Data omissions undermine transparency

Summary: Fracking chemical quantities reported in FracFocus can range from less than one pound to over one million pounds. Open-FF has found that data records for chemical quantity are incomplete in over 35,000 disclosures and more are added monthly. These FracFocus data omissions prevent users from understanding the scope of the environmental and public health risks associated with hydraulic fracturing operations. This post is part of Open-FFโ€™s effort to make FracFocus limitations known to stakeholders.

Background

FracFocus serves as a primary source of information about the chemicals used in hydraulic fracturing operations across the United States. As a national registry, it compiles disclosures from operating companies, detailing the types and amounts of chemicals used in each fracking job. While FracFocus holds the potential to be a powerful tool for understanding the environmental and health implications of fracking, its usefulness is significantly limited when critical values are missing or incomplete. This problem undermines the registry’s ability to provide a comprehensive picture of fracking chemical use.

Open-FF has found an extensive number of fundamental values are missing. These FracFocus data omissions hide the quantity of materials used and restrict what stakeholders can learn from FracFocus. To illustrate the issue, we focus here on one central variable: the quantity of a chemical in an individual record.

Records of Chemical Quantity

One necessary column of data anchors reported chemicals in the reality of a fracking job: chemical quantity (endnote 1). The range of quantities found in FracFocus is mind-bending. For a single chemical, hydrochloric acid (HCl), FracFocus registers more than 400 records under one pound – 2 cups or less. But we also find more than 350 records over one million pounds, enough to fill more than 12 tanker trucks – the kind you see on the highway – with the acid. Being able to distinguish between small and large quantities in FracFocus is clearly important.

In FracFocus, the standard quantity value is conveyed by “PercentHFJob,” which tells us how much of the total fracking fluid is made up of that specific chemical. For example, if a fracking job used 1,000,000 pounds of fluid, and hydrochloric acid made up 1% of that fluid, then 10,000 pounds of hydrochloric acid were used.

Missing PercentHFJob

However we find that many FracFocus records have either no data or a zero in this field (example).  A zero here does not make sense because if a chemical is reported, it has a non-zero weight. FracFocus is able to represent very small numbers – there are over 1000 records that register weights less than one grain of rice. Therefore, a value of zero is essentially the same as “no data.”

Currently about 6% of all FracFocus records don’t have this quantity value. We estimate the mass of those records to be over 50 billion pounds (2). Furthermore, this pattern has become worse in recent years:

The red horizontal line is the average percent of records across all FracFocus data.

Which chemicals

More than 800 distinct chemicals are on the affected list. Below is a selection of materials with either a large number of records without quantity or are never reported with quantity data. We highlight the name of the material when it is on a list of chemicals of concern (3).

CASRNNameNumber of omitted recordsPercent of records are omitted
“proprietary”(declared “trade secret”)656857%
1310-73-2Sodium hydroxide1053913%
7681-52-9Sodium hypochlorite826738%
64742-47-8Distillates (petroleum), hydrotreated light79304%
14808-60-7Quartz (SiO2)75652%
107-21-1Ethylene glycol68447%
12125-02-9Ammonium chloride683710%
67-56-1Methanol59472%
64-19-7Acetic acid49807%
79-06-1Acrylamide423013%
7647-01-0Hydrochloric acid33372%
10049-04-4Chlorine dioxide293923%
68551-12-2Alcohols, C12-16, ethoxylated29385%
68648-89-5Benzene, ethenyl-, polymer with 2-methyl-1,3-butadiene, hydrogenated280956%
81741-28-8Tributyltetradecylphosphonium chloride265918%
50-00-0Formaldehyde264113%
127-08-2Potassium acetate252655%
13463-67-7Titanium dioxide250715%
1305-78-8Calcium oxide245086%
12136-45-7Potassium oxide244790%
16481-66-6Potassium metaborate (KBO2) hydrate(3:4)232154%
7758-19-2Sodium chlorite21229%
13840-56-7Boric acid (H3BO3), sodium salt (1:x)207666%
68424-85-1C12-16-Alkylbenzyldimethylammonium chlorides20434%
93-83-4Oleic acid diethanolamide203314%
69011-36-5Polyoxyethylene trimethyldecyl alcohol202919%
108-20-3Isopropyl ether1163100%
110-44-12E,4E-Hexadienoic acid54100%
68991-48-0Alcohols, C7-21, ethoxylated52100%
10124-31-9Ammonium acid phosphate50100%
63428-86-4Polyethylene glycol hexyl ether sulfate, ammonium salt41100%
68036-95-3Oxirane, 2-methyl-, polymer with oxirane, ether with 2-(chloromethyl)oxirane polymer with 4,4โ€ฒ-(1-methylethylidene)bis[phenol]28100%
30704-64-4p-tert-Butylphenol-ethylene oxide-formaldehyde-propylene oxide copolymer21100%
94525-01-6Aspartic acid, homopolymer, sodium salt15100%
113221-69-52-Propenoic acid, ethyl ester, polymer with ethenyl acetate and 2,5-furandione, hydrolyzed12100%
652-18-62,3,5,6-Tetrafluorobenzoic acid11100%
76006-33-23-Bromo-2-methylbenzoic acid11100%
328-89-22-Bromo-4-(trifluoromethyl)benzoic acid11100%
320-31-04-Bromo-2-(trifluoromethyl)benzoic acid11100%
28314-83-25-BROMO-2,4-DIFLUOROBENZOIC ACID11100%
123-17-12,6,8-Trimethyl-4-nonanol8100%
28178-42-92,6-Diisopropylphenyl isocyanate5100%
28906-96-9Bisphenol A-epichlorohydrin-polyformaldehyde copolymer3100%
7440-42-8Boron2100%
68458-49-1Polyphosphoric acids, esters with polyethylene glycol nonylphenyl ether2100%
68037-49-0Sodium C10-18-alkanesulfonates2100%
74499-34-6Alcohols, C12-15, propoxylated1100%
108-31-62,5-Furandione1100%
64741-66-8Naphtha, petroleum, light alkylate1100%
68514-28-3Humic acids, potassium salts1100%
68424-51-1Fatty acids, tall-oil, compds. with salicylic acid 2-(aminoiminomethyl)hydrazide1100%

