PtX-BOA: PtX Business Opportunity Analyser

PtX-BOA is a tool that aims to promote the export of a wide range of PtX molecules, including amongst others, green ammonia, e-methanol and synthetic fuels. Users can calculate the delivered cost of PtX molecules from an export country to an import country, with a detailed cost breakdown comparison highlighting the competitive edge of one country against another.

Development

Setup

After cloning the repository, create a virtual python environment and install the necessary dependencies:

pixi install --locked -e dev

The code is autoformatted and checked with pre-commit. If you make changes to the code that you want to commit back to the repository, please install pre-commit with:

pre-commit install

If you have pre-commit installed, every file in a commit is checked to match a certain style and the commit is stopped if any rules are violated. Before committing, you can also check your staged files manually by running:

pre-commit run

In order to run the tests locally run pytest in the root directory:

pixi run --locked -e dev pytest -vv

To start the app locally, run the following command:

pixi run --locked app

To start the app with development mode, run the following command. Development mode has debugging tabs in blue and green version, and no flh optimization in blue version.

pixi run --locked dev-app

Download optimization cache for local development

cd ptxboa\cache
scp -r ptxboa2:ptx-boa_offline_optimization/optimization_cache/* .

Release Procedure

• merge all relevant branches into develop
• create a relase branch
• change and commit CHANGELOG.md with description of changes
• update version (pixi run bumpversion patch|minor|major). This creates automatically a commit
• create pull requests to merge release into main
• merging this will automatically (via git action) create and publish the new docker image wingechr/ptx-boa:<VERSION>-<BRANCH>
• merge main back into develop

Update docker image in production

# connect to server
ssh ptxboa

# set variables
VERSION=3.0.5
APP=app
PORT=9000
MODE=prod
CACHEDIR=/home/ptxboa/ptx-boa_offline_optimization/optimization_cache

# is preview?
IS_PREVIEW=1
if [[ "$IS_PREVIEW" == "1" ]]; then
  PORT=9001
  APP="$APP-preview"
  MODE=preview
  # optionally: separate CACHEDIR
  # CACHEDIR=/home/sshfs_remote/optimization_cache
fi

# pull latest image from dockerhub
docker pull wingechr/ptx-boa:$VERSION
# stop and delete the currently running container "app"
docker stop $APP || true
docker rm $APP || true
# start the latest image as "app"
docker run -d -p $PORT:80 -v $CACHEDIR:/mnt/cache -e PTXBOA_MODE=$MODE --name $APP --restart unless-stopped wingechr/ptx-boa:$VERSION

# see logs
docker logs --follow $APP

Cleanup docker images from old versions

# check which docker images are downloaded
docker image ls
# show running containers
docker ps
# delete all unused objects
docker system prune -a

Testing / code coverage

pytest --cov=ptxboa --cov-report=term-missing --cov-report=html:htmlcov

Internal documentation

This section contains internal documenation on data flows, structure of the code base etc.

flowchart LR
    subgraph app [PtX-BOA WebApp]
        streamlit[ptxboa_streamlit.py]
    end

    subgraph script [PtX-BOA WebApp]
        offline_optimization[offline_optimization_script.py]
    end

    subgraph api [PtX-BOA Calculation Api]
        PtxboaAPI[ptxboa/api.py:PtxboaAPI]
        PtxCalc[ptxboa/api_calc.py:PtxCalc]
        DataHandler[ptxboa/api_data.py:DataHandler]
        PtxOpt[ptxboa/api_optimize.py:PtxOpt]
    end

    subgraph opt [FLH Optimizer]
        api_opt[flh_opt/api_opt.py:api_opt]
    end

    subgraph files [Files]
        profiles[flh_opt/renewable_profiles/*.csv]
        cache[ptxboa/cache/XX/XX/*.pickle]
        data[ptxboa/data/*.csv]
    end

    streamlit <-->|user data, settings| PtxboaAPI
    profiles --> api_opt
    cache <-->|cache| PtxOpt
    data --> DataHandler
    PtxOpt --> api_opt
    PtxboaAPI <--> PtxCalc
    DataHandler <--> PtxOpt
    PtxboaAPI <-->|user data| DataHandler
    offline_optimization --> cache
    offline_optimization <--> api_opt

Loading

Structure of input data

• csv file(s) with scalar data
• RE profiles (flh and weighting coefficients)

Caching optimization results

• location of cached files
• hashing

Creating renewable generation profiles

• atlite
• methodology

The PyPSA model

The pypsa optimization model is created and solved via the flh_opt.api_opt.optimize() function

ptx-boa/flh_opt/api_opt.py

Line 148 in 61a5915

def optimize(

.

Function parameters are a dictionary with all required parameters, and the path to the folder with the renewable profiles data.

Function output is a dictionary with the results of the optimization, and the pypsa network object that contains the solved model.

Example input dict

{
  "SOURCE_REGION_CODE": "GYE",
  "RES": [
    {
      "CAPEX_A": 0.826,
      "OPEX_F": 0.209,
      "OPEX_O": 0.025,
      "PROCESS_CODE": "PV-FIX"
    }
  ],
  "ELY": {
    "EFF": 0.834,
    "CAPEX_A": 0.52,
    "OPEX_F": 0.131,
    "OPEX_O": 0.2,
    "CONV": {
      "H2O-L": 0.677
    }
  },
  "DERIV": {
    "EFF": 0.717,
    "CAPEX_A": 0.367,
    "OPEX_F": 0.082,
    "OPEX_O": 0.132,
    "PROCESS_CODE": "CH4SYN",
    "CONV": {
      "CO2-G": 0.2,
      "HEAT": -0.2,
      "H2O-L": -0.15
    }
  },
  "H2O": {
    "CAPEX_A": 0.07726085034488815,
    "OPEX_F": 0.0356900588308774,
    "OPEX_O": 0,
    "CONV": {
      "EL": 0.003
    }
  },
  "CO2": {
    "CAPEX_A": 0.07726085034488815,
    "OPEX_F": 0.0356900588308774,
    "OPEX_O": 0,
    "CONV": {
      "EL": 0.4515,
      "HEAT": 1.743,
      "H2O-L": -1.4
    }
  },
  "EL_STR": {
    "EFF": 0.544,
    "CAPEX_A": 0.385,
    "OPEX_F": 0.835,
    "OPEX_O": 0.501
  },
  "H2_STR": {
    "EFF": 0.478,
    "CAPEX_A": 0.342,
    "OPEX_F": 0.764,
    "OPEX_O": 0.167
  },
  "SPECCOST": {
    "H2O-L": 0.658,
    "CO2-G": 1.0
  }
}

Example output dict

{
  "RES": [
    {
      "SHARE_FACTOR": 0.519,
      "FLH": 0.907,
      "PROCESS_CODE": "PV-FIX"
    }
  ],
  "ELY": {
    "FLH": 0.548
  },
  "DERIV": {
    "FLH": 0.548
  },
  "EL_STR": {
    "CAP_F": 0.112
  },
  "H2_STR": {
    "CAP_F": 0.698
  }
}

Example flowchart

This flowchart shows an example model as being created by the optimize function: