Engineers are rethinking the standard engineering design paradigm, using lifecycle cost analysis (LCCA) to compare system alternatives that fulfill the same functional requirements but differ in terms of initial and operating costs. 
The functional requirements of the project often include sustainability goals. This new “price of admission” creates design tension between tried-and-true and emerging technologies. The consideration of new technology or novel system concepts has historically required the owner to trade higher first costs for the promise of reduced operating costs. The following case study for the American University, Washington College of Law (AU WCL) set out to evaluate two high-performance HVAC systems, one representing a paradigm shift and the other based on a conventional approach with sustainable enhancements. The design team wanted to demonstrate to the client, and to themselves, that a well-designed innovative system could stand up to a high-performance conventional system over the life of the building. The team discovered that sometimes innovation saves money — no tradeoffs required.


The new law school was conceived of as a campus comprising two new building wings connected below grade, two renovated buildings, and an infill atrium for a total building area of approximately 310,000 sq ft, not including underground parking. The two new buildings comprised approximately 80% of the total project area and were the focus of the HVAC system analysis during schematic design.  
Compared to the existing buildings, the new buildings offered the greatest flexibility in the development of multiple system concepts and their integration with architectural elements. This portion of the project houses instructional facilities, faculty and student offices, a library, law clinic, and food service. The law school is located on AU’s Tenley Campus, remote from the main campus. It will be served by standalone heating and cooling plants located in the new buildings. One of the project’s goals was to achieve LEED Gold certification.


The initial HVAC concept arose from an analysis of the 2003 CBECS energy end-use profiles of educational facilities, coupled with the author’s experience designing higher education facilities. Space heating accounts for 47% of energy end-use, the largest share of consumption. Space heating is the sum of the energy required to offset losses from transmission, infiltration, and ventilation. 
While loads associated with transmission and infiltration remain fixed in an HVAC comparative analysis, ventilation can vary depending on the HVAC system type.
This is contrary to the ASHRAE Standard 90.1, Building Performance Rating Method, used for LEED, which requires ventilation be equal for the proposed and baseline systems. According to S.A. Mumma, outside airflow rates for overhead VAV systems, calculated using the ASHRAE Standard 62.1 Ventilation Rate Procedure, can be 20-30% higher than for DOAS (Note 1). In an effort to reduce ventilation fan power and the heating (and cooling) energy required to condition outside air, the team chose to evaluate a DOAS as one of the HVAC system alternatives.  
There are many types of DOAS configurations commercially available, but the key features incorporated into the basic system concept include the following.
  • 100% outside air delivered directly to each zone
  • Decoupled space sensible and latent loads 
  • Flow rate based on ASHRAE Standard 62.1 or greater as required to meet the latent load 
  • Total energy recovery per ASHRAE Standard 90.1
  • Low-temperature supply air (48°F) via high induction diffusers
A DOAS is commonly coupled with a parallel, sensible-only conditioning system. Conventional options for parallel conditioning in-clude air handling units, fan coil units, and water-source heat pumps.  Somewhat more novel technologies such as VRF units, chilled beams, and radiant ceiling panels are also in use, though still more common in Europe than in the U.S. Both the stringent acoustical requirements and physical conditions of the project (i.e., the expansive nature of the buildings and program) supported the selection of a hydronic-based, parallel conditioning system. Hydronic systems are relatively quiet compared to air distribution systems, especially those with fans and/or compressors in or near the space served. In addition, hydronic systems consume less transport energy and take up less mechanical and plenum space than an equivalent air distribution system.  
Chilled beams were dismissed early in design based on a simple study of a 90-person lecture hall.  The primary air through the beam was calculated as the greater of that required to ventilate per ASHRAE Standard 62.1, offset space latent loads, or meet space sensible loads based on a chilled beam air induction ratio of 3.5 (typical induction ratios range between 2.5 and 4.0). The airflow rate required to meet space sensible loads exceeded that required to offset space latent loads by a factor of 1.2 and exceeded the ASHRAE minimum ventilation rate by a factor of 2.4. This would have resulted in an oversized DOAS or a recirculating air distribution system, obviating the benefits of DOAS altogether. 
Next, the team evaluated and ultimately selected radiant ceiling panels (RCPs) for the parallel sensible conditioning system. The following advantages were noted.
  • Use of lower heating and higher cooling indoor design temperatures than all-air system (operative temperatures remain comparable) 
  • High level of thermal comfort (Note 2)
  • Low noise (air terminal units supplying outside air directly to space are 60-80% lower in capacity than for all-air system)
  • Small plenum space requirements
  • Low maintenance (fewer air terminal units and virtually maintenance-free multi-function control valves)
  • Excellent humidity control 
  • Ease of reconfiguration
The high-performance VAV system represented a conventional HVAC approach but included outside air energy recovery and magnetic frictionless chillers to further improve efficiency. Both alternatives included condensing boilers. System alternatives were developed in conjunction with the client’s stakeholders and were considered viable approaches to optimizing energy performance.



