A S H RA E JOURNAL
1 can be applied to any multiple-space system, but it does not VAV Rooftop Unit
account for the local recirculation that occurs in fan-powered or Air Handler
Outdoor
VAV systems or in dual-fan, dual-duct VAV systems. Conse-
Air
quently, it results in conservatively low ventilation system effi-
ciency for VAV systems that provide more than one ventilation Exhaust Recirculated Primary
path to each space. Generalized equations for ventilation sys- Air Air Air
tem effectiveness, applicable to a wider range of multiple-space
Return
systems, have been derived but are not discussed here.2, 3
Air
In addition, real buildings often include a mixture of ventila-
tion systems. Total ventilation airflow (Vot) must be found sepa- VAV
rately for each system in a building. Boxes
Supply
Challenge 2: Air
Space 1 Space 2 Space 3
Operate with Required Ventilation at All Loads Vd =5000 cfm Vd =5000 cfm Vd =5000 cfm
The design procedure described earlier identifies worst-case Rp =20 cvm/P Rp =20 cvm/P Rp =20 cvm/P
total outdoor airflow, but what happens to ventilation require- Pd =25 people Pd =30 people Pd =47 people
ments during normal operation? Table 1, Column A illustrates a DVR =500 cfm DVR =600 cfm DVR =700 cfm
VAV system at full load and at two part-load conditions. Eac =1.0 Eac =1.0 Eac =1.0
Vm =1000 cfm Vm =1000 cfm Vm =1000 cfm
zm =0.50 zm =0.60 zm =0.70
Full Load
For illustration, assume that the critical space requires mini- Figure 1: Simple three-space, single-duct VAV system.
mum primary airflow for its thermal load, while the other spaces
require much more primary airflow and the system fan delivers
design primary airflow. This full-load condition corresponds to met with a fixed-position outdoor air damper? No. A fixed out-
the worst-case design scenario described above. Ventilation door air damper position results in relatively constant outdoor
system efficiency (Ev = 1 + 0.180 0.700 = 0.480) determines airflow percentage, but a variable outdoor airflow volume.
total outdoor airflow (Vot = 1,800 0.480 = 3,750 cfm [1770 L/s]). Table 1, Column B, illustrates the results of a common de-
sign mistake: fixing the OA damper position to deliver average
70% Load ventilation fraction (X = 0.18 or 18% outdoor air) at all operating
On a cool summer morning with a 70% cooling load, space conditions. It results in underventilation at full load (1,800 cfm
primary airflow values change, but assuming no change in space [850 L/s] versus 3,750 cfm [1770 L/s]) and at part load (1,260 cfm
population, space ventilation airflow values remain constant at [595 L/s] versus 2,380 cfm [1123 L/s] and 630 cfm [297 L/s]
design (Vo = DVR). The critical ventilation fraction (Z = 0.500) versus 2,210 cfm [1043 L/s]).
and the average ventilation fraction (X = 1800 7000 = 0.257) Some designers fix the OA damper position to deliver critical
change as shown. The ventilation system efficiency increases ventilation fraction (Z = 0.70 or 70% outdoor air) at all loads.
at this load condition (Ev = 1 + 0.257 0.500 = 0.757). Less This results in overventilation (7,000 cfm [3300 L/s] vs 3,750
overventilation occurs in non-critical spaces so less unused cfm [1770 L/s]) both at design and at part load.
dilution capacity exhausts. The system requires less total out-
door airflow (Vot = 1,800 0.757 = 2,380 cfm [1123 L/s]) even Fixed Outdoor Airflow
though required space ventilation airflow (Vo = DVR) remains It takes the right volume, not percentage, of outdoor air to
unchanged. meet the ventilation requirement in Standard 62-1989. Using
direct or indirect sensing, total outdoor airflow volume must be
35% Load maintained at or above the worst-case minimum total outdoor
On an autumn day with a 35% cooling load, primary airflow airflow (Vot) at all load conditions by adjusting the outdoor air
values change again. As was true for the 70% load condition, damper position as primary airflow changes. Table 1, Column
the current ventilation system efficiency dictates the total out- C, shows the effect of maintaining outdoor airflow at the design
door airflow requirement. At this condition, critical space venti- value (3,750 cfm [1770 L/s]), which is overventilation at all part-
lation fraction (Z = 0.700) and current average ventilation frac- load conditions.
