Permissible Exposure Limits vs TLVs and RELs: Which Number Governs a Laboratory Chemical

An OSHA permissible exposure limit is the only airborne exposure number a laboratory can be cited for exceeding. It is also, for most chemicals on a laboratory bench, the least protective number available and the oldest. The great majority of the PELs in 29 CFR 1910.1000 were adopted in 1971 under the OSH Act’s two-year start-up authority, taken from Walsh-Healey Act contractor standards which had themselves been taken from the American Conference of Governmental Industrial Hygienists’ 1968 Threshold Limit Values. OSHA’s single serious attempt to modernise them, the 1989 Air Contaminants Standard, was vacated in its entirety by the Eleventh Circuit in 1992 and never replaced. The numbers a compliance officer measures you against are, for hundreds of substances, fifty-eight years old.

That produces a specific and awkward split for anyone running a chemical hygiene programme: the number you must not exceed and the number you should design controls against are frequently not the same number, sometimes by a factor of ten or twenty-five. This page sets out where each limit comes from, what it does and does not bind, exactly what the Eleventh Circuit vacated (which is widely misdescribed), and how 29 CFR 1910.1450(d) turns exposure limits into a monitoring duty that is structurally different from the one a production facility owes.

It is the third of three pages on the same regulatory seam. Hazard Communication (1910.1200) in the research laboratory covers which HazCom duties survive the laboratory exemption, and the written hazard communication program covers who owes one and how it nests inside the Chemical Hygiene Plan. This one extends the same 1910.1200-versus-1910.1450 boundary into exposure limits, where the supersession rule at 1910.1450(a)(2) does most of the work.

The four numbers, and which of them is law

Limit	Issued by	Legal status	Averaging basis
PEL — permissible exposure limit	Federal OSHA, 29 CFR 1910.1000 Tables Z-1/Z-2/Z-3 and the substance-specific standards	Enforceable. Exceeding it is a citable violation	8-hour TWA, with ceiling and peak provisions where specified
State-plan PEL	An OSHA-approved State Plan, e.g. Cal/OSHA’s AC-1 table	Enforceable in that state only, and must be at least as effective as the federal limit	8-hour TWA, plus STEL and ceiling entries
REL — recommended exposure limit	NIOSH, published in the NIOSH Pocket Guide to Chemical Hazards	Not enforceable. A federal agency recommendation transmitted to OSHA for use in rulemaking	TWA for up to a 10-hour workday in a 40-hour week
TLV — Threshold Limit Value	ACGIH, a private not-for-profit scientific association	Not enforceable. ACGIH states it is not a standards-setting body and that TLVs are not intended for use as legal standards	8-hour TWA, with STEL and ceiling notations

Two points in that table are load-bearing and routinely lost. First, the NIOSH REL and the OSHA PEL are not computed over the same shift. A REL is a time-weighted average for up to a ten-hour workday; a PEL is an eight-hour average in a forty-hour week. Comparing the two bare numbers without stating the basis is an apples-to-oranges comparison that a competent industrial hygienist will pick up immediately. Second, ACGIH derives TLVs from health data alone, with no consideration of technical or economic feasibility, and says explicitly that they should not be adopted as standards without a separate analysis of control options and feasibility. A TLV is a health benchmark, not a ready-made compliance target.

Where the PELs came from, and why they stopped moving

Section 6(a) of the Occupational Safety and Health Act of 1970 gave OSHA two years to promulgate “start-up” standards on an expedited basis, without hearing or comment, by adopting existing federal standards or national consensus standards. Using that authority, OSHA in 1971 promulgated approximately 425 PELs for air contaminants at 29 CFR 1910.1000, derived principally from the federal standards applicable to government contractors under the Walsh-Healey Act. Those Walsh-Healey standards had in turn been adopted from the ACGIH’s 1968 TLV recommendations; a smaller number came from published American Standards Association consensus standards. Comparable limits were adopted for shipyards at 29 CFR 1915.1000 and construction at 29 CFR 1926.55.

Since 1970, OSHA has completed full section 6(b) rulemakings producing new PELs for 16 agents, plus standards without PELs for 13 carcinogens. OSHA states the position itself, on its own Annotated PELs pages: “OSHA recognizes that many of its permissible exposure limits (PELs) are outdated and inadequate for ensuring protection of worker health.”

