See changelog here
The Leak Detection and Repair Simulator (LDAR-Sim) is an open-source modeling framework for exploring the effectiveness of methane leak detection programs. The purpose of LDAR-Sim is to enable transparent, collaborative, flexible, and intuitive investigation of emerging LDAR technologies, methods, work practices, regulations, and deployment strategies.
LDAR-Sim has many potential uses, including:
- Test emissions reduction equivalence among distinct LDAR programs
- Evaluate performance and cost of methane sensing technologies and work practices
- Predict the emissions mitigation of proposed or existing fugitive methane policies
- Inform the development and niche of technologies and work practices
To learn more about LDAR-Sim, you can:
- User manual manual
- Read our story map (less technical introduction).
- Read Fox et al., 2020.
For first time users, we recommend attempting to reproduce the case study results in Fox et al. 2020 (see below).
Thomas Fox: thomas@highwoodemissions.com
Mozhou Gao: mozhou.gao@ucalgary.ca
Thomas Barchyn: tbarchyn@ucalgary.ca
Chris Hugenholtz: chhugenh@ucalgary.ca
LDAR-Sim was invented by Thomas Fox, Mozhou Gao, Thomas Barchyn, and Chris Hugenholtz at the University of Calgary's Centre for Smart Emissions Sensing Technologies.
LDAR-Sim is free software: you can redistribute it and/or modify it under the terms of the MIT License . LDAR-Sim is distributed in the hope that it will be useful, but without any warranty; without even the implied warranty of merchantability or fitness for a particular purpose.
NOTE: This applies to all versions following Commit 69c27ec, Made on March 1st, 2021, Previous versions of LDAR-Sim were licensed under the GNU Affero General Public License. All redistributions or modifications made on LDAR-Sim versions created before Commit 69c27ec (March 1st, 2021) are required to be in compliance with version 3 of the GNU Affero General Public License.
The Fox et al. 2020 release is immortalized in a separate branch that can be found by clicking here.
The Fox et al., 2020 release contains the exact code and inputs used in our LDAR-Sim synthesis paper. We recommend using this release, especially for first time users.
Citation for this release: Fox, Thomas A., Mozhou Gao, Thomas E. Barchyn, Yorwearth L. Jamin, and Chris H. Hugenholtz. "An agent-based model for estimating emissions reduction equivalence among leak detection and repair programs." Journal of Cleaner Production (2020): 125237.
This guide is intended to get a user running with LDAR-Sim, note that even though we have supplied default variables, these should be used with caution, as many are not fully understood, are dependent on specific company workpractices, and very by geographical region.
Read and understand the LDAR-Sim LICENSE (MIT License). Read the user manual manual Read Fox et al 2020 to familiarize yourself with LDAR-Sim fundamentals.
Using Conda (Conda-forge) and the requirements file included in the "install folder" Follow the directions included in the Setting Up LDAR Sim Dev Environment file. The requirements.txt file can also be used with PIP and pipenv, but Python should be installed seperately.
- Install Miniconda3 newest version
- From Conda Shell: cd into LDAR-Sim/install
conda config --add channels conda-forgeconda config --set channel_priority strict1conda create -n ldar_sim --file requirements.txtconda activate ldar_sim
Alternatively pip and pipenv can be used to install there requirements file with:
pip install ldar_sim -r requirements.txt
if you are using satellite modules orbit predictor needs to be added to environement
pip install orbit_predictor==1.14.2
The application requires both facility and weather data to run. We have included sample facilities and weather data for Alberta as an example. Checkout the user manual for more information on formating of facility data. Weather data can either be downloaded directly. ERA5 data can be downloaded directly from copernicus using the /module_code/weather/ERA5_downloader.py module (see file for instructions). Note the output data is in hourly format, therefore the flag weather_is_hourly should be set to True. Multiple ERA nc files can be concatinated with ERA5_concat.py.
The simulation files allow a user to set global / program/ and method parameters. If a parameter is not included in the file a default value will be used. One Global file and at least one Program File is required for running the program while method files are required for running a method.
A Global yaml file is required, the most basic setup is as follows (note , that P_OGI, and P_none are required): G_.yaml =>
parameter_level: global // Denotes the program level (used for input handling)
version: '2.0' // Denotes the version
reference_program: P_OGI // Denotes the regulatory reference program for relative differences
baseline_program: P_none // Denotes a baseline program for estimating program mitigation, usually in place of no formal LDARP_OGI.yaml => A Program yaml file is required, the most basic setup is as follows (where a method Label points at a global parameter):
program_name: P_OGI // Denotes program name (must be unique)
parameter_level: program // Denotes the program level (used for input handling)
version: '2.0' // Denotes the version
method_labels: // Denotes the associated methods
- OGI P_none.yaml =>
program_name: P_none
parameter_level: program
version: '2.0'
method_labels: []
M_OGI.yaml => A Method yaml file is required, the most basic setup is as follows:
parameter_level: method // Denotes program name (must be unique)
version: '2.0' // Denotes the version
label: OGI // Specify the label to link to an associated program
deployment_type: mobile // How the technology operates, 'mobile', 'stationary' or 'orbit'
measurement_scale: component // Does the sensor measure at a site level, equipement level or component level
is_follow_up: False // Does the technology survey sites after a screening technology flags the site.
sensor:
MDL: [0.0362] // Minimum detectable leak in g/s.
cost:
per_site: 600 // Cost per site survey ($)
t_bw_sites:
vals: [30] // Time to travel between sites (minutes)
RS: 2 // Surveys required per year per site (ie. 2 surveys per site every year)
time: 120 // Time to perform detection at a site (minutes)Checkout the user manual for more info on the parameters.
The main program is a python script called LDAR_Sim_main.py. Within the virtual environent (or where all py packages are installed) run:
Python LDAR_Sim_main.py {G_XXX} {P_XXX} {M_XXX} {M_YYY}
where each arguement is a path too a global, program, or method input parameter file. for example:
Python LDAR_Sim_main.py ./simulations/G_.yaml ./simulations/P_air.yaml ./simulations/M_aircraft.yaml ./simulations/M_OGI_FU.yaml
alternatively, an entire directory can be passed using the "-P", "--in_dir" flags where all files within the directory are added to the program. for example:
Python LDAR_Sim_main.py --in_dir ./simulations.py
will load all files in the simulations folder into the program.
Output files, including maps, charts and csv files will be generated and placed in the output folder.
Note: that you can use absolute references or relative, where the the root folder is this folder.
Several LDAR-Sim advances are not publicly available at this time, including more advanced equivalence scenario modeling, specific method modules, and cost-effectiveness comparisons.
The Included python code follows strict PEP8 Standards for formatting with the a modification to the Line Length rule, where lines cannot exceed 100 characters. Contributed code will be rejected if it does not meet this standard. We suggest using PEP8 autoformatters and Linting (Flake8 , Black) when making contributions.
The authors welcome all contributions and collaborations. Please reach out - we would love to hear from you and/or work with you!