Printer Friendly

Dedicated Outdoor air systems and building pressurization.


DOASs have become very common and are delivering, for the most part, superb performance compared to conventional all-air systems (1). However, the issue of building pressurization with DOAS is a detail that warrants careful attention. The two main adverse derivatives of building pressurization with DOAS are: unbalanced flow through the TER equipment causing its performance to be compromised on the OA delivery side under design conditions; and, second, the return air flow has frequently been observed in the field to be only 20 to 40% that of the design supply air flow leading to such severe TER performance degradation that the cooling plant is unable to meet the load.

In addition, there is always uncertainty during both the design and operational phases of a project concerning the latent load that must be removed from the occupied spaces in the field. This uncertainty is the result of 1) nebulous initial occupancy information and changing future space use and density, and 2) envelope integrity as a function of use and time. To avoid damaging condensation from chilled ceilings or beams, it is imperative that there be reserve capacity (2) in the DOAS to accommodate such latent load uncertainties. In fact, even with packaged terminal equipment such as fan coil units, water source heat pumps, and variable refrigerant flow split systems, reserve capacity in the DOAS is necessary to prevent their condensate pans from becoming distributed septic[3] amplifiers. This reserve capacity can be very economically made a part of the pressurization unit when compared to putting the reserve capacity into an enlarged unbalanced flow DOAS. Or perhaps better, the reserve capacity could be simply added to the pressurization unit at a later time in stages, thus avoiding unnecessary first cost investments. Note: achieving reserve capacity by introducing recirculated space air, in order to save energy and capacity, is highly discouraged by this author. To do so would eliminate the many benefits of DOAS which uses no recirculated air?


The migration path from all air systems, to DOAS with unbalanced flow, to the integrated balanced flow DOAS with a pressurization unit, is illustrated in Figure 1. The Figure focuses on the ventilation air components and their path back outside. Part a) illustrates that a leaky building is not much of a problem so long as the ventilation requirements exceed exfiltration plus toilet and building exhaust. Part b) illustrates that a typical DOAS experiences unbalanced flow at least equal to the exfiltration. And if the building is extra leaky, the unbalanced flow becomes excessive, adversely impacting the thermal performance and possibly the ability of the entire system to meet thermal loads. This is especially true when differential pressure is used in an attempt to achieve pressurization. Part c) illustrates the integrated concept where the pressurization component flow is held constant regardless of the building leakiness, and the air flow through the DOAS is intentionally held balanced.


It is desirable to limit infiltration in buildings for comfort and IAQ reasons. Infiltration avoidance can be achieved by either providing such a tight envelope that there is no air leakage (something we have yet to do cost effectively) or by building pressurization. Therefore, most designs intend to pressurize buildings sufficiently to 1) cause exfiltration during the cooling season to avoid comfort problems in the perimeter zones and 2) to limit IAQ problems from microbial growth associated with condensation that could occur when moist OA infiltrates into the cool building envelope. During the heating season, pressurization for the same reasons is desirable. However, the pressurization during the heating season need only stop infiltration, thus avoiding exfiltration of moist building air and potential condensation. During unoccupied periods pressurization is not necessary for occupant comfort, or during periods when the outdoor air Dew Point Temperature (DPT) is below an allowable upper space DPT. Therefore, during much of the unoccupied periods minimal or no pressurization is required.

At this point, it is important to establish building pressurization flow rates. This is an especially important issue with DOAS systems that employ heat recovery. Unlike conventional all air systems without heat recovery that can utilize all of the ventilation air for pressurization, DOAS systems must maximize return air flow to deliver the greatest reduction in OA treatment costs. In other words, the TER energy recovery is inversely proportional to the pressurization flow.

The pressurization flow is a function of many variables--building air tightness being a principal variable. To estimate the needed pressurization flow, the literature covering air tightness recommendations and actual test results (4) is useful. Most of the published data gives leakage rates per unit area of perimeter wall area at either 75 or 50 pa (0.3 or 0.2 in WG). For the most part, where pressurization controls are discussed, the differential pressure set point between inside and out is 0.03 inch WG (7.5 pa) (5).

When published building envelope target leakage rates (5 [m.sup.3]/hr * [m.sup.2] [0.2735 scfm/[ft.sup.2]] for offices and 9 m3/hr*[m.sup.2] [0.4922 scfm/[ft.sup.2]] for schools at 50 pa (6)) are adjusted for the control pressure differential (assuming flow is proportional to the square root of the pressure differential (5)), and the wall area leakage rate converted to a per unit floor area, the results are in the 0.04 to 0.06 scfm/[ft.sup.2] (0.2-0.3 L/s * [m.sup.2]) of floor area range. To put these numbers in perspective, that represents about 0.5 air changes per hour (ACH).

The 0.04 to 0.06 scfm/[ft.sup.2] (0.2-0.3 L/s * [m.sup.2]) of floor area range is of the same order of magnitude as the 0.06 scfm/[ft.sup.2] (0.3 L/s * [m.sup.2]) of floor area component of ASHRAE Standard 62.1 (Std. 62.1). That means the arrows labeled pressurization component of ventilation air in Figure 1 could be labeled the floor component of Std. 62.1. At this point the correlation will be considered purely coincidental, especially since envelope leakage is a function of many uncertainties such as; construction workmanship, building type, wall to floor area ratios, building height and climate. However, the floor and occupant components of Std. 62.1 for various occupancy categories may provide insight into the issue of an integrated balanced flow DOAS/pressurization package. Note: Standard Cubic Feet per Minute (scfm) is used throughout this paper since Std. 62.1 specifies ventilation air flows at room condition where a cfm closely equals a scfm. When dealing with OA flows at an air handling unit, the unconditioned OA temperatures are generally well above or below standard conditions where the specific volumes vary widely. For example, the mass flow rate of 10,000 cfm (4,700 L/s) of air at room conditions is 43,890 [lbm.sub.DA]/hr (5.7 kg/s). For the same mass flow rate at 0F (-18C) and 100% RH only 8,480 cfm (4,000 L/s)--15% lower flow rate--is needed, and at 95F(35C) and 40 gr/lbm(0.006 kg/kg) humidity ratio the required flow rate is 10,320 cfm (4,870 L/s)--3% greater flow rate.

While the pressurization flow requirement is generally independent of occupancy type (like the floor component of Std. 62.1), the balance of the OA flow is a highly variable function of the occupancy density and prescribed scfm/person. As a result, the ratio of the pressurization flow to the total OA flow, as summarized in Table 1, varies widely by occupancy category. That variation, illustrated by the ratio of pressurization flow to total ventilation flow rate, falls between 0.1 for lecture rooms to 0.7 for offices, with elementary classrooms in- between at around 0.3.
Table 1, Uncorrected Fraction of Floor to Total Ventilation Air Based
on Std. 62.1.

Occupancy category  floor  scfm/person  [scfm/[ft.sup.2]]  # occ/1000
                    frac      (L/s)      (L/s.[m.sup.2])   [ft.sup.2]

edu 5-8              0.3     10 (5)         0.12 (0.6)       25 (25)

edu 9+               0.3     10 (5)         0.12 (0.6)       35 (35)

lecture              0.1    7.5 (3.8)       0.06 (0.3)       65 (65)

conf                 0.2      5 (2.5)       0.06 (0.3)       50 (50)

office               0.7      5 (2.5)       0.06 (0.3)        5 (5)

retail sales         0.5    7.5 (3.8)       0.12 (0.6)       15 (15)

Occupancy    occ scfm/1000      floor scfm/1000      comb scfm/1000
category       [ft.sup.2]          [ft.sup.2]          [ft.sup.2]
           (L/s.100[m.sup.2])  [L/s.100[m.sup.2])  (L/s.100[m.sup.2])

edu 5-8        250 (125)            120 (60)            370 (185)

edu 9+         350 (175)            120 (60)            470 (235)

lecture        488 (247)             60 (30)            548 (277)

conf           250 (125)             60 (30)            310 (155)

office           25 (13)             60 (30)             85 (43)

retail          113 (57)            120 (60)            233 (117)

Occupancy category  Comb scfm/person (L/s-person)  scfm OA/[ft.sup.2]

edu 5-8                       14.8 (7.4)                0.4 (1.9)

edu 9+                        13.4 (6.7)                0.5 (2.4)

lecture                        8.4 (4.2)                0.5 (2.8)

conf                           6.2 (3.1)                0.3 (1.6)

office                        17.0 (8.5)                0.1 (0.4)

retail sales                  15.5 (7.8)                0.2 (1.2)


