-04-17 20:20:23Z ucecesf $
#
# This input file is for solving the well known 10sp1 heat exchanger
# network synthesis problem. One reference to it can be obtained in:
#
# Lewin, D R (1998). "A generalized method for HEN synthesis using
# stochastic optimization -- II. The synthesis of cost-optimal
# networks," Computers & Chemical Engineering 22(10) 1387-1405.
#
# It is a useful case study in that it severely penalizes the number
# of units and is different, therefore from the other case studies.
#
# Copyright (c) 2001-2002, Eric S Fraga, All rights reserved.
# -----------------------------------------------------------------
#
# Make it easy to reference object types by specifying the base
# package here:
#
use jacaranda.design
#
# the main input is the HEN synthesis problem definition which
# consists of five cold and five hot streams. The definition of each
# stream consists of the following information:
#
# htr task Tin Tout duty U
#
# where type is either "cool" or "heat" (you wish to cool hot streams
# and heat cold streams). I prefer using this terminology as cold and
# hot are relative terms and often the hot streams are colder than the
# cold ones!  The point is that the problem is one of cooling and
# heating. U is the heat transfer coefficient in kW/K/m^2.
#
hen.HEN 10sp1hen
  # first specify the hot streams, streams which need cooling
  htr H1 cool "433" "366" " 8.79*(433-366)*kW" "5.75*kW/K/m^2"
  htr H2 cool "522" "411" "10.55*(522-411)*kW" "5.75*kW/K/m^2"
  htr H3 cool "544" "422" "12.56*(544-422)*kW" "5.75*kW/K/m^2"
  htr H4 cool "500" "339" "14.77*(500-339)*kW" "5.75*kW/K/m^2"
  htr H5 cool "472" "339" "17.73*(472-339)*kW" "5.75*kW/K/m^2"
  # and now the cold streams, those that need heating.
  htr C1 heat "355" "450" "17.28*(450-355)*kW" "1.47*kW/K/m^2"
  htr C2 heat "366" "478" "13.90*(478-366)*kW" "1.47*kW/K/m^2"
  htr C3 heat "311" "494" " 8.44*(494-311)*kW" "1.47*kW/K/m^2"
  htr C4 heat "333" "433" " 7.62*(433-333)*kW" "1.47*kW/K/m^2"
  htr C5 heat "389" "495" " 6.08*(495-389)*kW" "1.47*kW/K/m^2"
end
#
# we also need to define the utilities that are available. The format
# of this input is:
# 
# type  description  Tin  Tout  TransferCoefficient  Cost in $/J
#
# For this problem, we have only two utilities: steam for heating and
# water for cooling. The steam exhibits no temperature change
# (available heat is latent) whereas the water will go up
# significantly.
ps.DiscreteUtilities utilities
  hot "Steam" "509" "509" "5.0 *kW/K/m^2" "37.64/kW/(hpy*3600)"
  cold "Water" "311" "355" "1.0 *kW/K/m^2" "18.12/kW/(hpy*3600)"
end

#
# The optimization criterion is the annualized cost, with capital cost
# amortized over a period of a single year. This penalizes the use of
# large numbers of units. The arguments to the criterion command are:
#
# criterion name operation expression units minimum format
#
ps.Criteria annualized
  criterion OneYear sum "opercost+capcost" "$/yr" 0.0 "#,##0"
end
#
# We now define the base heat exhanger model for utility exchange. The
# main input is the actual cost correlation to use for capital
# costs. The operating costs come from the utilities used, as defined
# above.
#
ps.HeatExchanger hex
  model
    deltain = abs(T1in-T2out)
    deltaout = abs(T1out-T2in)
    lmtd = (deltain - deltaout)/log(deltain/deltaout)
    A = Q/U/lmtd
    opercost = Cu*Q*hr*hpy
    capcost = 145.63 * A^0.6
  end
end
#
# tell the system the settings defined above. In particular, the
# criterion for optimization, the heat exchanger, and the list of
# utilities must be specified.
#
ps.Data data
  criteria annualized		# ranking criterion
  hex hex			# heat exchanger model
  utils utilities		# utilities for other heating/cooling
  print
end
#
# Finally, we can bring up the graphical display which will allow us
# to investigate the different HEN options. The input to this object
# is the HEN object (defined above) which has the definition of the
# hot and cold streams. The utilities etc are automatically retrieved
# from the global data object defined immediately above. Finally, the
# default height and width of the display panel is changed to ensure a
# reasonable discretization for this problem. In particular, you will
# see that the two streams C3 and H3 are very close in duty amounts
# and smaller panel widths will lead to these being treated as
# identical in size. Of course, for early design, this may not be a
# problem. However, the user is able to control the size of the panel,
# and hence the level of discretization, easily.
#
visual.QDeltaT 10sp1
  # the problem is defined by the heating and cooling requests in the
  # HEN object defined earlier.
  hen 10sp1hen
  # we now present the graphical view of the problem. This is the
  # basis for both user interaction and for optimization.
  width 800
  height 600
  view
  # having created the graphical view, we can now request an
  # optimization. the progress of the optimization will be shown
  # graphically! to enable this, uncomment the following line. The
  # line is commented out because, on slow systems, this can take an
  # appreciable amount of time and the optimization option is
  # available on the menu in any case.
  ### optimize
end