New peak-equipment-load data intended to help take the guesswork out of sizing HVAC systems for new laboratories
Although equipment such as autoclaves, glass washers, refrigerators, and computers accounts for a significant portion of laboratory energy use, measured laboratory-equipment-load data can be difficult to come by, leaving laboratory designers to estimate equipment loads based on nameplate data or experience on past projects. Consequently, peak equipment loads frequently are overestimated, leading to oversized HVAC systems, increased construction costs, and inefficiency at low-part-load operation.
Labs for the 21st Century (Labs21), a voluntary partnership program sponsored by the U.S. Environmental Protection Agency and the U.S. Department of Energy, analyzed peak-equipment-load measurements from 39 laboratory spaces in nine buildings across five institutions (Duke University; Lawrence Berkeley National Laboratory [Berkeley Lab]; the University of California, Davis [UC Davis]; the University of California, Santa Cruz [UC Santa Cruz]; and the University of Wisconsin–Madison). This article documents those data. While a given laboratory may have unique loads and other design considerations, the data in this article can serve as a "sanity check" for design assumptions.
Clamp-on meters were used to measure equipment electrical loads. Measurements were taken when the laboratory spaces were nominally fully occupied and in use. In most cases, measurements were taken over a period of one to four weeks. Measured was one or more of the following:
- Apparent instantaneous power--the product of voltage and current at any instant.
- Actual instantaneous power, which becomes a thermal load to a space.
- Average interval power--the average of actual instantaneous power over a time interval, typically 15 min.
Maximum apparent instantaneous power divided by laboratory area is the key metric for sizing an electrical distribution system, while maximum average interval power divided by
laboratory area is the key metric for sizing an HVAC system. (For more on sizing, see "Right-Sizing Laboratory Equipment Loads."1)
Table 1 summarizes the data, as well as the sizes and types of spaces. The footnotes provide additional information on the measurement approach for each building. Laboratory type was determined by the measurement personnel and does not necessarily refer to a standard designation.
Figure 1 shows maximum average interval power per square foot of biology-laboratory space. The data show a wide range of loads, from just over 1 w per square foot to nearly 10 w per square foot. Figure 2 shows the correlation between laboratory area and maximum average interval power per square foot of biology-laboratory space. The data suggest that smaller laboratories tend to have higher equipment-load densities--possibly because smaller laboratories may have more equipment per unit area, as well as less diversity in loads because of less total equipment.
Figure 3 shows maximum average interval power per square foot of chemistry-laboratory and equipment-room space. All of the chemistry laboratories had loads of less than 4 w per square foot. Loads in equipment rooms, which typically contain shared equipment with little or no bench space, can exceed 15 w per square foot.
Figures 4 and 5 compare instantaneous loads and average interval loads in buildings A and D, respectively. The data show that peak interval loads are significantly lower than peak instantaneous loads. This suggests that sizing HVAC systems based on instantaneous loads can result in significant oversizing.
This article presented equipment-load data from various laboratory spaces. For HVAC sizing, the key metric is maximum average interval power per square foot. In biology laboratories, data for that metric were wide ranging, from just over 1 w per square foot to nearly 10 w per square foot. In all of the chemistry laboratories, loads were less than 4 w per square foot. In equipment rooms, loads exceeded 15 w per square foot. While these data can serve as a "sanity check" for design assumptions, laboratories tend to have unique loads and operational characteristics. Therefore, for new-design and major retrofit projects, it is strongly recommended that measurements be taken in comparable facilities with similar functions and operational characteristics.
Measurement data were provided by William Brewer of Duke University, Steve Greenberg and David Heinzerling of Berkeley Lab, William Starr of UC Davis, Patrick Testoni of UC Santa Cruz, and Mike Walters of Affiliated Engineers Inc.
1) Frenze, D., Greenberg, S., Mathew, P., Sartor, D., & Starr Jr., W. (2005). Right-sizing laboratory equipment loads. Available at http://www.labs21century.gov/pdf/bp_rightsizing_508.pdf
A staff scientist with Lawrence Berkeley National Laboratory, Paul Mathew, PhD, conducts applied research and market-transformation activities related to energy use in buildings and teaches courses on energy-efficient design.