EPA-600/R-97-131
November 1997
ENERGY COSTS OF IAQ CONTROL THROUGH INCREASED VENTILATION
IN A SMALL OFFICE IN A WARM, HUMID CLIMATE:
¦¦k m ¦ ¦ ¦ ¦ * m t w"k. r" ^ ¦¦ a ¦ ¦
Parametric Analysis Using the DOE-2 Computer Model
by
D. Bruce Henschel
Air Pollution Prevention and Control Division
National Risk Management Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
Prepared for
U. S. Environmental Protection Agency
Office of Research and Development
Washington, D. C. 20460
-------
TECHNICAL REPORT DATA ... .,
(Please read Instructions on the reverse before comple [|| |||[ || ||||| || |[|| 11 ||| || 11
1. REPORT NO. 2.
EPA-600/R-97-131
3, in mi it iiaai ii i ¦ ii 11 in mi
PB98-113368
4. TITLE AND SUBTITLE
Energy Costs of IAQ Control Through Increased Ven-
tilation in a Small Office in a Warm, Humid Climate:
Parametric Analysis Using the DOE-2 Computer Mode
5. REPORT DATE
November 1997
6. PERFORMING ORGANIZATION CODE
1
7. AUTHOR(S)
D. Bruce Henschel
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
See Block 12
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
NA (Inhouse)
12. sponsoring agency name and address
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NO 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 9/95 - 3/97
14. SPONSORING AGENCY CODE
EPA/600/13
15. supplementary notesAPPCD pro;ject 0fficer is Bruce Henschel, Mail Drop 54, 919/
541-4112.
is. abstractTke rep0rt gives results of a series of computer runs using the DOE-2. IE
building energy model, simulating a small office in a hot, humid climate (Miami).
These simulations assessed the energy and relative humidity (RII) penalties when the
outdoor air (OA) ventilation rate is increased from 5 to 20 cfm/person to improve
indoor air quality. The effect was systematically assessed of each building and
mechanical system parameter on the energy penalty resulting from increased OA.
The cost and effectiveness were also assessed of methods for reducing, elevated-RH
hours. The parameters offering the greatest practical potential for energy savings
are conversion to: very efficient lighting and equipment; very efficient cooling coils;
to a variable air volume (from a constant volume) system; cold-air distribution; or
improved glazing or roof resistance to heat transfer. If the OA increase were ac-
companied by any one of these modifications, the energy penalty would be significant-
ly reduced (comparing the modified system at 20 against the baseline at 5 cfm/per-
son). The number of occupied hours above 60% RH could be dramatically reduced
(with a minimal energy cost impact) if the economizer were eliminated. Conversion
to a system that controlled office humidity would eliminate all of the elevated-RH
occupied hours, at an energy cost of $90/year.
17. 1 KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COS ATI Field/Group
Pollution
Energy
Expenses
V entilation
Humidity
Computerized Simulation
Pollution Control
Stationary Sources
Indoor Air Quality
13 B
14G
05C, 14A
13 A
04B
18. distribution statement
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
192
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
FOREWORD
The U. S. Environmental Protection Agency is charged by Congress with pro-
tecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions lead-
ing to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research
program is providing data and technical support for solving environmental pro-
blems today and building a science knowledge base necessary to manage our eco-
logical resources wisely, understand how pollutants affect our health, and pre-
vent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites and groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regulations
and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by EPA's Office of Re-
search and Development to assist the user community and to link researchers
with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
ii
-------
ABSTRACT
A series of computer runs has been completed using the DOE-2.1E building
energy model, simulating a small (4,000 ft2) strip mall office cooled by two packaged
single-zone systems, in a hot, humid climate (Miami). These simulations assessed the
energy penalty, and the impact on indoor relative humidity (RH), when the outdoor air
(OA) ventilation rate of the office is increased from 5 to 20 cfm/person in this
challenging climate to improve indoor air quality. One objective was to systematically
assess how each parameter associated with the building and with the mechanical
system impacts the energy penalty resulting from increased OA. Another objective
was to assess the cost and effectiveness of off-hour thermostat set-up (vs. system
shut-down), and of humidity control (using overcooling with reheat), as means for
reducing the number of hours that the office space is at an RH above 60% at the 20
cfm/person ventilation rate.
With the baseline set of variables selected for this analysis, an OA increase
from 5 to 20 cfm/person is predicted to increase the annual cost of energy consumed
by the heating, ventilating, and air-conditioning (HVAC) system by 12.9%. The
analysis showed that the parameters offering the greatest practical potential for
energy savings are conversion to very efficient lighting and equipment (1.5 W/ft2) and
conversion to very efficient cooling coils (electric input ratio = 0.284). If the increase
to 20 cfm/person were accompanied by either of these conversions, the 12.9% HVAC
energy penalty for the increased OA rate would be eliminated; the modified system
at 20 cfm/person would have a tower annual HVAC energy cost than the baseline
system at 5 cfm/person. Other parameters offering significant practical potential for
energy savings are: conversion from packaged single-zone units to a variable air
volume system; conversion to cold-air distribution (minimum supply air temperature =
42 °F); or improvements in the glazing or in the roof resistance to heat transfer. If the
OA increase were accompanied by any one of these modifications, the 12.9% penalty
would be reduced to between 2 and 7% (the modified system at 20 compared against
the baseline at 5 cfm/person).
According to the DOE-2.1E model, the increase in ventilation rate could be
achieved with an 85% reduction in the number of occupied hours above 60% RH,
compared to the baseline system at 5 cfm/person — with only a $19/year increase in
energy cost - if the economizer were eliminated. That is, most of the elevated-RH
hours in the baseline case were predicted to be the result of economizer operation.
If the control system were modified so that it controlled the humidity as well as the
temperature in the office space, all of the elevated-RH occupied hours would be
eliminated, at an energy cost of $90/year.
Neither economizer elimination nor humidity control would address i/rtoceupied
periods, when most of the elevated-RH hours occur. Building operators concerned
about biological growth at elevated RH should consider operation of the cooling
iii
-------
system during unoccupied hours, perhaps with the thermostat set up, rather than
system shut-down off-hours. Off-hour set-up from 75 to 81 °F would add only
S1O/'year to energy costs, and would provide some modest reduction in unoccupied
eievated-RH hours. Set-up to 79 °F would provide a greater reduction, at an energy
cost of $38/year.
DOE-2.1E underestimates the number of elevated-RH hours because it does not
address the moisture capacitance of building materials and furnishings, or re-
evaporation off the cooling coils when they cycle off with the air handler operating.
As a result, the performance of the RH reduction steps above may be overestimated,
or the costs of the steps underestimated,
iv
-------
TABLE OF CONTENTS
Page
Abstract to
List of Figures i*
List of Tables x
Metric Conversion Factors
Section 1. Objectives and Approach 1-1
1.1 Background 1-1
1.2 Overall Approach 1-3
1.3 Objectives 1-3
1.4 Limitations 1-4
Section 2. Executive Summary 2-1
2.1 Objectives and Approach 2-1
2.2 The Baseline Building and HVAC System 2-2
2.3 The Impact of Building and HVAC Parameters on the
Penalties Associated with Increased Ventilation 2-2
2.3.1 Parameters Creating the Greatest Reductions
in HVAC Energy 2-9
2.3.2 Parameters Creating the Greatest Reductions
in Hours at Elevated RH 2-13
2.4 The Impact of Steps to Reduce Indoor Humidity 2-16
2.4.1 Thermostat Set-Up vs. System Shut-Down 2-17
2.4.2 Humidity Control by Overcooling and Reheat 2-18
Section 3. Description of the Study 3-1
3.1 The Software 3-1
3.2 The Baseline Building 3-2
3.3 The Baseline HVAC System 3-3
3.4 The "Plant" 3-8
3.5 Energy Costs 3-8
3.6 Weather Data 3-8
3.7 Parametric Variations from the Baseline Values 3-9
3.8 Strategy for the Calculations 3-9
V
-------
TABLE OF CONTENTS (continued)
Paae
%
Section 4. Baseline Results 4-1
4.1 Summary Tables 4-1
4.2 Discussion 4-1
Section 5. The Effects of Parametric Variations 5-1
5.1 Building Orientation 5-1
5.1.1 The Effect of Orientation at 5 cfm/person 5-1
5.1.2 The Effect of Orientation on Increased Ventilation Rates 5-3
5.2 Building Shade 5-4
5.2.1 The Effect of Shade at 5 cfm/person 5-4
5.2.2 The Effect of Shade on Increased Ventilation Rates 5-4
5.3 Occupant Density 5-6
5.3.1 The Effect of Occupancy at 5 cfm/person 5-8
5.3.2 The Effect of Occupancy on Increased Ventilation Rates 5-8
5.4 Lighting and Office Equipment Power Consumption 5-9
5.4.1 The Effect of Lighting/Equipment Power at 5 cfm/person 5-9
5.4.2 The Effect of Lighting/Equipment Power
on Increased Ventilation Rates 5-11
5.5 Infiltration Rate 5-12
5.5.1 The Effect of Infiltration at 5 cfm/person 5-12
5.5.2 The Effect of Infiltration on Increased Ventilation Rates 5-12
5.6 Exterior Wall Resistance to Heat Transfer 5-14
5.6.1 The Effect of Wall Resistance at 5 cfm/person 5-16
5.6.2 The Effect of Wall Resistance
on Increased Ventilation Rates 5-16
5.7 Amount of Glazing 5-17
5.7.1 The Effect of Glazing Amount at 5 cfm/person 5-17
5.7.2 The Effect of Glazing Amount
on Increased Ventilation Rates 5-20
5.8 Glass Type 5-21
5.8.1 The Effect of Glass Type at 5 cfm/person 5-23
5.8.2 The Effect of Glass Type on Increased Ventilation Rates 5-24
5.9 Roof Resistance to Heat Transfer 5-24
5.9.1 The Effect of Roof Resistance at 5 cfm/person 5-24
5.9.2 The Effect of Roof Resistance
on Increased Ventilation Rates 5-28
vi
-------
TABLE OF CONTENTS (continued)
Page
5.10
Elimination of All Exterior Surfaces
5-29
5.10.1 The Effect of Infinite Exterior Resistance
at 5 cfm/person
5-29
5.10.2 The Effect of Infinite Exterior Resistance
on Increased Ventilation Rates
5-29
5.11
Thermostat Set-Up Rather Than System Shut-Down
During Unoccupied Hours
5-31
5.12
Alternative HVAC Systems
5-34
5.12.1 The Effect of Alternative HVAC Systems
at 5 cfm/person
5-37
5.12.2 The Effect of Alternative HVAC Systems
on Increased Ventilation Rates
5-47
5,13
Ducted Air Returns
5-52
5.13.1 The Effect of Ducted Returns at 5 cfm/person
5-52
5.13.2 The Effect of Ducted Returns
on Increased Ventilation Rates
5-55
5.14
Use of Cold-Air Distribution
5-56
5.14.1 The Effect of Cold-Air Distribution at 5 cfm/person
5-57
5.14.2 The Effect of Cold-Air Distribution
on Increased Ventilation Rates
5-63
5.15
Modifications to the Economizer
5-70
5.15.1 The Effect of Eliminating the Economizer
5-72
5.15.2 The Effect of an Enthalpy-Controlled Economizer
5-72
5.16
Alternative Cooling Electric Input Ratios
5-74
5.16.1 The Effect of EIR at 5 cfm/person
5-74
5.16.2 The Effect of EIR on Increased Ventilation Rates
5-77
5.17
Alternative Cooling Capacities and Sensible Heat Ratios
5-78
5.17.1 The Effect of Cooling Capacity and SHR
at 5 cfm/person
5-78
5.17.2 The Effect of Cooling Capacity
on Increased Ventilation Rates
5-83
5.18
Alternative Weather Input Files
5-86
Section 6.
The Impact of Humidity Control
6-1
6.1
Introduction
6-1
6.1.1 Impacts of Increased OA on RH When
No RH Reduction Steps Are Taken
6-1
6.1.2 Approaches for Reducing the Impact
of Increased OA on RH
6-2
6.1.3 Limitations of the DOE-2 Model for RH Analysis
6-3
vii
-------
TABLE OF CONTENTS (concluded)
Page
6.2 Reducing RH Through Adjustments to HVAC Design
Using Temperature Control Alone 6-3
6.2.1 Results from Other Investigators 6-3
6.2.2 Cooling Coil Set-Up vs. Shut-Down
During Unoccupied Hours 6-7
6.3 Reducing RH Using Enthalpy Control 6-12
6.3.1 The Effect of Enthalpy Control at 5 cfm/person 6-13
6.3.2 The Effect of Enthalpy Control
on Increased Ventilation Rates 6-19
Section 7. References 7-1
APPENDICES
Appendix A. Sample Input DOE-2 File: 4,000 ft2 Office in Strip Mall A-1
Appendix B. Rationale for Values Selected for Variables in
DOE-2.1E Input File: 4,000 ft2 Office in Strip Mall B-1
viii
-------
LIST OF FIGURES
Page
Figure 1. Floor plan for the baseline 4,000 ft2 office in a Miami strip mall 3-5
Figure 2. Front and rear views of the baseline office 3-6
-------
LIST OF TABLES
Page
Table S-1, Effects of Building arid HVAC Variables on HVAC Capacity and
Energy Cost, and on Occupied Hours Above 60% RH 2-3
Table S-2. Building and HVAC Variables Creating the Greatest Reductions
in Annual HVAC Energy Cost at 20 cfm/person (from
Table S-1} 2-8
Table S-3. Building and HVAC Variables Creating the Greatest Reductions
in Hours Above 60% RH at 20 cfm/person (from
Table S-1) 2-10
Table S-4. Approximate Contribution of the Various Heat Sources
to the Annual HVAC Energy Consumption
in the Baseline Building 2-11
Table S-5. Effect of Humidity Control by Overcooling and Reheat 2-20
Table 1. Baseline Building Design and Operating Conditions:
Small Office in Miami Strip Mall 3-4
Table 2. Baseline HVAC Design and Operating Conditions:
Small Office in Miami Strip Mall 3-7
Table 3. Results from DOE-2.1E Modeling of the Baseline Small office
in Miami 4-2
Table 4. Annual Electric Energy Consumption in the Baseline Small
Office, Broken Down According to End Use 4-3
Table 5. Effect of Building Orientation: Increase Compared to Baseline
Case with 5 cfm/person 5-2
Table 6. Effect of Building Shading: Increase Compared to Baseline
Case with 5 cfm/person 5-5
Table 7. Effect of Building Occupancy: Increase Compared to Baseline
Case with 5 cfm/person 5-7
Table 8. Effect of Lighting and Office Equipment Power Consumption:
Increase Compared to Baseline Case with 5 cfm/person 5-10
Table 9. Effect of Infiltration Rate: Increase Compared to Baseline
Case with 5 cfm/person 5-13
Table 10. Effect of Exterior Wall Resistance: Increase Compared to
Baseline Case with 5 cfm/person 5-15
Table 11, Effect of Amount of Glazing: Increase Compared to Baseline
Case with 5 cfm/person 5-18
Table 12. Effect of Glass Type: Increase Compared to Baseline
Case with 5 cfm/person 5-22
X
-------
LIST OF TABLES (continued)
Page
Table 13. Effect of Roof Resistance: Increase Compared to Baseline
Case with 5 cfm/person 5-25
Table 14, Effect of Infinite Resistance for All Exterior Surfaces:
Increase Compared to Baseline Case with 5 cfm/person 5-30
Table 15. Effect of Thermostat Set-Up Rather than System Shut-Down:
Increase Compared to Baseline Case with 5 cfm/person 5-32
Table 16. Effect of Assuming Alternative HVAC Systems: Increase
Compared to Baseline Case with 5 cfm/person 5-35
Table 17. Effect of Ducted Air Returns: Increase Compared to Baseline
Case with 5 cfm/person 5-53
Table 18. Effect of Cold-Air Distribution: Increase Compared to
Baseline Case with 5 cfm/person 5-58
Table 19. Effect of Modifying Economizer Operation: Increase
Compared to Baseline Case with 5 cfm/person 5-73
Table 20. Effect of Cooling EIR: Increase Compared to Baseline
Case with 5 cfm/person 5-75
Table 21. Effect of Varying Cooling Capacity and SHR at 5 cfm/person 5-79
Table 22. Effect of Varying Cooling Capacity and SHR at 20 cfm/person 5-84
Table 23. The Effect of the WYEC vs. the TMY Weather File on the
Results for the Baseline Small Office in Miami 5-88
Table 24. Effect of Enthalpy Control During Occupied Hours, Under
Various Conditions 6-14
xi
-------
METRIC CONVERSION FACTORS
Although it is EPA's policy to use metric units in its documents, non-metric
units have been used in this report consistent with common practice in the heating,
ventilating, and air-conditioning industry. Readers more accustomed to the metric
system may use the following factors to convert to that system.
Non-Metric
foot (ft)
square foot (ft2)
cubic foot per minute
(cfm)
pound (lb)
degrees Fahrenheit (°F)
British thermal unit
(Btu)
Times
0.305
0.0929
37
0.454
5/9 (°F - 32)
0.293
British thermal unit per hour 0.293
(Btu/h)
ton (of refrigeration)
(12,000 Btu/h)
3,520
Yields Metric
meter (m)
square meter (m2)
liters per second (L/s)
kilogram (kg)
degrees Celsius (°C)
watt-hour (W-hr)
watt (W)
watts (of cooling capacity)
xii
-------
SECTION 1
OBJECTIVES AND APPROACH
1.1 BACKGROUND
In 1989, the American Society of Heating, Refrigerating and Air-Conditioning
Engineers (ASHRAEJ issued ANSI/ASHRAE Standard 62-1989 (ASHRAE, 1989a)
recommending increased outdoor air (OA) ventilation rates to maintain acceptable
indoor air quality (IAQ) inside the full array of commercial, institutional, and residential
structures. This standard recommends, for example, an OA ventilation rate of 20 cfm
per person in the office space within office buildings, and 15 cfm/person in the
classroom space within schools. These ventilation rates are greater than those which
had been recommended in the earlier version of Standard 62 (ASHRAE, 1981), which
had recommended 5 cfm/person in non-smoking offices and classrooms.
Increased OA ventilation rate, and improved distribution of the ventilation air,
are perhaps the most commonly utilized techniques for improving IAQ. Accordingly,
IAQ researchers are concerned with the costs associated with increased ventilation
rates, and with possible approaches for reducing these costs, as a means for
increasing the acceptance and effective utilization of this IAQ control technique.
Since the time that increased outdoor air ventilation rates were being consid-
ered during the development of Standard 62-1989, a number of computer simulation
studies have been conducted to estimate the energy consumption and energy cost
impacts of this increase in various building types (Eto and Meyer, 1988; Eto, 1990;
Steele and Brown, 1990; Ventresca, 1990; Mudarri and Hall, 1993; Rojeski et al.,
1995; Shirey and Rengarajan, 1996). Most, although not all, of the published
simulation studies have involved the use of DOE-2, the building energy simulation
software developed for the U. S. Department of Energy (York et al., 1981; U. S.
Department of Energy, 1994a). These studies estimate that increasing OA from 5 to
20 cfm/person in office buildings would increase total annual building energy
consumption and cost by 0 to 5% in large office buildings, and by 2 to 15% in
medium to small office buildings, depending upon climate.
Some of the simulation studies cited above addressed a single, specific building
with a specific heating, ventilating, and air-conditioning (HVAC) system. Other
studies considered multiple buildings. Those that addressed multiple buildings
generally addressed multiple specific buildings with specific HVAC systems; for any
one building, the building parameters and the HVAC design and operating parameters
were generally fixed for the analysis (except for the variation in OA rates).
1-1
-------
Therefore - while these several studies on a variety of realistic buildings consistently
show the energy cost of increased OA to be limited - the studies do not quantify the
extent to which the selection of the many specific building and HVAC parameters
impacts the cost of increasing OA from 5 cfm/person to the higher ASHRAE 62-1989
value. Perhaps the cost of increasing OA might be at least partially offset by judicious
selection of building and HVAC design and operating conditions.
The Mudarri and Hall study did vary a few of the building parameters (internal
configuration, shell heat transfer resistance, and occupant density) and HVAC
parameters (HVAC system type, presence or absence of an economizer, and boiler and
chiller efficiencies) in a large office building. But this study was conducted for the
moderate climate of Washington, D. C., where the cost of increasing OA in a large
office is relatively small. Thus, the incremental effects of these building/HVAC
parameters on the costs of increased OA were difficult to distinguish. Steele and
Brown also addressed the effect of one building parameter, namely, occupant density,
in ten different building types in the Seattle and Richland, WA, climates. It would be
of interest to systematically determine the effects of varying a broader array of
building and HVAC design and operating parameters, in a climate where any effects
of these variations would be as pronounced as possible.
Most of the energy modeling to date has focussed on cold (or temperate)
climates. However, the greatest energy cost increases resulting from increased OA
tend to occur in hot, humid climates, due to the high sensible and latent cooling loads.
In addition to its impact on energy costs, increased OA in humid climates can necessi-
tate added efforts to control the indoor relative humidity (RH) below the upper limit
of 60% recommended by ASHRAE, for purposes both of thermal comfort (ASHRAE,
1992a) and of reduced allergenic/pathogenic organism growth (ASHRAE, 1989a). For
these reasons, some of the greatest concerns about the increased OA rates recom-
mended by ASHRAE 62-1989 have been expressed in regions having hot, humid
climates. Eto and Meyer (1988) and Eto (1990) did include the humid Miami, FL,
climate among the 13 cities that they considered; these analyses predicted that,
among the cities studied, Miami would experience the greatest increase in energy
consumption and costs resulting from the OA increase, in medium-sized and large
offices. However, these two studies did not address the impact of increased OA on
indoor RH levels, or any energy penalty resulting from an effort to control RH
increases created by the increase in ventilation rate.
Among the published simulation studies, only Shirey and Rengarajan have
attempted to rigorously address the cost impacts of controlling indoor humidity when
increasing OA in humid climates. This latter study showed that an increase from 5
to 20 cfm/person in a small office in Miami would indeed increase annual total building
energy costs by roughly 6%, consistent with Eto (1990), if one addresses only the
control of temperature in the office. But if one wishes to simultaneously control RH
to less than or equal to 60%, special steps would be required which could cause the
1-2
-------
energy cost increase to rise to 10 to 15%, depending upon the humidity control
approach that is taken. Taking into consideration the increased installation costs (as
well as the energy costs), Shirey and Rengarajan estimate that an increase from 5 to
20 cfm OA/person could increase the life cycle costs of the HVAC system (including
installed hardware plus energy) by 7 to 32%, if indoor RH is to be controlled to 60%
or less. Because standard over-cooling and reheat of the supply air to control
enthalpy is generally prohibited by Florida code (FDCA, 1993), Shirey and Rengarajan
did not address this approach.
1.2 OVERALL APPROACH
This study involved a systematic series of DOE-2.1E computer simulations to
estimate indoor conditions (i.e., the number of hours at elevated temperature and RH),
and the energy consumption and cost, in a small (4,000 ft2) office in Miami
conditioned by packaged, direct expansion HVAC equipment. This office is similar,
although not identical, to the office modeled by Shirey and Rengarajan.
These simulation runs comprised a parametric analysis to systematically
quantify how each of the potentially important building design and operating variables
(in DOE-2 terminology, the LOADS variables), and each of the HVAC design and
operating variables (the SYSTEMS variables), impact the computed indoor conditions
and the computed energy consumption and cost in that office. For each building and
HVAC parameter, simulations were run to determine the incremental effect of that
parameter on the increase in energy use, and on the change In indoor conditions,
resulting from an increase in OA ventilation rate from 5 to 20 cfm/person.
Particular attention was devoted to assessing the effectiveness and costs of
possible means for controlling RH < 60% using the conventional direct expansion
HVAC systems typical of small offices. This analysis was not as rigorous as that of
Shirey and Rengarajan, but did address the over-cool/reheat approach that was not
analyzed in the other study.
1.3 OBJECTIVES
This effort had three objectives.
1) To assess the extent to which the selection of the baseline building conditions
impacts the computed indoor conditions and energy usage/cost, and the
computed penalty caused by an OA ventilation rate increase from 5 to 20 cfm/
person.
1-3
-------
2) To assess whether variations in specific building and HVAC design and
operating parameters might at least partially offset the increase in energy costs
resulting from the increase in ventilation rate, under the challenging conditions
of a hot, humid climate.
3} To develop a better understanding of the practical issues and costs involved in
controlling indoor RH when the ventilation rate is increased in small offices in
hot, humid climates.
If the costs and the indoor temperature/RH impacts associated with an increase in the
OA rate can be reduced through judicious selection of building and HVAC design and
operating conditions, increased ventilation might be more widely accepted and more
effectively utilized for IAQ control in humid climates.
Another, secondary objective of the study was to develop familiarity with the
DOE-2 software.
1.4 LIMITATIONS
The cost analysis conducted in this report addressed only the energy cost
impacts of the various building and HVAC system modifications. No attempt was
made here to address the impacts on the installation cost of the building or of the
HVAC system, or to address any impacts on maintenance costs.
The standard DOE-2.1E software does not address the moisture capacitance
of the building materials and furnishings; nor does it address condensed moisture that
is re-evaporated off of the cooling coils when the compressor cycles off during air
handler operation (Birdsall, 1995). Shireyand Rengarajan (1996) utilize a customized
version of DOE-2 to conclude that these two moisture-related issues that can have
significant impacts on the computed indoor RH values in humid climates. On worst-
case summer days in Miami, the RH values predicted by DOE-2.1E during HVAC
operating hours might be as much as 10 to 20 percentage points lower than those
predicted by a model that does incorporate these moisture considerations, according
to Shirey and Rengarajan.
If this is true, the DOE-2.1E calculations presented in this report will under-
estimate the humidity levels in the small office, including the number of hours having
RH _> 60% (one of the parameters reported here as a measure of indoor conditions).
This potential problem would also impact the assessment here of the equipment
performances and the energy costs that would be required in order to control indoor
RH levels below 60% as OA is increased from 5 to 20 cfm/person, since the
assessment may be being made using artificially low RH values.
1-4
-------
Despite this potential problem, this analysis still meets its basic objectives,
utilizing what is arguably the most widely used software in the U. S. for modeling
building energy consumption and costs. Specifically, this analysis quantifies how
individual building and HVAC parameters impact energy consumption and costs, and
whether the selection of these parameters might significantly affect the estimated
costs of increased ventilation, even though there might be some question regarding
the accuracy of the predicted RH values. The analysis also provides useful perspec-
tive regarding the effectiveness and relative energy costs of various options for
controlling RH, again despite the uncertainty in the RH values.
1-5
-------
[This page intentionally blank.]
1-6
-------
SECTION 2
EXECUTIVE SUMMARY
Among the three basic techniques for improving IAQ -- improved ventilation,
air cleaning, and source management - improved ventilation is perhaps the most
commonly utilized. In ANSI/ASHRAE Standard 62-1989, ASHRAE recommended that
outdoor air ventilation rates in office space be increased from 5 to 20 cfm/person to
maintain acceptable IAQ. There will be an energy penalty associated with an OA
increase, which will usually be most pronounced in hot, humid climates. Also of
particular concern in humid climates, an OA increase can result in increased indoor RH
levels, which can be of concern both from the standpoint of occupant comfort, and
from the standpoint of fungal growth.
2.1 OBJECTIVES AND APPROACH
To assess these energy and RH penalties associated with increased ventilation,
a systematic series of computer simulations have been run using the DOE-2.1E
software to model a small (4,000 ft2) office in a hot, humid climate (Miami). These
simulation runs comprised a parametric analysis to systematically quantify how each
of the building and HVAC system variables impacts energy consumption and cost, and
HVAC performance (in particular, indoor RH levels), at ventilation rates of both 5 and
20 cfm/person.
By defining the building and HVAC parameters having the greatest impact on
HVAC energy consumption and cost, this assessment was intended to suggest those
parameters which -- if modified in conjunction with the increase in OA -- could at least
partially offset the energy and cost penalties associated with the increased ventilation
rate. Likewise, by defining the parameters having the greatest impact on indoor RH,
the assessment was intended to suggest parametric modifications which could reduce
the RH impacts of the OA increase.
As part of this analysis, the DOE-2.1E model was used to further assess the
energy penalty and the effectiveness of two specific approaches for reducing the
number of hours at RH levels greater than 60%. These approaches are: 1) turning
the thermostat up (rather than shutting the HVAC system down) during unoccupied
cooling hours; and 2) use of a humidity controller on the HVAC system, employing
over-cooling and reheat as necessary to maintain the RH below 60% during occupied
hours.
This analysis did not address the equipment/installation costs associated with
the parametric variations, or any impact of the variables on maintenance costs.
2-1
-------
2.2 THE BASELINE BUILDING AND HVAC SYSTEM
The building type selected for this analysis was a small, one-story office in a
strip mall, with adjoining space (occupied by other tenants) on either side. The office
had a frontage of 40 ft and a depth of 100 ft for a total floor area of 4,000 ft2, and
was subdivided into two 2,000 ft2 zones (of 40 by 50 ft). A small office was
selected because the U. S. population spends a substantial number of hours inside
offices, and Government statistics indicate that approximately half of the office
buildings in the U. S. are 5,000 ft2 and smaller.
The front and rear exterior walls are concrete block with exterior stucco and
interior insulation, with 46% glazing in the front and 20% in the rear. The side walls
are considered to be thermally neutral interior walls, adjoining neighboring offices.
The roof is assumed to be built-up roofing over insulated decking, and the floor is a
carpeted concrete slab.
Full occupancy is 27 persons (150 ft2/person). The occupancy varies
throughout the day on weekdays, between 6 am and 7 pm. The building is
unoccupied overnight (7 pm to 6 am), and all day on weekends and holidays.
The baseline HVAC system consists of two rooftop, constant-volume, packaged
single-zone (PSZ) units, one dedicated to each of the 2,000 ft2 zones. The units
included electric resistance heating; annual heating requirements are minimal in the
Miami climate. Ventilation rates of both 5 and 20 cfm/person were considered. The
cooling setpoint was 75 °F during occupied hours; the cooling was shut down
overnight and on weekends. The heating setpoint was 70 °F, set back to 55 °F during
off-hours. The cooling electric input ratio (EIR) was 0.341 Btu/h of electric input per
Btu/h of cooling output, considered to be representative of modern PSZ units.
Further details regarding the baseline building and HVAC system are presented
in Section 3 and in Appendix B.
2.3 THE IMPACT OF BUILDING AND HVAC PARAMETERS ON THE PENALTIES
ASSOCIATED WITH INCREASED VENTILATION
Table S-1 summarizes how each of the building and HVAC system parameters
impacts the computed cooling coil capacity, the annual HVAC energy cost, and the
percentage of occupied hours having an RH above 60%.
For ease in comparison, the impact of each parameter is presented in Table S-1
as the percentage change from the baseline building and baseline system operating
at a ventilation rate of 5 cfm/person. Under baseline conditions at 5 cfm/person, the
cooling capacity computed by the software is 103.6 kBtu/h (8.6 tons of refrigera-
2-2
-------
TABLE S-1
Effects of Building and HVAC Variables on
HVAC Capacity and Energy Cost, and on Occupied Hours Above 60% RH
Annual Occupied
Cooling HVAC Hours
Coil Energy with RH
Capacity Cost1 > 60%
OA Rate = 5 cfm/person
Baseline system with OA rate
of 5 cfm/person
103.6
kBtu/h
$2,510
40 hr/yr
OA Rate = 20 cfm/person
Baseline system with OA rate
of 20 cfm/person
Results below are expressed as the
percentage change from the baseline
numbers at 5 cfm/person, above
15.1
(incr. to
119.2 kBtu/h)
12.9
(incr. to
$2,835)
-25
(deer, to
29 hr/yr)
Building fLOADS) Variables
Effect of building orientation
(baseline building faces north)
- Building faces south
15.5
10.2
-25
Effect of building shading
(baseline has door, window overhangs)
- Delete all overhangs
Effect of occupant density
(baseline is 150 ft2/oerson)
- Reduce density to 300 ft2/person
- Increase to 100 ft2/person
21.0
-0.2
29.6
16.0
-1.5
26.9
-25
-31
-25
Effect of lighting/equipment power use
(baseline is 2.55 W/ft2)
- Reduce to 1.5 W/ft2
- Increase to 4.0 W/ft2
0.4
36.1
-5.1
38.7
+ 7
-48
(continued)
2-3
-------
TABLE S-1 (continued)
Percentage Increase Over
Baseline at 5 cfm/person
Cooling
Coil
Capacity
Annual
HVAC
Energy
Cost1
Occupied
Hours
with RH
> 60%
OA Rate = 20 cfm/person (continued)
Effect of infiltration rate
(baseline is 0.1 ACH)
- Decrease to 0 ACH 13.1 11.4 -25
Effect of exterior wall resistance
(baseline U. = 0.16 Btu/h ft2 F°i
- Decrease to 0 Btu/h ft2 F°
- Decrease to 0.6 Btu/h ft2 F°
Effect of amount of glazing
(baseline is 33% of exterior walls)
- Decrease to 0%
Effect of glass type
(baseline U„ = 0,94 Btu/h ft2 F°,
shading coefficient = 0.55)
- Improve to U0 = 0.32, S-C = 0.16
Effect of roof resistance
(baseline U.= 0.066 Btu/h ft2 F°)
- Reduce to U0 = 0
Effect of total insulation of office
(baseline has exterior walls, roof)
- Eliminate all exterior surfaces
(hypothetical)
HVAC (SYSTEMS) Variables
14.2 11.4 -25
14.8 12.3 -25
10.4 3.5 -43
13.3 6.1 -40
15.1 6.7 -34
7.6 -8.4 -60
Effect of thermostat set-up off-hours
(baseline shuts down off-hours)
- Cooling setpoint 81 °F off-hours 15.1 13.3 -25
(continued)
2-4
-------
TABLE S-1 (continued)
Percentage increase Over
Baseline at 5 cfm/person
Cooling
Coil
Capacity
Annual
HVAC
Energy
Cost1
Occupied
Hours
with RH
> 60%
OA Rate = 20 cfm/person (continued)
Effect of alternative HVAC systems
(baseline is 2 PSZ units/2 zones)
- 1 PSZ unit/1 zone 12.1 11.4 -30
- 1 PSZ unit/1 zone + 1 subzone 15.1 10.1 -33
- 1 PVAVS unit/2 zones 16.8 4.9 +135
- 2 PTAC units/2 zones 7.9 9.1 -100
Effect of ducted return air
(baseline is plenum return)
- Air return via ducts 15.1 12.4 -33
Effect of cold-air distribution
(baseline is PSZ/55°F min. supply T)
- PSZ/42°F minimum supply T 22.2 6.2 -72
- PVAVS/42°F minimum supply T 23.3 1.6 -52
Effect of economizer modifications
(baseline is T-controlled econo.)
- No economizer 15.1 13.7 -85
- Enthalpy-controlled economizer 15.1 13.0 -55
Effect of cooling electric input ratio
(baseline is EIR = 0.341)
- Cooling EIR = 0.284 15.1 -1.8 -25
- Cooling EIR = 0.427 15.1 105.5 -25
Effect of cooling capacity and SHR
(baseline is 8.6 tons/SHR = 0.75)
- 10 tons/SHR = 0.78 15.8 13.8 -25
- 10 tons/SHR = 0.73 15.8 13.7 -25
- 11 tons/SHR = 0.78 27.4 15.4 -25
- 11 tons/SHR=0.73 27.4 15.6 -25
(continued)
2-5
-------
TABLE S-1 (concluded}
Percentage Increase Over
Baseline at 5 cfm/oerson
Cooling
Coil
Capacity
Annual
HVAC
Energy
Cost1
Occupied
Hours
with RH
> 60%
OA Rate = 20 cfm/person (continued)
Weather File Variables
Effect of alternative weather files
(baseline-Typical Meteorological Year)
- Weather Year for Energy Calcs, 30.8 12,2 -8
Note:
1 Energy costs include electricity for: the air-conditioning compressor and
condenser fan; the electric resistance heating coils; the motor for the central
air handling fan; and auxiliaries (compressor crankcase heaters). Cost of
electricity is $0.0473/kWh plus a demand charge of $9.96/kW above 10 kW.
2-6
-------
tion), the annual HVAC energy cost is $2,510, and the percentage of occupied hours
above 60% RH is 40 hours per year (1.2% of 3,276 occupied hours), as shown by
the first entry in the table.
All of the other entries are for operation at 20 cfm/person.
The second entry in the table shows the predicted impacts when the baseline
building and system are simply operated at 20 cfm/person, without any other
variations in the building and HVAC variables. For example, this entry shows that
operation at the increased ventilation rate increases HVAC energy cost by 12.9% (an
increase of $325, to $2,835 per year). Use of the HVAC energy costs in this table
is intended to emphasize the impact on the HVAC system. If one instead used the
total building energy costs - which are $4,273 per year, including lighting and
equipment, at 5 cfm/person -- the $325 increment caused by the OA increase would
correspond to only a 5.4% increase.
The second entry in the table also shows that the OA increase in the baseline
system is computed to decrease the percentage of elevated-RH occupied hours by
25%, from 40 to 29 hours per year. (The explanation for this effect is discussed
later.)
The remainder of the entries show the predicted impacts (at 20 cfm/person) as
each of the building and HVAC parameters is systematically varied from its baseline
value. The percentage change with each parameter should be compared with the
corresponding percentage increase experienced by the baseline system at 20 cfm/
person, discussed in the preceding two paragraphs. If, for example, the percentage
increase in annual HVAC energy cost becomes less than 12.9% when a given
parameter is varied, this parametric variation is predicted to consume less HVAC
energy at 20 cfm/person than would the baseline at 20 cfm/person. In concept, the
HVAC energy penalty associated with increasing the baseline from 5 to 20 cfm/person
could be correspondingly reduced if the OA increase could practically be accompanied
by this variation in this parameter.
For some parameters, the percentages become negative. This means that a
building or HVAC system incorporating that parametric variation could operate at 20
cfm/person at a savings compared to the baseline at 5 cfm/person.
Section 5 of this report presents a detailed analysis of why each of these
parameters is predicted to have the result that it does.
Table S-2 -- presented in the same format as Table S-1 -- lists those entries
from Table S-1 that are predicted to offer the greatest potential reductions in HVAC
energy cost at 20 cfm/person, compared to the baseline at 5 cfm/person. These
entries are listed in descending order, with the parametric variation offering the
greatest reduction listed first.
2-7
-------
TABLE S 2
Building and HVAC Variables Creating the Greatest Reductions
in Annual HVAC Energy Cost at 20 cfm/person (from Table S-1)
Annual Occupied
Cooling HVAC Hours
Coil Energy with RH
Capacity Cost > 60%
OA Rate = 5 cfm/person
Baseline system with OA rate 103.6
of 5 cfm/person kBtu/h $2,510 40 hr/yr
OA Rate = 20 cfm/person Results below are expressed as the
percentage change from the baseline
numbers at 5 cfm/person, above
Baseline system with OA rate
of 20 cfm/person 15.1 12.9 -25
Variables giving the greatest reduction in HVAC energy cost at 20 cfm/person. in
descending order
Eliminate all exterior surfaces (ideal) 7.6 -8,4 -60
Reduce lighting/equipment to 1.5 W/ft7 0.4 -5.1 7
Reduce cooling electric input ratio to 0.284 15.1 -1.8 -25
Reduce occupant density to 300 ft2/pcrson -0.2 -1.5 -31
Convert to PVAVS with cold-air distribution
(minimum supply air T = 42 °F) 23.3 1.6 -52
Decrease glazing to 0% of wall area 10.4 3.5 -43
Convert from 2 PSZ units to 1 PVAVS unit -
standard minimum supply air T (55 °F) 16.8 4.9 135
improve glass type to UQ = 0.32, S-C = 0.16 13.3 6.1 -40
Convert PSZ to cold-air distribution (42°F) 22.2 6.2 -72
Increase roof resistance to U0 = 0 15.1 6.7 -34
2-8
-------
Similarly, Table S-3 lists those entries from Table S-1 that are predicted to offer
the greatest potential reductions in hours at elevated RH at 20 cfm/person, compared
to the baseline at 5 cfm/person. Again, the entries are listed in descending order.
2.3.1 Parameters Creating the Greatest Reductions in HVAC Energy
Six of the ten parameters listed in Table S-2 are parameters associated with the
building: elimination of all exterior surfaces (a hypothetical consideration); reduced
lighting/equipment wattage; reduced occupant density; decreased glazing; improved
glass type; and increased roof insulation.
That each of these parameters would significant.y reduce annual HVAC energy
cost, of course, is not surprising. However, it is instructive to explore why these
parameters fall in the order they do in Table S-2.
Table S-4 summarizes the contribution of each of the individual heat sources
to the annual HVAC energy consumption as predicted by the D0E-2.1E model, at
ventilation rates of both 5 and 20 cfm/person.
As shown, lighting and equipment are the largest individual contributors to the
HVAC load, contributing about half of the total load from all sources. Thus, it is not
surprising that a 40% reduction in lighting plus equipment wattage (from 2.55 to
1.5 W/ft2) would provide the greatest reduction in HVAC energy costs among the
practical alternatives in Table S-2. (Only the hypothetical scenario of eliminating all
exterior surfaces provided a greater reduction.) This reduction in lighting plus
equipment wattage could be achieved by converting from the prescriptive or average
wattages in ASHRAE 90.1-1989 (ASHRAE, 1989b) to very efficient lighting (e.g.,
including daylighting) and more efficient (or more limited) equipment usage.
As shown in Table S-2, the HVAC energy cost savings from more efficient
lighting/equipment would more than offset the increase in HVAC energy costs
resulting from an increase in OA from 5 to 20 cfm/person. The building with efficient
lighting/equipment could operate at 20 cfm/person with a HVAC cost savings of 5.1 %
compared to the baseline 5 cfm/person case. Of course, efficient lighting/equipment
would provide even greater savings in total building energy costs, by reducing the
energy costs for lighting and equipment as well as for the HVAC system.
As shown in Table S-4, occupants are tied with glazing and (at 20 cfm/person)
with OA as the second largest contributor to HVAC energy consumption. According-
ly, it is not surprising that cutting occupancy in half (from 150 to 300 ft2/person)
would provide the next greatest reduction among the 6 building parameters in Table
S-2. Of course, reducing occupant density will not generally be a viable option for
reducing energy costs.
2-9
-------
TABLE S-3
Building and HVAC Variables Creating the Greatest Reductions
in Hours Above 60% RH at 20 cfm/person (from Table S-1)
Annual Occupied
Cooling HVAC Hours
Coil Energy with RH
Capacity Cost1 > 60%
OA Rate = 5 cfm/person
Baseline system with OA rate 103.6
of 5 cfm/person kBtu/h $2,510 40 hr/yr
OA Rate = 20 cfm/person Results below are expressed as the
percentage change from the baseline
numbers at 5 cfm/person, above
Baseline system with OA rate
of 20 cfm/person 15.1 12.9 -25
descendina order
Eliminate economizer
15.1
13.7
-85
Convert PSZ unit to cold-air distribution
(minimum supply air T = 42 aF)
22.2
6.2
-72
Eliminate ail exterior surfaces (ideal)
7.6
-8.4
-60
Convert to enthalpy-controlled economizer
15,1
13.0
-55
Convert to PVAVS with cold-air distribution
23.3
1.6
-52
Decrease glazing to 0% of wall area
10.4
3.5
-43
Improve glass type to UQ •« 0.32, S-C-0.16
13.3
6.1
-40
Increase roof resistance to UQ = 0
15.1
6.7
-34
Note: The DOE-2.1 E model used here does not account for moisture capacitance of the
building materials/furnishings, or for re-evaporation of moisture from the cooling coils
when the coils cycle off with the air handler operating. As a result, the number of
hours computed to have RH > 60% at any given set of conditions will usually be low.
2-10
-------
TABLE S-4
Approximate Contribution of the Various Heat Sources
to the Annual HVAC Energy Consumption in the Baseline Building
Percentage Contribution to
Annual HVAC Energy Consumption1
Heat Source OA = 5 cfm/person OA = 20 cfm/person
Conduction and Radiation Through Exterior Surfaces (sensible)
Exterior walls - conduction 2 2
Glazing - conduction and radiation 14 12
Door - conduction 0.5 0.4
Roof - conduction 8 6
Slab - conduction -0.5 -0.4
Infiltration
(sensible and latent) 2 2
Mechanically Introduced Outdoor Air
(sensible and latent) 2 15
Internal Sources
Occupants (sensible and latent) 15 13
Lighting (sensible) 40 35
Equipment (sensible) 17 15
Domestic hot water heater (sensible) ~0 ~0
TOTAL 100 100
Note:
1 Annual HVAC energy consumption for the baseline building is 26,145 kWh/year
for a ventilation rate of 5 cfm/person, and 29,390 kWh/year for 20 cfm/person.
2-11
-------
Table S-4 shows that -- among the exterior surfaces - conduction and radiation
through the glazing are the most important contributors to HVAC energy consumption
in this office. As a result, it is not surprising that adjustments to the glazing --
eliminating it altogether, or increasing its resistance to conduction and radiation -
should show up on Table S-2 as the next most effective building parameters for
reducing HVAC energy costs. Eliminating (or substantially reducing) the glazing might
not often be a viable option. However, improving the glazing resistance is a viable
option, if the building owner is prepared to accept the increased construction costs.
The results shown in Table S-2 were computed when the baseline case - single-pane
glass with a moderate combination of tinting, coatings, or shading, having an overall
heat transfer coefficient (U„) of 0.94 Btu/h ft2 F°, and a shading coefficient (S-C) of
0.55 — was converted to double pane glass with tinting and a highly reflective
coating, having U0 = 0.32 Btu/h ft2 F° and S-C = 0.16.
Finally, Table S-4 shows that -- among the exterior surfaces - roof conduction
is the second most important contributor, about half as important as glass conduction
and radiation. The roof is important because it represents such a large exterior
surface area for this building (4,000 ft2, compared to only 700 ft2 for the unglazed
portion of the exterior walls), and it has the most consistent direct exposure to solar
radiation. Consequently, it is not surprising that hypothetically increasing the roof
resistance to infinity (i.e., reducing the roof UQ from 0.066 Btu/h ft2 F° to zero) is the
exterior surface parameter that provides the next greatest reduction in HVAC energy
cost in Table S-2 (cutting the cost penalty from the OA increase about in half, from
12.9% to 6.7%).
Of course, reducing the roof U0 all the way to zero is not practical. However,
these results show that -- if additional resources are going to be expended to better
insulate the shell of this particular office configuration -- one is better served directing
those resources towards improved glazing and increased roof resistance, rather than
towards increased wall or slab resistance.
The other four of the ten parameters listed in Table S-2 are parameters
associated with the HVAC system: improving the cooling system efficiency;
converting from a constant-volume PSZ system to a packaged variable-air-volume
system (PVAVS); and conversion to cold-air distribution (i.e., a minimum supply air
temperature of 42 °F rather than 55 °F), with either the PSZ system or the PVAVS.
Of these four, the parameter providing the greatest reduction in HVAC energy
cost is improved efficiency of the PSZ cooling coils. In this calculation, the EIR was
decreased from the baseline value of 0,341 - corresponding to an energy efficiency
ratio (EER) of 10 Btu/h per W), representing a typical efficiency -- to an EIR of 0.284
(EER = 12 Btu/h per W), representing a high-efficiency unit. If the building owner
were prepared to invest in high-efficiency cooling units, this office could operate at
20 cfm/person while simultaneously saving 1.8% of the HVAC energy cost compared
to operation at 5 cfm/person with the baseline, moderate-efficiency system. This
1.8% savings corresponds to a modest $46/year.
2-12
-------
As shown in Table S-2, conversion of the pair of PSZ units to a single two-zone
PVAVS (operating at the standard minimum supply air temperature of 55 °F) would
reduce by 60% the HVAC energy cost penalty associated with the OA increase. That
is, the penalty would drop from 12.9% to 4.9%. PVAVS can be slightly more
complicated and more expensive than the PSZ units, and hence do not appear to be
as widely used in strip mall space of the type being modeled here. However, PVAVS
of this capacity are commercially available, and can reasonably be considered as a
means to reduce the energy penalty in this application.
PVAVS reduce energy consumption and cost by reducing the volume of supply
air being delivered. Most of the savings result from reduced power consumption by
the central air handling fan, since power consumption varies with the cube of the
volumetric flow rate. A small portion of the savings results from reduced cooling coil
consumption, since reduced central fan operation results in less heat being added to
the circulating air stream by the fan motor.
Finally, Table S-2 shows that conversion to cold-air distribution (with either the
PVAVS or the PSZ system) will provide a significant reduction in the HVAC energy
penalty associated with the OA increase. Operation at a minimum supply air tempera-
ture of 42 °F instead of 55 °F reduces volumetric flow rates, thus reducing fan power
consumption as well as the amount of heat added to the air stream by the fan motor.
Superimposing cold-air distribution and a PVAVS — for which volumetric flows are
already significantly reduced - provides the greater reduction in HVAC energy costs,
among the two HVAC types.
The use of cold-air distribution creates a number of design and operating
complications that could make such an approach impractical for small offices such as
the one modeled here, where simplicity in maintenance is important. Among these
complications are the need for: a) increased care to reduce the risk of moisture
condensation on the ductwork and the diffusers; and b) possible powered terminals
to provide adequate throw of the reduced volume of air out through the diffusers (a
step which would offset part of the energy savings achieved through the reduction
in volumetric flow).