The details of DOAS equipment arrangements are beyond the scope of this paper; however, most will utilize TER (enthalpy wheel) as depicted simply (7) in Figure 2. When the DOAS is used for pressurization it leads to unbalanced flow through the TER device, increasing its effectiveness ([epsilon]) as defined in Figure 3, but decreasing the heat recovery rate (it decreases the apparent effectiveness ([[epsilon].sub.apparent]) on the OA stream). As the flow imbalance increases, even when all of the building toilet exhaust air is returned to the TER, the recovered energy decreases.



Eight examples are presented in Table 2. These were selected (8) as representative examples, but are by no means exhaustive.
Table 2. Eight Examples, Including Balanced and Unbalanced TER Flow

Example #  SA Flow, scfm (L/s)  RA Flow, scfm (L/s)   Wheel D,
                                                     inches (m)

    1        19,500 (9,200)       19,500 (9,200)     130 (3.3)

    2        19,500 (9,200)       15,000 (7,080)     130 (3.3)

    3        15,000 (7,080)       15,000 (7,080)     130 (3.3)

    4        15,000 (7,080)       15,000 (7,080)     114 (2.9)

    5        19,500 (9,200)        5,850 (2,760)     130 (3.3)

    6         5,850 (2,760)        5,850 (2,760)      92 (2.3)

    7        19,500 (9,200)       17,550 (8,260)     130 (3.3)

    8        17,550 (8,260)       17,550 (8,260)     124 (3.1)

Example #    FV, sfpm (m/s)     [DELTA]P, in WG (pa)

    1          510 (2.6)             0.35 (87)

    2      510/390 (2.6/2.0)    0.35/0.27 (87/67)

    3          390 (2.0)             0.27 (67)

    4          510 (2.6)             0.35 (87)

    5      510/152 (2.6/0.8)    0.35/0.11 (87/27)

    6          321 (1.6)             0.22 (55)

    7      510/456 (2.6/2.3)    0.35/0.31 (87/77)

    8          510 (2.6)             0.35 (87)

Example #  [epsilon], %  Apparent [epsilon], %  Q, MBH (kW)

    1          74.5              74.5            929 (272)

    2           86                66             822 (241)

    3           79                79             753 (221)

    4          74.5              74.5            715 (210)

    5          99.7              29.9            364 (107)

    6          81.7              81.7            304 (89)

    7          79.4              71.5            889 (261)

    8          74.5              74.5            836 (245)

The data in this Table assumes OA conditions of 85F (29C) DBT and 140
gr/lbm (0.02 kg/kg) humidity ratio and return air conditions of 75F
(24C) DBT and 50% RH. They also assume no seal leakage or purge. The
presence of either will cause a reduction in effectiveness, heat
recovered, and an increase in both face velocity and pressure drop.
Wheels with excellent seals and no purge are recommended for those
reasons. Where purge is considered necessary to safeguard against toxic
cross contamination, it is probably not the correct application for a
wheel style TER.

For example #1, a 130 inch (3.3 m) diameter TER wheel with balanced flow of 19,500 scfm (9,200 L/s) has an effectiveness of 74.5% and a heat recovery rate of 929 thousand Btu/hr (MBH) (272 kW).

When unbalanced flow of 19,500 scfm (9,200 L/s) supply air (SA) and 15,000 scfm (7,080 L/s) return air (RA), example #2 (representative of facility with a pressurization (floor) ventilation rate to total ventilation rate ratio of 0.23), is applied to the same wheel, it's effectiveness shoots up to 86%. This seems very good, but the actual heat recovered is only 88% that for the balanced flow of example #1. The drop in recovered heat is consistent with the poor apparent effectiveness, i.e. 66%, on the OA path side.

Example #3 illustrates the performance for the same 130 inch (3.3 m) wheel at a balanced 15,000 scfm (7,080 L/s) flow integrated with a 4,500 scfm (2,125 L/s) pressurization unit (look ahead to Figure 4 for an illustration). In example #3 the effectiveness is 20% higher than the apparent effectiveness of example #2, but the recovered energy is only 8% less.


Example #4 is for a balanced flow of 15,000 scfm (7,080 L/s)., but with a smaller wheel (114 inch [2.9 m]) selected to match the 510 sfpm (2.6 m/s) face velocity of the larger wheel under 19,500 scfm (9,200 L/s] flow conditions. The effectiveness for the 23% smaller wheel area is 94% that of the larger wheel for the same flow conditions. The recovered heat is, similarly, 95% that of the larger wheel, A 5% energy penalty for a 23% reduction in the DOAS equipment first cost, if realized, may be a good compromise. Those implications will be discussed further in this paper.

Examples #5 and #6 compare selections for an office type application where the pressurization (floor) ventilation component to total ventilation ratio is 0.7--which means only 30% of the total ventilation air is available for return to the TER and energy recovery. Seventy percent of the required ventilation flow exfiltrates for pressurization, Example #5 is for a single DOAS with unbalanced flows of 19,500 scfm (9,200 L/s) of supply ventilation air and 5,850 scfm (2,760 L/s) of return air. With such unbalanced flow, the TER [[epsilon].sub.apparent] drops to a dismal 29.9 % and transfers just 364 MBH (107 kW)--far below the performance with the more nearly balanced flow of example #2, Example #6, a balanced 5,850 scfm (2,760 L/s) flow DOAS in an integrated package, has a 50% smaller TER wheel, but still recovers 85% as much energy. This also represents a great fan energy savings since so much less SA, i.e. 13,650 scfm (6,440 L/s) less, must pass through the flow resistance of the TER wheel.

Examples #7 and #8 reflect a pressurization (floor) component to the total ventilation flow ratio of only 10%--perhaps typical of a lecture room facility. Even in this example, the TER wheel size can step down two sizes and still deliver 95% as much recovered energy.

At this point, one might question the potential to down size the TER in the integrated balanced flow DOAS, in order that a cost effective reduction can be realized.

TER wheels in the diameter range of 92 to 174 inches (2.3 to 4.4 m) come in about 3 inch (0,076m) increments, or approximately 5% change in area. That would suggest that if the unbalanced flow difference is at least 5%, the smaller balanced flow TER could use at least one step smaller wheel, From Table 1, one might conclude that the balanced flow TER could be two sizes smaller when handling lecture hall type conditions, six sizes smaller for the classroom conditions, and up to fourteen step sizes smaller for office applications.

For large wheels between 174 and 251 inches (4.4 and 6.4 m) in diameter (these are big wheels), the wheel area increments are about 16%, Flows in the range where such large wheels are used would still lend themselves to wheel selections for balanced flow of two to five step sizes smaller for the occupancy categories presented in Table 1.


The justification for reserve capacity has been discussed above. The implementation could be either the selection of a larger unbalanced flow DOAS or the selection of an integrated balanced flow DOAS/pressurization package. All of the reserve capacity in the integrated package would be assigned to the pressurization unit. The hypothesis implicit in this paper is that, for most cases, the integrated approach would be the best choice to accommodate the need for reserve capacity as well as pressurization.

Reserve capacity could be considered optional for some terminal equipment selections that are fitted with condensate pans, provided distributed septic (3) amplifiers are acceptable. On the other hand, it is hard to imagine a situation where the ventilation system could be allowed to come up short of latent load capacity where the terminal equipment included chilled ceilings or beams (9).


With the background provided above, no doubt there are many possible configurations. It is hoped that the industry will soon provide a number of options from which the design community mat select.

At the risk of both simplicity and stifling creativity, one possible arrangement is as illustrated in Figure 4. There, the pressurization unit is integrated in parallel with a simple DOAS, In an effort to promote minimal fan energy consumption, bypass dampers are included around the cooling coil(s) and TER. The separate pressurization fan further minimizes the operating cost.

In the simplest sense, two separate pieces of equipment (10) could be used to meet the technical objectives addressed in this paper. However, for this approach to provide maximum cost effectiveness, the duplicate costs of installation and controls need to be eliminated via integration. Note: when surplus ventilation air is employed for a LEED point or a favorable supply air DPT, it should pass through the balanced flow DOAS, not the pressurization unit.


