A laboratory ventilation system that does not work effectively may not keep personnel safe and can produce significant adverse impacts on operations
Laboratory personnel treat their ventilation systems as a given. They expect them to be designed right, to always work properly, and to require no real effort or understanding on their part. Most end users have only the most rudimentary understanding of how their own systems work, alarm, and can fail. Yet they implicitly rely on these systems to keep them safe.
The convergence of these two factors often results in laboratory ventilation systems that are poorly designed, marginally functional and not as safe as assumed.
Laboratory exhaust ventilation is expensive. Estimates range from $20–70/(ft3/min) to install and $3–12/(ft3/min) to operate depending on local costs and system design. Hoods (Figure 1) are an added expense, ranging from $200–400 per linear foot to purchase and install. Hood controls add another $1,000-$3,000 per hood. This means it is not uncommon for ventilation systems to account for 15–40% of the total cost of a new laboratory. Hence, it is not surprising that designers, contractors and management all want to minimize the amount of exhaust provided to lower the capital costs of a new or modified laboratory.
The trick to minimizing these costs while ensuring safety and operability is to avoid these common mistakes.
Ten common mistakes
1. Lack of adequate laboratory exhaust to meet operational requirements. A laboratory is first constructed based on a conceptual design with an associated budget and schedule. Too often this conceptual design fails to address all the laboratory exhaust requirements. Many exhaust requirements need a time-consuming and detailed survey to identify; some will even require taking actual measurements. Neither is something usually amenable to the schedule and resource constraints of a conceptual estimate.
Another issue is that some installations may have too little ventilation for their current (and future needs). This requires a careful evaluation by the organization’s safety personnel to identify and articulate their revised basis, another time-consuming and resource-intensive task. As a result, conceptual exhaust requirements, and their attendant costs, are almost always underestimated — often badly.
The best way to avoid this issue is to provide adequate exhaust contingency in the initial scoping design and cost. Contingency is an allowance for historically predictable but currently unknown factors. While contingency is routinely applied to cost estimates, commonly but less frequently to schedules, it is rarely if ever applied to conceptual exhaust requirements. As a result, the need for several local exhausts not identified before, that extra hood (or more often hoods) that got missed, the new hood for a new program, and routine growth all combine to make almost every laboratory short of exhaust before it is even built.
Compounding this problem is that ventilation is the major expense for a new laboratory. So, when budgets run tight, the first place value engineering (the politically correct term for cost cutting) always starts is by trying to cut some ventilation. The result of these efforts is a laboratory that does not have the exhaust it needs to work properly. This problem sometimes manifests itself as soon as the new laboratory is built; it almost always surfaces within a few years. The organization must ensure that enough ventilation is included in all stages of the project. Typically, I recommend at least 20% extra exhaust be added as contingency in the conceptual design stage to try and account for these factors.
2. Too optimistic an assumed diversity factor. The diversity factor is the number of hoods or other exhaust systems likely to be operational at any one time. A 70% diversity factor assumes that no more than 70% of all the hoods or other exhausts on a fan will ever be in use at any one time. It is used to size the capacity of modern variable exhaust systems. These systems automatically vary the speed of the fan to match the currently in use hood exhaust needs. As more hoods are opened for use, the fan speeds up. As hoods are closed and not in-use, the fan slows down.
These systems, while costly, pay for themselves within a few years through energy savings. If the design diversity is ever exceeded, the fan cannot supply enough exhaust and so hood performance will begin to suffer and personnel may be exposed — unknowingly — to hazardous materials.
A common problem is developing a diversity factor based on either general guidelines, cursory surveys, limited experience at other facilities, or limited monitoring. It is very easy to compile data that show 50% of all hoods are not in use over the course of a year. However, it requires more effort to recognize that these data are skewed by holidays, vacations and similar general periods of lower use, which obscures the fact that on many days, the actual use will reach or exceed 70%. Diversity has to always be more than is ever needed, rather than just the average.
Other issues can acerbate this problem. Are hood doors routinely left open when not in use? This will increase the apparent exhaust requirements and required diversity factor. Conversely, a design based on a reduction in exhaust assuming all hoods will always be fully closed when not in use even for a moment is not realistic, as people will not always comply. More use (during a growth phase or unusually busy period) or less use (during a downturn or slow period) skews the results. Surveys are often notoriously inaccurate as they are based on people’s perceptions, which are often very different from the reality, or affect the outcome by the very act of looking at it. (Walking into a laboratory and mentioning one is looking at hood use is very likely to result in the occupants deciding to await one’s departure before doing anything.)
When the diversity factor is breached, meaning more hoods are in use at a given time than the system can support, there is no easy fix and safety is likely to be compromised. Most laboratories, when asked how they know this breaching of the diversity factor has occurred and what is their plan to deal with it, respond with a resounding silence. And a tight budget promotes a downward diversity estimate with often attendant problems later.
3. Limiting hood opening sizes. Designers suggest limiting hood openings to save capital and operating costs. The most common reason is that the cost of the laboratory has been underestimated. When the design is complete and the costs have risen above the budget, the design firm is stuck with trying to find ways to reduce the costs back to within budget. The easiest way to do this is to reduce the size of the laboratory or reduce the size of the ventilation system. Owners accept the latter suggestion as it is much more palatable at that stage of the project to agree to limit the hood opening than to explain to everyone that they must reduce their laboratory space and/or the number of hoods they will have. The owners convince themselves that they can come up with a procedure their folks will follow and make it workable without any real operational impact.
Less often, but still often enough, the problem arises in an established facility that has run out of exhaust due to modifications and new installations. In this case, the designer’s options are very few. Either spend a large amount of money and time, and create significant disruption to expand the ventilation system or get agreement to limit the hood openings. Again, the owner feels that the easiest solution is to limit the size of the hood opening as the best of only awful options.
The most common approach is to install a mechanical stop or warning label at the maximum height the hood sash is designed to be open (Figure 2). Neither approach works in real life. Mechanical stops are modified or removed. Labels are ignored or removed. Almost every survey I have ever conducted, even when they know I am coming, shows numerous sashes open well above their design limits. Explaining the potential hazards to operating personnel rarely corrects the problem for more than a few days, at most. In many cases, the maximum design openings are so small as to realistically make normal operations cumbersome to impossible. Hence, operating personnel must ignore them to get their work done.
4. Lack of adequate supply air. All researchers eventually require more exhaust for new equipment or operations. Providing this exhaust is always expensive so many design firms will not provide the matching amount of supply air to feed the exhaust. (Doing so effectively doubles the costs.) As a result, over time, many (I dare say most) laboratories become much too negative as the total exhaust significantly exceeds the total supply. This means the exhaust is working harder to find the air to exhaust because it is too constricted. As a result, a fan that should be able to exhaust 1,000 ft 3 /min may only exhaust 900 ft3/min or less. This adversely reduces hood effectiveness, makes passing a hood face velocity test more difficult, often draws air in from other laboratories that are not as negative, and makes it harder to open doors. In fact, some laboratories get so negative that they waste a significant amount of the new exhaust they so expensively added.
The problem is that the effect of a 1% supply shortfall is almost always acceptable. However, if the 1% shortfall is actually 2%, and if the process is repeated several times so that the total shortfall is now 10% or more, or if the design basis fails to recognize where the shortfall will be drawn from, major operating problems can — and usually do — arise. Any increase in exhaust must be matched to an increase in supply to be effective.
5. Assuming alternative ventilation approaches will perform as effectively and safely as a hood. Purchasing and installing hoods and their attendant supply and exhaust systems are so expensive, there is a tendency to try and move some operations from a capital-intensive hood to less costly alternatives. These may include ductless hoods, ventilated enclosures, canopy hoods, laminar-flow cabinets and other alternatives. While all these devices may be effective if a hazard analysis and risk assessment shows that they are suitable for the envisioned operations and if they are properly selected, designed, installed and used, few meet all these requirements. Worse, none of them will perform as well as a hood in all operations all the time.
Ventilated enclosures (Figure 3) are fabricated with no real design work and fail to capture effectively because they do not have a properly designed plenum to distribute the exhaust. Ductless hoods can work effectively for specific uses but require procedures to confine their use to the design basis and require routine costly filter changes. Canopy hoods fail to capture all but the lightest gases under the most optimum conditions and usually are useless. Laminar-flow hoods, useful for solids handling, are used for the same operations as regular hoods with gases and liquids and fail to work effectively due to a face velocity that is too low. All these issues revolve around a failure to thoroughly analyze the required operations and evaluate what mitigative measures are suitable for the risks involved.
