Editor's note: From the Archives is a special section of the digital Synergist that brings previously published articles to our current audience. This article, which originally appeared in the February 2011 issue, is once again timely given the publication of OSHA's Final Rule on Confined Sp​aces in Construction in May 2015. For more information about the Final Rule, read the article in this month's NewsWatch sec​tion.

 

OSHA’s long-awaited Notice of Proposed Rulemaking (NPRM) for Confined Spaces in Construction was released for review and comment three years ago. Politics, the economy and resources being what they’ve been, the January 2008 deadline for comments has been followed by an unsurprising silence. The matter has just now resurfaced on OSHA’s agenda for planned action in 2011. As we wait, however, it’s worthwhile considering how this pending change will affect construction in the U.S. What would be an appropriate model for forming a construction industry confined spaces (CICS) standard, and what would such a standard mean for industrial hygienists? ACCUMULATING RISKS I’ve gotten the impression that CS work is often considered a safety issue rather than an IH issue. This is a huge misconception. Consider a tragic fatality that occurred in the early 1990s. A worker in East Providence, R.I., was overcome and killed by toluene (CAS: 50646-98-5) exposure while cleaning ink from a tank. He was a trained and experienced contracted tank cleaner. His company had performed the same job at the same facility five times previously, using an array of PPE. Though this accident happened a few weeks before OSHA’s general industry confined spaces, or GICS, standard (29CFR1910.146–Permit-required confined spaces) went into effect, the cleaning company stated afterward they were nevertheless trying to conduct the job in compliance with the pending standard. The volume of the tank was about 7.5 m³. As a matter of mechanics, the space passed through the following stages as liquid toluene in the ink evaporated into the tank, slowly poisoning, incapacitating, and eventually killing the cleaner:

