Reading a Safety Data Sheet for What It Actually Tells You
Why this matters
A pail of coil cleaner and a jug of loop inhibitor both arrive with a safety data sheet, and in most shops both sheets go into a binder that gets opened after somebody is already hurt. Read that late, the sheet is a legal record. Read before the container is opened, it is the only document on the truck that names which glove material survives what is inside, what the product must never share a drain or a measuring jug with, and whether the hazard reaches you through skin or through air. Those are three different controls and picking the wrong one leaves you fully equipped and fully exposed.
The sheet fails as a working document for a structural reason, not a writing one. Its section order was fixed for emergency responders, so the questions a responder asks are at the front and the questions a user asks are in the middle. If you read it front to back and stop when you get bored, you stop one section before the part you needed.
The order is responder order, not user order
The 16-section format is standardized under the OSHA hazard communication standard, and because 29 CFR 1910.1200(g) requires the sheet to be readily accessible to employees in their work area during each work shift, a sheet that lives only in the shop office does not satisfy it for a tech opening the pail on a roof. Carry it or carry a device that reaches it.
Map the sections to the question each one answers, and the reading order stops being a mystery:
| Section | Answers | Who it was written for |
|---|---|---|
| 2 Hazards | What class of harm, and the signal word | Everyone, which is why it is over-read |
| 3 Composition | What is actually in it, and at what range | You, for whether the dilution rules apply and whether your sheet is current |
| 4 First aid | What to do in the first minutes | Responder, and the injured person |
| 7 Handling and storage | What it must not sit next to | You, before the truck is loaded |
| 8 Exposure controls and PPE | Glove material class, eye and face, respiratory | You, before the lid comes off |
| 9 Physical and chemical properties | pH, density, flash point, vapor pressure | You, for everything downstream |
| 10 Stability and reactivity | What it must never contact | You, and the drain |
| 11 Toxicological information | Which route actually matters | You, when picking a control |
Sections 12 through 15 are environmental, disposal, transport and regulatory. Sections 12 and 13 are the ones that decide whether the rinse water is allowed in a floor drain, which is a real operating question and is covered in the companion card on discharge.
Section 8 is the operating section, and it is incomplete on purpose
Section 8 will name a glove material class, eye and face protection, and a respiratory control where one is needed. What it will almost never give you is a breakthrough time, which is the number that decides whether the glove is right for your job rather than for a splash.
Breakthrough time is measured under a standardized permeation test: a sample of the glove material is exposed to the chemical on one side and the time until detectable permeation on the other side is recorded, on a flat sample at room temperature with continuous contact. Two conditions in that sentence are load-bearing. The rating is for a flat unstretched sample, and the material at a knuckle is thinner and stretched. The rating is at room temperature, and permeation rate climbs steeply with temperature, so a glove rated in a lab bench test is optimistic on a hot descaling job.
So a glove class from section 8 plus a breakthrough time from the glove manufacturer's permeation chart is one control. Section 8 alone is half of one. This is not a gap in the sheet, it is a division of labor: 29 CFR 1910.138 requires hand protection to be selected for the specific task, the specific chemical and the duration of contact, and only your job knows the duration.
The same split applies to respiratory protection. If section 8 names a cartridge type, you are being told a program is required, not that a cartridge solves it. Respirator use for employees sits under 29 CFR 1910.134, which requires medical evaluation before use and annual fit testing at 1910.134(f)(2), so a cartridge mask handed out of the truck box without either is not a control, it is a costume.
The sheet describes the product as supplied
This is the single most-missed line on the document, and everything downstream turns on it. Every value in sections 9 and 11 was measured on the product in the container. If your procedure dilutes it, heats it, sprays it, or lets it sit open in a closed mechanical room, you are working with something the sheet did not measure.
Dilution moves some properties and not others, and the direction is worth being exact about. For a strong acid that is fully dissociated, a tenfold dilution raises pH by about one unit, so a product listed below pH 2 lands near pH 3 at one part in ten. That is still acidic enough to injure an eye. For a buffered blend or a weak acid, the same tenfold dilution barely moves pH at all, because the buffer resists it, so do not carry the one-unit rule across to a product whose section 3 lists a buffering salt.
What dilution does not reduce at all is the hazard of the concentrate you are still handling while you dilute. The most dangerous minute of a dilution is the minute the concentrate is in the air and in motion. Add acid to water and never water to acid. The heat of mixing is the same either way; what differs is what absorbs it. A drum of water takes that energy and warms. A few inches of water hit by a stream of acid takes the same energy into a fraction of the mass, boils at the interface, and throws concentrated acid back out. Wear the chemical splash goggle and face shield plus the glove class section 8 names for the concentrate, not for the finished dilution, and do it where an eyewash meeting the 15-minute flush duration in ANSI/ISEA Z358.1 is within seconds of walking distance, which 29 CFR 1910.151(c) requires wherever a person may be exposed to injurious corrosive materials.
