[coffeeshopgoth]

I have used the FracFocus database for about 6 or 7 years now (been a petroleum/reservoir engineer for ~20 years), and it is quite depressing to look at some of the chemicals they use. While I applaud the study, it would be interesting to see how much cleanup they have tried to do on the data to yield their numbers (OpenFF isn't great). I have written a fairly long cleanup in Python to get some of the data valuable to the industry. Hydraulic Fracturing recipes differ from basin to basin based on geologies/depths/and major product being produced, just to name a few variables. Doing some math, because they have decided to report the numbers in a very convoluted way (they don't give it in barrels or lbs, they give it in volume percentage based on bulk water volume reported - this leads to inconsistent numbers because the carrier fluid isn't just water, it is water and some other chemical, which differs well to well, that increases the density - which will mess with final volume/weight counts on the various chemicals). The other major issue is what you call a chemical. For example, one chemical can have 10 different names, and pumpers (field personnel) are entering the data by hand a lot of time. Some guy, taking mud samples for an oil company, working at a site 50 miles from Midland, isn't really caring how he is spelling "2-acryloyloxyethyl(benzyl)dimethylammonium chloride" at 4 in the morning. So, on top of chemicals with multiple names (and just field reference names), you have spelling errors galore. Also, when going through the database, there is no standardized way to enter percentages...nor do they care if those percentages add up to 1 (100?). The study does what it is supposed to do - identifies "bad stuff goes into the ground", but we really can't judge what the real amount is given how error full the database is. To finish a thought from before, the thing that E&P companies get from the database on competitors is how much sand (what kind of sand/proppant, too) and water they are using to fracture the rock in a well. The completion is the more expensive part of the drilling/completion/facilities cost of a well. If you are interested in seeing what the database looks like, here is the location. I haven't cleaned it in a bit, and it looks like they changed up the site some, so take this link FWIW. https://dev.fracfocus.org/index.php/data-download

[coffeeshopgoth]

Also worth mentioning, where the article talks about "proprietary recipes" - that isn't just referencing one chemical in many cases. They will just enter that phrase as the full list of chemicals used.

[loa_in_]

If I got it correctly, the units they use represent some upper/lower bounds on chemical mix percentage?

[coffeeshopgoth]

Sample Calculation: Total Base Water Volume (gal) (Total Bulk Volume): 16,489,452

Max Ingredient Concentration in HydroFracture Fluid for water (% by mass): 91.0

Max Ingredient Concentration in HF Fluid for silica (% by mass): 8.6

Gallon of water weight (lbs./gal): 8.34

Barrel (bbl) = 42 gallons

Total weight of water (lbs.) = TBV * Gallon of water weight = 137,522,030

Total weight of frac job (lbs.) = Total weight of Water/Water concentration in frac = 137,522,030 / (91.0/100) = 151,123,110

Proppant weight, Silica, (lbs.) = 151,123,110 * (8.6/100) = 12,996,588

Barrels of water = 16,489,452 / 42 = 392,606

So, 99.6% of the frac job is sand (or something like it) and water. The rest is "chemicals". Of course, this number varies wildly based on recipe - 92-99% is usually the range. You would figure out the other chemicals the same way. Problem is percentages could be in percentage or decimal format and/or those numbers could just be straight up wrong, or you aren't provided other info that is relevant in the calc.