Table of contents
  1. Story
    1. Data Science for EPA Fracturing Data
    2. Slides
      1. Slide 1 Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0
      2. Slide 2 Outline
      3. Slide 3 Research Questions
      4. Slide 4 Research Products
      5. Slide 5 General Information
      6. Slide 6 Hydraulic Fracturing Water Cycle
      7. Slide 7 Additives
      8. Slide 8 Example FracFocus Disclosure
      9. Slide 9 Hydraulic Fracturing Fluid Composition
      10. Slide 10 Disclosure Locations
      11. Slide 11 Ingredients
      12. Slide 12 Confidential Business Information (CBI) Ingredients
      13. Slide 13 Additive Ingredients
      14. Slide 14 Most Frequently Reported Additive Ingredients for Gas Wells
      15. Slide 15 Most Frequently Reported Additive Ingredients for Oil Wells
      16. Slide 16 Base Fluids
      17. Slide 17 Non-Aqueous Fluids in Base Fluid
      18. Slide 18 Cumulative Water Volume, by State
      19. Slide 19 Per-Disclosure Water Volume, by State
      20. Slide 20 Water Sources
      21. Slide 21 Analysis Summary 1
      22. Slide 22 Analysis Summary 2
      23. Slide 23 State Summaries 1
      24. Slide 24 State Summaries 2
      25. Slide 25 State Summaries 3
      26. Slide 26 Project Database
      27. Slide 27 What Did We Learn?
      28. Slide 28 Questions?
  2. Slides
    1. Slide 1 EPA Big Data Analytics: Data Science for EPA Fracturing Data
    2. Slide 2 Announcement
    3. Slide 3 EPA's Study of Hydraulic Fracturing for Oil and Gas and Its Potential Impact on Drinking Water Resources
    4. Slide 4 The Hydraulic Fracturing Water Cycle
    5. Slide 5 Data Mining for Data Science
    6. Slide 6 Data Science for EPA Fracturing Data: MindTouch Knowledge Base
    7. Slide 7 Data Mining Files and Attachments
    8. Slide 8 Data Science for EPA Fracturing Data: MindTouch Knowledge Base Find
    9. Slide 9 Data Science for EPA Fracturing Data: Spreadsheet Knowledge Base
    10. Slide 10 Data Science for EPA Fracturing Data: Spotfire Cover Page
    11. Slide 11 Data Science for EPA Fracturing Data: Spotfire Data Management Tables
    12. Slide 12 Data Science for EPA Fracturing Data: Spotfire Oil & Gas by County
    13. Slide 13 Data Science for EPA Fracturing Data: Spotfire Oil & Gas by Well by County
    14. Slide 14 Data Science for EPA Fracturing Data: Spotfire Disclosures by County
    15. Slide 15 Data Science for EPA Fracturing Data: Spotfire Ingredients by County
    16. Slide 16 Conclusions and Recommendations
  3. Spotfire Dashboard
  4. Research Notes
  5. EPA's Study of Hydraulic Fracturing for Oil and Gas and Its Potential Impact on Drinking Water Resources
    1. The Hydraulic Fracturing Water Cycle
      1. Stage 1: Water Acquisition 1
      2. Stage 2: Chemical Mixing
      3. Stage 3: Well Injection
      4. Stage 4: Flowback 6 and Produced Water 7 (Hydraulic Fracturing Wastewaters)
      5. Stage 5: Wastewater Treatment and Waste Disposal
  6. Published Scientific Papers
    1. Analysis of Existing Data
    2. Laboratory Studies
      1. Analytical Method Development
    3. Scenario Evaluations
      1. Subsurface Migration Modeling
  7. EPA Analysis of FracFocus 1 Data
  8. Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0
    1. Overview
    2. Background
    3. State Summaries
    4. Results
    5. Study Limitations
  9. Data Management and Quality Assessment Report
    1. Cover Page
    2. Disclaimer
    3. List of Tables
    4. List of Figures
    5. Preface
    6. Acknowledgements
    7. List of Acronyms
    8. 1. Introduction
    9. 2. Source Data
    10. 3. Database Development
      1. 3.1. Downloading and Conversion
        1. Figure 1. Example FracFocus 1.0 disclosure
      2. 3.2. Extraction and Parsing
        1. Table 1 Summary of parsing success
      3. 3.3. Output Data Structure
    11. 4. Assignment of Hydrocarbon Regions to Disclosures
      1. Table Five publicly available datasets with national coverage
    12. 5. Quality Assurance Process for Locational Data
    13. 6. Chemical Name Standardization
    14. 7. Data Field Descriptions
      1. 7.1. Data Fields in Main Tables
        1. 7.1.1. Well Header Field Descriptions
        2. 7.1.2. Ingredient Field Descriptions
      2. 7.2. Data Fields in Tables Associated with Standardizations
        1. 7.2.1. Chemical Name Standardization
        2. 7.2.2. Operator Standardization Information
        3. 7.2.3. Trade Name Standardization
        4. 7.2.4. Ingredient Purpose Standardization
      3. 7.3. Data Fields in Other Tables
        1. 7.3.1. Proppant Identification
        2. 7.3.2. Resin Coating Identification
        3. 7.3.3. CBI Identification
        4. 7.3.4. Water Source Identification
        5. 7.3.5. Purpose Categorization
        6. 7.3.6. State Regulation Information
        7. 7.3.7. County Information
        8. 7.3.8. Water Synonyms
        9. 7.3.9. Unparsed PDFs
    15. 8. Summary
    16. References
    17. Footnotes
      1. 1
      2. 2
      3. 3
      4. 4
      5. 5
      6. 6
      7. 7
      8. 8
      9. 9
  10. Reproducing FracFocus 1.0 Data Analysis Tables and Figures, March 2015
    1. Background
    2. Table/Figure from FracFocus 1.0 Data Analysis
    3. R Code to Reproduce Table ES-1
  11. NEXT

EPA Fracturing Data

Last modified
Table of contents
  1. Story
    1. Data Science for EPA Fracturing Data
    2. Slides
      1. Slide 1 Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0
      2. Slide 2 Outline
      3. Slide 3 Research Questions
      4. Slide 4 Research Products
      5. Slide 5 General Information
      6. Slide 6 Hydraulic Fracturing Water Cycle
      7. Slide 7 Additives
      8. Slide 8 Example FracFocus Disclosure
      9. Slide 9 Hydraulic Fracturing Fluid Composition
      10. Slide 10 Disclosure Locations
      11. Slide 11 Ingredients
      12. Slide 12 Confidential Business Information (CBI) Ingredients
      13. Slide 13 Additive Ingredients
      14. Slide 14 Most Frequently Reported Additive Ingredients for Gas Wells
      15. Slide 15 Most Frequently Reported Additive Ingredients for Oil Wells
      16. Slide 16 Base Fluids
      17. Slide 17 Non-Aqueous Fluids in Base Fluid
      18. Slide 18 Cumulative Water Volume, by State
      19. Slide 19 Per-Disclosure Water Volume, by State
      20. Slide 20 Water Sources
      21. Slide 21 Analysis Summary 1
      22. Slide 22 Analysis Summary 2
      23. Slide 23 State Summaries 1
      24. Slide 24 State Summaries 2
      25. Slide 25 State Summaries 3
      26. Slide 26 Project Database
      27. Slide 27 What Did We Learn?
      28. Slide 28 Questions?
  2. Slides
    1. Slide 1 EPA Big Data Analytics: Data Science for EPA Fracturing Data
    2. Slide 2 Announcement
    3. Slide 3 EPA's Study of Hydraulic Fracturing for Oil and Gas and Its Potential Impact on Drinking Water Resources
    4. Slide 4 The Hydraulic Fracturing Water Cycle
    5. Slide 5 Data Mining for Data Science
    6. Slide 6 Data Science for EPA Fracturing Data: MindTouch Knowledge Base
    7. Slide 7 Data Mining Files and Attachments
    8. Slide 8 Data Science for EPA Fracturing Data: MindTouch Knowledge Base Find
    9. Slide 9 Data Science for EPA Fracturing Data: Spreadsheet Knowledge Base
    10. Slide 10 Data Science for EPA Fracturing Data: Spotfire Cover Page
    11. Slide 11 Data Science for EPA Fracturing Data: Spotfire Data Management Tables
    12. Slide 12 Data Science for EPA Fracturing Data: Spotfire Oil & Gas by County
    13. Slide 13 Data Science for EPA Fracturing Data: Spotfire Oil & Gas by Well by County
    14. Slide 14 Data Science for EPA Fracturing Data: Spotfire Disclosures by County
    15. Slide 15 Data Science for EPA Fracturing Data: Spotfire Ingredients by County
    16. Slide 16 Conclusions and Recommendations
  3. Spotfire Dashboard
  4. Research Notes
  5. EPA's Study of Hydraulic Fracturing for Oil and Gas and Its Potential Impact on Drinking Water Resources
    1. The Hydraulic Fracturing Water Cycle
      1. Stage 1: Water Acquisition 1
      2. Stage 2: Chemical Mixing
      3. Stage 3: Well Injection
      4. Stage 4: Flowback 6 and Produced Water 7 (Hydraulic Fracturing Wastewaters)
      5. Stage 5: Wastewater Treatment and Waste Disposal
  6. Published Scientific Papers
    1. Analysis of Existing Data
    2. Laboratory Studies
      1. Analytical Method Development
    3. Scenario Evaluations
      1. Subsurface Migration Modeling
  7. EPA Analysis of FracFocus 1 Data
  8. Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0
    1. Overview
    2. Background
    3. State Summaries
    4. Results
    5. Study Limitations
  9. Data Management and Quality Assessment Report
    1. Cover Page
    2. Disclaimer
    3. List of Tables
    4. List of Figures
    5. Preface
    6. Acknowledgements
    7. List of Acronyms
    8. 1. Introduction
    9. 2. Source Data
    10. 3. Database Development
      1. 3.1. Downloading and Conversion
        1. Figure 1. Example FracFocus 1.0 disclosure
      2. 3.2. Extraction and Parsing
        1. Table 1 Summary of parsing success
      3. 3.3. Output Data Structure
    11. 4. Assignment of Hydrocarbon Regions to Disclosures
      1. Table Five publicly available datasets with national coverage
    12. 5. Quality Assurance Process for Locational Data
    13. 6. Chemical Name Standardization
    14. 7. Data Field Descriptions
      1. 7.1. Data Fields in Main Tables
        1. 7.1.1. Well Header Field Descriptions
        2. 7.1.2. Ingredient Field Descriptions
      2. 7.2. Data Fields in Tables Associated with Standardizations
        1. 7.2.1. Chemical Name Standardization
        2. 7.2.2. Operator Standardization Information
        3. 7.2.3. Trade Name Standardization
        4. 7.2.4. Ingredient Purpose Standardization
      3. 7.3. Data Fields in Other Tables
        1. 7.3.1. Proppant Identification
        2. 7.3.2. Resin Coating Identification
        3. 7.3.3. CBI Identification
        4. 7.3.4. Water Source Identification
        5. 7.3.5. Purpose Categorization
        6. 7.3.6. State Regulation Information
        7. 7.3.7. County Information
        8. 7.3.8. Water Synonyms
        9. 7.3.9. Unparsed PDFs
    15. 8. Summary
    16. References
    17. Footnotes
      1. 1
      2. 2
      3. 3
      4. 4
      5. 5
      6. 6
      7. 7
      8. 8
      9. 9
  10. Reproducing FracFocus 1.0 Data Analysis Tables and Figures, March 2015
    1. Background
    2. Table/Figure from FracFocus 1.0 Data Analysis
    3. R Code to Reproduce Table ES-1
  11. NEXT

