Reprinting this proof for distribution or posting on web sites is not permitted.

Authors may request permission to reprint or post on their web site once the
final version has been published. A reprint permission form may be found at [Link].

CO2-Based Demand
Control Ventilation
Do Risks Outweigh Potential Rewards?
By David S. Dougan and Len Damiano, Member ASHRAE the required removal or dilution of more
than 50% of the pollutants.1,2 The second

E
nergy conservation strategies often distract attention from source is body odor, produced by the oc-
cupants as a result of their activities. This
other core design objectives, including occupant health, pro- latter source has provided many designers
with the opportunity to automatically
ductivity and avoiding threats to the building structure’s long-term reset outside airflow rates in facilities
with variable occupancy, and capture
integrity. CO2-based demand control ventilation (DCV) is an en-
the energy savings available compared
ergy-conserving strategy that, in some cases, has sacrificed several to continuous conditioning.
The concept of using CO2 input for
of these fundamental design objectives. Instead we have embraced DCV makes sense and can save money
on building operating costs under specific
short-term energy cost savings and accepted greater risks to oc- circumstances. Building managers can
see an energy benefit from reductions of
cupant health, diminished worker productivity, increased main- outside air intake rates as the occupant
density decreases. However, systems
tenance costs, and life-cycle cost for the structure. rarely are implemented that account for
the code-mandated “actual number” of
This article examines the sources of Demand Control Ventilation
people in a particular ventilation zone.3
risk using DCV, the components typically A typical building has two significant
used and possible ways to minimize risk contaminant sources that, without proper
ventilation, can lead to unsatisfactory About the Authors
without sacrificing potential energy sav-
David S. Dougan is president of Ebtron and
ings from dynamically resetting intake indoor air quality. One source is the build- Len Damiano is vice president of sales and market-
rates based on occupancy changes. ing, which in many cases can result in ing at Ebtron in Loris, S.C.

2 ASHRAE Journal [Link] October 2004

 Space
1.50
Occupants

VOCO VoCs
1.25
VeCe
VeCs N
Ce
Heavy Work
Cs 1.00
(1.05)

CO2 Production, L/min.

Equation 1 – Outside Airflow Calculation
VO = N/(CS – CO )
where 0.75
VO = outdoor airflow rate per person
VE = breathing rate
Light Machine Work (0.6)
N = CO2 generation rate per person
Walking (0.55)
CE = CO2 concentration in exhaled breath 0.50
CS = CO2 concentration in the space
CO = CO2 concentration in outdoor air

Figure 1: Two chamber model and outside airflow calculation. Office Work (0.3)
0.25 Seated Quiet (0.25)
Sleeping (0.2)

