Energy impacts of duct design in homes
Many central residential heating and air-conditioning systems in the U.S. have substantially higher external static pressures than recommended by standards and ratings organizations, often due to undersized and constricted ductwork. (Allison Bailes at EnergyVanguard has written extensively on problems associated with flex duct design). Because excess external static pressures can have negative impacts on energy consumption, lower resistance ductwork designs (which may be achieved by a combination of lower resistance materials, larger ductwork diameter, and proper field installation) are typically considered best practices within the industry. However, the impacts of high external static pressures on energy consumption are complex, as the relationships between pressure, fan efficiency, fan power draw, airflow rates, heating and cooling capacities and efficiencies, and system runtimes are also complex and depend in part on the type of blower motor used in the air handling unit (AHU). Moreover, there is a lack of information on optimal operational static pressures for central residential heating and air-conditioning systems and, importantly, the overall life cycle energy and cost impacts of utilizing lower pressure duct designs compared to higher pressure designs.
Therefore, in this work we performed whole building energy simulations and a life cycle cost analysis to compare the total life cycle costs of centralized space conditioning in two new single-family model homes in two separate climates in the United States (Austin, TX and Chicago, IL), both operating under a range of assumptions for operating external static pressures and with real ductwork design configurations as determined by local HVAC contractors. Energy simulations followed a framework of scenarios with specified assumptions for low, medium, and high external static pressures paired with blowers utilizing both permanent split capacitor (PSC) motors and electronically commutated motors (ECMs) in each modeled home. Local heating and air-conditioning contractors in each location provided actual duct designs and cost estimates for both flexible and rigid sheet metal ductwork materials to meet each specified pressure in each home. These designs varied in upfront costs (due to design details, material costs, and labor costs), material type (flexible duct and rigid sheet metal), and ductwork lengths, diameter, and overall layout (and therefore surface areas of ductwork installed in unconditioned space) in order to achieve the predetermined levels of static pressure. Each duct design was assumed to be correctly installed according to standard industry practices (e.g., with minimal compression or sag). The contractors provided their design and installation cost estimates as if the duct systems were to actually be designed and installed in each home. The predetermined external static pressure values were used to estimate the impacts on airflow rates, fan power draws, fan efficiencies, and overall heating and cooling capacities in each scenario using existing fan curve and system performance data for nationally representative residential PSC and ECM blowers, which were then combined with ductwork characteristics from the contractors (e.g., duct UA values) to simulate annual energy consumption for each ductwork design and static pressure level in each home using EnergyPlus.
The Chicago home had a floor area of 2100 ft2 and utilized a 15 SEER 3-ton central air-conditioning unit with a 92.5% AFUE 68 kBTU/hr gas-fired furnace with a nominal airflow rate of 1200 CFM at the lowest system pressure of 0.50” w.c. (125 Pa). The Austin home had a floor area of 3150 ft2 and utilized a 15 SEER, 8.5 HSPF 4-ton air-source heat pump with a nominal airflow rate of 1600 CFM at the lowest system pressure of 0.55” w.c. (138 Pa). The Chicago home had ducts installed in an unconditioned basement and the Austin home had ducts installed in an unconditioned attic. Each home was modeled at three static pressure conditions with both PSC and ECM blowers and with both flexible and rigid sheet metal ductwork designs specified to achieve each static pressure level. Low, medium, and high pressures for the Chicago home were 0.50” w.c. (125 Pa), 0.80” w.c. (200 Pa), and 1.10” w.c. (275 Pa), respectively; the same pressure levels were 0.55” w.c. (138 Pa), 0.85” w.c. (213 Pa), and 1.15” w.c. (288 Pa) in the larger Austin home. In general, lower pressure duct designs were assumed to increase airflow rates and fan power draws in systems with PSC blowers, which was expected to primarily decrease system runtimes and reduce overall energy consumption. Lower pressure duct designs with ECM blowers were assumed to decrease fan power draws while keeping airflow rates nearly constant, which was expected to primarily decrease fan energy consumption, all else being equal. However, differences in contractor duct designs (which primarily affected duct UA values) complicated these expected results somewhat because of heat transfer across ductwork in unconditioned spaces.
