Integrated Design and Operation Model for NGCC-PCC-DAC Retrofit

This repo contains the optimization model for the NGCC-PCC-DAC retrofit design modeled in Julia/Jump and Python/Pyomo.

Main features

• co-optimizes the design and operational decisions
• considers multiple future energy scenarios (varying electricity prices for a year + varying CO2 prices)
• formulates a novel DAC model to account for sorbent dynamics
• contains logic for system start-up/shutdown

For the full model description, please contact Pengfei Cheng.

Branches

Currently there are 4 branches:

• main: the most up-to-date one with all the features
• base-NGCC: contains the optimization model for the operations of B31A (unretrofitted NGCC)
• pyomo-ver: the identical model formulated in Python/Pyomo
• pyomo-SP: the stochastic version of the model in in Python/Pyomo

main Branch Layout

.
├── FLECCS_env                          # Julia environment configuration
├── src                                 # source files
│   ├── models                          # optimization model files
│       ├── constraints
│       ├── output
│       ├── params
│       ├── model.jl                    # main file
│       ├── obj.jl
│       └── variables.jl
│   ├── post_processing                 # post-processing scripts
│   ├── regression                      # regression scripts for data from simulation
│   └── resources                       # other resource files
├── .gitignore
├── README.md
└── requirements.txt

Installation

Julia Environment

The model code is based on Julia 1.7.0. To install packages, run the following commands in a Julia REPL:

# press "]" to enter Julia's Pkg (package manager) REPL
# cd to the project directory
(@1.7) pkg> activate FLECCS_env
(FLECCS_env) pkg> instantiate

For more information, see the instruction.

Python Environment

Python scripts/Jupyter notebooks are used for post processing, and only basic packages are needed. To install these packages, run

pip install -r requirements.text

Optimization solver

To use a certain MILP solver, please install it separately. Please make sure that the solver has accessible API to Julia/Jump.

Usage

To run the model in the Julia REPL (in terminal):

# make sure the environment has been configured
include("src/models/model.jl")
# remember to set the output directory for the first-time usage in src/models/output/output.jl

Data Sources

Location	File name	Function	Source
src/regression/	coefs.csv	coefficients from simulation (only considers 50% and 100% load), currently in use	linear regression
src/resources/	NGCC_performance.xlsx	system performance coefficients	Howard Hendrix
src/resources/	overall-price-signals.csv	all sets of electricity price series	Princeton team and NREL team
src/models/params	.jl files	other data	Georgia Tech FLECCS team

Credits

The code was initially developed by David Thierry and further developed and maintained by Pengfei Cheng. The development was funded by the ARPA-E's FLECCS program and supported by the Georgia Tech FLECCS Team (Matthew Realff, Joseph Scott, Fani Boukouvala, Howard Hendrix, Christopher Jones, Ryan Lively, Frank Kong, and Trimeric staff).