This Commentary discusses the most common safety issues observed in a variety of academic and industrial research laboratories over the author’s 45-plus year career. It highlights the issues and provides additional references for further information. The list can serve as a good tool to help laboratories identify potential safety issues in areas commonly overlooked.

No abstract

Placing equipment in a hood does not always guarantee a safe installation. This Article discusses the various issues associated with proper functioning of a hood with significant research equipment installed in it. The Article briefly discusses hood test methods and issues with assessing hood performance. It then highlights some common problems including storage, large solid objects, equipment which does not belong in a hood and/or will not fit in the hood, horizontal blockage, and some operational issues. The paper then suggests some approaches to avoiding these issues. The focus is understanding what reduces the effectiveness of a hood so that it can be adequately addressed as part of the hazard analysis and risk assessment to ensure a safe operation.

Spring‐loaded relief valves are one of the most common safety devices installed in pilot plants and laboratory bench‐top units. They are typically used in sizes much smaller than those in process units, but their performance is assumed to be equivalent. Most organizations provide only limited preventive maintenance and re‐inspection for these valves under the assumption that they are very reliable devices and operation is fairly assured. Our experience and detailed test data with small size (1/2‐inch and less) spring‐loaded relief devices indicates that most will not perform as reliably as expected due to adhesion of the elastomer seal over time. This leads to initial relief pressures well over the 10% overpressure considered routine, with 30% overpressure being fairly common among certain sizes and conditions. Failing to recognize this problem can lead to otherwise preventable accidents, and even injuries. Data on more than 1,000 relief devices in actual research service for several years was analyzed and the performance measured over time. The data clearly indicate the problem is endemic across all manufacturers and not easily solved. Results are presented in summary form to allow evaluation of the risk inherent in the use of these devices. It indicates a lower level of confidence in actual valve performance, as well as a need for more detailed hazard analysis and risk assessment of the potential for significant relief valve overpressure.

With the increased cost of research and the improvements in instrumentation and analytical devices, the size of pilot plants has decreased significantly over the past several decades. The smaller the pilot plant, the less costly it is to construct and operate, and the faster it is to build, start up, and commission reducing the lead time to productive results. It is now common to have pilot plants small enough to fit in a standard laboratory, even in a fume hood. This is desirable as it often reduces staffing requirements, shrinks the building footprint, and reduces facility operating costs. Shared equipment and resources are more feasible. Staffing often is more flexible. Overall, it seems to be a very good idea particularly for a newer organization without the deep pockets to afford a larger and more expensive facility. Placing pilot plants inside a research laboratory, however, creates some issues unfamiliar to many laboratory designers and raises issues that pilot plant designers may not recognize exist. First and foremost is code compliance without which you cannot get a permit to construct the lab. This paper will try and discuss some of the implications of designing a laboratory which will have pilot plants incorporated in its use to meet these code requirements. The paper will try to address the issues that need to be evaluated and addressed in the overall design of the laboratory to ensure that installing pilot plants is done safely and in compliance with the code. While this is very feasible, it does require the laboratory designer to address several aspects of the codes that can significantly affect either the design of the laboratory or how to best design the pilot plants to fit safely into an existing laboratory.