2.3.2 Parameters Creating the Greatest Reductions in Hours at Elevated RH
According to the DOE-2 model, occupied hours having RH levels greater than
60% occur on cool mornings in Miami. During the first hours after system startup on
cool mornings, the outdoor RH can be high (over 90%), but the indoor and outdoor
temperatures can be sufficiently low that the cooling coils operate at greatly reduced
capacity (or remain off altogether). As soon as the cooling coils begin operating at
a significant fraction of their capacity - usually within 2 or 3 hours after startup - the
indoor RH drops below 60%. (On warm summer mornings, the coils begin operating
near full capacity immediately upon startup; thus, elevated-RH indoor hours never
occur during warm weather, despite the high outdoor RH levels that exist.)
2-13
-------
On some cool morning hours, when the economizer is able to provide all of the
sensible cooling required by the space, the economizer will activate in lieu of coil
operation. (The economizer and the cooling coils cannot operate simultaneously in the
PSZ units.) During economizer operation - when a large amount of untreated,
potentially high-moisture-content outdoor air can be introduced into the building --
there is an increased potential for indoor RH levels to exceed 60%. In practice, the
economizer on the HVAC systems being modeled here does not operate often.
As shown in Table S-3, simply increasing the OA rate from 5 to 20 cfm/person
using the baseline system (with no changes in any other variables) is predicted by
DOE-2 to reduce the number of elevated-RH occupied hours by 25%. Although this
percentage may seem significant, the actual number of hours involved is small,
corresponding to a reduction from 40 hours per year at 5 cfm/person (1.2% of all
occupied hours) to 29 hours at 20 cfm/person (0.9%).
The decrease in the number of elevated-RH hours occurs because, on average,
the increased OA rate increases the sensible load. In addressing this increased load,
the PSZ coils to operate at a lower temperature during the cool morning periods when
elevated-RH hours occur. This increases the latent cooling provided by the system.
This increase in latent cooiing at 20 cfm/person is predicted by DOE-2 to more than
offset the increase in latent load caused by the increased OA rate.
Other researchers have made similar calculations using a model that includes
factors not addressed by DOE-2, namely, moisture capacitance and re-evaporation off
the cooling coils (Shirey and Rengarajan, 1996). These analysts predict that - in
contrast to the DOE-2 predictions - an increase in OA rate in Miami would signifi-
cantly increase, not decrease, hours at elevated RH. Also, when capacitance and re-
evaporation are considered, it is predicted that some of the elevated-RH hours will
occur during warm weather, not just on cool mornings.
Table S-3 lists the eight parametric variations predicted by DOE-2 to provide the
greatest reductions in the number of elevated-RH occupied hours.
Two of the most effective of these eight variations involve adjustments to the
economizer. This is not surprising, since - in the Miami climate, as discussed above -
the economizer is likely to cause elevated indoor RH during those hours when it
operates.
When the economizer is eliminated altogether (and the system is operating at
20 cfm/person), as shown in the table, occupied hours above 60% RH are reduced
by 85% compared to the baseline 5 cfm/person case (from 40 to 6 hours/year). This
result confirms that, in this humid climate, the bulk of the elevated RH hours are
caused by the economizer.
2-14
-------
If the economizer were converted to enthalpy control, rather than standard
temperature control, hours above 60% RH are reduced by 55% (from 40 to 18
hours/year). Economizer enthalpy control prevents the economizer from operating if
the outdoor enthalpy is greater than the indoor enthalpy {even if the outdoor
temperature is lower). But this controller does not make any effort to control the
indoor humidity. Thus, if the outdoor enthalpy were lower, the controller would allow
the economizer to operate - and hence allow the cooling coils to shut down - even
if this meant that indoor RH values would exceed 60%. Thus, enthalpy control would
eliminate only some of the economizer-induced elevated-RH hours.
Elimination of the economizer altogether is predicted by DOE-2 to almost
eliminate occupied hours above 60% RH in this climate, and it does so with only a
modest energy cost penalty. The annual HVAC energy cost for the no-economizer
case at 20 cfm/person is only $23/year greater than that for the baseline temperature-
controlled-economizer case at 20 cfm/person. Thus, this is a viable option to consider
for reducing indoor RH. By comparison, the option of economizer enthalpy control is
less attractive, since the cost and maintenance requirements make such controllers
less desirable for small office applications, and since enthalpy control is less effective
in reducing elevated-RH hours.
Two of the other parametric variations in Table S-3, offering significant
reductions in the number of elevated-RH hours, involve conversion to cold-air
distribution. These include conversion of the baseline PSZ units to cold-airdistribution
(providing a 72% reduction, from 40 to 11 hours), and conversion to a PVAVS with
cold-air distribution (providing a 52% reduction, from 40 to 19 hours). This occurs
largely because - at the very low coil temperatures in cold-air systems - the amount
of latent cooling increases significantly relative to the standard (55 °F supply air
temperature) case. Thus — after the coils activate on cool mornings, when the
elevated-RH hours occur - RH levels in the office space drop more rapidly with the
cold-air system.
However, due to the operating complications and likely increased maintenance
of cold-air systems, it is not likely that this approach would often be considered for
use in a small strip mall office such as the one modeled here.
The remaining four parameters listed in Table S 3 involve efforts to make the
building shell more heat resistant: hypothetical total isolation of the space; elimination
of the glazing; improving the glazing; and increasing the roof resistance. These four
parameters appear in Table S-3 in the same order that they appeared in Table S-2.
The reason why these parameters have this effect on the number of elevated-
RH hours is that -- the better insulated the building - the less it cools off over cool
winter nights and weekends. Consequently, the cooling coils see a greater cooling
load more quickiy after startup on the cool mornings, when the elevated RH occupied
2-15
-------
hours occur. The temperature-activated coils come on earlier after startup, and
provide greater total (and hence latent} cooling during these morning hours, thus
reducing the number of elevated-RH hours. The more effective the shell insulating
step, the better the building retains its heat overnight, and the greater the resulting
latent heat removal in the morning. For this reason, the insulation steps that provide
the greatest reduction in total HVAC energy cost (Table S-2) also provide the largest
reduction in elevated-RH hours (Table S-3).
These results show that resources devoted toward improved glazing and
increased roof resistance will have the greatest impact, not only on reducing HVAC
energy cost, but also in reducing (modestly) the number of hours at elevated RH.
It is interesting to note that Table S-3 does not include any of the parameters
that involve latent heat entry into, or generation inside, the building. Reducing
occupant density to 300 ft2/person reduces the number of elevated-RH hours by 31 %,
just below the cut-off used in preparing the table. Reducing outdoor air infiltration
from 0.1 air change per hour to zero has essentially no impact on the number of
elevated-RH hours.
2.4 THE IMPACT OF STEPS TO REDUCE INDOOR HUMIDITY
As indicated previously, the DOE-2 model does not incorporate the moisture
capacitance of building materials and furnishings, nor moisture re-evaporation off the
cooling coils when the coils cycle off with the air handler operating. As a result -
unlike a model that does include these phenomena (Shirey and Rengarajan, 1996) --
DOE-2 does not predict an increase in elevated-RH occupied hours when the OA rate
is increased. On this basis, DOE-2 might not be expected to precisely simulate the
actual energy and performance impacts that would result when steps are taken to
reduce the number of hours at elevated RH.
Despite this shortcoming, it is still felt that a DOE-2 analysis can provide useful
perspective regarding the possible magnitude of the effects of steps to reduce RH.
For example, the conclusion in the preceding section — that elimination of the
economizer would substantially reduce the number of occupied hours above 60%
RH -- is felt to be valid, despite the fact that the absolute number of computed
elevated-RH hours might be low.
A variety of steps can be taken to reduce the number of hours at elevated
indoor RH in warm, humid climates. These steps fall into two categories.
a) Utilize an HVAC control system that relies solely on temperature control, as is
typical for office space. But design and operate the HVAC system such that --
2-16
-------
as the system operates to control temperature in the space - there will be as
few hours as possible having RH levels above 60%.
b) Incorporate humidity control as well as temperature control into the HVAC
control system, which is not common for an office of this type. The humidity
control could be achieved, e.g., using over-cooling with reheat, or using
desiccants.
The RH results presented in Tables S 1 and S-3 can be viewed as an assess-
ment of a wide range of building and HVAC parameters that might serve as steps that
would fall into Category a) above. As discussed in Section 2.3.2, the most practical
conclusion apparent from Table S-3 is that occupied hours at elevated RH can be
substantially reduced at 20 cfm/person if the economizer is deleted in warm, humid
climates.
Two additional RH reduction steps are considered in further detail here. One --
which falls into Category a) above - involves setting the thermostat temperature up
to 81 °F during off hours (overnight, weekends, and holidays) rather than turning the
system off altogether during cooling periods. The second -- which falls into Category
b) -- involves using a humidity controller on the system, over-cooling and reheating the
supply air as necessary. The humidity control approach was considered in order to
assess the energy penalty associated with this procedure, recognizing that humidity
control is not commonly used in small offices, and that reheat is generally prohibited
by Florida code (FDCA, 1993).
2.4.1 Thermostat Set-Up vs. System Shut-Down
According to the DOE-2 simulation, setting the thermostat up to 81 °F during
off-hours, rather than shutting the system off, will have no impact on the number of
occupied hours above 60% RH. This result occurs because elevated-RH occupied
hours are predicted by DOE-2 to occur during the first hours after startup on cool
mornings. During such cool weather, the overnight temperatures will not have been
sufficiently high to cause the overnight office temperature to exceed 81 °F. Thus,
even if the thermostat is set up rather than the system being turned off, the cooling
coils will not activate overnight. No latent cooling will be provided overnight, and, as
a result, the latent load encountered by the system upon startup in the morning will
remain unchanged. Accordingly, the number of elevated-RH occupied hours will
remain unchanged.
This predicted result could be different if the DOE-2 model had addressed
moisture capacitance and coil re-evaporation. In that case, some elevated-RH
occupied hours occur during hot weather, when temperatures sometimes can be
sufficient to activate the coils overnight, This would reduce the latent load seen by
the system upon startup, by reducing the re-evaporated moisture that remains in the
2-17
-------
air overnight, and by reducing the amount of sorbed moisture, it could thus reduce
the number of elevated-RH occupied hours occurring during warm weather.
With or without consideration of capacitance and re-evaporation, switching to
thermostat set-up rather than system shut-down will reduce the number of elevated-
RH unoccupied hours during warm weather. On some hot nights and weekends in
Miami, with the system off, the indoor RH can be above 60% much of the time, due
to infiltration alone (even in the absence of re-evaporation effects). With thermostat
set-up, when the off-hour office temperature exceeds 81 °F and the coils come on,
the RH drops below 60%, at least for the hours when the coils are activated. On
some warm days, this can represent 15% to 20% of the unoccupied hours that would
otherwise be at elevated RH. The same reduction in elevated unoccupied hours will
be achieved regardless of the OA rate during occupied hours, since OA ventilation is
not provided during unoccupied hours.
Thermostat set-up will not impact the elevated-RH unoccupied hours that occur
during cool weather in Miami, since the office temperatures will generally not exceed
the 81 °F level that would trigger coil operation.
Of course, reducing the number of elevated-RH unoccupied hours will not
improve occupant comfort, since no one will be in the building. But it will reduce the
risk of biological growth. Since most of the elevated-RH hours in this building occur
during unoccupied hours - regardless of whether the OA rate during occupied hours
is 5 or 20 cfm/person - switching from system shut-down to thermostat set-up would
appear to be an important step for any building operator concerned about micro-
biologicals.
The operating cost associated with set-up vs. shut-down is low, according to
the DOE-2 predictions. As shown in Table S-1, switching to 81 °F thermostat set-up
at 20 cfm/person would increase the annual HVAC energy cost by only 0.4%
(amounting to only $10 per year) compared to the baseline shut-down case at 20
cfm/person. Selecting an even lower set-up temperature of 79 °F - which would
reduce the number of unoccupied elevated-RH hours by an even greater amount -
would increase annual HVAC energy costs by only $38.
The detailed analysis of thermostat set-up is presented in Section 6.2.2.
2.4.2 Humidity Control by Overcooling and Reheat
Overcooling the supply air to condense moisture, then reheating to achieve the
proper supply air temperature, has historically been a method for controlling humidity.
Although humidity control is not commonly utilized in small offices (except in special
cases), and although Florida codes now generally prohibit reheat due to the energy
penalty involved, it is of interest to assess the costs and effectiveness of this
approach, in comparison with the other approaches considered here.
2-18
-------
It is re-emphasized that - since DOE-2 does not include the moisture
capacitance and re-evaporation phenomena - DOE-2 underestimates the number of
occupied hours where the RH exceeds 60%. Correspondingly, the computations here
will necessarily underestimate the energy consumption and costs for a system that
is designed to control these elevated-RH hours.
The results of this analysis of a reheat-based humidity control system are
summarized in Table S-5, presented in the same format as Table S-1.
As shown in the table, the humidity control system operating at 20 cfm/person
is predicted by DOE-2 to increase the HVAC energy cost by 16.5% compared to the
baseline (temperature-control) system at 5 cfm/person, and by 3.6% (i.e., 16.5 vs.
12,9%) compared to the baseline system at 20 cfm/person. But the humidity control
system does achieve its objective, of eliminating aii occupied hours above 60% RH.
Comparing Tables S-3 and S-5, it is apparent that - among the parametric
variations predicted to provide the greatest reductions in elevated-RH hours --
conversion to a humidity controller involves the largest increase in HVAC energy cost
(16.5%), but provides the greatest reduction in hours above 60% RH (100%).
Second to the humidity controller in both these categories - at least as estimated by
DOE-2 - is elimination of the economizer (13.7% increase in HVAC energy cost, 85%
reduction in elevated-RH hours).
This comparison suggests that elimination of the economizer would eliminate
85% of the elevated-RH occupied hours at 20 cfm/person, at an energy cost increase
of S19/year (compared to the baseline case at 20 cfm/person). To eliminate the
remaining 15% of the elevated-RH occupied hours, one could convert to a humidity
controller, at an energy cost increase of $90/year (compared to the baseline at 20
cfm/person). Conversion to a humidity controller automatically prevents economizer
operation during those hours when the economizer is responsible for the elevated RH,
and provides the additional cooling/reheat required to address the remaining elevated-
RH hours.
This DOE-2 comparison would change if one included moisture capacitance and
coil re-evaporation in the model. In that more rigorous case, the effectiveness of
economizer elimination at 20 cfm/person would decrease - i.e., the percentage
reduction in elevated-RH occupied hours would be much less than 85% -- since the
new model would show a much greater number of elevated-RH hours being caused
by factors other than economizer operation. Conversion to a humidity controller
would remain 100% effective, but the energy cost would increase, since, again, the
new model would show many more elevated-RH occupied hours.
Most of the energy penalty incurred by the humidity-controlled system results
from additional sensible and latent cooling on cool, humid days, when the elevated-RH
2-19
-------
TABLE S-5
Effect of Humidity Control by Overcooling and Reheat
Cooling
Coil
Capacity
Annual
HVAC
Energy
Cost
Occupied
Hours
with RH
> 60%
OA Rate = 5 cfm/person
Baseline system {temperature
control only) with OA rate
of 5 cfm/person
103.6
kBtu/h
$2,510
40 hr/yr
OA Rate = 20 cfm/person
Baseline system (temperature
control only) with OA rate
of 20 cfm/person
Results below are expressed as the
percentage change from the baseline
numbers at 5 cfm/person, above
15.1
12.9
-25
Humidity control system (tempera-
ture plus RH control! with
OA rate of 20 cfm/person
15.1
16.5
-100
2-20
-------
occupied hours occur according to the DOE-2 model, As would be expected, the
penalty is relatively small during mild and hot weather. And the contribution of reheat
to the total penalty is small, on the order of 10% of the total; the increased sensible
and latent cooling is responsible for the remainder.
As indicated above, preventing economizer operation during elevated-RH hours
reduces the number of elevated hours by 85% at an energy cost of $19/year.
Considering that the humidity controller does prevent economizer operation under
these conditions, it seems surprising that conversion to humidity control raises the
energy cost penalty to $90/year simply to address the remaining 15% (only 6 hours/
year). The reason is that, during many cool-weather hours, the humidity-controlled
system — as modeled by DOE 2 -- cools the office space down toward the heating set-
point (70 °F) instead of the cooling set-point (75 °F), at a significant energy penalty.
This seems to occur because, during cool weather, the moisture content of the office
air (lb moisture per lb dry air) can sometimes be so close to 60% RH that the office
temperature can determine whether the office is above or below the 60% set-point.
The humidity controller tends to operate the offices at a lower temperature - where
a given moisture content would result in a higher RH - perhaps as the result of
occasional supply air over-cooling required to prevent elevated-RH hours. This seems
to establish a cycle, whereby the simulated system has to continue to over-cool the
supply air in order to maintain 60% RH in the cooler offices.
It is emphasized that humidity control will maintain the RH in the offices only
during occupied hours, when the HVAC system is operating. Regardless of which
simulation model is used, a large fraction of the total elevated-RH hours in the space
occur during unoccupied hours. Thus, if biological growth is a concern, some off-hour
operation would be required even if a humidity control system were used to eliminate
all elevated-RH hours during occupied periods. This is true regardless of the OA rate
during occupied hours.
A detailed discussion of the humidity controller analysis is presented in Sec.
6.3.
2-21
-------
[This page intentionally blank.]
2-22
-------
SECTION 3
nCCrDIDTIAM AC tljc cti inv
UCwLrnli I IvJIM Ui I iitl O I UUT
3.1 THE SOFTWARE
The computer program used for this analysis was DOE-2.1 E, the current version
of the DOE-2 building energy simulation software, which evolved from efforts begun
by the predecessor of the U. S. Department of Energy, and by the State of California,
in the mid-1970's (York et al., 1981; U. S. Department of Energy, 1994a; Ayres and
Stamper, 1995). The DOE-2 program is well documented and supported, and is
among the most widely used software in the country for simulating building energy
consumption and energy costs.
In the DOE-2 program, the building and HVAC system to be modeled are
described by the user in an input file written in the Fortran-based "Building Description
Language". A weather file describing the key climatic conditions of the location of
the building is selected from one of three sources.
Appendix A presents an example of a DOE-2.1 E input file used for the analysis
presented in this report. This particular input file is for the "baseline" set of office and
HVAC conditions, as discussed later.
The DOE-2 energy simulation involves four primary steps.
1) The LOADS calculation, a meticulous, hourly heat transfer and heat balance
calculation which determines the loads in each space within the defined
building. This calculation considers external loads (created by ambient
temperature, solar angle, etc.), internal loads (resulting from occupants,
lighting, equipment, etc.), and infiltration. The heat transfer resistance and
heat capacitance of the building shell, and the characteristics of the glazing, for
example, have significant impacts on the external loads.
2) The SYSTEMS calculation, which computes the hourly performance of the
"secondary" HVAC equipment (the coils, air handler, and air distribution
system) in response to the calculated loads, and the conditions (temperature,
RH) which this equipment is able to maintain in the defined zones within the
building. For an all-air HVAC system, for example, a central element of the
SYSTEMS calculation is the computation of what mass of supply air that must
be supplied at what temperature to each zone in order to maintain the user-
specified temperature setpoint in that zone - a computation which will be
3-1
-------
impacted by the control method used by the particular HVAC system being
modeled (modulation of supply air volume, modulation of supply temperature,
or both). DOE-2 allows the user to select from among about two dozen
different HVAC system types for modeling; for a given HVAC system type,
users may allow the program to use default system design values (e.g., for coil
efficiencies and for air handler performance curves), or may specify their own
design values.
3} The PLANT calculation, which computes the hourly performance of the
"primary" HVAC equipment (e.g., boilers and chillers) required to provide the
coil temperatures, heat extraction rates, etc., determined under SYSTEMS. The
PLANT is responsible for providing the total actual energy requirements of the
building. This includes not only the fuel and electricity requirements for running
the boilers and chillers, but also, e.g., any purchased electricity computed under
SYSTEMS for operating the air handler, and calculated under LOADS for
powering the building lighting and internal equipment.
4) The ECONOMICS calculation, where the total fuel and electricity requirements
tallied under PLANT are converted to costs, based upon specified energy cost
rates.
3.2 THE BASELINE BUILDING
A small office was the building type selected for this analysis.
Office space was selected because the U. S. population spends a substantial
number of person-hours inside offices. Only residential structures would seem to have
a clearly greater number of person-hours of occupancy.
Government statistics (U. S. Department of Commerce, 1994; U. S. Depart-
ment of Energy, 1994b) indicate that approximately half of the office buildings in the
U. S. are 5,000 ft2 and smaller; almost 90% of the office buildings are 25,000 ft2 and
smaller. Expressed in terms of square footage of office space, something less than
10% of the total square footage in the U. S. is in offices of 5,000 ft2 and less, and
about one-third in offices of 25,000 ft2 and less. On this basis, it was decided to
focus this study on small offices, less than 25,000 ft2. In addition to the population
of small offices, another reason for selecting small offices is that they typically utilize
packaged, direct expansion HVAC systems which offer the challenge of somewhat
reduced flexibility for humidity control, compared to the built-up systems common in
large office buildings. The specific office size selected for this study — 4,000 ft2,
toward the lower end of the size range — was selected for simplicity, and to be
consistent with Shirey and Rengarajan.
3-2
-------
The baseline office defined for this simulation is depicted in Figures 1 and 2,
and further described in Table 1. The parametric analysis was conducted by
systematically varying these baseline conditions, as discussed later. To model a hot,
humid climate, this office is located in Miami, Florida. The office has 4,000 ft2 of
total space, and is subdivided into two 2,000 ft2 zones (one zone in the front, and one
in the rear). This office is located in the middle of a strip shopping mall, such as an
office occupied by an insurance or travel agency. As a result, only the front and rear
walls are exterior walls; the two side walls adjoin conditioned space occupied by
neighboring tenants, and are assumed to be thermally neutral.
The baseline values summarized in Table 1 for the building design and operating
parameters are presented more completely in the LOADS section of Appendix B.
Appendix B also presents the rationale for the selection of these baseline values, and,
as discussed later, indicates the alternative values that are considered for each of
these parameters during this parametric analysis. The baseline values are incorporated
into the LOADS portion of the baseline input file presented in Appendix A.
The baseline values for the building variables are believed to represent typical
values for space in strip malls, based upon inspections of a number of malls. This
office is similar to the one considered by Shirey and Rengarajan, although some of the
baseline values have been changed from their values based upon the rationales
presented in Appendix B, and in an effort to more precisely describe the building.
3.3 THE BASELINE HVAC SYSTEM
The baseline HVAC system is summarized in Table 2, with further definition in
the SYSTEMS sections of Appendices A and B. The baseline system consists of two
packaged single-zone (PSZ) rooftop units, one for each of the two zones in the office.
These constant-volume units provide direct expansion cooling, and have electric
resistance heating.
PSZ units -- which can be either unitary (rooftop or outside-the-wall) systems,
as in this case, or split systems - are a common choice for small offices. They can
be among the simplest and least expensive of the central systems (Birdsall, 1995).
Although other types of systems (such as packaged VAV systems) can also be
considered, it is not uncommon for small offices, much larger than the one being
considered here, to be conditioned by multiple PSZ units, one for each of the
building's control zones. The use of two PSZ units in the baseline building here -
which results in each unit having approximately 5 tons of refrigeration capacity ~ is
consistent with the configuration used by Shirey and Rengarajan, and represents a
commonly utilized capacity for this type of unit. From a practical standpoint, the use
of two units will also improve the comfort in the offices; the north side of the building
will have a much lower solar load than the south side, with the result that some
occupants on one side could be too cool and/or some on the other side too warm if
3-3
-------
TABLE 1
Baseline Building Design and Operating Conditions;
Small Office in Miami Strip Mall
Parameter
Baseline Value
Total size
Zones
Height
Orientation
Building shade
Maximum occupancy
Occupancy pattern
Lighting
Equipment (elec. outlets)
Lighting/equipment schedule
Infiltration rate
Exterior walls
Window area
Window glass type
Roof
Floor slab
40 ft frontage by 100 ft depth (4,000 ft2).
Subdivided into 2,000 ft2 front and rear zones (each
40 by 50 ft).
9 ft ceiling height, with 4-ft-high plenum overhead
for utilities and return air.
Front faces north.
A 10-ft overhang along the front at ceiling height
(9 ft) to protect mall customers; a 3-ft high
parapet along the front roof to visually shield
rooftop equipment, protect service personnel.
27 persons (1 50 ft2/person).
Full occupancy at 8-1 lam and 2-5pm weekdays;
80% at 11 am-noon and 1 -2pm; 40% at noon-
1pm; 30% at 6-7am and 5-6pm; 10% at 6-
8pm; zero at 8pm-6am weekdays, and at all
times on weekends and holidays.
1.8 W/ft2.
0.75 W/ft2.
Full power at 7am-5pm weekdays; 30% at 6-7am and
5-6pm; 10% at 6-8pm; 5% at all other times
on weekdays, weekends, and holidays.
0.1 air changes per hour.
Concrete block with exterior stucco and interior
insulation; overall heat transfer coefficient
(including interior and exteriorfilm resistances)
UQ = 0.16 Btu/h ft2 F°.
46% of front wall area, 20% of rear wall area.
Single pane with tinting, coating, and/or interior
shading [shading coefficient = 0.55, U0 (incl.
aluminum frame) = 0.94 Btu/h ft2 F°],
Built-up roofing over mineral board insulation and
metal deck; U0 (incl. film resistances) = 0.066
Btu/h ft2 F°.
4-in. thick concrete slab with carpeting.
3-4
-------
Adjacent |
Office \
Adjacent
Office
Rear Wa
Interior Wai
Front
Zone 50 ft
Front Wa
Front Overhang
Figure 1. Floor plan for the baseline 4,000 ft2 office in a Miami strip mall.
3-5
-------
Overhang
Glass
18ft.-
1 ft.
L—
Parapet
Glass
Door
2 ft,
" 7.5 ft. |
3 ft.
Glass
_P
4 ft.
9 ft.
-7.5fhzr)
1 ft.
40 ft.
7 ft.
^5ft.j
J**—^
Front View
T
6|ft.
2,5 ft,|
Glass
Glass
Door
T
4 ft.
9 ft.
5 ft-
•8.5 ft. —»»4.5 ft.-
3 ft.
Rear View
4*4.5 ft.-*
Figure 2. Front and rear views of the baseline office.
3-6
-------
TABLE 2
Baseline HVAC Design and Operating Conditions:
Small Office in Miami Strip Mall
Parameter
Baseline Value
HVAC system type
Office cooling setpoint
Office heating setpoint
OA ventilation rate
Minimum supply air
temperature for cooling
Maximum supply air
temperature for heating
Maximum humidity in offices
Economizer
Cooling capacity and
sensible heat ratio
Cooling efficiency
Heating efficiency
Return air method
Two rooftop direct expansion, constant volume,
packaged single-zone (PSZ) units, one dedicat-
ed to each of the 2,000 ft2 zones.
75°F during occupied hours (6am-7pm weekdays);
cooling is shut off at all other times.
70°F during occupied hours; heating is set back to
55°F at all other times.
5 (and, as warranted, 20) cfm/person
55°F
100°F
Not set.
Economizer present, controlled by indoor vs. outdoor
temperatures.
Calculated by program.
Electric input ratio (EIR) = 0.341 Btu/h electric input
per Btu/h cooling output [corresponding to
energy efficiency ratio (EER) = 10 Btu/h
cooling output per W electric input].
EIR = 1.0 Btu/h electric input per Btu/h cooling
output (electric resistance heating).
Return via overhead plenum (unducted).
3-7
-------
one attempted to condition the entire 4,000 ft2 with a single, 10-ton PSZ system.
Technically, though, individual PSZ units are commercially available in capacities
greater than 10 tons.
3.4 THE "PLANT"
Packaged systems, such as PSZ systems, incorporate their "primary" and
"secondary" HVAC equipment into a single unit. The classical "primary" equipment
required for built-up systems, addressed in the PLANT portion of DOE-2 — e.g.,
boilers, chillers, and cooling towers - are absent. These components are replaced by,
e.g., electric resistance heating, an electric-driven compressor, and an air-cooling
condenser fan within the packaged unit, defined under SYSTEMS (Section 3.3 above).
As a result, the PLANT portion of the DOE-2 computation in this modeling of
packaged systems may be viewed simply as the supplier of purchased electricity for:
the lighting and office equipment in the building; the compressor, outdoor condenser
fan, electric resistance heaters, and the air handling unit associated with the packaged
HVAC system; and the electric domestic hot water (DHW) heater assumed for this
office.
3.5 ENERGY COSTS
The electric rate structure used for this analysis was the GSD-1 schedule for
Florida Power and Light Company, which serves the Miami area. This is the rate
structure used by Shirey and Rengarajan.
This structure involves a basic energy charge of $0.0473 per kWh, plus a peak
demand charge of $9.96/kW for each kilowatt above 10 kW. There is no demand
charge for the first 10 kW.
3.6 WEATHER DATA
The Typical Meteorological Year (TMY) weather file for Miami was used as the
baseline file, for essentially all of these calculations. For comparison, one run was
made with the Weather Year for Energy Calculations (WYEC) file for Miami.
The DOE-2 user has three choices of weather files for Miami (each of which
provides one year of hourly data): the Test Reference Year (TRY) files and the TMY
files, available from the National Climatic Data Center, and the WYEC files, available
from ASHRAE. TRY files provide data from a single, selected year; as such, these
files might result in a somewhat greater variation between predicted and observed
energy consumption, since no single year is likely to be "typical" during all 12 months.
3-8
-------
TMY files improve on the TRY files by assembling 12 "typical" months from different
years {e.g., a typical January from one year, a typical February from another). WYEC
files attempt to provide even further improvement by assembling "weather events"
(of a duration shorter than a month) from different years (e.g., a typical first two
weeks of July from one year, a typical second two weeks from another).
3.7 PARAMETRIC VARIATIONS FROM THE BASELINE VALUES
The baseline values presented in Tables 1 and 2 (and in Appendix B) for the
building and HVAC design and operating parameters were selected because they were
felt to be reasonably typical for an office of the type being considered.
For this sensitivity analysis, alternative values were selected for most of these
parameters, varying from the baseline values. These alternative values are presented
in Appendix B. These alternative values commonly include a lower value and a higher
value, bracketing the baseline value. The range defined by these alternative values
usually define the broadest reasonable range that might be expected in practice; for
example, the EER of the cooling coils is varied through a range extending from 8 to
12 Btu/h cooling output per W electric input. In some cases, extreme values are
selected in order to demonstrate maximum effects (for example, perfect total insula-
tion of the walls and roof, or the total absence of glazing).
In the semantics used in Appendix B, the "low" alternative value shown in the
appendix is the value expected to result in reduced energy consumption, and the
"high" value is that expected to result in increased consumption.
3.8 STRATEGY FOR THE CALCULATIONS
As the first step, runs were made with the baseline building and HVAC system
to define baseline HVAC performance, HVAC capacity requirements, energy consump-
tion, and energy costs. Baseline runs were made at OA ventilation rates of both 5
and 20 cfm/person, to estimate the impact of increased ventilation. The baseline
results are presented in Section 4 of this report.
As the next steps, each of the individual building and HVAC parameters was
varied in turn, through the values discussed in Section 3.7, at a ventilation rate of 5
cfm/person. For each parametric variation, the predicted incremental impacts on
performance, capacity requirements, and energy use were calculated. Where a given
parameter appeared to have a significant effect, or where otherwise of interest, the
calculation for a given parametric variation was repeated at a ventilation rate of 20
cfm/person. The results of this sensitivity analysis are reported in Section 5.
3-9
-------
As the final step, calculations were made to assess the practical ability of
packaged HVAC systems to control RH in this office, and the approximate energy
costs involved in doing so as ventilation rate is increased from 5 to 20 cfm/person.
These calculations were made recognizing the potential uncertainties in using DOE-2
to estimate indoor RH in humid climates, as discussed in Section 1.4. The results of
this assessment of RH control are discussed in Section 6.
In the sensitivity analysis calculations reported in Sections 4 and 5, the HVAC
capacities reported in the tables are those calculated by the computer program as
being necessary to meet the load. [Only in Section 5.17, where the effects of
capacity are explored, were HVAC capacities and sensible heat ratios (SHRs) specified
for the runs, simulating commercially available equipment.] The computed capacities
never correspond exactly to the capacities available in commercial cooling units,
which are commonly marketed in increments of one-half to one ton of refrigeration (6
to 12 kBtu/h). Thus, a variation in a building or HVAC parameter that increased the
computed cooling capacity requirement by some fraction of a ton of refrigeration
could, in practice, necessitate the installation of a unit at the next highest tonnage
increment. Installation of a higher-capacity unit would impact the system operation
in a number of ways, impacting system performance and energy use.
This issue could have been addressed by specifying, in the DOE-2 input file for
each run, the capacity and SHR of the commerically available unit that would be
needed for that parametric variation, rather than accepting the program's computed
(fractional tonnage} default value. It was decided to use the program-calculated
default capacities because that approach was felt to provide a clearer measure of how
the parametric variations impact capacity requirements.
HVAC performance is measured in terms of the predicted percentage of total
hours throughout the year that the office space is undercooled, and in terms of the
percentage of the occupied hours when the RH exceeds 60%. With the zone design
temperatures specified in Appendices A and B, and with the decision discussed above
to allow the program to calculate the cooling capacities, the percentage of hours
undercooled is generally less than 0.5%, and the percentage of occupied hours
calculated as having RH > 60% is generally about 1 %.
3-10
-------
SECTION 4
BASELINE RESULTS
4.1 SUMMARY TABLES
The results of the computations for the baseline building and HVAC system (as
defined in Tables 1 and 2) are shown in Table 3, for OA ventilation rates of both 5
and 20 cfm/person.
As discussed in the previous section, system performance is summarized in
terms of the percentage of hours undercooled and at RH values > 60% (averaged over
the two zones to represent the entire office space). System design requirements are
summarized in terms of the program-calculated HVAC cooling capacity (summed for
the two PSZ units combined). Building and system operating requirements are
summarized in terms of the energy consumption and energy costs for the HVAC
system by itself, and for the total building (including the HVAC system, the lighting
and office equipment, and the DHW heater).
To show how the total electric energy consumption in the office is distributed,
Table 4 breaks down the total annual building energy consumption figures in Table 3,
according to the various end uses.
4.2 DISCUSSION
End uses of energy. Two of the significant conclusions from Table 4 are:
1) The HVAC system consumes something less than half of the total power
required by the office, a fraction that remains essentially unchanged as the OA
ventilation rate is increased from 5 to 20 cfm/person. The largest single power
consumer -- requiring half of the office's consumption - is the lighting and
office equipment (computers, copiers, fax machines, etc.).
2) In the Miami climate, power consumption for heating is almost negligible in this
office. The bulk of the HVAC power consumption is for sensible and latent
cooling. For this reason, the HVAC capacity addressed in Table 3 (and in
subsequent tables) is the cooling capacity.
HVAC capacity requirements. As shown in Table 3, the combined cooling
capacity of the two PSZ units would have to increase by 1 5% as the OA ventilation
4-1
-------
TABLE 3
Results from DOE-2.1E Modeling of
the Baseline Small Office in Miami
Output Variable
Value of Output Variable when: Increase caused by increase
OA = 5 OA = 20 in OA from 5 to 20 cfm/person
cfm/oerson cfm/person Variable Units %
Total required cooling
capacity1 (kBtu/h)
Annual HVAC energy
consumption1 (kWh)
Annual HVAC energy
costs1 {$)
Annual building energy
consumption1 (kWh)
Annual building energy
costs1 ($)
% of all hours {8760
hr/year} undercooled1
% of occupied hours {3276
hr/year) when RH>60%1
103.558
26,145
2,510
60,161
4,273
0.4
1.2
119.193
29,390
2,835
63,406
4,598
0.4
0.9
15.635
3,245
325
3,245
325
-0.3
15.1
12.4
12.9
5.4
7.6
¦0
•0
HVAC capacity, energy consumption, and energy cost numbers represent the sum for
the two PSZ units combined; building energy figures include both zones. Zone
undercooling and humidity performance numbers represent the total 4,000 ft2 of office
space, i.e., the average for the two zones.
4-2
-------
TABLE 4
Annual Electric Energy Consumption in the Baseline Small Office,
Broken Down According to End Use
End Use
Annual Electric Energy
Consumption for that
End Use (kWh)
For OA = 5 For OA = 20
cfm/oerson cfm/person
Percentage of Total
Annual Consumption
in the Office
For OA = 5 For OA = 20
cfm/person cfm/person
HVAC System
Cooling (compressor
and condenser
fan)
Heating (electric
resistance)
Air handling fan
Auxiliaries (comp.
crankcase heater)
Subtotal - HVAC
18,778
30
7,328
22,001
52
7,328
31
1
12
~0
34
1
11
~0
26,145
29,390
44
46
Other than HVAC
Lighting and office
equipment
Domestic hot water
heater
30,429
3,587
Subtotal - Non-HVAC 34,016
TOTAL FOR OFFICE 60,161
30,429
3,587
34,016
63,406
50
56
100
48
54
100
4-3
-------
rate is increased, if similar temperature and RH performances are to be maintained in
the zones.
Given that one ton of refrigeration corresponds to 12 kBtu/h, and that off the-
shelf packaged units are typically marketed in ton or half-ton increments, one would
likely choose a pair of units adding up to at least 9 tons (108 kBtu/h) for this office
if the ventilation rate were to be 5 cfm/person, and at least 10 tons (120 kBtu/h) if
the rate were to be 20 cfm/person. As shown later in Section 5, specification of
those higher capacities in the DOE-2 calculation (rather than using the program-
calculated values of 103.558 and 119.193 kBtu/h) would have notably improved the
temperature control performance shown in Table 3, would have had almost no effect
on the RH performance, and would have slightly increased energy consumption and
costs.
Energy consumption and costs. Table 3 indicates that increasing the ventilation
rate increases electric consumption by 3,245 kWh per year. (As shown in Table 4,
this increase results essentially entirely from an increase in HVAC cooling costs, as
expected.} Expressed as a percentage of HVAC energy consumption, this 3,245 kWh
increase represents an increase of over 12%; expressed as a percentage of total office
consumption, it represents an increase of over 5% [consistent with the results of Eto
(1990) in modeling small offices in Miami]. Since HVAC energy consumption is
something less than half the total office consumption (in this baseline case and in the
parametric variations covered in Section 5), the percentage increase in HVAC
consumption will always be something greater than twice the percentage increase in
total office consumption.
The corresponding energy cost increase resulting from increased ventilation is
$325 per year. This amount corresponds to a 12.9% increase in the HVAC energy
costs, and a 7.6% increase in total office energy costs. The percentage increase in
energy costs (for the baseline and in Section 5) is consistently greater than the
percentage increase in consumption, due to the effect of the electric demand charge.
Performance. As shown in Table 3, there is no significant change in the
percentage of hours undercooled as the ventilation rate increases, remaining at 0.4%
of the 8,760 hours in the year. This is not surprising. The computer program designs
the capacity of each PSZ unit to maintain the temperature in the zone that it is
conditioning. Accordingly, at 20 cfm/person, the program has increased the capacity
of each unit as necessary to meet the additional load created by the increased OA
flow.
The percentage of occupied hours >60% RH also remains almost unchanged,
at about 1 %, as the ventilation rate increases. As a matter of fact, according to these
DOE-2 computations, there is a small reduction in the number of elevated-RH hours
with the increase in OA ventilation rate -- from 1.2% of the 3,276 occupied hours (40
4-4
-------
hours per year) to 0.9% (29 hours per year), While this decrease in elevated-RH
hours is tiny, it is useful to note conceptually why the program would predict a
decrease.
Elevated-RH hours tend to occur during the first several hours after startup on
cool, humid mornings. This occurs because the cooling coils are operating at a small
fraction of capacity (or are off altogether) due to the low sensible load during those
cool hours, so that there is little or no latent cooling. Increasing the OA rate to 20
cfm/person causes, on average, an increase in the sensible load during these morning
hours. Since constant-volume PSZ systems handle load changes by modulating
supply air temperature, this increased sensible load triggers a reduction in the PSZ
cooling coil temperature. The reduction in coil temperature increases the latent
cooling provided by the system at 20 cfm/person. The increased latent cooling can
more than offset the increased latent load caused by the increased influx of humid
OA. Table 3 indicates that the increase in latent cooling exceeds the increase in latent
load during cool morning hours sufficiently to eliminate 11 elevated-RH hours when
the OA is increased. (This effect is discussed in greater detail in Section 5.12.2.)
As discussed in Section 6, Shirey et al. (1995), and Shirey and Rengarajan
(1996), suggest that, in fact, RH levels could increase at increased OA flows, rather
than decreasing (or remaining about the same) as predicted here. According to their
calculations, RH increases resulting from moisture capacitance and condensate re-
evaporation effects at 20 cfm/person would more than compensate for the RH
decreases due to the two effects discussed above. There is no way to independently
verify the Shirey and Rengarajan results here using the DOE-2 model.
4-5
-------
[This page intentionally blank.]
4-6
-------
SECTION 5
THE EFFECTS OF PARAMETRIC VARIATIONS
5.1 BUILDING ORIENTATION
In the baseline configuration, the front of the office was facing north. The
predicted effects of orientation in the other directions are summarized in Table 5.
For this comparison, the effects of building orientation are shown as
incremental changes from the baseline configuration with an OA ventilation rate of 5
cfm/person. Runs at 20 cfm/person are shown only for the cases where the building
is facing north (since that is the baseline direction), and where it is facing south (since
that is the direction which, at 5 cfm/person, resulted in the greatest variation in
energy consumption and cost relative to the baseline).
5.1.1. The Effect of Orientation at 5 cfm/person
The first three rows in Table 5 (for the ventilation rate of 5 cfm/person) show
that building orientation has only a minor impact on energy consumption in this
building.
Among the four compass directions, a south orientation results in the lowest
required cooling capacity and energy consumption/cost. When the front office faces
south, the program predicts a 3% reduction in HVAC energy consumption and costs,
relative to the north-facing baseline. Review of the program reports confirms that the
reason for this result is that both the window solar load and the wall conduction load
for the building are distinctly lower for the south orientation, relative to any other
orientation. With the south orientation, the 10-foot front overhang is on the south
side, providing exterior shading in the direction from which the maximum solar gain
would be received. The unshaded rear glass and wall is on the north side, and hence
will be shaded by the building itself. As a result, annual cooling (compressor and
condenser) power consumption is the lowest for any of the four orientations. Also,
with the shading, peak loads are lower with the result that system flows are lower;
hence, power consumption by the air handling fan is reduced.
The greatest capacity and energy requirements result when the front of the
building is facing east (with HVAC energy consumption and costs increasing about
1 % relative to the baseline). East and west are the two orientations providing the
greatest window solar load, since the rising and setting suns, respectively, have a low
incident angle which allows the sun to hit the highly glazed front of the building
5-1
-------
TABLE 5
Effect of Building Orientation:
Increase Compared to Baseline Case with 5 cfm/person1
Case
Increase in
total required
cooling capacity
In
kBtu/h As %
Increase in
annual energy consumption
As % of As % of
HVAC building
In kWh energy energy
Increase in annual energy cost
As % of As % of
HVAC building
In $ energy cost energy cost
Hours/yr Occupied
under- hours/yr
cooled, RH> 60%,
% %
Baseline values: 103.5582
Bldg faces north,
OA = 5 cfm/p
(absolute values)2
OA ventilation rate = 5 cfm/person
Bldg faces south -0,196 -0.2
Bldg faces east 5.050 4.9
Bldg faces west 1.276 1,2
OA ventilation rate = 20 cfm/person
Bldg faces north 15.635 15.1
Bldg faces south 16,009 15.5
26,1452
(HVAC)
60,1 612
(building)
-820 -3.1 -1.4
186 0.7 0.3
24 0.1 -0
3,245 12.4 5.4
2,451 9.4 4.1
2,5102
(HVAC)
4,2732
(building)
-73
31
4
325
256
-2.9
1.2
0.2
12.9
10.2
-1.7
0.7
0.1
7.6
6.0
0.4
0.4
0,5
0.7
0.4
0.4
1.2
1.3
1.2
1.2
0.9
0.9
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
underneath the overhang; and, during latter or initial part of the day, respectively, the
sun will be hitting the unshaded rear windows. In fact, for this building, the west
orientation results in a slightly higher window solar load and wall conduction load than
does the east orientation. However, the east orientation has the somewhat higher
peak load -- probably because it receives the early morning solar gain on the front
glazing on summer Monday mornings when peak loads will occur. As a result, the
east orientation has somewhat higher air flows, and hence higher power consumption
by the air handling fan, causing this orientation to have an annual power consumption
(and a cooling capacity requirement) slightly greater than that for the west orientation.
For all four orientations, the percentage of hours undercooled remains at about
0.5%, and the percentage of occupied hours >60% RH remains at about 1.2%.
5.1.2 The Effect of Orientation on Increased Ventilation Rates
The last two rows in Table 5 show the effects of north vs. south orientation
at the increased OA ventilation rate of 20 efm/person.
At 5 cfm/person, the south orientation results in a reduction of about 800
kWh/year and $70/year, relative to the north orientation. This same differential exists
between the two orientations at 20 cfm/person. For example, increasing the ventila-
tion rate in the north orientation results in an increased energy consumption of 3,245
kWh/year relative to the north-facing/5 cfm baseline; increased ventilation in the south
orientation results in an increase of only 2,451 kWh/year from the north-facing/5 cfm
baseline, 794 kWh less.
Stating the above point in another way, the difference in annual energy
consumption between the south-facing/20 cfm case and the south-facing/5 cfm case
is (2,451) - (-820) = 3,271 kWh. This is essentially identical to the 3,245 kWh
difference between the north-facing/20 cfm case and the north-facing/5 cfm case.
For a given climate and HVAC system, the incremental cost of treating an additional
15 cfm/person of outdoor air will not be heavily dependent on the orientation of the
building. Therefore, if one had happened to choose a south-facing office rather than
a north-facing office to assess the effects of increased ventilation, this choice would
not have significantly affected the conclusions.
For either the north or south orientation, an increase to 20 cfm/person would
require approximately a 16 kBtu/h (1.3 ton) increase in cooling capacity. In either
case, from a practical standpoint (with PSZ units assumed to be available only in 0.5-
ton increments), this would translate into the need to install units totalling 10 tons of
cooling capacity rather than 9 tons.
5-3
-------
5.2 BUILDING SHADE
In the baseline configuration, the front of the office was shaded with a 10-foot
overhang at ceiling height along its entire frontage, whereas the rear was not shaded
at all. Table 6 shows the effects for the cases where: a) the rear is also shaded, with
overhangs that extend 6 ft outward at ceiling height over the rear windows and door;
and b) there are no shading overhangs at all, either on the front or the rear.
5.2.1 The Effect of Building Shade at 5 cfm/person
From Table 6, adding shading on the rear decreases HVAC energy consumption
and costs by 214 to 3%. The small reduction in cooling capacity requirements {0.713
kBtu/h) would result in no change in the capacity of the commercial units that would
have to be purchased (9 tons total), since the units will likely have to be purchased
in 0.5-ton increments. This rear shading is having its maximum effect in this baseline
orientation, which has the rear facing south. Computations confirm that the energy
consumption and cost savings resulting from the addition of the rear shading would
have been slightly less had the rear been facing one of the other compass directions,
where the solar gain through the t/zishaded rear windows would be less.
Removing the front shading from the baseline case increases HVAC energy
consumption and costs by 21/z to 3%. The 5.213 kBtu/h increase in cooling capacity
requirements could result in the need to increase the installed capacity by 0.5 ton, to
9.5 tons. Deletion of the front shading has its minimum effect in the baseline
orientation, since the front is facing north. The large amount of front glazing is thus
receiving substantial shading from the building itself even in the absence of the front
overhang. If the front office were facing south, the penalty for deleting the front
overhang would jump from 667 kWh and $69 per year, shown in Table 6 for the north
orientation, to 1,641 kWh and $159 per year, an increase of about 6% over the
baseline.
The changes in building shading have essentially no impact on system
performance, in terms of hours undercooled or at elevated RH. Again, major changes
in these output parameters would not be expected, especially for the hours under-
cooled, since the program should be sizing the system capacity in each case to
achieve consistent thermal performance.
5.2.2 The Effect of Shade on Increased Ventilation Rates
From Table 6, in a building where all overhangs are deleted, the added energy
consumption resulting from increasing the ventilation rate from 5 to 20 cfm/person
in the north-facing building is {3,950 - 667) = 3,283 kWh/year, and the added cost
is ($401 - $69) = $332/year. In the worst-case situation where the overhangs were
deleted in a south-facing building, these differentials from increased ventilation would
5-4
-------
TABLE 6
Effect of Building Shading:
Increase Compared to Baseline Case with 5 cfm/person1
Case
Increase in
total required
cooling capacity
In
kBtu/h As %
Increase in
annual energy consumption
As % of As % of
HVAC building
In kWh energy energy
Increase in annual energy cost
As % of As % of
HVAC building
energy cost energy cost
In $
Hours/yr Occupied
under- hours/yr
cooled, RH> 60%
% %
Baseline values: 103.5582
1 Q-ft overhang in
front, none in rear,
OA = 5 cfm/p
(absolute values)2
OA ventilation rate = 5 cfm/oerson
Add 6-ft overhang
over rear door
and windows -0.713 -0.7 -734 -2.8 -1.2 -65 -2.6 -1.5 0.3 1.2
Delete all over-
hangs (front and
rear) 5.213 5.0 667 2.6 1.1 69 2.7 1.6 0.4 1.2
OA ventilation rate = 20 cfm/person
Delete all overhgs 21.718 21.0 3,950 15.1 6.6 401 16.0 9.4 0.4 0.9
26,145 2,510 0.4 1.2
(HVAC) -- - (HVAC)
60,1 612 4,273'
(building) -- — (building)
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
be 3,175 kWh and $311. These impacts of increased ventilation are similar to those
shown for the baseline building in Table 3, and for different orientations in Section
5.1,2. Again, for a given climate and HVAC system, the incremental energy to treat
an additional 15 cfm/person of OA does not vary significantly as a function of the
building overhang configuration. Even if the modeler had happened to choose the
worst-case situation of a south-facing office with no overhang, this choice would not
have significantly affected the calculated effects of increased ventilation,
However, the calculations do show that -- especially when the building is facing
south -- such shading can have a meaningful impact in reducing energy consumption
{in addition to protecting clientele from the weather). In a south-facing office without
an overhang, increasing the ventilation rate from 5 to 20 cfm/person would increase
annual energy consumption by 3,175 kWh at a cost of $311, as indicated above. But
if a front overhang were installed at the same time that the ventilation rate were
increased, the net increased annual consumption would be only 810 kWh at a cost
of $97. Thus, adding an overhang in that case could nominally recover more than
two-thirds of the energy penalty resulting from the increase in OA rate.
In terms of practical cooling capacity, the addition of rear overhangs results in
no effective change in the required increase in capacity when ventilation rate is
increased from 5 to 20 cfm/person; units totalling 9 tons must be increased by 1 ton,
to 10 tons, with or without the rear overhang. If the front overhang is deleted,
increased ventilation would again result in a capacity increase of 1 ton, except in this
case, the increase would be from 9.5 tons to 10.5 tons {whether the building were
facing north or south). For the hypothetical case suggested in the previous para-
graph -- where a south-facing building with no front overhang and a ventilation rate
of 5 cfm/person is increased to 20 cfm/person and an overhang added - the required
practical capacity increase would be only 0.5 ton, from 9.5 to 10 tons.
5.3 OCCUPANT DENSITY
In the baseline configuration, the occupant density was that specified by
ASHRAE 62-1989 (7 persons per 1,000 ft2, or approximately 150 ft2/person), corres-
ponding to about 27 persons in the 4,000 ft2 office. Occupant density is particularly
important in the calculations, because it not only determines the sensible and latent
load from people, but it also significantly impacts the actual cfm of outdoor air that
is brought into the building in response to the specified cfm per person {and hence the
sensible and latent load for conditioning the OA).
Table 7 shows the calculated effects of: a) decreasing occupancy by 50%, to
13 persons (300 ft2/person); and b) increasing occupancy by 50%, to 40 persons
{100 ft2/person).
5-6
-------
TABLE 7
Effect of Building Occupancy:
Increase Compared to Baseline Case with 5 cfm/person1
Case
Increase in
total required
coolina caoacitv
In
kBtu/h As %
Increase in
annual enerav consumotion
As % of As % of
HVAC building
In kWh enerav enerav
Increase in annual enerav cost
As % of As % of
HVAC building
In $ enerav cost enerav cost
Hours/yr
under-
cooled,
%
Occupied
hours/yr
RH > 60%,
%
Baseline values:
103.5582
„
26,1452
2,5102
0.4
1.2
150 ft2/person.
(HVAC)
--
--
(HVAC)
—
—
OA = 5 cfm/p
60,1612
4,2732
(absolute values)2
(building) —
(building)
—
OA ventilation rate
= 5 cfm/oerson
300 ft2/person
-9.171
-8.9
-2,151
-8.2
-3.6
-207
-8.2
-4.8
0.3
1.0
100 ftJ/person
8.387
8.1
2,152
8.2
3.6
204
8.1
4.8
0.3
1.2
OA ventilation rate
= 20 cfm/oerson
300 ft2/person
-0.252
-0.2
-436
-1.7
-0.7
-37
-1.5
-0.9
0.4
0.8
100 ft2/person
30.678
29.6
6,703
25.6
11.1
674
26.9
15.8
0.4
0.9
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
5.3.1 The Effect of Occupancy at 5 cfm/person
The first two rows in Table 7 show that, at 5 cfm/person, decreasing and
increasing the number of occupants by 50% have meaningful, equal effects in
opposite directions, as expected. HVAC energy consumption, HVAC energy cost, and
cooling capacity are each decreased by about 8% by decreasing the number of
occupants, and increased by 8% by increasing occupancy.
At 300 ftVperson, the predicted percentage of hours above 60% RH decreases
slightly relative to the baseline, due to the reduced latent load from people and the
reduced flow of moisture-containing OA.
In the extreme, if the number of occupants were decreased to zero, HVAC
energy consumption would be reduced by approximately 15%. Stated another way,
roughly 15% of the HVAC energy consumption in the baseline building at 5 cfm/
person is due to the sensible and latent load contributed by the occupants.
5.3.2 The Effect of Occupancy on Increased Ventilation Rates
The last two rows in Table 7 show that increasing the ventilation rate to 20
cfm/person at 300 ft2/person results in HVAC energy consumption, costs, and cooling
capacity requirements that are slightly less than the 1 50 ft2/person baseline at 5
cfm/person. At 100 ftVperson, increasing the ventilation rate results in more than a
25% increase in HVAC energy and capacity over the baseline.
At 300 ft2/person, increasing the OA ventilation from 5 to 20 cfm/person
increases energy consumption by (-436) - {-2,151} = 1,715 kWh, and increases
energy costs by (-37! - (-207) = $170. By comparison, at 100 ft2/person, the
increase from 5 to 20 cfm/person results in increases of 4,551 kWh and §470.
Occupancy is one of the few parameters that causes these incremental increases from
increased ventilation to vary dramatically from the baseline increases shown in Table
3 (3,245 kWh and $325).
In terms of required cooling capacity, at 300 ft2/person, the increase in
ventilation rate would necessitate an increase of 1 ton in the practical installed
capacity, from 8 to 9 tons. This is the same 1-ton increment as required by the
baseline, except in that case, the increase is from 9 to 10 tons. At 100 ftVperson,
the increase is 1.5 tons, from 10 to 11.5 tons.
This result underscores the obvious conclusion: The fewer the number of
people in the space, the lower the cost of increasing the flow of OA per person.
Clearly, the assumed occupant density can have a significant impact on the energy
penalty that a modeler would predict from increasing ventilation from 5 to 20 cfm/
person.
On this same basis, the use of demand-controlled ventilation (DCV) could
conceptually reduce the incremental penalty of increased ventilation. DCV would
effectively reduce the number of occupants used by the HVAC system in the OA cfm
per person calculation during periods of reduced occupancy.
5-8
-------
5.4 LIGHTING AND OFFICE EQUIPMENT POWER CONSUMPTION
In the baseline building, overhead lighting was assumed to consume 1,8 W/ft2
and office equipment 0.75 W/ft2 (for a total of 2.55 W/ft2), consistent with ASHRAE
90.1-1989. Increased lighting and equipment power consumption in the building
would have two effects: a) it would increase the sensible heat load that would have
to be addressed by the HVAC system; and b) it would increase the non-HVAC
component of building power consumption, causing the incremental energy penalty
of increased ventilation to seem to be less when expressed as a percentage of total
building energy.
Table 8 shows the estimated effects of: a) decreasing the lighting +
equipment power consumption to 1.0 + 0.5 = 1.5 W/ft2; and b) increasing this
consumption to 2.25 + 1.75 = 4.0 W/ft2. The rationales for these selections are
summarized in Appendix B. The lower value for power consumption would corres-
pond to the use of efficient modern lighting, daylighting, and Energy Star appliances.
An increase in the lighting/equipment power consumption causes an increase
in the HVAC energy consumption and cost estimates (due to increased sensible heat
load), plus a much larger increase in total building energy consumption and costs (due
to the increased HVAC energy plus increased consumption by the lights and
equipment). In Table 8, the increases/decreases shown for energy consumption and
costs include only the HVAC energy variations, not those for the total building.
5.4.1 The Effect of Lighting/Equipment Power at 5 cfm/person
The results in Table 8 for the two runs at 5 cfm/person show dramatic
decreases and increases in HVAC energy and capacity requirements as lighting/
equipment power requirements are decreased/increased. These large decreases/
increases demonstrate the importance of this load on HVAC operation. No other
building or HVAC parameter studied here has had as great an impact on HVAC energy
consumption, HVAC energy costs, and required cooling capacity {_±_ 14-27%) at the
low ventilation rate of 5 cfm/person.
The range of lighting/equipment power consumptions considered here (1.5 to
4.0 W/ft2) admittedly covers the extremes, causing the size of the observed
variations; modelers generally do not use these extreme values. However, the
selected power consumptions do vary between published modeling studies from about
2.5 to 3.5 W/ft2, a differential that could still create a significant difference between
estimates. This would appear to be a variable which needs to be selected with some
care.
In the extreme, if the lighting wattage were reduced to zero, HVAC energy
consumption would be reduced by approximately 39%, If equipment wattage were
reduced to zero, HVAC energy would be reduced by about 17%. Thus, sensible heat
from the lighting and equipment is responsible for approximately 39% and 17%,
respectively, of the total HVAC energy consumption in the baseline building at 5 cfm/
person -- the largest contributions from any individual source.
5-9
-------
TABLE 8
Effect of Lighting and Office Equipment Power Consumption:
Increase Compared to Baseline Case with 5 cfm/person1
increase in
Increase in
total required
annual enerav consumption
Increase in annual enerav cost
Hours/yr
Occupied
coolina capacity
As % of
As % of
As % of
As % of
under-
hours/yr
In
HVAC
building
HVAC
building
cooled,
RH > 60%,
Case
kBtu/h As %
In kWh enerav
enerav
In $ enerav cost
enerav cost
%
%
Baseline values:
103.5582
26,1452
2,5102
0.4
1.2
2.55 win2,
(HVAC) --
~
(HVAC)
—
OA-5 cfm/p
60,1 612
4,2732
(absolute values)2
(building) --
—
(building!
OA ventilation rate
= 5 cfm/oerson
1.5 W/ft2
-14.695 -14.2
-5,020s -19.2
-8.3
-437s -17.4
-10.2
0.3
2.0
4.0 W/ft2
20.961 20.2
7,0383 26.9
11.7
6323 25.2
14.8
0.4
0.8
OA ventilation rate
= 20 cfm/oerson
1.5 W/ft2
0.404 0.4
-1,9583 -7.5
-3.3
-1273 -5.1
-3.0
0.3
1.3
4.0 W/ft2
37.350 36.1
10,4103 39.8
17.3
9713 38.7
22.7
0.5
0.6
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase irom the baseline.
Increase (or decrease) in HVAC energy and costs; the increase (or decrease) in total building energy and costs is much greater.
-------
The _±_ 17-27% variation in energy consumption and costs from baseline values
shown in Table 8 is for HVAC energy. If one considered the impacts on total building
energy consumption and costs - considering the impact of lighting/equipment power
on non-HVAC, as well as HVAC, energy requirements -- the results are even more
dramatic. At a ventilation rate of 5 cfm/person, varying lighting/equipment power
between 1.5 and 4.0 W/ft2 causes total building energy consumption and cost to vary
by an impressive +_ 29-50% from the baseline (2.55 W/ft2) total building energy
values.
From the standpoint of performance, it is noted that the percentage of occupied
hours above 60% RH is slightly higher than the baseline when lighting/equipment
power drops to 1.5 W/ft2, and slightly lower than the baseline when this value rises
to 4.0 W/ft2. This effect occurs because the higher lighting/equipment power
consumption results in a consistently higher sensible heat load on the HVAC system.
As discussed in Section 4.2, increased sensible load tends to increase the latent
removal of this system - including during cool mornings after startup, when elevated-
RH occupied hours occur. Thus, higher lighting/equipment power consumption results
in fewer elevated-RH hours. The lower lighting/ equipment power case has reduced
sensible loads, and hence increases the number of elevated-RH hours.
5.4.2 The Effect of Lighting/Equipment Power on Increased Ventilation Rates
At 1.5 W/ft2, increasing the OA ventilation rate from 5 to 20 cfm/person
increases HVAC energy consumption by (-1,958) - (-5,020) = 3,062 kWh/year, and
HVAC energy costs by (-127) - (-437) = $310/year. The comparable values for
increasing ventilation at 4.0 W/ft2 are 3,372 kWh and $339. Thus, the absolute
values for the incremental costs of increased ventilation are approximately the same
for both of these cases, and for the baseline case (Table 3), as would be expected.
Expressed as percentages of the total building energy consumption at 5 cfm/
person, these impacts of increased ventilation are: 7.2% increase at 1.5 W/ft2 (total
building energy at 5 cfm/person = 42,611 kWh); 5.4% increase at the baseline value
of 2.55 W/ft2 (total building energy at 5 cfm/person = 60,161 kWh, from Table 3);
and 4.0% increase at 4.0 W/ft2 (total building energy at 5 cfm/person = 84,502
kWh). So the selected value of lighting/equipment power requirements could make
a couple percentage point difference in the impact of increased ventilation, when
expressed in this manner. Except for occupant density (Section 5.3), no other single
building or HVAC parameter makes this percentage vary so much from the 5.4%
baseline value.
From Table 8, it is noted that HVAC energy consumption and cost at 1.5 W/ft2
and a ventilation rate of 20 cfm/person are 1,958 kWh and $127 lower (5 to 8%
lower) than they are for the baseline case (at 2.55 W/ft2 and 5 cfm/person). This
effect is even more pronounced when total building energy consumption is considered:
total building energy consumption is 45,674 kWh/year (at a cost of $3,040/year) at
1.5 W/ft2 and 20 cfm/person, which is about 25% less than the 60,161 kWh and
5-11
-------
$4,273/year at 2.55 W/ft2 and 5 cfm/person. There would be some increase in
installed cost associated with the efficient lighting systems, which cannot be
addressed here. However, in concept, if one designed a new building with highly
efficient lighting and conserving appliances, one could more than offset the HVAC
energy cost penalty of increasing the ventilation rate from 5 to 20 cfm/person.
A similar point can be made regarding HVAC cooling capacity. The required
capacity at 1.5 W/ft2 and 20 cfm/person is 103.962 kBtu/h, which would translate
into a practical installed capacity of 108 kBtu/h (9 tons). For the 2.55 W/ft2 and 5
cfm/person case, the required capacity is 103.558 kBtu/h, which would also translate
to 9 tons. Thus, an improvement in lighting and equipment efficiency could nominally
offset any increase in cooling capacity that would otherwise be necessitated by an
increase in ventilation rate.
5.5 INFILTRATION RATE
The baseline building is assumed to have an infiltration rate (i.e., an uncontrol-
led entry rate for outdoor air through leaks around windows, doors, etc.) of 0.1 air
changes per hour (ACH). This was felt to represent a reasonably typical rate for a
well-performing building {see Appendix B). Table 9 shows the effects when the
assumed infiltration rate is changed to: a) 0 ACH (i.e., a building that is pressurized
everywhere, preventing infiltration); and b) 0.3 ACH, one of the higher infiltration
rates commonly assumed for modern offices.
5.5.1 The Effect of Infiltration at 5 cfm/person
As shown in the 5 cfm/person rows of Table 8, in the narrow ACH range
considered here, HVAC energy consumption and cost appear to vary by 1 to 2%
(either upward or downward, depending on the direction of the ACH change) for each
0.1 ACH change in the assumed infiltration rate. Required cooling capacity appears
to vary by perhaps 2 to 4% per 0.1 ACH change.
Stating the results at 0 ACH in another manner, sensible and latent heat from
infiltration are responsible for approximately 2% of the total HVAC energy consump-
tion in the baseline building.
5.5.2 The Effect of Infiltration on Increased Ventilation Rates
At 0 ACH, increasing ventilation rate from 5 to 20 cfm/person would increase
cooling capacity requirements by 16.129 kBtu/h, energy consumption by 3,288 kWh,
and energy cost by $329. These increments are essentially identical to the
increments experienced at the baseline infiltration rate of 0.1 ACH, as expected.
One question of interest is: If an increased ventilation rate pressurizes a
building and hence reduces infiltration, to what extent might the savings from reduced
infiltration compensate for the energy penalty resulting from increased ventilation?
As shown in Table 3, relative to the baseline (0.1 ACH, 5 cfm/person), the energy
5-12
-------
TABLE 9
Effect of Infiltration Rate:
Increase Compared to Baseline Case with 5 cfm/person1
Increase in
Increase in
total required
annual enerav consumption
Increase in annual enerav cost
Hours/yr
Occupied
coolina capacity
As % of
As % Of
As % of
As % of
under-
hours/yr
In
HVAC
building
HVAC
building
cooled,
RH >60%,
Case kBtu/h As %
In kWh enerav
enerav
In $ enerav cost
enerav cost
%
%
Baseline values: 103.558 s
26,1452
2,5102
0.4
1.2
0.1 ACH,
(HVAC) --
—
(HVAC)
-
OA = 5 cfm/p
60,1612
4,2732
(absolute values)2
(building) --
—
(building)
OA ventilation rate = 5 cfm/oerson
0 ACH -2.567 -2.5
-385 -1,5
-0.6
-43 -1.7
-1.0
0.4
1.2
0.3 ACH 8.577 8.3
895 3.4
1.5
106 4.2
2.5
0.2
1.1
OA ventilation rate = 20 cfm/oerson
0 ACH 13.562 13.1
2,903 11.1
4.8
286 11.4
6.7
0.4
0.9
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
penalty of increasing the ventilation rate to 20 cfm/person (0.1 ACH/20 cfm/person)
is 3,245 kWh, and the cost penalty is $325. By comparison, from Table 9, the
energy penalty of increased ventilation at 0 ACH (0 ACH/20 cfm/person) relative to
the baseline (0.1 ACH/5 cfm/person) is 2,903 kWh (342 kWh, or 10%, less than the
3,245 kWh at 0.1 ACH), and the cost penalty is $286 ($39, or 12%, less than the
$325). Thus, if a building with an infiltration rate of 0.1 ACH and a ventilation rate
of 5 cfm/person had its infiltration rate ideally reduced to zero when its ventilation rate
was increased to 20 cfm/person, the reduction in infiltration would offset only 10 to
12% of the energy penalty associated with the increased ventilation.
If the initial building had had an infiltration rate of 0.3 ACH at 5 cfm/person
(rather than the baseline 0.1 ACH), and if this 0.3 ACH had been reduced to zero by
the increase in ventilation rate to 20 cfm/person, then the reduction in infiltration
would offset approximately 40% of the ventilation energy penalty. This is probably
about the maximum benefit that might ideally be anticipated.
In terms of commercial cooling capacity, the baseline case with 20 cfm/person
(0.1 ACH/20 cfm/person) would require a pair of PSZ units totalling 10 tons, rounded
upward to the nearest half ton, as indicated previously. Even if the increase in
ventilation rate caused the infiltration to drop to zero (0 ACH/20 cfm/person), the
required commercial capacity would still round upward to 10 tons.
5.6 EXTERIOR WALL RESISTANCE TO HEAT TRANSFER
The baseline building has exterior walls constructed of heavy-weight hollow
concrete block with stucco on the exterior and with insulation and gypsum board on
the interior. This wall construction is typical, and represents an overall heat transfer
coefficient for the wall (including the inside and outside film coefficients) of U0 =
0.16 Btu/h ft2 F°.
Table 10 shows the estimated effects of the following.
a) Reducing the exterior wall U0 to zero (i.e., providing infinite wall resistance),
representing the extreme case. The glazing and doors associated with the
exterior (front and rear) walls remain in place, so that there is still conduction
and solar gain through those sources.
b) Replacing the block walls with 4-in. stud walls insulated with R-11 batts (Uc =
0.064 Btu/h ft2 F°), a realistic wail construction having better resistance to heat
transfer than does the baseline biock wall, but not the ideal infinite resistance
assumed in a) above.
c) Deleting the inside insulation on the exterior walls, increasing UQ to 0.34
Btu/h ft2 F°, representing a case of poor heat transfer resistance.
5-14
-------
TABLE 10
Effect of Exterior Wall Resistance:
Increase Compared to Baseline Case with 5 cfm/person1
Increase in
total required
coolina capacity
In
Case kBtu/h As %
increase in
annual enerav consumption
As % of As % of
HVAC building
In kWh enerav enerav
Increase in annual enerav cost
As % of As % of
HVAC building
In $ enerav cost enerav cost
Hours/yr
under-
cooled,
%
Occupied
hours/yr
RH > 60%,
%
Baseline values: 103.5582
26,1452
2,5102
0.4
1.2
U„ = 0.1 6\
(HVAC)
-
--
(HVAC)
—
—
OA = 5 cfm/p
60,1 612
4,2732
(absolute values)2
(building) --
—
(building)
—
OA ventilation rate = 5 cfm/person
U0 = 0 -0.755 -0.7
-401
-1.5
-0.7
-38
-1.5
-0.9
0.3
1.1
U0 = 0.063 -0.307 -0.3
-185
-0.7
-0.3
-16
-0.6
-0.4
0.3
1.2
UQ = 0.343 2.07 9 2.0
527
2.0
0.9
54
2.2
1.3
0.4
1.3
OA ventilation rate = 20 cfm/oerson
U0 = 0 14.731 14.2
2,834
10.8
4.7
287
11.4
6.7
0.3
0.9
UQ = 0.063 1 5.31 7 1 4.8
3,046
11.7
5.1
308
12.3
7.2
0.3
0.9
Capacity, energy consumption, and cost numbers represent the two PSZ units combined- Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
Units of U0 are Btu/h ft2 Fa. Wall U0 includes interior and exterior film resistances.
-------
5.6.1 The Effect of Wall Resistance at 5 cfm/person
As shown in Table 10, varying the wall resistance at 5 cfm/person had only a
limited effect. Decreasing the wall heat transfer coefficient by a factor of more than
2.5 from the baseline, to 0.064 Btu/h ft2 F°, reduces HVAC energy consumption and
costs by only 0.6 to 0.7%. In the extreme, reducing this coefficient all the way to
zero would reduce HVAC energy consumption/costs only by 1.5%. On the other
hand, doubling U0 to 0.34 Btu/h ft2 F° (an extreme achieved by deleting all insulation
from the baseline wall) has a slightly greater impact in the opposite direction,
increasing HVAC energy consumption/costs by about 2%.
The results at a wall UQ of zero indicate that sensible heat conduction through
the exterior walls is responsible for approximately 2% of the total HVAC energy
consumption in the baseline building at a ventilation rate of 5 cfm/person.
On the basis of the results shown in Table 10, one would have to look closely
at the cost-effectiveness before incurring any significant additional construction costs
in an effort to improve wall resistance to heat transfer beyond the baseline value.
This limited benefit from improved exterior wall insulation is consistent with results
reported by others (Parker, 1996). As shown in later sections, improvements to the
glass and the roof offer somewhat greater potential.
5.6.2 The Effect of Wall Resistance on Increased Ventilation Rates
As shown in Table 3, increasing OA ventilation rate from 5 to 20 cfm/person
in the baseline building - without any change in the wall resistance -- resulted in an
increase of 3,245 kWh/year in HVAC energy consumption. As shown in Table 10,
increasing ventilation rate from 5 cfm/person in the baseline building with the
insulated block walls (5 cfm/0.16 Btu/h ft2 F°) to 20 cfm/person in a building with
frame walls (20 cfm/0.06 Btu/h ft2 F°) increases consumption by 3,046 kWh. That
is, if one accompanied the increase in ventilation rate with a re-design of the walls in
this manner to improve the wall resistance, one might expect to reduce the energy
penalty resulting from the increased ventilation by only 199 kWh/year, or 6%.
In the extreme, if one accompanied the increase in ventilation rate with a wall
re-design that ideally raised wall resistance to infinity, one could reduce the energy
penalty to 2,834 kWh/year -- a modest reduction of 411 kWh/year, or 13%.
Thus, improving exterior wall resistance to heat transfer does not appear to be
the most promising avenue for achieving major reductions in energy consumption.
5-16
-------
5.7 AMOUNT OF GLAZING
The baseline building has a front wall area 46% of which consists of glazing
(including the glass front door), and a rear wall area which is 20% glazed. Table 11
shows the estimated effects of:
a) deleting all of the glazing, front and rear; and
b) increasing the glazing in the rear such that it is identical to the glazing on the
front.
5.7.1 The Effect of Glazing Amount at 5 cfm/person
As shown in Table 11 for the ventilation rate of 5 cfm/person, deleting all
glazing decreases energy consumption and costs, and decreases design cooling
capacity, by 8 to 10%. Increasing the glazing on the rear from 20% to 46%
increases these parameters by approximately the same amount.
The energy savings resulting from deleting the glazing are reduced by the fact
that the front windows are shaded. Likewise, the energy penalty associated with
increasing the rear glazing is increased by the fact that there is no rear shading. For
example, deleting the shading overhang on the front would have increased the energy
savings due to glazing elimination from the 2,535 kWh/year shown in Table 11 to
3,135 kWh/year, or about 12%.
The 9.7% reduction in HVAC energy consumption shown in Table 11 when the
windows are deleted assumes that, realistically, the glazing is replaced by exterior wall
area that conducts a lesser amount of heat. If one instead assumes, as an extreme,
that the glazing is replaced by a surface that is completely thermally neutral, the
results (with other adjustments! indicate that heat conduction and solar radiation
through the glass are responsible for roughly 14% of the total HVAC energy consump-
tion in the baseline building.
Complete elimination of the glazing is an extreme measure that would often not
be a practical option from an aesthetics standpoint. However, these figures indicate
the maximum savings that might be achieved by reducing the amount of glazing. The
savings appear meaningful but modest.
Reasons for observed effects on hours undercooled and at elevated RH.
Variations in the amount of glazing have only minor effects on the performance of the
system (the percentage of hours undercooled or at elevated RH). Although these
effects are small, it is of interest to understand conceptually why the observed
variations occur. The hours undercooled (typically 20 to 30 hours per year) and the
hours above 60% RH (typically 20 to 50 hours per year) commonly occur during the
first occupied hours on warm weekday mornings, especially on summer Monday
mornings. The HVAC system has been off overnight or over the weekend; thus,
5-17
-------
TABLE 11
Effect of Amount of Glazing:
Increase Compared to Baseline Case with 5 cfm/person1
Case
Increase in
total required
cooling capacity
In
kBtu/h As %
Increase in
annua! energy consumption
As % of As % of
HVAC building
in kWh energy energy
Increase in annual energy cost
As % of As % of
HVAC building
$ energy cost energy cost
In
Hours/yr Occupied
under- hours/yr
cooled, RH> 60%,
% %
Baseline values: 103.5582
46% glazing in
front, 20% in rear
OA = 5 cfm/p
(absolute values)2
OA ventilation rate - 5 cfm/oerson
0% in front,
0% in rear
46% in front,
46% in rear
-8.574 -8.3
9.922 9.6
OA ventilation rate = 20 cfm/oerson
0% in front,
0% in rear
10.805 10.4
26,1452
(HVAC) -
60,1 612
(building) --
-2,535
2,069
779
-9.7
7.9
3.0
-4.2
3.4
1.3
2,5102
(HVAC)
4,273*
(building)
-258 -10.3
215
88
8.6
3.5
-6.0
5.0
2.1
0.4
0.2
0.3
0.2
1.2
0.8
1.3
0.7
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, nor the increase from the baseline.
-------
temperatures and RH levels are elevated, and heat has gotten stored in the building
materials and furnishings during this warm unconditioned period, depending upon their
heat capacities. During the first few hours on a warm Monday morning in Miami, the
air entering the cooling coils has the highest temperature and moisture content that
the system will see during that day. As a result, even though the cooling compressor
is operating continuously during those hours, it might be unable to reduce the warm
entering air down to the supply air temperature desired by the zone, depending upon
the amount of over-design in the system. The warm entering air will also cause the
coil surface temperature to be higher, with the result that less of the moisture in the
air will be condensed.
When the glazing is deleted, the calculations predict that the interior of the
building can be 1 F° or more cooler just before system startup on a summer Monday
morning, relative to the baseline case with glazing. Thus, the air initially entering the
cooling coils on startup is cooler, with the result that the supply air temperature
leaving the coils is cooler by up to 1 F°, helping to avoid undercooling of the space.
Also, the coil surface temperature is lower by up to 1 F", condensing more moisture
and thus helping to reduce the RH of the space. The lower space temperature prior
to startup also results in less heat buildup in the building materials and furnishings,
with the result that the space cools faster; on a Monday where the baseline building
with glazing remains undercooled for the first four hours, the building with no glazing
might remain undercooled for only the first two hours. Contributing to the faster
cooling is the fact that, with the glazing absent, there is no glass conduction and glass
solar radiation load on the space during these first hours.
As a result of the lower supply air temperatures achievable by the PSZ systems
in the building with no glazing, Table 11 shows that, at the ventilation rate of 5
cfm/person, the percentage of hours undercooled drops from the 0.4% with the
baseline building, to 0.2% for the no-glazing case. Similarly, due to the lower coil
surface temperatures in the no-glazing case, the percentage of occupied hours above
60% RH drops to 0.8% from the baseline 1.2% at 5 cfm/person. These reductions
in supply air and coil temperatures occur despite the reduction in cooling capacity
shown in Table 11 for the no-glazing case; the reduced load more than compensates
for the reduced capacity.
When the amount of glazing is increased from the baseline at 5 cfm/person,
Table 11 shows that hours undercooled decrease slightly relative to the baseline
values {to 0.3% from 0.4%), and hours at elevated RH increase slightly (from 1.2%
to 1.3%). The reason for these effects is that the increased cooling capacity designed
for the increased-glazing case more than compensates for the increased entering air
temperatures on warm weekday mornings, but does not quite compensate for the
increased humidity. With the glazing increased, the temperature in the rear zone is
perhaps 0.7 F° warmer on a summer Monday morning just before system startup,
but - whereas the baseline rear zone might remain undercooled for two hours on that
5-19
-------
morning -- the greater capacity in the increased glazing case is able to adequately
reduce the zone temperature within only one hour, despite the elevated initial coil inlet
temperature.
This ability of the higher-capacity system to reduce the number of hours
undercooled in the increased-glazing case does not translate into a similar ability to
reduce hours at elevated RH. The higher-capacity system achieves its increased
cooling in this case to a large extent by providing a greater flow rate of supply air, and
to a lesser extent by reducing the temperature of that supply air relative to baseline
values. Thus, coil surface temperatures in the higher-capacity system are only a
fraction of a degree lower than in the lower-capacity baseline case; sometimes the coil
temperatures are the same in both cases. Thus, the higher-capacity system is not
always able to condense enough additional moisture to compensate for the greater
initial moisture content in the increased-glazing case.
Similar small variations in hours undercooled and at elevated RH will be
observed in the other tables in Section 5. While exact scenarios will vary from case
to case, the explanations will always involve the same factors discussed above: the
entering air temperature (and moisture content) at startup on warm weekday morn-
ings; the cooling capacity; and the constant PSZ design flows. These factors will
determine; the temperature (and flow) of the supply air exiting the coils, and hence
the ability to reduce hours undercooled; and the coil surface temperature, hence the
ability to reduce hour at elevated RH.
5.7.2 The Effect of Glazing Amount on Increased Ventilation Rates
Hypothetically, if one deleted the glazing in the design of a new building and
designed the HVAC system to provide 20 rather than 5 cfm OA/person, the deleted
glazing would reduce the net energy penalty from the 3,245 kWh/year shown in Table
3 for the ventilation increase alone, to the 779 kWh/year shown in Table 11. Thus,
deleted glazing would nominally recover about 75% of the energy penalty associated
with the increased ventilation. Reducing rather than eliminating the glazing would, of
course, recover a smaller percentage of the energy penalty.
This observation is not intended as a recommendation that glazing be eliminated
or significantly reduced. The decision regarding the glazing configuration will be
determined by aesthetics and cost-effectiveness considerations on a site-specific
basis. The observation here is intended only to illustrate the relative energy impacts
of increased ventilation versus substantially reduced amounts of glazing.
In terms of performance, Table 11 shows that the hours undercooled and at
elevated RH in the no-glazing case at 20 cfm/person are about the same (or are
slightly less than) as in the no-glazing case at 5 cfm/person (and, correspondingly, are
lower than in the baseline building at 5 cfm/person). This decrease occurs despite the
5-20
-------
increase in both sensible and latent heat resulting from the increased OA flow rate.
The reason is that the increase in cooling capacity in the 20 cfm/person case
compensates for the increase in incoming sensible and latent heat. Thus, after startup
on a summer Monday morning, zone temperature is reduced more rapidly, despite the
elevated temperature of the air entering the coils. Where the zone was undercooled
during the first two hours in the no-glazing/5 cfm case, it might be undercooled for
only one hour in the no-glazing/20 cfm case.
The design air flow rate in the 20 cfm/person case is the same as in the
corresponding 5 cfm/person case, due to the manner in which the systems are
designed. Thus, the increase in capacity at 20 cfm/person must be achieved by
operating at lower coil temperatures rather than by increasing flow. The coil
temperature in the no-glazing/20 cfm case is up to 1.5 F° lower than in the no-
glazing/5 cfm case {and up to 2.5 F° lower than in the baseline glazing/5 cfm case).
As a result, the elevated initial humidities on Monday mornings are more rapidly
reduced in the 20 cfm/person case {despite the increased latent load added by the
incoming OA), and the number of hours at elevated RH decreases.
This improved performance for 20 cfm/person systems, relative to the
corresponding 5 cfm/person systems, is seen consistently in the tables throughout
Section 5, for the same reasons discussed above.
5.8 GLASS TYPE
The baseline building has single-pane glass in a thermally broken aluminum
frame, providing an overall heat transfer coefficient (including the frame) of U0 = 0.94
Btu/h ft2 F°, determining heat conduction through the glass. It has a shading
coefficient (S-C) of 0.55, determining the transmittance, reflectance, and absorptance
of solar radiation. A S-C of 0.55 can be achieved with single-pane glass through
various combinations of glass tinting, reflective coatings, and interior shading.
Table 12 shows the estimated effects of:
a) Decreasing the fenestration Uc to 0.32 Btu/h ft2 F° by switching to double-pane
glass with a coating, and decreasing S-C to 0.16, about the lowest value that
can reasonably be achieved through the use of tinting, highly reflective coat-
ings, and interior shading.
b) Increasing the S-C to 0.94 (with a U0 of 0.94 Btu/h ft2 F°), by switching to
simple single-pane glass with no tinting, coating, or interior shading
(representing glass having minimum resistance).
An even more extreme case than a) above would be to delete the glazing altogether,
a case considered in Section 5,7.
5-21
-------
TABLE 12
Effect of Glass Type:
Increase Compared to Baseline Case with 5 cfm/person1
Case
Increase in
total required
cooling capacity
In
kBtu/h As %
Increase in
annual energy consumption
As % of As % of
HVAC building
In kWh energy energy
Increase in annual energy cost
As % of As % of
HVAC building
In $ energy cost energy cost
Hours/yr Occupied
under- hours/yr
cooled, RH > 60%,
% %
Baseline values: 103.5582
U0 = G.94 Btu/h-
ft2F°, S-C = 0.55,
OA = 5 cfm/p
(absolute values)2
OA ventilation rate = 5 cfm/person
Uc = 0.32,
S-C = 0.16
-5.755 -5.6
26,145z
(HVAC)
60,1 612
(building)
-1,936 -7.4
-3.2
2,5102
(HVAC)
4,2732
(building)
-194 -7.7
-4.5
0.4
0.2
1.2
0.9
UD = 0.94,
S-C = 0.94 7.366 7.1
OA ventilation rate = 20 cfm/person
U0 = 0.32,
S-C = 0.16
13.725 13.3
1,930 7.4
1,405 5.4
3.2
2.3
188
153
7.5
6.1
4.4
3.6
0.5
0.2
1.1
0.7
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
5.8.1 The Effect of Glass Type at 5 cfm/person
As shown in Table 12, increasing the resistance of the glass to conduction and
solar radiation as described in a) above reduces HVAC energy consumption and costs
by 7 to 8%. Decreasing resistance as described in b) increases consumption and
costs by a similar amount. The computed required cooling capacity also varies by
similar amounts.
As discussed in Section 5.7.1, increasing glass resistance in the extreme -- by
deleting the glass altogether -- reduces energy consumption and costs by about 10%.
As with changes in the amount of glazing, the incremental effects observed due
to changes in the characteristics of the glazing are impacted by the configuration of
the exterior shading around this office. The greater the amount of exterior shading,
the less the incremental benefits of improving glazing (or the less the incremental
penalty of using less resistant glazing). For example, if the baseline building had had
no front overhang, the reduction in energy consumption achieved by using the glazing
in a} above would have been 2,374 kWh (rather than 1,936 kWh), a 9% reduction
(rather than 7%).
The 7 to 10% energy savings shown in Tables 11 and 12 for improvement in
(or elimination of) the glazing compare with the maximum savings of less than 1 to
2% shown in Table 10 for improvements to exterior wall resistance, and the
maximum savings of about 6% that will be discussed in Table 13 for improvements
to roof resistance. It is not surprising that glazing improvements are more effective
than wall improvements, since, as indicated previously, the walls account for only 5%
of the baseline building's annual sensible heat load, whereas the glazing accounts for
14%.
But it is less obvious why glazing improvements should be more effective than
roof improvements, since the baseline roof accounts for an even larger portion of the
sensible load, 21 %. The reason is probably that the heat conducted through the roof
into the plenum area goes largely to increasing plenum temperature, with only a
relatively small portion being conducted across the ceiling into office space. This heat
in the plenum is channeled directly into the cooling coils with the return air, where the
increased entering air temperature creates a larger AT driving force between the air
and the coils, giving more effective heat removal. By comparison, the heat introduced
by conduction and radiation through the glazing enters the office space, where it is
mixed with the larger volume of office air before being routed to the cooling coils.
Thus, it has less effect in increasing the temperature of the air entering the coils,
makes less contribution to the AT driving force, and thus is not removed as efficiently.
Hours undercooled and at elevated RH vary slightly with the changes in glass
type, for reasons similar to those discussed in Section 5.7.1.
5-23
-------
5.8.2 The Effect of Glass Type on Increased Ventilation Rates
If one replaced the baseline glass with the high-performance glass considered
here at the same time that ventilation rate were increased from 5 to 20 cfm/person,
the improved glazing would decrease the net energy penalty from the 3,245 kWh/year
shown in Table 3 for the ventilation increase alone, to the 1,405 kWh/year shown in
Table 12 for the 20 cfm/person case. Thus, improving the glazing in this manner
would nominally recover 57% of the energy penalty associated with the increased
ventilation.
Of course, if the baseline glazing were replaced with improved glazing that was
not as resistant as the high-performance glazing assumed here, the energy benefits
from improving the glazing would be less.
No effort is made here to assess the installation costs, and hence the overall
cost-effectiveness, of improving the glazing. The goal here is simply to indicate the
variables having the greatest impact on energy consumption.
The percentage of hours undercooled for the 20 cfm/person case in Table 12
are about the same as that for the corresponding 5 cfm/person case; the percentage
of hours at elevated RH are slightly less. These are consistent with the results
discussed in Section 5.7.2, and have the same explanation.
5.9 ROOF RESISTANCE TO HEAT TRANSFER
The baseline building has a roof construction consisting of built-up roofing over
R-12 mineral board and a metal deck. This roof represents an overall heat transfer
coefficient (including the inside and outside film coefficients) of UQ = 0.066 Btu/
h ft2 F°.
Table 13 shows the estimated effects of the following.
a) Reducing the roof U0 to zero (i.e., providing infinite roof resistance),
representing the extreme case.
b) Replacing the R-12 mineral board with lesser insulation, approximately doubling
Uc to 0.12 Btu/h ft2 F°).
5.9.1 The Effect of Roof Resistance at 5 cfm/person
As shown in Table 13, eliminating all heat transfer through the roof (U0 = 0)
at 5 cfm/person reduces HVAC energy consumption and costs by about 6 to 7%.
Approximately doubling the heat transfer increases energy consumption by about a
similar amount.
5-24
-------
TABLE 13
Effect of Roof Resistance:
Increase Compared to Baseline Case with 5 efm/person1
Increase in
Increase in
total required
annual enerov consumotion
Increase in annual
enerav cost
Hours/yr
Occupied
coolina capacity
As % of
As % of
As % of
As % of
under-
hours/yr
In
HVAC
building
HVAC
building
cooled,
RH> 60%,
Case kBtu/h As %
In kWh e
enerav
In $ enerav cost enerav cost
%
%
Baseline values: 103.5582
26,1452
2,5102
0.4
1.2
UD = 0.066s,
(HVAC)
-
-
(HVAC)
-
OA = 5 cfm/p
60,1612
4,2732
(absolute values)2
(building]
I ™
—
(building)
OA ventilation rate = 5 cfm/oerson
c
o
II
o
o
o
-1,609
-6.2
-2.7
-165 -6.6
-3.9
-0
0.8
U0 = 0.123 9.725 9.4
1,619
6.2
2.7
206 8.2
4.8
0.4
1.7
OA ventilation rate = 20 cfm/oerson
U0 = 0 15.635 15.1
1,673
6.4
2.8
168 6.7
3.9
-0
0.8
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
Units of Uc are Btu/h ft2 F°. Roof U0 includes interior and exterior film resistances.
-------
The results for the case where roof U0 is zero, when adjusted, indicate that
sensible heat conduction through the roof is responsible for approximately 8% of the
total HVAC energy consumption in the baseline building. This modest contribution
from the roof is greater than the 2% from the exterior walls (Section 5.6.1), but is
less than the 14% estimated from the glazing (Section 5.7.1).
Table 13 shows no change in required cooling coil capacity when UD is
decreased to zero, a result that might seem curious. But when Uc is increased to 0.12
Btu/h ft2 F°, capacity requirements increase by 9%, consistent with intuition. The
explanation for these results is as follows.
DOE-2 computes the cooling coil capacity based upon design temperatures for
the office space and the plenum (specified in the input file as DESIGN-COOL-T).
These design temperatures are generally selected to provide some selected degree of
over-design. As discussed in Appendix B, for the calculations here, the design cooling
temperature for the office space is 74°F, a little below the actual cooling setpoint of
75°F, so that the coils are somewhat over-designed. Likewise, the design temperature
for the plenum is 90°F, whereas, for the baseline building with the baseline roof U„ of
0,066 Btu/h ft2 F°, the peak hourly temperature actually encountered in the plenum
is above 89°F. Accordingly, the cooling coils are slightly over-designed, assuming that
these coils will have to cool the entering air from 90°F to the desired supply air
temperature, when in fact, the entering air will have to be cooled only from about
89°F or less.
Since the plenum never quite reaches the design temperature at the baseline
roof U0, it does not, either, when U0 is reduced to zero and the actual plenum
temperature becomes even lower. Accordingly, in both cases, the program designs
the coil capacity assuming a plenum temperature of 90"F, and there is no change in
capacity when the roof U0 is decreased below the baseline.
But when the roof U0 is increased to 0.12 Btu/h ft2 F°, more heat can enter the
plenum, and actual plenum temperatures do reach and exceed 90°F. When this
occurs, the program adjusts the plenum design cooling temperature upward, so that
the coils will not be under-designed. As a result, when the roof U0 is increased to this
extent over the baseline value, the computed capacity does increase as intuitively
anticipated.
When U„ is reduced to zero, the percentage of hours undercooled drops to
about zero. At U0 = 0, the actual plenum temperatures are further below the design
temperatures than they were in the baseline case, and the cooling coils are thus more
over-designed. Accordingly, the coils now have sufficient excess capacity to handle
the spikes in entering air temperature that they encounter at startup on warm week-
day mornings, as discussed in Section 5.7.1, and these early-morning undercooled
hours are essentially eliminated.
5-26
-------
But when UD is raised to 0.12 Btu/h ft2 F°, the percentage of undercooled hours
remains about the same as in the baseline case. This is the result of two competing
phenomena.
a) The program has adjusted the plenum design cooling temperature upward, so
that the cooling capacity is no longer over-designed as it was in the case where
U0 = 0. However, because of the heat now conducting through the roof in the
afternoon, the design cooling capacity is 9% greater than in the baseline case,
with the result that the capacity at U0 = 0.12 is better able than the baseline
to handle the first morning hours on warm weekday startups. As a result, a
summer Monday morning that might have been undercooled for the first four
hours in the baseline case and the first one hour in the U0 = 0 case, is
predicted to be undercooled for the first two hours in the U0 = 0.12 case.
b) But because so much more heat conducts through the roof in the U0 = 0.12
case, this case experiences some undercooled hours around 4 pm in the
afternoon, a situation not encountered in the baseline case. On the particular
Monday morning in a) above, the UD = 0.12 case experiences two undercooled
hours at 4 and 5 pm in the afternoon, with the result that this case, like the
baseline, experiences a total of four undercooled hours on that day. The hours
are just spread between the morning and the afternoon, rather than all being
in the morning.
The causes for the reduced percentage of occupied hours at RH>60% at
reduced U„ are also similar to those discussed in Section 5.7.1. At U0 = 0, the
plenum and office space remain cooler over nights and weekends; the lower entering
air temperatures during startup on warm weekday mornings result in lower coil
temperatures, and the number of early-morning elevated RH hours is thus reduced
compared to the baseline.
When UD is increased to 0.12 Btu/h ft2 F°, the percentage of occupied hours at
RH> 60% increases to 1.7%, among the highest predicted for any of the cases
considered in this report. At first, this result seems counter-intuitive. The cooling
coils in the U0 = 0.12 case commonly operate at temperatures 1 to 2 F° colder than
those in the baseline case during warm weather, so that they condense more
moisture. (The increase in capacity shown in Table 13 for this case is achieved by
lowering supply air temperature while maintaining the same constant PSZ flow as that
in the baseline case.) Correspondingly, hourly reports for the particular summer
Monday considered above show that the latent cooling is greater and the moisture
levels in the offices {expressed in lb moisture/lb dry air) are lower for the U0 = 0.12
case than for the baseline case, as expected.
The reason for the apparent anomaly is that this improved RH control during
warm weather in the UD = 0.12 case is apparently more than offset by poorer RH
5-27
-------
control during cooler weather. On a coo! winter Monday, increased heat loss through
the roof in the Uc = 0.12 case makes it unnecessary for the cooling coils to come on
(to handle internal heat gains) until 11 am, whereas the baseline coils have to come
on at 9 am. Then, for the first two hours after the coils switch on in the U0 = 0.12
case, the coils are operating at a smaller part load ratio than are the baseline coils (due
to the increased free roof cooling), with the result that the Btu/h of latent cooling is
lower. As a result, RH levels in the offices (from internal generation) during the first
four hours of this winter Monday morning are greater for the UQ = 0,12 case than for
the baseline case.
A similar effect occurs on mild spring and fall Mondays. In the baseline case,
the economizer opens to a greater extent in the hour or two before the cooling coils
switch on; and, once the coils do switch on, they are operating closer to full load and
are achieving more Btu/h of latent cooling. So again, the baseline case achieves lower
RH in the offices than does the U0 = 0.12 case.
The net effect of these phenomena is that the Uc = 0.12 case has a much
greater percentage of occupied hours at RH > 60% than does the baseline case, with
fewer of these hours occurring during the summer, and more occurring during cool
and mild seasons.
5.9.2 The Effect of Roof Resistance on Increased Ventilation Rates
Using the same rationale as that in Section 5.6.1, Table 13 shows that -- if an
increase in ventilation rate in the baseline building were accompanied by steps that
reduced the roof heat transfer to zero -- the increase in energy consumption would be
1,673 kWh/year, rather than the baseline increase of 3,245 kWh/year. The improve-
ment in roof resistance would thus conceptually recover 48% of the energy penalty
associated with the increased ventilation.
Of course, reducing roof heat transfer to zero is an extreme case that would not
be achievable in practice. In a more "practical" extreme case, where R-19 batts are
added beneath the roof deck (decreasing Uc to 0.028 Btu/h ft2 F°), the improved roof
insulation would compensate for 28% (rather than 48%) of the energy penalty of the
increased ventilation.
Again, no effort is made here to assess the installation costs, and hence the
overall cost-effectiveness, of improving roof insulation.
At UQ = 0, the hours undercooled and at elevated RH remain essentially the
same with a ventilation rate of 20 cfm/person as in the 5 cfm/person case. The
reasons are essentially the same as those discussed in Section 5.7.2.
5-28
-------
5.10 ELIMINATION OF ALL EXTERIOR SURFACES
The previous four sections have addressed the individual effects of varying the
heat transfer resistances of the three types of above-grade exterior surface: the
wails, the windows, and the roof. In each case, the extreme value was to assume
that the surface had infinite resistance.
In Table 14, for illustration, the extreme case is considered where all three
types of exterior surface have infinite resistance simultaneously. In practice, such a
case could be approached only if the space being considered were completely
enclosed within a larger structure, i.e., if the space had no exterior walls, and all
boundaries of the space were interior walls adjoining other conditioned space.
5.10.1 The Effect of Infinite Exterior Resistance at 5 cfm/person
In the extreme, with all external loads except for infiltration and slab conduction
deleted, Table 14 shows that energy consumption and cost would decrease by about
20%, and design cooling capacity would decrease by about 10%.
Thus, of the energy consumed by the baseline HVAC system, about 20%
results from loads on the walls, windows, and roof. (The sum of the individual
percentages from Sections 5.6.1, 5.7.1, and 5.9.1, which have been adjusted,
actually add up to 24%.) The remainder of the HVAC power consumption results
from: internally generated loads (occupants, lighting, and equipment); mechanically
introduced outdoor ventilation air (5 cfm/person for the percentage shown here);
infiltration; and very small contributions from conduction through the slab and door.