The scale of the freeze is easy to underestimate. Table Z-1 as currently published runs to roughly 610 substance rows, of which about 90 are not limits at all but cross-references pointing at a substance-specific standard elsewhere in Subpart Z (“Benzene; see 1910.1028”, “Formaldehyde; see 1910.1048”). The remainder are, with few exceptions, the 1971 values.

What the Eleventh Circuit actually vacated

This is the part most secondary summaries get wrong, so it is worth stating precisely.

In January 1989 OSHA issued a revised Air Contaminants Standard setting permissible exposure limits for 428 toxic substances in a single multi-substance rulemaking (54 Fed. Reg. 2332). Ten consolidated petitions for review followed, brought both by industry petitioners and by the AFL-CIO — the union argued the new limits were insufficiently protective and the compliance timetable too generous, while industry argued the opposite. In AFL-CIO v. OSHA, 965 F.2d 962 (11th Cir. 1992), decided 7 July 1992, the court agreed with both sides on procedure and held:

“OSHA’s overall approach to this rulemaking is so flawed that we must vacate the whole revised Air Contaminants Standard.”

Three distinctions matter for anyone citing this case:

• The 1989 revision was vacated, not “OSHA’s PELs.” Striking the revision left the pre-1989 values — the 1971 start-up limits — standing as the enforceable numbers. The practical effect was reversion, not repeal.
• The court did not find the substances harmless. It expressly held that the record “adequately explains and supports OSHA’s determination that the health effects of exposure to these 428 substances are material impairments.” What OSHA had not adequately supported, substance by substance, was the threshold finding under section 3(8) that a significant risk of those impairments existed at prior exposure levels (the requirement the Supreme Court drew from the OSH Act in the Benzene case), or the section 6(b)(5) finding that the new limits reduced that risk to the extent feasible.
• The defect was the generic method, not the science. The court accepted that OSHA may combine multiple substances in one rulemaking and set priorities under section 655(g). What it would not accept was substance-by-substance findings replaced by an across-the-board rationale: “OSHA has lumped together substances and affected industries and provided such inadequate explanation that it is virtually impossible for a reviewing court to determine if sufficient evidence supports the agency’s conclusions.” The only exceptions to section 6(b)’s findings requirements, the court noted, are the section 6(a) start-up provisions — “the applicability of which has long since passed” — and the section 6(c) emergency provisions.

That last sentence is why the 1971 limits have proved so durable. The mechanism that produced them in bulk expired in 1973, and the mechanism that remains requires a full risk-and-feasibility record per substance.

How to read a Z-table entry

Three tables, three grammars, at 29 CFR 1910.1000:

• Table Z-1. A limit with no prefix is an 8-hour time-weighted average not to be exceeded in any 8-hour shift of a 40-hour week. A limit prefixed “C” is a ceiling that must not be exceeded at any time; where instantaneous monitoring is not feasible, 1910.1000(a)(1) directs that the ceiling be assessed as a 15-minute TWA that must not be exceeded at any point in the working day. Both ppm and mg/m³ columns are given.
• Table Z-2. A three-part structure: an 8-hour TWA, an acceptable ceiling concentration, and an acceptable maximum peak above that ceiling for a stated maximum duration. OSHA’s own worked example at (b)(3): for a substance with a 10 ppm TWA, 25 ppm ceiling and 50 ppm peak, an employee may be above 25 ppm (never above 50) for at most 10 minutes, and that excursion must be offset by time below 10 ppm so the shift average still comes in at or under 10 ppm. Several common laboratory solvents sit here rather than in Z-1 — toluene, methylene chloride, carbon tetrachloride, trichloroethylene, and mercury in all its forms.
• Table Z-3. Mineral dusts, as an 8-hour TWA, with the respirable-fraction formulae for silica and the accompanying aerodynamic-diameter table.

The mixture rule almost nobody applies

Paragraph 1910.1000(d)(2) is the provision most laboratories never reach and most often breach in principle. Where an employee is exposed to a mixture of air contaminants, the equivalent exposure is computed as the sum of the ratios of each measured concentration to its own limit:

E_m = (C₁ ÷ L₁) + (C₂ ÷ L₂) + … + (C_n ÷ L_n)

where C is the measured concentration of each contaminant and L is that contaminant’s own 8-hour TWA limit. The value of E_m must not exceed 1. A bench running a three-solvent extraction can therefore be over the regulatory limit while every individual solvent reads comfortably under its own PEL. Paragraph (d)(1) gives the separate single-substance formula, E = (C_aT_a + C_bT_b + …) ÷ 8, for a shift with varying concentrations. Neither formula appears in most institutional chemical hygiene plans.

The cross-walk: how far apart the numbers actually are

The table below is drawn from OSHA’s own Annotated Table Z-1, which publishes the federal PEL, the Cal/OSHA PEL and the NIOSH REL side by side. The ACGIH TLV column is deliberately omitted here: TLVs are copyrighted by ACGIH, are not publicly available, and may not be reproduced without permission. For TLV values, consult ACGIH directly; OSHA’s annotated tables link out to ACGIH for each substance.

Substance	Federal OSHA PEL (8-hr TWA)	Cal/OSHA PEL	NIOSH REL (up to 10-hr TWA)	Ratio, PEL : REL
n-Hexane	500 ppm	50 ppm	50 ppm	10×
Chloroform	C 50 ppm (ceiling)	2 ppm	Ca; STEL 2 ppm [60 min]	25× (ceiling vs STEL)
Acetone	1,000 ppm	500 ppm; STEL 750; ceiling 3,000	250 ppm	4×
Acetonitrile	40 ppm	40 ppm; STEL 60 ppm	20 ppm	2×
Ammonia	50 ppm	25 ppm; STEL 35 ppm	25 ppm; STEL 35 ppm	2×
Methyl alcohol	200 ppm	200 ppm; STEL 250; ceiling 1,000	200 ppm; STEL 250 ppm	1×
Ethyl alcohol	1,000 ppm	1,000 ppm	1,000 ppm	1×
Dimethylformamide	10 ppm	10 ppm	10 ppm	1×
Pyridine	5 ppm	5 ppm	5 ppm	1×
Hydrogen peroxide	1 ppm	1 ppm	1 ppm	1×
Phenol	5 ppm	5 ppm	5 ppm; ceiling 15.6 ppm [15 min]	1× TWA, ceiling added
Sulfuric acid	1 mg/m3	0.1 mg/m3; STEL 3 mg/m3	1 mg/m3	1× federal, 10× in California
Ethyl ether	400 ppm	400 ppm; STEL 500 ppm	No numeric REL (NIOSH Appendix D)	STEL added by Cal/OSHA

“Ca” is NIOSH’s potential-occupational-carcinogen notation. Two things fall straight out of this table and both are decision-relevant. The gap is not uniform — for ethanol, DMF, pyridine, hydrogen peroxide and methanol the three bodies agree, and treating “the PEL is obsolete” as a blanket rule will waste control money on substances where it is false. Where the gap is large it is very large, and it clusters on the neurotoxins and the carcinogens: n-hexane at ten times, chloroform at twenty-five. And the Cal/OSHA column is a reminder that in a State Plan jurisdiction the stricter number is not a recommendation at all — sulfuric acid mist at 0.5 mg/m³ is compliant federally and a violation in California.

The actual decision: which number to design controls against

The honest answer is that these are two different questions with two different answers, and conflating them is what produces both over-engineering and under-protection.

1. Your compliance floor is the PEL. Always.

For laboratory uses of OSHA-regulated substances, 1910.1450(c) requires the employer to assure that exposures do not exceed the PELs specified in Subpart Z. That is the citable line, it is non-negotiable, and no consensus limit displaces it in either direction. A lower TLV does not raise your legal obligation; a higher one would not lower it.

2. Your design target should be the lowest applicable published OEL.

The Chemical Hygiene Plan requirement is not written as a PEL-compliance requirement. 1910.1450(e)(1) requires a plan “capable of protecting employees from health hazards” and capable of keeping exposures below the (c) limits; (e)(3)(ii) requires criteria the employer will use to determine and implement control measures “to reduce employee exposure,” with particular attention to chemicals known to be extremely hazardous. Nothing in that text stops at the PEL. In practice, the defensible design target is the lowest of: the NIOSH REL, the ACGIH TLV, the Cal/OSHA PEL (or your own state-plan limit), and any manufacturer OEL appearing in Section 8 of the safety data sheet — noting the ten-hour versus eight-hour basis difference before you compare.

Section 8 of every SDS you receive already carries these. 1910.1200 Appendix D, Table D.1 requires the OSHA PEL, the ACGIH TLV, and any other exposure limit used or recommended by the manufacturer, importer or employer preparing the sheet. If you are reading safety data sheets and skipping Section 8, the cross-walk you need is already in the document.

3. If there is no PEL at all, the framework is the General Duty Clause

Section 5(a)(1) of the OSH Act requires each employer to furnish a workplace “free from recognized hazards that are causing or are likely to cause death or serious physical harm.” It reaches chemicals for which no PEL exists — and hundreds of chemicals in routine laboratory use have none. But the standard of proof is high. OSHA’s Field Operations Manual and its 2 November 2018 enforcement memorandum on respiratory hazards not covered by PELs set out four elements OSHA must establish: exposure to a hazard the employer failed to keep the workplace free of; recognition of that hazard; that it was causing or likely to cause death or serious physical harm; and a feasible and useful method of correction.

The memorandum is unusually explicit about what will not do: citations “are not based solely on evidence that a measured exposure exceeded a recommended occupational exposure limit (OEL), such as a Threshold Limit Value,” and proving serious physical harm “must include more than just the fact that a measured exposure exceeded a TLV or REL, because these recommended limits may be much lower than the level at which a serious health effect may be experienced.” Where the four elements are not all proven, the Area Office issues a hazard alert letter rather than a citation — a real and under-appreciated outcome that carries no penalty but does create a documented recognition record for any future inspection.

4. If a PEL exists but is inadequate, the bar is actual knowledge

The Field Operations Manual states that section 5(a)(1) “shall normally not be used to impose a stricter requirement than that imposed by the OSHA standard,” but that where an existing standard is inadequate a citation can be considered if all four elements are met and “there must be actual employer knowledge that the standard was inadequate to protect employees from death or serious physical harm,” citing Int’l Union UAW v. Gen. Dynamics Land Sys. Div., 815 F.2d 1570 (D.C. Cir. 1987). Its worked example is directly on point: with a PEL of 15 ppm and a recognised OEL of 3 ppm, a citation can be considered only for exposures between the OEL and the PEL where the data establishes likely death or serious physical harm at the measured level and the employer has actual knowledge that the PEL is inadequate. Such cases are subject to pre-citation review by OSHA’s national office, and they are rare — OSHA reported to Congress in 2003 that only five section 5(a)(1) citations referencing OELs had issued in the preceding two years.

There is an uncomfortable corollary that safety officers should face squarely. The 2018 memorandum lists, as evidence of employer recognition, “internal safety and health policies related to workplace operations involving the chemical that may refer to an OEL.” Writing a TLV or REL into your Chemical Hygiene Plan as your internal control limit is exactly such evidence. That is not an argument for leaving it out — the recognition it creates is the same recognition every SDS in the building already creates, and a plan that targets only the 1971 PEL is poor practice on its own terms. It is an argument for setting an internal limit you actually intend to meet, and for documenting the controls that meet it, rather than adopting a number aspirationally and leaving the fume hood at 60 fpm. See fume hood certification and inspection and ANSI/AIHA Z9.5 for the performance side of that commitment.

How the Laboratory Standard changes the monitoring duty

This is where laboratory practice diverges sharply from the rest of general industry, and it is governed by the supersession rule at 1910.1450(a)(2), not by the exposure limits themselves.

Where the Laboratory Standard applies, it “shall supersede, for laboratories, the requirements of all other OSHA health standards in 29 CFR part 1910, subpart Z,” with three exceptions:

1. (a)(2)(i) — for any OSHA health standard, “only the requirement to limit employee exposure to the specific permissible exposure limit shall apply for laboratories,” unless that standard says otherwise or (a)(2)(iii) applies.
2. (a)(2)(ii) — prohibitions on eye and skin contact specified by any OSHA health standard still apply.
3. (a)(2)(iii) — where the action level (or, absent an action level, the PEL) “is routinely exceeded for an OSHA regulated substance with exposure monitoring and medical surveillance requirements,” paragraphs (d) and (g)(1)(ii) of the Laboratory Standard apply.