tion (X = 0.514) result in increased ventilation system efficiency
(Ev = 1 + 0.514 0.700 = 0.814) with decreased total outdoor Challenge 3:
airflow (Vot = 1,800 0.814 = 2,210 cfm [1043 L/s]). Control to Minimize Ventilation Energy Impact
The design procedure discussed earlier yields a worst-case
Fixed Outdoor Air Damper Position total outdoor airflow. However, ventilation system efficiency in-
Traditional VAV systems set minimum outdoor airflow us- creases during normal operation since less unused dilution ca-
ing an outdoor air (OA) damper at a fixed minimum position; pacity exhausts at part load. Can the ventilation system be con-
actual outdoor air intake volume, of course, varies as total pri- trolled to assure proper ventilation without costly overventila-
mary airflow varies. Can the ventilation volume requirement be tion? Yes, using a ventilation reset scheme as explained later.
October 1998 ASHRAE Journal 41
Space 1 Space 2 Space 3 Sy stem
Design Ventilation Rate (cfm), DVR = 500 600 700 DVR = 1,800
ASHRAE Requirements
Standard 62-1989 states that when mechanical ventila- Population and Load Diversity
tion is used, provision for airflow measurement should be in-
cluded, and that for VAV systems, provision shall be Although not specifically included in Standard 62-1989,
made to maintain acceptable indoor air quality throughout some designers account for population diversity when
the occupied zone. And, with regard to multiple-space system designing ventilation systems. At any given time, the ac-
ventilation, it states: Where more than one space is served by tual population may be design value (P = Pd) in one space
a common supply system the system outdoor air quantity and less than design value in all other spaces. Since all
shall then be determined using Equation 6-1. spaces are not at design population simultaneously, it fol-
In other words, proper ventilation for multiple-space VAV lows that the design ventilation requirements for the sys-
systems must be determined using the MSE, outdoor airflow tem could be based on the highest population in any given
measurement and control. It follows that a control scheme that time block rather than sum-of-peak population (Pd).
senses both space primary airflow and total outdoor airflow To account for system population diversity some de-
and solves the MSE dynamically to reset the total outdoor signers estimate actual space population for each hour,
airflow setpoint (Vot) will satisfy Standard 62s ventilation re- then find a population diversity factorthe ratio of the
quirements. The term ventilation reset is used to describe this highest population in any hourly block to the sum-of-
control scheme. peak population (PF = Pb Pd) and use it to reduce
total ventilation airflow accordingly (Von = PF DVR).
Basic Ventilation Reset Conservatively assuming that actual population always
Assuming space population and ventilation requirements equals design population (P = Pd and PF = 1.0) results
remain constant (Vo = DVR), decreased primary airflow leads to in overventilation at all loads. On the other hand, aggres-
higher ventilation system efficiency at part load and a decreased sively estimating block population (Pb << Pd and PF
need for first-pass outdoor airflow. A basic ventilation reset << 1.0) can result in significant underventilation at all
control scheme senses actual primary airflow (Vp) and calcu- loads.
lates the current ventilation fraction (z) for each space, finds Most VAV system designers account for load diver-
the current total outdoor airflow setpoint (i.e. finds Z, Von and sity when finding the primary airflow. At any given time,
Vpt, then solves MSE to find Vot) for the system, and finally actual cooling load may equal design value (Q = Qd) in
modulates total outdoor airflow to maintain the new setpoint. some spaces and less than design value in other spaces.
