Editor’s note: This two-part article is one of a series of Best Practice Guides for laboratories, produced by Laboratories for the 21st Century (“Labs21”), a joint program of the U.S. Environmental Protection Agency and the U.S. Department of Energy. Geared toward architects, engineers, and facility managers, these guides provide information about technologies and practices to use in designing, constructing, and operating safe, sustainable high-performance laboratories. For more information about these free resources, see: www.Labs21century.gov/toolkit/bp_guide.htm. The Labs21 website also provides full information about the agency’s upcoming annual conference, to be held in North Charleston, S.C., Oct. 2-4. The meeting is co-sponsored by I2SL, the International Institute for Sustainable Laboratories.
The first article covers load variation and energy use in labs. The second part, which will be published in July, discusses specific strategies for minimizing reheat.
Fig. 1. Range of measured 15-min-interval power for various laboratory spaces in a building at UC-Davis. The upper and lower ends of the lines represent maximum and minimum, respectively. The upper and lower ends of the boxes represent 99th and 1st percentiles of the measurements, respectively. All figures: Labs21.Click to enlarge.
HVAC systems that are designed without properly accounting for equipment load variation across laboratory spaces in a facility can significantly increase simultaneous heating and cooling, particularly for systems that use zone reheat for temperature control. This best practice guide describes the problem of simultaneous heating and cooling resulting from load variations,
and presents several technological
and design process strategies to minimize it.
Load variation in laboratories A measurement study conducted in two laboratory buildings at the Univ. of California-Davis provides some insight into the extent of load variation across laboratory spaces in a typical university laboratory building. In each building, measurements were made for several laboratory spaces, representing the range of different uses within that building. Clamp-on meters were used to take continuous measurements of equipment electrical loads for each lab space. Each measurement period was typically about two weeks long. The measurements were taken when the labs were nominally fully occupied and used.
Fig. 1 (above) shows the range of 15-min-interval power for various laboratory spaces at UC-Davis. Consider peak loads: The peak for most spaces is under 6 W/ft2; a few are between 6 and 10 W/ft2; and one space is high-intensity, at about 17 W/ft2.
This is a fairly common situation, in which one or two labs have very high equipment loads compared to the others. The problem arises when all these labs are served by a single air-handling unit with zone reheat coils for temperature control (a widely used HVAC strategy). The high-intensity labs then drive the supply air temperatures and flows to handle their high equipment loads, and, as a result, all the other labs have to use reheat to maintain desired temperatures. This issue usually does not come up during design, because designers assume a uniform equipment load intensity for all laboratory spaces served by an air handler and assume no variation between those spaces. Energy simulations conducted during the design phase that reflect this assumption will not show the increased reheat energy use that is due to load variation.
Fig. 2. Simulation model used to analyze the energy impact of load variation. Boundary conditions for all zones were set to be adiabatic (no heat gain or loss) to eliminate envelope-related variations in loads for each zone. Zone 3 is about 12.5% of the total area. Click to enlarge
Fig. 1 also shows that there is significant variation across time within each laboratory. For example, for the first space, 3L2A, the maximum is just over 6 W/ft2, and the minimum is just over 1 W/ft2. The 99th percentile is at about 3.5 W/ft2; in other words, for only 1% of the time does the load in this space exceed 3.5 W/ft2, even though the peak is almost double that amount. The wide range within each laboratory suggests that even if peak load variations across labs are accounted for in sizing airflows to these spaces, the variations within each space across time would still provide adequate cause for simultaneous heating and cooling.
Impact of load variation on energy use To analyze the increase in reheat energy use arising from equipment load variation, several parametric energy simulations were conducted using the DOE-2.2 energy simulation tool. The simulation model consisted of a set of five laboratory spaces served by a single air-handling unit
(Fig. 2, above). To eliminate
envelope-related load variations across these spaces, the boundary conditions of all the spaces were assumed to be adiabatic (no heat gain or loss). The lighting and occupancy load profiles in all the spaces were identical.
Each parametric case consisted of two simulations:
Fig. 3. Equipment load profiles used for simulation with load variation and simulation with uniform loads (calculated with the DOE-2.2 energy simulation tool). “Variation-High Intensity” and “Variation-Typical” represent high-intensity and typical space load profiles in the simulation with load variation. “Uniform” represents the area-weighted load profile in all spaces for the simulation with uniform loads. (The total equipment loads for the building in each simulation are identical.)Click to enlarge.
• Simulation with load variation: One zone has a “high-intensity” equipment load profile, while the remaining zones have a “typical” load profile.
• Simulation with uniform loads: All zones have the same uniform equipment load profile, which represents an area-weighted average of the “high-intensity” and “typical” load profiles.
These profiles are indicated in Fig. 3 (left). The total building equipment load in any given hour is identical for both simulations, as are all other parameters. Thus, energy impacts of load variation can be isolated and analyzed.
The base-case model has a VAV system with hot-water reheat, a water-cooled chiller plant, and a natural gas boiler. HVAC component and system efficiencies were set to be consistent with good practice. None of the HVAC component and system parameters were varied in the parametric simulations. The minimum outdoor air ventilation rate for these spaces was set at 1 cfm/ft2.
Fig. 4 (below) shows the base-case source energy use intensity in three different climates in the U.S. The increase in total source energy intensity resulting from load variation ranges from 10% in San Francisco to 14% in Atlanta. An analysis of the simulation results showed that the bulk of this increase is due to additional heating. The increase in heating energy use by zone reheat coils was 48% in Washington, D.C., 50% in San Francisco, and 68% in Atlanta.
Fig. 4. Base-case source energy use intensity in three different U.S. climates (SF = San Francisco; DC = Washington, D.C.; AT = Atlanta; “Uni” = uniform load simulation; “Var” = simulation with load variation). The percentages are the increase in total source energy relative to the “Uni” for each case. Click to enlarge.
The increase in reheat energy due to load variation depends on the minimum ventilation rate. Higher ventilation rates will increase the total energy use; however, as ventilation rates increase, the heating and cooling requirements are less “internal load-driven” and more “ventilation-driven,” thereby reducing the impact due to load variation. Fig. 5 (below) shows that if the ventilation rate were doubled to 2 cfm/ft2, the percentage increase caused by load variation in Washington, D.C., would be 7% (vs. 11% for 1 cfm/ft2). At 3 cfm/ft2, the impact of load variation on reheat energy use is minimal.
Another factor that affects the increase in reheat energy is the extent of the differential between the loads in the high-intensity space and the other spaces. In the base case, the high intensity load was 12 W/ft2, while the peak of the typical load profile was about 3 W/ft2. If the differential is reduced, the amount of increase in reheat energy use will correspondingly reduce. To explore this effect, a parametric analysis was done with the high-intensity load halved to 6 W/ft2. The results for two different climates are shown in Fig. 6 (below). In San Francisco, the increase in reheat energy due to load variation drops from 10% to 7%, while in Washington, D.C., it drops from 11% to 6%.
Fig. 5. Sensitivity analysis of source energy use intensities for different ventilation rates (“Uni” = uniform load simulation; “Var” = simulation with load variation). All results are for the Washington, D.C., climate. Click to enlarge.
In summary, the sensitivity analysis shows that the relative (%) increase in reheat energy due to load variation will (a) decrease with higher minimum ventilation rates and (b) decrease with smaller load differentials.
Note that in all the cases the simulation models assume that the HVAC controls are working as intended. However, experience from re-commissioning laboratories indicates that HVAC controls often deviate from design intent, and that consequently the energy use resulting from simultaneous heating and cooling can increase dramatically.
Part 2 of this article, to be published in July, lays out multiple specific strategies for resolving problems with reheat in laboratory facilities.
Fig. 6. Sensitivity analysis of degree of load differential between high-intensity and typical zone (“Uni” = uniform load simulation; “Var” = simulation with load variation). 12 W refers to the base case load of 12 W/ft2 in the high-intensity zone. 6 W refers to an alternative with 6 W/ft2 in the high-intensity zone. (SF = San Francisco; DC = Washington, D.C.)Click to enlarge.
The authors of this document were David Frenze, PE, formerly with Earl Walls Associates (now X-nth), San Diego; Paul Mathew, Lawrence Berkeley National Laboratory (LBNL), Berkeley, Calif.; Michael Morehead, PE, Flack+Kurtz Inc., San Francisco; Dale Sartor, PE, LBNL; and William Starr Jr., Univ. of California-Davis. Reviewers and contributors included Dan Amon, PE, U.S. Environmental Protection Agency, Washington, D.C.; Phil Bartholomew, PE, CUH2A, Princeton, N.J.; Sheila Hayter, National Renewable Energy Laboratory (NREL), Washington, D.C.; Chris Lawrence, Trox USA Inc., Alpharetta, Ga.; Will Lintner, PE, U.S. Dept. of Energy, Washington, D.C.; Peter Morris, Davis Langdon, Sacramento, Calif.; and Otto Van Geet, PE, NREL, Golden, Colo.
Production assistance for the original Best Practices Guide was provided by Jim Miller, editor, LBNL, Berkeley, and Alice Ramirez, production, LBNL, Berkeley.
U.S. Department of Energy
Energy Efficiency and Renewable Energy
Federal Energy Management Program
www.eere.energy.gov
Laboratories for the 21st Century
U.S. Environmental Protection Agency
Office of Administration and Resources Management
www.labs21century.gov
References
• Bartholomew, P., 2004. “Saving Energy in Labs,” ASHRAE Journal, February 2004. pp 35-40.
• Labs21, 2004. “Case Study: Marian E. Koshland Integrated Natural Science Center at Haverford College, Haverford, Pa.,” published by Laboratories for the 21st Century Program. Available at: www.labs21century.gov/toolkit/case_studies.
htm.
• Morehead, M., 2003. “The Problem With Single-Duct VAV: The Built-in Inefficiency of a Common Lab HVAC System,” presented at the Laboratories for the 21st Century annual conference, October 2003, Denver (see www.labs21century.gov/conf/past/2003/abstracts/h1_morehead.htm
For more information
• On minimizing reheat energy use in laboratories: Paul Mathew, National Renewable Energy Laboratory, 510-486-5116, PAMathew@lbl.gov.
• On Laboratories for the 21st Century: Dan Amon, PE, U.S. Environmental Protection Agency, 202-564-7509, amon.dan@epa.gov, or Will Lintner, U.S. Dept. of Energy, Federal Energy Management Program, 202-586-3120, william.lintner@ee.doe.gov.