The consequence is concrete. A production facility using formaldehyde owes 1910.1048 in full, and 1910.1048(d)(1)(i) requires it to monitor employees, full stop — initial monitoring is mandatory, not conditional. A research laboratory using formaldehyde at laboratory scale owes, from that standard, the PEL (0.75 ppm 8-hour TWA, 2 ppm 15-minute STEL) and the eye and skin provisions. Its monitoring duty comes instead from 1910.1450(d).

What 1910.1450(d) actually requires

Paragraph (d)(1) reads: the employer “shall measure the employee’s exposure to any substance regulated by a standard which requires monitoring if there is reason to believe that exposure levels for that substance routinely exceed the action level (or in the absence of an action level, the PEL).”

Two conditions, both of which must be satisfied before any obligation to sample arises:

• The substance must be regulated by a standard which requires monitoring. A Z-1-only substance has a PEL but 1910.1000 imposes no monitoring requirement, so (d)(1) does not command initial monitoring for acetone, acetonitrile, methanol or n-hexane no matter what your exposure looks like. The duty attaches to the substances with their own standards — formaldehyde (1910.1048), benzene (1910.1028), methylene chloride (1910.1052), lead, cadmium, chromium(VI), and the rest of the Subpart Z substance-specific list.
• There must be reason to believe the trigger is routinely exceeded. The trigger is the action level where the substance has one, and the PEL where it does not. “Action level” is defined at 1910.1450(b) as “a concentration designated in 29 CFR part 1910 for a specific substance, calculated as an eight (8)-hour time-weighted average, which initiates certain required activities such as exposure monitoring and medical surveillance.” It is a number OSHA sets in the substance standard, not one the laboratory chooses. Formaldehyde: 0.5 ppm. Benzene: 0.5 ppm. Methylene chloride: 12.5 ppm — in each case half the 8-hour PEL, which is the usual but not universal pattern.

The rest of paragraph (d) follows mechanically. Under (d)(2), if initial monitoring discloses exposure over the action level (or PEL), the employer must immediately comply with the exposure monitoring provisions of the relevant substance-specific standard — the full 1910.1048 or 1910.1028 monitoring regime switches on at that moment, which is what (a)(2)(iii) means in practice. Under (d)(3) monitoring may be terminated in accordance with that relevant standard. Under (d)(4) the employer must notify the employee of any monitoring results in writing, individually or by posting, within 15 working days of receiving them — a hard deadline with no exposure-level threshold attached, and the paragraph laboratories most often miss.

Paragraph (g)(1)(ii) is the parallel medical trigger: where exposure monitoring reveals a level routinely above the action level (or PEL) for a regulated substance with monitoring and medical surveillance requirements, medical surveillance must be established for the affected employee as prescribed by that particular standard.

The judgment this places on the Chemical Hygiene Officer

“Reason to believe” is not a sampling schedule. Nothing in 1910.1450 sets a monitoring frequency, and there is no periodic obligation until initial monitoring finds something. What the standard does is place an assessment on the employer, discharged in practice by the Chemical Hygiene Officer, that has to be defensible after the fact. The inputs a compliance officer will look for are the ordinary ones: quantities and volatility, whether the work is contained in a fume hood with documented performance, frequency and duration of the task, the results of any job hazard analysis, and prior sampling in comparable operations. Recording that assessment — including the reasoned conclusion that there was no reason to believe the action level is routinely exceeded — is what turns a judgment call into a documented determination, and it belongs in the Chemical Hygiene Plan rather than in an inspector’s notebook.

Note also the boundary condition. 1910.1450(a)(3)(i) removes the whole supersession benefit where the use does not meet the definition of “laboratory use of hazardous chemicals” — a repetitive single-assay operation functioning as a step in a production process is not laboratory use, and the full substance-specific standard, mandatory monitoring included, applies even though the work is happening at a bench. The same four-condition test governs the HazCom laboratory exemption.