  1. Story
    1. Data Science for EPA Fracturing Data
    2. Slides
      1. Slide 1 Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0
      2. Slide 2 Outline
      3. Slide 3 Research Questions
      4. Slide 4 Research Products
      5. Slide 5 General Information
      6. Slide 6 Hydraulic Fracturing Water Cycle
      7. Slide 7 Additives
      8. Slide 8 Example FracFocus Disclosure
      9. Slide 9 Hydraulic Fracturing Fluid Composition
      10. Slide 10 Disclosure Locations
      11. Slide 11 Ingredients
      12. Slide 12 Confidential Business Information (CBI) Ingredients
      13. Slide 13 Additive Ingredients
      14. Slide 14 Most Frequently Reported Additive Ingredients for Gas Wells
      15. Slide 15 Most Frequently Reported Additive Ingredients for Oil Wells
      16. Slide 16 Base Fluids
      17. Slide 17 Non-Aqueous Fluids in Base Fluid
      18. Slide 18 Cumulative Water Volume, by State
      19. Slide 19 Per-Disclosure Water Volume, by State
      20. Slide 20 Water Sources
      21. Slide 21 Analysis Summary 1
      22. Slide 22 Analysis Summary 2
      23. Slide 23 State Summaries 1
      24. Slide 24 State Summaries 2
      25. Slide 25 State Summaries 3
      26. Slide 26 Project Database
      27. Slide 27 What Did We Learn?
      28. Slide 28 Questions?
  2. Slides
    1. Slide 1 EPA Big Data Analytics: Data Science for EPA Fracturing Data
    2. Slide 2 Announcement
    3. Slide 3 EPA's Study of Hydraulic Fracturing for Oil and Gas and Its Potential Impact on Drinking Water Resources
    4. Slide 4 The Hydraulic Fracturing Water Cycle
    5. Slide 5 Data Mining for Data Science
    6. Slide 6 Data Science for EPA Fracturing Data: MindTouch Knowledge Base
    7. Slide 7 Data Mining Files and Attachments
    8. Slide 8 Data Science for EPA Fracturing Data: MindTouch Knowledge Base Find
    9. Slide 9 Data Science for EPA Fracturing Data: Spreadsheet Knowledge Base
    10. Slide 10 Data Science for EPA Fracturing Data: Spotfire Cover Page
    11. Slide 11 Data Science for EPA Fracturing Data: Spotfire Data Management Tables
    12. Slide 12 Data Science for EPA Fracturing Data: Spotfire Oil & Gas by County
    13. Slide 13 Data Science for EPA Fracturing Data: Spotfire Oil & Gas by Well by County
    14. Slide 14 Data Science for EPA Fracturing Data: Spotfire Disclosures by County
    15. Slide 15 Data Science for EPA Fracturing Data: Spotfire Ingredients by County
    16. Slide 16 Conclusions and Recommendations
  3. Spotfire Dashboard
  4. Research Notes
  5. EPA's Study of Hydraulic Fracturing for Oil and Gas and Its Potential Impact on Drinking Water Resources
    1. The Hydraulic Fracturing Water Cycle
      1. Stage 1: Water Acquisition 1
      2. Stage 2: Chemical Mixing
      3. Stage 3: Well Injection
      4. Stage 4: Flowback 6 and Produced Water 7 (Hydraulic Fracturing Wastewaters)
      5. Stage 5: Wastewater Treatment and Waste Disposal
  6. Published Scientific Papers
    1. Analysis of Existing Data
    2. Laboratory Studies
      1. Analytical Method Development
    3. Scenario Evaluations
      1. Subsurface Migration Modeling
  7. EPA Analysis of FracFocus 1 Data
  8. Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0
    1. Overview
    2. Background
    3. State Summaries
    4. Results
    5. Study Limitations
  9. Data Management and Quality Assessment Report
    1. Cover Page
    2. Disclaimer
    3. List of Tables
    4. List of Figures
    5. Preface
    6. Acknowledgements
    7. List of Acronyms
    8. 1. Introduction
    9. 2. Source Data
    10. 3. Database Development
      1. 3.1. Downloading and Conversion
        1. Figure 1. Example FracFocus 1.0 disclosure
      2. 3.2. Extraction and Parsing
        1. Table 1 Summary of parsing success
      3. 3.3. Output Data Structure
    11. 4. Assignment of Hydrocarbon Regions to Disclosures
      1. Table Five publicly available datasets with national coverage
    12. 5. Quality Assurance Process for Locational Data
    13. 6. Chemical Name Standardization
    14. 7. Data Field Descriptions
      1. 7.1. Data Fields in Main Tables
        1. 7.1.1. Well Header Field Descriptions
        2. 7.1.2. Ingredient Field Descriptions
      2. 7.2. Data Fields in Tables Associated with Standardizations
        1. 7.2.1. Chemical Name Standardization
        2. 7.2.2. Operator Standardization Information
        3. 7.2.3. Trade Name Standardization
        4. 7.2.4. Ingredient Purpose Standardization
      3. 7.3. Data Fields in Other Tables
        1. 7.3.1. Proppant Identification
        2. 7.3.2. Resin Coating Identification
        3. 7.3.3. CBI Identification
        4. 7.3.4. Water Source Identification
        5. 7.3.5. Purpose Categorization
        6. 7.3.6. State Regulation Information
        7. 7.3.7. County Information
        8. 7.3.8. Water Synonyms
        9. 7.3.9. Unparsed PDFs
    15. 8. Summary
    16. References
    17. Footnotes
      1. 1
      2. 2
      3. 3
      4. 4
      5. 5
      6. 6
      7. 7
      8. 8
      9. 9
  10. Reproducing FracFocus 1.0 Data Analysis Tables and Figures, March 2015
    1. Background
    2. Table/Figure from FracFocus 1.0 Data Analysis
    3. R Code to Reproduce Table ES-1
  11. NEXT

Story

Slides and Script for TIBCO Spotfire Webinar, September 1, 2015

Data Science for EPA Fracturing Data

fracking-happening.jpg

The announcement said: The EPA released its peer-reviewed analysis of over two years of data from the FracFocus Chemical Disclosure Registry 1.0. FracFocus is a publicly accessible website, managed by the Ground Water Protection Council (GWPC) and Interstate Oil and Gas Compact Commission, where oil and gas production well operators can disclose information about ingredients used in hydraulic fracturing fluids at individual wells.

On March 2015: U.S. Department of the Interior Releases Final Rule to Support Safe, Responsible Hydraulic Fracturing Activities on Public and Tribal Lands. See Press release and Final rule.

The five stages of The Hydraulic Fracturing Water Cycle are shown below. The newest features are the Published Scientific Papers and  Analysis of Existing Data. The latter were mined for data sets for analytics and visualizations in Spotfire and PDF files to be converted to MindTouch as follows:

  • Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0 - See Attachments below
    • Four xlsx files totaling 23 MB
  • Data Management and Quality Assessment Report - See Attachments below
    • 2 MB PDF and 96 MB ZIP file with Access Database and PDF documentation
  • Project database developed from FracFocus 1.0 disclosures
    • Same as second item above
  • Selected data tables from the analysis of FracFocus 1.0
    • Same as first item above
  • State-level Summaries of FracFocus 1.0 Hydraulic Fracturing Data - See Attachments below
    • 7 MB PDF Guide and 22 PDF files
  • EPA Analysis of FracFocus Data Fact Sheet - See Attachments below
    • 0.375 MB PDF
  • EPA Webinar: Analysis of FracFocus Data - See Slides below
    • 2 MB PDF

The data mining process, tools, and sample results are shown in the Slides below.

This is the second Data Science for EPA Big Data Analytics Use Case. Also see previous Data Science for EPA EnviroAtlas and Data Science for EPA Air Data).

Two more Data Science for EPA Big Data Analytics Uses Cases are:

More Uses Case are in process for an EPA Big Data Analytics MOOC (Massive Open Online Course) for the Federal Big Data Working Group Meetup.

Slides

PDF

Slide 1 Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0

EPAFracturing032014Slide1.png

Slide 2 Outline

EPAFracturing032014Slide2.png

Slide 3 Research Questions

EPAFracturing032014Slide3.png

Slide 4 Research Products

EPAFracturing032014Slide4.png

Slide 5 General Information

EPAFracturing032014Slide5.png

Slide 6 Hydraulic Fracturing Water Cycle

EPAFracturing032014Slide6.png

Slide 7 Additives

EPAFracturing032014Slide7.png

Slide 8 Example FracFocus Disclosure

EPAFracturing032014Slide8.png

Slide 9 Hydraulic Fracturing Fluid Composition

EPAFracturing032014Slide9.png

Slide 10 Disclosure Locations

EPAFracturing032014Slide10.png

Slide 11 Ingredients

EPAFracturing032014Slide11.png

Slide 12 Confidential Business Information (CBI) Ingredients

EPAFracturing032014Slide12.png

Slide 13 Additive Ingredients

EPAFracturing032014Slide13.png

Slide 14 Most Frequently Reported Additive Ingredients for Gas Wells

EPAFracturing032014Slide14.png

Slide 15 Most Frequently Reported Additive Ingredients for Oil Wells

EPAFracturing032014Slide15.png

Slide 16 Base Fluids

EPAFracturing032014Slide16.png

Slide 17 Non-Aqueous Fluids in Base Fluid

EPAFracturing032014Slide17.png

Slide 18 Cumulative Water Volume, by State

EPAFracturing032014Slide18.png

Slide 19 Per-Disclosure Water Volume, by State

EPAFracturing032014Slide19.png

Slide 20 Water Sources

EPAFracturing032014Slide20.png

Slide 21 Analysis Summary 1

EPAFracturing032014Slide21.png

Slide 22 Analysis Summary 2

EPAFracturing032014Slide22.png

Slide 23 State Summaries 1

EPAFracturing032014Slide23.png

Slide 24 State Summaries 2

EPAFracturing032014Slide24.png

Slide 25 State Summaries 3

EPAFracturing032014Slide25.png

Slide 26 Project Database

EPAFracturing032014Slide26.png

Slide 27 What Did We Learn?

EPAFracturing032014Slide27.png

Slide 28 Questions?

EPAFracturing032014Slide28.png

Slides

Slides

Slide 3 EPA's Study of Hydraulic Fracturing for Oil and Gas and Its Potential Impact on Drinking Water Resources

http://www2.epa.gov/hfstudy

BrandNiemann04172015Slide3.PNG

Slide 4 The Hydraulic Fracturing Water Cycle

http://www2.epa.gov/hfstudy/hydrauli...ng-water-cycle

BrandNiemann04172015Slide4.PNG

Slide 5 Data Mining for Data Science

http://www2.epa.gov/hfstudy/epa-anal...acfocus-1-data

BrandNiemann04172015Slide5.PNG

Slide 6 Data Science for EPA Fracturing Data: MindTouch Knowledge Base

http://semanticommunity.info/Data_Sc...Data_Analytics
http://semanticommunity.info/Data_Sc...racturing_Data

BrandNiemann04172015Slide6.PNG

Slide 7 Data Mining Files and Attachments

BrandNiemann04172015Slide7.PNG

Slide 8 Data Science for EPA Fracturing Data: MindTouch Knowledge Base Find

http://semanticommunity.info/Data_Sc...Data_Analytics
http://semanticommunity.info/Data_Sc...racturing_Data

BrandNiemann04172015Slide8.PNG

Slide 9 Data Science for EPA Fracturing Data: Spreadsheet Knowledge Base

http://semanticommunity.info/%40api/...?origin=mt-web

BrandNiemann04172015Slide9.PNG

Slide 10 Data Science for EPA Fracturing Data: Spotfire Cover Page

Web Player

BrandNiemann04172015Slide10.PNG

Slide 11 Data Science for EPA Fracturing Data: Spotfire Data Management Tables

Web Player

BrandNiemann04172015Slide11.PNG

Slide 12 Data Science for EPA Fracturing Data: Spotfire Oil & Gas by County

Web Player

BrandNiemann04172015Slide12.PNG

Slide 13 Data Science for EPA Fracturing Data: Spotfire Oil & Gas by Well by County

Web Player

BrandNiemann04172015Slide13.PNG

Slide 14 Data Science for EPA Fracturing Data: Spotfire Disclosures by County

Web Player

BrandNiemann04172015Slide14.PNG

Slide 15 Data Science for EPA Fracturing Data: Spotfire Ingredients by County

Web Player

BrandNiemann04172015Slide15.PNG

Slide 16 Conclusions and Recommendations

BrandNiemann04172015Slide16.PNG

Spotfire Dashboard

For Internet Explorer Users and Those Wanting Full Screen Display Use: Web Player Get Spotfire for iPad App

Error: Embedded data could not be displayed. Use Google Chrome

Research Notes

EPA's Study of Hydraulic Fracturing for Oil and Gas and Its Potential Impact on Drinking Water Resources

Source: http://www2.epa.gov/hfstudy

At the request of Congress, EPA is conducting a study to better understand any potential impacts of hydraulic fracturing for oil and gas on drinking water resources. The scope of the research includes the full lifespan of water in hydraulic fracturing. 