This represents the hypothetical maximum energy savings that could be
achieved with improved insulation.
5.10.2 The Effect of Infinite Exterior Resistance on Increased Ventilation Rates
The results for the 5 cfm/person case in Table 14 show that the combined
contribution of conduction and radiation through the walls, windows, and roof to
HVAC energy consumption in the baseline building is about 5,135 kWh/year. Table
3 showed that the incremental energy penalty of increasing the baseline ventilation
rate from 5 to 20 cfm/person is 3,245 kWh/year (which gets adjusted slightly, to
3,257 kWh/year when modeling the case of infinite exterior resistance).
Thus, the nominal energy savings of deleting the walls, windows, and roof is
greater than the energy penalty of increasing the ventilation rate, by a difference of
1,878 kWh/year. One could hypothetically increase the ventilation rate and save
1,878 kWh/year at the same time, if one could make the walls, windows, and roof
thermally neutral in the process.
5-29
-------
TABLE 14
Effect of Infinite Resistance for All Exterior Surfaces:
Increase Compared to Baseline Case with 5 cfm/person1
Increase in
Increase in
total required
annual enerav consumption
Increase in annual enerav cost
Hours/yr
Occupied
coolina caoacitv
As % of
As % of
As % of As % of
under-
hours/yr
In
HVAC
building
HVAC building
cooled.
RH > 60%,
Case
kBtu/h As %
In kWh enerav
enerav
In $ enerav cost enerav cost
%
%
Baseline values:
103.5582
26,145?
2,510Z
0.4
1.2
2 ext. walls, Uc =
(HVAC) --
~
(HVAC)
0.163, with glass;
60,1 612
4,2732
roof Uo = 0.06 63;
(building) ~
~
(building)
OA = 5 cfm/p
(absolute values)2
OA ventilation rate
= 5 cfm/Derson
No exterior
surfaces
-11.204 -10.8
-5,135 -19.6
-8.5
-561 -22.4 -13.1
0
0.4
OA ventilation rate
= 20 cfm/Derson
No exterior
surfaces
7.922 7.6
-1,878 -7.2
-3.1
-211 -8.4 -4.9
0
0.5
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
Units of U0 are Btu/h ft2 F°. Values of U0 include interior and exterior film resistances.
-------
Of course, completely eliminating heat gain through walls, windows, and the
roof is an extreme which is not achievable in practice (except by enclosing the space
within another building). A more "practical" extreme case would be to combine the
most resistive alternatives considered in Sections 5.6, 5.8, and 5.9: insulated frame
walls replacing the hollow-block walls, switching to highly reflective double-pane
glass, and adding R-19 insulation underneath the baseline roof. These steps together
would recover roughly 90% of the energy penalty associated with the increase in
ventilation rate. But there would still be some energy penalty, not a savings.
Table 14 shows that, with infinite exterior resistance, the number of hours
undercooled drops to zero with either ventilation rate. This result is not surprising.
With infinite resistance, the temperature buildup in the offices over warm nights and
weekends will be greatly reduced, limited to that resulting from infiltration, from the
5% of the lighting that remains on during unoccupied periods, and from the small
amount of conduction through the slab and door. As a result, the HVAC systems are
not overwhelmed by a high entering air temperatures upon startup on warm weekday
mornings. They always have sufficient capacity to reduce the supply air to the
temperature required by the zones.
5.11 THERMOSTAT SET-UP RATHER THAN SYSTEM SHUT DOWN DURING
UNOCCUPIED HOURS
The baseline HVAC system operating procedure involves complete shut-down
of the cooling coils overnight, and over weekends and holidays. This saves energy,
but allows temperatures and humidity levels in the offices and plenums to build up
during warm weather off-hours. As a result, the first operating hours on warm
weekday mornings (and especially on Monday mornings) can experience elevated
temperatures and RH levels.
Table 15 shows the results when - instead of shutting the cooling coils off on
nights and weekends - the office thermostats are set up to a cooling setpoint
temperature of 81 °F, 6 F° above the 75° setpoint used during occupied hours. For
these runs, the outside air dampers were kept closed during unoccupied hours, so that
the temperature control was achieved by circulating building air with no OA
ventilation.
As shown in the table, operation at 81 °F during unoccupied hours has the
desired effect of eliminating the percentage of hours undercooled to zero, for both
ventilation rates. On summer Sunday afternoons in Miami, the temperatures in the
baseline office space (with the cooling coils completely off) can reach 85°F, and those
in the plenums can approach 90°F. The temperatures entering the cooling coils upon
startup on Monday morning can be as high as 84°F in the baseline case. Keeping
these temperatures near 81 °F by leaving the coils on at the higher setpoint has two
5-31
-------
TABLE 15
Effect of Thermostat Set-Up Rather than System Shut-Down:
Increase Compared to Baseline Case with 5 cfm/person1
Case
Increase in
total required
cooling capacity
In
kBtu/h As %
Increase in
annual energy consumption
As % of As % of
HVAC building
In kWh energy energy
Increase in annual energy cost
As % of As % of
HVAC building
In $ energy cost energy cost
Hours/yr Occupied
under- hours/yr
cooled, RH >60%,
% %
Baseline values: 103.5582 — 26,1452 2,5102 0.4 1.2
Cooling coils (HVAC) - -- (HVAC)
shut down during 60,1 612 4,2732
off hours; (building) -- — (building)
OA = 5 cfm/p
(absolute values)2
OA ventilation rate = 5 cfm/person
Cooling setpoint
81°F off hours 0 0 622 2.4 1.0 6 0.2 0.1 0 1.2
OA ventilation rate = 20 cfm/person
Cooling setpoint
81CF off hours 15.635 15.1 3,892 14.9 6.5 335 13.3 7.8 0 0.9
Capacity, energy consumption, and cost numbers represent the two PSZ units combined. Undercooling and RH performance
numbers represent the average for the two zones.
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
effects which have been successful in eliminating undercooled hours, First, by
reducing the temperature of the air entering the cooling coils by several degrees at
weekday morning startups, this approach has made it possible for the systems of the
given capacities to reduce the supply air temperature to the levels demanded by their
zones during these crucial first hours of the day. Second, by keeping the tempera-
tures of the zones down over the night or weekend, this approach has reduced the
zones' demand at startup, again making it easier for the systems of the given
capacities to meet the demand.
No reductions are achieved in the percentage of occupied hours at elevated RH;
the 1.2% and 0.9% shown in Table 15 for the 81° set-up case are identical to the
values seen at 5 and 20 cfm/person, respectively, in the baseline (shut-down) case.
As discussed later, most of the occupied hours with elevated RH occur after startup
during cool weather, when the humidity can be high but when temperatures are so
low that the cooling coils operate at greatly reduced capacity (or are off altogether).
In such cool weather, off-hour ambient temperatures are often below 81 °F. With an
81 °F set-up temperature, the coils would not come on when the building is
unoccupied. As a result, the behavior of the building and the system during those
elevated-RH hours after startup on cool mornings is identical, regardless of whether
the cooling coils have been shut off during the unoccupied hours, or whether they
have been set up. Accordingly, there is no net effect on RH levels in the offices
during occupied hours.
However, there is a significant impact on RH levels during unoccupied hours in
warm weather. When the coils are shut off entirely during unoccupied hours, as in
the baseline case, the RH levels commonly reach 60 to 75% over warm summer
weekends, and also overnight in mid-week, due to the latent content of infiltrating
outdoor air. When the coils are instead left on at the 81°F setpoint, the coils tend to
cycle on during warm weekend afternoons, and sometimes on weekday mornings;
with this cycling, peak RH levels during unoccupied summer hours are reduced to
perhaps 40-50%. Since indoor biocontaminant growth can be significantly increased
at elevated RH levels, with 60% RH being the upper level recommended by ASHRAE
(ASHRAE, 1989a), the elevated RH levels experienced in humid climates when the
cooling coils are turned off over a weekend should be of concern. The relatively high
indoor temperatures over the weekends (up to almost 90°F in the plenum), combined
with the high RH, would help expedite biocontaminant growth.
Shirey and Rengarajan (1996) suggest that actual indoor RH levels during
occupied hours in humid climates might be higher than those predicted by DOE-2,
because DOE-2 does not take into consideration either the moisture capacitance of
furnishings and building materials, or the re-evaporation of moisture off the cooling
coils when the coils switch off with the air handler operating. At least in some
circumstances, moisture capacitance is calculated by those authors to be the more
significant of these two causes of the discrepancy between the DOE-2 RH predictions
and their own. That is, moisture in the infiltrating air is absorbed into the furnishings
and building materials over nights and weekends, and then takes some time to desorb
when the HVAC system is started up again, contributing to the occupied hours at
elevated RH in the Shirey and Rengarajan computations. To the extent that this is
5-33
-------
happening, it would add further importance to the need to keep RH levels down on
nights and weekends by leaving the coils on at a set up setpoint.
As would be expected, since DOE-2 calculates system capacities based upon
the design temperatures, the cooling coil capacity in the 81° set-up cases are
unchanged from their baseline {shut down) counterparts. The 15.635 kBtu/h increase
shown in Table 15 for the 81° set-up/20 cfm case, relative to the shut-down/5 cfm
baseline, is identical to the increase for the shut-down/20 cfm case.
As would be expected, Table 15 shows an increase in energy consumption in
the 81° set-up cases, relative to their baseline counterparts. At 5 cfm/person, the
increase is 622 kWh/year; at 20 cfm/person, the increase is 647 kWh/year relative to
the shut-down/20 cfm case.
But interestingly, the increases in cost are much smaller than the increases in
consumption would initially suggest - only $6/year at 5 cfm/person, and only $10/
year at 20 cfm/person (relative to the shut-down/20 cfm case). This is smaller than
the cost that would be calculated by simply applying the $0.0473/kWh electricity
charge to the increased kilowatt-hours of usage. The increase in usage is being
largely offset by reduced peak demands during warm months, and hence reduced
demand charges. Keeping the plenums (and offices) at more moderate temperatures
over warm nights and weekends shaves the peak load that is experienced during
startup hours on weekday mornings.
Based on these calculations, the operator of an office such as this would be
well advised to consider operation of the HVAC at a set-up cooling setpoint during
unoccupied periods, rather than turning the cooling system off. Such an approach
could significantly reduce the risk of biocontaminant growth over nights and
weekends (and would be of added importance if elevated RH levels during unoccupied
hours leads to elevated RH levels during occupied hours due to the moisture capaci-
tance of materials). It would also improve occupant comfort on warm mornings, from
the standpoints both of temperature and humidity. And it would achieve these
benefits at minimal cost.
5.12 ALTERNATIVE HVAC SYSTEMS
The baseline configuration assumes that the office is subdivided into two 2,000
ft2 zones, each conditioned by a dedicated packaged single-zone (PSZ) unit. Under
this assumption, each of the two PSZ units requires approximately 5 tons of cooling
capacity.
Table 16 shows the results when several poss ble alternative HVAC system
configurations are considered instead. The alternative HVAC system configurations
5-34
-------
TABLE 16
Effect of Assuming Alternative HVAC Systems:
Increase Compared to Baseline Case with 5 cfm/person
Increase in
Increase in
total required
annual enerav consumotion
Increase in annual
enerav cost
Hours/yr
Occupied
coolina capacity
As % of
As % of
As % of
As % of
under-
hours/yr
In
HVAC
building
HVAC
building
cooled.
RH>60%,
Case
kBtu/h As %
In kWh
enerav
enerav
In $
enerav cost enerav cost
%
%
Baseline values:
103.5581
26J451
2,5101
0.4
1.2
Two PSZ units.
(HVAC
) --
--
(HVAC)
—
—
OA = 5 cfm/p
60,1611
4,2731
(absolute values)1
(building) ~
—
(building)
—
OA ventilation rate
= 5 cfm/oerson
One PSZ unit
-3.060 -3.0
-377
-1.4
-0.6
-38
-1.5
-0.9
0.5
1.2
One PSZ, subzone3
0 0
-761
-2.9
-1.3
-43
-1.7
-1.0
0.3
1.2
One PSZ, subzone4
0 0
-1,105
-4.2
-1.8
-67
-2.7
-1.6
0.3
1.0
One PVAVS
1.479 1.4
-3,149
-12.0
-5.2
-186
-7.4
-4.4
-0
1.5
Two PTAC units
-6.718 6.5
-2,719
<*
0
*—
1
-4.5
-232
-9.2
-5.4
0.2
~0
OA ventilation rate
= 20 cfm/oerson
Two PSZ units2
15.635 15.1
3,245
12.4
5.4
325
12.9
7.6
0.4
0.9
One PSZ unit
12,512 12.1
2,858
10.9
4.8
285
11.4
6.7
0.5
0.8
One PSZ, subzone3
15.639 15.1
2,461
9.4
4.1
279
11.1
6.5
0.3
0.8
One PSZ, subzone4
15.639 15.1
2,067
7.9
3.4
254
10.1
5.9
0.3
0.8
One PVAVS
17.413 16.8
-78
-0.3
-0.1
124
4.9
2.9
~0
2.9
Two PTAC units
8.152 7.9
1,944
7.4
3.2
228
9.1
5.3
0.2
0
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
The baseline case at 20 cfm/person (from Table 3), repeated here for comparison.
The front zone is the control zone; the rear zone is the subzone.
The rear zone is the control zone; the front zone is the subzone.
-------
considered are:
1) a single PSZ unit conditioning the entire 4,000 ft2 office, with the entire space
being considered as a single zone. The required size of this single PSZ unit --
approximately 10 tons of cooling capacity - is well within the range of
commercially available units.
2} a single PSZ unit conditioning the entire space, but with the front (more
shaded) 2,000 ft2 zone being used as the control zone (determining compressor
operation and hence supply air temperature), and with the rear (more exposed)
zone being a variable air volume (VAV) subzone. Because the rear subzone is
no longer able to control the temperature of the supply air entering the sub-
zone, subzone space temperature is now controlled instead by modulating the
flow rate of the air entering the subzone.
3) a single PSZ unit conditioning the entire space, as in 2) above, except with the
rear 2,000 ft2 zone being used as the control zone, and the front zone being a
VAV subzone.
4) a single packaged variable air volume system (PVAVS) serving the two 2,000
ft2 zones. PVAVS would be a commercially reasonable choice for this applica-
tion, although perhaps less common than PSZ. Consistent with typical design
of packaged VAV systems, it is assumed that the PVAVS supplies an
essentially constant supply air temperature (which is as close to the minimum
supply temperature, 55°F, as the on/off compressor can deliver); and that zone
temperatures are controlled solely by modulating supply air flow rate to each
zone.
5) two through-the-wall packaged terminal air conditioner (PTAC) units, one
conditioning each zone. The roughly 5-ton units required for this application
are somewhat larger than PTAC units commonly available commercially; PTAC
units would not usually be considered for an office such as that being studied
here. PTAC units are being considered in this study solely for illustration.
Overall, Table 16 shows that the greatest reductions in energy consumption
occur with the PVAVS. In fact, the table indicates that a PVAVS could provide 20
cfm OA/person while consuming 78 kWh/year less than the baseline dual PSZ units
providing only 5 cfm/person. One could increase the ventilation rate and
simultaneously save energy if one installed the PVAVS instead of the two PSZ units.
As discussed further later, this effect results because the PVAVS moves much less
air and thus consumes less energy in operating the central air handling fan.
5-36
-------
Table 16 also suggests that a pair of PTAC units would require less capacity,
and reduce energy consumption and costs, relative to two PSZ units, despite the fact
that the PTAC units are assumed to have a poorer cooling efficiency. In part, this
result occurs because the PTAC units are not conditioning the overhead plenum space
to the extent that the PSZ units are; the return air to the PTAC units flows directly
from the offices, rather than through the plenum. Also, the air handler fan power for
the PTAC fans (and the air temperature rise through the fans) are only 12% of that
for the PSZ units according to the DOE 2 default values, because the PTAC fans do
not have to move the air through either supply or return ductwork. However, these
very factors that are causing PTACs to appear advantageous in these calculations
would also cause PTACs to inadequately distribute cooling and ventilating airthrough-
out the two 2,000 ft2 zones. Thus, one would not likely select individual PTAC units
for this application, despite the computed energy savings.
5.12.1 The Effect of Alternative HVAC Systems at 5 cfm/person
Single PSZ system conditioning entire 4.000 ft2 as one zone
Capacity. As shown in Table 16, the cooling capacity of the single PSZ unit
at 5 cfm/person is calculated to be 3.060 kBtu/h less than the combined capacities
of the two individual units in the baseline case at that ventilation rate. This result
occurs because the single unit is being sized for the peak load of the total space
(which occurs at 5 pm on October 11), rather than for the peak load of the front
office plus the peak load of the rear (which occur at 5 pm on June 14 and at 4 pm on
January 25, respectively).
DOE-2 computes cooling capacities by first calculating the maximum supply air
flows required to a given space (at the minimum supply temperature) in order to
handle the peak load for that space, determined in the LOADS part of the program.
(With PSZ systems, of course, this maximum flow becomes the constant volume that
is provided whenever the air handler is operating.) The required system cooling
capacity is then computed by multiplying this maximum flow times the required
reduction in the enthalpy of the air entering the cooling coils during the hour that this
peak load occurs, in order to reduce the air leaving the coils to the minimum supply
temperature. Thus, changes either in the flow rate, or in the entering air enthalpy at
peak load, create the differences in the calculated capacities.
Because the peak loads for the two PSZ systems in the baseline case occur at
different times than the peak for the single-PSZ case, the flows and the entering
enthalpies are both different in the two cases. In the baseline case, the combined
maximum air handier design flow for the front plus rear units is 3,810 cfm; by
comparison, the design flow for the air handler in the single-PSZ case is 3,689 cfm,
about 3% lower. Presumably, the entering enthalpies during the peak-load hours also
vary somewhat between the two cases. If one specifies a design flow of 3,810 cfm
for the single-PSZ case, so that the two cases now have the same flow rates (but
different peak-hour entering enthalpies), the computed cooling capacities are similar.
5-37
-------
Thus, the difference in design air flow - and not the difference in peak-hour entering
enthalpies - is primarily responsible for the difference in the computed capacities
between the two cases. This conclusion is also apparent from the fact that the
percentage reduction in cooling capacity shown in Table 16 for the single-PSZ case
(3%) is essentially the same as the percentage reduction in total flow for this case.
Not surprisingly, this reduction in total capacity in the single-PSZ case results
in a tiny increase in the percentage of hours undercooled (0.5% vs. 0.4%).
Energy consumption and cost. The 377 kW.n reduction in annual energy
consumption in the single-PSZ case, and the $38 reduction in annual energy cost, also
result almost entirely from the reduction in air handler design flow. As discussed
previously, electrical power to the air handling fan is a significant contributor to
system electric consumption. If one specifies that fan for the single-PSZ unit have the
same design flow (3,810 cfm) as the combined flows of the baseline fans, the single-
PSZ case would consume only 37 kWh/yr less than the baseline, and the annual
HVAC energy costs of the two cases would be exactly the same ($2,510).
Hours at elevated temperature or RH. Although treating the entire 4,000 ft2 as
a single zone in this manner appears on paper to provide about the same performance
at a slightly lower cost, there would be a comfort penalty in practice. Occupants in
the rear (south) office would likely be too warm, and/or occupants in the front (north)
office would likely be too cool.
Single PSZ system with 2,000 ft2 control zone and 2.000 ft2 subzone
Capacity. As shown in Table 16, the calculated cooling capacities for the cases
of a single PSZ unit with a control zone and a subzone are exactly the same as that
for the baseline case with two PSZ units. This holds true regardless of which of the
two zones is the control zone. This result would be expected. As in the baseline
case, these control zone/subzone cases compute design capacity based on the peak
load in the front zone plus the peak load in the rear (unlike the case of the single
4,000 ft2 zone discussed above). As a result, the total design flows in the control
zone/subzone cases are identical to that for the baseline (3,810 cfm); the dates and
hours for which the peak loads are computed are the same as for the baseline, so that
the enthalpies of the air entering the cooling coils during the peak hours are the same;
and, consequently, the design capacities are identical.
Energy consumption and cost. Energy consumptions and costs are moderately
lower for the two control zone/ subzone cases, relative to the baseline. Diagnosis
shows that 65 to 75% of this reduction is caused by reduced power consumption by
the fan, and that almost all of the remainder results from reduced power consumption
by the cooling coils.
The reduced fan power consumption occurs because one of the two zones (the
subzone) is now being operated in a VAV mode. Thus, during most of the operating
hours, flows to that subzone will be reduced relative to what they would be in the
5-38
-------
baseline, constant-volume case. Because the fan is moving less air, the power
required to operate the fan decreases by about 500 to 800 kWh/year, depending upon
which of the two zones is used as the control zone.
The reduced cooling coil power consumption results from a variety of reasons.
One reason is that - since fan heat dissipation is assumed to increase the temperature
of the circulating air by 1.8 F° -- the reduced flow through the fan results in less heat
being introduced into the air stream from this source. Another reason is that the
average temperature in one or both zones will sometimes be a couple tenths of a
degree higher in the control zone/subzone cases, relative to the baseline, so that
sometimes slightly less cooling is being provided. In addition, the efficiency of the
cooling coils varies slightly when the coils are operated at different fractions of full
load - and the fraction of full load operation during a given hour will vary somewhat
between the baseline case and the control zone/subzone cases.
Table 16 indicates that energy consumption and costs are slightly lower when
the rear zone is the control zone, compared to when the front zone is the control
zone. For example, consumption is 1,105 kWh/year less than the baseline when the
rear zone controls, and only 761 kWh/year less when the front zone controls, a
difference of 344 kWh/year. Although this difference is small, it is of interest to
understand why it occurs.
The control zone determines what the supply air temperature will be. Regard-
less of which zone controls, the supply air temperature determined by the control zone
will sometimes be lower than that required for the other zone (the subzone). When
the supply temperature is too low, flows to the subzone are reduced, thus reducing
fan power consumption. On the other hand, regardless of which zone controls,
sometimes the supply air temperature selected by the control zone will be higher than
that required for the subzone. When supply temperature is too high, flows to the
subzone rise to their design maximum, but the subzone is not cooled to the extent
that it would have been had it been able to control its own supply temperature. When
this occurs, energy consumption by the cooling coils is reduced relative to the
baseline, because the subzone is no longer receiving the amount of cooling that it did
with the baseline two-PSZ system.
Thus, when either one of the zones is treated as a subzone, there will some-
times be an energy savings due to reduced fan power consumption, and sometimes
a savings due to reduced cooling coil consumption. Which of the two zones results
in the greatest savings when considered as the subzone depends upon the mix and
magnitude of the savings from these two sources. As it turns out, in this case,
treating the front (north) zone as the subzone — and the rear (south) zone as the
control zone -- results in somewhat greater savings. Of the 344 kWh difference in
energy consumption between these two cases, almost the entire savings results
because the rear-control case requires less fan power. That is, the south-facing rear
5-39
-------
zone tends more often to demand cooler supply air than is required by the north-facing
front zone, so that front zone flows are reduced to a greater degree. This result is
intuitively reasonable.
Of the 344 kWh/year additional savings between the rear-control-zone and
front-control-zone cases, only 10 kWh/year results due to reduced cooling coil (and
heating coil) energy consumption. Both the rear-control-zone and the front-control-
zone cases result in a cooling coil energy savings of about 250 kWh/year relative to
the baseline. For some hours on some days, the front-control-zone case requires less
cooling energy than the rear-control-zone case; for other hours, the situation is
reversed. These hours apparently offset each other, so that the net effect is that both
control-zone cases reduce cooling energy consumption to the same extent relative to
the baseline.
Hours at elevated temperature or RH. For both of the control-zone/subzone
cases, the hours undercooled (and some-times the hours at elevated RH) are lower,
by a tiny amount, compared to the baseline case. This occurs despite the fact that,
during occasional hours when the supply air temperature is too high, the subzone is
receiving less cooling than it does in the baseline case, and is hence a couple tenths
of a degree warmer. The reason for the slight net reduction in hours undercooled
must be that the one large PSZ system in these cases, with about 10 tons of cooling
capacity, sometimes has the reserve capacity necessary to handle peak loads in one
of the zones - peak loads that could not have been handled if that zone were
conditioned by a dedicated 5-ton unit.
Taking one summer Monday morning (July 8) to provide an example, in the
baseline case with two 5-ton PSZ units, the front zone is undercooled for the first four
hours after the system begins operating. But when there is a single PSZ system with
the front system serving as the control zone, the front zone is undercooled during only
one hour. During the first two hours on that Monday morning, the supply air tempera-
ture to the front zone averages 1.8 F° cooler in the front-control-zone/rear-subzone
case than in the baseline case, because the 10-ton coils have sufficient capacity to
provide the additional cooling. On this same summer Monday morning, the number
of undercooled hours in the rear zone is also reduced, from two hours (in the baseline
case) to one hour (in the front-control-zone case).
Single PVA VS system conditioning both zones
With the two constant-volume PSZ units assumed in the baseline case, the flow
to each zone remains constant; zone temperature is controlled by modulating the
average supply air temperature. With the PVAVS, it is the supply air temperature that
remains (relatively) constant (at about the minimum supply temperature of 55°F); in
this case, zone temperature is controlled primarily by modulating the flow to each
zone. In larger and more complex VAV systems, it is possible to modulate supply air
temperature as well as flow rate; but that ability is not commonly included (or needed)
in simple PVAVS.
5-40
-------
Capacity, Referring to Table 16, the design cooling capacity for the two-zone
packaged VAV system is 1.479 kBtu/h (1.4%) greater than the sum of the capacities
for the two PSZ units in the baseline case. The sole reason for the higher PVAVS
design capacity is that the default fan that DOE-2 assumes for PVAVS increases
supply air temperature by 2.1 F° (due to dissipation of fan heat), whereas the
temperature rise assumed for the PSZ fan is only 1.8 F°. The added cooling capacity
computed for PVAVS is solely to remove this incremental additional fan heat during
peak load. If one specifies a PVAVS fan that raises the supply air temperature by
1.8 F°, identical to the PSZ fan, the computed design cooling capacity becomes
identical for PVAVS and PSZ.
This result would be expected. All of the parameters used by DOE-2 in
calculating design cooling capacity are the same for both HVAC systems; i.e., system
flows at peak load, and the date and hour of peak loads in each zone (and hence the
enthalpies of the air entering the coils, as determined by weather, internal loads, and
ventilation rate). The heat added by the air handling fan is the only variable used in
the calculation that varies.
Energy consumption and cost. As shown in Table 16, the PVAVS results in a
12% reduction in HVAC energy consumption compared to the baseline two-PSZ
system, the greatest reduction of any of the HVAC alternatives considered here.
PVAVS also results in a 7% reduction in HVAC energy cost, again the greatest
reduction, with the exception of the anomalous PTAC result. The only other
parameters which enable an energy savings this great at 5 cfm/person are:
dramatically reducing lighting and equipment power consumption (19% reduction in
energy consumption, see Section 5.4); elimination of (or achieving infinite thermal
resistance in) the windows, walls, and roof (20% reduction, Section 5.10); and the
use of cold-air distribution (11% reduction, Section 5.14). Of these parameters
offering the greatest potential, switching from a PSZ to a PVAVS is the most practical
and easily achievable.
Diagnosis of these results indicates that, of the 3,149 kWh/year energy savings
shown in Table 16 for PVAVS compared to the baseline two-PSZ system, 87% of this
savings results from reduction in power consumption by the air handling fan. Almost
all of the remaining 13% results from reduced energy consumption by the cooling
coils.
The substantial reduction in fan power consumption results because, with the
VAV system, the fan is almost always moving less air than in the baseline (constant-
volume) case. Only at peak loads do variable flows reach the maximum value (3,810
cfm) that the constant-volume baseline fan is always moving whenever it operates.
The power consumption of a given fan decreases with the cube of the volumetric flow
rate through the fan (ASHRAE, 1992c). As shown in Table 4, the air handler repre-
sents 28% of the HVAC energy requirements in the baseline case. Hence, a major
reduction in fan power consumption through reductions in flows would be expected
to have an important impact on total system consumption. Switching from the
baseline PSZ configuration to the PVAVS reduces fan power consumption by 2,742
kWh/year, or 37% (from 7,328 to 4,586 kWh/year).
5-41
-------
The remaining 13% of the 3,149 kWh/year total reduction in HVAC energy
consumption -- i.e., 407 kWh/year - results primarily from a modest reduction in
cooling coil energy requirements for the PVAVS relative to the two PSZ units. This
reduction in cooling coil energy occurs because the reduced PVAVS flows result in
less heat being introduced into the air through dissipation of heat from the air handler,
causing a net reduction in the load on the coils.
As indicated previously, heat dissipation from the PVAVS supply fan is assumed
in the computations to increase air temperature by 2,1 F°, compared to a 1.8 F°
increase for the PSZ fans. The total heat added to the air stream by the fan is
proportional to the volume of air flow multiplied by this increase in temperature across
the fan. During hours when PVAVS flows are between about 85% and 100% of the
combined PSZ flows in the baseline system, the total heat added by the PVAVS fan -
and hence the contribution of the fan to cooling coil energy consumption - will be
greater than the contribution from the PSZ fans. But when the PVAVS flows are less
than 85% of the PSZ flows, the contribution of the PVAVS fan will be less than that
of the PSZ fans. The net effect over the course of a year is for the PVAVS fan to
contribute /ess heat to the air stream, creating the net reduction in cooling coil energy
consumption observed here.
Hours at elevated temperature or RH. Table 16 shows that, with the PVAVS,
the total number of hours undercooled each year is reduced to essentially zero. By
comparison, with the baseline two-PSZ configuration, 0.4% of all hours are
undercooled (about 30 hours per year in each zone).
This difference results because the DOE-2 simulation models the PSZ and
PVAVS control systems in different ways. Upon startup on a warm summer morning,
the PVAVS immediately begins operating at peak capacity, providing supply air near
the minimum supply temperature (55 °F); the supply air temperature remains near this
MIN-SUPPLY-T all day, with supply flows varying as necessary to accommodate the
varying load. For example, on Monday morning, July 8, the PVAVS operates at
maximum capacity for the first 3 hours after startup, and the space temperatures in
each zone - which were above 84 °F before startup -- average below 76 °F during the
first hour after startup. This is within the 2 F° throttling range of the cooling setpoint
(75 °F), and there are thus no undercooled hours on that day with the PVAVS.
By comparison, the two PSZ systems start up on that same Monday morning
(July 8) operating at only 83 to 91 % of maximum capacity over the first two hours,
even though the zones are telling the system that they are undercooled. As a result,
the 84 °F pre-startup zone temperatures are reduced only to an average of 77.7 °F
during the first two hours of July 8, more than 2 F° above the 75 °F setpoint. Each
zone is thus considered to be undercooled during those hours with the baseline PSZ
configuration.
5-42
-------
Table 16 also indicates that the percentage of occupied hours with relative
humidities above 60% are about the same with the PVAVS and the baseline dual PSZ
configuration. Although these percentages of hours at elevated RH are almost
identical (1.5% vs. 1.2%), it is of interest to understand how these similar
percentages result from two different control systems.
The hours at elevated indoor RH tend to occur during winter mornings in Miami,
when the outdoor RH can be high (over 90%), but when the indoor temperature is low
enough (about 70 °F) such that the cooling coils are operating at greatly reduced
capacity. Sometimes in the winter, the cooling coils do not operate at all during the
first few hours. (On summer mornings, the coils are operating close to maximum
capacity beginning at startup, so that indoor RH levels are consistently below 60%.)
Take Monday morning, January 14, as an example. The office spaces are at
70 to 71 °F just before startup, well below the cooling setpoint of 75 °F that should
trigger cooling. But anticipating that occupation will cause the temperature to rise,
the PVAVS control equipment starts the system operating immediately in the cooling
mode, an approach that maintains the offices well below 75 °F until early afternoon.
Typical of PVAVS operation, the system provides supply air near to the minimum
supply temperature beginning at startup; but because of the reduced cooling load
(technically, the offices are over-cooled during the morning), the PVAVS supply air is
provided at the minimum variable-volume flow throughout the morning hours.
The low PVAVS cooling coil temperature condenses a meaningful fraction of
the moisture, causing a decrease in the indoor RH from its pre-startup value of over
80%. However, because the flows are at their minimum, this moisture removal is not
sufficient to reduce the office RH below 60%. The RH ranges between 70 and 72%
for the first four operating hours, and does not drop below the 60% target level until
mid-afternoon on this particular Monday. The first nine occupied hours are above
60% RH on January 14 with the PVAVS, and the RH range during this 9-hour period
is 60 to 72%.
By comparison, due to a different control approach, the baseline PSZ systems
have their cooling coils entirely off during the first two to three hours on this Monday,
allowing office temperatures to rise to about 74 °F before cooling begins. During
those first occupied hours with the cooling coils off, the office RH rises significantly
as a result of: a) latent heat release by the occupants; and b) mechanical supply of
outdoor air by the activated HVAC system, containing moisture equivalent to >.80%
RH at office temperature. During these initial hours with the cooling coils inactive, the
office RH rises to 90 to 100%, substantially higher than the maximum values with the
PVAVS.
But once the cooling coils for the two PSZ units do come on, their combined
hourly latent cooling rate soon surpasses that of the PVAVS. Even though the PSZ
5-43
-------
coil surface temperature is not as low as in the PVAVS during the initial hours, the
fact that the PSZ units are treating a much greater quantity of air results in
significantly greater total moisture removal. As a result, with the baseline PSZ
system, the office RH drops below 60% within 6 hours in the front office, and within
4 hours in the rear office, compared to 9 hours with the PVAVS.
In summary, the PSZ units result in elevated office RH for fewer hours, as
indicated in Table 16. However, during those fewer hours, the PSZ RH levels can
reach much greater values than those seen with the PVAVS, since the PSZ coils can
be off completely.
One PTAC unit serving each zone
As shown in Table 16 -- in comparison with the baseline dual-PSZ configuration
at 5 cfm/person - the use of PTAC units would appear to decrease the cooling
capacity requirement by 6.5% (i.e., by 6.718 kBtu/h), the annual HVAC energy use
by 10% (by 2,719 kWh), and the HVAC energy cost by 9% (by $232). But as
indicated previously, this result is an artifact. An attempt in practice to cool a 2,000
ft2 zone with an unducted, through-the-wall PTAC unit would result in the occupants
near the unit being uncomfortably cool, and those remote from the unit being
uncomfortably warm.
Most of the apparent savings with the PTAC units results because the default
PTAC air handler is a low-power, low-static-pressure fan, since this fan does not have
to move air through either supply or return ducting. This default PTAC fan provides
a static pressure rise of only 0.3 in. WG (compared to 3.0 in. WG for the default PSZ
air handler), consumes only 0.000070 kW per cfm of air moved (12% of the value for
PSZ, 0.000587 kW/cfm), and raises the air temperature by only 0.2 F° (compared to
1.8 FD for the PSZ fan).
The PTAC computations were repeated with the power consumption and the
temperature rise for the PTAC fan set equal to the values for the PSZ fan {0.000587
kW/cfm, 1.8 F°), rather than being allowed to default to the low PTAC values. With
these fan values:
a) the required PTAC cooling capacity rose almost to the value of the baseline PSZ
system. That is, the 6.718 kBtu/h decrease shown in Table 16 dropped to
almost zero.
b) the annual HVAC energy consumption by the PTAC system rose dramatically.
The 2,719 kWh decrease shown in Table 16 (relative to the PSZ baseline)
became a 6,150 kWh increase over baseline consumption, since the PTAC
units (having a default electric input ratio of 0.438) are inherently less efficient
than the PSZ units (with an EIR set equal to 0.341).
5-44
-------
c) correspondingly, the annual HVAC energy cost for the PTAC system rose
substantially, from a $232 savings compared to PSZ costs, to a $515 increase
over PSZ.
These figures indicate that the very low default performance (and hence very
low power consumption) of the PTAC fans is largely responsible for the apparent
PTAC savings shown in Table 16. It is this low fan power consumption {and the
resulting lack of static pressure to provide effective air distribution) that makes the
PTAC system generally not a good choice for this application.
In addition to low PTAC fan power consumption, another, secondary
contributor to the reduced cooling load and power requirements of PTAC systems
might be that the PTAC system is not conditioning the overhead plenum to the same
extent.
With the baseline PSZ configuration, the 4-ft-high space above the offices is
used as the return air plenum. As such, it is being swept with conditioned office air
whenever the HVAC fan is operating. Consequently, it is always within a couple
degrees of the office temperature - and sometimes within a couple tenths of a
degree -- during fan operation. By comparison, with the PTAC units -- which draw the
return air directly from the offices - this overhead space is simply dead space which
exchanges heat with the offices only via conduction through the uninsulated ceiling,
but with no air exchange. As a result, with the PTAC units, the unused plenum can
average perhaps 3 F° warmer on a summer day than it does when it serves as the PSZ
plenum. This increased PTAC plenum temperature represents heat that does not have
to be removed by the PTAC units, and hence, a potential savings in PTAC cooling
capacity and power requirements. On the other hand, the heat that does conduct
through the ceiling into the offices is removed less efficiently by the PTAC units;
plenum heat swept directly into the PSZ cooling coils by return air is removed more
efficiently.
In Section 5.13, results are reported for the case where return air to the PSZ
system is accomplished using (insulated) ducts, rather than using the overhead
plenum as in the baseline PSZ case. Those results indicate that net effect of avoiding
the partial conditioning of the plenum — but of then having to remove some of that
plenum heat after it has conducted into the office space -- seems to be a small
savings in energy consumption. Extrapolating that PSZ result to the PTAC case, it is
likely that the fact that the PTAC units are avoiding partial conditioning of the
overhead space is only a minor part of the reason why the PTAC units require less
power than the baseline PSZ case. The default low power requirements of the PTAC
fans are the predominant explanation.
Hours at elevated temperature or RH. Table 16 indicates that the PTAC
systems result in slightly fewer hours undercooled than do the PSZ systems (0.2%
5-45
-------
compared to 0.4%). This effect results because the PTAC units are calculated to
provide cooler supply air during the first couple hours after startup on warm summer
mornings. For example, during the first two hours after startup on Monday, July 8,
the computed temperature of the air leaving the PTAC cooling coils is 1 to 5 F° cooler
than the air leaving the PSZ coils. The PTAC units are operating at a greater
percentage of full capacity during those first two hours, compared to the PSZ units.
Consequently, while the PSZ systems result in the zones being undercooled for the
first four hours on July 8, the PTAC units result in undercooling for only one or two
of those first four hours (depending on the zone}. This effect might result from
differences in the manner by which the PTAC vs. PSZ control systems are simulated
by the program, and/or from the fact that the less powerful PTAC fans add less heat
to the air stream.
Table 16 also indicates that, according to the calculations, the PTAC systems
reduce the number of occupied hours above 60% RH to about zero. The PTAC coils
do tend to operate at a lower temperature than the PSZ coils under some conditions,
consistent with this prediction that they might remove more moisture. However,
diagnosis indicates -- if PTAC units do result in fewer elevated RH hours than do PSZ
units -- the difference is not as great as that suggested in Table 16, and PTAC units
do not in fact reduce elevated-RH hours to zero. The apparent PTAC vs. PSZ RH
effects in Table 16 are believed to result from errors in the DOE-2 simulation and
reporting.
To illustrate, the morning of Monday, January 14, is taken as an example.
During the first three to four hours after startup on that day, the cooling coils are not
operating in the front zone with either PTAC or PSZ; the system is either in the
heating mode, or is just ventilating with neither the heating nor cooling coils operating.
Thus, neither system is removing moisture from the air during those hours. Yet, the
RH in the front zone rises to 96 to 100% with the PSZ system (reaching a high of
0.0210 lb moisture per lb dry air) — whereas, with the PTAC system, RH is calculated
to hold at 77 to 84% (its maximum being 0.0144 lb/lb, lower than the moisture
content of the outdoor air, despite indoor latent sources). Since the DOE-2 program
cannot address the moisture capacitance of indoor materials, moisture capacitance is
not contributing to these apparently anomalous results. Accordingly, it seems clear
that the program is underestimating the moisture levels with the PTAC system, and/or
is overestimating moisture levels with the PSZ system.
It is also noted that there must be an error in DOE-2's relative humidity scatter
plot report for the PTAC system (the "SS-N" report), on which Table 16 is based. The
SS-N report suggests that, with the PTAC system, the RH will be above 60% for only
one occupied hour (in the front zone) during the entire year. But hourly reports for
January 14 with the PTAC system show RH's above 60% (actually, above 70%) for
four occupied hours in the front zone, and for one hour in the rear zone, on that day
alone. [By comparison, the PSZ system is computed to be above 60% (and above
5-46
-------
70%) for five hours in the front zone on that day, and for three hours in the rear
zone,]
5.12.2 The Effect of Alternative HVAC Systems on Increased Ventilation Rates
As discussed in Section 4, and as repeated in Table 16, increasing the OA
ventilation rate from 5 to 20 cfm/person with the baseline two-PSZ/two-zone system
increases the required cooling capacity by 15.635 kBtu/h, the energy consumption by
3,245 kWh/yr, and the energy cost by $325/yr.
Alternative PSZ configurations
For the three alternative HVAC systems in Table 16 that are PSZ-based -- the
one-PSZ/one-zone configuration and the two one-PSZ/zone/sub-zone configurations -
increasing the ventilation rate to 20 cfm/person increases capacity requirements,
energy consumption, and costs by essentially identical increments above each
configuration's respective values at 5 cfm/person. This is consistent with the
expectation that such similar HVAC configurations will respond to the increased OA
sensible and latent loads in the same manner.
For example, at 20 cfm/person, the one-PSZ/one-zone configuration requires
an increase in capacity of 15.572 kBtu/h relative to this same configuration at 5
cfm/person, essentially identical to the 15.635 kBtu/h OA-induced increase for the
baseline system. But the capacity of this one-PSZ/one-zone configuration at 5 cfm/
person was 3.060 kBtu/h less than that required for the baseline two-PSZ/two-zone
system at 5 cfm/person, for the reasons discussed in Section 5.12.1. As a result, the
capacity for the one-PSZ/one-zone configuration at 20 cfm/person is increased by only
(15.572 - 3.060 = ) 12.512 kBtu/h relative to the baseline two-PSZ/two-zone system
at 5 cfm/person, the figure shown in Table 16, Similarly, the increase in energy
consumption for the one-PSZ/one-zone configuration at 20 vs. 5 cfm/person is 3,235
kWh/yr (essentially identical to the baseline's 3,245 kWh/yr), and the increase in
energy cost is $323/yr (identical to the baseline's $325/yr).
Single PVA VS conditioning both zones
Capacity. For the PVAVS, the required increase in capacity for the increase
from 5 to 20 cfm/person is 15.934 kBtu/h, slightly greater than the OA-induced
increase for the baseline two-PSZ system. (As a result, the required PVAVS capacity
at 20 cfm/person is 15.934 + 1.479 = 17.413 kBtu/h greater than that for the PSZ
system at 5 cfm/person, as shown in Table 16.) This minor difference results from
the greater temperature rise in the air stream assumed across the default PVAVS air
handler (2.1 vs. 1.8 F° for PSZ), the same issue discussed for PVAVS in Section
5.12.1. Reducing the assumed PVAVS fan temperature rise to 1.8 F° results in a
computed capacity increase for the PVAVS at 20 vs. 5 cfm/person of 15.614 kBtu/h,
essentially identical to the baseline PSZ system.
5-47
-------
Energy consumption and cost. The OA-induced increase in energy consumption
for the PVAVS is only 3,070 kWh/yr, somewhat lower than the 3,245 kWh/yr
increase for the baseline case. This 175 kWh/yr difference results almost entirely
because cooling coil energy consumption (i.e., consumption by the compressor and
the condenser fan) increases to a lesser degree with increasing OA in the case of the
PVAVS, compared to the PSZ system. Changes in consumption by the air handler
and the heating element are not significant contributors to this difference.
PVAVS cooling coil energy consumption is impacted less by an increase from
5 to 20 cfm/person because the hourly EIRs for the PVAVS are impacted slightly
differently by the OA increase than the EIRs for the PSZ system are impacted. The
EIR during any given hour for a given system is determined, in part, by the part-load
ratio (i.e., the fraction of full capacity) at which the system is operating. The default
equation relating EIR to part-load ratio is the same for the PVAVS and the PSZ
system. However, due to the inherent characteristics of PVAVS vs. PSZ, an increase
in OA from 5 to 20 cfm/person impacts the hourly part-load ratios (and hence the
EIRs) slightly differently. Therefore - even though an increase from 5 to 20 cfnn/
person should increase the load to the same extent for both systems - the differences
in EIRs results in a different number of kilowatt-hours being required to address this
load. This is the cause of the net 175 kWh/yr difference indicated in the preceding
paragraph.
Due to the lesser increase in PVAVS energy consumption resulting from the
increased OA — and since PVAVS at 5 cfm/person was consuming 3,149 kWh/yr less
than PSZ at 5 cfm/person — Table 16 estimates that a PVAVS could operate at 20
cfm/person in this building, while consuming 78 fewer kWh/yr than the baseline dual
PSZ system at 5 cfm/person. Thus, switching from a dual PSZ system to a PVAVS
in designing a new building could theoretically enable an increase in OA ventilation
rate without an increase in energy consumption.
Increasing the PVAVS from 5 to 20 cfm/person increases energy cost by $310
per year. This is slightly less than the OA-induced increase in PSZ energy costs
($325), consistent with the fact that, with PVAVS, the OA-induced increase in energy
consumption is 175 kWh/yr less, as discussed above. Thus, the PVAVS is estimated
to be able to deliver 20 cfm/person at an annual cost only $124 greater than that of
the baseline PSZ system delivering only 5 cfm/person, as indicated in Table 16.
Hours at elevated temperature or RH. Table 16 shows that, at 20 cfm/person,
the PVAVS is estimated to result in essentially no occupied hours being undercooled,
just as at 5 cfm/person. The reasons why PVAVS gives no undercooled hours,
whereas the PSZ-based systems do result in some undercooling, are the same as
those given previously (Section 5.12.1). Due to its control system, the PVAVS
operates closer to full capacity than the PSZ units during startup on summer weekday
mornings, when the PSZ undercooled hours occur; correspondingly, PVAVS delivers
5-48
-------
more power to its cooling coils (i.e., to the compressor and the condenser fan) during
those hours.
Table 16 indicates that, at 20 cfm/person, the occupied hours at RH levels
above 60% approximately doubles with the PVAVS (to 2.9% of all occupied hours,
or 94 hours per year), compared to the 5 cfm/person case. By comparison, the PSZ-
based systems all show a tiny decrease in elevated-RH hours with the increase from
5 to 20 cfm/person. On winter weekday mornings, when the elevated-RH hours
occur, the increase in the PVAVS latent cooling resulting from the increased OA does
not compensate for the increase in latent load with the increased OA. By comparison,
with the PSZ systems, the increased latent cooling does compensate for the increase
in load. This effect results from the differences between the PVAVS and PSZ control
systems.
Consider Monday, January 14 - the same day addressed previously for the 5
cfm/person case. In terms of flows and temperatures, the PVAVS performs the same
at 20 cfm/person as at 5 cfm/person on that day. The hourly flows to each zone --
and the hourly temperatures of each zone, the supply air, and the cooling coil
surface - are essentially identical, regardless of the OA rate. This similarity results
because the externally induced and occupant-induced thermal loads on the zones are
independent of OA rate, and because of the characteristics of the PVAVS control
system (which modulates supply air flow at a fairly steady temperature near the
minimum supply temperature). Because the mixed air entering the coils has a higher
latent load with the increased intake of humid outdoor air at 20 cfm/person, the
identical total flows and coil temperatures result in an increase in the amount of
moisture removed relative to the 5 cfm/person case. Increasing the PVAVS OA rate
from 5 to 20 cfm/person increases latent cooling by over 50% (from 82,600 to
125,000 Btu/day) on January 14. Latent cooling is increased from 19% of total
cooling at 5 cfm/person, to 25% of total cooling at 20 cfm/person.
But this increase in PVAVS latent cooling is insufficient to compensate for the
increased latent load. At 20 cfm/person, the hourly moisture content of the supply
air leaving the PVAVS coils (lb moisture per lb dry air) is 5 to 15% higher than the
content at 5 cfm/person. The RH in the office space during occupied hours reaches
higher maximum levels during the initial hours after startup (78% at 20 cfm/person,
compared to 72% at 5 cfm/person). And the office RH remains above 60% during
all 13 occupied hours on January 14, whereas at 5 cfm/person, the RH was above
60% only during the first 9 hours with the PVAVS.
The PSZ-based systems react differently to the increase in OA. As with the
PSZ system at 5 cfm/person, the PSZ cooling coils remain off in each zone for the
first two to four hours after startup at 20 cfm/person, allowing the humidity to
increase as the result of occupant latent heat. But due to the increased in-flow of
outdoor air at 20 cfm/person -- outdoor air whose moisture content corresponds to
5-49
-------
about 80% RH at office temperature - the maximum humidity levels reach only about
90% RH at 20 cfm/person, compared to 90 to 100% at 5 cfm/person.
Like the PVAVS, the total PSZ flows remain unchanged as OA is increased, as
do the office temperatures and the supply air temperatures. But since the PSZ control
system is designed to modulate supply air temperature at constant flow, the PSZ
system -- unlike the PVAVS -- reduces coil temperature (sometimes by several degrees
or more) in response to the greater sensible and latent loads in the mixed air entering
the coils when OA is increased.