The National Fire Protection Association (NFPA) standard NFPA 45 Fire Protection for Laboratories Using Chemicals clearly warns against using any of these alternatives as a replacement for a standard hood without careful analysis. A detailed hazard analysis and risk assessment usually ends up with an assessment that “we really need a hood.” Instead, the specter of going over budget often drives research organizations to incredibly poor decisions based on naïve hopes that these less expensive options will work adequately. Usually, they create long-term operating problems, difficult exposure issues and problems costly to repair later.
6. Failing to understand the purpose of a hood alarm and not setting it properly. A hood alarm is required by NFPA 45 Fire Protection for Laboratories Using Chemicals section 7.8.7 “for indicating that the hood airflow remains within safe design limits.” It is not intended as a device that tells the operator that their average 100 ft/min face velocity is now 99 ft/min due to changes in wind against the fan exhaust, minor perturbations in a variable exhaust air system, sudden sash changes, motion near the face of the hoods or dozens of other potential reasons for a small, short-term upset. A hood alarm should only go off when the total exhaust air has fallen off significantly (at least more than 10%) for a reasonably significant period (usually a minute or more), indicating a serious problem affecting the hood’s safe use has occurred. Instead hood alarms trip continually and are eventually permanently silenced, ignored or modified Sub Rosa (secretly) to stop going off.
7. Not maintaining a proper laboratory negative pressure or too much negative pressure. NFPA 45 7.3.3 requires all laboratories to be negative with respect to non-laboratory areas. The American Industrial Hygiene Association (AIHA) standard AIHA Z9.5 Laboratory Ventilation recommends the higher flow be from the lower, non-laboratory area to the higher hazard laboratory area in 5.2.1 unless there is no credible risk.
The most common problems I have seen involve reverse airflow, no negative pressure and excessive negative pressure. Reverse airflow almost always occurs because another nearby area is so starved of supply air that it overwhelms the laboratory’s ventilation system and sucks air from an adjacent laboratory. This is a difficult problem to solve except by providing more supply air to the starved laboratory, which is often costly.
Other, less-viable solutions include requiring some doors only be used for emergency exits and providing airlocks or ante chambers. Other times reverse airflow is due to the supply exceeding the exhaust, pressuring the laboratory. This may be due to a design error, a poorly designed or improperly functioning control system, or a removal of exhaust within the laboratory without any corresponding adjustment to the supply. No negative pressure usually arises only when the design firm has designed the system improperly. Usually, although not always, it can be corrected by simply turning down the supply air to the laboratory. Sometimes this is not possible as the designers used the air from the laboratory to supply other areas, a practice no experienced professional would likely support. Excessive negative pressure usually arises due to a biohazard or cleanroom specification being needlessly applied to a standard laboratory. Again, turning down the supply air rate often works. Biohazard facilities and cleanrooms usually need a much higher defined air flowrate into the laboratory facility for safety, but these same requirements are often inappropriately applied to standard wet chemistry laboratories where all they do is create operating and maintenance problems and increase costs.
8. Overly complicated pressurization schemes. Both NFPA 45 and AIHA Z9.5 are clear that momentary loss of negative pressure when opening a door is acceptable. Nevertheless, many designers provide complicated systems to try and maintain the negative pressure at all times. These systems almost invariably create operating problems from too much pressure or too little pressure. A simple offset in the exhaust and supply air rates (so the supply air tracks lower than the exhaust at all times) is all that is necessary. It may be momentarily disturbed, even reversed, during a door opening or closing, but this has never been shown to be a major issue in a standard laboratory.
9. Getting too focused on total exhaust rate. Total exhaust rates for a laboratory are usually expressed in air changes per hour (ACH). NFPA 45 recommends at least 4 ACH for an unoccupied laboratory and suggests that most occupied (in use) laboratories should be over 8 ACH. The Federal Occupational Safety and Health Administration (OSHA) suggests 4–12 ACH, the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) 6–12, the National Research Council 6–12, and the American Institute of Architects (AIA) 4–12. AIHA Z9.5 frankly notes this is “a subject of controversy.”
In general, overall exhaust rates, while critical for safety during an accident or a release, are not effective in protecting the laboratory occupants from routine releases or emissions, since they will be breathing the air with the hazardous releases until it has been diluted to negligible levels. Depending on the amount released, this can be for some time. Since all operations that have a credible release of hazardous materials should take place in a hood, properly designed ventilated enclosure, sealed system, or under appropriately designed and properly used local ventilation, releases into the general laboratory environment should be rare.
NFPA 45 requires all hazardous materials be captured at the source. Spills, broken containers, momentary releases from a hood due to transfer operations, and similar minor and infrequent events are the most common causes of where the general laboratory exhaust is required. While this suggests that some level of ACH is required, it should be set more by the number of hoods, ventilated enclosures, and local exhaust the operations require rather than by an overall exhaust rate. I, personally, would be very uncomfortable with a laboratory without at least 6 ACH or, better yet, 8 ACH but higher values are not necessary. If odors are common and conditions resulting in releases prevalent, then the solution usually lies in fixing the laboratory operations or a specific issue, rather than the overall laboratory ventilation design.
10. Relying on local exhausts to capture vapors and limit exposures. Local exhaust (Figure 4) when properly designed, installed and — most importantly — properly used is a great mitigative measure. To work properly, the local exhaust must be positioned within the design basis. This basis is often much closer to the work that is releasing vapors than is operationally desirable. If the distance from the release is increased, significantly more exhaust is required or effectiveness becomes substantially worse. All too often, the design economics dictate an exhaust that needs to be within a few inches above the top of the operation. This position is often inconvenient and invariably ends up being used 2–3 times further away with attendant significant loss in effectiveness. Worse, most local exhausts are often simply flexible hoses to lower costs. To position these properly requires a third hand, second person, or some sort of jury rigged holder. The first is sadly not a part of human anatomy, the second rarely available, and the last usually much further away than what the exhaust was designed for. Most, although not all, operations that rely on a local exhaust really need to take place in a hood.
What is the bottom line of all these common recurring problems? A laboratory ventilation system that does not work effectively may not keep personnel safe or can produce significant adverse impacts on operations. I strongly suggest you review your ventilation design very carefully before signing off on the design. This is particularly important if the designer or design firm is less experienced in laboratory ventilation. (And many who claim to be, sadly are not.) I suggest that a cold-eyes review by a laboratory ventilation specialist is often the best money you can spend in a project, because correcting most of these issues after construction ranges from the very expensive to the impossible.
Edited by Gerald Ondrey
Note: All photos courtesy of the author, unless otherwise indicated.
Richard P. Palluzi of Richard P Palluzi LLC (72 Summit Drive, Basking Ridge, NJ 07920; Phone: 908-285-3782; Email: [email protected]) is a consultant to the pilot plant and laboratory research community on safety, design and research project management. He retired as a Distinguished Engineering Associate after 40 years at ExxonMobil Research and Engineering, where he was involved in the design, construction and support of pilot plants and laboratories for ExxonMobil affiliates worldwide. Palluzi is the author of two books, over 180 articles and 60 presentations. He was also one of the core team contributors to CCPS’s new book, “Handbook For Process Safety In Laboratories And Pilot Plants, A Risk-Based Approach.” He has chaired the AIChE Pilot Plant Committee, ExxonMobil’s Pilot Plant and Laboratory Safety Standards Committee, and ExxonMobil’s Clinton site Safe Operation Team. Palluzi is a member of AIChE’s Process Development Division and was elected an AIChE Fellow in 2021. He is on the Technical Committees for National Fire Protection Association NFPA-45 Fire Protection for Laboratories Using Chemicals, NFPA-55 Industrial and Medical Gases, NFPA-2 Hydrogen Technologies Code, and Z9.5 Laboratory Ventilation. He has a BE(ChE) and an ME(ChE) from Stevens Institute of Technology and is a licensed Professional Engineer (PE) and Certified Safety Professional (CSP). Palluzi now consults full time on all aspects of safety, design, and operation for pilot plants, laboratories and research facilities. He also develops and teaches numerous public and custom training on pilot plants and laboratories.