More sophisticated control schemes that account for changes It follows that total primary airflow for the VAV system at
in population and recalculate ventilation requirements (Vo = P design conditions can be based on the highest space
Rp) can also be used. For instance, systems with predictable cooling load in a given time block rather than sum-of-
space population can use a simple time-of-day schedule to es- peak space cooling loads.
timate population, recalculate Vo and reset total outdoor air- To account for system load diversity, hour-by-hour
flow (Vot) as described above. space load can be used to find the highest system cool-
Alternatively, actual space population may be determined, ing load (Qb) during any hourly time block. Then, load
either directly by people-counting or indirectly by sensing a diversity factor can be found as the ratio of the highest
tracer gas (e.g. differential CO2). Actual population can then be load in any time block to the sum-of-peak load at design
used to recalculate Vo and reset Vot, again, by solving MSE conditions (DF = Qb Qd) and used to reduce total
dynamically. ventilation airflow accordingly (Vpt = DF Vd).
Implementation. Basic ventilation reset is easily implemented
42 ASHRAE Journal October 1998
Outdoor Air, Vot, Delivered
A B C
P rimary Airflow, Vp 5,000 cfm 4,000 cfm 1,000 cfm
37% or 18% or 37% or
100% Ventilation Rate, DVR 500 cfm 600 cfm 700 cfm
3,750 cfm 1,800 cfm 3,750 cfm
Avg. Ventilation Fraction, z 0 .1 0 0 .1 5 0 .7 0
using the data-rich environment afforded by communicating with sufficient capacity, operated to ventilate properly, and
controls throughout the VAV system. Specifically, use pres- controlled to avoid significant operating cost increases. Are
sure-independent VAV terminal units with direct digital con- there any other concerns related to VAV system ventilation?
trols (DDC/VAV), a building automation system (BAS) with Absolutely.
simple equation-solving capability, an air handler with DDC
controls and a means for measuring and maintaining outdoor Freeze Protection and Mixing
airflow. Figure 3 illustrates this implementation. Proper VAV system ventilation leads to a high percentage of
In operation, individual DDC/VAV controllers sense primary outdoor air at part load. During cold weather, mixed air tempera-
airflow (Vp) in the spaces they serve and calculate current ture can approach outdoor air. The mixed airstream can stratify
space ventilation fraction (z = Vo Eac Vp) using values and water in coils can freeze.
entered or calculated for Vo and Eac. The BAS continuously Water in coils can be protected from freezing in many ways.
polls each DDC/ VAV controller, collecting current primary air- For example, drain the coils, keep the water in motion, add gly-
flow Vp, required space ventilation airflow Vo and current ven- col to the water, use face-and-bypass dampers to limit cold
tilation fraction z from each. It then calculates total primary airflow or preheat the outdoor air.
airflow (Vpt = Vp), total design ventilation airflow (Von = Streams of air at very different temperatures are difficult to mix.
Vo), and average ventilation fraction (X = Von Vpt), deter- Air blending devices offer effective mixing but usually require
mines critical space ventilation fraction (Z = largest z), and uses more space (longer air handlers) and more fan horsepower. As an
these values to find ventilation system efficiency (Ev = 1 + X alternative, preheating the outdoor air before the mixing box de-
Z) and a new outdoor airflow setpoint (Vot = Von Ev). The air creases the airstream temperature differential, allowing good mix-
handler DDC controller receives the new setpoint and modu- ing without blending and offering freeze protection as well.
lates outdoor airflow as required to maintain it.
Whats It Worth? Basic ventilation reset control reduces Airflow Sensing
the outdoor airflow at part load and saves operating costs. Can Outdoor airflow must be determined or sensed for proper
it be justified? That depends on building type and location. VAV system ventilation. Many indirect airflow-sensing schemes
The outdoor air load in an office building in Miami, for instance, (e.g. fan tracking, differential temperature sensing4) are unable
may be reduced by 8% while the same building in Toronto or to meet the 10% accuracy criterion typical for air balancing
Minneapolis may realize a 30% reduction. An economic analy- measurements. Although direct-sensing methods using inde-
sis program can be used to quantify the savings associated pendent airflow-monitoring stations can be designed to meet
with ventilation reset. accuracy needs, they tend to be costly (requiring field installa-
tion and field calibration) and difficult to apply (straight duct-
Other Challenges? length requirements). Injection fan schemes with airflow sens-
As shown earlier, VAV ventilation systems can be designed ing and fan speed control can be used successfully although
44 ASHRAE Journal October 1998
VENTILATION