Thus -- once the PSZ coils do come on, after remaining off during the first few
hours -- the lower coil temperatures at 20 cfm/person result in greater moisture
removal than occurred at 5 cfm/person. The moisture content (lb of moisture per lb
of dry air) of the supply air leaving the PSZ cooling coils during the first one to two
hours of operation is 5 to 15% lower at 20 cfm/person than at 5 cfm/person. (By
comparison, with PVAVS, the moisture content was 5 to 15% higher at 20 than at
5 cfm/person.) As a result, the RH in the office space drops slightly more rapidly at
20 than at 5 cfm/person during the first hour after the coils come on. However - on
January 14, at least - the offices remain above 60% RH for exactly the same number
of hours at 20 cfm/person with the PSZ system as they did at 5 cfm/person (6 hours
in the front, 4 hours in the rear). This is consistent with the results shown in Table
16: The percentage of occupied hours above 60% RH with the PSZ system at 20
cfm/person is not significantly different than the percentage with the PSZ system at
5 cfm/person, and the slight change that does occur is in the direction of reduced
elevated-RH hours at the higher OA rate.
Further comparing the effect of increased OA on the PVAVS vs. the PSZ
system, with the baseline two-PSZ configuration, increasing the OA rate from 5 to 20
cfm/person increases the latent cooling by 87% on January 14, from a combined total
of 78,000 to 146,000 Btu/day. By comparison, as indicated previously, the increased
OA increased PVAVS latent cooling by only 50%. At 20 cfm/person, the dual PSZ
configuration provides greater latent cooling than the PVAVS (146,000 vs. 125,000
Btu/day). In the PSZ system, latent cooling is increased from 20% of total cooling at
5 cfm/person, to 32% of total cooling at 20 cfm/person (compared to 25% with
PVAVS). As a result, the increase in latent cooling by the PSZ system at 20
cfm/person compensates for the increase in latent load caused by the higher OA rate.
In summary, when the OA rate is increased from 5 to 20 cfm/person on
January 14, the PVAVS experiences an increase in the number of occupied hours at
RH > 60% (13 rather than 9), and an increase in the maximum RH levels (78% rather
than 72%). By comparison, the PSZ system experiences no significant change in the
number of elevated-RH hours, and a decrease in the average RH during those elevated
hours (about 78% rather than 85%).
5-50
-------
One PTAC unit serving each zone
Capacity. For an increase from 5 to 20 cfm/person, the required increase in the
total PTAC capacity is 14.870 kBtu/h, somewhat less than the 15.635 kBtu/h
increase for the baseline PSZ configuration. Diagnosis indicates that this somewhat
reduced impact of OA in PTAC capacity is due to the different approach used by the
DOE-2 software in computing capacities for zonal systems such as PTAC vs. central
systems such as PSZ (Winkelmann et al., 1993). It is not the result of differences
between the PSZ and PTAC default values for the key variables that influence the
capacity calculation. Specifically, the reduced impact of increased OA on estimated
PTAC capacity is not due to the reduced temperature gradient across the PTAC supply
fan (as discussed in Section 5.12.1), the higher PTAC coil bypass factor, or the
different standard equations expressing bypass factor, total capacity, and sensible
capacity as functions of wet- and dry-bulb temperatures. Even if the values for ail of
these parameters are set the same for PTAC as for PSZ, the impact of 5 vs. 20
cfm/person on the computed PTAC system capacity remains at about 14.7 kBtu/h.
Correspondingly, even with these "corrections", the computed PTAC capacity at 20
cfm/person remains about 1.5 kBtu/h lower than PSZ capacity at 20 cfm/person. (By
comparison, in Table 16, PTAC is 15.635-8.152 « 7.5 Btu/h lower).
The difference in computed PTAC capacities is also not due to additional heat
picked up in the plenum by the return air in the PSZ case. Except for heat released
into the plenum by the office lights (of which there is none in this case), plenum heat
does not factor into the capacity calculation.
Energy consumption and cost. As shown in Table 16, two PTAC units
operating at 20 cfm/person use less power than the baseline PSZ system at 20
cfm/person {by 3,245 - 1,944 = 1,301 kWh/yr), and have a lower energy cost (by
S325 - $228 = $97). But the increases in power consumption and cost for PTAC at
20 cfm/person vs. PTAC at 5 cfm/person are significantly greater than the increase
for PSZ with that increase in OA. Specifically, increased OA causes PTAC power
consumption to rise by 2,719 + 1,944 = 4,663 kWh/yr (compared to 3,245 kWh/yr
for PSZ), and energy cost to rise by $232 + $228 = $460/year (compared to $325
for PSZ). This result is no surprise. The PTAC unit cools less efficiently than does
PSZ (cooling EIR = 0.438 for PTAC, vs. 0.341 for PSZ), so that the incremental
increase in power usage resulting from the OA load increase is correspondingly
greater.
As discussed in Section 5.12.1, the primary reasons why PTAC at 20 cfm/
person is consumes less power than PSZ at 20 cfm/person is that the default PTAC
supply fan is assumed to consume far less power and to add far less heat to the
supply air. If the PTAC calculations are repeated assuming PSZ fan characteristics,
the inefficient PTAC system at 20 cfm/person consumes about 25% more power than
the PSZ system at 20 cfm/person, at about the same percentage increase in energy
cost.
5-51
-------
5.13 DUCTED AIR RETURNS
The baseline configuration assumes that the office air returns to the HVAC
system by way of the overhead plenum spaces, without return ducting. This common
approach results in some effective conditioning of the plenum space by the return air,
such that the temperatures in the plenums are generally within a degree or two of the
temperatures in the conditioned offices.
It is also possible to model the case where ducting is provided for the return air.
In a small office such as this, it would be common for such return ducting to be
routed through the overhead spaces, although, for the purposes of the DOE-2
modeling, it is not necessary to specify exactly where the ducting is located. The
model assumes that the ducts are perfectly insulated, i.e., that there is no heat
transfer between the return air (inside the ducting) and the space outside the ducting.
The one exception is that any energy from the office lighting that is not released into
the conditioned space is assumed to be released into the return ducting. But in this
study, all energy consumed by the lights has been assumed to be released into the
office space (LIGHT-TO-SPACE = 1); consequently, no heat would be added to the
air in the return ducting.
Table 17 shows the results when the office air is assumed to return to the
baseline PSZ units via return ducting rather than via the overhead plenum.
5.13.1 The Effect of Ducted Returns at 5 cfm/person
Capacity. As shown in Table 17, ducting the returns has no effect on the
computed cooling capacity of the PSZ system at 5 cfm/person. Ducting has no effect
on capacity because — even when the return air flows through the overhead plenum -
- the capacity calculations do not take into consideration any of the heat picked up
in the plenum (Winkelmann et al., 1993). The one exception is that the lighting
energy that is not released into the office space is added to the return air stream in
the capacity calculation -- and that exception applies regardless of whether or not
there is return ducting. Thus, the presence of perfectly insulated return air ducting
does not influence the cooling capacity calculation.
Energy consumption and cost. But as shown in Table 17, there is some small
reduction in HVAC energy consumption and costs as a result of the return ducting
(298 kWh and $35 per year). This modest savings is the net result of two offsetting
phenomena.
The first of these phenomena is that ducting the returns avoids the partial
conditioning of the plenum space by the return air. The overhead space now serves
as a 4-ft-thick layer of air insulation between the outdoors and the conditioned space,
thus reducing the load on the cooling coils. When that space is not serving as a
return plenum, and is thus not being swept by return air from the offices, the tempera-
5-52
-------
TABLE 17
Effect of Ducted Air Returns;
Increase Compared to Baseline Case with 5 cfm/person
Case
Increase in
total required
cooling capacity
In
kBtu/h As %
increase in
annual energy consumption
As % of As % of
HVAC building
In kWh energy energy
Increase in annual energy cost
As % of As % of
HVAC building
energy cost energy cost
In $
Hours/yr Occupied
under- hours/yr
cooled, RH>60%,
% %
Baseline values'. 103.558' - 26,145' 2.5101 0.4 1.2
Air return (HVAC) - — (HVAC)
via plenum, 60,1611 4,273'
OA = 5cfm/p {building) -- -- (building)
(absolute values)1
OA ventilation rate = 5 cfm/person
Air return
viaduct 0 0 -298 -1.1 -0.5 -35 -1.4 -0.8 0.5 1.2
OA ventilation rate = 20 cfm/person
Air return 15.635 15.1 3,089 11.8 5.1 311 12.4 7.3 0.6 0.8
via duct
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
ture of the overhead space can average perhaps 3 F° warmer on a summer day than
in cases where this space does serve as a return plenum. That 3 F° represents a
sensible cooling load from which the coils are being spared.
The second phenomenon is that some portion of this sensible heat in the
overhead space flows into the offices via conduction through the ceiling, which is not
insulated in the baseline building. Once the conducted heat is in the office space, it
must be removed by circulating the bulk office air through the cooling coils. The
removal of this heat from the larger mass of office air will be less efficient than it
would have been had the warm plenum air been swept directly into the cooling coils,
as it would have been had the overhead space been being used as a return plenum.
So the offsetting phenomena are: a) return air ducting reduces the amount of
overhead heat that has to be removed by the system; but b) the portion that does
have to be removed is removed less efficiently. The relatively small difference in
energy consumption and cost between the plenum- and ducted-return cases shows
that these two phenomena just about cancel each other.
The modest 298 kWh energy savings shown in Table 17 for the ducted-return
case would increase significantly if the office ceiling were insulated, so that less of
the heat in the overhead space could conduct down into the conditioned space. For
example, if it were assumed that the uninsulated office ceiling were instead insulated
with R-30 batt insulation (in the extreme), the savings from switching from plenum
return to {perfectly insulated) ducted returns at 5 cfm/person would increase to 1,388
kWh and $178 per year.
Hours at elevated temperature or RH. Table 17 shows that ducting the return
air results is a tiny increase in the percentage of hours undercooled (0.5% vs. 0.4%).
During startups on summer weekday mornings, the return air in the baseline unducted
case is warmer, due to the heat in the plenum. This causes the PSZ units to have a
slightly higher actual capacity, and to operate at a greater fraction of total capacity
(a higher part-load ratio), than is the case with the identical PSZ units in the ducted
case. Under these circumstances, without ducting, the plenum space cools rapidly,
reducing heat conduction through the ceiling from the plenum into the office space.
The net result of these phenomena is that the office temperatures are consistently
computed to be a couple tenths of a degree cooler in the unducted case, compared
to the ducted case.
This small temperature difference explains why the baseline unducted plenum
case has a tiny percentage fewer undercooled hours. For example, on Monday, July
8, the rear zone in the ducted case is undercooled during the first three hours after
startup (with the third hour being a tenth of a degree above the 77 °F limit that
defines the zone as being undercooled). But in the baseline unducted case, the rear
5-54
-------
zone is undercooled only during the first two hours (with the third hour being a couple
tenths of a degree below the 77 °F maximum, and thus not undercooled).
Table 17 indicates that, at 5 cfm/person, the percentage of occupied hours
above 60% RH remains unchanged at 1.2% as the air return is changed from
unducted to ducted. Actually, ducting the returns does have a small impact on the
office RH values, but not enough to change the percentage of hours that are above
60%.
Elevated-RH hours tend to occur after startup on cool, humid winter mornings.
When the plenum is used for the unducted return underthese circumstances, it serves
to cool the return air during the initial hours, until the outdoor temperature warms up
later in the morning. But with ducted returns - insulated from the cool plenum -- the
return air does not receive this plenum cooling. Accordingly, with the ducted returns,
slightly more power is put into the cooling coils during these initial hours, compared
to the baseline unducted case, and the humidity levels in the supply air (and in the
offices) are somewhat lower during these hours.
Take, for example, the morning of Monday, January 14. The front office is at
an RH level above 60% for the first six hours after startup, regardless of whether the
return air is ducted or unducted. This is consistent with the indication in Table 17
that the percentage of occupied hours above 60% RH remain unchanged when the
returns are ducted. However, the average RH in the front office during those first six
hours decreases from 88% in the baseline unducted case, to 82% in the ducted-return
case.
Of course, on warm summer mornings, the reverse occurs. The baseline
unducted return air now picks up (rather than loses) heat in the plenum. So under
these circumstances, it is now the unducted case that consumes more cooling coil
power, and provides slightly lower-humidity supply air, during the initial hours.
However, during the summer, the RH levels in the offices are generally well below
60% with or without return air ducting, so that this effect has no impact on the
number of occupied hours above 60%.
5.13.2 The Effect of Ducted Returns on Increased Ventilation Rates
The relationship of the ducted-return results in Table 17 to the unducted results
is the same at 20 cfm/person as it is at 5 cfm/person, and for the same reasons
discussed in Section 5.13.1.
Capacity. With ducted returns at 20 cfm/person, the required cooling capacity
increases by 15.635 kBtu/h above the unducted 5 cfm/person baseline. This is
exactly the same increase as for the unducted 20 cfm/person case, for the same
reasons discussed previously.
5-55
-------
Energy consumption and cost. With ducted returns at 20 cfm/person, annual
HVAC energy consumption and costs increase by 3,089 kWh and $311, respectively,
relative to the unducted 5 cfm/person baseline. As expected, these figures are
slightly less than the 3,245 kWh and $325 annual increases predicted for the
t/flducted 20 cfm/person case.
Hours at elevated temperature or RH. The percentage of hours undercooled for
the ducted 20 cfm/person case (0.6% in Table 17) is slightly higher than the 0.4%
for the unducted 20 cfm/person case, for the same reasons discussed in the preceding
section.
The percentage of occupied hours above 60% RH for the ducted case at 20
cfm/person (0.8% in Table 17) is slightly lower than the 0.9% for the inducted case
at 20 cfm/person. The reason for this is that - in the ducted case -- slightly more
power is consumed by the cooling coils during the first few hours after startup on
winter weekday mornings. The ducted return is not cooled by the cool plenum during
these early hours, as occurs in the unducted case. Thus, in the ducted case, there
is a slightly increased cooling coil load during the early hours, and hence slightly
increased latent cooling.
For example, on the morning of Monday, January 14, operating at 20 cfm/
person, the cooling coils provide over 1,000 Btu of additional latent cooling during the
first 4 to 5 hours after startup in the ducted case, relative to the unducted case. The
front and rear plenum spaces range between 69 and 72 °F during these early hours
in the ducted case, indicating that the relatively cool plenums could have contributed
to cooling the office space below the cooling setpoint of 75 °F had the ducting not
kept the plenum air isolated.
The reason why the percentage of occupied hours at elevated RH is slightly
reduced at 20 vs. 5 cfm/person with the ducted PSZ system is the same as that for
the t/Aiducted PSZ system, as discussed in Section 4.2 (see Performance) and Section
5.12.2 (see Single PVA VS conditioning both zones).
5.14 USE OF COLD-AIR DISTRIBUTION
The previous sections have assumed that the minimum temperature of the
supply air being distributed to the offices during cooling is 55 °F. This is a typical
value for the minimum supply temperature, selected in an effort to remain above the
dew point of the supply air and thus to reduce the risk of moisture condensation on
the ductwork, and at the diffusers.
In some applications, consideration is sometimes given to designing the HVAC
system to operate at cooling supply air temperatures well below 55 °F (i.e., to the use
of "cold-air distribution"). Operation at lower supply air temperatures would have the
advantage of reducing the volume of cooling air that must be distributed. Since
5-56
-------
power consumption by a given fan is proportional to the cube of the volumetric flow
rate through the fan, a decrease in fan flow rate could result in a significant decrease
in fan power consumption. The reduced flow rates to the zones would also offer the
advantage of smaller duct size.
The possible disadvantages of cold-air diffusers include: a) the need for more
sophisticated designs and controls to reduce the risk of moisture condensation on the
ductwork and at the diffusers; and b) the possible need for fan-assisted ("powered")
terminals in order to achieve adequate "throw" of the reduced volume of air out
through the diffusers into the office space.
For this assessment, the use of cold-air diffusers was considered, with a
minimum supply air temperature of 42 °F, for both the baseline dual-PSZ configuration
and for the PVAVS (which substantially reduces fan power consumption even at
55 °F). The temperature of 42 °F is a typical value for cold-air diffuser designs, high
enough to keep cooling coil surface temperatures above 34 to 35 °F, thus avoiding
frosting on the outside of the coils,
It must be emphasized that this analysis of cold-air diffusers in this application
is academic, simply to determine what the effect could be. It is unlikely that cold-air
distribution would be used in an application such as this in practice. Strip mall offices
of the type assumed here tend to have simple mechanical systems to reduce
installation cost and to simplify maintenance. The complications that would be added
by a cold-air diffusion system would be inconsistent with the usual desire for a simple
system in this application. Moreover, in such a small office, reduction in duct installa-
tion cost (one of the benefits of cold-air distribution) might not be fully realized.
The results of the computations regarding cold-air distribution are presented in
Table 18.
5.14.1 The Effect of Cold-Air Distribution at 5 cfm/person
Capacity - PSZ system. Table 18 indicates that the use of cold-air distribution
results in a very small (0.514 kBtu/h) reduction in the cooling capacity of the two-PSZ
configuration, relative to the baseline case with 55 °F supply air. This occurs because
the increase in latent cooling capacity required for the cold-air system (since more
moisture is condensed at the lower supply temperature) is slightly more than offset
by the decrease in required sensible cooling capacity resulting from the reduction in
the amount of heat added to the air stream by the supply fan (since the fan is now
having to move less air).
Because the supply air is cooled by an additional 13 F° in the cold-air case, the
required flow drops from the baseline 3,810 cfm value to 2,263 cfm, a 40% reduc-
tion, based on a straightforward thermal balance. At 5 cfm/person, the colder coils
5-57
-------
TABLE 18
Effect of Cold-Air Distribution:
Increase Compared to Baseline Case with 5 cfm/person
Case
Increase in
total required
cooling capacity
In
kBtu/h As %
Increase in
annual energy consumption
As % of As % of
HVAC building
In kWh energy energy
Increase in annual energy cost
As % of As % of
HVAC building
In $ energy cost energy cost
Hours/yr Occupied
under- hours/yr
cooled, RH > 60%,
% %
Baseline values: 103.5585
Two PSZ units,
MIN-SUPPLY-T
= 55 °F,
OA = 5 cfm/p
(absolute values)1
OA ventilation rate = 5 cfm/person
Two PSZ units,
MINSUP-T = 42°F -0.514
-0.5
One PVAVS unit,
MIN-SUP-T = 42°F 0.339 0.3
OA ventilation rate = 20 cfm/person
Two PSZ units,
MIN-SUP-T = 42°F 23.051 22.2
One PVAVS unit,
MIN-SUP-T =42°F 24.119 23.3
26,1451
(HVAC) --
60,1 611
(building) -
-2,930 -11.2 -4,9
-4,941 -18.9 -8.2
1,160 4.4
-823 -3.1
1.9
-1.4
2,510'
(HVAC)
4,2731
(building)
0.4
1.2
-268 -10.7
-392 -15.6
157
39
6.2
1.6
-6.3
-9.2
3.7
0.9
0.1
•0
0.7
0.2
0.7
0.3
0.3
0.6
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
in the cold-air case can remove perhaps 2.6 additional pounds of water per hour from
that reduced air stream on a typical summer day, relative to the baseline case. At a
latent heat of vaporization of approximately 1,050 Btu/lb at coil temperatures, the
additional 2.6 lb of water removal would correspond to a required increase in latent
capacity of about 2.7 kBtu/h in the cold-air case. But the reduction in flow rate would
decrease the sensible heat added by the fan land hence the required sensible capacity)
by about 3.3 kBtu/h. This simple analysis shows why switching the dual-PSZ
configuration to cold-air operation would be expected to reduce the required total
cooling capacity by < 1 kBtu/h for the building being considered here.
Changing to cold-air operation also impacts some of the other parameters used
by the DOE-2 program to compute cooling capacity, including: a) the coil bypass
factor; and b) the wet-bulb temperature in the offices (used to convert capacities to
values at the conditions specified by the Air-Conditioning and Refrigeration Institute).
These factors lead to the precise value of 0.514 kBtu/h as the computed reduction in
cooling capacity for the cold-air case at 5 cfm/person.
Capacity - PVAVS. A PVAVS operating with cold-air distribution requires a
cooling capacity of 103.897 kBtu/h. As shown in Table 18, this is very slightly
(0.339 kBtu/h) larger than the 103.558 kBtu/h computed for the baseline PSZ system
with the standard 55 °F supply air temperature. It is 0,853 kBtu/h greater than the
103.044 kBtu/h computed for the PSZ system with the 42 DF supply, and 1.140
kBtu/h less than the 105.037 kBtu/h for the PVAVS with 55 °F supply.
The PVAVS at 42 °F requires less capacity than the PVAVS at 55 °F for the
same reasons that the PSZ at 42 °F requires a lower capacity than the PSZ at 55 °F.
The PVAVS at 42 °F condenses more moisture than its 55 °F PVAVS counterpart, and
thus requires a greater latent capacity. But this latent increase is more than offset by
a reduction in the sensible capacity requirement, resulting from reduced supply fan
heat due to the reduced air flows.
The reduction in supply air temperature causes a slightly greater reduction in
the computed capacity for a PVAVS (1.140 kBtu/h) than for a PSZ system (0.514
kBtu/h). This occurs because the default PVAVS supply fan is assumed to add more
sensible heat to the air stream (2.1 F°) than the PSZ fan (1.8 F°). Hence, a given
reduction in peak air flow rate results in a somewhat greater reduction in fan-imparted
sensible heat for the PVAVS. If the PVAVS computer runs were made specifying a
supply fan temperature rise of 1.8 F°, identical to the PSZ system, then the PVAVS
capacity at either supply temperature would become essentially equal to the PSZ
capacity at the same temperature. The reductions in capacity resulting from a
reduction in supply temperature would then necessarily become the same for both
systems.
5-59
-------
The PVAVS at 42 °F requires a slightly greater cooling capacity (by 0.853
kBtu/h) than the PSZ system at that same temperature, as is the case when the two
systems are compared at 55 °F (see Section 5,12.1). Again, this relationship results
because of the increased amount of sensible heat added by the PVAVS supply fan,
compared to the PSZ fan. But at the reduced flows in the 42 °F case, the capacity
difference between the two systems has shrunk -- from 1.479 kBtu/h in the 55 °F
case (see Table 16) to 0.853 kBtu/h in the 42 °F case.
Energy consumption and cost - PSZ system. With cold-air distribution, Table
18 shows that the energy consumption by the PSZ system decreases by 2,930 kWh/
yr compared to the baseline PSZ system with 55 °F supply. Annual energy costs
decrease by $268.
This 2,930 kWh reduction in energy consumption approximately equals the
reduction in power consumption by the supply fan {a little less than 3,000 kWh,
resulting from the reduced flows at 42 °F). The reduction in fan consumption is
partially offset by an increase in cooling coil energy consumption in the 42 °F case.
However, at 5 cfm/person, the increase in cooling coil consumption is small (about 60
kWh/yr). That is, at 5 cfm/person, the increase in annual cooling coil power
consumption for latent cooling at 42 °F (due to increased moisture condensation) is
just slightly greater than the reduction in annual coil power consumption for sensible
cooling (due to reduced sensible heat added by the supply fans at the reduced flows).
Thus, the overall 2,930 kWh reduction is explained by the roughly 3,000 kWh
reduction in supply fan consumption, slightly offset by the roughly 60 kWh increase
in cooling coil consumption.
The 2,930 kWh reduction with the cold-air PSZ configuration corresponds to
11 % of the baseline energy consumption by the HVAC system, and 11 % of the
HVAC energy cost. If cold-air distribution were practical in this application, it would
represent one of the single most effective steps for reducing energy consumption in
this building, following: achievement of infinite thermal resistance in the windows,
walls, and roof (20% reduction); dramatic reduction of power consumption by lighting
and equipment (19%); and conversion from PSZ to PVAVS, with or without cold-air
distribution (12 to 19%).
Energy consumption and cost - PVAVS. As shown in Table 18, with cold-air
distribution, the PVAVS requires 4,941 kWh/yr less than the baseline PSZ system at
55 DF, at a cost savings of §392. The cold-air PVAVS requires 1,792 kWh/yr (and
$206/yr) less than the 55 °F PVAVS, and 2,011 kWh/yr (and $124/yr) less than the
42 °F PSZ system.
Of the 4,941 kWh/yr reduction in energy consumption relative to the 55 °F PSZ
baseline, over 90% results from a reduction in supply fan power consumption (which
5-60
-------
decreases by 62%, from 7,328 to 2,758 kWh/yr). The remaining percentage results
from a reduction in cooling coil power consumption. With the 42 °F PVAVS, the
reduction in annual coil power consumption for sensible cooling (due to reduced
sensible heat added by the supply fan) more than offsets the increase in annual coil
consumption for latent cooling, relative to the baseline.
The 1,792 kWh/yr reduction for the 42 °F vs. the 55 °F PVAVS is due entirely
to reduction in fan power consumption between the two cases. Annual cooling coil
power consumption is almost identical for PVAVS at the two supply temperatures,
and does not contribute to the difference. Apparently, over the course of the year,
the increased energy consumption for latent cooling at the lower supply temperature
is just about exactly offset by the reduced consumption for sensible cooling due to
reduced fan heat at the lower flows.
Of the 2,011 kWh/yr reduction for the 42 °F PVAVS compared to the PSZ
system at that supply temperature, 80% is due to a reduction in supply fan power
consumption due to the lower flows in PVAVS vs. PSZ. The remaining 20% results
from reduced cooling coil power consumption, due to net reduced fan heat generation
at the reduced PVAVS flows.
In practice, the net energy savings with cold-air distribution -- and especially
with the cold-air PVAVS -- would not be as great as the figures shown here. The low
flows resulting at this low supply temperature might well necessitate the use of
powered terminals - i.e., auxiliary fans at the terminal boxes to ensure adequate
throw of the supply air out of the diffusers, for proper mixing in the offices. Such
auxiliary fans would consume some portion of the power savings computed above.
Hours at elevated temperature and RH - PSZ system. As shown in Table 18,
the percentage of hours undercooled with the PSZ system at 5 cfm/person drops
slightly -- from 0.4% to 0.1% (a reduction of 20 hours/yr) -- when the minimum
supply air temperature is reduced to 42 °F. This reduction occurs because, relative
to the 55 °F case, the control system for the PSZ system in the 42 °F case (as
modelled by DOE-2) causes it to operate at closer to full power -- providing greater
sensible and latent cooling -- immediately after startup on warm summer mornings,
when undercooled hours tend to occur.
Taking the morning of Monday, July 8, as an example, the baseline 55 °F PSZ
system operates at between 83 and 91 % of maximum capacity during the first two
hours after startup, providing an average total cooling of 96,000 Btu/h (both PSZ
units combined). The two offices average 77,7 °F during those two hours, more than
2 F° above the 75 °F setpoint. Hence, the offices are both considered to be under-
cooled during these two hours.
5-61
-------
By comparison, with the 42 °F supply air case, the PSZ system (as modelled by
DOE-2) operates at between 93 and 100% of full capacity during the first two hours
after startup, providing an average of 102,000 Btu/h. The office temperatures now
average 76.9 °F during these two hours -- i.e., within the 2 F° throttling range of
75 °F - and hence no undercooled hours are recorded.
Table 18 shows that the percentage of occupied hours above 60% RH
decreases slightly -- from 1.2% to 0.7% (a decline of about 15 hours/yr) - when the
supply temperature is reduced to 42 °F. Elevated-RH hours occur during the first few
hours after startup on relatively cool winter mornings with high outdoor humidity.
With both the 55 and the 42 °F supply temperatures, the cooling coils remain off
during the initial 2 or 3 hours after startup, leading to RH values above 60% in both
offices during those hours. But after the coils do come on, the 42 °F supply case
provides more latent cooling than does the 55 °F case (while providing less sensible
cooling and somewhat less total cooling). Thus, in the 42 °F case, the humidity levels
drop more rapidly, and there are thus fewer total hours above 60% RH, as reflected
in Table 18.
Take the morning of Monday, January 14, as an example. In the baseline 55 °F
case, the front office is above 60% RH during the first 6 hours after startup (with the
coils off totally during the first 3 hours); the latent cooling provided by the front PSZ
unit during those 6 hours is 6,566 Btu. By comparison, in the 42 °F case, the front
office is above 60% RH only during the first 5 hours, and the latent cooling by the
front unit during the first 6 hours is somewhat higher, 6,717 Btu. The rear office is
above 60% RH during the first 4 hours with the 55 °F supply, with latent cooling of
6,211 Btu by the rear unit during those hours. But with the 42 °F supply, the rear
office is above 60% RH only during the first 3 hours, with latent cooling of 7,828 Btu
during the first 4 hours.
Hours at elevated temperature and RH - PVAVS. Like the PVAVS with a 55 °F
supply temperature (Table 16), Table 18 shows that the PVAVS with the 42 °F supply
temperature at 5 cfm/person results in no hours undercooled during the year. The
reasons are the same as those given in Section 5,12.1 (see Single PVAVS condi-
tioning both zones). As with the 55 °F PVAVS, the 42 °F PVAVS begins operating
at full capacity during the first hours after startup on warm summer mornings, when
undercooled hours tend to occur. As a result, there are no undercooled hours.
As shown in Table 18, there are fewer occupied hours above 60% RH with the
42 °F PVAVS at 5 cfm/person (0.3%) than there are with the 42 °F PSZ system
(0.7%). This is a reversal of the situation with the 55 °F PVAVS, where the
percentage of elevated-RH hours (1.5%) were slightly greater than the percentage
with the 55 °F PSZ system (1.2%).
5-62
-------
The first issue that will be discussed is why there are so many fewer elevated-
RH hours with the 42 °F PVAVS compared to the 55 °F PVAVS (0.3 vs. 1.5%,
corresponding to a difference of 40 occupied hours per year). During the first hour
after startup on cool, humid summer mornings, the 42 °F system -- operating at a low
coil surface temperature of 40 °F and experiencing relatively high-humidity inlet air,
but operating at minimum flow - consistently provides more Btu's of latent cooling
than does its 55 °F PVAVS counterpart. This drops office humidity rapidly, although
not always to RH _< 60% during that first hour. During subsequent hours, the 55 °F
PVAVS is commonly providing more Btu's of latent cooling; although its coil
temperature is higher, its flows are greater and the moisture content of the entering
air is higher. But because of the high level of moisture removal during that first hour
with the 42 °F system, the RH levels in the office spaces drop below 60% sooner
with that system.
On Monday, January 14, for example, the 55 °F PVAVS reduces the office RH
(which is above 80% prior to startup) to 70% during the first hour, and to below 60%
after 9 hours. The average RH during those 9 hours is about 67%. By comparison,
with the 42 °F PVAVS, the office RH drops to 62% during the first hour, drops below
60% after only 5 hours (instead of 9), and averages about 64% during those 5 hours.
The trend is similar on Tuesday, January 15: the offices are at 61 to 62% RH during
the first two hours with the 55 °F system, but are never above 60% (on an hourly
average) with the 42 °F system.
The reason that the PVAVS at 42 °F results in fewer elevated-RH hours than
the PSZ system at 42 "F (0.3% vs. 0.7%, corresponding to about 15 hours per year)
is more difficult to explain. As discussed in the previous subsection, on January 14,
the 42 °F PSZ system results in 5 hours above 60% RH in the front zone (ranging
between 72 and 100% RH), and 3 hours in the rear zone (ranging between 65 and
100% RH). By comparison, on that same day, the 42 °F PVAVS results in 6 hours
above 60% RH in both zones (ranging between 62 and 65%, approximately). On the
next day (Tuesday, January 15), neither system has any hours above 60% RH. Thus,
the 42 °F PVAVS would seem to be having more elevated-RH hours than the cold-air
PSZ system, and it is unclear why PVAVS is being reported as having fewer hours.
Possibly, the hours above 60% RH with the 42 °F PVAVS are so close to the 60% RH
target, that the program is not counting them as being above 60% RH.
5.14.2 The Effect of Cold-Air Distribution on Increased Ventilation Rates
Capacity - PSZ system. As shown in Table 18, increasing the OA rate to 20
cfm/person with the 42 °F PSZ system increases the computed cooling capacity by
23.051 kBtu/h, relative to the baseline 55 °F PSZ system at 5 cfm/person. This
increase is larger than the 15.635 kBtu/h increase required when OA rate is raised to
20 cfm/person with the 55 °F PSZ system (see Table 3).
5-63
-------
This larger OA-induced increase in required capacity with the 42 °F system
results from increased latent cooling. The increase in OA to 20 cfm/person in this
humid climate significantly increases the latent load on the system, and the colder
cooling coils in the 42 °F case will remove a greater amount of this additional latent
heat.
intuitively, one might expect that - in this warm climate - increased OA flows
might also increase the sensible cooling requirements for the 42 °F system, since the
increased amount of (generally warm) OA will have to be cooled to a lower tempera-
ture than in the 55 DF case. However, analysis indicates that, in fact, the impact of
increased OA on sensible cooling is small. Of the increase in cooling capacity
resulting from the increase in OA to 20 cfm/person in the 42 °F case, no more than
a few percent appears to be required for additional sensible cooling of the OA;
essentially all of the added capacity is required for latent cooling of the OA.
As indicated in the earlier discussion of the cold-air system at 5 cfm/person, the
increased latent load at 42 "F is partially offset by a decrease in the fan-induced
sensible load. At the reduced flows in the cold-air system, the amount of sensible
heat added to the air by the supply fan is reduced. As discussed in Section 5.14.1,
at 5 cfm/person, the increase in peak latent capacity is entirely offset by the decrease
in peak sensible capacity. But with the cold-air system at 20 cfm/person, the
significant increase in peak latent capacity {resulting from the increased OA inflow)
is much greater than the decrease in peak sensible capacity (which remains almost
unchanged from the 5 cfm/person cold-air case, since the total 42 °F airflow remains
unchanged). Therefore, the total required cooling capacity experiences a net increase
for the 42 °F PSZ system at 20 cfm/person, relative to the 55 °F PSZ system at 20
cfm/person, by an amount equal to 23.051 - 15.635 = 7.416 kBtu/h. This 7.4
kBtu/h differential results from an increase of roughly 9.5 kBtu/h in latent capacity,
partially offset by a decrease of roughly 2 kBtu/h in sensible capacity, in the 42 °F
case.
Capacity - PVAVS. At 20 cfm/person, the PVAVS with 42 °F supply air has a
computed capacity requirement of 127.677 kBtu/h. This capacity is 24.199 kBtu/h
greater than that of the baseline 55 °F PSZ system at 5 cfm/person, as indicated in
Table 18.
This computed capacity of the 42 °F PVAVS at 20 cfm/person is 23.780 kBtu/h
greater than that of the 42 °F PVAVS at 5 cfm/person. This is almost exactly the
same differential as exists between the 42 °F PSZ systems at 20 vs. 5 cfm/person
(23.565 kBtu/h), and occurs for the same reason. This 23.780 kBtu/h difference
equals the increase in latent cooling load created by the increased inflow of humid
outdoor air.
5-64
-------
This capacity of the 42 °F PVAVS at 20 cfm/person is 6.706 kBtu/h greater
than that required for the 55 °F PVAVS at 20 cfm/person. Diagnosis indicates this
6.7 kBtu/h increase is the net result of an increase in latent capacity by over 7 kBtu/h,
partially offset by a decrease in sensible capacity by about 1 kBtu/h.
Energy consumption and cost - PSZ system. As shown in Table 18, the 42 °F
PSZ system at 20 cfm/person consumes 1,160 kWh/yr more power (at an increased
cost of $157/yr), compared to the baseline 55 °F PSZ system at 5 cfm/person. Thus,
if an increase from 5 to 20 cfm/person were accompanied by a switch from 55 °F to
cold-air distribution, the energy penalty of the OA ventilation rate increase would
theoretically decline from 3,245 kWh and $325./yr (see Table 3) to 1,160 kWh and
$157/yr, decreases of 50% or more.
Compared with the 55 °F PSZ system at 20 cfm/person, the 42 °F PSZ system
at 20 cfm/person consumes 2,085 kWh/yr less power, at a cost savings of $168/yr.
This 2,085 kWh/yr reduction for the 42 ° vs. the 55 °F system at 20 cfm/person is
the net effect of roughly a 3,000 kWh reduction in fan power consumption (due to
reduced air flows), partially offset by roughly a 900 kWh increase in cooling power
consumption (due to the increased latent load at the lower supply temperature).
As discussed in Section 5.14.1 (Energy consumption and cost - PSZ system),
the increase in PSZ cooling coil consumption at 42 ° vs. 55 °F was only about 60
kWh/yr at 5 cfm/person. At that lower OA rate, the increased latent cooling at 42 °F
was almost entirely offset by a decrease in sensible cooling due to reduced sensible
heat addition by the reduced-flow 42 °F supply fan. But at 20 cfm/person, the
required latent cooling energy increases, such that the latent increase at 42 °F
becomes 900 kWh/yr (rather than only 60 kWh/yr) greater than the sensible decrease
due to the reduced fan power. Thus, the 3,000 kWh reduction in fan power
consumption at 42 °F vs. 55 °F -- a reduction which remains the same, regardless of
the OA rate -- is partially offset by a 900 kWh increase in 42 °F cooling coil power
consumption at 20 cfm/person, while it had been offset only by a 60 kWh increase
in cooling coil power at 5 cfm/person.
Compared with the 42 °F PSZ system at 5 cfm/person, the 42 °F PSZ system
at 20 cfm/person consumes 4,090 kWh/yr more power at an increased cost of
$425/yr. These increases caused by increased OA with the 42 °F PSZ system are
greater than those predicted for the OA increase with the 55 °F PSZ system (3,245
kWh and $325/yr, from Table 3). An analysis of the discussion in the preceding
paragraph shows that the incremental effect of 20 vs. 5 cfm/person is greater in the
42 °F case because the OA increase causes an added (900 - 60 = > 840 kWh/yr
increase in cooling coil power consumption in the 42 °F PSZ system. Thus, 4,090
kWh = 3,245 kWh + -840 kWh.
5-65
-------
Energy consumption and cost - PVAVS. From Table 18, the 42 °F PVAVS at
20 cfm/person consumes 823 kWh/yr /ess power (but at an energy cost increase of
$39/yr)f compared to the baseline 55 °F PSZ system at 5 cfm/person, (The fact that
cost goes up while consumption goes down indicates that the 42 °F PVAVS results
in an increase in kW demand charges that more than offsets the decrease in kWh
usage charges.)
Thus -- hypothetical^ -- an increase in OA ventilation from 5 to 20 cfm/person
would nominally result in an energy savings of 823 kWh/yr, if this increase were
accompanied by a switch from a 55 °F PSZ system to a cold-air PVAVS. The energy
cost penalty associated with the increase in ventilation would be negligible ($39/yr).
Compared with the 55 °F PVAVS at 20 cfm/person (see Table 16), the 42 °F
PVAVS at 20 cfm/person consumes 745 kWh/yr less power, at a cost savings of
$85/yr. This 745 kWh savings with the 42 °F system is the net effect of a roughly
1,800 kWh reduction in fan power consumption, partially offset by a 1,050 kWh
increase in cooling coil power consumption.
This 745 kWh and $85/yr savings from operating a 20 cfm/person PVAVS at
42 ° rather than 55 °F is fairly modest. By comparison, the 55 °F, 20 cfm/person
PVAVS had offered a 3,323 kWh and $201 /yr savings compared to the 55 °F, 20
cfm/person PSZ system. Thus, if one is prepared to convert from a 55 °F PSZ
system, the greatest incremental energy savings is to achieved by switching to a
standard 55 °F PVAVS. Converting further, from a 55 °F PVAVS to a 42 °F PVAVS,
would seem to buy only a modest additional incremental energy savings, considering
the added complexities that the cold-air distribution system would introduce.
As discussed in the preceding subsection (Energy consumption and cost - PSZ
system), reducing the supply air temperature of a 20 cfm/person PSZ system from
55 ° to 42 °F results in a savings of 2,085 kWh and $168/yr. By comparison, the
predicted savings are smaller (745 kWh and $85/yr) for reducing the temperature of
the 20 cfm/person PVAVS. The relatively modest reductions achieved by converting
a PVAVS to 42 °F occur because -- even at 55 °F -- the PVAVS has already
substantially reduced flow rates (and hence fan power consumption). The ability to
further reduce PVAVS flows by reducing supply temperatures is thus more limited.
This situation is illustrated by the fact that -- with the 20 cfm/person PSZ system --
reducing the supply temperature reduces fan power consumption by about 3,000
kWh/yr, as discussed previously. However -- with the 20 cfm/person PVAVS —
reducing the temperature reduces fan power consumption by only about 1,800
kWh/yr.
5-66
-------
With either the PSZ system or the PVAVS at 20 cfm/person, the total annual
savings in power consumption (745 to 2,085 kWh) and cost ($85 to $168) achieved
by using cold-air distribution would be at least partially consumed by the auxiliary fans
that would likely be required to power the terminals for proper throw of the supply air
out of the diffusers. This is especially true with the PVAVS.
Hours at elevated temperature and RH - PSZ system. Table 3 shows that, as
OA is increased from 5 to 20 cfm/person with the 55 °F PSZ system, the percentage
of hours undercooled remains essentially unchanged, at 0.4% of all hours. By
contrast, Table 18 shows that as OA is increased with the 42 °F system, the
percentage of undercooled hours increases, from 0.1 % to 0.7% of all hours.
The reason for this effect is that - with the 42 °F system - an increase in OA
actually results in a decrease of the amount of sensible cooling provided by the
system during summer morning startups. Although the increase in OA increases the
total cooling provided by the 42 °F system's coils, it decreases the sensible cooling
by the coils, in part because the increased flow of OA (which is slightly cooler than
the return air at startup on summer mornings) provides an economizer sensible cooling
effect. But this increase in "economizer" sensible cooling does not compensate for
the reduction in sensible cooling by the coils, as modeled. As a result - with the
42 °F system at 20 vs. 5 cfm/person - the office space operates at higher tempera-
tures during the first few hours after startup on summer mornings, and there are more
undercooled hours.
Take the morning of Monday, July 8, as an example. At an OA rate of 5
cfm/person, the 42 °F PSZ system provides total cooling of 101,358 Btu to the front
zone during the first 2 hours after operation begins. Of this total, 89,289 Btu are
sensible cooling and 12,069 Btu are latent. At 5 cfm/person, the average temperature
in the front zone is 77.0 °F during those 2 hours, and neither of those hours is
recorded as undercooled (i.e., all are within 2 F" of the 75 °F setpoint).
At an OA rate of 20 cfm/person, the total cooling in the front zone during the
first 2 hours on July 8 with the 42 °F PSZ system is 110,724 Btu, 9,000 Btu greater
than the total cooling at 5 cfm/person. However, of this total cooling at 20
cfm/person, only 77,684 Btu is sensible cooling -- almost 12,000 Btu less than the
sensible cooling at 5 cfm/person. Latent cooling at 20 cfm/person jumps to 33,040
Btu -- almost 21,000 Btu greater than at 5 cfm/person. The reduction in sensible
cooling by the coils results in part because the increased flow of outdoor air ~ which
averages 76.5 °F during those 2 hours — provides some "economizer" cooling of the
return air, which averages 82.4 °F prior to OA mixing during those hours. But the
reduction in sensible cooling by the coils at 20 cfm/person over-compensates for the
increase in "economizer" sensible cooling, because, at 20 cfm/person, the front office
temperature averages 78.2 °F — well above the 77.0 °F average at 5 cfm/person. As
a result, while the 42 °F PSZ/5 cfm case had no undercooled hours in the front office
5-67
-------
on July 8, the 42 °F PSZ/20 cfm case is undercooled for the first 5 hours, according
to the DOE-2 model.
This effect does not occur with the OA increase in the 55 °F PSZ system. With
the 55 °F system, the sensible cooling in the front office during the first 2 hours on
July 8 is 86,507 Btu at 5 cfm/person and 86,910 Btu at 20 cfm/person -- essentially
unchanged with the OA increase. As a result, the front office temperatures in the
55 °F PSZ/20 cfm case are within 0.1 to 0.2 F° of those in the 55 °F PSZ/5 cfm case.
With the 55 °F system, the front office is undercooled for the first 2 hours after
startup on July 8 at both OA rates. Consequently, Table 3 shows the percentage of
undercooled hours for the 55 °F PSZ to be the same at both 5 and 20 cfm/person.
It would appear that the substantial increase in latent cooling in the 42 °F
system with the increase in OA is confounding either the PSZ unit's control system,
or the DOE-2 model, resulting in insufficient sensible cooling. The 42 °F system at 20
cfm/person is not operating at full capacity, according to the model; during the first
2 hours on July 8, the front PSZ unit is operating only at 77 to 90% of capacity, an
even lower percentage than the 55 °F/20 cfm unit (which operated at 82 to 93%).
And the supply air is at a temperature above the 42 °F minimum supply temperature
during the first 2 hours. Thus, the 42 °F/20 cfm system had the capacity to cool the
office more effectively than it did.
Table 18 shows that operating the 42 °F PSZ system at increased OA slightly
reduces the number of occupied hours above 60% RH, from 0.7% at 5 cfm/person
to 0.3% at 20 cfm/person. The difference between 0,7% and 0.3% corresponds to
12 hours per year per zone. The apparent explanation is that -- in the first few
occupied hours on cool winter mornings, before the cooling coils have activated but
when the indoor latent sources have caused indoor RH to exceed outdoor RH -- the
increased OA flow reduces indoor RH.
To illustrate this impact on RH, take the morning of Monday, January 14.
During the first 4 hours after startup on this particular morning, the outdoor humidity
ratio in Miami is about 0.014 lb moisture per lb dry air, corresponding to RH at office
temperature ranging between 80 and 90%. The cooling coils in the PSZ units do not
activate for the first 4 hours in the front zone, or the first 2 hours in the rear zone.
As a result, during those hours, there is no moisture removal by the HVAC system.
When the PSZ units are operating at 5 cfm/person, latent heat released by the
occupants causes the RH in the office space to average 99% during the first 4 hours
in the front zone, and 94% during the first 2 hours in the rear zone. But at 20
cfm/person, these averages drop by 10 percentage points, to 89% in the front zone
and 84% in the rear zone. Note that these decreases are independent of the design
cooling supply air temperature for the system, since the coils are off.
5-68
-------
When the cooling coils do come on, RH levels drop faster for the 5 cfm/person
system. Even though the 20 cfm/person systems achieve many more Btu's of latent
cooling, as discussed previously, this increase is insufficient to compensate for the
increased latent heat in the incoming air at the higher OA flow. As a result, during
the several hours after the coils have activated, the 42 °F PSZ at 5 cfm/person
reduces zone RH to levels 0 to 10 percentage points lower than the RH levels
predicted for the 42 °F PSZ at 20 cfm/person.
From the figures above, it is clear that January 14 is not one of the mornings
when the increased OA causes RH to drop below 60% during the initial hours before
the cooling coils come on. Presumably, the 12 fewer elevated-RH hours per year
occur on other days during cool-weather months.
As noted above, the impact of increased OA in reducing indoor RH before the
coils come on is independent of the minimum supply air temperature that the system
is designed to deliver once the coils are activated. Accordingly, this apparent
explanation for why increased OA causes the decrease in elevated-RH hours shown
in Table 18 for the 42 °F PSZ system, must also be the explanation for the reduction
shown in Table 3 for the 55 °F PSZ system. The decrease in elevated-RH hours for
the 55 °F system with the increase in OA also corresponds to 12 hours per zone per
year (a reduction from 1.2% to 0.9% of occupied hours, as shown in Table 3).
Hours at elevated temperature and RH - PVAVS. Table 18 shows that, as the
42 °F PVAVS is increased from 5 to 20 cfm OA/person, the percentage of hours
undercooled increases from about zero to 0.2% of all hours. This slight increase --
corresponding to about 11 hours per zone per year -- occurs for the same reason
indicated previously for the PSZ system. That is - perhaps confounded by the high
latent load at 20 cfm/person, and/or the anticipated "economizer" cooling by the
additional OA - the 42 °F PVAVS provides insufficient sensible cooling at the higher
OA rate after startup on summer mornings, even though it has sufficient capacity to
provide the needed cooling.
Take the morning of Monday, July 29, as an example. During the first 4 hours
after startup on that morning, the 42 °F PVAVS operating at 5 cfm/person provides
371,500 Btu of sensible cooling; the front zone is undercooled during 1 of those 4
hours at 5 cfm/person, and the rear zone is not recorded as undercooled during any
of the hours. But when OA is increased to 20 cfm/person, the system provides only
351,500 Btu during those first 4 hours, 20,000 Btu less than in the 5 cfm/person
case; the front zone is undercooled for 2 of the 4 hours, and the rear zone for 1 of the
4 hours. This occurs even though - for 3 of the 4 hours (including all of the hours
when one or both zones are undercooled) -- the 20 cfm/person system is operating
below its available capacity (i.e., at part-load ratios of 0.93 to 0.97), and thus had the
capability of providing more sensible cooling.
5-69
-------
Regarding hours at elevated RH, Table 18 shows that -- with the 42 °F PVAVS,
the percentage of hours at RH > 60% increase slightly {from 0.3 to 0.6%) as OA is
increased. This result is consistent with that for the 55 °F PVAVS, where the OA
increase also resulted in an approximate doubling of elevated-RH hours (from 1.5%
to 2.9%, in Table 16).