Overall, these combinations provided a total of 48 annual building energy simulations, results of which were used to compare the expected annual heating and air-conditioning energy costs between each duct design and system configuration over an assumed 15-year lifespan for each home. Finally, these 15-year life cycle energy cost differences were used alongside the contractor cost estimates for each duct design to compare differences in total life cycle costs of each scenario in terms of net present values (NPVs) using standardized industry assumptions for discount rates, cost of inflation, and future electricity and gas prices.
Key sections of the full report are summarized below. The full report for AHRI can be downloaded here.
AHU performance characteristics
We relied on virtual models of dozens of fan manufacturers to summarize the likely airflow rate and fan power draw responses to the external static pressures specified herein. For both PSC and ECM blowers, nominal airflow rates of 1200 CFM and 1600 CFM were assumed in the Chicago and Austin homes at the lowest external static pressures of 0.50” w.c. and 0.55” w.c., respectively. Increases in external static pressure to 0.80” w.c. (Chicago medium) or 0.85” w.c. (Austin medium) were expected to yield 20% and 18% reductions in flow for the PSC blowers and 3% and 1% reductions in flow for ECM blowers, respectively. Similarly, increases in external static pressure to 1.10” w.c. (Chicago high) and 1.15” w.c. (Austin high) were expected to yield 48% and 43% reductions in flow for PSC blowers and 8% and 2% reductions in flow for ECM blowers, relative to the lowest pressure cases. For the Chicago home, these flow changes corresponded to as much as a 41% reduction in fan power draw (PSC) and as much as a 42% increase in fan power draw (ECM) at the highest pressure. Similarly for the Austin home, the highest pressure yielded a 36% decrease in fan power draw for the PSC blower and a 55% increase in fan power draw for the ECM blower. These pressure, flow, and power draw changes are generally consistent not only with manufacturer data but with data from both laboratory and field tests. Changes in heating and cooling capacities and efficiencies at each of these airflow rates were then captured using built-in polynomial functions in EnergyPlus and used to predict annual space conditioning energy requirements in each scenario. Overall, these AHU characteristics represent values under rather extreme changes in external static pressures, which serve to provide an estimate of the likely bounds of energy impacts involved. In reality, contractors may simply increase the fan size or change fan speed settings to overcome excess pressure in the field, but these simulations do not explore that possibility.
Costs and characteristics of duct designs
For both the Austin and Chicago home duct designs by the Chicago contractor, lower pressure ducts were more expensive than higher pressure ducts, with costs of the lowest pressure designs ranging from 3% to 26% higher than the highest pressure designs, depending on home, target pressure, and material selection. These cost differences largely stemmed from using larger diameter duct materials to achieve lower target pressures. Cost differences in the Austin contractor’s designs were smaller and not as straightforward, with some lower pressure designs even being slightly less expensive than higher pressure designs (although the magnitude of differences were also smaller). These differences are attributed in part to very different designs between the two contractors to meet the same goals. For example, designs by the Chicago contractor resulted in duct surface areas that were typically 20-40% higher than the Austin contractor’s designs for a given target pressure, reflecting large differences in material use efficiencies. The Chicago contractor typically used a radial flex duct design where each supply register was served by an individual branch beginning at the AHU, while the Austin contractor typically used more material-efficient trunk and branch designs. These differences yielded substantial differences in duct UA values (assuming R-6 ductwork insulation for all scenarios), which are important to capture to account for heat transfer across ducts installed in unconditioned space. Finally, there were also large differences in costs for rigid sheet metal ducts compared to flexible duct designs according to both contractors. Rigid duct designs were estimated to cost as much as ~$6000 more than flex duct designs for some configurations, which had a large impact on the life cycle cost estimates herein.
Annual energy simulation results
Lower airflow rates with PSC blowers at high system pressures were predicted to yield large increases in space conditioning energy use relative to lower system pressures due primarily to lower capacities and longer system runtimes. Higher system pressures were predicted to yield only slight increases in space conditioning energy use with ECM blowers due primarily to higher fan power draws at nearly constant airflow rates (although fan energy is only a small portion of the total amount of energy used for space conditioning). More specifically, the lowest pressure ductwork designs by both contractors were predicted to decrease annual energy costs for space conditioning in the Chicago home relative to the highest pressure design by ~5-7% with a PSC blower. These savings were 0-3% in most cases in the Chicago home with ECM blowers, and even led to very slight increases in some scenarios due primarily to higher ductwork UA values with the lower pressure designs. Somewhat more drastically, the Austin home results for both contractors suggest that in this home with these duct designs, the combined effects of the lowest duct pressures will likely decrease space conditioning costs relative to the highest pressure designs by 22-25% with a PSC blower installed, but could either increase (as much as +4%) or decrease (as much a -4%) space conditioning costs with an ECM blower installed, depending on duct UA values stemming from individual contractor design details. These results suggest that lower pressure duct designs can yield substantial annual energy savings relative to high pressure duct designs, particularly for PSC blowers. The energy impacts of lower pressure duct systems with ECM blowers were smaller because fan energy is a small fraction of the total amount of energy used for HVAC purposes.
Life cycle cost-benefit
Three different sets of comparisons were then performed to estimate life cycle costs or benefits using both the simulation results and initial design and installation cost estimates from both contractors: (1) comparing low and medium pressure flex duct scenarios to the highest pressure flex scenario, (2) comparing low and medium pressure rigid sheet metal scenarios to the highest pressure rigid sheet metal scenario, and (3) comparing the same pressure designs with both flex duct and rigid sheet metal scenarios to the highest pressure flex duct scenario alone (the latter representing what is typically thought to be the least expensive duct design option). These three comparisons were made separately to provide comparisons between designs that were as realistic as possible; for example, some locations do not allow flexible ductwork so comparing rigid designs to flex designs is not always reasonable. Life cycle cost comparisons were also conducted separately for PSC and ECM blowers because AHU fan costs were not factored into this analysis.
Flex only. For the PSC+flex combinations, lower pressure duct designs were predicted to have 15-year net present values (NPVs) relative to the highest pressure PSC+flex combination ranging from approximately $430 to $1670 (positive values represent life cycle savings), depending somewhat on target pressure and more so on contractor design (i.e., the combined effects of duct UA and initial cost estimates). For the Chicago contractor’s designs, the medium pressure PSC+flex combinations yielded the highest NPVs; for the Austin contractor’s PSC+flex combinations, the lowest pressure PSC+flex combination yielded the highest NPV in the Austin home and results were similar to the medium pressure results in the Chicago home. For ECM+flex systems, 15-year NPVs ranged from a savings of $37 to an excess cost of $1435 with the Chicago contractor’s designs; the Austin contractor’s designs yielded savings in all lower pressure scenarios ranging from $109 to $419, again with the medium pressure duct system in the Chicago home having a higher NPV than the low pressure and vice versa in the Austin home. These results suggest that within flexible duct systems only, both medium and low pressure duct systems can result in life cycle costs savings over a 15-year period, particularly for PSC systems and often for ECM systems, although the savings may vary depending on actual duct design characteristics and design and installation costs. In total, the lowest pressure flex duct systems yielded 15-year savings relative to the highest pressure flex duct systems in 6 of 8 model scenarios comparing across two homes, two fan types, and two contractors’ designs, while medium pressure flex duct systems yielded 15-year savings in 7 of 8 model scenarios. These results suggest that lower pressure ductwork systems are generally more cost effective if the analysis is restricted to flexible ductwork materials alone.
Rigid only. Limiting life cycle cost comparisons to within rigid sheet metal systems alone, the lower pressure rigid duct designs also generally yielded life cycle cost savings over the highest pressure rigid designs in most of the modeled scenarios. Five out of 8 model scenarios resulted in life cycle savings for the lowest pressure rigid systems relative to the highest pressure rigid systems, and 6 out of 8 scenarios resulted in life cycle savings for the medium pressure rigid duct systems, again summarizing across both homes, both fan types, and both contractors’ designs. These results suggest that if one is constrained to using rigid ductwork alone, lower pressure duct designs can also generally lead to life cycle cost savings in these two model homes, particularly for PSC fans, but also for some ECM scenarios. However, the magnitude (and sometimes direction) of savings may vary depending on fan type, level of pressure, and individual contractor cost estimates and duct design details (that primarily reflect differences in UA values). More specifically, all of the lower pressure duct designs from the Austin contractor yielded life cycle cost savings (ranging from $460 to $1510 for PSC+rigid combinations and from $64 to $244 for ECM+rigid combinations). The only scenarios that did not yield life cycle savings were those using the Chicago contractor’s estimates. Contractor designs alone thus can have a large impact on the economics of lower pressure duct systems in residences.
Flex versus rigid. A final comparison was made across both types of ductwork materials with the highest pressure flex duct design as the baseline scenario but again treating PSC and ECM blowers separately. As mentioned, most of the low and medium pressure flex duct designs yielded life cycle cost savings relative to the high pressure flex designs across both homes and both contractor designs. However, none of the rigid duct scenarios yielded life cycle savings over the highest pressure flex systems; initial cost estimates from both contractors were too high relative to any expected life cycle energy cost savings. However, these results should be interpreted with caution because they assume that both flexible and rigid sheet metal duct systems are equally likely to achieve the same target pressures. In reality, flexible ductwork materials are less likely to be able to achieve the lowest system pressures used herein.
Overall, lower pressure flexible ductwork systems combined with PSC blowers were shown to yield life cycle cost savings relative to high pressure flexible duct systems. Lower pressure flexible duct systems with ECM blowers were also shown to yield life cycle cost savings, although the magnitude of savings is lower than with PSC blowers and can vary depending on individual duct design details and contractor cost estimates.
Results herein were also explored for their sensitivity to a number of important input parameters. For one, extending the duct system life cycle length to 30 years did not drastically affect the results. Second, although some differences in annual energy consumption were predicted to stem from large differences in duct UA values based on the different contractors’ designs, controlling for duct UA values also did not drastically influence the outcomes. Third, the modeled homes utilized relatively high efficiency space conditioning equipment, which may have under-estimated savings relative to homes modeled with lower efficiency equipment. However, we explored this sensitivity by decreasing the efficiency of air-conditioning units to SEER 13, decreasing the HSPF of the heat pump to 7.7, and decreasing the AFUE of the gas furnace to 80, and demonstrated that the magnitude of savings involved would indeed increase for scenarios with predicted savings, but the number of simulation cases resulting in life cycle savings would not change. These outcomes all suggest that the results and conclusions herein are not highly sensitive to these particular assumptions.
There are a number of important limitations to this work that should be mentioned. For example, this work is limited to the particular homes, duct designs, cost estimates, and choices of input parameters used herein. This work also does not capture any changes in system pressures over time; pressures are assumed constant throughout the year (e.g., filters are changed regularly and coil fouling is minimal). This work also assumes that both flexible and rigid sheet metal ductwork have the same likelihood of being installed according to industry quality standards and therefore can meet the specified design pressures. In reality, flexible ductwork materials are more likely to be constricted during construction due to installation with excessive compression, excessive sag, or being pinched by wires and cables. However, these impacts are not captured herein, which is a very important limitation to these findings. Additionally, this work focuses only on energy consumption impacts and does not explore other factors such as air distribution effectiveness, occupant comfort, indoor air quality, or noise associated with different pressures, fans, and ductwork designs. Also, the NPV analyses herein focuses solely on the duct design and installation costs and modeled energy impacts, and does not account for differences in costs between PSC and ECM blowers. Additionally, duct leakage fractions also remained the same in each model scenario (10% of air handler flow), and were not varied with system pressures. Finally, this work does not explore differences in equipment reliability and maintenance that may differ across the ductwork materials used or between the two blower types. For example, blowers may need to be replaced more often when subjected to excessive static pressures, but we are not aware of accurate ways to estimate replacement times under different operational conditions and thus these impacts remain beyond the scope of this study.
We are grateful to the Air-Conditioning, Heating, and Refrigeration Institute (AHRI) for sponsoring this project.