(599ao) Simulation-Optimization Approaches to the Solve Strategic and Tactical Problems of Biomass Utilization for Energy and Commodity Chemicals

Simulation-Optimization Approaches to the Solve Strategic
and Tactical Problems of Biomass Utilization for Energy and Commodity Chemicals

Ismail Fahmi, Selen Cremaschi

Department of Chemical Engineering, The University of Tulsa,
800 South Tucker Drive, Tulsa, Oklahoma 74104, USA

10C08 Poster Session: Computers In Operations and
Information Processing

Abstract

As the fossil-based material reserves
deplete, but its demand by energy and commodity chemicals are increasing,
alternative feedstocks need to be discovered. One alternative feedstock is the
renewable, locally available, and abundant biomass. Incorporating biomass as a
feedstock to energy and commodity chemicals industry requires a significant
amount of investments in capacity expansions and in the research and
development (R&D). In addition, for each selected technology, the optimal
configuration and operating conditions of unit operations should be identified.
As a result, there are two investigation levels in the problem of incorporating
biomass to both energy and commodity chemicals industries: (1) Strategic level,
which covers all possible technologies supporting the overall biomass
processing system and the optimal decision will answer in which technologies,
when, and how much to invest, (2) Tactical level, which covers all production alternatives
for a selected processing technology and the solution will identify the optimal
process flowsheet and operating conditions. In this poster, two different
simulation-optimization approaches are presented to address the strategic and
tactical problems of incorporating biomass to energy and commodity chemicals
production.

The first type of
simulation-optimization approach, also known as the surrogate-based
optimization, uses simulation to generate surrogate models to accurately
represent the original first-order principal relationship. The surrogate models
are then used in a deterministic optimization problem formulation, which is
solved to yield the optimum solution. In this work, surrogate-based
optimization is utilized to solve the process synthesis problems, which has
been demonstrated in [1] with biodiesel production as the case
study. The three process alternatives considered were using super critical
methanol as reactant, using base catalyst, and using acid catalyst. In [1], it was shown that using surrogate models to
replace the original first-order principal relationship of each and every unit
operation in the optimization formulation can reduce the computational costs to
obtain a solution.

The second type of
simulation-optimization approach uses deterministic optimization with
stochastic simulation in an integrated fashion to study the impact of
uncertainties on the system performance. In the second simulation-optimization
approach, deterministic optimization is used to produce the optimum decision set
of the system at its current state. Stochastic simulation is used to predict
the system performance under uncertainties. In a single simulation-optimization
run, the simulation proceeds through the time steps of a timeline based on the
decisions set determined by the deterministic optimization. Because of the
accumulated differences between the expected and the realized behavior of the
system, the system performance under these decisions may no longer be
applicable or feasible at one point in time. This situation is called a trigger
event. In case of a trigger event, the simulation is halted and deterministic
optimization is recalled to produce a new set of optimal decision for the
remainder of the timeline. The control is passed back to simulation along with
the new set of decisions. The iteration between the deterministic optimization
and stochastic simulation continues until the final time step. After the
completion of the final step, a unique controlled evolution of the system
performance under uncertainties (i.e., one unique timeline) is generated. Multiple
unique timelines are produced because of different realizations of the
uncertain parameters. The statistically significant number of timelines can be
used to study the system performance under uncertainties. In this work, the
second type of simulation-optimization approach is used to address the problem
of strategic investment planning to shift from fossil-based feedstocks to
biomass feedstock for the chemical process industry (CPI). This problem is
referred to as the Biomass to Commodity Chemicals (BTCC) investment planning problem.
A case study of ethylene production from biomass and naphtha has been presented
to demonstrate the performance of the second type of the simulation-optimization
[2]. However, generating a single timeline may require
significant amounts of computational costs. The major sources of the
computational costs are the number of timelines and the computational resources
necessary to solve the deterministic optimization problem, which may be solved
many times during a single timeline. To address the former problem, a
systematic analysis on different sampling methods to cover the uncertain
parameters space of the BTCC investment planning problem is presented. To
address the latter problem, a study of the development of a specialized global
solver to solve the optimization problem is presented.

Keywords: simulation-based optimization, biomass feedstock
incorporation, strategic investment planning and tactical process synthesis problem

References:

[1]        I. Fahmi and S. Cremaschi,
"Process Synthesis of Biodiesel Production Plant using Artificial Neural
Networks as the Surrogate Models," Computers and Chemical Engineering
(in review), 2011.

[2]        I. Fahmi and S. Cremaschi,
"A Prototype Simulation-based Optimization Approach to Model Feedstock
Development for Chemical Process Industry," 22nd European Symposium on
Computer Aided Process Engineering, 2012.