Since people produce CO2 as a direct result of respiration, it

has been an understandable DCV input choice. Unfortunately,
a significant number of designers and owners have not fully 0 1 2 3 4 5
Physical Activity, MET Units
understood the relationship between CO2 levels and ventilation.
The least of which is that CO2 is neither a pollutant nor a direct Figure 2: CO2 production vs. metabolic activity.
measure of occupancy.4,5
1. The occupants are generating CO2 at an assumed constant
CO2 Levels and Ventilation rate: N (cfm* or L/s of CO2/person), i.e., their metabolic rate,
CO2-based DCV often is implemented with little regard to the diet, and level of activity are identical.
actual relationship between ventilation rates and CO2 levels. 2. Outside air, of known CO2 concentration: Co, is introduced
The ventilation rate procedure in ASHRAE Standard 62-2001, into the space at a constant rate: Vo (cfm or L/s per person).
Ventilation for Acceptable Indoor Air Quality, specifies required 3. The indoor CO2 level: Cs , represents human occupancy
minimum ventilation rates for compliance with the standard, not within the ventilation zone and there is no allowance for inac-
CO2 levels, for acceptable indoor air quality.4,5 curacy in measurement.
Section 6.1.3 of Standard 62-2001 (which was replaced by Calculating Vo = 7.5 L/s (15 cfm per person) with an as-
Addendum 62n) states, “Indoor air quality shall be considered sumed CO2 generation rate (N ) of 0.31 L/min. per person will
acceptable if the required rates of acceptable outdoor air in Table result in an indoor CO2 level approximately 700 ppm greater
2 are provided for the occupied space.” than the level of CO2 in the outside air. (Solving for the CO2 dif-
Systems must provide adequate dilution airflow rates for ferential, Cs – Co = N / Vo or 0.31 / (7.5 × 60 s/min.) = 0.000689
compliance. However, even though the standard clearly does L·CO2 / L·air, or 700 ppm]. This is the underlying mathematical
not specify acceptable CO2 levels for compliance, many believe relationship and comfort justification for the ventilation rate
that maintenance of space CO2 setpoint levels will result in ac- tables in the original ASHRAE ventilation standard.8,10 Stud-
ceptable indoor air quality by indirectly regulating the amount ies that were used by ASHRAE have indicated that 15 cfm per
of dilution air provided. person is the rate of outside air required to dilute offensive body
What is the relationship between CO2 and the outside airflow odor and the calculated 700 ppm is the CO2 rise referenced in
rate into a space? To answer that question, we must first under- Standard 62.6–10 Therefore, the resulting statements appeared
stand the mathematical model that describes the use of CO2 and in Section 6.1.3 of the 2001 standard:
the assumptions required for the model’s validity. “Comfort criteria, with respect to human bioeffluents (odor)
The relationship between CO2 levels and outside air ventila- are likely to be satisfied if the ventilation results in indoor CO2
tion rates can be described using a simple, two-chamber model, concentrations less than 700 ppm above the outdoor air concen-
as shown in Figure 1, from Standard 62-2001, Appendix C. tration.” This may be addressed in a proposed appendix created
This model relates the differential CO2 level (inside minus by Addendum 62ah or possibly by reference in the recently
outside) to the airflow rate per person when the following steady- contracted User’s Manual for Standard 62.
state conditions are true. Section 6.2 of Standard 62-2001, which was replaced by
Addendum 62n states, “Using CO2 as an indicator of bioef-
* cfm × 0.4719 = L/s fluents does not eliminate the need for consideration of other

October 2004 ASHRAE Journal 3

 Cs setpoint = 947 ppm (20 cfm/person, Co = 400 ppm,
N = 0.31 L/min., Sensor Error = variable Cs setpoint = 947 ppm (20 cfm/person, Co = variable,
N = 0.31 L/min., Sensor Error = 0 ppm
25.0%
25.0%

15.0%
15.0%
cfm/person Error

cfm/person Error
5.0% 5.0%

–100 –75 –50 –25 0 25 50 75 100 –5.0% 300 350 400 450 500
–5.0%

–15.0% –15.0%

–25.0%
–25.0%
Outdoor CO2 Level, ppm
Indoor CO2 Sensor Error, ppm

Figure 3: Sensor error. Figure 4: Assumed outdoor CO2.

