Summary: Open-FF1 has found a large number of chemical disclosures that misreport concentrations of fracking sand and water. These errors seriously distort the chemical profiles of the fracking jobs.
A common feature of disclosures in FracFocus is the large amount of water and sand. Usually, water is about 88%2 of the fracking fluid, and sand3 is about 11%. The rest is made up of chemical additives like surfactants and biocides as shown in the first figure.
The second figure shows a random sample of disclosures with the typical sand/water ratio: the pattern is clear even with obvious variation4.
But within FracFocus, we find examples where that ratio is very different, where sand dominates the whole disclosure. The next figure shows a random sample of disclosures in which sand dominates.
Certainly, companies experiment with ingredients and FracFocus data contains many examples. But most likely, experimentation is not the case with these sand-dominant disclosures. Instead, they are mistakes.5
As we explore below, these mistakes undermine not just sand data, but all records in a disclosure.
Evidence of error
Besides the obvious physical improbability that such a “fluid” would work as a fracking medium, other evidence points to misreporting6.
First, there are occasions where higher sand percentages are used legitimately. We reported on how gas-foam fracking jobs may have more sand. We exclude those jobs from this analysis. However, sand use at 50% or greater is rare even in those situations.
More evidence comes from operator’s response when alerted to these disclosures. Operators in Pennsylvania submit disclosures to both FracFocus and to a PA DEP disclosure instrument7. Unlike most FracFocus disclosures that appear to be published without state review, PA disclosures are not accepted until they are reviewed.
We found that the PA DEP rejected some sand-dominated disclosures and that operators then resubmitted the disclosures with more typical water-to-sand ratios. For example, CNX Gas Company submitted a disclosure (37-059-28201) with the following numbers in April 2023 (with water and sand highlighted)8:
PA DEP rejected the disclosure in October 2023 with the note to CNX:
“Please verify and correct data in Chemical Component % Mass used in Total Base Fluid column on page 6. % volumes appear to be incorrect.”
Soon after, CNX resubmitted the disclosure with the following numbers9:
Although CNX changed its disclosure to PA DEP several months ago, the FracFocus disclosure has remained uncorrected (as of Aug 19, 2024).
Ripples
When the numbers for the chemical percentages are wrong, the reported mass of ingredients may be wrong, too. Below we compare the masses for the first three ingredients in the CNX disclosure as submitted to FracFocus (in the field MassIngredient) to the masses calculated from the CNX-corrected numbers submitted to PA DEP.
| Material | Mass submitted to FracFocus (pounds) | Mass based on corrected percentages (pounds) |
| Water | 226,073,410 | 225,930,000 |
| Sand | 853,871,650 | 38,580,000 |
| HCl | 90,283 | 22,140 |
The fracking sand errors in this CNX disclosure are modest compared to some others. COG Operating, LLC reported a fracking job with over 1,000 trillion pounds of sand:
Across FracFocus
We have found hundreds of these sand-dominant disclosures and they have been part of FracFocus since it started.
FracFocus represents one of the few public information sources about fracking chemicals. While the industry wants the public to trust this source, when there are major errors in the data set and those errors remain uncorrected for years, trust is impossible. If FracFocus is to be a “a one-stop, easy-to-understand public resource for consumers wishing to explore this data,”10 data integrity is essential.
FracFocus Support has informed us that they are preparing an alert to companies about these sand-based issues11. Disclosures from specific states and companies with this fracking sand problem will be the subject of future posts and will be linked here.
- Open-FF is a public service project to foster better understanding of fracking chemicals disclosed at FracFocus. See the Open-FF home for more information and links to resources. ↩︎
- These are the median percentages across all FracFocus disclosures. ↩︎
- CASRN: 14808-60-7 ↩︎
- The data for this analysis is from the Open-FF data repository created from a FracFocus bulk data download on June 11, 2024. Because of disclosure publishing delays, the most recent months are often not reflective of current activity. Therefore we filtered out all disclosures with an end date within 90 days of the download date. In addition, disclosures with a different fluid profile occur in the FracFocus data: where the use of nitrogen or carbon dioxide replaces part of the water carrier. Disclosures with this type of profile have been excluded from this analysis. ↩︎
- When state agency representatives answer our questions about these disclosures, they have agreed that these sand dominated disclosures must be mistakes. ↩︎
- We have been told by FracFocus Support that some forms of fracking can tolerate higher percentages of sand. In particular, foam-based carriers where carbon dioxide or nitrogen gases make up a large fraction of the carrier may be able to carry more sand. Disclosures that include those gases make up less than 2% of FracFocus data and we exclude them from our analysis here. ↩︎
- You can search the PA DEP completion disclosures at: https://www.ahs.dep.pa.gov/eSubmissionPublicSearch/ ↩︎
- Accessed June, 2024 ↩︎
- As of June 5, 2024, PA DEP has not accepted the disclosure; by Aug. 5, 2024, the disclosure was marked as “Accepted.” ↩︎
- https://fracfocus.org/ accessed June 6, 2024. ↩︎
- Email to Gary Allison, Aug 1, 2024. ↩︎
Photo citation: Ted Auch, FracTracker Alliance, 2020. (modified by author)
25 thoughts on “Sand-dominated disclosures: patterns and problems”