And the explanation in the 42 °F case is the same as that for the 55 °F PVAVS
(see Section 5.12.2, Single PVAVS conditioning both zones. Hours at elevated
temperature or RH). Although the 20 cfm/person system at a given supply
temperature removes substantially more latent heat than does the 5 cfm/person
system at the same temperature, this increased removal does not compensate for the
increased latent inflow due to the increased OA rate.
Take Monday, January 14, as an example. Because the externally induced and
occupant-induced loads on the office spaces are independent of OA rate, the flows
to each zone from the 42 °F PVAVS are unchanged with increasing OA (remaining at
or near the minimum flow rate throughout that day). Because PVAVS character-
istically supply air near the minimum temperature, the supply air temperatures and the
coil surface temperatures are essentially unchanged with increasing OA; the coil
surface temperature averages 37.8 °F with the 5 cfm/person 42 °F PVAVS, and 37.9
°F with the 20 cfm/person system on January 14. The coil bypass factor decreases
from an average of 0.67 at 5 cfm to 0.61 at 20 cfm, a decrease of under 10%. That
is, at 20 cfm, more of the air nominally comes into contact with the coil surface,
reflecting, e.g., a deeper tube bank or reduced velocity, as necessary to cool the
warmer entering air down to the same coil outlet temperature.
And, correspondingly, the 20 cfm/person system removes 16% more latent
heat on January 14 (501,000 Btu compared to 431,000 Btu at 5 cfm/person). But
this increase is insufficient to compensate for the increased moisture present in the
increased OA flow. The moisture content of the air leaving the cooling coils increases
15%, from an average of 0.0087 lb of moisture per lb of dry air at 5 cfm to an
average of 0.0100 lb/lb at 20 cfm. As a result, the office space — which is above
60% RH only for the first 5 hours on January 14 at 5 cfm/person - is above 60% RH
for 10 of the 13 hours at 20 cfm/person.
5.15 MODIFICATIONS TO THE ECONOMIZER
The systems discussed to date have all included an economizer (with the
exception of the PTAC unit in Section 5.12, which is not designed to accommodate
an economizer). Economizer operation has been controlled by the outdoor air
temperature, with an economizer limit temperature of 68 °F. That is, the economizer
ceases to function when the outdoor air temperature rises above 68 °F, and the OA
flow rate drops back to its minimum value (of 5 or 20 cfm/person).
5-70
-------
To assess the effect of the economizer on energy consumption and costs in the
study building, two modifications to this baseline economizer were considered;
a) the economizer is eliminated altogether; or
b) economizer operation is controlled by a controller which senses the enthalpy
rather than the temperature of the outdoor air.
It would not be anticipated that the baseline economizer would be in operation
very many hours for the building being considered here. For one thing, in the warm
climate of southern Florida, the outdoor temperature is not often below the econo-
mizer limit temperature of 68 °F. Even during winter months in Miami, the outdoor
temperature will generally be at 68 °F or below only during the first hour or two after
startup in the morning.
A second factor limiting economizer operation is that PSZ systems commonly
operate with economizer lockout. That is, the economizer will operate only when it
is able to provide all of the required cooling, so that the air conditioning compressor
remains off entirely. The hermetic motors in packaged compressors are cooled by the
circulating refrigerant flow. If the compressors operated at very low loads - as might
occur if the economizer were providing much but not all of the required cooling -- the
motors could overheat, thus reducing motor lifetime. Economizer lockout is intended
to avoid this problem by ensuring that the compressor and the economizer are never
operating at the same time.
With these two limits on economizer performance, the economizer will typically
operate, at most, only during the first hour after startup, even during the coolest
weather. With economizer lockout, economizer usage remains low even when the
economizer limit temperature is raised to values well above typical values (e.g., to
80 °F).
If the economizer operation were controlled by the enthalpy rather than the
temperature of the outdoor air, economizer operation would be limited even further.
With enthalpy-based control, it would no longer be sufficient for outdoor temperature
to be below 68 °F (and the mixed air entering the cooling coils to be above 68 CF).
Now, in addition, the enthalpy of the outdoor air must be less than the enthalpy of the
mixed air. The high outdoor humidity levels during those morning hours immediately
after startup would be expected to result in elevated outdoor enthalpies during those
few hours when the economizer might otherwise be expected to operate, thus further
limiting economizer operation.
Enthalpy controllers are more expensive than dry-bulb temperature controllers,
and require more maintenance. Thus, enthalpy controllers would not be typical of
small HVAC units in the capacity range being considered here. Enthalpy control
5-71
-------
calculations will be made here despite this fact, in order to demonstrate what the
effect would be on energy and performance,
5.15.1 The Effect of Eliminating the Economizer
Table 19 shows the effect of eliminating the economizer, at both 5 and 20 cfm
OA/person. As would be expected, the effect is small.
As would be expected, elimination of the economizer has no impact on the
required peak cooling capacity, at either 5 or 20 cfm/person. The economizer will not
be operating when the cooling coils are operating at their peak, and hence does not
enter into the capacity calculation.
HVAC energy consumption increases by 1.9% when the baseline temperature-
controlled economizer is deleted at 5 cfm/person, and by 1.6 percentage points
(14.0% minus 12.4%) at 20 cfm/person. HVAC energy cost increases by 0.9% (by
S23) at 5 cfm/person, and by 0.8 percentage points (by 344 - 325 = $19) at 20
cfm/person. These small increases from deletion of the economizer are consistent
with the expectation that the economizer will be operating only a small portion of the
time.
The percentage of total hours undercooled remain unchanged when the
economizer is eliminated, as would be expected.
The percentage of occupied hours at RH > 60% decline when the economizer
is deleted (from 1.2 to 0.6% at 5 cfm/person, and from 0.9 to 0.2% at 20 cfm/
person). The economizer substantially increases the inflow of humid outdoor air
during the first hour of operation on cool mornings — exactly the times when hours
of elevated RH occur. Elimination of the economizer thus would be expected to
reduce the hours at elevated RH.
5.15.2 The Effect of an Enthalpy-Controlled Economizer
As shown in Table 19, the results with the enthalpy-controlled economizer are
in between those with the temperature-controlled economizer and those with no
economizer. They are almost exactly equal to the results with the baseline
temperature-controlled economizer, indicating that the use of enthalpy control instead
of temperature control eliminates only a few hours of economizer operation.
The PSZ cooling capacity is unaffected by the switch to an enthalpy-controlled
economizer, as expected.
Use of an enthalpy-controlled economizer increases HVAC energy consumption
by 33 kWh/yr (from 3,245 to 3,278 kWh/yr), resulting in essentially no energy cost
5-72
-------
TABLE 19
Effect of Modifying Economizer Operation:
Increase Compared to Baseline Case with 5 cfm/person
Increase in
Increase in
total required
annual enerav consumption
Increase in annual enerav cost
Hours/yr
Occupied
coolina capacity
As % of
As % of
As % of
As % of
under-
hours/yr
In
HVAC
building
HVAC
building
cooled.
RH> 60%,
Case
kBtu/h As %
In kWh enerav
enerav
In $ enerav cost
enerav cost
%
%
Baseline values:
103.5581
26,1451
2,5101
0.4
1.2
T-controlled
(HVAC) --
--
(HVAC)
-
economizer.
60,1 611
4,273'
OA = 5 cfm/p
(building) --
~
(building)
-
(absolute values)1
OA ventilation rate
= 5 cfm/oerson
No economizer
0 0
505 1.9
0.8
23 0.9
0.5
0.4
0.6
Enthalpy-controlled
economizer
0 0
45 0.2
0.1
2 0.1
~0
0.4
0.8
OA ventilation rate
= 20 cfm/oerson
T-controlled
economizer
15.635 15.1
3,245 12.4
5.4
325 12.9
7.6
0.4
0.9
No economizer
15.635 15.1
3,654 14.0
6.1
344 13.7
8.1
0.4
0.2
Enthalpy-controlled
economizer 15.635 15.1 3,278 12.5 5.4 326 13.0 7.6 0.4 0.6
These baseline values for capacity, consumption, and cost are the absolute values from Table 3, not the increase from the baseline.
-------
increase ($326 - $325 = $1/yr} relative to the temperature-controlled economizer.
This result indicates that the conversion to enthalpy control has only a minimal impact
on the already-limited hours of economizer operation.
As expected, the enthalpy-controlled economizer has no impact at all on the
number of undercooled hours, relative to either the temperature-controlled economizer
or the case with no economizer.
Relative to the temperature-controlled economizer, the enthalpy-controlled
economizer decreases the percentage of occupied hours at RH > 60% from 1.2% to
0.8% at 5 cfm/person, and from 0.9% to 0.6% at 20 cfm/person. These reductions
correspond to 10 to 15 hours per year per zone. This is probably an approximate
indication of the number of hours by which the conversion to enthalpy control reduces
economizer operation. (Of course, eliminating the economizer altogether reduces the
number of elevated-RH hours even further.)
5.16 ALTERNATIVE COOLING ELECTRIC INPUT RATIOS
All of the preceding energy consumption and cost calculations assume that the
cooling system has an electric input ratio (EIR) of 0,341 Btu/h of electric input per
Btu/h of cooling output (except for the PTAC unit in Section 5.12, for which the EIR
is higher). This is a reasonably representative value, corresponding to an energy
efficiency ratio (EER) of 10 Btu/h cooling output per W of electric input.
Table 20 shows the effect on the baseline PSZ system of varying the cooling
EIR to: 0.284 Btu/h per Btu/h (EER =12 Btu/h per W), representing a very efficient
cooling system; and 0.427 Btu/h per Btu/h (EER = 8 Btu/h per W), representing a
relatively inefficient system.
5.16.1 The Effect of EIR at 5 cfm/person
Capacity. As expected, Table 20 shows that variations in EIR have no impact
on computed maximum cooling capacity. The efficiency with which the cooling
system converts electric input into cooling output has no impact on the parameters
that determine the amount of cooling output that is necessary. The weather- and
occupant-induced loads on the space remain unchanged, as do the coil bypass factor
and the heat added to the air stream by the air handler.
Energy consumption and cost. Improvement of cooling efficiency to an EIR of
0.284 results in a significant reduction in HVAC energy consumption (12.0%) and
cost (12.6%), compared to the baseline case (with EIR = 0.341). These reductions
at 5 cfm/person are greater than those from any of the other parameters considered
in this report, with the exception of:
5-74
-------
TABLE 20
Effect of Cooiing EIR:
Increase Compared to Baseline Case with 5 cfm/person
Case
Increase in
total required
coolina caoacitv
In
kBtu/h As %
Increase in
annual enercv consumDtion
As % of As % of
HVAC building
In kWh enerav enerav
Increase in annual enerav cost
As % of As % of
HVAC building
In $ enerav cost enerav cost
Hours/yr
under-
cooled,
%
Occupied
hours/yr
RH> 60%,
%
Baseline values:
103.5581
26J451
2,51c1
0.4
1.2
EIR = 0.341,
{HVAC) --
—
(HVAC)
—
—
OA = 5 cfm/p
60,1 611
4,27s1
(absolute values!1
(building) -
—
(building)
—
OA ventilation rate
= 5 cfm/oersori
EIR = 0.284
0 0
-3,139
-12.0
-5.2
-317
-12.6
-7.4
0.4
1.2
EIR = 0.427
0 0
4,736
18.1
7.9
477
19.0
11.2
0.4
1.2
OA ventilation rate
= 20 cfm/oerson
EIR = 0.341
15.635 15.1
3,245
12.4
5.4
325
12.9
7.6
0.4
0,9
EIR = 0.284
15.635 15.1
-433
-1.7
-0.7
-46
-1.8
-1.1
0.4
0.9
EIR = 0.427
15.635 15.1
8,794
33.6
14.6
2,647
105.5
61.9
0.4
0.9
These baseline values for capacity, consumption, and cost are the absolute values from Tabte 3, not the increase from the baseline.
-------
conversion to PVAVS, which also gives a 12% reduction in energy consump
tion, although only a 7% reduction in cost (Section 5.12.1);
- conversion to highly efficient lighting and equipment, giving 19% reduction in
consumption and 17% reduction in cost (Section 5.4.1); and
achieving infinite thermal resistance in the walls, windows, and roof (a
hypothetical situation), giving 20% reduction in consumption and 22%
reduction in cost (Section 5.10).
Conversion of the PSZ system to cold-air distribution (Section 5.14.1) gives similar
but slightly smaller reductions (11 % in energy consumption and 11 % in cost).
Thus, accepting an increase in installed costs for the more efficient cooling
equipment would result in a 12% to 13% reduction in energy consumption and cost
each year at 5 cfm/person. Of the parameters listed above that provide the greatest
energy savings, the conversion to a more efficient cooling system (and/or conversion
to PVAVS) may be the most readily and practically achieved. At the indicated energy
cost savings of $317 per year for the study building with the two PSZ units, it would
take perhaps 7 years to recover the increased equipment costs for the more efficient
units.
Just as a reduction in EIR results in one of the most significant decreases in
energy consumption and cost, an increase in EIR to 0.427 causes one of the most
significant increases in energy consumption and cost. As shown in Table 20, the
higher EIR results in an 18% increase in annual energy consumption and a 19%
increase in cost. No other single parameter causes such an increase in energy use at
5 cfm/person, with the exception of the use of inefficient lighting and equipment
(Section 5.4.1). With an energy cost penalty of $477 per year for using the less
efficient units in this building, the initial savings in equipment costs that would be
achieved by using the lower efficiency units would be consumed within perhaps 2
years. The use of lower-efficiency units would not appear cost-effective.
Hours at elevated temperature and RH. Table 20 shows that the percentages
of undercooled hours and of hours above 60% RH do not change at all as EIR is
varied. Variation of the EIR changes only the computed amount of electric power
required to achieve the computed amount of cooling. There are no other changes to
the space or the system: hourly zone temperatures and humidities remain the same,
as do the hourly values of, e.g., total and latent cooling provided by the system,
cooling coil and supply air temperatures, bypass factor, and supply air humidity.
Therefore, EIR is not calculated to create any change in hours undercooled or at
elevated RH.
5-76
-------
5.16.2 The Effect of EIR on Increased Ventilation Rates
Capacity. Table 20 shows that, at 20 cfm/person - as at 5 cfm/person - the
cooling EIR has no impact on the computed maximum cooling capacity, since none of
the parameters that impact the capacity calculation are influenced by the EIR.
Energy consumption and cost. As indicated in Table 20, the PSZ system with
EIR = 0.284 could provide 20 cfm OA/person while consuming 433 kWh/yr less
power and costing $46/yr less than the baseline PSZ system (EIR = 0.341) at 5
cfm/person. That is, OA ventilation rate could nominally be increased while achieving
a 1.7% to 1.8% reduction in HVAC energy consumption and cost, if the ventilation
rate increase were accompanied by conversion of the system to a higher-efficiency
system.
The only other parameters which provided a reduction in energy consumption
or cost (relative to the baseline) when the ventilation rate is increased are:
conversion to PVAVS, providing a 0.3% reduction in consumption, although an
increase in cost (Section 5.12);
conversion to highly efficient lighting and equipment, providing a 5% to 7.5%
reduction (Section 5.4);
reduction in the number of occupants by one-half, providing a 1.5% to 1.7%
reduction (Section 5.3); and
- the hypothetical case of achieving infinite thermal resistance in the walls,
windows, and roof, providing a 7% to 8% reduction (Section 5.10).
It is of interest to compare the consumptions and costs involved with providing
20 cfm/person using the more efficient PSZ (EIR = 0.284), with the values involved
with providing 20 cfm/person using the system having the baseline EIR (0.341).
Referring to Table 20, at 20 cfm/person, the energy consumption with the more
efficient system is 3,245 - (-433) = 3,678 kWh/yr less than the consumption with
the system having the baseline EIR (a 12.5% reduction in annual consumption). The
annual energy cost is $325 - (-$46) = $371 less for the study building (a 13%
reduction). With this annual energy cost savings at 20 cfm/person, it would take
perhaps about 6 years to recover the increased equipment costs for buying the more
efficient units. As would be expected, this recovery period is somewhat less than the
roughly 7 years at 5 cfm/person (Section 5.16.1).
With the less effective EIR of 0.427, Table 20 shows that, at 20 cfm/person,
energy consumption increases by one-third (by 8,794 kWh/yr) and energy cost
approximately doubles (increasing by $2,647/yr), compared to the baseline (EIR =
5-77
-------
0.341) system at 5 cfm/person. No other parameter causes the 20 cfm/person
system to increase in cost so dramatically relative to the baseline, with the exception
of the use of inefficient lighting and equipment (Section 5.4.2).
The use of the less efficient cooling equipment in this building would be very
cost-ineffective at the 20 cfm/person OA ventilation rate. Compared to the PSZ units
with the baseline EIR (0.341) at 20 cfm/person, the units with the poorer EIR (0.427)
at 20 cfm/person consume 8,794 - 3,245 = 5,549 kWh/yr more power at an
increased cost of $2,647 - $325 = $2,322/yr. With this increase in annual power
cost, the savings in equipment cost that would be achieved by buying the less
efficient equipment would be consumed in perhaps half a year.
Hours at elevated temperature and RH. Table 20 shows that -- at 20
cfm/person, as at 5 cfm/person - the percentages of undercooled hours and of hours
above 60% RH do not change at all as EIR is varied at constant OA flow rate.
5.17 ALTERNATIVE COOLING CAPACITIES AND SENSIBLE HEAT RATIOS
During the variation of the parameters discussed previously, the system cooling
capacity has been allowed to default to the value computed by the DOE-2 program;
i.e., to the capacity computed as being required to meet the peak cooling load. As
indicated in the preceding tables, these default capacities have varied somewhat,
depending upon the building or HVAC system parameters being varied.
For the baseline configuration with 5 cfm OA/person, as shown in Table 3, the
total capacity (both zones combined) at Air-Conditioning and Refrigeration Institute
-------
TABLE 21
Effect of Varying Cooling Capacity and SHR at 5 cfm/person
% of Total Hours % of Occupied Hours
Undercooled at RH > 60%
Total Cooling HVAC Energy Cost ($/vr) SHR = SHR = SHR = SHR = SHR = SHR =
Capacity1 (tons) SHR = 0.78 SHR = 0.752 SHR = 0.73 0.78 0.752 0.73 0.78 0.752 0.73
8.5 2,502 - 2,473
8.62 -- 2,510
9.0 2,541 -- 2,527
9.5 2,562 - 2,551
10.0 2,584 -- 2,581
0.4 -- 1.2 1.2 -- 1.2
0.4 -- -- 1.2
0.1 -- 0.3 1.2 - 1.2
-0 -- 0.1 1.2 -- 1.2
~0 -- -0 1.2 -- 1.2
1 The combined cooling capacities of the front and rear PSZ units, expressed as tons of refrigeration. With total capacities
of 9,0 and 10.0 tons, the front and rear units are of equal capacity; with total capacities of 8.5 and 9.5 tons, the rear unit
is 0.5 tons larger than the front unit. The rated capacities and SHRs are at the conditions specified by the Air-Conditioning
and Refrigeration Institute.
2 Default values of cooling capacity and SHR computed by DOE-2 when these values are not set in the input file. These are
the "baseline" values used elsewhere in this report. With the baseline system, the front and rear zones are each served
by a nominal 4.3-ton unit.
-------
HVAC energy cost. As shown in Table 21, the HVAC energy cost increases
slightly with increasing capacity. There are two reasons that this occurs.
First, the units having greater capacity are better able to address the peak
cooling loads that are encountered on summer Monday mornings, and they consume
slightly more power in doing so. But the baseline (8.6 ton) system already handles
most of the peak-load hours, with only 0.4% of hours being undercooled (correspond-
ing to about 30 hours per year per zone). As a result, the increase in energy costs
involved in going to greater-capacity systems (and reducing the undercooled hours to
even smaller percentages) are small.
The second reason that power consumption is greater with the larger-capacity
units is that, on a given summer day, the larger units tend to keep the office space
a couple tenths of a degree cooler.
These effects are small. As shown in Table 21, increasing the total capacity to
9.5 or 10.0 tons - by which point the number of undercooled hours is reduced almost
to zero - increases energy costs by only $41 to $74/yr above the $2,510 baseline
value, an increase of only 1.5% to 3%.
Of the increase in cost due to capacity increase, 70% to 80% results from
increases in the monthly power demand (per-kW) charges, and the remainder due to
an increase in energy (per-kWh) charges. The increase in demand charges consists
of a $1 to $2 per month increase during the winter months, and a $2 to $5 per month
increase during the summer months. The increase in energy charges consists of a $0
to $1 per month increase during the winter months, and a $1 to $2 per month
increase during the summer months.
As shown in Table 21, at a given total capacity, the system having the SHR of
0.73 consistently consumes less power annually than the one having the SHR of
0.78. This occurs because the system having the lower SHR has less sensible cooling
capacity. It thus performs somewhat less sensible cooling throughout the year,
resulting in more hours undercooled. Because system operation is controlled by
sensible demand, the lower-SHR system winds up performing less latent cooling as
well, on many days. Correspondingly, at the 5 cfm/person OA rate, the lower-SHR
system consumes slightly less power on an annual basis.
To illustrate this impact of SHR on power costs, consider the 9.5-ton system
during the mid-summer week of July 8 through 1 2. During this week, the entire
system (front and back units combined) would provide 1 5,700 Btu less total cooling
(and 8,700 Btu less latent cooling) if the SHR were 0.73 instead of 0.78. Of course,
the effect is small; given the EER assumed for the cooling unit (10 Btu/h of cooling
output per W of electric input), 15,700 Btu of cooling corresponds to only 4.6 kWh
of power consumption over an entire week. But it illustrates why, over the course of
5-80
-------
a year, the SHR value would create the small energy cost differences seen in Table
21 at 5 cfm OA/person.
This effect of SHR is smaller during milder weather. For example, during the
week of March 18 through 22, the 9.5-ton system operating at 5 cfm/person would
provide only 1,000 Btu less total cooling if the SHR were 0.73 instead of 0.78.
Latent cooling would be essentially identical with either SHR.
This impact of capacity and SHR on power consumption is further addressed
below in the discussion of hours undercooled.
Hours at elevated temperature and RH. As shown in Table 21, at constant SHR,
increasing the total capacity systematically reduces the percentage of total hours
undercooled. This results, of course, because -- as sensible capacity increases -- the
system is better able to respond to the high sensible loads immediately after startup
on warm mornings, when undercooled hours occur.
At constant capacity, decreasing the SHR from 0.78 to 0.73 consistently
increases the percentage of undercooled hours. As indicated previously, decreasing
the SHR decreases the sensible cooling capacity, rendering the system less able to
handle the high sensible loads on summer mornings.
To illustrate, consider the first four hours in the front office after startup on
Monday, July 8. In the baseline case at 5 cfm/person, the front office was under-
cooled during those four hours (averaging 77.5 °F). The 4.3-ton PSZ unit (SHR =
0.75) conditioning the front office in the baseline case - which is rated at 3.2 tons
sensible capacity and 1.1 tons latent capacity - provided a 194,400 Btu of total
cooling during that four-hour period, of which 21,400 Btu were latent cooling.
In comparison with the baseline, cooling increases significantly during these four
hours when total building cooling capacity is increased to 9.0 tons at SHR = 0.78.
In this case, the unit conditioning the front office is increased from 4.3 to 4.5 tons;
the rated sensible capacity of this larger unit is 3.5 tons, and the latent capacity 1.0
ton. With the larger unit, the front office is not undercooled during any of the first
four hours on July 8 (the average temperature being 76.4 °F), since the total cooling
provided by the system during those hours increases to 213,300 Btu (about a 10%
increase over the cooling provided by the baseline unit).
The latent cooling with the larger (4.5-ton/0.78 SHR) unit increases to 22,800
Btu, a 6.5% increase over the baseline (4.3-ton/0.75 SHR) case. This increase in
latent cooling occurs despite the fact that - with the lower SHR - the smaller,
baseline front unit actually has a slightly greater rated latent capacity (1.1 tons vs. 1.0
ton). The operation of the cooling coils is dictated by the sensible cooling demands,
and the ability of the larger unit to respond to the high sensible load during those first
5-81
-------
four hours results in increased latent cooling, despite the reduced latent capacity of
the larger unit.
When the SHR of the 9.0 ton system is reduced from 0.78 to 0.73 -- increasing
the latent capacity but reducing the sensible capacity - the larger system still has a
greater sensible capacity than does the smaller, baseline system at SHR = 0.75.
With SHR = 0.73, the rated sensible capacity of the 4.5-ton unit conditioning the
front office is 3.3 tons, and the latent capacity is 1.2 tons. With this slightly greater
sensible capacity relative to the baseline (3.3 vs. 3.2 tons), the 4.5-ton/0.73 SHR unit
conditioning the front zone is better able to reduce office temperature during the first
four hours on July 8; only two of the four hours are undercooled (average temperature
77.2 °F). This is somewhat better than the 4.3-ton./0.75 SHR baseline unit (four
hours undercooled, 77.5 UF average), but not as effective as the 4.5-ton/0.78 SHR
unit (zero hours undercooled, 76.4 °F average). Power consumption and cost
naturally follow this same relationship. The 4.5-ton/0.73 SHR unit provides total
cooling of 200,900 Btu to the front zone during the first four hours ~ 3% more than
the baseline 4.3-ton/0.75 SHR unit, but 6% less than the 4.5 ton/0.78 SHR unit.
Again, the latent cooling provided by the units tracks the sensible cooling that
is provided, rather than the rated latent capacity of the units. The latent cooling
provided to the front zone during the first four hours by the 4.5-ton/0.73 SHR unit
(latent capacity 1.2 tons) is 21,900 Btu - 2.5% more than the 21,400 Btu provided
by the 4.3-ton/0.75 SHR unit (latent capacity 1.1 tons), but 4% less than the 22,800
Btu provided by the 4.5-ton/0.78 SHR unit (latent capacity 1.0 ton). The greater the
sensible capacity of the unit, the more vigorously it responds to the high sensible
loads during startup on warm mornings, and, as a result, the greater the latent cooling
it provides. In this example, the unit with the least latent capacity winds up providing
the greatest amount of latent cooling. The unit with the greatest latent capacity is
only in the middle, in terms of latent removal during these hours. The 4.5-ton/0.73
SHR unit would appear to be over-designed for latent removal, and - at least in
comparison with the 4.5-ton/0.78 SHR unit - under-designed for sensible removal
during peak hours.
Table 21 shows no impact of capacity or SHR on the percentage of occupied
hours at RH > 60%. This effect results because the hours at elevated RH occur after
startup on cool, humid mornings, during the initial hours before the cooling coils have
come on (or when the cooling coils are operating well below capacity). Therefore,
increasing total cooling capacity (or increasing the latent capacity via a decrease in
SHR) would not be expected to impact the percentage of elevated-RH hours. Hours
above 60% RH are not being caused by inadequate total or latent cooling capacity.
Hours at elevated RH do not occur during hot months in the Miami climate,
when an increase in cooling capacity might otherwise have an impact. In the earlier
discussion, it was indicated that the latent cooling by the unit serving the front office
5-82
-------
during the first four hours on July 8 could vary between 21,400 and 22,800 Btu,
depending upon the capacity and SHR of the systems considered. But even the lesser
amount of cooling is sufficient to keep the front office below 50% RH during the first
hours after startup.
5.17.2 The Effect of Cooling Capacity on Increased Ventilation Rates
Table 22 shows the estimated effect of varying cooling capacity and SHR at a
ventilation rate of 20 cfm/person.
HVAC energy cost. Table 22 shows that -- at 20 cfm/person as at 5 cfm/
person -- energy costs increase slightly as cooling capacity increases. As at 5
cfm/person, this occurs for two reasons; a} additional cooling by the larger units in
reducing the percentage of undercooled hours below 0.4%; and b) additional cooling
because the larger units tend to keep the office a couple tenths of a degree cooler on
some summer days.
As at 5 cfm/person, the effects of increased capacity on power consumption
and cost at 20 cfm/person are small. Increasing the total cooling capacity of the two
PSZ units from the 9.9 tons to 10.5 or 11.0 tons increases annual energy costs by
only $40 to $66, or about 1.5% to 2.5% of the power cost for the 9.9-ton system.
This is about the same cost increase that was computed for a comparable capacity
increase in the 8.6-ton system at 5 cfm/person.
Of the increase in cost due to capacity increase at 20 cfm/person, 60% to 75%
results from increases in the monthly power demand (per-kW) charges, with the
remainder due to an increase in energy (per-kWh) charges. By comparison, at 5
cfm/person, 70% to 80% of the increase was due to demand charges. At 20 cfm/
person, the increase in demand charges consists of a $0 to $2 per month increase
during the winter months, and a $3 to $4 per month increase during the summer
months. The increase in energy charges consists of no monthly increase during the
winter months, but a $2 to $3 per month increase during the summer months.
One difference between the results at 5 cfm/person (Table 21} and those at 20
cfm/person {Table 22) is that - at 20 cfm/person - the annual power costs for the
0.73 SHR unit are slightly greater (by $2 to $5) than those for the 0.78 SHR unit at
the two higher capacities. At 5 cfm/person, power costs with the 0.73 SHR unit
were always slightly less.
The reason for this small effect is that - when the inflow of humid outdoor air
is increased from 5 to 20 cfm/person — the latent load on the system throughout the
year is increased disproportionately relative to the increase in sensible load. This large
increase in the latent load makes more effective use of the increased latent capacity
of the 0.73 SHR system. In doing so, it causes the 0.73 SHR system at 20 cfm/per-
5-83
-------
TABLE 22
Effect of Varying Cooling Capacity arid SHR at 20 cfm/person
Total Cooling
Capacity1 (tons)
HVAC Energy Cost ($/vr)
SHR =0.78 SHR =0.73 SHR=0.702
% of Total Hours
Undercooled
SHR = SHR = SHR =
0.78 0.73 0.702
% of Occupied Hours
at RH > 60%
SHR = SHR = SHR =
0.78 0.73 0.702
9.9?
10.0
10.5
11.0
2,835
0.4
0.9
2,856
2,875
2,896
2,853
2,877
2,901
0.1
~0
0
0.2
0.1
~0
0.9
0.9
0.9
0.9
0.9
0.9
1 The combined cooling capacities of the front and rear PSZ units, expressed as tons of refrigeration. With total capacities
of 10.0 and 11.0 tons, the front and rear units are of equal capacity; with a total capacity of 10.5 tons, the rear unit is
0.5 tons larger than the front unit. The rated capacities and SHRs are at the conditions specified by the Air-Conditioning
and Refrigeration Institute.
2 Default values of cooling capacity and SHR computed by DOE-2 for the baseline system operating at 20 cfm/person, when
these values are not set in the input file. With the baseline system operating at 20 cfm/person, a nominal 4.9-ton unit
serves the front zone and a 5.0-ton unit serves the rear.
-------
son to consume proportionally more power, relative to the 0.78 SHR system, than it
did at 5 cfm/person.
To illustrate this, consider the 10.5-ton system operating at 20 cfm/person
during the week of July 8 through 12. In previous discussion, it was indicated that -
with the 9.5-ton, 5 cfm/person system - decreasing the SHR from 0.78 to 0.73
reduced the total cooling provided by the system by 15,700 Btu during that summer
week {and reduced the latent cooling by 8,700 Btu). By comparison -- with the 10.5-
ton, 20 cfm/person system - decreasing SHR from 0.78 to 0.73 decreases the total
cooling by only 3,500 Btu during that same week (and actually increases the latent
cooling by 1,600 Btu).
During the week of March 18 through 22, decreasing the SHR of the 9.5-ton,
5 cfm/person system from 0.78 to 0.73 reduced the total cooling by only 1,000 Btu
(and did not change the latent cooling at all). With the 10.5-ton, 20 cfm/person
system, reducing the SHR results in an increase of 5,900 Btu in total cooling, and an
increase of 6,500 Btu in latent cooling, over the course of that March week.
Clearly, then, the explanation for why Table 22 shows the 0.73 SHR system
with slightly higher annual power costs than the 0.78 SHR system at 20 cfm/person
is that the decreases in total cooling with reduced SHR during high-cooling weeks
(e.g., July 8-12) are not as large as they were at 5 cfm/person. And, these smaller
summer decreases are offset - or slightly more than offset - by increases in cooling
by the 0.73 SHR system during milder weeks (e.g., March 18-22). The increases in
latent cooling by the 0.73 SHR system at 20 cfm/person are largely responsible for
why the 0.73 vs. 0.78 SHR power consumption relationship is different at 20 cfm/
person than it was at 5 cfm/person.
But as shown in Table 22, the power cost differences between the 0.73 and
0.78 SHR systems are tiny. At an EER of 10, the 3,500 Btu reduction in total cooling
provided by the 0.73 SHR system at 20 cfm/person during July 8-12 corresponds to
a reduction in power consumption of only 0.35 kWh over the course of the week.
The 5,900 Btu increase in total cooling provided by the lower-SHR system at 20 cfm/
person during March 18-22 corresponds to an increased power consumption of only
0.59 kWh for the week. Over the course of a year, these tiny differences add up to
only a few dollars in power costs.
Hours at elevated temperature and RH. The trends shown in Table 22 for the
20 cfm/person cases are the same as those in Table 21 for 5 cfm/person.
At a given SHR, an increase in total capacity {and hence in sensible capacity)
always results in a reduction in the number of hours undercooled. A unit having
higher sensible capacity is better able to handle the peak-load hours immediately after
startup on warm Monday mornings, when the undercooled hours occur. And, at a
5-85
-------
given total capacity, the percentage of undercooled hours is lower for the higher-SHR
system, since the higher-SHR system has greater sensible capacity.
Also, increases in the total capacity or adjustment of the SHR has no impact on
the percentage of occupied hours at RH > 60%. These elevated-humidity hours
occur on relatively cool winter days, when the system is operating well below
capacity. Thus, increases in total or latent capacity will not significantly reduce the
number of elevated-RH hours.
5.18 ALTERNATIVE WEATHER INPUT FILES
For computation of the atmospherically induced loads on the building being
modeled, the DOE-2 program requires weather on an hourly basis for the entire year
at the building's location. The weather parameters of concern include, for example,
dry- and wet-bulb temperatures, wind speed, cloud cover, and total horizontal solar
radiation).
Over the years, hourly weather data have been compiled via several different
procedures, for use in various applications (such as DOE-2 modeling). These alterna-
tive procedures vary according to how they attempt to represent typical weather
conditions at the location.
Typical Reference Year (TRY) - The data in TRY weather files describe a single year
that has been selected from all of the years for which data are available. This
one year has been selected as being the most typical overall for the particular
location. For example, 1964 is the selected reference year for Miami. Since
any one year is almost certain to have some weeks or months that vary
significantly from the multi-year average, TRY weather data are viewed as being
potentially useful for comparing one building or HVAC design against another,
but as not being suitable for estimating average energy requirements over
several years.
Typical Meteorological Year (TMY) - The TMY file contains a typical year that has
been compiled by selecting the 12 most typical months from the various years
covered by the available weather data. For example, the TMY file for Miami
incorporates the January weather data from 1962, the February data from
1974, the March data from 1967, etc.
Weather Year for Energy Calculations (WYEC} - WYEC files take the TMY concept
a step further. The most typical months are still selected from different years,
but - when a given individual day within a typical month is considered not to
be representative -- the weather data for that day is replaced by the data for
that same day from another year (ASHRAE, 1993). WYEC files thus best
reflect the long-term mean weather conditions.
5-86
-------
All of the previous results in this report were computed using the TMY weather
file for Miami. To assess the impact of the weather file, the baseline conditions were
re-run using the WYEC file for Miami.
The results using the two weather files are compared in Table 23.
Capacity. As shown in Table 23, the WYEC weather file results in higher
computed cooling capacities for both OA flow rates (by 5.6% to 12.7%). This seems
to occur in part because the WYEC file incorporates a higher level of solar radiation
on the days that are used by the program in computing the capacity.
To illustrate this effect, consider the front office. The cooling capacity of the
unit serving this office is computed for the hour when -- according to the LOADS
portion of the DOE-2 program - the peak cooling load is imposed on this space by the
weather, by internal sources (occupants, lights, equipment), and by infiltration.
For the WYEC weather file, this hour for the front office is 5 pm on August 2
(with both OA flow rates). During that hour, the WYEC file predicts that the outdoor
dry-bulb temperature will be one degree warmer than the TMY file predicts (88 vs.
87 °F). As a result, conduction through the walls, windows, and roof is a few percent
greater with WYEC (in terms of kBtu conducted into the office and overhead plenum
during that hour). But the increase in the solar component with WYEC (radiation
through the window glass, and absorption by the walls and roof) is even greater; solar
gain through the front windows increases 20% relative to the TMY conditions (from
2.13 to 2.57 kBtu during that hour). With the WYEC file, the total heat entering the
front office and plenum by conduction and radiation from outdoors is 16.26 kBtu
during the hour. The increased temperature of the outdoor air also increases the
sensible heat added by infiltration, and by the OA mechanically supplied by the
system.
Under these conditions at 5 pm on August 2, based on the WYEC file, the
program computes a required ARI-rated cooling capacity for the front system of 53.61
kBtu/h at 5 cfm/person, or 66.11 kBtu/h with the higher OA load at 20 cfm/person.
By comparison, with the baseline TMY weather file, the hour with the peak load
in the front office (from LOADS) is 5 pm on June 14. This hour is selected primarily
because the TMY file records the outdoor dry-bulb temperature as being 91 °F during
that hour {one of the hottest hours during the year). The WYEC file records an
ambient temperature of only 82 °F during that hour. The high TMY temperature
increases heat conduction through the walls, windows, and roof, relative to the WYEC
case. With the TMY file, the total heat entering the front office and plenum by
conduction and radiation from outdoors is 15.67 kBtu during the 5 pm hour on June
14 {slightly less than the 16.26 kBtu indicated above for the WYEC case at 5 pm on
August 2). The increased temperature of the outdoor air in the TMY case on June 14
5-87
-------
TABLE 23
The Effect of the WYEC vs. the TMY Weather File on the Results
for the Baseline Small Office in Miami
Output Variable
Value of Output Variable at 5 cfm/person
%
TMY File WYEC File Difference
Value of Output Variable at 20 cfm/person
%
TMY File WYEC File Difference
Total required cooling
capacity (kBtu/'h)
Annual HVAC energy
consumption (kWh)
Annual HVAC energy
costs($)
Annual building energy
consumption (kWh)
Annual building energy
costs{$)
% of all hours (8760
hr/year) undercooled
% of occupied hours (3276
hr/year) when RH>60%
103.558 109.336
26,145
2,510
60,161
4,273
0.4
1.2
25,338
2,477
59,406
4,244
0.3
1.4
+ 5.6
-3.1
¦1.3
-1.3
-0.7
-0.1
+ 0,2
119.193 134.308 +12.7
29,390
2,835
4,598
0.4
28,531
2,817
63,406 62,599
0.9
4,584
0.2
1.1
-2.9
-0.6
-1.3
-0.3
-0.2
+ 0.2
-------
also increases the sensible heat added by infiltration, and by the OA mechanically
supplied by the system.
Under these conditions at 5 pm on June 14, based on the TMY file, the
program computes a required ARI-rated cooling capacity for the front system of 51.42
kBtu/h at 5 cfm/person, or 59.18 kBtu/h with the higher OA load at 20 cfm/person.
Consistent with the results in Table 23, these TMY capacities for the front unit are
4% to 12% smaller than the WYEC capacities indicated earlier.
It is of interest to note that the increased conduction of heat into the space in
the TMY case at 5 pm on June 14 (resulting from the high TMY outdoor air
temperature) is partially offset by increased solar radiation in the WYEC case. Solar
gain through the front windows during that hour with the WYEC weather file is 6%
greater than in the TMY case (2.62 vs. 2.48 kBtu). And even though there is a much
greater nominal temperature differential between the outdoor air and the air inside the
plenum in the TMY case at 5 pm on June 14 (19 vs. 10 F°), there is nevertheless
18% more heat conduction across the roof during that hour in the WYEC case (10.08
vs. 8.57 kBtu) due to increased solar absorption on the roof.
Energy consumption and cost. Table 23 indicates that -- despite the greater
capacities of the PSZ units in the WYEC case - the computed HVAC power consump-
tions are lower by about 3% at both OA flow rates, and HVAC power costs are lower
by about 1 %, when the WYEC weather file is used.
An analysis of hourly data from a variety of days over the course of the year
confirms that the reduced power consumption with the WYEC file results because, on
balance over the year, the WYEC weather data impose a slightly lesser load on the
system during the hours of system operation. Thus, while the load during the single
peak hour in Miami is greater for WYEC — resulting in a greater computed capacity for
the system under WYEC - the load over the entire year is slightly less with the WYEC
file.
To illustrate this effect, consider the week of July 8 through 12. On Monday,
July 8, the TMY data predict higher outdoor dry-bulb temperatures in Miami than do
the WYEC data (averaging 1.6 F° higher over the 13-hour period while the system is
operating). As a result, the TMY file results in greater total cooling by the system on
July 8. At 5 cfm/person, the total cooling with the TMY file is 1,225 kBtu (front and
rear units combined), about 5% greater than the 1,170 kBtu predicted by the WYEC
file. At 20 cfm/person, the TMY file also estimates a 5% larger number (1,470 kBtu
vs. 1,400 kBtu).
The following Friday, July 1 2, the situation is reversed. The WYEC file now
predicts the higher outdoor dry bulb temperature (by 4.3 F° on average), and the
WYEC file now results in 6% to 8% greater total cooling on that day (950 vs. 900
kBtu at 5 cfm/person, 1,210 vs. 1,110 kBtu at 20 cfn/person).
5-89
-------
For the entire week of July 8 through 12, it turns out that the TMY file results
in 2% greater total cooling provided at 5 cfm/person (5,010 vs. 4,930 kBtu), while
the WYEC file results in 0.2% greater cooling at 20 cfm/person (6,120 vs. 6,110
kBtu).
A more dramatic effect is seen during the winter week of January 14 through
18. During that week, the TMY file assumes mid-day outdoor dry-bulb temperatures
above 75 °F. The cooling coils are typically off - and, as necessary, the economizer
operating -- for perhaps the first couple hours after startup. But after the first couple
hours, the coils come on and the economizer is locked out. With the TMY file, the
total cooling provided by the system during that week (front and rear units combined)
is 2,960 kBtu at the 5 cfm/person OA rate.
By comparison, with the WYEC weather file, outdoor dry-bulb temperatures
during the week of January 14 through 18 are much lower, hitting highs of 46 °F on
Monday and 64 °F on Friday. For three of the days that week, the cooling coils never
come on at all in either office, with the economizer operating all day. With the WYEC
file, the total cooling provided during that week at 5 cfm/person is only 1,190 kBtu,
only 40% of the value estimated with the TMY file.
The WYEC file generally results in a greater amount latent cooling during any
given warm day or week. This occurs even on days when the TMY file assumes a
higher moisture content in the outdoor air. The greater latent cooling occurs because,
during warm weather, the WYEC system tends to operate at coil surface temperatures
averaging 1 to 2 F° colder than the TMY system, thus condensing more moisture,
even when the computed average supply air temperature leaving the coils is similar
for the two weather files.
Hours at elevated temperature and RH. As shown in Table 23, the WYEC file
results in a small reduction in the percentage of hours undercooled at both OA flow
rates (from 0.4% of all hours, to 0.2-0.3%). This results because the PSZ units have
a somewhat greater capacity in the WYEC case, and are thus better able to handle the
elevated coil inlet temperatures during summer Monday morning startups, when the
undercooled hours occur.
For example, consider the morning of Monday, July 8. With the 103.558
kBtu/h system at 5 cfm/person computed using the baseline TMY weather file, the
front office remains undercooled -- i.e., more than 2 F° above the 75 °F setpoint
temperature - for the first 4 hours after startup, and the rear office remains
undercooled for the first 2 hours. But with the 109.336 kBtu/h system at 5
cfm/person computed using the WYEC file, the front office is undercooled for only 1
hour that morning, and the rear office is not undercooled during any hour.
5-90
-------
However, with the exception of these initial hours after startups on warm
Monday mornings, the PSZ units maintain the office space at almost exactly the same
temperatures during the other hours, regardless of which weather file is used. Office
temperatures maintained by the system during any given hour remain within 0.3 P
of each other as the weather files are changed, with the higher office temperature
usually being predicted by the weather file that assumes the higher outdoor
temperature.
As shown in Table 23, the WYEC file results in a slight increase in the
percentage of occupied hours when the RH is greater than 60% (e.g., 1.4% vs. 1.2%
of occupied hours at 5 cfm/person). This small effect appears to result because the
WYEC file often assumes a lower outdoor temperature on cool mornings than does
the TMY file, with the result that the economizer is operating more often in the WYEC
case (bringing in additional outdoor moisture with the cooling coils off). However, this
effect is partially offset by the fact that the WYEC file also often assumes a lower
moisture content in the outdoor air, so that this additional unconditioned outdoor air
does not contribute to indoor RH to the same extent that it would under TMY
conditions.
As discussed in earlier sections, the hours of elevated indoor RH tend to occur
during the first few hours after startup on cool mornings, when the outdoor
temperature in Miami is sufficiently low such that increased OA flow can provide all
of the cooling required by the space. The economizer lockout feature on these
packaged units dictates that the economizer will operate only during hours when the
cooling coils can be off entirely, so that there is an increased influx of potentially
humid outdoor air during these hours without any removal of the moisture content.
Whether this economizer-induced OA flow will result in indoor RH values above 60%
will depend upon, among other things, the amount of OA flow and the moisture
content of that OA.
As one example of why Table 23 would be showing additional elevated-RH
hours with the WYEC file, consider the morning of Monday, January 28. During the
first three hours after startup, the WYEC file assumes an average outdoor temperature
of 62.7 °F, and an average moisture content of 0.0120 lb moisture/lb dry air. With
the WYEC file, the economizer operates for those first three hours (i.e., the cooling
coils do not come on until the fourth hour) for the PSZ system serving the front office.
And, during those hours, the moisture content of the indoor air in the front office
ranges between 0.0119 and 0.0125 lb/lb — corresponding to indoor RH values of
63% to 67%.
But with the TMY file on January 28, the average outdoor temperature during
those first three operating hours is 69.7 °F, 7 F° warmer than in the WYEC case. (The
outdoor moisture content also happens to be higher, 0.0143 lb/lb.) Because of this
higher outdoor temperature, the program computes that the economizer cannot come
5-91
-------
on at all on January 28 in the TMY case. Thus, while the WYEC file results in the
coils being off entirely during those first three hours, the TMY file results in the coils
providing an average of about 26,700 Btu/h of total cooling, including a reasonable
amount of iatent cooling. Thus, despite the higher moisture content of the outdoor
air in the TMY case, the maximum moisture level in the indoor air of the front office
during those three hours is 0.0089 lb/lb - corresponding to an RH of 48%.
Of course, this situation is not universally true on all days. In a few cases, the
TMY file predicts the lower outdoor temperature, and the situation is reversed.
The WYEC file often assumes a lower moisture content in the outdoor air,
consistent with the lower outdoor temperature, which can sometimes offset the RH
impact of the increased WYEC economizer operation. An illustration of this effect can
be seen in the front zone on January 14.
On January 14, the WYEC file assumes that the outdoor temperature remains
between 43 and 46 °F during the entire day while the PSZ system is operating, and
the outdoor moisture content remains between 0.0034 and 0.0047 lb moisture/lb dry
air. With this cool outdoor temperature, the economizer is operating all day, and,
consequently, the cooling coils never come on. But because of the low moisture
content in the outdoor air, the indoor moisture content in the front office never
exceeds 0.0102 lb/lb (corresponding to an indoor RH of 57%). The indoor RH never
reaches as high as 60% on that day with the WYEC file.
But with the TMY file, the outdoor temperature at startup on January 14 is
68 °F, increasing to 78 °F at mid-day. The outdoor moisture content is generally at
values above 0.0135 lb/lb. Under these conditions, the economizer never comes on
during that day. But the TMY outdoor temperature is low enough such that the
cooling coils also remain off for the first three hours after startup, and are on at only
a reduced level during the fourth and fifth hours. As a result of the latent heat
released by the building occupants, the high moisture content in the OA, and the
absence of any latent removal by cooling coils, the RH in the front office is well above
60% during the first five hours on January 14 with the TMY file. In fact, during the
first four hours, it is computed to be 95% to 100%.
Thus, despite the fact that the WYEC file often results in a greater degree of
economizer usage - and despite the fact that, in humid climates, increased
economizer usage might intuitively be expected to lead to increased indoor RH -- the
above example for January 14 shows that, under some circumstances, the WYEC file
can nevertheless result in fewer hours at elevated RH.
The above examples for January 14 and 28 illustrate that various offsetting
phenomena occur with the switch from the TMY to the WYEC weather files, and that
the impact on indoor RH is not straightforward. However, as shown in Table 23, the
net effect over the course of an entire year is for the WYEC file to result in a slightly
greater number of occupied hours above 60% RH.
5-92
-------
SECTION 6
THE IMPACT OF HUMIDITY CONTROL
6.1 INTRODUCTION
A primary concern with increasing the OA flow rate in humid climates is the
impact that such an increase might have on the indoor RH levels. Outdoor RH levels
in Miami can often be well above 60% [the indoor maximum recommended by
ASHRAE to reduce the growth of fungi, such as molds and mildew (ASHRAE, 1989a),
and to maintain occupant comfort (ASHRAE, 1992a)]. An increased influx of such
humid OA -- combined with the latent energy being generated inside the building -
could result in a high latent load on the HVAC system, possibly resulting in indoor RH
levels exceeding 60% during system operation.