contaminants.” preciate the differences between monitoring for evaluation and

Remember, more than 50% of the contaminants in the aver- monitoring for control.12
age office building are non-occupant generated and cannot CO2 generation rates can vary widely as indicated in Figure
be addressed by controlling CO2 levels alone. This appears to 2, based on activity levels. It may also vary based on diet and
be the single strongest motivator for the significant changes health of the occupants. As a result, significant error can exist
included in Addendum 62n and its subsequent adoption by in the cfm per person calculation (Table 1).
ASHRAE and ANSI. The model also is only valid under steady-state conditions.
We also should acknowledge that Appendix C covering CO2 CO2 DCV, by design, is intended to be used in dynamic situ-
in the current Standard 62 was included primarily to help ex- ations and implementation of this strategy often negates the
plain the origin of the rates used in the ventilation tables.4 It was validity of the model.13 In addition, the placement and reliability
not intended to support or to justify the use of CO2 for ventila- of the CO2 sensor is critical and the performance of today’s sen-
tion control. Yet, the steady-state, two-chamber mathematical sors still is reported to be questionable (Figure 3).5,8,14,15
model contained in the appendix has been used in applications CO2 sensors are reported to have noteworthy, technology-
and referenced to support the use of CO2 sensors for automatic specific sensitivities, unresolved issues and application con-
control purposes. siderations including:
As an indicator of ventilation adequacy for the dilution of • Drift;
body odor, the 700 ppm rise criteria is perfectly acceptable for • Overall accuracy;
a space being evaluated by a diagnostician, in accordance with • Temperature effects;
the requirements of the applicable ASTM standard.11 The CO2 • Water vapor;
generation rate assumed in the model is based on the average • Dust buildup;
generation rate for this minimum activity level (0.31 L/min. • Aging of the light source;
per person**). As a result, any increase in the average activity • Frequency of calibration;
level of the occupants (N) would tend to indicate a greater CO2 • Mechanical vibration;
differential than that calculated by the steady-state formula and • Electrical noise;
over ventilate the space, negatively impacting the expected • Sensor location in the space;
potential savings. • Number of sensors required;
Confidence in the calculated results can be increased if the • Method of averaging multiple sensors; and
ventilation rate into the building and space are held constant • Compounded error rates from multiple sensors.
during the evaluation, and the occupant density is maintained. Geographically and seasonally, outside CO2 levels vary
The technique is best suited for use with a single, handheld, widely.5,13 Outdoor levels are generally not measured, because
frequently-calibrated device in the hands of trained profession- CO2 sensors (in varying degrees) have trouble with accuracy
als, for use in localized areas for time-specific diagnostics.9,11,12 above a relatively low velocity threshold, at low ambient air
Unfortunately, misunderstandings regarding the valid applica- temperatures and may be affected by changes in atmospheric
tion of the technique can be created by those who do not ap- pressure (Figure 4).5,13
Sensor manufacturers have developed several methods to
** L/min. × 0.2642 = gallons/min. improve the reliability of CO2 measurement. Specific models

4 ASHRAE Journal [Link] October 2004

 Activity N, L/min. Vo, cfm Total OA cfm Required
People cfm/Person Comments
0.20 10 Required Cs – Co
Sleeping
0.30 15 7 95 13.5 807 ppm Overventilated at
Office Work
Walking 0.55 28 6 90 15 700 ppm } 700 ppm

