FEAST Tutorial.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# An Introduction to FEAST v3.1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "FEAST v3.1 is the Decemeber 2020 release of the Fugitive Emmissions Abatement Simulation Toolkit (FEAST). The toolkit simulates emissions from natural gas fields and the efficacy of leak detection and repair programs. This tutorial gives an example of how to generate an example emissions scenario in FEAST 3.1, define LDAR programs and analyze results. The tutorial is interactive, so feel free to experiment with the code cells and discover how your changes affect the results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "import copy\n",
    "import feast\n",
    "import pickle\n",
    "from multiprocessing import Pool"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Defining an emission scenario\n",
    "FEAST requires all potential emission sources to be declared as Components. Components are characterized by their emission production rate, emission size distribution, and the fraction of components expected to be emitting at the beginning of the simulation. In order to improve transparency, FEAST 3.1 does not have a \"default\" scenario. All emitters must be defined directly in order to run the simulation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define fugitive components\n",
    "For this example, we will define exactly one type of component: a fugitive emitter. Additional types of components can be created in the same way. Venting components (that is, components that are intended to produce emissions) can be created by setting the \"base_reparable\" variable to False. In real production sites vented and fugitive emissions will both be present."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "comp_fug = feast.EmissionSimModules.infrastructure_classes.Component(\n",
    "    name='Fugitive emitters',\n",
    "    emission_data_path='ExampleData/DataObjectInstances/production_emissions.p',\n",
    "    emission_per_comp=0.00231,  # Fraction of components expected to be emitting at the beginning of the simulation.\n",
    "    # emission_production_rate is set to about 5 new emissions per well per year\n",
    "    emission_production_rate=5.4 / 650 / 365,  # number of new emissions per component per day\n",
    "    repair_cost_path='ExampleData/DataObjectInstances/fernandez_leak_repair_costs_2006.p',\n",
    "    base_reparable=True\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define a production Site\n",
    "A site is made of a collection of components. The components are stored in the variable comp_dict. The analyst may include many component types in the comp_dict.\n",
    "\n",
    "Using the copy library is good practice if a component type is going to be reused in multiple sites to avoid shared parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "wells_per_site = 2\n",
    "comps_per_well = 650\n",
    "basic_site = feast.EmissionSimModules.infrastructure_classes.Site(\n",
    "    name='basic site',\n",
    "    comp_dict={\n",
    "        # ------ The number of components is proportional to the number of wells, ind2\n",
    "        'Fugitive ': {'number': comps_per_well * wells_per_site,\n",
    "                     'parameters': copy.deepcopy(comp_fug)},\n",
    "    },\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define a Time object\n",
    "FEAST requires the simulation time settings to be declared in a time object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "time = feast.EmissionSimModules.simulation_classes.Time(delta_t=1, end_time=365)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define a GasField\n",
    "All of the sites to be simulated must be collected into a dict and then passed to a GasField object. As of FEAST 3.1, a GasField object may also contain meteorological data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "site_dict = {\n",
    "    'basic site': {\n",
    "        'number': 100, # The number of basic_site instances to simulate in the gas_field\n",
    "        'parameters': basic_site\n",
    "    }\n",
    "    # If additional site types are needed, they should be added here as separate dict entries\n",
    "}\n",
    "gas_field = feast.EmissionSimModules.infrastructure_classes.GasField(\n",
    "    sites=site_dict,\n",
    "    time=time,\n",
    "    met_data_path='ExampleData/TMY-DataExample.csv'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Define LDAR programs\n",
    "LDAR programs are made of detection methods, repair methods, and the connections between them. The first step in creating LDAR programs is defining detection and repair methods."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define a repair method\n",
    "A repair method takes one input parameter: the delay (days) between when an emission is detected and when it is repaired."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "rep3 = feast.DetectionModules.repair.Repair(repair_delay=3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define detection methods\n",
    "Detection methods accept a variety of variables that depend on the type of detection method to be simulated. In all cases, a detection method requires a set of conditions to be specified where the probability of detection (PoD) is known. Three examples are demonstrated below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# An OGI-like example where the PoD depends on emission rate only.\n",
    "def make_ogi(dispatch_obj):\n",
    "    points = np.array([0.5, 1, 2, 3, 7]) * 0.01157 # g/s\n",
    "    probs = np.array([0, 0.25, 0.5, 0.75, 1])\n",
    "    ogi = feast.DetectionModules.comp_survey.CompSurvey(\n",
    "        time,\n",
    "        survey_interval=180, # days\n",
    "        survey_speed=500, # comps/hr\n",
    "        ophrs={'begin': 8, 'end': 17},\n",
    "        labor=400, # $/hr\n",
    "        detection_variables={'flux': 'mean'},\n",
    "        detection_probability_points=points,  # g/s\n",
    "        detection_probabilities=probs,\n",
    "        dispatch_object=copy.deepcopy(dispatch_obj),\n",
    "        site_queue=[],\n",
    "    )\n",
    "    return ogi\n",
    "ogi = make_ogi(rep3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# An OGI-like example that does not automatically survey components but can be dispatched\n",
    "def make_ogi_no_survey(dispatch_obj):\n",
    "    points = np.array([0.5, 1, 2, 3, 7]) * 0.01157 # g/s\n",
    "    probs = np.array([0, 0.25, 0.5, 0.75, 1])\n",
    "    ogi_no_survey = feast.DetectionModules.comp_survey.CompSurvey(\n",
    "        time,\n",
    "        survey_interval=None, # days\n",
    "        survey_speed=500, # comps/hr\n",
    "        ophrs={'begin': 8, 'end': 17},\n",
    "        labor=400, # $/hr\n",
    "        detection_variables={'flux': 'mean'},\n",
    "        detection_probability_points=points,  # g/s\n",
    "        detection_probabilities=probs,\n",
    "        dispatch_object=copy.deepcopy(dispatch_obj),\n",
    "        site_queue=[],\n",
    "    )\n",
    "    return ogi_no_survey\n",
    "ogi_no_survey = make_ogi_no_survey(rep3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# A site level survey method\n",
    "def make_plane_survey(dispatch_obj):\n",
    "    points = np.array([25, 50, 100, 200, 400]) * 0.01157 # g/s\n",
    "    probs = np.array([0, 0.25, 0.5, 0.75, 1])\n",
    "    plane_survey = feast.DetectionModules.site_survey.SiteSurvey(\n",
    "        time,\n",
    "        survey_interval=180, # days\n",
    "        sites_per_day=200,\n",
    "        site_cost=100, # $/site\n",
    "        detection_variables={'flux': 'mean'},\n",
    "        detection_probability_points=points,\n",
    "        detection_probabilities=probs,\n",
    "        dispatch_object=copy.deepcopy(dispatch_obj),\n",
    "        site_queue=[],\n",
    "        ophrs={'begin': 8, 'end': 17}\n",
    "    )\n",
    "    return plane_survey\n",
    "plane_survey = make_plane_survey(ogi_no_survey)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Continuous monitor method where detection depends on emission rate and windspeed. \n",
    "def make_cont_monitor(dispatch_obj):\n",
    "    points = [[0.5, 1], \n",
    "              [1.0, 1], \n",
    "              [1.1, 1], \n",
    "              [0.5, 5], \n",
    "              [1.0, 5], \n",
    "              [1.1, 5],\n",
    "              [0.5, 5.1], \n",
    "              [1.0, 5.1],\n",
    "              [1.1, 5.1]]\n",
    "    time_to_detect_days = [np.infty, 1, 0, np.infty, 5, 0, np.infty, np.infty, np.infty]\n",
    "    cont_monitor = feast.DetectionModules.site_monitor.SiteMonitor(\n",
    "        time,\n",
    "        time_to_detect_points=points,\n",
    "        time_to_detect_days=time_to_detect_days,\n",
    "        detection_variables={'flux': 'mean', 'wind speed': 'mean'},\n",
    "        site_queue=list(range(gas_field.n_sites)),\n",
    "        dispatch_object=copy.deepcopy(dispatch_obj),\n",
    "        ophrs={'begin': 8, 'end': 17}\n",
    "    )\n",
    "    return cont_monitor\n",
    "cont_monitor = make_cont_monitor(ogi_no_survey)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Collect detection methods into LDARPrograms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define LDAR programs\n",
    "ogi_survey = feast.DetectionModules.ldar_program.LDARProgram(\n",
    "    copy.deepcopy(gas_field), {'ogi': copy.deepcopy(ogi)},\n",
    ")\n",
    "# tiered survey\n",
    "tech_dict = {\n",
    "    'plane': plane_survey,\n",
    "    'ogi': plane_survey.dispatch_object\n",
    "}\n",
    "plane_ogi_survey = feast.DetectionModules.ldar_program.LDARProgram(\n",
    "    copy.deepcopy(gas_field), tech_dict,\n",
    ")\n",
    "\n",
    "# continuous monitor\n",
    "tech_dict = {\n",
    "    'cm': cont_monitor,\n",
    "    'ogi': cont_monitor.dispatch_object\n",
    "}\n",
    "cm_ogi = feast.DetectionModules.ldar_program.LDARProgram(\n",
    "    copy.deepcopy(gas_field), tech_dict,\n",
    ")\n",
    "\n",
    "# All programs\n",
    "ldar_dict = {\n",
    "    # Note: continuous monitor simulations dominate the computation load.\n",
    "    # Comment out the next line to exclude the continuous monitor program from the simulation.\n",
    "    'cm': cm_ogi,\n",
    "    'ogi': ogi_survey,\n",
    "    'plane': plane_ogi_survey\n",
    "}\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Run simulation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The evaluation is 0% complete\n",
      "The evaluation is 10% complete\n",
      "The evaluation is 20% complete\n",
      "The evaluation is 30% complete\n",
      "The evaluation is 40% complete\n",
      "The evaluation is 50% complete\n",
      "The evaluation is 60% complete\n",
      "The evaluation is 70% complete\n",
      "The evaluation is 80% complete\n",
      "The evaluation is 90% complete\n"
     ]
    }
   ],
   "source": [
    "scenario = feast.EmissionSimModules.simulation_classes.Scenario(time=time, \n",
    "                                                                gas_field=gas_field, \n",
    "                                                                ldar_program_dict=ldar_dict)\n",
    "scenario.run(dir_out='TutorialResults', display_status=True, save_method='pickle')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from feast.ResultsProcessing.plotting_functions import time_series\n",
    "# The next line loads results from the most recently completed simulation.\n",
    "realn = len(os.listdir('TutorialResults')) - 1\n",
    "time_series('TutorialResults/realization' + str(realn) + '.p', line_width=4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluating many simulations\n",
    "FEAST is a stochastic model. Therefore, many iterations of the model should be considered to evaluate the variability between iterations. The cells below run 30 iterations of a simulation similar to that defined above, but without a continuous monitor method so that simulations can complete more quickly. The cell will take a few minutes to evaluate.\n",
    "\n",
    "Note that many objects in FEAST must be re-created during each iteration in order to avoid results from the previous iteration persisting into the next simulation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tutorial_helper import make_iteration\n",
    "from multiprocessing import Pool"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "nreal=30\n",
    "with Pool(10) as p:\n",
    "    p.map(make_iteration, [ind for ind in range(nreal)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analyzing results\n",
    "FEAST offers a few simple tools for analyzing multiple pickled results files. \n",
    "The results_analysis function calculates the net present value of the of finding leaks and repairing leaks. If the gas_price parameter is set greater than zero, it also accounts for the value of avoided emissions. The npv output of the function returns a dict containing one NxK array for each of these cash flows, where N is the number of iterations, and K is the number of technologies. The order of LDAR programs is specified by the *programs* output parameter.\n",
    "\n",
    "In addition, *em_timeseries* contains an array of emission timeseries with dimension (N_programs + 1, N_timesteps, N_iterations). The first dimension is N_programs+1 because it includes an entry for the Null scenario as well as each LDAR program.\n",
    "\n",
    "The abatement_cost_plotter function uses the results_analysis function to calculate and plot the mitigation cost of each LDAR program."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "ra = feast.ResultsProcessing.results_analysis_functions.results_analysis\n",
    "npv, em_timeseries, programs = ra('TutorialResults-MC', gas_price=0, discount_rate=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "feast.ResultsProcessing.plotting_functions.abatement_cost_plotter('TutorialResults-MC', \n",
    "                                                                  gwp=34, discount_rate=0, gas_price=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Utilizing the JSON save method\n",
    "When running many simulations, the 'pickle' save method results in unnecessarily large results files. The JSON save method provides a more efficient alternative. The resulting files can be loaded without importing FEAST."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The evaluation is 0% complete\n",
      "The evaluation is 10% complete\n",
      "The evaluation is 20% complete\n",
      "The evaluation is 30% complete\n",
      "The evaluation is 40% complete\n",
      "The evaluation is 50% complete\n",
      "The evaluation is 60% complete\n",
      "The evaluation is 70% complete\n",
      "The evaluation is 80% complete\n",
      "The evaluation is 90% complete\n"
     ]
    }
   ],
   "source": [
    "time = feast.EmissionSimModules.simulation_classes.Time(delta_t=1, end_time=365)\n",
    "gas_field = feast.EmissionSimModules.infrastructure_classes.GasField(\n",
    "    sites=site_dict,\n",
    "    time=time,\n",
    "    met_data_path='ExampleData/TMY-DataExample.csv'\n",
    ")\n",
    "ogi = make_ogi(rep3)\n",
    "ogi_no_survey = make_ogi_no_survey(rep3)\n",
    "plane_survey = make_plane_survey(ogi_no_survey)\n",
    "# LDAR programs\n",
    "ogi_survey = feast.DetectionModules.ldar_program.LDARProgram(\n",
    "    copy.deepcopy(gas_field), {'ogi': ogi},\n",
    ")\n",
    "# tiered survey\n",
    "tech_dict = {\n",
    "    'plane': plane_survey,\n",
    "    'ogi': plane_survey.dispatch_object\n",
    "}\n",
    "plane_ogi_survey = feast.DetectionModules.ldar_program.LDARProgram(\n",
    "    copy.deepcopy(gas_field), tech_dict,\n",
    ")\n",
    "ldar_dict = {\n",
    "    'ogi': ogi_survey,\n",
    "    'plane': plane_ogi_survey\n",
    "}\n",
    "scenario = feast.EmissionSimModules.simulation_classes.Scenario(time=time, \n",
    "                                                            gas_field=gas_field, \n",
    "                                                            ldar_program_dict=ldar_dict)\n",
    "scenario.run(dir_out='TutorialResults-json', display_status=True, save_method='json')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "with open('TutorialResults-json/realization0.json') as f:\n",
    "    res = json.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['ogi', 'plane', 'Null'])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# A json result file is imported as a multilevel dict. The top level of the dict contains one entry for each LDAR program.\n",
    "res.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['emission timeseries', 'vent timeseries', 'repair cost', 'plane', 'ogi', 'ogi repair'])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The second level of the dict contains emission and vent timeseries, \n",
    "# as well as repair costs, and keys to additional levels of the dict\n",
    "# For complete documentation of the data structure, \n",
    "res['plane'].keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['deployment costs', 'deployment count', 'op env site fails', 'op env field fails'])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# To learn more about what is available in each layer, check out the keys!\n",
    "res['plane']['ogi'].keys()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}