Thus, two primary questions arise when considering increased OA in humid
climates.
a) What will be the impact of such an OA increase on the RH levels in the space,
if no steps are taken to reduce RH?
b) What will the costs be of trying to maintain the space below 60% RH when the
OA is increased to 20 cfm/person?
6.1.1 Impacts of Increased OA on RH When No RH Reduction Steps are Taken
Table 3 in Section 4 predicts that -- when OA is increased from 5 to 20 cfm/
person with the baseline PSZ units - the percentage of occupied hours above 60%
RH will actually decrease, from 1.2% to 0.9%. The DOE-2.1 E modeling predicts that
elevated-RH hours will remain essentially unchanged - actually, decrease slightly --
when OA is increased.
As discussed in Section 5.12.2, this tiny predicted decrease in elevated-RH
hours with the PSZ units results because - at increased OA -- the cooling coils
operate at lower temperatures, and provide increased latent cooling that more than
offsets the increase in latent load created by the increased OA flow. Moreover, as
discussed in Section 5.17, the number of elevated-RH hours (at either 5 or 20 cfm/
person) is not impacted by increases or decreases in either the HVAC system capacity
or SHR, because the elevated-RH hours occur immediately after startup on cool
6-1
-------
mornings when the coils are operating far below capacity {or perhaps are not
operating at all}.
Some investigators indicate that this apparent non-effect on increased OA
results because the DOE-2.1E model does not properly simulate certain parameters
that will impact the indoor humidity (Shirey et al., 1995; Shirey and Rengarajan,
1996). Specifically, DOE-2.1E does not account for two phenomena.
a) The moisture capacitance of the of the building materials and furnishings. DOE-
2 does not account for moisture adsorbed due to the infiltration of humid air
when the system is off overnight and over weekends and holidays. This
moisture would be released when the system comes back on, providing an
incremental additional latent toad throughout the day.
b) The condensed moisture on the surfaces of the cooling coils, which can re-
evaporate into the circulating air stream when the coils cycle off (but when the
air handler continues to operate) during operating hours.
Shirey et al. utilize a model -- a broad building flow and energy model
incorporating a variation of the DOE-2 model - which addresses these moisture
sources. Their model predicts that, when OA in this small Miami office is increased
to 20 cfm/person, the percentage of occupied hours above 60% RH - rather than
remaining essentially unchanged - would increase to 11-23%. And these elevated-RH
hours would occur not only on cool days, but on hot summer days as welt.
Because the elevated-RH hours are predicted to sometimes occur during hot
days, when the system will be operating near capacity, the calculations by Shirey et
al. predict that changes in system capacity and SHR will impact the percentage of
occupied hours at elevated RH.
Because the DOE-2.1E model used in this study does not address moisture
capacitance and re-evaporation, it is not possible to independently assess the
significance of these phenomena under different conditions.
6.1.2 Approaches for Reducing the Impact of Increased OA on RH
In concept, two general approaches might be considered for reducing the
impact of the increased OA on space RH.
a) Incorporate an HVAC control system that relies solely on temperature control,
as usual. But design and operate the HVAC system such that - as the system
operates to control the temperature in the space — there will be as few hours
as possible having RH levels above 60%.
6-2
-------
b) Alternatively, incorporate RH control as well as temperature control into the
HVAC control system. The RH control could be achieved using, e.g.,
desiccants, or overcooling with reheat.
HVAC systems in general -- and especially systems serving small spaces, as
considered in this study -- generally function only with temperature control, and not
with RH control. Humidity control involves somewhat greater system installed cost,
greater energy costs, and higher maintenance requirements (which the owners of a
strip mall or small office building might not have the personnel to conveniently
provide). Thus, in practice, RH control is considered only in cases where it is critical
to the occupants of the space.
6.1.3 Limitations of the DOE-2 Model for RH Analysis
In this section, the DOE-2.1E model is used to assess some alternatives for
reducing the RH impacts of increased OA, based on the two general approaches listed
in Section 6.1.2. Primarily, the assessment here addresses the incorporation of an
enthalpy controller into the system, with over-cooling and reheating of the air to
control moisture (an option which falls under the second of the two approaches
above).
As indicated in Section 6.1.1, it is recognized that the DOE-2 model might not
fully model RH effects, since it does not simulate building moisture capacitance or re-
evaporation off the cooling coils. Thus, in the analysis here using that model, it must
be understood that the RH levels during a given hour could be higher than those
predicted here - i.e., the RH performance of the control option being evaluated could
be poorer, and the energy consumption/costs higher -- if the capacitance and re-
evaporation phenomena were considered. Despite these limitations, the DOE-2 results
here are felt to show, at least qualitatively, the relative magnitude of the effects of
RH control, and the possible potential of enthalpy control.
6.2 REDUCING RH THROUGH ADJUSTMENTS TO HVAC DESIGN USING
TEMPERATURE CONTROL ALONE
6.2.1 Results from Other Investigators
In the study conducted by the Florida Solar Energy Center, or FSEC (Shirey et
al., 1995; Shirey and Rengarajan, 1996), six HVAC design modifications were
considered to assess their potential for reducing the impact of increased OA on indoor
RH, on energy consumption and costs, and on system installed cost. This study
addressed a small office in Miami having a PSZ system, essentially identical to the
office and the system considered in this report.
6-3
-------
The six design modifications considered in that study were as follows.
a) Reduction of the PSZ sensible heat ratio (SHR) to 0.73.
b) An increase in the system cooling efficiency, from an EER of 10 to an EER of
12 {combined with reduction in the SHR to 0.73).
c) Use of demand-controlled ventilation (DCV), so that OA flow rate is reduced
accordingly to maintain the 20 cfm/person ratio when the number of occupants
decreases. This calculation assumed an accurate sensor of actual building
occupancy, based on C02 monitoring.
d} Use of an enthalpy recovery wheel, exchanging sensible and latent heat
between the incoming OA and the outgoing building exhaust, in an effort to
reduce the load on the system.
e) Use of a heat pipe, whereby a very efficient recuperative heat exchanger (based
on heat pipe technology} is used to reduce the temperature of the warm return
air entering the coils, by exchanging it against the cool supply air leaving the
coils. The principle is that - with the cooler entering air - the coils provide a
colder outlet air, thus condensing more moisture. The cooling coils provide
about the same amount of sensible cooling as they would without the heat
pipe, but operate at a lower temperature.
f) Addition of a 100% OA cooling system. In this approach, the two PSZ units
treating the zones of the office space are supplemented by a third high-
efficiency, very low-SHR unit designed specifically to cool and dehumidify the
inlet OA stream. This cooled/dehumidified OA is then introduced into the
supply air downstream of the two main PSZ units.
In all cases, the system operation was controlled by space temperature alone,
not by enthalpy.
Hours at elevated RH. In summary, the computations showed that - at 20
cfm/person — DCV, the enthalpy wheel, the heat-pipe system, and the 100% OA unit
were able to keep elevated-RH occupied hours down to about the same number as
that experienced by the baseline system (conventional PSZ, SHR = 0.78, EER = 10)
at 5 cfm/person (i.e., about zero). That is, if the increase from 5 to 20 cfm/person
were accompanied by one of these design modifications, the number of occupied
hours above 60% RH would hardly increase (or would not increase at all), even with
a temperature-based (rather than an enthalpy-based) control system.
By comparison, simply increasing the OA flow in the baseline system without
any modifications (other than to increase capacity) increased the percentage of
6-4
-------
occupied hours above 60% RH from about zero to about 16%. This is dramatically
different from the DOE-2 results presented in Table 3 (where the OA increase has
essentially no effect on elevated-RH hours), due to the capacitance and re-evaporation
issues discussed previously.
Reducing the system SHR to 0.73 would reduce the impact of the increased OA
somewhat, according to FSEC, so that - instead of 16% of occupied hours being
above 60% RH -- the value falls to 11%. But there is still a substantial increase in
elevated-RH hours at 20 cfm/person relative to the baseline case at 5 cfm/person
(zero hours). By comparison, the DOE-2 modeling in the current report predicts that
a decrease in SHR would have no effect on elevated RH hours at all (see Section
5.17), with the percentage of elevated-RH occupied hours remaining at 0.9-1.2%.
This results because, according to DOE-2, all of the elevated-RH occupied hours occur
after startup on cool Monday mornings, when the coils are operating at greatly
reduced capacity (if at all), and their SHR accordingly does not fully come into play.
This differs from the model used by FSEC, where - due to moisture capacitance and
re-evaporation — many of the elevated-RH occupied hours occur during warm weather,
when the SHR of the equipment plays a more important role.
In the calculations by FSEC, an increase in the EER in combination with a
decrease in the SHR provides no additional impact on the percentage of elevated-RH
hours. The high-EER case (with SHR = 0.73) results in the same percentage of
elevated-RH occupied hours (11 %) as the standard-EER case with SHR = 0.73. This
is consistent with the predictions of DOE-2, that EER (or EIR) should not have a
significant impact on elevated-RH hours (see Section 5.16).
Energy consumption and cost. According to the FSEC predictions, an increase
in OA from 5 to 20 cfm/person would increase annual HVAC energy consumption and
cost by approximately 9% in this Miami building, if no modifications were made to the
baseline system (other than to increase capacity). This is somewhat less than the
12.4% and 12.9% increase predicted by DOE-2 for HVAC energy consumption and
cost (respectively), shown in Table 3.
If the increase in OA were accompanied by installation of an enthalpy wheel,
HVAC energy consumption and costs would actually decline compared to the 5 cfm/
person baseline case, according to the FSEC estimates. Thus, the enthalpy wheel
would not only avoid any increase in elevated-RH hours resulting from the OA
increase, as discussed above, but would save energy in the process.
If the increase in OA were accompanied by a switch to the higher-EER coils,
FSEC's HVAC energy consumption and costs would again decline relative to the 5
cfm/person baseline. The DOE-2 results provided a similar prediction for the case of
the improved-efficiency coils (see Section 5.16.2). Although the improved-efficiency
6-5
-------
coils have no impact on RH performance, they do, of course, reduce energy
consumption and cost.
If the increase in OA were accompanied by a switch to DCV, the HVAC energy
penalty associated with the OA increase would decline from the 9% (with no
modifications), to 6-8%. Thus, DCV, if it could be implemented rigorously, would
reduce the energy penalty while, at the same time, substantially reducing the increase
in elevated-RH hours accompanying the increased OA.
If the OA increase were accompanied by a switch to lower-SHR coils, the
HVAC energy penalty would increase above that for the unmodified system, to 13-
14%, Thus, the modest reduction in the increase in elevated-RH hours with increased
OA is offset by an increase in the energy penalty of the reduced-SHR system.
The increased energy penalty predicted by FSEC with the lower-SHR coils
differs from the DOE-2 results presented in Section 5.17.2. DOE-2 predicted almost
no change in power consumption and costs at 20 cfm/person as SHR was decreased
from 0.78 to 0.73. Actually, DOE-2 did predict a tiny increase in energy cost at lower
SHR, due to increased latent cooling energy consumption during mild weather. But
this increase was insignificant compared to the effect predicted by FSEC.
If the increase in OA were accompanied by the addition of a heat pipe system,
or of a 100% OA unit, the HVAC energy penalty would be substantially greater than
the 9% with the unmodified baseline system, according to the FSEC estimates. The
heat pipe system would increase the penalty to about 27%, the 100% OA unit to 21-
25%. Thus, while both of these options are very effective at reducing or eliminating
hours at elevated RH, they have a significant energy penalty.
Installed cost. The installed cost estimates prepared in conjunction with the
FSEC modeling predicted that an increase from 5 to 20 cfm/person for the baseline
system (with no modifications other than a capacity increase) would increase the
installed cost of the system by less than 2%, resulting from the capacity increase.
All six of the system modifications considered by FSEC involved an increase in
the installed cost at 20 cfm/person significantly greater than the baseline 2%.
The enthalpy wheel had the least increase in installed cost (11%). Thus, the
only one of the six options predicted to simultaneously eliminate elevated-RH hours
and save energy when OA is increased, also turned out to be the least expensive to
install, under FSEC's assumptions.
Reduction of the SHR to 0.73 had the next lowest installed cost increase, 14%.
But, as discussed earlier, this option gave only a modest reduction in the increase in
6-6
-------
elevated-RH hours, and had a larger energy penalty than did the 20 cfm/person
baseline.
Simultaneously increasing the EER, along with the SHR decrease, resulted in an
installed cost increase of 21 %. The high-EER/low-SHR case resulted in an energy cost
savings at 20 cfm/person, due to the improved coil efficiency, but it still provided only
the modest reduction in the increase in elevated-RH hours. And the installation cost
penalty is high.
Conversion to DCV resulted in an installed cost increase of just over 14%,
about the same as the reduced-SHR option. So DCV -- if it could be implemented
rigorously -- would greatly reduce the elevated-RH hours resulting from the OA
increase, would slightly reduce the associated energy penalty, and would result in a
moderate installed cost penalty.
The heat pipe system and the 100% OA unit (at 20 cfm/person) both resulted
in significant increases in installed cost (relative to the baseline at 5 cfm/person);
21 % for the 100% OA unit, 43% for the heat pipe system. Thus -- although both of
these approaches would be effective at reducing or eliminating the elevated-RH hours
resulting from the OA increase -- they would significantly increase both the energy
penalty and the installed cost penalty.
6.2.2 Cooling Coil Set-Up vs. Shut-Down During Unoccupied Hours
The analysis in Section 6.2.1 assumes that the coils are shut off altogether
overnight and during weekends and holidays, when the office is unoccupied. This
allows the indoor RH to increase during off-hours, resulting from the infiltration of
humid outdoor air. Alternatively, rather than shutting the system down altogether,
the thermostat might instead be set up to, e.g., 81 °F, as discussed in Section 5.11.
If the thermostat were set up, off-hour RH levels would be reduced, at least during
hot weather when the temperature became sufficiently high to cause the cooling coils
to activate during unoccupied hours.
Hours at elevated RH - DOE-2 model predictions. Switching from shut-down
to 81 °F set-up will not have any impact on RH levels — during either occupied or
unoccupied hours - when the weather is cool, and when the set-up temperature is
thus too high to activate the coils when the office is unoccupied.
As discussed in Section 5.11, DOE-2 predicts that all elevated-RH occupied
hours will occur on coo! mornings. Thus, switching to 81 °F set-up has no impact on
the number of occupied hours annually above 60% RH, as shown in Table 15. In
addition, the indoor RH will sometimes exceed 70% during w/ioccupied hours during
cool weather (because the outdoor RH will exceed 90%). Switching to 81 °F set-up
6-7
-------
will not reduce the number of t/rtoccupied elevated-RH hours during cool weather,
either.
But during warm weather, there is a significant impact. As discussed
previously, DOE-2 predicts that no elevated-RH occupied hours occur during warm
weather. The cooling coils are predicted to operate at sufficient power after startup
on warm Monday mornings to remove sufficient moisture from the air as soon as
occupancy begins. But the RH levels commonly reach 60 to 75% during imoccupied
hours overnight and on weekends in the baseline case when the system is shut down.
Switching to 81 °F set-up substantially reduces these warm weather elevated-RH
unoccupied hours. It eliminates them altogether on the warmer days when the
temperature is sufficiently high such that the coils are activated for even a modest
fraction of the unoccupied hours. Although the controller is controlling office
temperature rather than humidity, the activity by the cooling coils to maintain the
81 °F office temperature is predicted by DOE-2 to remove sufficient moisture to
maintain RH below 60% during many warm-weather unoccupied hours.
Thus -- according to the DOE-2 predictions — switching from shut-down to
81 °F set-up would not impact RH during occupied hours, and hence would not
contribute to the goal of maintaining occupant comfort (ASHRAE, 1992a). But, by
greatly reducing that portion of the elevated-RH hours that occur when the office is
unoccupied during warm weather, it would make a major contribution toward reducing
fungal growth (ASHRAE, 1989a).
This statement is true regardless of whether the system is operating at 5 or at
20 cfm OA/person. As modeled here, there is no outside air flow when the cooling
coils cycle on during unoccupied hours in response to the 81 °F set-up temperature;
the system is simply recirculating office air. Hence, the OA rate during occupied
hours has no impact on the ability of the system to eliminate elevated-RH hours during
unoccupied periods.
As discussed in Section 6.1.1, DOE-2 predicts that increased OA will not
increase elevated RH hours during occupied periods. Thus, the conclusion above
could be stated in another manner: If an increase in OA from 5 to 20 cfm/person
were accompanied by a switch from off-hour shut-down to 81 °F set-up, there would
be a significant reduction in the total number of elevated-RH hours (with this reduction
occurring during warm-weather unoccupied hours).
Hours at elevated RH - the FSEC model. As indicated in Section 6.1.1, FSEC
utilizes a mode! which addresses the moisture capacitance of building materials and
furnishings, and moisture re-evaporation from the coils when they cycle off. These
features - not incorporated in the DOE-2.1E model used here - can result in
significantly more hours above 60% RH, when the system is operating at 20 cfm/
6-8
-------
person. And many of the additional elevated RH hours occur during occupied periods
in warm weather, when DOE-2 predicts that elevated RH will not occur.
As with the baseline configuration used in the current study, the FSEC study
assumed that the cooling coils were shut down completely during occupied hours.
With the DOE-2 model available here, it is not possible to reproduce the FSEC results,
or to quantify what the impact would have been had the FSEC study considered off-
hour set-up rather than shut-down. However, based upon the underlying principles,
it is possible to consider qualitatively how thermostat set-up vs. system shut-down
would have impacted the FSEC results.
As with the DOE-2 model, it would not be expected that the FSEC model would
predict any significant effect of set-up vs. shut-down on elevated-RH hours (occupied
or unoccupied) during cool weather. The impact will occur only during warm weather,
when the set-up temperature would trigger coil operation during unoccupied hours.
To assess the effects of thermostat set-up on the FSEC modeling during hot
weather, it is necessary to understand the effects that the moisture capacitance and
re-evaporation phenomena are having on the FSEC model predictions. These effects
are presented in the FSEC publications (Shirey et al., 1995; Shirey and Rengarajan,
1996), and are summarized below.
On a typical hot summer weekday, the DOE-2 model - which excludes the
effects of capacitance and re-evaporation - predicts that office RH steadily increases
overnight when the cooling coils are shut down, as humid outdoor air infiltrates. The
RH will commonly exceed 60% for some portion of the night, and be above 60%
when the coils come on in the morning. When the system starts up in the morning,
the RH drops steeply, and remains well below 60% during the main occupied portion
of the day, when the sensible cooling loads are high and the coils are operating near
capacity (so that latent removal is also high). During the last three hours of coil
operation during the day -- between 6 and 9 p.m. -- RH increases (though commonly
remaining below 60%), because the reduced sensible cooling load during those hours
result in the coiis operating at a lower part-load ratio, with resulting lower latent
removal. When the coils go off, the RH begins its steady rise that will continue
throughout the night due to infiltration.
FSEC predicts that the moisture capacitance phenomenon will impact this
pattern in two ways. During the off-hours, when DOE-2 predicts a steady increase
in RH, moisture capacitance tends to hold RH levels steady, preventing them from
rising above 60%. The reason is that the infiltrating moisture from outdoors is being
adsorbed by certain building materials and furnishings. Then, throughout the main
portion of the day after startup the next morning -- when DOE-2 predicts a significant
decrease in RH - capacitance tends again to keep the RH levels steady, at levels well
above what DOE-2 would predict, because this sorbed moisture is being desorbed.
6-9
-------
Thus, moisture capacitance has the effect of decreasing the number of unoccupied
hours at elevated RH, relative to the DOE-2 predictions, and of increasing the RH
during occupied hours (although not necessarily to levels above 60%).
The coil re-evaporation phenomenon also impacts the DOE-2 pattern in two
ways. First, during the off-hours, there appears to be a continuing source of moisture
from the drain pans associated with the cooling coils. At least for the one summer
day {August 22) analyzed by Shirey et al., this off-hour evaporation added about 5
percentage points to the RH at 20 cfm/person, relative to what the RH would
otherwise have been. Second, when the coils are operating at sufficiently reduced
capacity during the day -- i.e., when the sensible cooling load is sufficiently reduced
such that the coils cycle off for significant periods -- re-evaporation can contribute
significant spikes in the RH. These spikes are much more severe at 20 than at 5
cfm/person. Comparing DOE-2 results for August 22 against the curve presented by
Shirey et al., major RH contributions from re-evaporation seem to occur when the coil
part-load ratio drops below about 0.7. Depending upon the ambient conditions, this
ratio can occur during the final hours of operation (6 to 9 p.m.), and during the initial
hours after startup at 6 a.m., on certain days during warm weather.
In summary, DOE-2 -- without capacitance and re-evaporation included - can
commonly show a certain number of elevated-RH hours overnight on warm weekdays,
but no elevated-RH hours after system startup. This is true at either 5 or 20 cfm/
person. FSEC - with capacitance and re-evaporation — commonly shows RH levels
below 60% almost all the time at 5 cfm/person. But, at 20 cfm/person — depending
on ambient conditions on a given day — the FSEC model can show not only a
significant number of unoccupied hours above 60% RH, but also some portion of the
occupied hours, in the morning and early evening.
Switching to 81 °F thermostat set-up during unoccupied hours, rather than
system shut-down, would likely have the following impacts on the FSEC predictions
(on those days when the ambient temperature is sufficiently high to trigger coil
operation during at least a modest fraction of the unoccupied hours).
Moisture capacitance effects would be reduced. Moisture removal during
occasional operation of the coils overnight and on weekends would reduce moisture
adsorption by the building materials and furnishings during those off hours. As a
result, less sorbed moisture would be available for re-release into the office space
after the coils start up in the morning.
Likewise, coil and drain pan re-evaporation effects would be reduced over
nights and weekends. As indicated, the FSEC model predicts a substantial re-
evaporation effect during off-hours when the coils are shut down, presumably from
water standing in drain pans. This off-hour re-evaporation can be sufficient to raise
the RH above 60%, even while the capacitance phenomenon is adsorbing all of the
6-10
-------
moisture contained in the off-hour infiltrating air. Operation of the coils for a couple
of the overnight hours would be sufficient to remove a significant amount of this
evaporated moisture from the office air, reducing average RH. This would be true
even if it were assumed that all of the re-evaporated moisture thus condensed
remained in the drain pans, where it would potentially have the opportunity to re-
evaporate a second time.
Thus, even with the FSEC model (incorporating capacitance and re-evapora-
tion), off-hour set-up rather than shut-down would significantly reduced the number
of elevated-RH hours during warm weather. Most of the elevated-RH unoccupied
hours during warm weather seem to be due to re-evaporation from the drain pans;
much of this re-evaporated moisture would be removed from the air, reducing
elevated-RH unoccupied hours.
The eievated-RH occupied hours that can occur soon after startup result
because moisture desorption, along with re-evaporation from cycling, reduced-load
coils, create a moisture spike which is superimposed on top of the aiready-elevated
off-hour RH that resulted from off-hour drain pan re-evaporation, if the coils had
come on for a couple hours overnight triggered by the 81 °F set-up temperature, the
off-hour RH level that existed at startup would be reduced, and the portion of the
moisture spike contributed by desorption would also be reduced. Thus, set-up (rather
than shut-down) would also be expected to reduce the number of elevated-RH
occupied hours predicted by the FSEC model at 20 cfm/person.
With either the DOE-2 or the FSEC models, thermostat set-up would not
eliminate all warm-weather elevated-RH hours. It would reduce or eliminate only
those hours occurring on days sufficiently warm such that the set-up temperature
triggered off-hour coil operation. The example day shown in FSEC's publications --
Thursday, August 22 - is one warm day where set-up vs. shut-down would not have
an impact on RH (at least not with an 81 °F set-up temperature). On that particular
day, the highest overnight temperature in the office space is 79 °F (with the TMY
weather file), so the coils are never activated during the unoccupied period.
Energy consumption and cost. As discussed in Section 5.11, and shown in
Table 15, the reduction in elevated hours achievable with thermostat set-up can be
achieved with a minimal energy consumption and cost penalty.
As shown in Table 3, increasing the OA flow from 5 to 20 cfm/person without
changing the temperature control strategy - i.e., with total system shut-down during
unoccupied hours - increases annual HVAC energy costs by $325/yr. As shown in
Table 15, if the increase in OA were accompanied by a switch to 81 °F set-up, this
would increase energy cost only by an additional $ 10/yr (a total cost increase of
$335/yr). Thus, a significant reduction in total elevated-RH hours could be achieved
at a minimal additional cost.
6-11
-------
6.3 REDUCING RH USING ENTHALPY CONTROL
In applications where humidity control is critical, enthalpy controllers are
installed on the system. One common approach for controlling humidity is to overcool
the gas stream to condense moisture, followed by reheat to return the supply air to
the temperature needed for proper thermal control. Another approach is to install
desiccant units in the system. These steps increase energy consumption and cost,
and, especially for the desiccant units, can increase installed cost and maintenance
requirements.
Enthalpy control is most commonly used in selected zones where some special
need — e.g., rare document storage, special manufacturing processes, etc. - is felt to
warrant the costs involved. Enthalpy control does not appear to be commonly utilized
solely for the purposes of occupant comfort and prevention of fungal growth in
offices. Recognizing this reality, the assessment here of the use of enthalpy control
was undertaken to define what the costs and benefits of using this approach would
be, relative to the other approaches presented in Section 6.2, if it were decided that
occupant comfort and indoor microbial control were sufficiently important to warrant
this step.
In this assessment, the humidity control approach considered is the traditional
procedure of overcooling followed by reheat. Reheat has been generally discouraged
for years due to the energy inefficiencies involved, and is generally prohibited by the
Florida code (FDCA, 1993). It is being considered in this assessment despite these
concerns, on the basis that ~ if indoor microbial control is considered to be
important - it is necessary to weigh the effectiveness and the energy penalties of the
reheat approach for comparison against the other approaches.
It was recognized at the outset that -- since the DOE-2.1E model does not
incorporate the moisture capacitance and re-evaporation phenomena -- the estimates
here will not be addressing the full latent load, and hence underestimating the full
energy impacts of enthalpy control. But it is anticipated that the DOE-2 analysis here
will suggest at least the relative impacts of the variables influencing enthalpy control.
Moreover - when steps are taken to maintain RH at reduced levels -- the impacts of
capacitance and re-evaporation will become less, and the difference between the
FSEC and DOE-2 RH predictions will shrink.
Unfortunately, the DOE-2 model is able to model enthalpy control only during
occupied hours. Setting a maximum allowable relative humidity — e.g., MAX-
HUMIDITY = 60 -- will not cause the coils to cycle on during unoccupied hours when
the humidity exceeds that level, Humidity would be controlled during unoccupied
hours only if there were simultaneous off-hour temperature control -- e.g. thermostat
set-up to 81 °F during unoccupied hours - which caused the coils to cycle on during
6-12
-------
certain unoccupied hours. And, even in that case, the humidity control would occur
only during those particular unoccupied hours when the temperature controller cycled
the coils on. Thus, it is not possible with the standard DOE-2 model to simulate
around-the-clock RH control.
Recognizing the limitations of the model, the results of the DOE-2 calculations --
controlling RH at 60% or less during occupied hours - are presented in Table 24,
6.3.1 The Effect of Enthalpy Control at 5 cfm/person
Energy consumption and cost. For three different coil capacity/SHR
combinations, Table 24 shows the predicted effect at 5 cfm/person of switching from
standard temperature-based system control ("No Humidity Control"! to enthalpy
control during occupied hours ("Maximum Humidity = 60%").
In all three capacity/SHR cases, conversion to enthalpy control at a constant
OA rate of 5 cfm/person is estimated to increase HVAC energy cost by about $145
per year, an increase of less than 6%.
As shown in Table 24, adjustments to the total refrigeration capacity and the
sensible heat ratio have only a minor impact on energy cost, with or without enthalpy
control. This is consistent with the results discussed in Section 5.17. The
explanation for this minor effect of capacity and SHR — presented in that earlier
section for the temperature-controlled case — also applies for the enthalpy-controlled
case.
The manner in which enthalpy control impacts energy consumption and costs
at a constant OA rate of 5 cfm/person can be illustrated by considering hourly
performance on selected winter and summer days.
First, consider the winter days of Thursday, January 10, and Friday, January
11.
At the end of the occupied period on Thursday, January 10, the temperatures
of both zones are 74-75 °F in the temperature-controlled case, and a little below 71 °F
in the enthalpy-controlled case. This difference results because the enthalpy-
controlled case has provided substantial additional cooling of the supply air during the
day in order to remove moisture, and the reheat coils have supplied sufficient heat
only to maintain the zone temperatures at the heating set-point (70 °F) throughout the
day. By comparison, in the temperature-controlled case, the system has provided no
heating during the day; the zones were at 75-77 °F just before startup, so that only
cooling was ever called for. As a result, with temperature control, the zones are
maintained at the cooling set-point (75 °F) all day. But in both cases, the moisture
content of the zone air at the end of the day is the same (0.008 lb moisture/lb dry
air), corresponding to an RH of about 43% in the 75 "F temperature-controlled case,
and about 49% in the 70 °F enthalpy-controlled case.
6-13
-------
TABLE 24
Effect of Humidity Control During Occupied Hours, Under Various Conditions
System Conditions
HVAC Energy Cost
($/yr)
No Maximum
Humidity Humidity
Control = 60%
% of Total Hours
Undercooled
No Maximum
Humidity Humidity
Control = 60%
% of Occupied Hours
at RH > 60%
No Maximum
Humidity Humidity
Control = 60%
Program-Calculated Capacities/SHRs
OA = 5 cfm/person
- 8.6 tons refrigeration capacity
- SHR = 0.75
OA = 20 cfm/person
- 9.9 tons refrigeration capacity
- SHR = 0.70
Increased Capacity/SHR = 0.78
OA = 5 cfm/person
- 9.0 tons capacity, SHR = 0.78
OA = 20 cfm/person
- 10.0 tons capacity, SHR = 0.78
Increased Capacity/SHR = 0.73
OA = 5 cfm/person
- 9.0 tons capacity, SHR = 0.73
OA = 20 cfm/person
- 10.0 tons capacity, SHR = 0.73
2,510 2,656
2,835 2,925
2,541
2,856
2,527
2,853
2,683
2,945
2,669
2,940
0.4
0.4
0.1
0.1
0.3
0.2
0
0
0
0
0
0
1.2
0.9
1.2
0.9
1.2
0.9
0
0
0
0
0
0
-------
Overnight between Thursday and Friday, the outdoor temperature drops from
72 °F at shut-down (according to the TMY weather file) to a low of 62 °F at startup
the next morning. The ambient humidity ranges between 0.011 and 0.013 lb/lb (with
ambient RH values varying between 77% and 94%). Between shut-down on
Thursday and startup on Friday -- due to the combined effects of infiltration, shell heat
transfer, and release of heat stored in the building materials and furnishings - the
zone temperatures rise by 1 to 1.5 F° in the temperature-controlled case (so that the
temperature just before Friday startup is 75-76.5 °F). In the enthalpy-controlled case,
the temperature rise is slightly greater -- about 2 F° -- so that the temperature before
startup is 72.5-73 °F. The moisture in the infiltrating air causes overnight office
humidity levels to reach a peak of 0.013 ib/lb, corresponding to an RH of roughly
70%, varying slightly depending upon office temperature.
According to the TMY weather file, Friday, January 11, is a fairly mild day, with
ambient temperatures holding in the low to mid 60's throughout the morning, and
reaching a high of 78 °F in mid-afternoon. The ambient RHs are in the range of 85%
to 95% in the morning, and 65% to 80% in the afternoon.
When the air handler starts up on Friday morning in the temperature-controlled
case, the cooling coils do not cycle on because the office temperature is below the
cooling set-point of 75 °F. In fact, in the temperature-controlled case at 5 cfm/person,
the coils do not cycle on until noon. As a result, no moisture is removed, and the
relative humidity remains above 60% all morning (a total of 5 occupied hours), but
drops below 60% immediately when the coils finally cycle on. The total cooling coil
energy consumption during that day in the 5 cfm/person temperature-controlled case
is just over 428,000 Btu (of which 57,000 Btu is latent cooling). Of course, there is
no heating energy consumption. As on Thursday, the office space is just below 75 °F
at the end of the occupied period on Friday.
When the system starts up on Friday morning in the enthalpy-controlled case,
the coils cycle on immediately in response to the elevated RH that exists. In the first
hour after startup, the average RH in the office space is reduced to about 52% (and
the office temperature is reduced below 71 °F). None of the occupied hours on Friday
are above 60% RH in the enthalpy-controlled case, as would be expected. Of course,
this added moisture removal is reflected in the energy consumption, On Friday, the
enthalpy-controlled system at 5 cfm/person has a total cooling coil energy consump-
tion of 864,000 Btu (of which 89,000 is latent), and a total reheat energy consump-
tion of 132,500 Btu. The total cooling and heating energy consumption that day -
996,500 Btu — is 2.3 times greater than that for the temperature-controlled case.
This relationship between energy consumptions in the temperature- and
enthalpy-controlled cases is representative of what is generally experienced on such
cool winter days when the elevated-RH occupied hours tend to occur (in the absence
of enthalpy control). The increase in consumption for the enthalpy-controlled case is
6-15
-------
not so great on certain winter days when the TMY weather file predicts less humid
conditions.
This energy consumption relationship is very different during hot weather.
Since the temperature-controlled case controls occupied-hour RH levels effectively
during hot weather, the difference in consumption between the two cases is small.
To illustrate this, consider the week beginning on Monday, July 8. Overnight,
prior to startup on Monday, the outdoor temperature has dropped from 82 to 76 °F.
The overnight ambient humidity has ranged from 0.018 to 0.019 lb/lb, corresponding
to an RH of about 90%.
Under these conditions, the indoor temperature is about 85 °F just before
Monday morning startup in the temperature-controlled case, and 83 to 83.5 °F in the
enthalpy-controlled case. Interestingly, even after the coils have been off over the
summer weekend, a couple-degree office temperature differential still exists between
the two cases, resulting from the office space having been 4 F° cooler in the enthalpy-
controlled case - 71 vs. 75 °F - when the system shut down on Friday evening. The
indoor humidity has also ranged between 0.018 and 0.019 lb/lb overnight between
Sunday and Monday, corresponding to an RH greater than 70%.
When the system starts up on Monday morning, July 8, in the temperature-
controlled case, the cooling coils cycle on at almost full capacity, since the space is
well above the 75 °F cooling set-point at startup. As a result, the office RH averages
below 60% starting with the first occupied hour, and there are no elevated-RH
occupied hours during the day. During this day - when the outdoor temperature
holds in the mid to upper 80's throughout most of the morning and afternoon - the
5 cfm/person temperature-controlled system consumes 1,234,000 Btu of cooling
energy, 134,000 Btu of which is latent cooling.
When the enthalpy-controlled system starts up on Monday morning, it actually
consumes slightly less cooling energy than does the temperature-controlled case. This
presumably occurs because the enthalpy-controlled office space is at a slightly lower
temperature at startup, as indicated above. Of course, the RH is below 60%
throughout the day with this system, as it is with the temperature-controlled system.
During the course of this day, the cooling energy consumption by the enthalpy-
controlled system is 1,225,000 Btu (of which 135,000 Btu is latent cooling) at 5
cfm/person. The reheat coils never have to activate on this day; the system is able
to maintain the RH below 60% without having to over-cool the supply air. Thus, on
this day, the enthalpy-controlled system actually consumes slightly less cooling/
heating energy than does the temperature-controlled system.
Because the enthalpy-controlled system is able to provide the needed moisture
removal without reheat on July 8, the system is controlled by the cooling set-point.
6-16
-------
Hence, the office space is at about 75 °F during most of the day, just as is the
temperature-controlled system. Later in the day, as the sensible heat load decreases,
the enthalpy-controlled system does perform a small amount of supply air over-cooling
in order to remove moisture, and this reduces office temperature to 73 to 74 DF. But,
since this over-cooling is not sufficient to drop the office temperature below 70 °F,
the reheat coils do not come on.
The high sensible load created by the weekend shutdown is largely addressed
on Monday. Peak loads are less severe during the remainder of the week, and, as a
result, energy consumption by the enthalpy-controlled system again becomes greater
than that for the temperature-controlled system. For example, on Friday, July 12, the
enthalpy-controlled system is having to over-cool the supply air in order to remove
sufficient moisture, and the office temperature ranges between 71 and 72 °F, close
to the heating set-point (compared to a steady 75 °F with the temperature-controlled
system). The enthalpy-controller reheat coils actually have to come on during the last
operating hour on Friday, to prevent office over-cooling. As a result, on Friday, total
cooling and heating energy consumption by the enthalpy-controlled system is
1,019,000 Btu, 13% greater than the 900,000 Btu in cooling energy required by the
temperature-controlled system.
The preceding examples for selected days in January and July illustrate why --
over the course of the year - the enthalpy-controlled case increases HVAC energy
consumption (i.e., cooling, heating, and air handler consumption! by about 11 % at a
constant OA rate of 5 cfm/person, and HVAC energy cost by about 6%.
Over 85% of the increase in HVAC energy consumption with the enthalpy-
controlled system over the course of the year results from increased cooling coil
(compressor and condenser fan) energy. Of the 2,820 kWh increase in annual HVAC
consumption caused by enthalpy control at 5 cfm/person, only 380 kWh (13%) is the
result of increased heating energy caused by the reheat coils. Increased cooling coil
energy is responsible for the remainder. Of course, air handler energy requirements
are unchanged by the switch to enthalpy control. Also, the small heating energy
requirements to maintain the heating set-point during normal heating hours (excluding
reheat! remain essentially unchanged.
Hours at elevated temperature and RH. As shown in Table 24, switching to
enthalpy control (with MAX-HUMIDITY = 60) eliminates all undercooled hours at 5
cfm/person. This occurs because on warm Monday mornings - when the undercooled
hours occur with the temperature-controlled system - the enthalpy-controlled system
encounters a cooler zone temperature upon startup, since the enthalpy system had
left the zones several degrees cooler on Friday evening. Thus, less cooling is
demanded from the enthalpy-controlled system upon startup, and undercooled hours
are avoided.
6-17
-------
For illustration, again consider the morning of Monday, July 8. Just before
startup on this morning, the office space temperature is a little below 85 °F in the
temperature-controlled case, and about 83 to 83.5 °F in the enthalpy-controlled case,
because the enthalpy-controlled zones had been about 4 F° cooler at shut-down on
Friday evening, as indicated earlier. Starting at 83 °F, the 5 cfm/person enthalpy-
controlled case is able to reduce the temperature of both zones sufficiently close to
the 75 °F cooling set-point -- i.e., to 77 °F or less - beginning with the first hour after
startup on Monday. Thus, the enthalpy system is not charged with any undercooled
hours on that day. But starting at 85 °F, the temperature-controlled case is slightly
above 77 °F for the first 4 hours in the front zone (with temperatures of 77.1 to
78.3 °F), and the first 2 hours in the rear zone (77.2 to 77.9 °F). The 1.5 to 2 F°
warmer zone encountered at startup by the temperature-controlled system would
appear to explain why this system resulting in undercooled hours while the enthalpy-
controlled system did not.
Due to the lower startup temperature, the enthalpy-controlled system is able
to avoid undercooled hours while consuming slightly less cooling energy than the
temperature-controlled system. During the first 4 hours on July 8 — when the
temperature-controlled system is encountering undercooled hours in one or both
zones - the enthalpy-controlled cooling coils are consuming a total of 387,000 Btu
{front and rear units combined), while the temperature-controlled cooling coils are
consuming 391,000 Btu, about 1% more energy.
Table 24 also confirms that - as commanded - the enthalpy-controlled case
eliminates occupied hours above 60% RH at the 5 cfm/person OA rate. This occurs,
of course, because the enthalpy-controlled system provides more cooling (and, as
necessary, reheat) on cool mornings when elevated-RH hours occur.
Consider the morning of Monday, January 14. Just before startup on this day,
the zones in the enthalpy-controlled case are at 68.5 to 69.5 °F, 1 to 2.5 F° cooler
than the 69.5 to 72 °F in the temperature-controlled case. Again, this differential at
startup exists because the enthalpy-controlled zones were 3.5 to 4 F° cooler at shut-
down on Friday evening. In both cases, the humidity before startup is 0.013 lb/lb,
corresponding to an RH of about 87% in the cooler, enthalpy-controlled zones, and
about 82% in the warmer, temperature-controlled zones.
When the temperature-controlled system starts up on January 14, the air
handler comes on, but neither cooling nor heating are provided. The zone temperature
is at or above the heating set-point, and below the cooling set-point, so that neither
of the coils are activated. As latent heat is generated inside the offices, and as the
outdoor humidity increases (in terms of lb/lb), the indoor humidity increases to a peak
of 0.02 Ib/ib during the morning. At 5 cfm/person, the cooling coils do not come on
until the third hour after startup in the rear zone, and the fourth hour in the front
6-18
-------
zone -- and then at greatly reduced capacity. As a result, the RH is above 60% during
the first 6 hours in the front zone, and the first 4 hours in the rear.
By comparison, when the enthalpy-controlled system starts up on January 14,
the cooling coils cycle on immediately — and the reheat activates as well. As a result,
there are no elevated-RH occupied hours during the course of the day. The reheat
coils maintain the office space just below 71 °F throughout the day, in response to the
heating set-point.
Of course, as emphasized previously, there is an energy penalty associated with
this dehumidification. During the first 6 hours, when one or both of the zones are at
elevated RH in the temperature-controlled case, the enthalpy-controlled system
consumes a total of 570,000 Btu for cooling and reheat in the two zones combined
{with about one-third of this total being energy for reheat), at 5 cfm/person. By
comparison, the temperature-controlled system consumes only 79,000 Btu during this
same 6-hour period — only 14% of the enthalpy-controlled total. For the entire day,
the temperature-controlled system consumes 395,000 Btu, 40% of the 997,000 Btu
consumed in the enthalpy-controlled case.
6.3.2 The Effect of Enthalpy Control on Increased Ventilation Rates
Energy consumption and cost. As shown in Table 24, at a constant OA rate
of 20 cfm/person, the increase in energy cost resulting from a switch from tempera-
ture to enthalpy control is slightly less than it is when this switch is made at a
constant OA rate of 5 cfm/person. At 5 cfm/person, the energy cost increase of
switching to enthalpy control is predicted to be about $145 per year for each of the
capacity/SHR combinations shown in the table (e.g., $2,656/yr with enthalpy control
vs. $2,510/yr with temperature control). At 20 cfm/person, this increase is reduced
to about $90 ($2,925/yr vs. $2,835/yr).
Another way of looking at this difference is that - with temperature control --
increasing the OA rate from 5 to 20 cfm/person results in a predicted energy cost
increase of about $325/yr (from $2,510 to $2,835 for the first case in Table 24). By
comparison, if this switch to 20 cfm/person were accompanied by a switch from
temperature control to enthalpy control, the energy cost increase would be $415/yr
(from $2,510 to $2,925), $90 greater than if the accompanying switch to enthalpy
control were not made.
As with the increase in energy cost, the increase in HVAC energy consumption
caused by the switch to enthalpy control is less when the OA is 20 cfm/person. At
a constant OA rate of 5 cfm/person, the switch from temperature control to enthalpy
control results in added energy consumption of about 2,800 kWh/year, for all three
capacity/SHR combinations considered in Table 24. More than 85% of this increase
at 5 cfm/person is due to increased compressor and condenser fan consumption, and
6-19
-------
less than 15% by reheat coil operation. By comparison, at a constant OA rate of 20
cfm/person, this penalty is only about 1,800 kWh/year. And only about 3% of this
1,800 kWh increase at 20 cfm/person is due to reheat coil operation.
This modest 1,000 kWh/yr reduction in the energy penalty (1,800 instead of
2,800 kWh/yr), achieved by switching to enthalpy control at 20 instead of 5
cfm/person, results from two phenomena.
a) About one-third of the 1,000 kWh/yr reduction results because the enthalpy-
controlled system sometimes consumes /ess reheat energy at 20 cfm/person
than it does at 5 cfm/person, in cool and in mild weather. The enthalpy-
controlled system tends to provide less sensible cooling (and more latent
cooling) at the higher OA rate during cool and mild weather, with the result that
less reheat is required. Reheat energy is also reduced at 20 cfm/person for the
reason described in b) below.
b) The remainder of the 1,000 kWh reduction results because the enthalpy-
controlled system sometimes consumes less total cooling energy at 20 than at
5 cfm/person, in mild weather. Typically — when over-cooling and then
reheating -- the enthalpy-controlled system maintains the offices at the heating
set-point of 70 °F. But at mild conditions when the RH in the office is just
slightly above 60%, the enthalpy-controlled system (as modelled by DOE-2)
sometimes reduces RH instead by simply letting the office temperature rise to
the cooling set-point of 75 °F, at a substantial savings in cooling (and reheat)
energy. This occurs during a greater number of hours when operating at the
increased OA rate.
To illustrate these effects, it is useful to consider three separate weeks
throughout the year - a cool week (ending January 18), a mild week (ending March
22), and a hot week (ending July 12).
During each of these weeks, the temperature-controlled system at 20 cfm/
person consumes more power for total cooling -- and more power for latent cooling —
than does the temperature-controlled system at 5 cfm/person, as would be expected.
Also, during each of these weeks, the enthalpy-controlled system at either OA
rate consumes more power for total cooling, more power for latent cooling, and (of
course) more power for reheat, than does the temperature-controlled system at the
same OA rate. Again, this would be expected.
But during the week ending January 18, the enthalpy-controlled system at 20
cfm/person consumes fess energy for reheat than does the enthalpy-controlled system
at 5 cfm/person. And during the week ending March 22, the enthalpy-controlled
system at 20 cfm/person consumes Iess energy for total cooling and /ess energy for
6-20
-------
reheat than does the enthalpy-controlled system at 5 cfm/pcrson. But during the
week ending July 12-- when the system tends to be operating near full capacity, and
essentially no reheat is needed -- the enthalpy-controlled system at 20 efm/person
consumes more power for total (and latent) cooling than does the enthalpy-controlled
system at 5 cfm/person.
Thus, the largest energy penalty for converting from temperature to enthalpy
control occurs when the switch is made at the lower OA rate. When the switch to
enthalpy control is instead implemented at a higher OA rate, the energy penalty
associated with the conversion is less during cool and mild weeks, and is about the
same during hot weeks. As a result, the net effect over the course of the entire year
is a modest reduction in penalty for conversion to enthalpy control as the OA rate
increases.
Consider the week ending January 18. During this week, the outdoor
temperature is generally mild, and outdoor humidity is high. As a result, when the OA
rate is increased from 5 to 20 cfm/person with the enthalpy-controlled system, the
latent cooling for the week more than doubles, from 484 kBtu to 1,010 kBtu for the
two zones combined. But at these outdoor temperature and RH conditions with these
PSZ units, the sensible cooling with the enthalpy-controlled system actually decreases
with the increase in OA (from 3,982 to 3,757 kBtu), partially off-setting the increase
in latent cooling. And because there is less sensible cooling at 20 cfm/person, less
reheat is required in order to maintain office temperature at the heating set-point (or
above); the increase in OA decreases reheat requirements for the week from 648 to
367 kBtu.
In summary, for the week ending January 18, an increase from 5 to 20
cfm/person with the enthalpy-controlled system increases total cooling requirements
by just over 300 kBtu, and decreases reheat requirements by just over 280 kBtu.
Thus, the OA increase results in only a minor 20 kBtu increase in total cooling plus
reheat energy for the week.
Consider now the week ending March 22. During this week, the outdoor
temperature is mild; it averages less than 1 F° warmer than during the January week
considered above, according to the TMY file for Miami. But the outdoor humidity is
much lower during this March week, averaging only 0.010 lb moisture per lb dry air
(compared to 0.012 lb/lb during the January week). Under these conditions, the
moisture content of the office space (in terms of lb/lb) often corresponds to an RH just
above, or just below, 60%. As a result, a small change in office temperature can
sometimes determine whether the office is above or below the RH set-point for the
enthalpy controller.
The effect of this "borderline" situation is that the enthalpy control system will
over-cool and reheat the supply air during some hours of the day, in order to maintain
6-21
-------
the RH below 60% at the heating set-point of 70 °F. But then, during other hours,
the system -- as modelled by DOE-2 -- will allow office temperature to rise to the
cooling set-point of 75 °F, where the warmer temperature will cause the RH to drop
below 60% without the substantial energy penalty associated with the over-cooling
and reheating. Increasing the OA rate from 5 to 20 cfm/person with the enthalpy-
controlled system increases the number of hours during this March week when the
RH control is accomplished by letting the office temperature rise. Hence, the increase
in OA rate reduces both the total cooling energy and the reheat energy during this
week.
First, considerthe 5-hour period between 7 am and noon on Wednesday, March
20. During this period, the system is maintaining the office space at about 71 °F,
near the heating set-point, with either OA rate. The outdoor temperature during this
period ranges between 70 and 74 °F, and the outdoor humidity ranges between
0.0089 and 0.0096 lb/lb {corresponding to RHs of about 55% to 60% at office
temperatures).
Under these conditions, for this 5-hour period in the morning, an increase in the
enthalpy-controlled system's OA rate from 5 to 20 cfm/person increases sensible
cooling by 2.5% (from 298 to 306 kBtu), and latent cooling by one third (from 28 to
39 kBtu). But, for reasons that are not clear, the model calculates that this increase
in OA decreases reheat requirements substantially (from 40 kBtu to zero). As a result,
the total cooling plus reheat energy requirements for the enthalpy-controlled system
during these 5 hours are somewhat reduced (from 366 kBtu to 345 kBtu) by virtue
of increasing the OA from 5 to 20 cfm/person.