What the sheet will not answer, and where those answers live
The sharpest use of an SDS is knowing when to stop reading it and go somewhere else. Five questions techs routinely try to answer from the sheet, and the actual source for each:
Will this attack my system's materials? Not on the sheet. Section 10 lists incompatible materials in broad classes for storage purposes, not a corrosion rate against your specific alloy or elastomer at your temperature. Material compatibility is its own investigation and the library has a card on which compatibility checks produce evidence and which ones falsely pass.
Is my ventilation adequate? Not on the sheet. Section 8 may list an exposure limit, which is a number to compare a measurement against, not a substitute for the measurement. In a mechanical room with no measurement, treat any vapor-generating task as requiring outdoor air movement and get the task out of the confined space rather than reasoning about it.
Can I mix it with the other product on my truck? Partly. Section 10 gives classes to avoid. It does not know your other product's formulation. The reliable read is the active-ingredient class of each, covered in the companion card on chemical incompatibility.
How long can my glove stay in it? Not on the sheet. Glove manufacturer permeation data, matched to your contact duration.
What is the residual after I rinse? Not on the sheet. That depends on your rinse volume and the surface, and it is the question that decides whether a potable-adjacent surface is safe to return to service.
Noticing that a question is off-sheet is faster than hunting through 16 sections for an answer that was never printed there.
Worked example: one sheet, one decision
A tech is quoted a descaling pass on a heat exchanger and picks up an acidic descaler. The sheet is illustrative, and the values here are the kind a sheet in this class carries rather than a specific product's:
- Section 2: corrosive to metals, causes severe skin burns and eye damage, may cause respiratory irritation.
- Section 3: an inorganic acid blend, stated in the 20 to 30 percent range, plus an inhibitor package.
- Section 8: chemical splash goggles and face shield, butyl or a named acid-resistant polymer glove, respiratory protection where mist is generated.
- Section 9: pH below 2 as supplied.
- Section 10: reacts with bases, with hypochlorite, and with active metals to release hydrogen.
The procedure calls for a 1 part product to 9 parts water charge, circulated warm for about 3 hours, then a neutralize and rinse.
Read it in the order the job needs. Section 10 first, because it is the one that kills people: hypochlorite is named, and the shop's coil cleaning kit and the customer's pool shed both carry hypochlorite. That is a segregation decision made before the truck is loaded, not on site.
Then section 9 against the dilution. Below pH 2, diluted tenfold, lands near pH 3 for a fully dissociated acid, and the inhibitor package tells you this blend is not a bare strong acid so the shift may be smaller. Either way the circulating solution is corrosive to eyes and to skin over time, which means the goggle and glove do not come off after mixing.
Then section 8 against the clock. A 3-hour circulation with hands in and out of the loop is not a splash exposure, it is repeated contact. A glove class with a 30-minute breakthrough time against that acid class covers 30 minutes of continuous contact under bench conditions, so on a 3-hour job you either move up a material class or change gloves at a fraction of the rated time. Running one pair for the full 3 hours is a six-fold overrun of the only number that mattered, and the failure is silent: permeation does not look like a hole.
Then section 2 against the route. "May cause respiratory irritation" plus a warm circulating acid is an inhalation route, and warm liquid raises vapor pressure, so the mist question is live in a way it would not be on a cold pass. That is a ventilation and respiratory decision under a 1910.134 program, and a glove does nothing for it. If the shop has no program, the honest call is to do this pass with the space ventilated to outdoors and the tech out of the vapor path, or to subcontract it.
The failure mode if this reading is skipped: the tech reads section 2, sees "corrosive," puts on the nitrile gloves already in the truck, mixes near a floor drain, and finishes the job. Nothing visibly happens. Three months later the same tech does the same job in a smaller room and has a coughing fit, and the shop learns the control was never in place on either job.
How to verify you read it right
Before the container opens, you should be able to answer four questions out loud without reopening the sheet: what route hurts me here, what does this never touch, what is different about the diluted or heated form, and how long can my protection actually last. If any answer is "the sheet says be careful," you have read section 2 and stopped.
One more check that catches stale sheets: compare the product name and the manufacturer on the container against the sheet in your hand. Reformulations happen and the sheet in the binder can be a version behind, which usually shows up as a changed percentage range in section 3. If the ranges disagree, the container wins and you get a current sheet before you use it.
References
- 29 CFR 1910.1200, hazard communication, including the accessibility requirement at (g) and the 16-section format in Appendix D
- 29 CFR 1910.138, hand protection selection based on task, chemical and duration; 29 CFR 1910.133, eye and face protection
- 29 CFR 1910.134, respiratory protection program requirements, including annual fit testing at (f)(2)
- 29 CFR 1910.151(c), quick drenching and flushing facilities; ANSI/ISEA Z358.1 for the 15-minute flush duration
- See related: Chemical Compatibility and the Checks Worth Running; Why Two Treatment Chemicals Can Be Incompatible; Acid + Caustic Chemical Handling Reference