Week 1 — Confined Space Awareness: Why "Just a Quick Look" Keeps Killing Workers

As a safety professional and expert witness, I get the call about confined‑space incidents more often than any other category of fatality I review. The fact pattern is almost always the same. A worker enters a manhole, vault, tank, lift station, or pit. Something goes wrong inside — usually a low‑oxygen atmosphere or a buildup of hydrogen sulfide. A coworker sees them collapse and climbs in to help. Within minutes, two people are dead instead of one.

OSHA has been telling us how to prevent this since 1993. The standard is 29 CFR 1910.146 (Permit‑Required Confined Spaces) for general industry, and 29 CFR 1926 Subpart AA (specifically 29 CFR 1926.1200–1213) for construction. In plain language, both standards say the same thing: before anyone enters a space large enough to fit a body but not designed for continuous occupancy, you have to identify the hazards, control them, test the atmosphere, post an attendant, and have a rescue plan ready that does not require entry by an untrained bystander.

When I review these cases, the gaps are rarely exotic. They are the same gaps, on every site.

The Three Things Almost Every Fatality Has in Common

Across the cases I have testified on, three failures show up over and over:

• The space was never formally identified as a permit‑required confined space. Crews called it "just a vault" or "the cleanout." Without that classification, no permit, no testing, and no rescue plan ever existed.

• The atmosphere was never tested before entry. Or it was tested once at the opening, but not in the bottom of the space where heavier‑than‑air gases collect. Oxygen, then flammables, then toxics — that is the testing order required by 1910.146(c)(5)(ii)(C), and it is constantly skipped.

• The "rescue plan" was 911. Calling the local fire department is not a rescue plan. By the time a non‑specialized fire response arrives, sets up, and effects a non‑entry rescue, the worker inside has been without oxygen for fifteen to twenty minutes. The standard at 1910.146(k) requires evaluation of the rescue service's ability to respond in a timeframe that matches the hazards in the space — and "in a timeframe" almost always means minutes, not the half hour or more it takes a typical 911 response to stage.

If a program has all three of these gaps, the next incident is not a question of if. It is a question of when.

What 29 CFR 1910.146 Actually Requires

Stripped of legalese, 1910.146 says an employer must do the following before anyone enters a permit‑required confined space:

1. Evaluate every space on the worksite and decide whether it meets the definition of a confined space, and whether it meets the additional criteria that make it permit‑required (hazardous atmosphere, engulfment hazard, internal configuration that could trap or asphyxiate, or any other recognized serious safety or health hazard).

2. Develop a written permit program that covers entry procedures, atmospheric testing, ventilation, communication, attendants, entrants, entry supervisors, and rescue.

3. Issue a written permit for every entry , signed by an entry supervisor, listing the hazards present, the acceptable atmospheric ranges (typically oxygen 19.5–23.5%, flammables under 10% LEL, toxics within published limits like 10 ppm H₂S), and the controls in place.

4. Test the atmosphere before entry, and continuously during entry if conditions could change. Test in the order: oxygen, flammables, toxics. Test at multiple depths.

5. Maintain a trained attendant outside the space at all times during entry. The attendant cannot leave, cannot enter, and is responsible for monitoring and summoning rescue.

6. Have a rescue capability that can perform non‑entry retrieval where feasible , and where it is not feasible, have trained entry rescuers who can be on scene within the time the hazards demand.

In construction, 29 CFR 1926.1200–1213 mirrors most of this, with extra emphasis on multi‑employer worksites — the controlling contractor has affirmative duties to coordinate confined‑space entries among subs, share hazard information, and ensure each sub knows what other crews are doing in the same or adjacent spaces.

Water and Wastewater: The Sector Where This Bites Hardest

I do a lot of work in water and wastewater, and confined‑space incidents drive a disproportionate share of fatalities in that sector. A few reasons:

• Process atmospheres are inherently hazardous. Hydrogen sulfide from anaerobic decomposition, methane from digesters, depleted oxygen in headspaces filled with nitrogen blanketing — these are normal operating conditions in wastewater systems, not exotic upset scenarios.

• Geographic dispersion. Lift stations, valve vaults, and force‑main cleanouts are scattered across miles of collection system. Crews drive to a remote vault, pop the lid, and the rescue resources are an hour away.

• Routine work feels routine. A two‑person crew has opened that same wet well five hundred times. Familiarity breeds the assumption that "today will be like all the other days." Until it isn't.

If your operation involves any of these spaces, your program should be tested by someone who has reviewed the failures, not just the standard. We do permit‑program audits, attendant and entrant training, and rescue capability assessments — and we provide expert testimony when an incident has already occurred and the questions are now legal ones.

Where Programs Look Compliant on Paper but Aren't

Three failure modes I see during program audits, even at organizations that have a written program:

• The permit is filled out after entry, not before. Sometimes literally backfilled to satisfy a records request. The standard requires the permit to be issued before entry begins.

• The four‑gas meter is not bump‑tested or calibrated on the schedule the manufacturer specifies. A meter that reads "all clear" on a sensor that died six months ago is worse than no meter at all — it gives the crew false confidence.

• The "rescue plan" names the local fire department without anyone ever asking the fire department whether they actually have confined‑space rescue capability. Many municipal departments do not. Many that do are understaffed and cannot guarantee response times. 1910.146(k)(1)(ii) requires the employer to evaluate the service's ability and proficiency — that means an actual conversation, not an assumption.

What to Do This Week

If you are a safety leader reading this, three concrete steps for the next seven days:

• Pull every confined‑space permit issued in the last 90 days and check whether the atmospheric readings were recorded before entry began. If not, that is a finding.

• Call the rescue service named on your program's rescue page. Ask them, on the record, what their response time is and what confined‑space rescue equipment they bring. Document the answer.

• Pick one space — your worst one — and walk it with a competent person. Re‑evaluate the classification. If anything in the space, the work, or the surrounding environment has changed since the original evaluation, your hazard assessment is stale.

If your team needs help building, auditing, or defending a confined‑space program, that is exactly the work IamWright Safety Consulting does, both as consultants and as expert witnesses when an incident has already happened. Call us before the next entry — or before the next deposition.

For consulting, training, or expert witness inquiries, please contact us using the contact form.

David Wright is the owner of IamWright Safety Consulting and serves as Director of Safety & Health, Construction at CDM Smith. He is retained as an expert witness in confined‑space, fall‑protection, excavation, and utility‑safety litigation.