Light Machine Work 0.60 30 5 85 17 644 ppm Underventilated at

Heavy Work 1.05 53 3 75 26 438 ppm } 700 ppm

Table 1: Calculation of Vo at CO2 production levels, C = 700 ppm. Table 2: Required CO2 level at various population densities in an
office space (area = 1,000 ft2). Total OA cfm required = 0.06 cfm/ft2
+ 5 cfm/person (Standard 62n, offices).
that automatically reset to the overnight ambient level have
helped reduce the frequency of required calibrations, if the CO2–based DCV has limited the amount of outside air intro-
ambient night levels in your area are valid base CO2 levels, duced into a building. Without a positive pressurization flow (the
which do not change over time. Others use methods to protect difference between the outside air intake and the total exhaust
the sensing elements from the environment, while still others flow rates), a building cannot be pressurized.
apply an internal light-source reference to assist stability.5 Designers must carefully consider space pressurization con-
However, the cumulative uncertainties of the hardware and trol when using demand control ventilation strategies (CO2 or
methodology remain. The total uncertainty’s impact on intake others). The amount of control error allowable at the intake for
rate control can result in significant risk for the designer and pressurization diminishes when the total amount of intake air
building owner who have chosen to implement a CO2 DCV is reduced, making the accuracy of intake control more impor-
strategy, unless deliberate care is taken and supplementary tant. DCV systems may satisfy their differential CO2 setpoints
actions are used (see Figures 3, 4).5,13 The practitioner must while ignoring differential flow (pressurization) and the amount
have more than just basic knowledge of the strategy’s proposed of dilution air needed to mitigate the effect of non-occupant
requirements and limitations. sources of contamination.16
Appendix B of NISTIR Report 6729 for the California En-
ergy Commission of March 2001 concluded, in part, that good Complications From Changes by Addendum 62n
practice usage of CO2-based DCV also would incorporate “other Changes to the ventilation rate procedure of ASHRAE Stan-
IAQ control technologies.”5 The report recognized most of the dard 62 resulted in outside airflow rates that vary significantly
limitations of DCV and felt that supplementing this technol- on a “per person” basis (Table 2). Addendum 62n recognizes
ogy would provide more overall reliability. Because CO2-based this and has modified the table’s structure to address the mag-
demand control ventilation has a tendency to overstate changes nitude of building-generated pollutants. Under ideal conditions,
in occupancy, direct measurement and control inputs for the CO2 levels can only relate to the rate that outside air enters the
actual intake rates may be useful in preventing overventilation building on a per person basis (i.e., cfm/person). Therefore, it is
and intake shutdowns.9,16 difficult to envision how CO2–based DCV can be implemented
Mounting airflow measurement devices in the outside air under the new requirements of Standard 62 and simultaneously
intake to limit the lowest reset point of outside airflow rates “maintain” the required minimum intake rates “under all load
during periods of lowered density (or diminished occupancy), conditions,”4 without significant energy cost impacts (in contrast
can reduce IAQ risk from underventilation. This modification to the savings expected from its use).
also may reduce the risk of wasting energy by allowing the SSPC 62.1 publicly announced the intention to develop a
establishment of an upper limit, never to be exceeded.16 CO2 appendix to Standard 62 this year. Addendum 62ah might
be considered by some to be the magic bullet for intake rate
DCV and Building Pressure control. Some rationale and application criteria is expected to
When the outside dew point exceeds 65°F,† humidity levels in be offered in this appendix, allowing the use of CO2 inputs for
negatively pressurized building envelopes can exceed 70% RH, indirect ventilation control and compliance with the require-
the minimum humidity level in which many molds can grow. ments of the ventilation rate procedure.
High humidity conditions in and near the building envelope Also the possibility exists that CO2-based DCV methodology
will result in mold growth.17 may be introduced in a user’s manual to Standard 62, whose
Some molds may be toxic to humans while most molds contract award was announced last January. The difficulty we
produce allergens. Many can damage the building’s structure have with a user’s manual being used to endorse or validate
and can be extremely expensive to remove from spaces that systems and equipment not addressed in the standard is that
are difficult or impossible to access, e.g., inside exterior walls. the content requires review and approval only by Standing
Recent publications have recognized the relationship between Standards Project Committee (SSPC) 62.1 and Society board.
building pressure and mold growth.17–23 The widespread use of The public will not be aware of the contents until it is ready for
publication. Until then, we are left with many questions and
†
(°F – 32) × 1.8 = °C little in the way of scientifically verifiable data to support the