Which Companies

Responsibility for these omissions is not spread evenly across the 1,800 operators reporting in FracFocus. We find that more than half of companies have never omitted quantity data.

If we plot just those companies with some records with omissions, we get a better picture of the range of offenders.

Showing the omission statistics of all companies that have some quantity data omissions. The 10 operators with the most flawed records are highlighted in orange.

In this chart, companies above the red horizontal line account for just 25% of all companies with omissions. Similarly, companies right of the red vertical account of just 25% of the all non-zero companies. Most companies are in the lower left quadrant. In contrast, the companies in the upper right quadrant have particularly high levels on data omission, either by total number or by percent of disclosures.

As part of this series, we will analyze many of the companies in that quadrant in upcoming posts. (See endnote 4 for companies with good reporting records.) Visit this page to see all posts related to this issue. The ten operators with the most flawed records are:

Operator
(and link to report)
Number of flawed records
Pioneer Natural Resources127,329
EOG Resources16,188
Blackbeard Operating10,697
Continental Resources10,057
DE3 Operating LLC7,966
Marathon Oil7,711
Chesapeake Operating7,626
Anadarko Petroleum Corporation7,353
SandRidge Energy7,323
BHP Billiton Petroleum6,253

Open-FF’s data response

For Open-FF data sets, although we detect and flag omitted data, we do not attempt to fill it with estimated values. Instead we leave the values as “no data” so that any aggregate calculations are not compromised by data not directly reported by the industry. In cases in which a user wishes to use an estimate, we suggest using the median PercentHFJob value for that chemical as a proxy. From there, a mass can be calculated from the “job_mass” in the data set.


Endnotes

  1. The name used in FracFocus for this value is “PercentHFJob” which is the chemical concentration within the entire fracking fluid. (On individual PDFs, this value is named “Maximum Ingredient Concentration in HF Fluid”). While FracFocus does sometimes directly provide the mass of the chemical (in pounds), it is not required data and therefore not available for many disclosures, it is not displayed on the PDF versions of disclosures, and we have found that in many cases, it is internally inconsistent and not reliable. PercentHFJob, on the other hand is a core piece of data for all disclosures and the mass of chemicals can be calculated from this value when disclosures are complete. To learn more about how mass is reported in FracFocus and is used in Open-FF, visit this page.
  2. Because the data for these records is missing, we cannot know the mass of any of them. However, we can roughly estimate it by using the median PercentHFJob for all FracFocus records that have values. Combining all of those rough estimates gives us a value of over 50 billion pounds.
  3. The lists of Chemicals of Concern we used here are:
    – the Clean Water Act list as compiled in EPA’s CompTox
    –  EPA’s Drinking Water Safety and Health Advisory list
    –  The Endocrine Disruption Exchange list
    – EPA’s comprehensive list of PFAS related compounds
    – the TSCA list of “Unknown, Variable composition, Complex reaction products and Biological” materials, meaning “these chemical substances cannot be represented by unique structures and molecular formulas.” These materials are typically poorly understood for the hazard profile.
    We note that the fracking industry is exempted from regulation of the chemicals on these lists. Nevertheless, the lists can help us understand potential hazards.
  4. At the other end of the spectrum from the big offenders are the companies with better reporting. The table below indicates companies at least 5,000 total records in FracFocus but less than 0.5% of their disclosures have quantity omissions.
OperatorTotal number of disclosuresPercent of records with omissionsNumber of flawed records
Aera Energy LLC22450.35187
Chief Oil & Gas34500
Citation Oil and Gas Corp.5090.580
Clayton Williams Energy, Inc.1830.4731
Crestone Peak Resources46400
EXTRACTION OIL & GAS LLC8750.4175
Grayson Mill Operating, LLC2400.1611
Great Western Oil & Gas Company, LLC4770.3869
Hilcorp Energy Company6950.4481
Kodiak Oil & Gas Corp1350.3125
Mack Energy Co.19300
Merit Energy Company LLC2110.0875
PennEnergy Resources, LLC24400
QEP Energy Company12830.22121
RKI Exploration & Production2390.1917
Range Resources Corporation12370.33121
SN EF Maverick, LLC4480.534
SN Operating, LLC24300
Samson Exploration, LLC1280.3218
Scout Energy Partners5900.1829
Sheridan Production Company, LLC3310.4660
Statoil Texas Onshore Properties LLC20900
Stronghold Energy II Operating LLC2050.0141
Summit Petroleum LLC3350.09611
Three Rivers Operating Co II LLC1750.2316
Ultra Resources8970.2970
Verdad Oil and Gas Corporation2230.138

5. All statistics in this post are calculated from a FracFocus data download from Jan.17, 2025.

5. Title image by Becky Mansfield, modified by author

Published by gwallison

I am a data analyst and programmer. I am interested in making "public" data more accessible.

54 thoughts on “Data omissions undermine transparency

Leave a Reply