The worker was killed by toluene’s toxicity, not fire or explosion. At room temperature, however, toluene’s vapor pressure would have driven its atmospheric concentration to about 2.7 percent—well within its flammable range (1.1 to 7.1 percent in air). Given the tools and activities associated with the tank cleaning operation, it’s a miracle there was no fire or explosion. When considering this accident, at what point does the job of an industrial hygienist start, and when does it end? Obviously, it should have started well before the cleaner entered the tank, with scrutiny and risk assessment of all activities, materials and equipment, and continued throughout. What happened here demonstrates how quickly forces routinely found in confined space work drive exposure from levels of chronic concern to immediately life-threatening conditions. CONTROL VS. ELIMINATION OF HAZARDS This accident demonstrated a fundamental flaw in the GICS. As the concentration of toluene in the tank rose, at no point was the entry compelled to be canceled. Once the tank achieved a “hazardous atmosphere” it became a permit-required space, which merely compelled another level of regulatory control for dealing with quickly accumulating risks. The cleaning company chose to make the worker safe for the space rather than eliminate the hazards, at least in part because the then-pending GICS standard allowed such an approach. As they had in the past, the cleaners addressed the increasing hazards with PPE, and it failed. The ability of confined space work to complicate hazards so quickly is supposed to be well recognized throughout industry. Confined spaces are everywhere: in factories, terminals, mines, trenches, shipyards, and the many devices built during construction. The question OSHA should consider is, which industry provides the best model for construction on how to deal with confined space work? There is no question that a CICS standard is needed. In its 1994 publication "Worker Deaths in Confined Spaces," NIOSH analyzed 10 years of CS accidents, cross referencing them to the U.S. Bureau of Labor Statistics (BLS) Standard Industrial Classification (SIC) codes, and found construction to have the third-highest rate of CS fatalities (behind the oil and gas industry, and agriculture). Construction’s fatality rate was shown by the report to be twice that of the (now CS-regulated) manufacturing sector. It also indicated that hazardous atmospheric conditions were responsible for well over half the overall number of fatalities. According to the BLS, there were over 6 million construction wage and salary workers in 2009. Despite the current economic situation, it’s safe to say there are 10 to 15 preventable CS fatalities per year in the construction industry. A fatality, however, is at the extreme end of the range of hazard exposure. As industrial hygienists, we can imagine the opportunities for hazard evaluation, elimination and control in virtually any CS work. Of course, that evaluation, elimination and control must start well before IDLH conditions are reached. This begs the question: at what point should the CS job be not just controlled, but cancelled? The U.S. Chemical Safety and Hazard Investigation Board (CSB) recently reviewed 105 CS accidents between 1993 and mid-2010. Fifty-three were the result of a flammable atmosphere. CSB also found that most of these accidents have occurred since 2003, which is 10 years after the GICS standard came into effect. Between February 2009 and April 2010 alone, CSB identified seven incidents involving six fatalities and four injuries. The CSB report questions why, as part of the GICS standard, OSHA technically allows work in spaces that have reached such an immediately deadly condition. Once a space meets the standard’s definition of “hazardous atmosphere,” it is required to be addressed as a permit space. Assuming the employer takes necessary precautions, work can proceed— but how far? What necessary precautions can an employer take to protect its workers from an ignitable/explosive atmosphere? A MODEL FOR CONSTRUCTION The model for the GICS was, to a large extent, the NIOSH “Guide to Safety in Confined Spaces.” This publication focused on the NIOSH finding that more than 50 percent of workers who die in confined spaces are attempting to rescue other workers. NIOSH went looking for an industrial standard that seemed to have the rescue situation under control, such as the American Petroleum Institute’s “API-2017, Guidelines for Work in Inert Confined Spaces in the Petroleum Industry,” which was developed to protect workers conducting removal of refining industry materials under intentionally prepared oxygen-deficient conditions. (These conditions were intentional to protect workers on supplied-air respirators from fire while removing pyrophoric residues from petroleum refining cracking equipment—so-called “change-outs.”) The API guidelines included a well-developed system for retrieval and rescue of workers using lifting devices, tripods, harnesses, lanyards, dedicated attendants, written permits, etc. This system became a part of the GICS—primarily because it directly addressed the question of quick and effective rescue. The success of this system depends on the relatively static nature of petroleum facilities, equipment and manufacturing materials. In such a setting, those responsible for IH and workplace safety can hone their program to a well-practiced routine. In addition, hazards and spaces in this environment are dependably finite and familiar. The training necessary to deal with these hazards can be collected in the facility-centered document (the written GICS program). Industrial hygiene and safety can be directed, therefore, toward a very specific checklist-oriented approach. The NPRM for confined spaces safety in construction bares a strong resemblance to GICS but with modifications associated with space classification. As perceived hazards increase, the classification of a space would change, and additional preparation would be required. However, as with the GICS, the NPRM for CICS includes no stop-work/cancel-entry aspects. Of the other models available to those developing the NPRM, the one governing confined space work in the shipbuilding industry was perhaps the most promising. Ships are a maze of enclosed and confined spaces that can carry hazardous materials in bulk and undergo a host of incompatible industrial processes during construction and repair. The marine construction standard (29CFR1915–Shipyard Industry) covers all aspects of shipyard work. Developed in 1958, 13 years before the passage of the OSH Act and 35 years before the GICS, the shipyard standard was absorbed by the newly founded OSHA in 1971. Its first substantive section, Subpart B, addresses work in confined and enclosed spaces, and it does not allow work in confined spaces that have less than 19.5 percent oxygen or more than 22 percent; that meet or exceed IDLH levels of toxicants; or that have concentrations above 10 percent of a flammable or combustible gas or vapor’s LFL. These spaces are classified as “not safe for workers”—entry and CS work is simply forbidden. Yet, ships get built. FORBIDDING ENTRY So, what necessary precautions can an employer take to protect its workers from a potentially explosive atmosphere? The only acceptable answer is to prevent work in or near one until the hazard is eliminated. Maritime construction realized early on that work in spaces that can quickly kill or incapacitate should not be controlled by a permit. Work there should simply be forbidden. (There is an exception for cases of rescue or placement of ventilation: see 29CFR§191512 sections a3, b3, and c4.) By taking this position, shipbuilders must move first to make (and keep) spaces safe for workers, not to make workers safe for spaces. Shipyards are at the bottom of an ocean of air some 65 miles deep; there is plenty of fresh air to remove atmospheric hazards. In addition, variety characterizes CS work in both construction and shipbuilding. It is one thing to expect a facility-centered written CS program to deal with CS hazards in a factory where, if you are fortunate, nothing much will change for decades. It’s something else to deal with CS hazards where the spaces can change overnight. Safety and IH must take a fresh but experienced look at every new CS job, and the CS standard should accommodate that process without allowing undue risk-taking by labor. Marine construction deals with this problem by relying on trained and experienced employer-designated competent persons. They must be familiar with the standards, the testing requirements, instrumentation and procedures, and with ships, shipbuilding and ship repair processes. Testing for all sealed spaces is called for, and shipyard workers are trained not to enter sealed spaces that haven’t been tested. In contrast to the GICS and proposed CICS standards, work known or expected to adversely affect the safety of vessel construction or repair must first be approached by hazard removal, not hazard control. The CSB recognizes the danger associated with addressing hazards in GICS work with permits rather than mandated elimination of IDLH conditions at critical levels. While addressing a recent CS disaster that killed five workers in Colorado in 2007, CSB recommended that OSHA

 cupancy in permit-required confined spaces. Why, then, would OSHA recommend a modified version of 29CFR1910.146 as a model for a new construction CS standard? AIHA’s Confined Spaces Committee submitted a substantial body of discussion, comments and recommendations to OSHA in response to the NRPM for its CICS regulations. Whatever it eventually looks like, it should include at least three key elements: