Moving from closed- to open-plan labs: A metrics analysis By Victor J. Cardona, AIA For many, open laboratory design has become the preferred way to plan new biomedical and other research buildings. These facilities are expected to provide generic lab space, increase flexibility and adaptability, and promote collaboration and interaction among users. Are they hitting the marks? Are they meeting administrator and researcher expectations? What metrics have we used, and are there any “lessons learned” that we can apply to future projects? Fig. 1. At Texas Tech Univ., an open-lab “ghost corridor” boosts space efficiency. Photo: Fred Golden Photography. Click to enlarge. Researcher vs. administrator preference As design teams plan labs, we often observe conflicts between researcher and administrator preferences. Most researchers want space ownership, with some level of decision-making in their space customization, bench location and configuration, and to refrigeration. The need for a quiet thinking/writing area is also important. For some, perception is everything. The reigning mentality is: “If I can’t see it, it’s too far away.” Administrators, on the other hand, take a more long-term view of these lab spaces. Important issues for them are flexibility and adaptability of the space; thus they prefer generic, modular layouts. They ask for designs that facilitate space assignment and re-assignment, and ease of changing space size and function. Space efficiency and optimization are also vital to the design, as is a wise and balanced use of resources. As laboratory planners, our job is to find equilibrium between these often competing needs. We aim to create collaborative, exciting, and inspiring spaces for researchers to work: spaces that are functional, safe, flexible, and adaptable, accommodating future changes in use and/or technologies. This is not a new quest for my firm, SmithGroup. We experimented with “semi-open” lab environments in the late 1980s and early 1990s, and are now embracing most of our lab design utilizing an open-plan lab concept. Despite the overriding popularity of this strategy, it’s worth examining it closely to make sure the reality lives up to the hype. Fig. 2. The Translational Genomics Institute, Phoenix, features neighborhoods of open-lab modules. Photo: Timmerman Photography. Click to enlarge. Open-plan labs We define open-plan labs as spaces with few walls, where principal investigators (PIs) share space assignment in a “neighborhood” concept. These spaces are large enough to accommodate interdisciplinary teams in one location. Contrary to the all-inclusive, closed-plan lab environment, open-plan labs segregate work activities. Routine bench work and data entry occur in the open-plan lab itself, and specialized work such as tissue culture and fume hood activities occurs in lab support spaces. Depending on the institution’s policies, desks can be located in the open-plan lab, or outside near the PI offices. The perceived advantages of this concept are many. The openness of these spaces promotes interaction and collaboration among residents, and facilitates space assignment and re-assignment. If properly planned, open labs can be extremely flexible and efficient. And, as an added feature, they also promote lab safety, with fume hoods and other chemical and/or biologically intensive activities segregated from the main work areas. On the other hand, these spaces undeniably provide less “ownership” to the researcher, since there is less definition of the space. Users perceive less privacy, and less intellectual and physical security. And if the open-plan lab design is rigorous, the spaces are more challenging to personalize. No set prescription dictates how large (or small) an open-plan lab “needs” to be. On the contrary, the key question is instead how large the lab “wants” to be. We find that the institutional culture and the researchers’ previous experience with open-plan lab planning are key factors in the acceptance of open-plan laboratories. For example, SmithGroup designed open-plan lab “neighborhoods” for Texas Tech Univ., El Paso, with as few as five open-lab modules, accommodating two to three researchers per neighborhood. On the other end, at the Translational Genomics Institute in Phoenix, we designed “neighborhoods” with 12 open-lab modules. Although the lab goals of both institutions are similar, their culture is quite different, resulting in environments that embrace the same concept, but are expressed in two unique solutions of different scales (Figs. 1 and 2,). Space allocation strategies In the open-plan lab, space is assigned by allocating an equivalent linear ft (ELF) of bench per each individual working in the lab (bench position or BP), rather than assigning a ft2 allocation per PI. Principal investigator space allocation then occurs by assigning each PI a number of these bench positions, based on their level of research funding or need. It is called an “equivalent” measurement since the linear footage not only includes the dedicated bench and desk, but also a portion of all shared areas such as sinks, equipment locations, and movable benches for shared resources and instrumentation. The amount of assigned bench space typically varies by the science housed in the lab. Most of our academic, biomedical research projects include an open-plan lab ELF allocation of 12 to 14 linear ft of bench per BP, consisting of a 4- to 5-ft-wide desk, a 6- to 8-ft-wide dedicated bench, and a 2- to 4-ft length used for shared functions. Most corporate and government labs dedicate more bench than their academic counterparts, and place the desk outside the open-plan lab. Depending on the lab module width, length, and bench configuration, one ELF requires from 6.5 to 7.5 net ft2 of floor space (see Fig. 3, below, for an ELF comparison chart). Science Chemistry/materials Computers/electroincs Biomedical/biotechnology   Univ   Corp./Gov. Univ   Corp./Gov. Univ   Corp./Gov.   NIH Equivalent linear ft(ELF)       Open lab Personal bench Desk Bench 4.0 lf   6.0 lf 4.0 lf   8.0 lf 6.0lf    8.0lf 6.0 lf   12.0 lf 4.0lf   6.0lf    6.5lf 6.0 lf   12.0 lf   8.0 lf Shared bench (sink, bench, equipment, storage) 8.0 lf   12.0 lf 6.0 lf    8.0 lf 2.0 lf    2.0 lf    2.5 lf Lab Support 8.0 lf  12.0 lf   12.0 lf  12.0 lf  14.0 lf Total 24.0 lf   38.0 lf 18.0 lf   28.0 lf 24.0 lf   32.0 lf   31.0 lf                  Fig. 3. Equivalent linear ft comparison chart. Source: SmithGroup. As analytical equipment becomes smaller and more affordable, we see a trend in the increased use of “personalized equipment” at the dedicated lab bench, especially among our corporate clients. This trend requires flexible open lab benches, capable of accommodating and adjusting from desk surface to instrumentation bench. It also requires flexible utility distribution systems that can be brought to the bench surface when and where required (Fig. 4, below). Lab to lab support space ratios Another variable in open-plan laboratories is the ratio between open-lab and lab support areas. The industrywide trend toward increased lab support space could be attributed to two factors. The first is the shift of some functions from a traditionally dedicated, closed-plan lab to the lab support area. The second is the increased use of equipment and instrumentation. Fig. 4. Pfizer Building 20 W bench configuration with overhead utilities (Ann Arbor, Mich.). Photo: William Schumann Photography. Click to enlarge. Lab to lab support space ratios also vary by science, with biomedical work requiring the largest amount of support space. In today’s biomedical labs we find ratios of similar amounts of open-plan lab to lab support space—in other words, a 1:1 ratio. But at the Univ. of California-San Francisco’s Genentech Hall building, on the Mission Bay campus, SmithGroup provided 1.2 ft2 of lab support space for every 1 ft2 of open lab. We usually find six types of dedicated lab support spaces required to support the biomedical open-plan lab: fume hood alcoves, tissue/cell culture rooms, linear equipment rooms, cold environmental rooms, flexible space where light-tight microscopy or specialized instrumentation can be located, and computational rooms. To provide some level of user accountability, researchers typically prefer to “dedicate” or share these spaces only among members of a designated “neighborhood.” These six types of spaces typically account for 60 to 70% of the total lab support space (Fig. 5, below). Ratios Chemistry/materials Computer/electroincs Biomedical/biotechnology   Range Range Range Open Lab: lab support 4:1 to 3:1 5:1 to 4:1 1:1 Open lab: research offices 3.5:1 2.5:1 3:1 to 2.5:1 Open lab: interaction areas 5:1 6:1 3.5:1 Fig. 5. Space ratios comparison chart. Source: SmithGroup. The remaining 30 to 40% includes shared functions such as additional environmental (warm) rooms, analytical instrumentation areas, radioisotope fume-hood labs, dark rooms, glassware cleaning and autoclave rooms, and the ever-increasing need for “freezer farms” (long-term storage rooms). This ratio is usually less in corporate labs, since they usually centralize some of the shared support functions such as glassware cleaning and autoclave rooms. Important to the open-lab concept is the connectivity among the “neighborhoods.” At the Univ. of Michigan Life Science Institute, in Ann Arbor, we used linear equipment rooms as entry points into each “neighborhood” (or quadrant, as they call them at U-M). In this project, these spaces were located on either side of the central, double-loaded corridor, promoting direct connectivity between quadrants (Fig. 6, below). Fig. 6. At the Univ. of Michigan’s Life Sciences Institute, an 83% floor efficiency was achieved using a rectangular floor plate. Design: SmithGroup. Click to enlarge. Lab to office space ratios The lab to office space ratio also varies between open-plan and closed-plan lab buildings, and generally results in fewer dedicated PI offices in today’s open-plan buildings. We attribute the change to both the increased efficiency of the open-plan lab space and to the trend for larger, PI-based research teams. A factor to consider is the emerging trend of housing “write-up” desk space outside the lab environment. Corporate clients have done this for many years, and now SmithGroup is seeing this trend in some of our academic projects. The benefits include safer work environments, without the potential for chemical and/or biological contamination, and quieter workplaces for post-docs, graduate students, and technicians. Locating desks outside the lab allows for closer proximity to the PI offices, potentially increasing interaction between users. At the Arizona Biomedical Collaborative (ABC) Building, presently under construction in Phoenix, SmithGroup provided post- doctorate desk work space outside the lab environment. This concept shifted more than 5% of the open-plan lab space to the office area, reducing the amount of space that required once-through ventilation—but it also resulted in larger workstation space allocations for the post-doctorate/fellows. The space allocation became a 5- 3 8-ft office cubicle instead of a 5-ft-long desk at the open-plan lab. Interaction spaces A major premise of open-lab planning is to increase user interaction and collaboration. We also realize that there are multiple other opportunities for this interaction to occur outside the open-plan lab. Consequently, most of today’s design teams are incorporating multiple formal interaction spaces such as conference and seminar rooms, and informal spaces such as break rooms and lounge areas. Since corporate research campuses usually provide for central food service amenities, we usually find these informal interaction areas more prevalent in academic environments. Fig. 7. SmithGroup designed a “family room” on each lab floor at Cal Tech’s Broad Center. Photo: Tim Hursley Photography. Click to enlarge. At Cal Tech’s Broad Center for the Biological Sciences, in Pasadena, SmithGroup provided a first floor café to support that area of the campus, in addition to “family rooms,” consisting of a combination of break room, lounge, kitchen and library space, on each typical lab floor (Fig. 7, left). Each floor also contains an informal, non-scheduled interaction area near the PI offices. It is not unusual to allocate 1 ft2 of interaction space per every 3.5 ft2 of open lab area. Flexibility A key concept in planning effective interdisciplinary lab environments is to allow for multiple sciences to be collocated next to or near to each other. Predicting which sciences and the right mix is the most difficult part of this effort. How do we “right-size” lab systems to accommodate their needs? One solution is to provide highly flexible lab environments that, in each area, can accommodate a variety of uses, from dry computational lab space to heavy chemistry. In this model, we provide for the worst-case scenario, with a “just-in-case” approach to the problem. A second option is the “just-in-time” approach, which provides additional capacity in the main mechanical and electrical systems, and creates ample pathways to upgrade the systems distribution when required, such as additional mechanical shaft and ceiling space. This approach creates highly flexible lab environments that can accommodate unforeseen requirements in the future, without excessive front-end costs. A third option is a combination of the previous two. At the Univ. of Illinois Beckman Institute I in Champaign-Urbana (the 1990 R&D Magazine Lab of the Year), SmithGroup located both wet and dry lab functions on every floor. We also provided for accessible shaft space at every lab “pod” or block, thus facilitating accessibility, pathways, and future systems upgrades. Space efficiency Space efficiency is another advantage of open-plan labs. These labs entail a greater space efficiency (ratio of dedicated, assigned space to the overall space on the floor) since there are fewer walls (wall space usually accounts for more than 6% of the overall gross ft2 of a building). In addition, the large, open lab environments include most of the circulation internal to the lab, in the “ghost corridor.” Typically open-lab facilities achieve 70 to 78% floor efficiency, compared with 65 to 73% floor efficiency in closed-plan labs. At the Univ. of Michigan Life Sciences Institute, where a rectangular floor plate was provided, an 83% floor efficiency was achieved. This factor is extremely important since floor efficiency is the main contributor to the overall building efficiency. Most of our open-plan lab buildings achieve a 57 to 62% building efficiency, compared with a 55 to 58% building efficiency in most closed-plan labs. Do open-plan labs cost less? One advantage of the open-plan lab is that it allows us to provide generic and less service-intensive open lab space, since the labs are supplemented by the service/utility-intensive lab support spaces. Another factor contributing to reduced cost is the provision of fewer walls and reduction of the mechanical control systems (which must be more complex to regulate the environments of multiple individual small labs). Open-plan labs have the potential to save money; in fact, some firms estimate that this scheme cuts up to 5% of the lab construction cost. Theoretically this is certainly possible, but in real life the savings can be elusive. For instance, open-plan labs usually incorporate more flexible (and costlier) casework systems. Since the labs are open and highly visible, teams tend to design them with larger exterior windows and improved finishes. The principles driving the open-plan trend often drive clients to provide a greater number of non-lab interaction areas than would be typical in a closed-plan facility. These improvements create richer environments but can cancel any potential savings. In making the “closed to open” transition we have found that open-plan labs are not for everyone. But given the right design—and if researchers can move past the inevitable question of “Are my supplies safe?”—the increased benefits outweigh any hesitations. Victor J. Cardona, AIA, is a VP and director of lab planning at SmithGroup (www.smithgroup.com), an 800-person architecture, engineering, interiors and planning firm with 10 offices across the U.S. SmithGroup, headquartered in Detroit, specializes in the science and technology, healthcare, learning, and workplace markets.