October 2004 ASHRAE Journal 5

validity of CO2 measurement input’s as suitable and otherwise pressure changes;
comparable in reliability to other available methods of ventila- 6. Consideration must be given to selecting only for the ap-
tion rate control. propriate range of operation;
This lack of supporting test data and the absence of applica- 7. Drift is still an issue and calibration recommendations
tions guidance were deficiencies identified as a few of the many must be followed;
problems in the blanket use of CO2–based ventilation control 8. The preferred locations for sensors are multiple ones placed
schemes by the recent California Air Resources Board report to in the occupied zones;
the writers of the state ventilation code, the California Energy 9. Do not use sensors that are not intended for control pur-
Commission (CEC).24 poses;
However, we should expect that the SSPC 62.1 commit- 10. Do not use sensors near doors, windows, intakes or ex-
tee and the contractor for the user’s manual will be thorough hausts, or in close proximity to occupants;
when identifying the limitations and requirements for the use 11. Single sensors in the return air should not be used for
of CO2-based DCV, the realistic performance expectations multiple spaces with very different occupancies;
and published limitations for specific applications, hopefully 12. Economizers should be allowed to override DCV; and
substantiated by both laboratory and full-scale testing before 13. Higher outdoor levels of CO2 will result in overventilation
including it as a Society-endorsed method of control. when levels are assumed (not measured) and an outdoor sensor
may be required by applicable standards or codes.
Other Defficiencies of Indirect Ventilation Control Persily included a list of questions that remain, suggesting
In literature reviews on CO2-based DCV, it was apparent that that the current use of the CO2 -based DCV control strategy is
many opinions exist regarding the validity of the control strategy more risky and less predictable than other, more direct methods,
in relation to the specifics of an application. The “voice of rea- and that reliable applications research is still lacking. Some of
son” originates from a qualified source—specifically, Andrew the questions also indicate limitations that designers and users
Persily, Ph.D., Fellow ASHRAE, who is a former chair of SSPC should seriously consider before implementing this control
62.1 and an employee of the National Institute for Standards strategy:
and Technology (NIST). 1. Is it acceptable to use a single sensor in a common return
As leader of the Indoor Air Quality and Ventilation Group, for multiple zones with similar expected occupancies?
Building Environment Division at the Building and Fire Re- 2. Can a lower setpoint compensate for differences in con-
search Laboratory in Gaithersburg, Md., Persily is in the middle centrations between zones?
of almost everything IAQ. He has been prolific and his work in 3. How much could this approach reduce energy savings?
the field has been exemplary. 4. Are there significant advantages to using a single sensor
In particular, the NIST report for the CEC mentioned earlier with multiple measurement locations (eliminating the com-
included a number of conclusions based on the reference ma- pounding of multiple-sensor error rates)?
terials available prior to publication. An examination of them 5. Should CO2 setpoints be varied for buildings with occu-
highlights many of the limitations of current CO2 sensor tech- pants whose CO2 generation is expected to vary?
nology and those inherent in the use of indirect measurements 6. Is a control algorithm that maintains a constant ventilation
for control (controlled values tend to grow larger or smaller rate per person necessary for acceptable IAQ?
very quickly, due to the magnification of combined errors and 7. What level of minimum ventilation is needed?
uncertainties). Climate, occupancy, operating hours and other 8. Can scheduled purges replace the minimum ventilation
building and HVAC system features make the savings expecta- rate?
tions extremely variable and not guaranteed.5 9. Is displacement ventilation an appropriate and compatible
Some conclusions about CO2 sensor technology and applica- distribution design with DCV? If so, where should the sensors
tions were made from a NIST review of current literature on the be located and can the setpoint be lowered?
subject.5 Some of the relevant ones include the following:
1. The greatest savings likely are to occur in buildings with Analysis of Risk and Benefits
large heating or cooling loads and with dense and unpredict- Compared to the potential benefit, implementing a strategy
able occupancies; with significant assumption-flaws is extremely risky. Literature
2. DCV may not be appropriate in mild climates; suggests that such a gain could be realized if outdoor airflow
3. Avoid DCV in spaces with significant sources other than rates were maintained at acceptable levels.8,10,25–29 Table 3 il-
people; lustrates the potential benefit of a 5% productivity gain. When
4. Avoid buildings with CO2 removal mechanisms; compared to potential energy savings, improved ventilation and
5. Both non-dispersive infrared detection (NDIR) and photo- Indoor Environmental Quality (IEQ) significantly outweighs
metric detection can be affected by light source aging—NDIR “minimized” ventilation.30
by particle buildup and photometric by vibration or atmospheric Based on Addendum 62n changes, the reduction in outside