The Hydraulic Fracturing Water Cycle

Source: http://www2.epa.gov/hfstudy/hydrauli...ng-water-cycle

EPA's study will look at potential impacts of hydraulic fracturing at each stage of the Hydraulic Fracturing Water Cycle.

Click on the image below to learn more about the Hydraulic Fracturing Water Cycle.

The hydraulic fracturing water cycle.  Please click on the image for an explanation of each step.1. Water AcquisitionNatural gas flows from fissures into well.2. Chemical Mixing3. Well Injection4. Flowback and Produced Water (Wastewater)5. Wastewater Treatment and Waste Disposal
 

Stage 1: Water Acquisition 1

  • Large volumes of water are withdrawn from ground water 2 and surface water 3 resources to be used in the hydraulic fracturing process.
  • Potential Impacts on Drinking Water Resources
    • Change in the quantity of water available for drinking
    • Change in drinking water quality

    Recently, some companies have begun recycling wastewater from previous hydraulic fracturing activities, rather than acquiring water from ground or surface resources.

    Ground water is the supply of fresh water found beneath the Earth’s surface, usually in aquifers, which supply wells and springs. It provides a major source of drinking water.

    Surface water resources include any water naturally open to the atmosphere, such as rivers, lakes, reservoirs, ponds, streams, impoundments, seas, estuaries, etc. It provides a major source of drinking water.

Stage 2: Chemical Mixing

  • Once delivered to the well site, the acquired water is combined with chemical additives 4 and proppant 5 to make the hydraulic fracturing fluid.
  • Potential Impacts on Drinking Water Resources
    • Release to surface and ground water through on-site spills and/or leaks

    Chemical additives are used for a variety of purposes (see examples in Table 4 on page 29 of the hydraulic racturing Study Plan (PDF)). A list of publicly known chemical additives found in hydraulic fracturing fluids is provided in Appendix E, Table E1 of the hydraulic fracturing Study Plan (PDF) .

    Proppant is a granular substance such as sand that is used to keep the underground cracks open once the hydraulic fracturing fluid is withdrawn.

Stage 3: Well Injection

  • Pressurized hydraulic fracturing fluid is injected into the well, creating cracks in the geological formation that allow oil or gas to escape through the well to be collected at the surface.
  • Potential Impacts on Drinking Water Resources
    • Release of hydraulic fracturing fluids to ground water due to inadequate well construction or operation
    • Movement of hydraulic fracturing fluids from the target formation to drinking water aquifers through local man-made or natural features (e.g., abandoned wells and existing faults)
    • Movement into drinking water aquifers of natural substances found underground, such as metals or radioactive materials, which are mobilized during hydraulic fracturing activities

Stage 4: Flowback and Produced Water (Hydraulic Fracturing Wastewaters)

  • When pressure in the well is released, hydraulic fracturing fluid, formation water, and natural gas begin to flow back up the well. This combination of fluids, containing hydraulic fracturing chemical additives and naturally occurring substances, must be stored on-site—typically in tanks or pits—before treatment, recycling, or disposal.
  • Potential Impacts on Drinking Water Resources
    • Release to surface or ground water through spills or leakage from on-site storage

    After the hydraulic fracturing procedure is completed and pressure is released, the direction of fluid flow reverses, and water and excess proppant flow up through the wellbore to the surface. The water that returns to the surface is commonly referred to as “flowback.”

    After the drilling and fracturing of the well are completed, water is produced along with the natural gas. Some of this water is returned fracturing fluid and some is natural formation water. These produced waters move back through the wellhead with the gas.

Stage 5: Wastewater Treatment and Waste Disposal

  • Wastewater is dealt with in one of several ways, including but not limited to: disposal by underground injection, treatment followed by disposal to surface water bodies, or recycling (with or without treatment) for use in future hydraulic fracturing operations.
  • Potential Impacts on Drinking Water Resources
    • Contaminants reaching drinking water due to surface water discharge and inadequate treatment of wastewater
    • Byproducts formed at drinking water treatment facilities by reaction of hydraulic fracturing contaminants with disinfectants

Published Scientific Papers

Source: http://www2.epa.gov/hfstudy/publishe...entific-papers

The study's individual research projects will be published as peer reviewed journal articles or as EPA reports that undergo peer review by external technical experts. The papers will support the draft assessment report on the potential impacts of hydraulic fracturing for oil and gas on drinking water resources. The following is a list of papers published to date. Additional papers will be posted as they are published.

Analysis of Existing Data

Scenario Evaluations

Subsurface Migration Modeling

EPA Analysis of FracFocus 1 Data

Source: http://www2.epa.gov/hfstudy/epa-anal...acfocus-1-data

The EPA conducted an analysis to better understand the chemicals and water used to hydraulically fracture oil and gas production wells in the United States and how chemical and water use vary in different locations across the country.

The EPA compiled and analyzed over two years of data from the FracFocus Chemical Disclosure Registry 1.0 provided by the Ground Water Protection Council (GWPC) and the Interstate Oil and Gas Compact Commission (IOGCC). FracFocus Exitis a publicly accessible website managed by GWPC and IOGCC where oil and gas production well operators can disclose information about ingredients used in hydraulic fracturing fluids at individual wells.

This work was done as part of the EPA’s Study of the Potential Impacts of Hydraulic Fracturing for Oil and Gas on Drinking Water Resources.

Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0

Source: http://www2.epa.gov/sites/production...e_fracfocu.pdf (PDF)

Overview of the EPA’s Study of the Potential Impacts of Hydraulic Fracturing for Oil and Gas on Drinking Water Resources

Overview

The EPA is conducting an assessment of the potential impacts of oil and gas hydraulic fracturing activities on the quality and quantity of drinking water resources in the United States. Expected in spring 2015, it will provide much needed information to states, industry, and most importantly our communities, to safeguard our water resources and protect public health.

The EPA’s assessment is based upon extensive review of literature, results from EPA research projects, and technical input from state, industry, non-governmental organizations, the public, and other stakeholders. Ultimately, it will help advance the state of our science and provide a new lens to help our states and communities understand the potential impacts on our drinking water resources from hydraulic fracturing. Part of this effort includes analyzing data from the FracFocus Chemical Disclosure Registry 1.0 to better understand the chemicals and water used to hydraulically fracture oil and gas production wells in the United States.

Background

FracFocus is a publicly accessible website (http://www.fracfocus.org) managed by the Ground Water Protection Council (GWPC) and the Interstate Oil and Gas Compact Commission (IOGCC). Oil and gas production well operators can disclose information at this website about ingredients and water used in hydraulic fracturing fluids at individual wells. The GWPC and IOGCC provided the EPA with over 39,000 PDF disclosures submitted by well operators to FracFocus 1.0 before March 1, 2013. The disclosures identified 20 states with reported well locations that were hydraulically fractured during the study period. Data in the disclosures were extracted from individual PDF files and compiled in a project database, which was used to conduct analyses on chemical and water use for hydraulic fracturing. Analyses were conducted on over 38,000 unique disclosures for wells hydraulically fractured between January 1, 2011, and February 28, 2013. The Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0 summarizes chemical and water use data and looks at how chemical and water use vary in different locations across the United States.

In addition to the Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0, the EPA is releasing the project database developed from FracFocus 1.0 disclosures, Microsoft Excel spreadsheets with the datasets that were used in the analysis, user guides, and state-level summaries for the well locations represented in FracFocus 1.0 for the period of study. The accompanying Data Management and Quality Assessment Report describes the structure of the database, data fields, and the quality assessment of the data. The database and the Data Management and Quality Assessment Report can be used to better understand the EPA’s methods or to perform additional data analyses.

State Summaries

The EPA is releasing state-level summaries of the chemicals and water used for hydraulic fracturing in the 20 states represented during the time of this study. The state-level summaries and a User’s Guide that describes the information available in the summaries are available at www2.epa.gov/hfstudy/published-scientific-papers. The information found in the state summaries reflects data found in the FracFocus 1.0 disclosures and information from publicly available sources. Neither the states nor GWPC verified the data shown in the summaries. The information and figures presented in the state summaries may differ from information held by the states.) The 20 states are:

  • Alabama
  • Alaska
  • Arkansas
  • California
  • Colorado
  • Kansas
  • Louisiana
  • Michigan
  • Mississippi
  • Montana
  • New Mexico
  • North Dakota
  • Ohio
  • Oklahoma
  • Pennsylvania
  • Texas
  • Utah
  • Virginia
  • West Virginia
  • Wyoming

Results

Hydraulic fracturing fluids were generally found to contain 88% (by mass) water, 10% quartz used as proppant, and <1% additive ingredients. 698 unique ingredients (i.e., chemicals) were reported by 428 operators in 20 states. The median number of additive ingredients per disclosure was 14. Hydrochloric acid, methanol, and hydro-treated light petroleum distillates were reported in more than 65% of all disclosures analyzed. Seventy percent of the disclosures analyzed included at least one ingredient that was claimed to be confidential business information (CBI), and 11% of the ingredient records were identified as CBI.

In more than 93% of the disclosures, water was used as the base fluid, with reported volumes ranging from 30,000 to 7.2 million gallons per disclosure. Information related to water sources was reported in 29% of the disclosures. Some of these terms indicate a condition of water quality, such as “fresh,” rather than a specific identification of the source of the water such as ground water or surface water. The most commonly reported source of water used for base fluid was listed as “fresh” (68% of disclosures with water source information).

Study Limitations

Despite the challenge of adapting a dataset originally created for local use and single-PDF viewing to answer broader questions, the project database provided substantial insight into water and chemical use for hydraulic fracturing. All data that met the project’s quality assurance criteria were included in the analyses. The project database represents the data reported to FracFocus 1.0 rather than all hydraulic fracturing that occurred in the United States during the study time period. The project database is an incomplete picture of all hydraulic fracturing due to the voluntary reporting in some states for certain time periods (in the absence of state reporting requirements), the omission of information on CBI ingredients from disclosures, and invalid or erroneous information created during the development of the database or found in the original disclosures. The development of FracFocus 2.0, which became the exclusive reporting mechanism in June 2013, was intended to increase the quality, completeness, and consistency of data submitted by operators by providing dropdown menus, warning and errors messages during submission, and automatic formatting of certain fields. GWPC has announced additional changes and upgrades for FracFocus 3.0 to enhance data searchability, increase system security, provide greater data accuracy, and further increase data transparency.

For more information, please visit http://www.epa.gov/hfstudy

Contact: Dayna Gibbons, Office of Research and Development, gibbons.dayna@epa.gov

Data Management and Quality Assessment Report

Source: http://www2.epa.gov/sites/production...032015_508.pdf (PDF)

Cover Page

Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0: Data Management and Quality Assessment Report

U.S. Environmental Protection Agency Office of Research and Development Washington, DC

March 2015 EPA/601/R-14/006

Disclaimer

This document has been reviewed in accordance with U.S. Environmental Protection Agency policy and approved for publication. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

Preferred Citation: U.S. Environmental Protection Agency. 2015. Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0: Data Management and Quality Assessment Report. Office of Research and Development, Washington, DC. EPA/601/R-14/006.

Preface

The U.S. Environmental Protection Agency (EPA) is conducting a Study of the Potential Impacts of Hydraulic Fracturing for Oil and Gas on Drinking Water Resources. The study is based upon an extensive review of the literature; results from EPA research projects; and technical input from state, industry, and non-governmental organizations, as well as the public and other stakeholders. A series of technical roundtables and in-depth technical workshops were held to help address specific research questions and to inform the work of the study.

In Fiscal Year 2010, Congress urged the EPA to examine the relationship between hydraulic fracturing and drinking water resources in the United States. The EPA’s Plan to Study the Potential Impacts of Hydraulic Fracturing on Drinking Water Resources was reviewed by the agency’s Science Advisory Board (SAB) and issued in 2011. The Study of the Potential Impacts of Hydraulic Fracturing on Drinking Water Resources: Progress Report, detailing the EPA’s research approaches and next steps, was released in late 2012 and followed by a consultation with individual experts convened under the auspices of the SAB.

This report, Evaluation of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0: Data Management and Quality Assessment Report, is the product of one of the research projects conducted as part of the EPA’s study. It has undergone independent, external peer review, which was conducted through the Eastern Research Group, Inc. All peer review comments were considered in the report’s development. The report has also been reviewed in accordance with agency policy and approved for publication.