A working checklist

1. For each hazardous chemical in routine use, record the federal PEL, the state-plan PEL if you are in a State Plan jurisdiction, the NIOSH REL and the ACGIH TLV. Section 8 of the SDS gives you the PEL and TLV; OSHA’s Annotated Z-tables give you the Cal/OSHA PEL and NIOSH REL.
2. Note the averaging basis alongside each value. Eight-hour TWA, up-to-ten-hour TWA, 15-minute STEL and ceiling are not interchangeable.
3. Flag every chemical where the lowest OEL is materially below the PEL. Those are the substances where PEL compliance and adequate protection genuinely diverge and where control decisions need a rationale.
4. Set the internal control target in the Chemical Hygiene Plan at the lowest applicable OEL, and be able to show the controls that meet it. Do not set a target you have no intention of engineering to.
5. Identify which of your chemicals are covered by a substance-specific standard that requires monitoring. Those, and only those, can trigger 1910.1450(d)(1).
6. For each of those, record the action level and the reasoned determination of whether there is reason to believe it is routinely exceeded. Date it, name the assessor, and revisit it when the process, scale or personnel change.
7. If you monitor, diarise the 15-working-day written notification under (d)(4) as a hard deadline.
8. Where a chemical has no PEL, document the recognised-hazard analysis and the control basis on the same footing — the General Duty Clause is where that exposure lives.

For the wider programme these obligations sit inside, see the laboratory compliance hub, the particularly hazardous substances requirements at 1910.1450(e)(3)(viii), and PPE selection for chemical handling. Where controls cannot bring exposure below the PEL and respiratory protection is used, respirator fit testing under 1910.134 applies in full.

Frequently asked questions

Are ACGIH TLVs legally enforceable?

No. ACGIH is a private, not-for-profit scientific association, not a standards-setting body, and states that TLVs are health-based values not intended for use as legal standards. OSHA cannot cite an employer for exceeding a TLV as such. A TLV can appear as evidence supporting a General Duty Clause case, but OSHA’s own enforcement policy states that a section 5(a)(1) violation would not be based solely on a TLV.

Are NIOSH RELs enforceable?

No. NIOSH is a research agency with a statutory mandate to recommend standards to OSHA; RELs are those recommendations, published in the NIOSH Pocket Guide to Chemical Hazards. They carry federal-agency authority as scientific recommendations and OSHA lists NIOSH ahead of ACGIH among the sources it consults, but they are not regulatory limits.

Why are OSHA PELs so out of date?

Most were adopted in 1971 under the OSH Act’s section 6(a) start-up authority from Walsh-Healey contractor standards, which derived from the 1968 ACGIH TLVs. Updating them now requires full section 6(b) rulemaking with substance-by-substance findings of significant risk and feasibility. OSHA’s one attempt to do 428 substances at once, the 1989 Air Contaminants Standard, was vacated in AFL-CIO v. OSHA in 1992 precisely for not making those findings substance by substance.

Did AFL-CIO v. OSHA strike down OSHA’s permissible exposure limits?

No. It vacated the 1989 revision of the Air Contaminants Standard, which had the effect of reverting the enforceable limits to their pre-1989 values. Table Z-1 remains in force and enforceable today; it simply contains the older numbers. The court also expressly accepted OSHA’s finding that the health effects at issue were material impairments.

Which limit should I use if the PEL and the TLV differ?

Both, for different purposes. The PEL is the compliance floor you must not exceed. The lowest applicable OEL — REL, TLV, or state-plan PEL — is the sensible design target for engineering controls, and 1910.1450(e)(3)(ii)’s obligation to implement measures that reduce exposure is written to support exactly that. Where they differ by an order of magnitude, document why your control decision landed where it did.

Can OSHA cite a laboratory when exposure is below the PEL?

Rarely, and only under the General Duty Clause. OSHA’s Field Operations Manual permits a section 5(a)(1) citation for exposures between a recognised OEL and the PEL only where all four general-duty elements are proven and the employer has actual knowledge that the PEL is inadequate to protect employees, following Int’l Union UAW v. Gen. Dynamics. Such cases require pre-citation national-office review. The more common outcome is a hazard alert letter, which carries no penalty.

Does a research lab have to do exposure monitoring under 1910.1450?

Only conditionally. 1910.1450(d)(1) requires initial monitoring where the substance is regulated by a standard that itself requires monitoring and there is reason to believe exposure routinely exceeds the action level, or the PEL where no action level exists. For a Z-1-only substance there is no monitoring standard to trigger. If initial monitoring does find exposure over the trigger, (d)(2) switches on the full monitoring regime of the substance-specific standard immediately.

What is an action level, and can a laboratory set its own?

No. Under 1910.1450(b) an action level is “a concentration designated in 29 CFR part 1910 for a specific substance, calculated as an eight (8)-hour time-weighted average, which initiates certain required activities such as exposure monitoring and medical surveillance.” It is set by OSHA in the substance-specific standard — 0.5 ppm for formaldehyde and benzene, 12.5 ppm for methylene chloride. A laboratory may adopt a stricter internal control limit, but that is a control target, not an action level, and it does not alter the regulatory trigger.

Do state OSHA plans have stricter exposure limits?

Some do. An OSHA-approved State Plan must be at least as effective as the federal programme and may be stricter. California has the most extensive list of state occupational exposure limits, published as the Cal/OSHA AC-1 table and enforceable in workplaces under Cal/OSHA jurisdiction. Check your own state plan before assuming the federal Z-table number is the operative one.

Do PELs apply to laboratory-scale work at all?

Yes. 1910.1450(c) requires that for laboratory uses of OSHA-regulated substances, employee exposures do not exceed the PELs specified in Subpart Z. Laboratory scale changes the surrounding regime — monitoring, medical surveillance, regulated areas — but not the exposure limits themselves.

Primary sources

• 29 CFR 1910.1000, Air contaminants — eCFR (Table Z-1 ceiling/TWA rules at (a); Table Z-2 peak provisions and worked example at (b); Table Z-3 at (c); computation and mixture formulae at (d))
• 29 CFR 1910.1450, Occupational Exposure to Hazardous Chemicals in Laboratories — eCFR (supersession at (a)(2); action level definition at (b); PELs at (c); employee exposure determination at (d); Chemical Hygiene Plan at (e); medical consultation at (g))
• Substance-specific standards, action levels and PELs verified against eCFR: 1910.1048 formaldehyde (AL 0.5 ppm; PEL 0.75 ppm TWA / 2 ppm STEL), 1910.1028 benzene (AL 0.5 ppm; PEL 1 ppm TWA / 5 ppm STEL), 1910.1052 methylene chloride (AL 12.5 ppm; PEL 25 ppm TWA / 125 ppm STEL)
• AFL-CIO v. OSHA, 965 F.2d 962 (11th Cir. 7 July 1992) — full opinion text, vacating the 1989 Air Contaminants Standard, 54 Fed. Reg. 2332
• OSHA, Permissible Exposure Limits — Annotated Tables — osha.gov/annotated-pels (origin of the 1971 PELs; OSHA’s own statement that many PELs are outdated; descriptions of Cal/OSHA PELs, NIOSH RELs and ACGIH TLVs)
• OSHA Field Operations Manual, CPL 02-00-163, Chapter 4 — general duty clause elements and Example 4-25
• OSHA memorandum, “Enforcement Policy for Respiratory Hazards Not Covered by OSHA Permissible Exposure Limits,” 2 November 2018 — osha.gov
• OSHA letter of interpretation, “Use of occupational exposure levels (OELs) for substances not covered by an OSHA Permissible Exposure Limit,” 24 January 2003 — osha.gov
• ACGIH TLV values are copyrighted and not reproduced here — acgih.org

Regulatory text and OSHA guidance summarised as published and read on 26 August 2026. The NIOSH Pocket Guide pages at cdc.gov returned HTTP 403 to automated retrieval during preparation of this page; NIOSH REL values quoted here are taken from OSHA’s own Annotated Z-tables rather than from the Pocket Guide directly, and readers should consult NIOSH for current values and skin/sensitisation notations. OSHA-approved state plans may impose limits at least as effective as, and in some cases stricter than, the federal standard. This is a reference for research administrators and laboratory safety personnel, not legal advice.