6 ASHRAE Journal [Link] October 2004

airflow rates with decreasing occupancy in low density zones intake. This additional control input would allow you to verify
(i.e., offices) is small. A 57% reduction in the occupancy (seven and maintain the design minimum levels of outside air “under
to three people per 1,000 ft2)†† only results in a 21% decrease any load condition,”4 as required by Section 5.3 of ASHRAE
in the required outside air. When we consider that the entire Standard 62-2001. The user could guarantee an operational
energy bill for a typical owner-occupied structure amounts to ceiling, a maximum intake rate based on design calculations
about 1% of the annual cost to the building owner,30 is this or preference, never to be exceeded and which should not be
limited savings really worth the risk of potential liability and overridden by other control inputs. Any reliable method of
the loss of productivity benefits from an improved working or occupancy determination may then be used to reset the intake
learning environment? flow rates between the predetermined minimum base rate (in-
In addition, recent changes to commercial general liability cluding differential CO2) and the design maximum, for energy
insurance policies exclude optimization and verifiable
compensation for mold dam- Annual Annual Annual compliance with Standard 62
Income Benefit/Person Benefit/ft2 *
age. Lowering outside air or any code-mandated ventila-
$20,000 $1,000 $6.94 tion rate requirement.16
ventilation rates decreases the
$40,000 $2,000 $13.89
margin of error in pressure Direct measurement of out-
$60,000 $3,000 $20.83
control. Effective and stable side air intake rates has been
$80,000 $4,000 $27.78
pressurization strategies, such $5,000 $34.72
demonstrated to eliminate
$100,000
as those that directly control * Based on 144 ft per person.
2
the uncertainty of indirect
the pressurization flow, or measurement, even though the
Table 3: Potential benefit of 5% productivity gain.
the intake/exhaust and sup- application may be considered
ply/return air volumetric “too difficult” by those unfa-
differentials, will have to be miliar with the latest research
implemented to minimize designer and owner risk, bolstering findings for the application of the most recognizable velocity
the benefits to the occupants. measurement technologies.
The tradeoff between direct vs. indirect measurements usu-
Conclusions and Commentary ally tends to be between accuracy and effort, respectively. The
Clearly, energy benefits can be realized by implementing a cost differences can sometimes be exchanged for accuracy
demand controlled ventilation strategy, if the number of oc- gained or lost. When both are similar in total cost, the decision
cupants and the actual ventilation rate (cfm/person) can be should always go to direct measurement, which delivers less
determined with a reasonable degree of accuracy; and if build- uncertainty and less risk of error, and therefore less risk of the
ing pressure can be maintained simultaneously. Unfortunately, resultant impacts from under- or overventilation.
assumptions needed to justify the CO2-DCV approach leave Improvement in indirect measurement accuracy can be made
designers and owners vulnerable to unnecessary risk. Changes in some situations by extending the time intervals involved
to the ventilation rate procedure of ASHRAE Standard 62 in and/or recalibrating sensors more frequently to account for drift.
Addendum 62n12 may result in diminished benefits from this When the risk of errors or the cumulative error rate becomes
technique, hence, higher risk. too large, the method itself must be questioned. Without direct
We have attempted to demonstrate that the use of the steady- measurement for comparison in the same system, only theory
state formula previously discussed, combined with suitable can be argued.
PID controls and one of the currently available types of CO2 Some velocity measurement devices or combinations of
sensors is not supported by sufficient scientific authority to velocity pressure components are just as unworthy of con-
avoid the risks associated with its use. Designers and owners sideration for outside air application. Some technologies can
should weigh the risks and benefits prior to implementing a claim to be more exact than others and possess documentation
CO2 -DCV strategy. If and when selected, CO2 -DCV applica- to support the claim. Others may be easier or more difficult to
tions should be limited to only those spaces with high densi- apply, less or more repeatable, susceptible or resistant to fouling,
ties and unpredictably variable or intermittent occupancy, and requiring regular and repetitive recalibration or never needing
only after having provided a reliable method to maintain a periodic adjustments.
continuous base ventilation rate, while preserving a minimum The key to the successful implementation of any automatic
pressurization flow. control strategy is the repeatability of the sensing system or
One way to limit the risk without sacrificing the potential method. Until CO2-based DCV can be realistically demonstrat-
energy benefits of DCV is to add a suitable and reliable, duct ed to challenge the performance of direct intake measurement,
or plenum mounted airflow measuring station in the outside air with scientific validity, it will always convey a larger risk of
noncompliance with any rate-based standard than with direct
†† 2
ft × 0.0929 = m2 methods of ventilation control.

October 2004 ASHRAE Journal 7

References Board to Mr. Brian Alcorn, Energy Efficiency and Demand Analysis
1. Fanger, P.O. 1988. “Introduction to the olf and decipol units to Division, California Energy Commission.
quantify air pollution perceived by humans indoors and outdoors.” 25. Djukanovic, R., P. Wargocki and P.O. Fanger. 2002. “Cost-benefit
Energy and Buildings 12. analysis of improved air quality in an office building.” Proceedings of
2. Fanger, P.O. 1990. “New principles for a future ventilation Indoor Air 2002. Monterey, Calif. 1:808–813.
standard.” Fifth International Conference on Indoor Air Quality and 26. Fisk, W.J. and [Link]. 1997. “Estimates of Improved
Climate. Toronto. Pgs. 353–363. Productivity and Health From Better Indoor Environments”, Law-
3. International Mechanical Code. 2003. Section 403.3.1. rence Berkeley National Lab & U.S. Department of Energy ISSN
4. ANSI/ASHRAE Standard 62-2001, Ventilation for Acceptable 095-6947
Indoor Air Quality. 27. Wargocki, P. 2002. “Making the case for IAQ.” ASHRAE IAQ
5. Persily, A., S. Emmerich. 2001. State-of-the-Art Review of CO2 Applications 3(4).
Demand Controlled Ventilation Technology and Application. NISTIR 28. Wargocki, P., [Link] and P.O. Fanger. 2000. “Pollution source
6729, National Institute of Standards and Technology. control and ventilation improve health, comfort and productivity.”
6. Cain, W., R. Leaderer, [Link]. 1983. “Ventilation requirements in Proceedings of Cold Climate HVAC, Pgs. 445–450.
buildings—I. Control of occupancy odor and tobacco smoke.” Atmo- 29. Wargocki, P., [Link]. 2002. “Subjective perceptions, symptom
spheric Environment 17(6):1183–1197. intensity and performance: a comparison of two independent studies,
7. Fanger, P.O. 1990. “New principles for a future ventilation both changing similarly the pollution load in an office.” Indoor Air
standard.” Fifth International Conference on Indoor Air Quality and 12:74–80.
Climate. Toronto. Pgs. 353–363. 30. BOMA, NIBS-WBDG, Minn. Dept. of Admin., et al. www.
8. Kohloss, F. 2003. History And Background of Ventilation Rates. [Link]/IHP/ihp_biblio.html.
Seminar 4. ASHRAE Annual Meeting.
9. Persily, A., 2000. “The relationship between indoor air quality
and carbon dioxide,” Indoor Air 1996 2:961–966. [Link]
gov/bfrlpubs/build96/[Link].
10. Taylor, S. 2003. Rationale for Minimum Ventilation Rates,
Seminar 4, ASHRAE Annual Meeting.
11. ASTM D 6245-98. 1998. “Standard guide for using indoor car-
bon dioxide concentrations to evaluate indoor air quality and ventila-
tion.” American Society for Testing Materials: Conshohocken, Pa.
12. Persily, A. and W.S. Dols. 1990. “The relation of CO2 concentra-
tion to office building ventilation.” ASTM Special Technical Publica-
tion 1067-1990:77–91.
13. Lawrence Berkeley National laboratory. 2003. “Ventilation rates
and technologies—How are CO2 concentrations related to ventilation
rates?” Indoor Environment Division Web site. Ventilation and Indoor
Environmental Quality. [Link]
Downloaded July 22, 2003.
14. Murray, S. Solving Roof Leaks with Fans. Roof Consultants
Institute, 14th International Conference and Trade Show, Canada.
15. Seppänen, O., W.J. Fisk and M.J. Mendel. 1999. “Association
of ventilation rates and CO2-concentrations with health and other
responses in commercial and institutional buildings,” Indoor Air 9:
226-252. Also published as LBNL document No. 43334.
16. Ebtron. 2000. IAQ by Design Brochure and Seminar Presenta-
tions, Loris, S.C.
17. Weber, B. 2003. “Unwrapping modern building envelopes.” A&E
Perspectives [Link]/news/ae/[Link].
18. Ask, A. 2003. “Building scientists meet in new england: air
barriers are more important than ever; vapor barriers are not.” Today’s
A/C & Refrigeration News September.
19. Ask, A.C. 2003. “Ventilation and air leakage.” ASHRAE Journal
45(11):28–36.
20. Harkins, P. 2003. “Moisture control, remediation are keys to
mold control in buildings.” National Underwriter 107(7):21.
21. MacPhaul, D. 2003. “The basics of hvac systems in humid cli-
mates - what goes wrong and why.” Energy Engineering 100(3).
22. Mold Litigation Task Force of the AGC of America. 2003.
“Managing the risk of mold in the construction of buildings.” Con-
structor Magazine. May.
23. Murray, S. “Solving roof leaks with fans.” Roof Consultants
Institute: 14th International Conference and Trade Show.
24. Jenkins, P. 2002. Letter dated Dec. 20, 2002. Manager of Indoor
Exposure Assessment Section of the California State Air Resources

8 ASHRAE Journal [Link] October 2004