The EPA is writing a state-of-the-science assessment that integrates a broad review of existing literature, results from peer-reviewed EPA research products (including this report), and information gathered through stakeholder engagement efforts to answer the fundamental research questions posed for each stage of the hydraulic fracturing water cycle:

  • Water Acquisition: What are the possible impacts of large volume water withdrawals from ground and surface waters on drinking water resources?
  • Chemical Mixing: What are the possible impacts of surface spills on or near well pads of hydraulic fracturing fluids on drinking water resources?
  • Well Injection: What are the possible impacts of the injection and fracturing process on drinking water resources?
  • Flowback and Produced Water: What are the possible impacts of surface spills on or near well pads of flowback and produced water on drinking water resources?
  • Wastewater Treatment and Waste Disposal: What are the possible impacts of inadequate treatment of hydraulic fracturing wastewaters on drinking water resources?

The state-of-the-science assessment is not a human health or an exposure assessment, nor is it designed to evaluate policy options or best management practices. As a Highly Influential Scientific Assessment, the draft assessment report will undergo public comment and a meaningful and timely peer review by the SAB to ensure all information is high quality.

Acknowledgements

The EPA would like to acknowledge the Ground Water Protection Council and the Interstate Oil and Gas Compact Commission for providing data and information for this report. Assistance was provided by The Cadmus Group, Inc., under contract EP-C-08-015. The contractor’s role did not include establishing agency policy.

List of Acronyms

API American Petroleum Institute

CASRN Chemical Abstracts Service Registry Number

CBI Confidential Business Information

CSV Comma-Separated Values

EIA U.S. Energy Information Administration

EPA U.S. Environmental Protection Agency

FIPS Federal Information Processing Standards

GIS Geographic Information System

GWPC Ground Water Protection Council

ID Identification

IOGCC Interstate Oil and Gas Compact Commission

NAD North American Datum

PDF Portable Document Format

QA Quality Assurance

TVD True Vertical Depth

USGS U.S. Geologic Survey

WGS World Geodetic System

XML Extensible Markup Language

1. Introduction

This report describes the procedures used to develop a database from data submitted to the FracFocus Chemical Disclosure Registry (subsequently referred to as “FracFocus”) by well operators. The resulting project database was used to conduct the analyses described in the Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0 (subsequently referred to as the “data analysis report;” US EPA, 2015). 1 This data management report can be used in conjunction with the project database and data analysis report to reproduce the results presented in the data analysis report and to conduct additional analyses, if desired.

2. Source Data

FracFocus is a publicly accessible website (www.fracfocus.org) managed by the Ground Water Protection Council (GWPC) and the Interstate Oil and Compact Commission (IOGCC) where oil and gas production well operators can disclose information about the composition of hydraulic fracturing fluids at individual wells. 2 Disclosures included in the project database were submitted to FracFocus by well operators using the FracFocus 1.0 format and were provided in portable document format (PDF) to the U.S. Environmental Protection Agency (EPA) by the GWPC in March 2013. 3 The PDF files were converted to Extensible Markup Language (XML) and parsed into a Microsoft Access database (Microsoft Corporation, 2012). Reviews of data quality were conducted on the project database to ensure that the results from analyses of the project database reflect the data contained in the original PDF disclosures, while identifying obviously invalid or incorrect data to exclude from analyses.

The source data provided by the GWPC were a bulk archive of 39,136 disclosures in PDF format that were submitted to the FracFocus 1.0 website prior to March 1, 2013. Each disclosure was initially submitted by the well operator to FracFocus in the form of a Microsoft Excel spreadsheet and contained information on one production well that was hydraulically fractured with a single fracture date. Each Excel spreadsheet was then converted into a PDF file by the FracFocus website.

3. Database Development

The initial development of the project database involved data conversion of disclosures from PDF format to XML files, parsing to extract information, and incorporation of the resulting data into a Microsoft Access database. The subsequent steps to conduct quality assurance (QA) and the resulting tables and fields that are suitable for data analysis are described in Sections 4, 5, 6, and 7. In describing the database development in this report, underline formatting denotes table names, bold formatting denotes field names, and italic formatting denotes data values.

3.1. Downloading and Conversion

The GWPC prepared a complete archive of all FracFocus 1.0 PDF disclosures (files) uploaded through February 28, 2013, and transferred the archive to the EPA. Adobe Acrobat Pro X (Adobe Systems Incorporated, 2011) was then used to convert all 39,136 PDF files in the archive to XML 2003 spreadsheet (Microsoft Excel 2003 XML) files. The conversion was performed because it is inherently difficult to extract data from PDF files, which are intended to provide consistent visual presentation across devices rather than structured representation of data for parsing and extraction. Tables of information in PDF files, in particular, can present a challenge for conversion. The source Microsoft Excel files, as uploaded by the operators, contained data in tables. However, in a PDF file, a table is essentially a series of lines and characters positioned on a page that, when assembled by PDF-reading software, appear as a table to the end user. To obtain tabular information from a PDF file, the PDF was converted to XML file format, which allows discrete data to be sorted into specific fields so that the data can be manipulated during analysis.

Each FracFocus 1.0 disclosure contains two tables of information. Figure 1 shows an example of an individual well disclosure available to the public as a PDF. At the top of each disclosure is the well header table (outlined in blue in Figure 1), which contains the fracture date, well identifiers [i.e., American Petroleum Institute (API) number and well name], locational data, production type, true vertical depth (TVD) of the well, and the total water volume used to hydraulically fracture the well. 4

Figure 1. Example FracFocus 1.0 disclosure

EPAFracturingDataManagementFigure1.png

The ingredients table (outlined in red in Figure 1) provides information on the trade names of the additives used in the hydraulic fracturing fluids, the supplier, and additive purpose. Each additive contains one or more ingredients, and the ingredients table includes the chemical name and Chemical Abstracts Service Registry Number (CASRN) for each ingredient, as well as the maximum concentrations as a percentage by mass in the additive and in the hydraulic fracturing fluid.

3.2. Extraction and Parsing

A script was used to read the XML files, parse the relevant data, and compile those data into a useable format. The parsing script was written in Python 2.7 (Python Software Foundation, 2012) and uses the Beautiful Soup 4 library (Richardson, 2013) to read the XML files.

The script first locates and extracts the well header information for a given file. Generally, the fracture date appears first in a PDF, followed by other parameters in order. The script locates the first cell in the file that is of cell type “DateTime.” 5 The script then reads the columns below the date with the assumption that the other well header fields are ordered as anticipated from the disclosure template provided to well operators. In some cases, text wrapping in the original PDFs will split values into multiple rows, resulting in extra header cells. To address this, the position of the longitude field, which is always a negative number for locations within the United States, is used as a “landmark” to recalibrate the ordering of data fields.

The script parses information from the ingredients table by locating individual columns of information and then reading cells in that column until the bottom of the table is reached. The bottom of the table is either the last row with more than one cell or the last row in the sheet. Columns are located by searching for text patterns that indicate the presence of a column header. In developing the script, the text patterns were refined based on experience; some operators represent the same column of information differently. For the data fields Purpose and Trade Name in the ingredients table of the disclosure (Figure 1), operators generally enter a value once to indicate that an additive trade name or purpose applies to all ingredients that follow (e.g., additive “Plexgel 907L-EB” in ingredients table of Figure 1). Thus, a purpose and trade name are applied to ingredients until a new trade name and purpose are encountered. Blank values in the purpose and trade name columns are replaced with the previous value as the column is parsed.

The parsing approach is highly sensitive to formatting. If an operator departed from the FracFocus 1.0 template when originally creating a disclosure, the disclosure may have been skipped or information from the disclosure may have parsed incorrectly. Most of the disclosures were prepared in a consistent format that enabled relatively easy parsing of data. However, some disclosures were uploaded using templates modified by the operators, with columns added or removed, fields left blank, or invalid data entered. The modified disclosures were problematic during parsing and QA.

Table 1 Summary of parsing success
Well header parsed Ingredient table parsed Number of disclosures Percentage of disclosures
Yes or No Yes or No 39,136 100%
Yes Yes or No 38,530 98.5%
Yes Yes 37,017 94.6%
Yes No 1,513 3.87%
No No 606 1.55%

Note: “Yes” and “No” indicate whether portions of the disclosures (well header or ingredient table) were successfully parsed. “Yes or no” indicates that the disclosure counts include disclosures that were parsed and those that were not.

As shown in Table 1, the well header table was successfully parsed from 98.5% of disclosures (38,530 of 39,136), and both the well header and ingredient tables were successfully parsed from 94.6% of disclosures (37,017 of 39,136).

3.3. Output Data Structure

The script parsed the resulting data into two comma-separated value (CSV) files that form the foundation of the project database. One file contains the well operator, well identifiers, production, and locational data from the well header; the other file contains the additive, additive purpose, chemical, and chemical concentration data from the ingredients table. The two-table structure was considered appropriate because a one-to-many relationship exists between the well header values for an individual disclosure and the multiple values from the ingredients table that correspond to that disclosure. The two tables are linked in the project database by a constructed unique identification (ID) field. The ID field is necessary because the combinations of API Well Number and Fracture Date for 228 disclosures were found to be duplicated in the dataset and, thus, cannot serve as unique identifiers. Unique disclosures—defined by the combination of API Well Number and Fracture Date—were selected from duplicate disclosures by choosing the file with the most recent modification date. The modification date associated with each PDF is not information found on the publicly available disclosure that may be downloaded from FracFocus. If two or more records shared the same values for API Well Number and Fracture Date, then the PDF file with the most recent modification date was flagged as the authoritative disclosure.

To maximize the transparency of the QA effort, the final database contains two versions of the data extracted from the FracFocus 1.0 disclosures. The first version contains data as originally parsed without any formatting, spelling corrections, or standardization—these tables are denoted with the “Original” prefix in their names. The values in these tables were taken directly from the CSV files produced by the parsing script and are stored verbatim as text. The second version contains data after formatting, corrections, and standardization were performed—these tables are denoted with the “Qa” prefix. The “Qa” tables also contain fields describing the adjustments made to each disclosure and whether the values met QA criteria. The two-version structure enabled straightforward review of all changes and streamlined tracing of disclosures back to the source data.

The primary tables in the project database are as follows:

  • OriginalWell. Well header data with verbatim (unadjusted) values as parsed to input data.
  • QaWell. Well header data with minor adjustments applied, including fixed typographical errors, removal of extraneous characters, and corrections of obvious transpositions (e.g., latitude and longitude swapped, state and county swapped). Columns accompanying each set of well header values, also referred to as QA flag fields, describe adjustments made to the OriginalWell data and whether the data met QA criteria as included in the QaWell table.
  • OriginalIngredient. Ingredients data with verbatim (unadjusted) values as parsed to input data.
  • QaIngredient. Ingredients data with minor adjustments applied, including corrected formatting of CASRNs and standardized suppliers. Similar to the table QaWell, the QaIngredient table includes QA flag fields that describe the adjustments made and whether the data met QA criteria for inclusion in analyses.

Additional tables in the database supporting the QA efforts and data analyses include the following:

  • IngredientNameStandardization. Ingredient names were standardized using a list of chemical names paired with CASRNs compiled by the EPA. These standardized names are used in the QaIngredient table.
  • PurposeStandardization. Additive purpose names were standardized and applied to the QaIngredient table to correct for spelling capitalization, spaces, and punctuation for most purpose entries. Synonyms for proppants and base fluids are also identified in this table.
  • PurposeCategorization. Categorization of related additive purposes was applied to the standardized purposes for ease of summarizing the data during analyses. Information from this table was used for queries in which summary information was compiled regarding additive purposes.
  • TradeNameStandardization. Standardized additive trade names were applied to values in the TradeName field to correct for spelling, capitalization, spaces, and punctuation and are used in the QaIngredient table.
  • OperatorStandardization. Standardized operator names were applied to values in the Operator data to consolidate different representations of operator names and are used in the QaIngredient table.
  • StateRegulation. This table lists effective dates for state laws that either mandate disclosure of hydraulic fracturing chemicals to FracFocus, allow FracFocus as an alternative to reporting to state agencies, or require reporting to state agencies. (This information was obtained through separate research and is not information reported by operators to FracFocus.)
  • Counties. This table provides a listing of all counties in the United States by state, name, and Federal Information Processing Standards (FIPS) code. This table also includes a separate identifier for the five case study counties included in the data analysis report.
  • CBISynonym. A list was compiled of terms interpreted to indicate confidential business information (CBI) in the Chemical Name and Cas fields of ingredient records. This table was used for analyses of ingredient data reported as CBI or an associated term (such as ‘proprietary,’ ‘trade secret,’ etc.).
  • Proppants. This table provides a listing of solid materials associated with proppant-related additive purposes and indicates whether these materials should be excluded from additive ingredient analyses conducted for the data analysis report. 6 The table is not associated with any changes or standardizations in the QaIngredient table, but was referenced in queries for chemicals.
  • ResinCoating. This list contains ingredients associated with proppant-related additive purposes; these are ingredients that are not minerals, but rather chemicals associated with resin coatings on proppants. The list was referenced in queries for the proppants and additive ingredients analyses discussed in the data analysis report and is not associated with any changes or standardizations in the QaIngredient table.
  • WaterSourceTerm. This list of terms is interpreted to indicate water sources reported by operators in the TradeName and Comments fields that are included in the QaIngredient table. These terms were used for the water source analysis described in the data analysis report.
  • UnparsedPDFs. This table lists the PDFs that were unable to be parsed. It is incorporated for transparency and reference.
  • WaterSynonyms. This list contains variations of operator entries (e.g., in the TradeName, Comments, or ChemicalName fields in QaIngredient) that indicate water but no other descriptors for the water source for base fluids. This list was used in querying for water sources. An ingredient record could match a term on this list only if it did not already match a term in WaterSourceTerm.

Section 7 describes the specific data fields found in these tables. Sections 4, 5, and 6, respectively, discuss the incorporation of geospatial data into the database, the QA procedures for well locational data, and the standardization of chemical names.

4. Assignment of Hydrocarbon Regions to Disclosures

Operators reported the production type (oil or gas) on FracFocus 1.0 disclosures, but not the specific producing formation. To offer basic geologic context for the locations of the disclosures, the hydrocarbon regions underlying each disclosure’s latitude and longitude coordinates were added to the QaWell table after conversion of the coordinates to the North American Datum 83 (NAD83) in Esri ArcGIS v. 10.1 geographic information system (GIS; Esri, 2012).

National-scale spatial data describing the areal extent of hydrocarbon regions are limited—local and regional studies are more common. Five publicly available datasets with national coverage were chosen to be spatially joined to well locations. The National Oil and Gas Assessment province boundaries shapefile was obtained from the U.S. Geological Survey (USGS; USGS, 1995), and shapefiles for coalbed methane basins, tight gas basins, and shale gas plays and basins were obtained from the U.S. Energy Information Administration (EIA; US EIA, 2007, 2011a, b). These datasets were used for general reference purposes and with the understanding that the boundaries are approximate and that production may not be occurring from the co-located pla y. The following text boxes describe the content of these databases and provide links to metadata and file download locations.

Table Five publicly available datasets with national coverage

Topic Field name USGSProvinces
USGS Oil and Gas Provinces Description This dataset includes 71 very large oil and gas provinces delineated as part of the USGS’s 1995 National Oil and Gas Assessment (USGS, 1995). Although this layer has coarse spatial resolution, it has the advantage of covering the entire lower 48 states plus Alaska, which means that (nearly) every disclosure in the project database will be located within a province.
  Metadata http://certmapper.cr.usgs.gov/geopor...B3E1E6B1CD9%7D
  Download http://certmapper.cr.usgs.gov/data/n...e/pr_natlg.zip
EIA Shale Basins Field name ShaleBasin
  Description This dataset includes 32 major sedimentary basins that contain hydrocarbon-bearing shales and correspond to the translucent pink “Basins” in the EIA “Lower 48 States Shale Plays” map.
  Metadata http://www.eia.gov/pub/oil_gas/natur.../maps/maps.htm
  Download http://www.eia.gov/pub/oil_gas/natur...legasbasin.zip
EIA Shale Plays Field name ShalePlay
  Description This dataset includes 45 shale plays that correspond to the translucent orange “Current Plays” and yellow “Prospective Plays” in the EIA “Lower 48 States Shale Plays” map.
  Metadata http://www.eia.gov/pub/oil_gas/natur.../maps/maps.htm
  Download http://www.eia.gov/pub/oil_gas/natur...alegasplay.zip
EIA Tight Gas Basins Field name TightGas
  Description This dataset includes 13 sedimentary basins that contain tight gas formations and correspond to the translucent pink “Basins” in the “Major Tight Gas Plays, Lower 48 States” map.
  Metadata http://www.eia.gov/pub/oil_gas/natur.../maps/maps.htm
  Download http://www.eia.gov/pub/oil_gas/natur...sbasinplay.zip
  File in ZIP archive: TightGasBasins_EIA_June2010.shp
EIA Coalbed Methane Basins Field name CoalBed
  Description This dataset includes 98 sedimentary basins that contain coalbed methane and correspond to the translucent pink “Coal Basins, Regions & Fields” in the “Coalbed Methane Fields, Lower 48 States” map.
  Metadata http://www.eia.gov/pub/oil_gas/natur.../maps/maps.htm
  Download http://www.eia.gov/pub/oil_gas/natur.../cbm_4shps.zip
  File in ZIP archive: CBMbasins_Reserv06_Prod06.shp

ArcGIS 10.1 software was used for the spatial join process. The ArcGIS for Desktop Basic license includes the “Spatial Join” geoprocessing tool, which is routinely used to link the attributes of multiple sets of spatial data. In this case, the hydrocarbon regions were “join features,” and the disclosure locations were the “target features.” The disclosure locations were determined by the latitude and longitude coordinates in the project database (after QA and conversion to NAD83 datum, as described in Section 5), corresponding to the NAD83_Lon and NAD83_Lat fields in the QaWell table of the database. The “Join Operation” parameter was “JOIN_ONE_TO_ONE” and the “Match Option” parameter was set to “INTERSECT,” such that a disclosure must spatially intersect the join feature in order to be assigned its value.

The assignment of a hydrocarbon region to a disclosure record in the database is meant to give context to the disclosure location and is likely to be more reliable at the basin scale than at the play scale. Interpretations of the analysis results do not assume that the wells at the disclosure locations are producing from any of the co-located shale plays assigned by this spatial join. Another limitation in accurately assigning plays to the disclosure locations is that the EIA geospatial data do not include boundaries for tight sand plays or coalbed plays; only basin boundaries are available from EIA for these two types of unconventional plays. Therefore, in areas with stacked plays that include sands or coalbeds in addition to shales, it is not possible to determine whether the producing formation is a shale play or another formation based solely on the locational data and the spatial join. Also, comparable EIA geospatial data were not available for oil basins.

For 4,644 disclosures (12% of 38,530 disclosures), the disclosure locations were within the surface boundaries of two EIA shale plays (i.e., plays with active production that are at different depths in the same general surface area, also known as “stacked plays”). Because operators do not report the play or formation that is being hydraulically fractured, there is ambiguity regarding the appropriate formation for the disclosure. Although operators provided TVDs, it is unknown if some of these values may include lateral lengths. Given the limitations of the TVD data, they were not used to interpret formation in regions with stacked plays in cases of shale play overlap for a location. Therefore, the value assigned to the ShalePlay field of the QaWell table is a combination of the individual shale play names, delimited by forward slashes (e.g., Avalon-Bone Spring/Barnett-Woodford.

Arthur et al. (2014) and Carter et al. (2013) summarized data from FracFocus by plays by assuming that the geographic placement of disclosures approximated the geologic placement in popular production plays. Before using the same strategy to categorize results in the data analysis report, the accuracy of geospatial information in identifying plays associated with disclosures was assessed. The results of the spatial join were compared with analogous information from the commercial database DrillingInfo (DrillingInfo, 2011). Because the EIA geospatial data used for the spatial join included play-level boundaries for shales but not for tight sands or coalbeds (these were only delineated at the basin level), the comparison was limited to shales. DrillingInfo is populated using state databases and includes information on producing formations. It includes API well numbers that correspond to 7,761 disclosures in the project database. Of the 7,761 disclosures, 7,153 are co-located with the EIA boundaries for shale plays. Among these 7,153 disclosures, 83% had EIA shale play designations generally consistent with the operator-identified formations in DrillingInfo. Among the 17% of disclosures for which the EIA shale plays did not match the DrillingInfo formations, the mismatches generally occurred where there are stacked plays that include shales in addition to tight sands or coalbeds, and the producing formation is a sandstone, limestone, or coal-bearing formation.

At this time, the basin designations provide useful context for the project database, but shale play designations should be regarded with care in areas with stacked producing plays. Ultimately, the data were not summarized by play in the data analysis report to be consistent with the analysis of the data “as is.”

5. Quality Assurance Process for Locational Data

The well header table in each disclosure includes three sources of locational data:

  • State name and county name information, as stored in the StateFFQA and CountyFFQA fields, respectively, of the QaWell table.
  • State and county information encoded in the first five digits of the API Well Number, as stored in the APIFFQA field of the QaWell table.
  • Latitude and longitude coordinates in the well header, as stored in the LatitudeFFQA and LongitudeFFQA fields, respectively, of the QaWell table. The datum of the coordinates is stored in the ProjectionFFQA field of the QaWell table.

Because the three locational sources were easily available and comparable, the location was determined to have met QA criteria if all three locational data fields agreed. 7

To validate the location of each disclosure, the state and county entries for each of these three fields were compared. First, the leading five digits from APIFFQA were converted to state and county names using lookup tables from the Society of Petrophysicists and Well Log Analysts (2010). Second, the states and counties that intersect the coordinates reported in the LatitudeFFQA and LongitudeFFQA fields were determined using ESRI ArcGIS 10.1 software. Due to the varying datums entered in the ProjectionFFQA field, four separate shapefiles were created:

  • Disclosures with a NAD83 projection were read into a point shapefile with the North American Datum of 1983 geographic coordinate system.
  • Disclosures with a WGS84 projection were read into a point shapefile with the World Geodetic System Datum of 1984 geographic coordinate system, and then transformed to NAD83 via the “NAD_1983_To_WGS_1984_1” datum transformation with the Project geoprocessing tool.
  • Disclosures with a NAD27 projection in the lower 48 United States were read into a point shapefile with the North American Datum of 1927 geographic coordinate system, and then transformed to NAD83 via the “NAD_1927_To_NAD_1983_NADCON” datum transformation with the Project geoprocessing tool.
  • Disclosures with a NAD27 projection with a StateFFQA listed as Alaska were read into a point shapefile with the North American Datum of 1927 geographic coordinate system, and then transformed to NAD83 via the “NAD_1927_To_NAD_1983_Alaska” datum transformation with the Project geoprocessing tool.

Following datum transformations to NAD83, these four shapefiles were merged into a single shapefile using the Merge geoprocessing tool. The final latitude and longitude coordinates (after transformation to NAD83, if needed) were stored in the NAD83_Lat and NAD83_Lon fields, respectively, in the QaWell table.

To join state and county names to each disclosure location, the Spatial Join geoprocessing tool was used with the 2010 TIGER/Line shapefile of counties from the US Census Bureau (USCB, 2011) with the “Join Operation” parameter set to “JOIN_ONE_TO_ONE” and the “Match Option” parameter set to “INTERSECT.” The resulting attribute table was exported to Microsoft Excel (Microsoft Corporation, 2002).

In Excel, the three sets of state and county locations were compared, resulting in six QA measures for the locational data. These comparisons were case-insensitive to avoid situations where, for example, the data values Mckee and McKee would not match. These comparisons also ignored spaces and hyphens to avoid situations where, for example, Mc Kee and McKee would not match. For each of the six comparisons, a QA flag field was added to the data table with True or False Boolean values:

  • StateMatchAPI_FF indicates whether or not the API code for the state (APIState) matches the state reported in the well header table (StateFFQA).
  • StateMatchGIS_FF indicates whether or not the state that contains the GIS-mapped disclosure location (GISState) matches the state reported in the well header table (StateFFQA).
  • StateMatchAPI_GIS indicates whether or not the API code for the state (APIState) matches the state that contains the GIS-mapped disclosure location (GISState).
  • CountyMatchAPI_FF indicates whether or not the API code for the county (APICounty) matches the county reported in the well header table (CountyFFQA).
  • CountyMatchGIS_FF indicates whether or not the county that contains the GIS-mapped disclosure location (GISCounty) matches the county reported in the well header table (CountyFFQA).
  • CountyMatchAPI_GIS indicates whether or not the API code for the county (APICounty) matches the county that contains the GIS-mapped disclosure location (GISCounty).

Based on these six fields, two additional flags were added:

  • AllStateOK is True if all three state comparison fields are True.
  • AllCountyOK is True if all six state and county comparison fields are True.

Locational data were used in the data analysis report for analyses in which information was needed at the state or county level. The QA-related fields were used as appropriate to either exclude disclosures that did not meet QA criteria from analyses or to categorize results with uncertain locational information.

6. Chemical Name Standardization

Ingredient names and CASRNs are entered by operators in the ingredients table, and the names can include a wide range of variations for a given ingredient, including synonyms, misspellings, different punctuations and formatting, and different alpha-numeric spacing. To identify ingredients used in hydraulic fracturing fluids, entries of both ingredient names and CASRNs were verified and standardized. The CASRNs were determined valid for analyses after being verified with the Chemical Abstracts Service (2014); ingredient records with invalid CASRNs were excluded from certain analyses presented in the data analysis report. Note that this approach assumes that the CASRN entered into the project database is correct.

Ingredient names for verified CASRNs were standardized using a list of unique chemical names paired with CASRNs developed by the EPA. This standardization was needed because of the above-noted range of presentations of ingredient names. Because the ingredient names were standardized, the names found in the data analysis report and the project database may differ from the names reported by operators in the original PDF disclosures.

The EPA used standardized chemical names from Appendix A in the agency’s Study of the Potential Impacts of Hydraulic Fracturing on Drinking Water Resources: Progress Report (2012) for the EPA-standardized chemical names used in the project database and in this report. 8 Chemical name and structure quality control methods were used to standardize chemical names for CASRNs found in the project database, but not included in Appendix A of the Progress Report. 9 The same methods were used in the development of Appendix A of the Progress Report and ensure correct chemical names and CASRNs.

7. Data Field Descriptions

The sections below provide a listing and descriptions of the data fields in the project database tables.

7.1. Data Fields in Main Tables

The primary tables that contain the data from the disclosures are:

  • OriginalWell
  • QaWell
  • OriginalIngredient
  • QaIngredient

The two “Original” tables contain the data as parsed from the original PDF disclosures. In the two “Qa” tables, data have undergone basic standardization, and a series of QA flag fields has been established to facilitate analyses. Fields with “QA” or “flag” in their names are in the “Qa” tables.

7.1.1. Well Header Field Descriptions

This section lists the fields in the OriginalWell and QaWell tables, which contain information derived from the 38,530 disclosures with successfully parsed well headers. For convenience, these are grouped into relevant categories based on the well header source field.

 

Well ID WellId   A unique identifier assigned to each disclosure that was parsed into the project database
Fracture Job Date DateFF   The verbatim fracture date from the parsed disclosure
  DateFFQA   DateFF after minor editing to correct obvious typos, incorrect formatting, and remove invalid values
  DateFFflag OK 38,277 disclosures (99.34%) with DateFF unchanged
    OK, formatted 2 disclosures (0.0052%) with DateFF reformatted to fix an obvious typo
    Early 222 disclosures (0.58%) with DateFF before 1/1/2011 (the first day of the study period), which resulted in a blank for these disclosures in the DateFFQA field
    Late 28 disclosures (0.073%) with DateFF after 2/28/2013 (the last day of the study period), which resulted in a blank for these disclosures in the DateFFQA field
    Unclear 1 disclosure (0.0026%) with DateFF that could not be read, which resulted in a blank for these disclosures in the DateFFQA field
State StateFF   The verbatim state name from the parsed disclosure
  StateFFQA   StateFF after minor editing to correct obvious typos and differences in formatting
  StateFFflag OK 33,699 disclosures (87.46%) with StateFF unchanged
    OK, misspelled 38 disclosures (0.099%) with StateFF corrected to fix an obvious typo
    OK, postal to full 4,793 disclosures (12.44%) with StateFF corrected to substitute postal code (e.g., TX changed to Texas)
County CountyFF   The verbatim county name from the parsed disclosure
  CountyFFQA   CountyFF after minor editing to correct misspelled County names, remove extraneous “County” and “Parish” suffixes, and remove invalid values
  CountyFFflag OK 36,758 disclosures (95.40%) with CountyFF unchanged
    OK, misspelled 563 disclosures (1.46%) with CountyFF corrected to fix an obvious typo
    OK, shortened 1,206 disclosures (3.13%) with CountyFF corrected to remove extraneous suffixes (e.g. County, Parish, Borough)
    Unclear 3 disclosures (0.0078%) with CountyFF that was omitted or otherwise erroneous, which resulted in a blank for these disclosures in the CountyFFQA field
API Well Number APIFF   The verbatim API well number from the parsed disclosure
  APIFFQA   APIFF after minor editing to include leading zeroes and add hyphens
  APIFFflag OK 29,168 disclosures (75.70%) with APIFF unchanged
    OK, formatted 9,352 disclosures (24.27%) with APIFF reformatted to include leading zeroes and add hyphens
    Different than filename 10 disclosures (0.026%) with APIFF different than the API well number embedded in the PDF filename
Operator OperatorFF   The verbatim well operator from the parsed disclosure
  OperatorFFQA   OperatorFF after minor editing to aggregate synonymous and misspelled operator names
  OperatorFFflag OK 9,935 disclosures (25.79%) with OperatorFF unchanged
    OK, mapped 28,595 disclosures (74.21%) with OperatorFF changed to a synonym based on the OperatorStandardization table
Well Name NameFF   The verbatim well name from the parsed disclosure
  NameFFQA   Matches NameFF because no values required editing
  NameFFflag OK 38,530 disclosures (100.0%) with NameFF unchanged
Longitude LongitudeFF   The verbatim longitude from the parsed disclosure
  LongitudeFFQA   LongitudeFF after minor editing to correct obvious typos and transpositions, and to remove invalid values
  LongitudeFFflag OK 38,394 disclosures (99.65%) with LongitudeFF unchanged
    OK, lat/lon swapped 4 disclosures (0.010%) with LongitudeFF clearly transposed with latitude
    OK, nonnegative 129 disclosures (0.33%) with LongitudeFF erroneously non-negative but otherwise valid
    Unclear 3 disclosures (0.0078%) with LongitudeFF likely erroneous based on the resulting map location, which resulted in a blank for these disclosures in the LongitudeFFQA field
Latitude LatitudeFF   The verbatim latitude from the parsed disclosure
  LatitudeFFQA   LatitudeFF after minor editing to correct obvious typos and transpositions, and to remove invalid values
  LatitudeFFflag OK 38,518 disclosures (99.97%) with LatitudeFF unchanged
    OK, lat/lon swapped 4 disclosures (0.010%) with LatitudeFF clearly transposed with longitude
    OK, negative 5 disclosures (0.013%) with LatitudeFF erroneously negative but otherwise valid
    Unclear 3 disclosures (0.0078%) with LatitudeFF likely erroneous based on the resulting map location, which resulted in a blank for these disclosures in the LatitudeFFQA field
Projection ProjectionFF   The verbatim projection (technically a datum) from the parsed disclosure
  ProjectionFFQA   Matches ProjectionFF because no values required editing
  ProjectionFFflag OK 38,530 disclosures (100.0%) with ProjectionFF unchanged
Production Type (oil or gas) TypeFF   The verbatim production type from the parsed disclosure
  TypeFFQA   Matches Type FF because no values required editing
  TypeFFflag OK 38,530 disclosures (100.0%) with TypeFFQA unchanged
True Vertical Depth DepthFF   The verbatim true vertical depth (in feet) from the parsed disclosure
  DepthFFQA   DepthFF after minor formatting to remove units, average ranges, and remove invalid values
  DepthFFflag OK 37,721 disclosures (97.90%) with DepthFF unchanged
    OK, formatted 81 disclosures (0.21%) with DepthFF formatted to remove units and other extraneous characters
    Range 5 disclosures (0.013%) with DepthFF given as a range, which resulted in the DepthFFQA value being averaged from the minimum and maximum range values
    High 14 disclosures (0.036%) with DepthFF greater than 25,000 feet, the upper threshold identified by the EPA for reasonable depths, which results in a blank for these disclosures in the DepthFFQA field
    Low 5 disclosures (0.013%) with DepthFF less than 500 feet, the lower threshold identified by the EPA for reasonable depths, which results in a blank for these disclosures in the DepthFFQA field
    Not given 704 disclosures (1.83%) with DepthFF not reported, which results in a blank for these disclosures in the DepthFFQA field
Total Water Volume VolumeFF   The verbatim total water volume (in gallons) from the parsed disclosure
  VolumeFFQA   VolumeFF after minor formatting to remove units and remove invalid values
  VolumeFFflag OK 38,108 disclosures (98.90%) with VolumeFF unchanged
    OK, formatted 27 disclosures (0.070%) with VolumeFF formatted to remove units and other extraneous characters
    OK, revised 140 disclosures (0.36%) with VolumeFF revised due to altered header format
    Empty, revised 32 disclosures (0.083%) with VolumeFF removed due to altered header format
    High 11 disclosures (0.029%) with VolumeFF greater than 50 million gallons (upper threshold set by the EPA), which results in a blank for these disclosures in the LatitudeFFQA field
    Not given 133 disclosures (0.35%) with VolumeFF not reported, which results in a blank for these disclosures in the LatitudeFFQA field
    Unclear 79 disclosures (0.21%) with VolumeFF given but not valid numbers, which results in a blank for these disclosures in the LatitudeFFQA field
Duplication APICount   In table QAWell, the number of disclosures with this API well number. A total of 2,283 disclosures (5.93%) shared an API well number with at least one other disclosure.
  Authoritative   In table QAWell, True if the disclosure is the authoritative disclosure among a set of duplicates with the same APIFFQA and DateFFQA, as determined by the folder date or file creation date. A total of 38,301 disclosures (99.41%) matched are authoritative.
Locational Data from API Well Number APIState   In table QAWell, the name of the State associated with the first two digits of the API Well Number in the APIFFQA field. The State associations were downloaded from http://www.spwla.org/technical/us-state-codes.
  APICounty   In table QAWell, the name of the County associated with the first five digits of the API Well Number in the APIFFQA field. The County associations were downloaded from http://www.spwla.org/xls/counties.xls.
Locational Data from GIS Spatial Join of Longitude/Latitude Coordinates NAD83_Lon   The LongitudeFFQA coordinate, after being converted to the NAD83 datum
  NAD83_La   The LatitudeFFQA coordinate, after being converted to the NAD83 datum
  GISState   The name of the state in which the NAD83_Lat and NAD83_Lon are located. Coordinates did not intersect a state in 56 disclosures (0.15%), resulting in blank values for GISState field.
  GISCounty   The name of the county in which the NAD83_Lat and NAD83_Lon are located. Coordinates did not intersect a county in 56 disclosures (0.15%), resulting in blank values for GISCounty field.
  USGSProvince   The name of the USGS Oil and Gas Province coincident with the disclosure's coordinates. Coordinates did not intersect a USGS province in 56 disclosures (0.15%), resulting in blank values for USGSProvince field.
  ShaleBasin   The name of the EIA Shale Basin coincident with the disclosure's coordinates. Coordinates did not intersect a shale basin in 1,120 disclosures (2.91%), resulting in blank values for ShaleBasin field.
  ShalePlay   The name of the EIA Shale Play coincident with the disclosure's coordinates. Coordinates did not intersect a shale play in 14,894 disclosures (38.66%), resulting in blank values for ShalePlay field.
  TightGas   The name of the EIA Tight Gas Basin coincident with the disclosure's coordinates. Coordinates did not intersect a tight gas basin in 4,170 disclosures (10.82%), resulting in blank values for TightGas field.
  CoalBed   The name of the EIA Coal Bed Methane Basin coincident with the disclosure's coordinates. Coordinates did not intersect a coalbed methane basin in 20,534 disclosures (53.29%), resulting in blank values for CoalBed field.
State Locational Matching StateMatchAPI_FF   True if APIState matches StateFFQA. The two field values matched for 38,476 disclosures (99.86%).
  StateMatchGIS_FF   True if GISState matches StateFFQA. The two field values matched for 38,390 disclosures (99.64%).
  StateMatchAPI_GIS   True if APIState matches GISState. The two field values matched for 38,381 disclosures (99.61%).
County Locational Matching CountyMatchAPI_FF   True if APICounty matches CountyFFQA. The two field values matched for 37,733 disclosures (97.93%).
  CountyMatchGIS_FF   True if GISCounty matches CountyFFQA. The two field values matched for 36,894 disclosures (95.75%).
  CountyMatchAPI_GIS   True if APICounty matches GISCounty. The two field values matched for 37,372 disclosures (96.99%).
Other locational fields AllStateOK   True if all three StateMatch fields are true. The three field values matched for 38,359 disclosures (99.56%).
  AllCountyOK   True if all three StateMatch and all three CountyMatch fields are true. The three field values matched for 36,754 disclosures (95.39%).
7.1.2. Ingredient Field Descriptions

This section lists the fields in the OriginalIngredient and QaIngredient tables, which provide information on additives and their ingredients, as well as base fluids and proppants.

 

IngredientId The unique identifier added to each ingredient record that was parsed into the database.
WellId The unique identifier added to each disclosure that was parsed into the database.
TradeName The ingredient trade name. A number of trade name values are comma-joined lists of multiple trade names for the entire disclosure. Microsoft Access cannot store many of these long values in a text field, but converting to Memo would increase database size.
Supplier The ingredient supplier. Supplier values (names) were standardized manually in QAIngredient.
Purpose The purpose assigned to a particular ingredient. In table QAIngredient, purpose entries were standardized manually to correct for misspellings, punctuation, hyphenation, and capitalization.
ChemicalName The original value parsed from the disclosures, in the OriginalIngredient table; or the standardized chemical name, where available, in the QaIngredient table.
Cas The CASRNs of the ingredient as parsed from the disclosures, in the OriginalIngredient table. In the QaIngredient table, CASRNs have been stripped of non-numeric characters and properly hyphenated, and CASRNs with invalid check digits have been removed.
EPAIngredientId The identifier that links ingredient name standardization in the QAIngredient table with the IngredientNameStandardization table. Records for 796,692 ingredients were matched to an EPAIngredientName.
AdditiveConcentration The original “maximum ingredient concentration in additive (% by mass)” parsed from FracFocus disclosures, in the OriginalIngredient table. In the QaIngredient table, entries expressed as a single decimal value were kept intact, while non-numeric values or ranges for 353,157 values were changed to Null.
FluidConcentration The original “maximum ingredient concentration in hydraulic fracturing fluid (% by mass),” in the OriginalIngredient table. Entries expressed as a single decimal value were kept intact, while non-numeric values or ranges for 291,293 values were changed to Null.
Comments Comments entered by the operator on the FracFocus disclosure. No changes were made to values in this field.
ValidTradeName True if the trade name should be regarded as valid. This flag is set based on the TradeNameStandardization table. Values of TradeName appear to not be trade names for 252,361 ingredients; these have been flagged in the QAIngredients table as having an invalid trade name (value of False).
ValidPurpose True if the purpose should be regarded as valid. This flag is set based on the PurposeStandardization table. Values of Purpose appear not be purposes for 204,123 ingredient records; these have been flagged in the QAIngredients table as having an invalid purpose (value of False).are clearly not purposes.
ValidAdditiveConcentration True if AdditiveConcentration is between 0 and 100. For 356,789 ingredients, this field has been flagged in the QaIngredients table as False (invalid value).
ValidFluidConcentration True if FluidConcentration is between 0 and 100. For 293,614 ingredients, this field has been flagged in the QaIngredients table as False (invalid value).
ValidCas True if Cas matches a standardized ingredient in the IngredientNameStandardization table. For 433,753 ingredients, this field has been flagged in the QaIngredients table as False (invalid value).

7.2. Data Fields in Tables Associated with Standardizations

Several tables store the corrections and standardizations used to develop the QAWell and QAIngredient tables. These standardizations have been conservatively developed to facilitate data analysis.

7.2.1. Chemical Name Standardization

The following table lists the fields in the IngredientNameStandardization table. Ingredient names for verified CASRNs were standardized using a list of unique chemical names paired with CASRNs that was developed by the EPA (Section 6).

 

EPAIngredientId The primary key for the table, which can be used to join the QaIngredient and IngredientNameStandardization tables.
EPAIngredientName The chemical name for the ingredient as determined by the EPA.
Cas The CASRN corresponding to an individual chemical. The EPA provided unique identifiers in the form of NOCAS_XXXXX (where XXXXX is a numerical identifier) for chemicals without CASRNs.
7.2.2. Operator Standardization Information

This section lists the fields of the OperatorStandardization table.

 

Original The original operator name, found in the Operator field of the OriginalIngredient table. The OperatorFF field in OriginalWell was joined to this table using this field during the standardization process.
Standardized The standardized name to use in the Operator field of the QaIngredient table
7.2.3. Trade Name Standardization

This section lists the fields of the TradeNameStandardization table, in which trade names were standardized to correct spelling and punctuation and evaluated to identify and flag entries that do not represent additives (e.g., numerical values, purposes, chemical names). Some fields were used in assigning a value to the ValidTradeName field in the QaIngredient table. Other fields provide additional categorization for reference.

 

ID A unique identifier for each row in this table.
Multiple Entries in Trade Name Field Checked if the trade name value appears to list multiple trade names. Some operators listed all additives used in one cell. This field is used to determine the value of the ValidTradeName field.
Ingredient (General name) - not proppant Checked if the value appears to be an ingredient. This field is used to determine the value of the ValidTradeName field.
Purpose Name Checked if the value appears to be an additive purpose. This field is used to determine the value of the ValidTradeName field.
Number that looks like possible concentration Checked if the value appears to be a chemical concentration (possibly the result of parsing errors). This field is used to determine the value of the ValidTradeName field.
Possible CASRN Checked if the value appears to be a CASRN. This field is used to determine the value of the ValidTradeName field.
Other Checked if there appears to be another type of problem with the trade name value. This field is used to determine the value of the ValidTradeName field.
Count A, B, C, D, E or F 1 if any of the above 6 fields are checked, otherwise 0.
May or may not be Trade Name Checked if it is not readily clear if the entry refers to something other than the trade name
Commodity Checked if the value of the trade name is a commodity name (e.g., water)
Proppant (generic or trade name) Checked if the value appears to indicate a proppant
Suggested spelling or punctuation correction The standardized value of the TradeName field of the QaIngredient table
Trade Name as Listed in FracFocus The original value of the TradeName field of the OriginalIngredient table. The TradeName field in OriginalIngredient was joined to this table using this field during the standardization process.
7.2.4. Ingredient Purpose Standardization

This section lists the fields of the PurposeStandardization table, in which purposes were evaluated to identify and flag entries that do not represent purposes (e.g., numerical values, chemical names, operator names). Some fields were used in assigning a value to the ValidPurpose field in the QaIngredient table. Other fields provide additional categorization for reference; the two fields referring to proppants were used in querying for proppants and in excluding proppants from additive ingredient analyses.

 

ID A unique identifier for each row in this table
Multiple Entries in Purposes Field Checked if the additive purpose value appears to list multiple purposes. Some operators listed the purposes of all additives used in one cell. This field is used to determine the value of the ValidPurpose field.
Ingredient (General Name)(excludes HCl) Checked if the value appears to be a chemical ingredient. This field is used to determine the value of the ValidPurpose field.
Commercial Product Name that doesn't include purpose and not IDd Checked if the value appears to be a trade name of an additive. This field is used to determine the value of the ValidPurpose field.
Purpose Can Be Inferred from Product Name or From Another Entry Checked if the purpose be inferred from an additive name or some other purpose entry for another ingredient record. This field is used to determine the value of the ValidPurpose field.
Item is Likely a Proppant Checked if the value appears to indicate a proppant, even though it does not use a common identifying term such as proppant or list one of the chemical names sand, silica, or quartz. This field is used to determine the value of the ValidPurpose field.
Other Checked if there is another type of problem with the additive purpose value. This field is used to determine the value of the ValidPurpose field.
Count B, C, D, E, F, or G 1 if any of the above 6 fields are checked, otherwise 0.
Proppant - uses word Proppant or other Identifying Term Checked if the value appears to indicate a proppant, using the word proppant or listing one of the chemical names sand, silica, or quartz or other identifying term
Purpose corrected for caps, spacing, dashes, misspellings The standardized value of the Purpose field of the QaIngredient table
Purpose as Listed in FracFocus The original value of the Purpose field of the OriginalIngredient table. The Purpose field in OriginalIngredient was joined to this table using this field during the standardization process.
Related to Base Fluid Checked if the additive purpose appears to be related to the base fluid
Related to Alternative Carrier Checked if the additive purpose appears to be related to a non-water base fluid. The relationship was determined by observation and used for analysis of non-water base fluids.

7.3. Data Fields in Other Tables

Several additional tables have been added to the database with lists that were used to support the analyses described in the data analysis report.

7.3.1. Proppant Identification

This section contains information about the Proppants table, which lists solids (e.g., minerals, ceramics) associated with proppant-related purposes (as parsed from disclosures). Information in this table assisted with excluding the minerals used as proppants from analyses of additive ingredients.

 

ChemicalName The chemical name of the proppant. The ChemicalName field in QaIngredient was joined to this table using this field to identify proppants.
Cas The CASRN of the proppant
OK to exclude Checked if the chemical can be excluded from the additive ingredient analyses
7.3.2. Resin Coating Identification

This section contains information about the ResinCoating table, which lists ingredients parsed from disclosures associated with the additive purpose of resin coatings. This list assisted in capturing the ingredients used for resin coatings on proppants in analyses of additive ingredients.

 

ChemicalName The chemical name of the resin coating. The ChemicalName field in QaIngredient was joined to this table using this field to identify resin coatings.
Cas The CASRN of the resin coating
7.3.3. CBI Identification

This section contains information about the CBISynonym table, which lists terms used to indicate that an operator has claimed CBI status for an ingredient in the ChemicalName and Cas fields of the OriginalIngredient table. This table was used for analyzing the numbers of ingredient records in the database that were listed by the operators as CBI.

 

Term A term indicating CBI.
7.3.4. Water Source Identification

This section contains information about the WaterSourceTerm table, which lists terms in the TradeName and Comments fields of the OriginalIngredient table that indicate the source of water used for the base fluid (e.g., fresh, recycled). This table was used to query the database for information on water sources.

 

Source A term indicating a water source
7.3.5. Purpose Categorization

This section contains information about the PurposeCategorization table, which lists the categories of purposes as found in the Purpose field of the QaIngredient table. This table was used to group ingredients by purpose category.

 

Category The category of the standardized purpose.
Purpose The standardized purpose.
7.3.6. State Regulation Information

This section contains information about the StateRegulation table, which contains information about state reporting requirements. A single state may have multiple rows when regulations are amended.

 

ID A unique identifier for each row in this table
State The name of a state
Reporting Requirement Type The recipient of required reporting, either the FracFocus registry (FracFocus), the state regulator (State), both FracFocus and the state (FracFocus AND State), or either FracFocus and the state (FracFocus OR State).
EffectiveDate The effective date of the state regulation.
Effective Date within FF DB Timeframe? Either Y if the date is between 1/1/2011 and 2/28/2013 or N otherwise.
Notes Notes about the regulation, including relevant limitations.
7.3.7. County Information

This section contains information about the Counties table, which contains information about counties.

 

STATE The state abbreviation
COUNTY The full name of a county (e.g., Clay County)
FIPS The county FIPS code
STATE_FIPS The state FIPS code
CountyName The short name of a county (e.g., Clay)
StateName The name of a state
CaseStudy Identifies whether the county is a focus county in the data analysis report
7.3.8. Water Synonyms

This section contains information about the WaterSynonyms table, which contains a list of synonyms for an unknown water source.

 

TradeName A synonym for an unknown water source
7.3.9. Unparsed PDFs

This section contains information about the UnparsedPDFs table, which lists the 606 PDF files that could not be successfully parsed (Table 1).

 

PDFName The PDF filename of the unparsed disclosure
API_Final The API well number, as extracted from PDFName
Data Storage Error Identifies 14 disclosures that GWPC indicated should be excluded from the project database because of a data storage error
State The state in which the disclosure is located, based on the API well number

8. Summary

The project database was developed from PDF disclosures given to the EPA by the GWPC and submitted to the FracFocus Chemical Disclosure Registry 1.0 before March 1, 2013. Data from the PDF files were converted to XML format, parsed, and incorporated into a Microsoft Access database. The data in the project database were then subject to QA procedures to ensure that the results from analyses of the project database reflect the data contained in the original PDF disclosures, while identifying obviously invalid or incorrect data to exclude from analyses. A conservative approach was used in all data handling; no records were deleted and the original data remain in the project database. To improve the results of analyses, data have been subject to minimal standardization of operator names, trade names, and purposes, as well as standardization of chemical names according to CASRNs. The standardized entries are included in the two “Qa” tables. During QA work on the project database, data limitations were encountered, and QA flag fields were developed to identify agreement among locational data and instances of problematic data. During data analysis, database queries and subsequent calculations were structured to compensate for these limitations. Results of analyses conducted on the project database are presented in the Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0 (US EPA, 2015).

References

Adobe Systems Incorporated. 2011. Adobe Acrobat X Pro 10.

Arthur, JD, Layne, MA, Hochheiser, HW, and Arthur, R. 2014. Spatial and Statistical Analysis of Hydraulic Fracturing Activities in US Shale Plays and the Effectiveness of the FracFocus Chemical Disclosure System. SPE Hydraulic Fracturing Technology Conference, The Woodlands, Texas, February 4-6. Society of Petroleum Engineers.

Carter, KE, Hakala, JA, and Hammack, RW. 2013. Hydraulic Fracturing and Organic Compounds - Uses, Disposal and Challenges. SPE Eastern Regional Meeting, Pittsburgh, Pennsylvania, August 20-22. Society of Petroleum Engineers.

Chemical Abstracts Service. 2014. Check Digit Verification of CAS Registry Numbers. Available at http://www.cas.org/content/chemical-...ances/checkdig. Accessed April 21, 2014.

DrillingInfo, Inc. 2011. DI Desktop December 2011 download.

Esri,Inc. 2012. ArcGIS 10.1.

Microsoft Corporation. 2013. Excel 2013.

Microsoft Corporation. 2012. Access 2013.

Python Software Foundation. 2012. Python 2.7.

Richardson, L. 2013. Beautiful Soup 4.

Society of Petrophysicists and Well Log Analysts. 2010. API Standards Information. Available at http://www.spwla.org/technical/api-codes. Accessed April 21, 2014.

US Energy Information Administration (US EIA). 2007. Data for the Coalbed Methane Panels. Oil- and Gas-Related Maps, Geospatial Data, and Geospatial Software. Available at http://www.eia.gov/ pub/oil_gas/natural_gas/analysis_publications/maps/maps.htm. Accessed April 18, 2014.

US EIA. 2011a. Data for the Tight Gas Plays Map. Oil- and Gas-Related Maps, Geospatial Data, and Geospatial Software. Available at http://www.eia.gov/pub/oil_gas/natural_gas/ analysis_publications/maps/maps.htm. Accessed April 18, 2014.

US EIA. 2011b. Data for the US Shale Plays Map. Oil- and Gas-Related Maps, Geospatial Data, and Geospatial Software. Available at http://www.eia.gov/pub/oil_gas/natural_gas/ analysis_publications/maps/maps.htm. Accessed April 18, 2014.

US Environmental Protection Agency (US EPA). 2012. Study of the Potential Impacts of Hydraulic Fracturing on Drinking Water Resources: Progress Report. EPA 601/R-12/011. US Environmental Protection Agency, Washington, DC. 278 pages.

US EPA. 2013. Distributed Structure-Searchable Toxicity (DSSTox) Database Network. Available at http://www.epa.gov/ncct/dsstox/index.html. Accessed April 21, 2014.

US EPA. 2014. Substance Registry Services. Available at http://ofmpub.epa.gov/sor_internet/ registry/substreg/home/overview/home.do. Accessed April 21, 2014.

US EPA. 2015. Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0. EPA/600/R-1/003. US Environmental Protection Agency, Washington, DC. 168 pages.

US National Library of Medicine (US NLM). 2014. ChemID Plus Advanced. Available at http://chem.sis.nlm.nih.gov/chemidplus. Accessed April 21, 2014.

US Census Bureau (USCB). 2011. Toplogically Integrated Geographic Encoding and Referencing (TIGER)/Line Shapefiles. Available at ftp://ftp2.census.gov/geo/tiger/TIGER2010/COUNTY/2010. Accessed September 16, 2013.

US Geological Survey (USGS). 1995. Province Boundaries shapefile. National Oil and Gas Assessment. Available at https://catalog.data.gov/dataset/199...nce-boundaries. Accessed April 18, 2014.

Footnotes

1

The project database and the data analysis report are available at http://www2.epa.gov/hfstudy/publishe...entific-papers.

2

Prior to February 28, 2011, six of the 20 states with data in the project database began requiring operators to disclose chemicals used in hydraulic fracturing fluids to FracFocus (Colorado, North Dakota, Oklahoma, Pennsylvania, Texas, and Utah). Three other states started requiring disclosure to either FracFocus or the state (Louisiana, Montana, and Ohio), and five states required or began requiring disclosure to the state (Arkansas, Michigan, New Mexico, West Virginia, and Wyoming). Alabama, Alaska, California, Kansas, Mississippi, and Virginia did not have reporting requirements during the period of time studied in the data analysis report. Between February 5, 2011, and April 13, 2012, Pennsylvania required reporting to the state. As of April 14, 2012, Pennsylvania required reporting to both the state and FracFocus.

3

FracFocus 2.0 became the exclusive disclosure mechanism in June 2013. More information on the FracFocus 1.0 FracFocus 2.0 formats may be found in the FracFocus 2.0 Operator Training materials available at http://fracfocus.org/node/331.

4

More information on the field descriptions may be found in Section 7.1.1.

5

Adobe Acrobat identified apparent dates and standardized them automatically. The standardization in this dataset was later reversed, because Acrobat occasionally “standardized” non-date values.

6

Additive ingredients are defined as ingredients reported for additives that have purposes other than base fluid or proppant.

7

Well locations in Alaska were not subject to county-level locational QA criteria, because the five-digit API well numbers in Alaska are not organized by counties. The coordinates for all disclosures from Alaska fall within the boundaries of the North Slope borough.

8

Table A-1 in the Progress Report.

9

In the majority of cases, valid CASRNs and the associated ingredient names in the project database were paired correctly for a given CASRN. If an ingredient name (whether specific or non-specific) did not match the CASRN reported by the operator, the CASRN was added to a chemical name standardization list and assigned a correct chemical name. The chemical standardization list consists of CASRNs paired with appropriate chemical names and was used to standardize chemical names in the project database based on the CASRNs reported by the operators. This process was undertaken because numerous synonyms and misspellings for a given chemical were present in the original data. Standardized, specific chemical names were identified using the EPA’s Distributed Structure-Searchable Database Network (US EPA, 2013), the EPA’s Substance Registry Services database (US EPA, 2014a), and the U.S. National Library of Medicine ChemID database (US NLM, 2014). Additional information on chemical name and structure quality control methods can be found at http://www.epa.gov/ncct/dsstox/Chemi...rocedures.html.

Reproducing FracFocus 1.0 Data Analysis Tables and Figures, March 2015

Source: http://www2.epa.gov/hfstudy/epa-proj...-1-disclosures (PDF)

Background

The data underlying the tables and figures in the Analysis of Hydraulic Fracturing Fluid Data from the FracFocus Chemical Disclosure Registry 1.0 (EPA/601/R14/003; referred to as FracFocus 1.0 Data Analysis) can be recreated using queries in the Microsoft Access 2013 database developed from FracFocus 1.0 disclosures. For 12 of the tables from the FracFocus 1.0 Data Analysis (see list below), the only action required to reproduce the tables and figures from the report is to conduct an “Access query” by double-clicking the name of the corresponding query in the project database (e.g., "Table 1 – FracFocus Disclosure Count"). For the 26 remaining tables and figures presented in the FracFocus 1.0 Data Analysis (see list below), additional calculations are needed using the statistical program R (R Core Team, 2013) in order to reproduce the results in the FracFocus 1.0 Data Analysis.

Reproduction of tables and figures that include data analyzed in R (e.g., Tables ES-1 and ES-2) requires R to connect to pre-built queries in Access and then summarize query results and calculate summary statistics. R may be downloaded at http://cran.us.r-project.org/. Reproduction of tables that use R code will require the following packages in R in addition to the provided code: RODBC, data.table, reshape2, plyr, ggplot2, gridExtra, gtable, scales. Additionally, Geographic Information System (GIS) software is needed to reproduce the five maps.

The following list indicates the method needed to reproduce each table or figure in the FracFocus 1.0 Data Analysis and the page on which the R code can be found. If additional analysis in R is required, the query that provides the raw data is identified in the column, “Corresponding Query in the EPA Project Database.” The R code used to generate specific tables is provided in this guide beginning on page 4.

Table/Figure from FracFocus 1.0 Data Analysis

Table/Figure from FracFocus 1.0 Data Analysis Corresponding Query in the EPA Project Database How To Reproduce Page Number for R Code
Table ES-1 Table ES-1 – Summary Raw data in Access query + R Code 4
Table ES-2 Table ES-2 – Most Frequently Reported Ingredients Raw data in Access query + R Code 5
Table 1 Table 1 -Geog Distrib Parsed Disclosures Access query  
Table 2 Table 1 -Geog Distrib Parsed Disclosures Access query  
Table 3 Table 2 -FracFocus Disclosure Count Access query  
Table 4 Table 4 -Chem CBI By State Access query  
Table 5 Table 5 -FracFocus Filters Access query  
Table 6 Table 6 -Disclosures Submitted Raw data in Access query + R Code 6
Table 7 Table 7 -Chemicals Per Disclosure By State Raw data in Access query + R Code 7
Table 8 Table 8 -Twenty Frequent Chemicals Oil Raw data in Access query + R Code 9
Table 9 Table 9 -Twenty Frequent Chemicals Gas Raw data in Access query + R Code 11
Table 10 Table 10 -Frequent Chemical Purposes Raw data in Access query + R Code  
Table 11 Table 11 -Counties Included Regional Diversity Access query 12
Table 12 Table 12 -Twenty Chemicals Selected Counties Raw data in Access query + R Code  
Table 13 Table 13 -Nonaqueous Base Fluids Raw data in Access query + R Code 13
Table 14 Table 14 -Usage Nonaqueous Ingredients By State Access query  
Table 15 Table 15 -Water Usage By State Raw data in Access query + R Code 15
Table 16 Table 16 -Water Usage 90th Percentile Counties Raw data in Access query + R Code 16
Table 17 Table 17a -Disclosures By Water Source Table 17b -Disclosures By Water Source Table 17c -Disclosures By Water Source Raw data in Access query + R Code 17
Table 18 Table 18a -Median Concentrations By Source Table 18b -Median Concentrations By Source Raw data in Access query + R Code 19
Table 19 Table 19 -Ten Frequent Proppants Raw data in Access query + R Code 21
Table B-1 Table B-1 -Chemical families CBI Access query  
Table B-2 Table B-2 -Most Frequently Reported CBI Purposes Raw data in Access query + R Code 23
Appendix C qryChemAnalysis Raw data in Access query + R Code 24
Table D-1 Table D-1a -Disclosures By Operator By State Table D-1 -Disclosures By Operator By State Access query  
Table E-1 Table E-1 -Reporting regulations Access query  
Table F-1 Table F-1 – Additive Purposes Access query  
Table G-1 Table G-1 -Twenty Chemicals Andrews Raw data in Access query + R Code 27
Table G-2 Table G-2 -Twenty Chemicals Bradford Raw data in Access query + R Code 29
Table G-3 Table G-3 -Twenty Chemicals Dunn Raw data in Access query + R Code 31
Table G-4 Table G-4 -Twenty Chemicals Garfield Raw data in Access query + R Code 33
Table G-5 Table G-5 -Twenty Chemicals Kern Raw data in Access query + R Code 35
Table H-1 Table H-1 -County Water Usage Raw data in Access query + R Code  
Figure 2 Maps Raw data in Access query + R Code + GIS 38
Figure 3 Maps Raw data in Access query + R Code + GIS 39
Figure 4 Figure 4 -Monthly Distribution Access query  
Figure 5 Maps Raw data in Access query + R Code + GIS 40
Figure 6 Maps Raw data in Access query + R Code + GIS 41
Figure 7 Maps Raw data in Access query + R Code + GIS 42

R Code to Reproduce Table ES-1

Table ES-1 includes state-specific information on the number of unique disclosures with a fracture date between January 1, 2011, and February 28, 2013; total water volumes reported per disclosure; and the number of unique additive ingredients reported per disclosure

EPAFRactureDataRCode Example.png

NEXT

Page statistics
2018 view(s) and 56 edit(s)
Social share
Share this page?

Tags

This page has no custom tags.
This page has no classifications.

Comments

You must to post a comment.

Attachments