The team began by determining indoor design conditions, thermal zones, and the quantity and distribution of air handling units for both alternatives. The air handling unit approach was influenced by the capacity of commercially available equipment and the occupancy schedules of the various program elements. Spaces were assigned to air handling units and ventilation rates were calculated in accordance with ASHRAE Standard 62.1-2007.
For the DOAS, ventilation rates for each thermal zone were based on the larger of that required by ASHRAE or to meet the space latent load. With a ceiling supply of cool air, the zone air distribution effectiveness (EZ) is equal to 1.0. For a DOAS, system ventilation efficiency is equal to 1.0. The resultant ventilation rate for the DOAS was 57,100 cfm, an increase of 37% over the ASHRAE minimum ventilation rate. 
For the VAV system, an Ez of 1.0 and 0.8 for interior and perimeter zones was applied. Supply airflow rates to each zone were cal-culated based on heating and cooling loads and a supply air temperature of 57°F. The maximum primary outdoor air fraction, Zp, for each air handling unit was 0.55 based on the prevalence of high occupancy density spaces such as classrooms, conference rooms, multi-purpose assembly rooms, and library study spaces. These space types, when located along the building perimeter, will experi-ence part-load conditions when fully occupied. As such, the minimum primary airflow rate was adjusted upwards when spaces were occupied and then reset to a lower damper position after occupants left the space. The system ventilation efficiency, Ev, was 0.60 for each air handling unit, resulting in a total design ventilation rate of 69,323 cfm, a 21% increase compared to the DOAS.

Cooling and Heating Capacity 

The cooling plant capacity for each system was sized for space cooling and outside air ventilation loads.  Cooling capacity was 420 tons for the DOAS/RCP system and 480 tons for the VAV system due to its higher ventilation rate. Both system chiller plants comprised two equally-sized, water-cooled centrifugal chillers operating in parallel, two induced-draft, cross-flow cooling towers with variable frequency drive (VFD) fans, and a variable-flow primary pumping scheme. Ideally, separate chillers would be used to produce low-temperature and high-temperature chilled water, but this configuration would result in a greater quantity of smaller capacity, less efficient chillers in order to achieve the same level of redundancy.  Two chillers operating at the same conditions were provided in parallel with a heat exchanger used to maintain the high-temperature chilled water loop. Chiller plant efficiency for the DOAS/RCP system was penalized as a result of having to produce 40°F chilled water to achieve the necessary dehumidification.  Conversely, the VAV system was provided with magnetic frictionless chillers with integral VFDs, resulting in improved chiller performance at part load.
The heating plant for each system was sized for space heating and outside air ventilation loads.  Heating capacity was 1,096 MBH for the DOAS/RCP system and 1,266 MBH for the VAV system. Both alternatives were based on equally-sized, natural gas, water-tube condensing boilers with integral in-line circulator and secondary heating hot water pumps. Similar to the RCP chilled water loop, a heat exchanger was used to maintain the low-temperature heating hot water loop.

Air and Water Distribution/Zoning

The DOAS/RCP system delivered 100% outside air to a combination of constant volume and VAV terminal units by way of three modular air handling units. VAV boxes were used for more densely occupied spaces with demand-based ventilation controls. Constant volume terminal units were distributed by floor and/or by program to enable zone isolation for non-simultaneous occupancy. Two high induction supply air diffusers and one return air diffuser per zone were assumed.
RCPs comprised 25% of the total ceiling area, displacing an equivalent area of ACT and gypsum ceiling material. RCP control valves were provided for each thermal zone.
The VAV system utilized six VAV air handling units, three provided with total energy recovery for a total supply air flow rate of 203,000 cfm. One or more single duct VAV boxes were provided for each zone. Space zoning was applied consistently across alternatives. Due to the higher supply air flow rate, an average of three supply air diffusers and two return air diffusers per zone was assumed.  

Energy Consumption, Cost, and Carbon Emissions

A whole building energy simulation was performed using EnergyPlus software to calculate energy consumption (Note 3). Monthly energy consumption by end use is shown in Figures 1 and 2.
The DOAS/RCP alternative results in higher pump energy and lower fan energy compared to the VAV system due to water being its primary energy transport medium. The VAV system has higher energy consumption for space heating, most likely due to its higher ventilation rate and limited box turndown for high occupancy density spaces when those spaces are fully occupied. While the DOAS/RCP alternative consumed 29% less annual energy and produced 28% less carbon emissions than the VAV system, annual energy cost was only 25.5% lower as a result of a considerably low natural gas utility rate of $0.53/therm.

First Cost

An independent cost estimate was provided based on an accounting of those aspects of each alternative’s system components that differed significantly between alternatives (Note 4). The cost estimate captured mechanical, electrical, and architectural characteristics and included the cost of TAB work, commissioning, and automatic temperature controls. To the surprise of the team, the DOAS/RCP system resulted in a first cost savings.

Maintenance Cost

Annual maintenance costs were estimated for each alternative, in part by reviewing AHU and chiller installation, operation, and maintenance (IOM) manuals, and assigning an hourly rate, duration, and frequency of occurrence to each maintenance task. Only maintenance tasks that differed significantly between alternatives were considered. Annual maintenance costs for the DOAS/RCP system were 25% less than that of the high-efficiency VAV system. While centrifugal chiller maintenance costs were significantly higher for the DOAS/RCP system compared to the magnetic frictionless chiller used in the high-efficiency VAV alternative, the fan and filter maintenance for the VAV system exceeded the maintenance reductions associated with the chiller. While many owners are understandably concerned about maintenance costs, it should be noted that the annual maintenance cost difference between alternatives ($4,740) is less than 3% of the difference in annual energy cost, and therefore has little to no effect on the outcome of the analysis.


An LCCA was performed according to the procedures outlined in the Federal Energy Management Program. Given the lower first, operating, and maintenance cost of the DOAS/RCP system, this alternative achieved lowest LCC. A summary of the relevant LCCA components is given in Table 1. A sensitivity study was conducted that tested the impacts of equal first costs and found the DOAS/RCP system remained the lowest LCC.  
Although the DOAS/RCP system was the clear victor, the team continued to look for opportunities to improve upon the original concept. For example, we tested our initial assumption that a low primary air temperature was preferable to limit active panel surface area. In fact, in a number of interior spaces, we found simultaneous cooling and reheating was occurring at part load. Additionally, we discovered that the heating and reheat load dictated radiant panel area in other spaces. Coincidentally, by adjusting the supply air upwards, we decreased total active surface area and were no longer tied to using high induction diffusers, an aesthetically limiting air device.
Of course, the higher leaving air condition sparked an evaluation of a number of downstream cooling coil reheat options including a heat pipe, a passive dehumidification wheel, and a sensible heat wheel with heat pipe being selected as the preferred configuration. This particular revision to the system concept resulted in an additional 8% energy savings and 3.5% energy cost savings. Replacing oil-lubricated compressor technology with a permanent magnet synchronous motor, magnetic bearings, and integral variable frequency controller further improves energy efficiency and cost by another 5% and 11.5%, respectively. 
Other improvements were incorporated into the DOAS/RCP concept but not specifically cost-analyzed, such as RCP pumps and an injecting/mixing valve for RCP heating and cooling loops in lieu of heat exchangers. Ultimately, the design team fine-tuned the architectural envelope with strategically reduced glazing, exterior shading, and improved glazing performance, which can and should be done regardless of the HVAC system choice. Anticipation of smaller plenums for HVAC distribution allowed the architect to optimize floor-to-floor heights and ceiling elevations.
In conclusion, the building design industry is moving beyond code-minimum energy compliance requirements. In order to meet more aggressive targets for energy efficiency and carbon reduction, small enhancements to business-as-usual system approaches may not suffice. As demonstrated by this study, there are HVAC system technologies that are in current use elsewhere in the world but have yet to achieve broad market penetration here in the U.S. that offer significant energy savings potential. For the American University Washington College of Law, what started out as a paradigm shift in HVAC system design became a holistic design solution that resulted in both first and life-cycle cost savings
Note 1:  Mumma, S.A., Dedicated OA Systems, IAQ Applications/Winter 2001.
Note 2:  Mumma, S.A., Comfort with DOAS Radiant Cooling System, IAQ Applications/Fall 2004.
Note 3:  Energy simulation performed by Ebert & Baumann Consulting Engineers, Washington D.C.
Note 4:  Cost estimating provided by TCT Cost Consultants, Washington D.C.