This reduction during this 5-hour period on the morning of March 20 might thus
be attributed to a phenomenon similar to that responsible for the effects during the
January week, discussed above. That is, at 20 cfm/person, more of the total cooling
is latent cooling, with the result that the temperature of the supply air is reduced to
a lesser extent and the need for reheat (to maintain the 71 °F office temperature) is
thus reduced. However, it is not apparent why the model is predicting the elimination
of reheat requirements at 20 cfm/person during this period. The mixed-air tempera-
tures entering the coils are about the same at both OA rates, essentially the same
amount of sensible cooling is provided by the coils at both rates, and, corres-
pondingly, exactly the same reduction in air temperature occurs across the coils at
both rates. It is curious that the 5 cfm/person case should be computed as providing
40 kBtu of reheat under these conditions, while the 20 cfm/person case is providing
none.
Now considerthe afternoon hours on March 20, when a different phenomenon
occurs. The outdoor air is somewhat drier in the afternoon compared to the
morning -- 0.0081 to 0.0089 lb/lb, corresponding to RHs of 45% to 55% at office
6-22
-------
temperatures. Outdoor temperatures are relatively mild during the afternoon, ranging
from 72 to 75 °F, according to the TMY weather file.
Presumably because of the increased influx of this relatively dry outdoor air at
20 cfm/person, the enthalpy-controlled system at this higher OA flow rate shifts from
the heating set-point of 70 °F to the cooling set-point of 75 °F during the entire
afternoon on March 20. That is, RH is reduced by raising the office temperature. By
comparison, at 5 cfm/person, the enthalpy controller keeps the system operating at
the heating set-point (i.e., reducing RH by over-cooling and reheating) during the entire
afternoon, until the last hour before shut-down in the evening.
The energy impact of this action by the controller is substantial. For the 7-hour
period from noon through 7 pm, the total cooling energy consumed by the enthalpy-
controlled system drops from 422 kBtu to 284 kBtu as OA is increased from 5 to 20
cfm/person (although the latent cooling increases modestly, from 25 to 36 kBtu).
Correspondingly, the OA increase decreases reheat energy consumption during this
7-hour period, from 29 kBtu to zero.
These results for Wednesday, March 20, explain the results observed for the
entire week ending March 22. For the entire week, the increase in OA from 5 to 20
cfm/person in the enthalpy-controlled system results in a 12% decrease in sensible
cooling requirements (from 3,849 to 3,371 kBtu), a 57% increase in latent cooling
(from 312 to 490 kBtu), and almost complete elimination of reheat requirements
(decreasing reheat consumption from 238 to 10 kBtu). The total cooling plus reheat
energy consumption is reduced by a little more than 500 kBtu (from 4,400 to 3,872
kBtu) by the increase in OA.
Finally, consider the week ending July 12. During this hot week, essentially no
reheat is needed in the enthalpy-controlled system; even the temperature-controlled
system, which operates near full capacity without reheat capability, maintains the RH
below 60%. Also, the outdoor conditions are so hot and humid that the enthalpy-
controlled system will always be controlled by the cooling set-point, Thus, neither of
the phenomena discussed above, that caused increased OA to reduce energy
consumption during the January and March weeks, would be expected to come into
play during July.
And indeed, this is the case. Increasing OA from 5 to 20 cfm/person with the
enthalpy-controlled system increases total cooling during this July week by 1,000
kBtu (975 kBtu of which is increased latent cooling). This is almost identical to the
1,100 kBtu increase in total cooling (940 kBtu increase in latent cooling) experienced
when OA is increased in the temperature controlled system. And there is essentially
no change in reheat energy consumption caused by the OA increase in the enthalpy-
controlled system; there is essentially no reheat energy consumed during July at either
OA rate.
6-23
-------
In summary, with the enthalpy-controlled system, an increase in OA from 5 to
20 cfm/person causes almost no change in energy consumption during the January
week, a 500 kBtu decrease in consumption during the March week, and a 1,000 kBtu
increase in consumption during the July week. By comparison, with the temperature-
controlled system, that same increase in OA causes an increase in energy consump-
tion during each of those three weeks. The net results are that:
a) with the enthalpy-controlled system, increasing OA from 5 to 20 cfm/person
increases annual energy consumption; but
b) the resulting increase in annual energy consumption with the enthalpy-
controlled system is less than the increase that would be caused by that same
OA increase with the temperature-controlled system.
Hours at elevated temperature and RH. As shown in Table 24, the enthalpy-
controlled system operating at 20 cfm OA/person eliminates all undercooled hours
and, as commanded, all occupied hours at elevated RH. This occurs for the same
reasons discussed in connection with the enthalpy-controlled system at 5 cfm/person
(see Section 6.3.1).
6-24
-------
SECTION 7.
REFERENCES
ASHRAE (1981) Ventilation for Acceptable Indoor Air Quality, Atlanta, GA, American
Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (ASHRAE
Standard 62-19815-
ASHRAE (1989a) Ventilation for Acceptable Indoor Air Quality, Atlanta, GA,
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.
(ANSI/ASHRAE Standard 62-1989).
ASHRAE (1989b) Energy-Efficient Design of New Buildings Except Low-Rise Residen-
tial Buildings, Atlanta, GA, American Society of Heating, Refrigerating and Air-
Conditioning Engineers, Inc. (ASHRAE/IES Standard 90.1 -1989).
ASHRAE (1992a) Thermal Environmental Conditions for Human Occupancy, Atlanta,
GA, American Society of Heating, Refrigerating and Air-Conditioning Engineers,
Inc. (ANSI/ASHRAE Standard 55-1992).
ASHRAE (1992b) User's Manual: ASHRAE/IES Standard 90.1-1989, Atlanta, GA,
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.
ASHRAE (1992c) 1992 ASHRAE Handbook: HVAC Systems and Equipment,
Atlanta, GA, American Society of Heating, Refrigerating and Air-Conditioning
Engineers, Inc.
ASHRAE (1993) 1993 ASHRAE Handbook: Fundamentals, Atlanta, GA, American
Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.
ASHRAE (1995) 1995 ASHRAE Handbook: HVAC Applications, Atlanta, GA,
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.
Ayres, J. M., and E. Stamper (1995) "Historical Development of Building Energy
Calculations," ASHRAE Journal, 37 (2), 47-55.
Birdsall, B. E. (1995) Personal communication, Berkeley, CA.
Eto, J. H., and C. Meyer (1988) "The HVAC costs of increased fresh air ventilation
rates in office buildings," ASHRAE Transactions, 94 (Pt. 2), 331-345.
Eto, J. H. (1990) "The HVAC costs of increased fresh air ventilation rates in office
buildings, Part 2," in Proceedings of the Fifth International Conference on
Indoor Air Quality and Climate, Toronto, Ontario, Vol. 4, 53-58.
7-1
-------
FDCA (1993) Energy Efficiency Code for Building Construction, Tallahassee, FL,
Florida Department of Community Affairs.
Mudarri, D. H., and J. D. Hall (1993) "Increasing outdoor air flow rates in existing
buildings," in Proceedings of the Sixth International Conference on Indoor Air
Quality and Climate, Helsinki, Finland, Vol. 5, 21-26.
Odom, J. D., and G. DuBose (1994) Preventing Indoor Air Quality Problems in Hot,
Humid Climates: Problem Avoidance Guidelines, Orlando, FL, CH2M HILL.
Parker, D. S. (1996) Modeled and Metered Energy Savings from Exterior Wall
Insulation, Cocoa, FL, Florida Solar Energy Center (Report No. FSEC-PF-308-
96).
Rojeski, P., H, Singh, and J. Jones (1995) "The costs associated with improving
indoor air quality using ventilation and/or filtration strategies," submitted to
ASHRAE Journal.
Shirey, III, D. B., K. Rengarajan, and R. A. Raustad (1995) Impacts of ASHRAE
Standard 62-1989 on Small Florida Offices, Cocoa, FL, Florida Solar Energy
Center (Report No. FSEC-CR-787-95).
Shirey, 111, D. B., and K. Rengarajan (1996) "Impacts of ASHRAE Standard 62-1989
on Small Florida Offices," presented at the 1996 ASHRAE Winter Meeting,
Atlanta, GA.
Steele, T., and M. Brown (1990) Energy and Cost Implications of ASHRAE Standard
62-1989, Portland, OR, Bonneville Power Administration.
Todesco, G. (1996) "Super-Efficient Buildings: How Low Can You Go?" ASHRAE
Journal, 38 (12), 35-40.
U. S. Department of Commerce (1994) Statistical Abstract of the United States:
1994 (114th edition), Washington, D. C.
U. S. Department of Energy (1994a) DOE-2 Basics Manual (Version 2. IE), Berkeley,
CA, Lawrence Berkeley Laboratory (NTIS Accession No. DE-940 13165).
U. S. Department of Energy (1994b) Commercial Building Characteristics, 1992:
Commercial Buildings Energy Consumption Survey, Washington, D. C., Energy
information Administration [Report No. DOE/EIA-0246(92)].
Ventresca, J. A. (1990) "Measuring the impacts of increased ventilation upon indoor
air quality and electric demand: a progress report," in Energy and Environ-
7-2
-------
mental Strategies for the 1990's, Atlanta, GA, Association of Energy Engineers,
331-336.
Winkelmann, F. C., B. E. Birdsall, W. F. Buhl, K. L, Ellington, A. E. Erdem, J. J.
Hirsch, and S. D. Gates (1993) DOE-2 Supplement, Version 2.1E, Berkeley,
CA, U. S. Department of Energy, pp. 3.129 through 3.132.
York, D. A., E. F. Tucker, and C. C. Cappiello (eds.) (1981) DOE-2 Reference Manual
(Version 2.1A) - Part I, Los Alamos, NM, U. S. Department of Energy (Report
No. LBL-8706 Rev. 2, NTIS Accession No. LBL-8706 Rev. 2).
7-3
-------
[This page intentionally blank.]
-------
APPENDIX A
Baseline Input D0E-2.1E File:
4,000 ft2 Office in Miami Strip Mall
INPUT LOADS
TITLE
LINE-1 * SMALL OFFICE STUDY #2 *
LINE-2 *4000 FT2 OFFICE IN STRIP MALL {MIAMI)*
LINE-3 *CASE 01A: — BASE CASE •—*
LINE-4 *HVAC: 2 PSZ UNITS WITH ECONOMIZERS*
LINE-5 *OA = 5 CFM/PERSON* ..
$ DOE—2.IE FILE BY BRUCE HENSCHEL
REVISED SEPTEMBER 6, 1995
$ THIS SMALL OFFICE IS MODIFIED FROM A SIMILAR ONE MODELED BY THE
$ FLORIDA SOLAR ENERGY CENTER (FSEC). IT IS A 4000 FT2 "SLICE" OUT
$ OF A LARGER, SINGLE-STORY STRIP MALL. THIS OFFICE IS IN TURN SUB-
$ DIVIDED INTO TWO 2000 FT2 ZONES, EACH COOLED WITH A DEDICATED
$ PACKAGED SINGLE-ZONE DIRECT EXPANSION AIR-CONDITIONER (PSZ), HEATED
$ WITH ELECTRIC RESISTANCE HEATING.
BECAUSE THIS OFFICE IS ADJOINED ON EACH SIDE BY OTHER OFFICES,
LEASED BY OTHER TENANTS WHO ARE CONDITIONING THEIR ADJOINING
SPACE USING THEIR OWN DEDICATED UNITS, THIS MODELING EFFORT
ADDRESSES ONLY THE FRONT AND REAR (EXTERIOR) WALLS. IT IS
ASSUMED THAT THE TWO SIDE WALLS ARE THERMALLY NEUTRAL, AND THE
MODEL THUS PRETENDS THAT THEY DO NOT EXIST.
DIAGNOSTIC
ABORT
RUN-PERIOD
CAUTIONS ..
ERRORS ..
JAN 1 1974 THRU DEC 31 1974
BUILDING-LOCATION
LATITUDE = 25.8 LONGITUDE =80.3
ALTITUDE = 7 TIME-ZONE = 5
AZIMUTH = 0 HOLIDAY = YES ..
LATITUDE, LONGITUDE, ETC., CORRESPOND TO MIAMI, FLORIDA.
-BUILDING SHADING-
'S
FRONT-OVERHANG
BUILDING-SHADE
X=40 Y=10Q
H=10 W= 40
Z=9 AZ=0
TILT=180 ..
FRONT-PARAPET
BUILDING-SHADE
X=40 Y=100
H=2 W= 40
Z=13 AZ=0
TILT=90 ..
-LOADS SCHEDULES-
OCCUPY- 1
SCHEDULE
THRU DEC 31
(MON,FRI)
(SAT,HOL)
(1,6)(0)
(8,11)(1)
(15,17)(1)
(21,24)(0)
(1,24)(0) .
(7)(.3)
(12,14)(,
(18,20)(,
8,
3,
4,
1,
8)
1)
A-l
-------
LIGHTS-1 = SCHEDULE THRU DEC 31
(MON,FRI) (1,6)(.05) (7)(.3)
(8,17)(1) (18,20)(.3,.1,.1)
(21,24)(.05)
(SAT,HOL) (1,24)(.05) ..
INFILT-1 = SCHEDULE THRU DEC 31
(ALL) (1,24)(1) ..
$ SPACE CONDITIONS $
OFFICE-ENVIRON= SPACE-CONDITIONS
TEMPERATURE
PEOPLE-SCHEDULE =
AREA/PERSON
PEOPLE-HEAT-GAIN =
LIGHTING-SCHEDULE=
LIGHTING-TYPE
LIGHTING-W/SQFT =
LIGHT-TO-SPACE
EQUIP-SCHEDULE =
EQUIPMENT-W/SQFT =
INF-METHOD
INF-SCHEDULE =
AIR-CHANGES/HR =
FLOOR-WEIGHT
ZONE-TYPE
(72)
OCCUPY-1
150
450
LIGHTS-1
1EC-FLUOR-NV
1.8
1
LIGHTS-1
0.75
AIR-CHANGE
INFILT-1
0.1
70
CONDITIONED .
$ CONSTRUCTIONS AND GLASS TYPE —$
EX-WALL-LAYER = LAYERS MATERIAL = (SC01,CB11,IN32,GP02)
INSIDE-FILM-RES = 0.68 ..
$ THIS LAYER OF MATERIALS IN THE EXTERIOR WALL INCLUDES 1-IN.
$ STUCCO, 8-IN, HEAVY-WEIGHT HOLLOW BLOCK, 3/4-IN. POLYSTYRENE
$ INSULATION, AND 5/8-IN. GYPSUM BOARD, PLUS THE INSIDE AND
$ OUTSIDE FILM RESISTANCES.
IN-WALL-LAYER = LAYERS MATERIAL = (GP02,AL21,GP02)
INSIDE-FILM-RES = 0.68 ..
CEILING-LAYER = LAYERS MATERIAL = (AG02)
INSIDE-FILM-RES = 0.61 ..
R-12—INSUL — MATERIAL TH=0.333 COND=0.024 DENS=15 S—H=0.17 .•
METAL-DECK = MATERIAL TH=0.021 COND=26 DENS=480 S-H=0.10 ..
ROOF-LAYER = LAYERS MATERIAL « (BR01,R-12-INSUL,METAL-DECK)
INSIDE-FILM-RES =0.61 ..
SLAB-LAYER = LAYERS MATERIAL = (CC03,CP02)
INSIDE-FILM-RES =0.92 ..
EXT-WALL
INT-WALL
CEILING
CONSTRUCTION
CONSTRUCTION
CONSTRUCTION
LAYERS = EX—WALL—LAYER
ABSORPTANCE =0.7 ROUGHNESS
LAYERS = IN-WALL-LAYER ..
LAYERS = CEILING-LAYER ..
A-2
-------
MALL-ROOF
SLAB
BACK-DOOR
FRNT-DOOR
CONSTRUCTION
CONSTRUCTION
CONSTRUCTION
GLASS-TYPE
LAYERS = ROOF-LAYER
ABSORPTANCE = 0.8
LAYERS = SLAB-LAYER ..
U-VALUE =0.59
ABSORPTANCE - 0.84 ..
PANES « 1
SHAD ING-COEF >=0.55
GLASS-CONDUCTANCE =1.47
$ DEFAULT GLASS-CONDUCTANCE WHEN PANES=1
FRAME-CONDUCTANCE = 1.245 ..
$ FRAME-COND. FOR THERMALLY BROKEN ALUMINUM
ALL-WINDOWS = GLASS-TYPE
PANES = 1
SHADING-COEF =0.55
GLASS-CONDUCTANCE =1.47
FRAME-CONDUCTANCE = 1.245 ..
$ SPACE DEFINITIONS $
$ THE 13-FT—TALL BUILDING INCLUDES 9-FT CEILINGS IN THE OFFICE
$ AREA, WITH A 4-FT PLENUM OVERHEAD.
FRONT-PLENUM = SPACE
X=0 Y=Q Z=0 AZ=0
AREA = 2000
VOLUME ¦ 8000
TEMPERATURE = (72)
$ PLENUM TEMPERATURE FOR LOADS CALCULATION IS SAME AS
$ THAT FOR OFFICE SO THAT CEILING IS THERMALLY NEUTRAL.
FLOOR-WEIGHT = 5
ZONE-TYPE = PLENUM ..
FRONT-P-WALL = EXTERIOR-WALL
X=40 Y=100 Z=9 AZ=0
H=4 W=40 TILT=90
CONSTRUCTION = EXT-WALL ..
FRONT-ROOF = ROOF
X=40 Y=100 2=13 AZ=0
H=50 W=40 TILT=0
CONSTRUCTION » MALL-ROOF
GND-REFLECTANCE = 0 ..
REAR-FLENUM = SPACE
LIKE FRONT-PLENUM ..
REAR-P-WALL = EXTERIOR-WALL
X=0 Y=0 Z=9 AZ=180
H=4 W=40 TILT=90
CONSTRUCTION = EXT-WALL ..
REAR-ROOF = ROOF
X=0 Y=0 Z=13 AZ=180
H=50 W=40 TILT=0
CONSTRUCTION = MALL-ROOF
GND-REFLECTANCE = 0 ..
$ AGAIN, NO SIDE WALLS ARE INDICATED FOR THE PLENUMS, ASSUMING
$ NO HEAT TRANSFER WITH THE PLENUMS OF ADJOINING OFFICES (OR
$ BETWEEN THE FRONT AND REAR PLENUMS).
A-3
-------
FRONT-OFFICE =
SPACE
x=o y=o z=o
AREA = 2000
VOLUME = 18000
SPACE-CONDITIONS
AZ=0
= OFFICE-ENVIRON ..
FRONT-O-WALL
EXTERIOR-WALL
X=40 ¥=100
H=9 W=40
CONSTRUCTION
Z=0 AZ=0
TILT=90
EXT-WALL
FRONT-WIN-1 - WINDOW X-l.l Y=2.6 H=6.3 W-17.8 SETBACK-0.33
GLASS-TYPE - ALL-WINDOWS FRAME-WIDTH =0.1 ..
FRONT-WIN-2 - WINDOW LIKE FRONT-WIN-1
EXCEPT X=21.1 W=7.3 . .
FRONT-WIN-3
WINDOW LIKE FRONT-WIN-1
EXCEPT X=28.6 Y=7.1 H=1.8 W=2.8
FRONT-WIN-4
WINDOW LIKE FRONT-WIN-1
EXCEPT X=31.6 W=7.3 ..
FRONT-DOOR
WINDOW
X=28.67 Y=0.17
GLASS-TYPE
FRNT-DOOR
H=6.66 W=2.66
FRAME-WIDTH
SETBACK=0.33
0.17 ..
FRONT-CEILING=
INTERIOR-WALL
AREA = 2000
X=40 Y=100 Z=9 AZ=0
CONSTRUCTION = CEILING
NEXT-TO FRONT-PLENUM ..
TILT=0
FRONT-SLAB = UNDERGROUND-FLOOR
AREA = 2000
X=40 Y=50 Z=0 AZ=0 TILT=180
CONSTRUCTION = SLAB
U-EFFECTIVE = 0.035 ..
$ TOE CALCULATED OVERALL U-VALUE FOR THE SLAB + CARPET + INSIDE-FILM-RESISTANCE
$ IS 0.39 BTU/HR-FT2-FO. THE REDUCED U-EFFECTIVE — USING DOE'S CONVENTION FOR
$ AVOIDING OVER-ESTIMATION OF HEAT TRANSFER THROUGH THE SLAB — IS CALCULATED
$ BY MULTIPLYING THIS U BY THE RATIO OF THE SLAB PERIMETER (180 FT) TO THE SLAB
$ AREA (2000 FT2).
REAR-OFFICE = SPACE
LIKE FRONT-OFFICE ..
REAR-O-WALL =
REAR-WIN-1
REAR-WIN-2
REAR-DOOR
EXTERIOR-WALL
X=0 Y=0 Z=0 AZ=180
H=9 W=40 TILT=90
CONSTRUCTION = EXT-WALL ..
= WINDOW X=5.1 Y=2,6 H=5.8 W=8.3 SETBACK=0.33
GLASS-TYPE = ALL-WINDOWS FRAME-WIDTH = 0.1 ..
WINDOW
EXCEPT
LIKE REAR-WIN-1
X=27.1 ..
DOOR
X=18 Y=0 H=7 W=3
CONSTRUCTION = BACK-DOOR
A-4
-------
mid-o-wall
REAR-CEILING =
REAR-SLAB
INTERIOR-WALL
X=0 Y=5Q Z=0 AZ=180
H=9 W=40
CONSTRUCTION » INT-WALL
NEXT-TO FRONT-OFFICE ..
INTERIOR-WALL
AREA = 2000
X=0 Y=0 Z=9 AZ=180
CONSTRUCTION = CEILING
NEXT-TO REAR-PLENUM ..
UNDERGROUND-FLOOR
AREA = 2000
X=0 Y=50 2=0 AZ=180
CONSTRUCTION « SLAB
U—EFFECTIVE ¦ 0.035 ..
TILT=0
TILT=180
LOADS-REPORT
$-
-REFORTS-
¦$
VERIFICATION=(LV-D)
SUMMARY=(LS-C)
END ..
COMPUTE LOADS
INPUT SYSTEMS
-SYSTEMS SCHEDULES-
OFFICE-COOL-T = SCHEDULE
OFFICE-HEAT-T = SCHEDULE
FAN-ON = SCHEDULE
OCCUPY-1 = SCHEDULE
THRU DEC 31
(MON,FRI)
(SAT,HOL)
THRU DEC 31
(MON,FRI)
(SAT,HOL)
THRU DEC 31
(MON,FRI)
(SAT,HOL)
THRU DEC 31
(MON,FRI)
(SAT,HOL)
(1,6)(99) (7,19)(75)
(20,24)(99)
(1,24)(99) ..
(1,6)(55) (7,19)(70)
(20,24)(55)
(1,24)(55) ..
(1,6)(0) (7,19){1)
(20,24)(0)
(1,24)(0) ..
(1,6)(0) (7)(.3)
(8,11)(1) (12,14)(.8,.4,.8)
(15,17)(1) (18,20)(,3,.1,.1)
(21,24)(0)
(1,24)(0) ..
5 DEFINITION OF ZONES- $
OFFICE-ZAIR = ZONE-AIR
OA-CFM/PER = 5 ..
A-5
-------
OFFICE-ZCONT = ZONE-CONTROL
DESIGN-COOL-T = 74
COOL-TEMP-SCH = OFFICE-COOL-T
DESIGN—HEAT-T =71
HEAT-TEMP-SCH = OFFICE-HEAT-T
THERMOSTAT-TYPE = PROPORTIONAL
TKROTTLING-RANGE = 2 ..
$ DEFAULT THROTTLING-RANGE VALUE WHEN THERMOSTAT-TYPE = PROPORTIONAL
FRONT-OFFICE = ZONE
ZONE-TYPE = CONDITIONED
ZONE-AIR = OFFICE-ZAIR
ZONE-CONTROL = OFFICE-ZCONT
SIZING-OPTION = ADJUST-LOADS ..
$ SIZING-OPTION MUST BE SET AT ADJUST-LOADS WHEN AN OVERHEAD PLENUM IS
$ PRESENT, SO THAT THE SYSTEMS CALCULATION WILL NOT IGNORE ROOF LOAD
$ WHEN SIZING EQUIPMENT.
REAR-OFFICE = ZONE LIKE FRONT-OFFICE .-
FRONT-PLENUM = ZONE
ZONE-TYPE = PLENUM
DESIGN-COOL-T = 90
DESIGN-HEAT-T =50
SIZING-OPTION = ADJUST-LOADS ..
REAR-PLENUM = ZONE LIKE FRONT-PLENUM ..
$ DEFINITION OF SYSTEMS $
$ THE FRONT AND REAR OFFICES EACH HAVE A PACKAGED SINGLE-ZONE, CONSTANT-VOLUME,
$ DIRECT EXPANSION COOLING SYSTEM WITH CENTRAL ELECTRIC HEATING.
PKG-AC-CONTROL = SYSTEM-CONTROL
MAX-SUPPLY-T = 100
MIN-SUPPLY-T =55
ECONO-LIMIT-T = 68
ECONO-LOCKOUT = YES ..
$ DEFAULT ECONO-LOCKOUT VALUE FOR PSZ SYSTEMS
PKG-AC-AIR = SYSTEM-AIR
OA-CONTROL = TEMP
MAX—OA—FRACTION = 1 ..
$ THESE SETTINGS INDICATE THAT THE SYSTEM HAS AN ECONOMIZER (CONTROLLED BY
$ SUPPLY AIR T), AND CAN OPERATE AT UP TO 100% OA.
PKG-AC-FAN = SYSTEM-FANS
FAN-CONTROL = CONSTANT-VOLUME
FAN-SCHEDULE = FAN-ON
NIGHT-CYCLE-CTRL = CYCLE-ON-FIRST
FAN-PLACEMENT = DRAW-THROUGH ..
$ THESE SETTINGS INDICATE THAT THE FANS ARE TURNED OFF OVERNIGHT AND ON
$ WEEKENDS/HOLIDAYS WHEN NO-ONE IS IN THE BUILDING, BUT THAT THEY WILL CYCLE ON
$ IF NEEDED TO MAINTAIN TEMPERATURE.
A-6
-------
PKG-AC-TERM = SYSTEM-TERMINAL
MIN-CFM-R&TIO = 1.0
REHEAT-DELTA-T = 0 ..
$ TRADITIONAL SETTINGS FOR CONSTANT VOLUME SYSTEMS.
PKG-EQUIP-FRONT = SYSTEM-EQUIPMENT
COOLING-EIR = 0.341
S THIS EIR CORRESPONDS TO E1R =10.0 BTU/H OUTPUT/WATT OP ELECTRIC INPUT
HEATING-1IR =1.0 ..
$ IF HEATING-EIR WERE ALLOWED TO DEFAULT, PROGRAM WOULD MODEL HEAT PUMP.
PKG-EQUIP-REAR = SYSTEM-EQUIPMENT
LIKE PKG-EQUIP-FRONT ..
FRONT-SYSTEM
SYSTEM
SYSTEM-TYPE
SYSTEM-CONTROL =
SYSTEM-AIR
SYSTEM-FANS
SYSTEM-TERMINAL =
SYS TEM-EQUIPMENT=
HEAT-SOURCE
RETURN-AIR-PATH =
PLENUM-NAMES
ZONE-NAMES =
PSZ
PKG-AC
PKG-AC
PKG-AC-
PKG-AC'
PKG-EQUIP
ELECTRIC
PLENUM-ZONES
(FRONT-PLENUM)
(FRONT-OFFICE,
CONTROL
AIR
FAN
TERM
FRONT
FRONT-PLENUM)
HEAT-ELEC-METER = M2
COOL-ELEC-METER = M3
HTREJ-ELEC-METER= M4
AUX-ELEC-METER = M4
VENT-ELEC-METER = MS
THE VARIOUS HVAC SYSTEM COMPONENTS ARE PLACED ON SEPARATE METERS,
TO TRACK HOW HVAC kW AND kWh ARE DISTRIBUTED.
REAR-SYSTEM = SYSTEM LIKE FRONT-SYSTEM EXCEPT
SYSTEM-EQUIPMENT= PKG-EQUIP-REAR
PLENUM-NAMES = (REAR-PLENUM)
ZONE-NAMES = (REAR-OFFICE, REAR-PLENUM) ..
$ PLANT ASSIGNMENTS $
TOTAL-OFFICE = PLANT-ASSIGNMENT
SYSTEM-NAMES = (FRONT-SYSTEM, REAR-SYSTEM)
DHW-GAL/MIN =0.1
DHW—SIZE = 30
$ DEFAULT DHW TANK SIZE FOR 0.1 GAL/MIN USAGE IS 30 GAL
DHW-SUPPLY-T = 140
$ DEFAULT SUPPLY T. ALSO, HIGH T PROTECTS AGAINST LEGIONELLA.
DHW-TYPE = ELECTRIC
DHW-SCH = OCCUPY-1
DHW-1LEC-METER = Ml . .
$ THIS ASSUMES THAT THERE IS A DEDICATED DOMESTIC HOT WATER (DHW) HEATER, AND
$ THAT MAXIMUM DHW CONSUMPTION IS 2 GAL/DAY PER PERSON DURING OCCUPIED HOURS.
A-7
-------
$ SYSTEM REPORTS $
SYSTEMS-REPORT VERIFICATION(SV-A)
SUMMARY=(SS-A,SS-F,SS-N) ..
END ..
COMPUTE SYSTEMS ..
input plant ..
TOTAL-OFFICE = PLANT-ASSIGNMENT
$ —DEFINE PLANT EQUIPMENT $
HT-WTR-HTR = PLANT-EQUIPMENT
TYPE = ELEC-DHW-HEATER
SIZE = -999
INSTALLED-NUMBER « 1 ..
PLANT-PARAMETERS
ELEC-DHW-LOSS = 0.03
$ PLANT REPORTS $
PLANT-REPORT SUMMARY = (PS-B,PS-E,BEPS) ..
END
COMPUTE PLANT ..
INPUT ECONOMICS ..
§ COST OF ELECTRICITY $
$ THE FOLLOWING CHARGES REPRESENT THE FLORIDA POWER & LIGHT GSD-1 RATE SCHEDULE.
ELEC-COST « UTILITY-RATE
RESOURCE = ELECTRICITY
ENERGY-CHG = 0.0473
BLOCK-CHARGES = (DEMAND-MIN) ..
DEMAND-MIN = BLOCK-CHARGE
BLOCK1-TYPE = DEMAND
BLOCKl-DATA = <10, 0,
1, 9.96) ..
$ THIS RATE INCLUDES A CONSTANT ENERGY CHARGE OF 4.73 CENTS PER kWh; THERE IS
$ NO DEMAND CHARGE FOR THE FIRST 10 kW, THEN A CHARGE OF 9.96 DOLLARS/kW FOR
$ EACH ADDITIONAL kW.
A-8
-------
$ ECONOMICS REPORTS $
ECONOMICS-REPORT SUMMARY = (ES-E)
END . .
COMPUTE ECONOMICS ..
STOP
A-9
-------
[This page intentionally blank.]
A-10
-------
APPENDIX B
Rationale for Values Selected for Variables in
D0E-2.1E Input File: 4,000 ft2 Office in Strip Mai
Variable Value Rationale for Selection
LOADS Variables
General Building
Description
DO
BUILDING-LOCATION
Building AZIMUTH
(AZIMUTH = 0 for
front wall)
BUILDING-SHADE
Baseline: One-story office
in slab-on-grade strip mall.
Office dimensions are 40 ft
frontage by 100 ft depth;
subdivided into two 40 x 50-
ft offices. Side walls are
interior walls adjoining
conditioned space occupied
by other tenants; the front
and rear walls are exterior.
Conditioned space has 9-ft
ceilings, with 4-ft uncondi-
tioned space overhead which
can serve as a return air
plenum if desired.
Parametric variations: None
Baseline: Miami, Florida
Parametric variations; None
Baseline: 0 (building
faces north)
Alternative values: 90,
180, 270 (building faces
east, south, and west)
Baseline: 10-ft exterior
overhang on front, at
ceiling height. No over-
hang on rear. 2-ft parapet
extending above front of
roof (but not rear).
Low value: 6-ft awning added
over windows and door on
rear, at ceiling height.
High value: No overhang on
either front or rear.
Represents one typical type of small office (e.g.,
insurance or travel agencies). The length of frontage
and the depth are typical of strip mall space. Sub-
division of total floor area into two spaces is repre-
sentative, and allows separate coolxng of the north and
south zones. The 9-ft ceilings are common for office
space, although higher ceilings are also commonly
encountered in strip malls, A plenum height of 4 ft is
reasonable, although shorter (3-ft) plenums are commonly
used by modelers, and plenums taller than 4 ft are
commonly encountered.
Desire to test hot, humid climate.
Arbitrary assumption.
Typical of strip malls containing offices.
Models extreme case.
(continued)
-------
APPENDIX B (continued)
Variable
Value
Rationale for Selection
AREA/PERSON
Baseline: 150 ft2/person
PEOPLE-HEAT-GAIN
Occupancy Pattern
(PEOPLE-SCHEDULE)
00
1
LIGHTING-TYPE
LIGHTING-W/SQFT
EQUIPMENT-W/SQFT
Low value; 300 ft2/person
High value: 100 ft2/person
Baseline: 450 Btu/h/person
Parametric variations: None
Baseline: Full occupancy at
8-llam and 2-5pm; 80% at
llam-noon and l-2pm; 40% at
noon-lpm? 30% at 6-7am and
5-6pm; 10% at 6-8pm; zero
at other times.
Parametric variations; None
Baseline: Recessed fluores-
cent non-vented.
Parametric variations: None
Baseline: 1.8 W/ft2
Low value: 1.0 W/ft2
High value: 2.25 W/ft2
Baseline: 0.75 W/ft2
Low value: 0.5 W/ft2
High value: 1,75 W/ft2
Approximately the maximum occupancy for offices (seven
persons per 1000 ft2) recommended for office space in
ASHRAE 62-1989 {ASHRAE, 1989a). The ASHRAE 62-1989
value for malls (20 persons/1000 ft2, or 50 ft2/person)
is not used since this space is deemed to represent
office space rather than typical mall retail space.
50% decrease in number of occupants, compared to ASHRAE
62-1989 recommendation.
50% increase in number of occupants.
Adjusted sensible plus latent heat gain per person indi-
cated in ASHRAE Fundamentals (ASHRAE, 1993) for moder-
ately active office work.
A reasonable occupancy schedule for a small strip-mall
office. The schedule used in prior modeling by FSEC
(Shirey and Rengarajan, 1996).
Representative of modern offices.
The prescriptive lighting power allowance for office
buildings of this size in ASHRAE 90.1-1989 (ASHRAE,
1989b).
Representing efficient lighting in modern offices, e.g.,
utilizing daylighting (Todesco, 1996; Birdsall, 1995).
A 25% increase in the prescriptive allowance from ASHRAE
90.1-1989. Representing less efficient modern lighting
(or better-lit space).
The average receptacle power density for offices speci-
fied in ASHRAE 90.1-1989.
More limited usage of computers, copiers, etc., relative
to the average receptacle power density in ASHRAE 90.1-
1989.
Heavier usage of computers, copiers, etc.
(continued)
-------
APPENDIX B (continued)
variable
Value
Rationale for Selection
LIGHTING- and
EQUIP-SCHEDULE
Baseline values Full power
at 7am-5pm? 30% at 6-7am
and 5-6pm; 10% at 6-8pm;
5% at all other times.
Parametric variations; None
A reasonable schedule for a small office. The schedule
used in prior modeling by FS1C (Shirey and Rengarajan,
1996).
AIR-CHANGES/HR
Baseline:
0.1 ACH
Low value; 0 ACH
High value; 0.3 ACH
Representing average air infiltration rate for offices
which are reasonably tight and/or which operate with
much of the interior near neutral pressure, or slightly-
pressurized.
Representing offices which are in fact pressurized at
all locations.
One of the higher infiltration rates assumed by DOE-2
modelers for modern offices.
CD
I
W
exterior wall
Construction
(front and
rear walls)
Baseline: 8-in. heavy weight
hollow concrete block wall;
1-in. stucco layer on exter-
ior, 3/4-in. expanded poly-
styrene insulation and 5/8-
in. gypsum board layers on
interior. U-value (including
hj and h„) =0.16 Btu/h ft2 F°
(h0 averages 2.5 Btu/h ft" F°
at Miami weather conditions).
Intermediate value; Frame
wall with 3.5-in. R-ll batt
insulation. U-value (incl.
^ and h„)=0.06 Btu/h ft2 F°.
Low value; No heat transfer
across exterior wall; infi-
nite wall resistance.
High value; Concrete block
wall, like baseline, except
delete 3/4-in. insulation.
U-value (incl. h; and h„) =
0.34 Btu/h ft2 F".
Construction used by Florida Solar Energy Center (FSEC)
in modeling a similar office (Shirey and Rengarajan,
1996). Representative of a construction encountered in
Florida (Odom and DuBose, 1994) . The U-value of the wall
is lower than the maximum (1 Btu/h ft2 F°) determined
using either the prescriptive or the performance criteria
for Miami in ASHRAE 90.1-1989 (ASHRAE, 1989b).
Representative of a construction encountered in Florida
(Odom and DuBose, 1994).
Models the extreme case.
Models extreme case.
(continued)
-------
APPENDIX B (continued)
Variable Value Rationale for Selection
WINDOW Area
(incl. frame)
Window Frame Type
Baseline; Windows comprise
33% of gross exterior wall
area.
- 46% of front wall area.
- 20% of rear wall area.
Low value: No windows on
either front or rear wall.
High value; Windows comprise
46% of gross exterior wall
area.
Baseline; Thermally broken
aluminum, 1.25 in. wide;
glass set back 4 in. from
exterior face of wall.
Parametric variations; None
Common for front walls in strip malls.
Occasionally observed in rear walls in strip malls,
depending on use of space adjoining rear wall. (More
commonly, rear walls have no window area.}
Models extreme case.
Assumes rear wall has same amount of fenestration as
front wall (extreme case). Towards upper end of range
considered by other modelers (Eto, 1990).
Representative of office space in strip malls.
03
GLASS-TYPE
Baseline; One pane, shading
coefficient = 0.55 (PANES =
1, SHADING-COEF = 0.55).
Overall D for fenestration
(incl. frame) = 0.94 Btu/
h ft2 F° at Miami weather
conditions (average h0 =
2. 5 Btu /h ft2 F°) .
Low value; PANES = 2,
SHADING-COEF = 0.16.
Fenestration U = 0.32 Btu/
h ft2 F" (Miami weather).
High value; PANES - 1,
SHADING-COEF = 0.94.
U = 0.94 Btu/h ft2 F°.
SHADING-COEF selected based on ASHRAE 90,1-1989 for Miami
climate for: front wall (48% glazing, significant exter-
ior shading); and rear wall (20% glazing, no exterior
shading). This SHADING-COEF can be achieved with single-
pane glass through various combinations of glass tinting,
glass reflective coatings, and interior shading (ASHRAE,
1993; ASHRAE, 1992b).
Double-pane glass with tinting and highly-reflective
coatxng (and/or interxor shading), representxng glass
having good resistance to conduction and superior
resistance to solar transmission (Todesco, 1996; ASHRAE,
1993; ASHRAE, 1992b).
A typical single-pane clear glass, 1/4-in. thick,
representing glass having minimum resistance.
DOOR Constructions Baseline; Front door is Reasonably typical.
glass with 2-in. aluminum
frame (treated as a window,
with glazing as above).
Rear door is brown steel
door with fiber core (U-
VALUE = 0.59 Btu/h ft2 F°,
ABSORPTANCE = 0.84).
Parametric variationss None
(continued)
-------
APPENDIX B (continued)
Variable
Value
Rationale for Selection
ROOF Construction
Baseline: Built-up roofing
over 4-in., R-12 mineral
board and metal deck. U-
Overall U-value of roof
(incl. hi and hj = 0.066
Btu/h ft2 F°.
Low value; No heat transfer
across roof; infinite roof
resistance»
High value: Replace R-12
mineral board in baseline
with 2-in., R-6.9 board.
Overall U = 0.12 Btu/
h ft2 F*.
The roof construction used by FSEC in modeling a similar
office (Shirey and Rengarajan, 1996).
Models extreme case.
Approximately doubles rate of heat transfer through roof
compared to baseline.
INTERIOR-WALL
Constructions
03
i
CJI
UNDERGROUND-FLOOR
Construction
Baseline; Walls adjoining
neighbors and subdividing
office are 5/8-in. gypsum
board on each side of 2-
by 4-in. studs. Ceiling
is acoustic tile.
Parametric variations; None
Baseline; 4-in. thick floor
slab with carpet. Overall
^EFFECTIVE (incl. h^ =
0.035 Btu/h ft2 F°.
Parametric variations: None
Typical.
Typical construction. The effective U-value was
estimated using the DOE-2 convention for avoiding over-
estimation of slab heat transfer {multiplying the actual
calculated U-value by the ratio slab perimeter:slab area).
SYSTEMS Variables
Cooling SCHEDULE
Heating SCHEDULE
Baseline; Cooling set point
in conditioned space is
75°F during occupied hours
(6am-7pm weekdays); off at
all other times.
Low value; Cooling set point
75°F during occupied hours;
set up to 81°F at all other
times during cooling season.
High value: None
Baseline: Heating set point
70°F during occupied hours;
set back to S5"F at all
other times (heating season)
Parametric variationsj None
Set point is a typical value, and is consistent with
ASHRAE 55-1992 (ASHRAE, 1992a). Turning cooling of f after
hours, and on weekends and holidays, is a common assump-
tion used by modelers, even in hot, humid climates (Shirey
and Rengarajan, 1996). In this case, office tempera-
tures reach a high of 85°F on summer Sunday afternoons.
Assess impact of night set-up in hot weather to protect
computers from high temperatures.
Typical values. Set point during occupied hours is
consistent with ASHRAE 55-1992.
(continued)
-------
APPENDIX B (continued)
Variable
Value
Rationale for Selection
Design Temperatures
for Zones
03
03
HVAC SYSTEM-TYPE
Baseline: Conditioned space:
DESIGN—COOL—T = 74°F,
DESIGN-HEAT-T = 71°P.
Plenum: DESIGN-COOL-T =
90°F, DESIGN-HEAT—T = 50°F,
Parametric variations;
Baseline:
None
Two rooftop direct
expansion, constant-volume,
packaged single-zone (PSZ)
units; one unit dedicated
to each of the two 2,000 ft2
zones in the office.
Alternative #1; A single PSZ
unit serving the entire
4,000 ft2 space.
Alternative #2: A single PSZ
unit serving the entire
space, with one of the two
sub-offices treated as a
VAV subzone. South office
is the control zone.
Alternative #3; Same as
Alternative #2, except that
north office is the control.
Alternative #4; A single two-
zone packaged VAV system
(PVAVS).
Alternative #5; Two packaged
terminal air conditioner
(PTAC) units, one in each
2,000 ft2 office.
These temperatures, which control the sizing of the
cooling and heating coils by DOE-2, are reasonable
selections for the conditioned space, in view of the
setpoints selected above under the cooling and heating
SCHEDULES.
Values sometimes selected by other modelers. Since the
plenum will in fact rarely get as hot as 90°F, even when
the system is off over weekends, selection of this
plenum DESIGN-COOL-T will result in moderate oversizing
of the cooling coils, since the conditioned space will
be estimated by LOADS as receiving artificially high
heat gain from the plenum. At 90°F, this oversizing
will be an amount which will limit the hours undercooled
and at RH > 60% to about 1%, considered to be a reason-
able percentage for these computations. [The selection
of DESIGN-HEAT-T is of little importance, since, in
Miami, heating requirements are so small that the cool-
ing load predominates in equipment sizing.]
Typical HVAC systems for offices of this size. Allows
reasonably effective control of temperatures on the
sunny vs. shaded sides of the office. Most computations
made with this HVAC configuration.
Also representative. Provides less effective temperature
control in the space, now treated as one large zone. PSZ
units of this capacity commonly available.
Included to illustrate the effect of this approach.
Also a reasonable choice; might be somewhat less common
than PSZ for this application.
Generally less representative of new construction.
(continued)
-------
APPENDIX B (continued)
Variable
Value
Rationale for Selection
RETURN-AIR-PATH
Baseline; PLENUM-ZONES
(all systems except PTAC;
DIRECT for PTAC).
Alternative value: DUCT
HVAC system air return via the overhead plenum is a
common choice. Through-the-wall units such as PTAC
necessarily draw the return air directly from the room.
Ducted returns are also a reasonable choice.
Supply Temperatures
for System Air
Economizer
03
Baseline; MIN-SUPPLY-T =
55°F; MAX-SUPPLY-T = 100°F
(for PVAVS, REHEAT-DELTA-T=
45 F") .
Low value:
42°F.
High value:
MIN-SUPPLY-T
None
Baseline: Economizer with
lockout; ECONO-LIMIT-T =
68°F. (Exception: PTAC
units have no economizer.)
Alternative value:
economizer.
No
Supply temperature for cooling air (55°F) is the commonly
selected value, being above the typical dew point and thus
avoxdxng condensation on the supply ducts. Supply
temperature for heating air (100°F) is lower than other
modelers commonly use (105-120"F), in view of Miami's
limited heating needs; MAX-SUPPLY-T has minimal impact
on these computations.
Models case of cold—air diffusers.
Economizers are increasingly common in packaged HVAC
units. Economizer lockout (preventing economizer from
functioning when cooling system must also operate) is
common in these small systems, to prevent the hermetic
compressor motors from overheating at times when econo-
mizer operation would create refrigerant flows too low
to adequately cool the motor. PTAC units are not con-
figured to accommodate an economizer. 68°F is one
typical value for the outdoor air temperature at which
the economizer returns to minimum outdoor air operation.
COOLING—EIR
Heating
Baseline: 0.341 (EER = 10).
(Exception; PTAC units
have EIR = 0.438, or EER =
7.8. )
Low value: 0.284 (EER = 12).
High value: 0.427 (EER = 8).
Baseline: Electric resistance
(HEATING-EIR =1). Central
heating coils for all sys-
tems except PVAVS; terminal
reheat for PVAVS.
Parametric variations: None
An electric input ratio of 0.341 Btu/h electric input
per Btu/h of cooling output (corresponding to an energy
efficiency ratio of 10 Btu/h of cooling output per W of
electric input), is representative of reasonably effi-
cient modern PSZ units. (The DOE-2 default value for
the COOLING—3IR of PSZ units is 0.360.) The EIR for the
less efficient PTAC units is the DOE-2 default value.
Represents about the most efficient commercial units.
Represents about the least efficient commercial units.
One reasonable selection.
Air-source heat pumps (HEATING-EIR < 1) not considered
in view of low heating load in Miami.
(continued)
-------
APPENDIX
Variable
Value
COOLING-CAPACITY
Baselines 103.6 kBtu/h at
5 cfm/person, 119.2 kBtu/h
at 20 cfm/person.
Low value: 102 kBtu/h (5
cfm/person)
Sensible heat ratio
(SHR)
High values; 108, 114, and
120 kBtu/h (5 cfm/person);
120, 126, and 132 kBtu/h
(20 cfm/person).
Baseline; 0.75 at 5 cfm/per-
son, 0.70 at 20 cfm/person
Alternative #1: SHR = 0.78
Alternative #2; SHR =0.73
00
00
PLANT Variables
Domestic Hot Water
(DHW)
Baseline; Electric resistance
30-gal hot water heater,
delivering up to 0.1 gal/min
(depending on occupancy) at
140°F.
Parametric variations; None
B (continued)
Rationale for selection
The default values computed by the DOE-2 program.
Slightly more than half of these capacities resides in
the unit conditioning the rear zone.
The commercially available capacity (corresponding to 8.5
tons of refrigeration) closest to the default capacity
computed by the program at 5 cfm/person.
Commercially available capacities providing some excess
capacity above the computed default values.
The default values computed by the program.
A value typical of commercial units containing three rows
of cooling coils.
A value typical of commercial units containing four rows
of coaling coils.
DHW delivery rate based on maximum rate of 2.0 gal/person/
day estimated for offices in the ASHRAE Applications Hand-
book (ASHRAE, 1995), based on an 8-hour work day. Storage
tank size is DQE-2 default minimum value. Supply tempera-
ture, higher than the 105-115'F recommended by ASHRAE
for office lavatories, is to protect against growth of
Legionella (ASHRAE, 1995). [Note; For simplicity, no
lavatory is included in the modeled office; a lavatory
would be modeled as part of one of the 2,000 ft2 sub-
offices, and would have minimal impact on the calcula-
tions. To enable these calculations, it is assumed that
the exhaust rate from such a lavatory would never be
greater than the rate at which outdoor air is supplied
by the HVAC system, even at 5 cfm/person (67 cfm per
sub-office), even though 67 cfm could be insufficient
bathroom exhaust by some codes.]
-------
APPENDIX B (concluded)
Variable Value Rationale for Selection
ECONOMICS Variables
Cost of Electricity Baseline: Energy charge of The GSD-1 rate structure for Florida Power and Light Co.
4.73 £/kWh; demand charge which serves the Miami area (Shirey and Rengarajan, 1996)
of zero up to 10 kW usage,
then $9.96/kW for each kW
above 10.
Parametric variations; None
Weather File Variables
Weather File Baseline: Typical Meteorolog-
ical Year (TMY) file for
Miami.
Alternat ive; Weather Year for
Energy Calculations (WYEC)
file for Miami.
------- |