The overview which follows is for the purpose of clearly describing how this author envisions the operation of the pressurization unit, This sequence is neither in complete detail, nor has it been field tested:

1) Pressurization unit to operate during all occupied periods,

2) Pressurization unit to operate during unoccupied periods provided dehumidification is required as indicated by the OA DPT (in excess of 60F (15.5C)--adjustable set point)

3) Damper A to modulate open in sequence (to assure the pressurization enclosure is not damaged by negative pressure) with the fan when the system is to operate.

4) When the pressurization air fan is to operate, set point (adjustable but initially set to the floor component of Std 62.1) shall be maintained with a VFD based upon the flow station ([FS.sub.p]). Set point can also be adjusted as necessary to accommodate unforeseen pressurization or reserve capacity needs.

5) When pressurization unit is to operate, the CC shall cool the air to set point (adjustable but initially set at 4SF [9C] DBT) provided the OA DPT > 48F (9C).

6) When pressurization unit is to operate and the OA DPT[less then or equal to] 48F (9C), the CC shall cool the air only as required to handle the space sensible load in cooperation with the DOAS.

7) When pressurization unit is to operate and cooling is not required, fully open the CC bypass damper. Otherwise the damper is to be fully closed.


Energy performance is sensitive to wheel selections such as size, effectiveness, pressure drop, leakage, purge, and desiccant employed. Also, an important variable in the energy performance is the fraction of total ventilation air needed for pressurization. The fraction impacts both the cooling energy use as well as the fan energy consumption, For representative TER equipment (balanced/unbalanced effectiveness/apparent effectiveness pairs of 0,73/0.64 and 0,64/0.53) with a pressurization fraction of 30% (representative of elementary classrooms), the results demonstrate that the cooling energy increases by 15% and 4% with the integrated package for a Columbus, OH climate (i.e. much of 40[degrees] N Lat). But the fan energy dropped by 12% when compared to an unbalanced flow DOAS, This amounts to between about $100/yr increased operating cost to no increase in operating cost for the integrated package compared to a similar unbalanced flow DOAS when the supply flow is 19,500 scfm (9,200 L/s]-4,500 scfm (2,125 L/s) of which is for pressurization.


Carrying forward the example (#2) of 19,500 scfm (9,200 L/s] of ventilation air, of which 4,500 scfm (2,125 L/s) of the flow is for pressurization, the following first cost observations can be made:

1) The DOAS part of the integrated package handles 23% less air than the unbalanced flow DOAS.

2) Since first cost of DOAS units are essentially on a scfm basis, at about $8/scfm ($17/L/s), the cost differential is about $36,000, The 4,500 scfm (2,125 L/s) pressurization unit adds about $2/scfm ($4.25/L/s), or $9,000.

3) For these flows, the integrated package, without regard for reserve capacity, saves about $27,000 over an unbalanced flow DOAS.

This paper has not provided specific guidance with respect to extent of needed reserve capacity. However if 20% reserve capacity were deemed necessary, the following economic impact would occur:

1) An additional 20% reserve capacity would add 3,900 scfm (1,840 L/s) at $8/scfm ($17/L/s), or $31,200 to the unbalanced flow unit.

2) Adding the reserve capacity to the pressurization unit of the integrated package would add about $7,800.

3) DOAS with unbalanced flow and 20% reserve capacity increases the first cost over the integrated package by:

* $27,000 savings over the unbalanced flow equipment for pressurization

* $31,200 savings over the unbalanced flow equip for 20% reserve capacity

* -$7,800 added cost of reserve capacity put into the pressurization unit.

* $50,400 estimated savings by selecting the integrated package for both pressurization and 20% reserve capacity. It represents an estimated 27% reduction in first cost.


The integrated package, when commercially available, has the potential to offer the following advantages:

1) Enables the first cost of the balanced flow DOAS to be reduced by nearly the fraction of pressurization flow.

2) Does not degrade the TER performance resulting from unbalanced flow

3) Allows reduced fan energy use since less combined supply air and purge air flow occurs on both sides of the wheel. Important since fan energy use is significant.

4) Allows lower operating cost, even though the cooling/dehumidification energy use may increase a little.

5) Eliminates the added installation first cost for two systems (DOAS and Pressurization).

6) Allows energy use for pressurization to be limited with flow measurement control.

7) Allows reserve capacity to be added to the pressurization unit, at much lower first cost than in the DOAS where expensive heat recovery is used.

8) Simplifies controls by dividing the duties.

No change of thinking is without disadvantages. At this time there are a few issues that may fall into this category related to the integrated package, They include:

1) May use more cooling energy.

2) Energy use results are sensitive to equipment selection, i.e. coil [DELTA]P, fan [eta], TER selections ([DELTA]P & effectiveness), air required for pressurization, cooling COP's.

3) May be falsely perceived as more complicated.

4) Not very beneficial for unbalanced flows of less than 15%.


Pressurized buildings improve comfort and IAQ by reducing or eliminating inward air leakage at the envelope. DOASs are capable of providing building pressurization and must do so. It was illustrated above that adequate pressurization can be achieved with about one half an air change per hour, or 0.06 scfm/[ft.sup.2] (0,3 L/s-[m.sup.2]). To assure this flow, it is recommended that the traditional hunting differential pressure control be replaced with measured constant flow control.

The adequate pressurization flow, drawn from a part of the combined Std, 62.1 ventilation requirement, is 30% of the total ventilation requirement for schools and 70% for office spaces. Consequently, the heat recovery equipment in DOAS can experience significant unbalanced flow. It was shown that such unbalance adversely impacts heat recovery.

Therefore, it is recommended that an integrated balanced flow DOAS/Pressurization package be developed and introduced into the market place. The first cost savings, significant in most cases, resulting from the integrated package is expected to cause an even greater expansion of DOAS penetration into the market, Significantly this can be achieved without an increase in operating costs over unbalanced flow DOAS. In addition the pressurization unit offers a relatively inexpensive place to accommodate necessary reserve capacity.

With respect to DOAS in general, and the integrated approach specifically, there are still 3 areas where more information/research would further support the conclusions above. They include:

1) Confirm the recommended limiting/bounding flow rate for pressurization, i.e. 0.5 ACH (0.06 scfm/[ft.sup.2] [0.3 L/s*[m.sup.2]]).

2) Confirm that the recommendation to employ fixed measured pressurization flow control, with the occasional few hours of moisture migration through the envelope when infiltration may occur, does not lead to IAQ or comfort problems.

3) Establish a systematic method for determining the appropriate reserve capacity, or a way to accommodate it in stages.

The intent of this paper has been to encourage the use of pressurization with DOAS and to motivate the industry to consider the following two situations: the need for pressurization and reserve capacity, and bringing integrated balanced flow DOAS/Pressurization packages into the market place, It will be most interesting to see which companies will be first to bring the integrated package into the market place.


(1) Dieckmann, J. K. Roth. J. Brodrick. 2007. "Dedicated Outdoor Air Systems Revisited". ASHRAE Journal Dec. 2007.

(2) Murphy, J. 2009. "Role of Safety Factors in the Design of Dedicated Outdoor Air Systems (DOAS)." ASHRAE Transactions 116(2).

(3) ASHRAE 2008 HVAC Systems and Equipment Handbook, Chapter 6, pg 6.2, Principal Advantages.

(4) INFILTEC Large Building Envelope Airtightness:

(5) Net Pressure to Control Moisture:

(6) The Air Tightness Testing and Measurement Association (ATTMA) Technical Standard 1.

(7) Semco, Inc. EPHC series. 2009.

(8) Thermotech Enterprises, Inc. 2006. "Heat Wheel Selection Software". Version 2.1.0.

(9) Mumma, S. A. 2002. "Chilled Ceilings in Parallel with Dedicated Outdoor Air Systems: Addressing the Concerns of Condensation, Capacity, and Cost" ASHRAE Transactions 108(2).

(10) Mescher, Kirk T. Private Communications.

S. A. Mumma is professor Emeritus of Architectural Engineering, College of Engineering, Penn State

S. A. Mumma, Ph.D., P.E.

Fellow, Life Member, ASHRAE
COPYRIGHT 2010 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Mumma, S.A.
Publication:ASHRAE Transactions
Article Type:Report
Geographic Code:1USA
Date:Jul 1, 2010
Previous Article:Evolutionary computation approach to heat exchanger design.
Next Article:Energy recovery in air handling systems in non-residential buildings - design considerations.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters