Evaluation of Variable Refrigerant

Flow (VRF) System Performance

Using an Occupancy Simulated
Research Building: Introduction and
Summer Data Analysis Compared
with a Baseline RTU System
Piljae Im, PhD

Jeffrey D. Munk

Member ASHRAE

Member ASHRAE

Kwanwoo Song

ABSTRACT
Variable refrigerantflow (VRF) heat pumps are relatively new to the U.S. market although these systems have been used in many countries in
Europe and Asiafor more than 25years. Although there are known benefits like easier modular installation and higher efficiency, several studies
show that there are concerns regarding the application of VRF systems in the U.S, including 1) lack of awareness of energy efficiency advantages,
and 2) higherfirst cost. VRF system performance has been measured mainly in laboratories, which are well controlled but cannot replicate how a
unitperforms in an actual installation. In real building installations, the performance measurement of this type of system would be a challenge due
to the complexty of the system and number of sensors required to calculate the precise performance. In addition, without controlling the human
behavior in the real building which is difficult, there are too many uncertainties to evaluate the operational performance of the system. Oak
Ridge National Laboratorys Flexible Research Platform (FRP) overcomes this hurdle by simulating occupancy through the use of controlled
lighting, plug loads, internal sensible and latent heat generation, and HVAC operational schedules. To this end, a 12 ton VRF system was
installed in the FRP in June of 2014 and was operated every other week alternately with a 12.5 ton conventional roof top unit. The hourly
energy consumption of both systems was characterized based on outdoor air temperature, and energy savings for the cooling season are estimated to
be 19.8% for the VRF system.

INTRODUCTION
Multi-split air conditioning systems or variable refrigerant flow (VRF) systems are relatively new systems in the
U.S. market although the systems have been used in many countries in Europe and Asia for more than 25 years. The
known benefits of the VRF systems include 1) easier modular installation (particularly beneficial for building retrofit),
2) space savings with multiple indoor units connected to one outdoor unit, 3) ability to respond to fluctuations in
space load conditions, 4) easier and cost effective maintenance and commissioning, and 5) energy efficiency. Although
VRF ststems are receiving increased attention in the U.S. marketplace due to these known benefits, several studies
(Goetzler 2007, Amarnath and Blatt 2008, Aynur et al 2009) show that there are several main concerns for the
Piljae Im and Jeffrey Munk are R&D staffs in Building Technologies Research and Integration Center (BTRIC), Oak Ridge National Laboratory, Oak Ridge, TN.
Kwanwoo Song is a principal research engineer in Smart Home Solution Laboratory at Samsung Electronics.

application of VRF systems in the U.S., which include 1) lack of awareness and studies of energy efficiency
advantages, 2) higher first cost - about 5% to 20% higher than the chilled water systems or 30% to 50% more than
single packaged ducted system (Amarnath and Blatt, 2008), 3) lack of optimized /integrated VRF control in energy
management control system (EMCS), 4) code compliance issues specific to the U.S., and 5) lack of training of the
complicated VRF installation and/or differences in construction practices.
There have been numerous studies on the multi-split VRF systems since the systems were introduced in Japan
about 30 years ago. Most studies include field and laboratory experiments and/or energy modeling analysis. In general,
field experiments were performed with a set of VRF indoor units and outdoor unit, and their energy consumption and
performance were analyzed in real buildings in several different scenarios (e.g., varying indoor loads, different control
modes, ventilation impact, etc.). The field studies verified the measured cooling/heating performance of the VRF
systems in these various conditions, while it was relatively hard to compare the VRF system performance with other
HVAC systems in the same building. In other studies, energy modeling has been used to compare the VRF system
performance with other HVAC systems in a building with the same conditions. Some modeling studies use generic
building models to compare the performance and energy consumption of VRF and other HVAC systems, while
others use calibrated models to do the same analysis. Although the latter case would provide better analysis results
than the former, it still has limited accuracy when predicting the precise energy consumption of the alternative system
in a real building due to uncertainties in the model algorithm itself and human behavior and operational schedules of
the building. Therefore, it is desired to have a research building that is representative of real buildings with control of
occupancy and other internal load profiles. The ability to operate multiple HVAC systems alternately under the same
climate conditions would allow for a more direct comparison of the systems as they would be operating in the same
building under similar outdoor weather conditions.

FLEXIBLE RESEARCH PLAFORMS

Recently, Oak Ridge National Laboratory (ORNL) constructed two light commercial building flexible research
platforms (FRPs) on its campus: one single-story FRP and one two-story FRP (Figure 1). The two-story FRP exposes
the test buildings to natural weather conditions for purposes of research and development (R&D) leading to
system/building-level advanced energy efficiency solutions for new and retrofit applications. The buildings occupancy
is emulated with portable heaters and humidifiers so that variations in energy use between the two systems are
attributable only to the system performance and weather. Variations from different weather conditions can be
normalized using weather normalization or calibrated simulation modeling. The two-story FRP, having a footprint of
40x40 (12 m x 12 m) (Figure 1), was used in this study and simulates light commercial buildings common in the
nations existing building stock.

RESEARCH PLAN
FRP with Baseline System
As a first step of the study, a baseline test building was designed, constructed on the permanent apparatus of the
two story FRP, and monitored in series based on the literature review (EIA 2003, ASHRAE 1989, Huang et al., 1991)
and communication with industry partners. The baseline building was designed to be representative of a typical
existing low-rise office building in U.S, which is built in the 1990s. Building materials and equipment that were
typically implemented in the 1990s were preferably applied to the permanent part of the FRP. The building has 10
conditioned zones including 2 core zones and 8 perimeter zones. The baseline 12.5 ton roof top unit (RTU) is rated at
9.7 EER, 10.3 IPLV, and 81% AFUE. Each room has a variable air volume (VAV) box with an electric resistant
heating element. The central fan in the air handling unit (AHU) draws the return air from each room. An exhaust fan
is located in each floor, and the exhaust fans are operating when the supply fan is running. The rooftop unit is
programmed to maintain a constant discharge air temperature (set to 58F (14.4C)). The natural gas furnace portion

Permanent Apparatus

tmm
Figure 1

FRP permanent apparatus (above), single-story test FRP (lower left) and 2 story test FRP (lower

right)

of the RTU would engage if the building return air temperature were to drop below 58F. As long as the discharge air
is at least 58F, the zone electric heat in the VAV boxes will activate to provide the reheat. The buildings setpoint
temperature was set to 70F (21.1C) and 75F (23.9C) for heating and cooling, respectively. During unoccupied
hours (6:00 pm to 6:00 am), the setback and setup temperatures are set to 60F and 86F, respectively. The baseline
building characteristics are summarized in Table 1.
The baseline systems performance is monitored using multiple sensors in the building to measure RTUs
capacities, electric power consumption and gas flow rate, and indoor condition. The performance of the system is
being monitored by both refrigerant and air side measurements. Room temperature and relative humidity sensors as
well as supply, return, mixed air temperature and relative humidity sensors are also being monitored. For the RTU
power, compressor and fan power have been measured individually, and the individual rooms reheat power was also
monitored. All was collected and stored every 30 seconds.

Table 1.
Building Component
Location

Baseline Test Building Characteristics

Value
Oak Ridge, TN

Building Width

40 feet (12.2 meter)

Building Length

40 feet (12.2 meter)

Story Height (floor to floor)

Number of Floors
Number of thermal zones

14 feet (4.3 meter)

2
10 (8 parameter & 2 core zones)

Wall Structure

Concrete Masonry Units (CMU) with face brick

Wall Insulation

Fiberglass Rus-11 (Rsi-1.9)

Floor

Slab-on-grade

Roof Structure

Metal deck with polyiso & ethylene proplylene diene monomer (EPDM)

Roof Insulation

Polyiso Rus -18(Rsi-3.17)

Windows

Double clear glazing

Window-to-Wall Ratio

28%

Lighting power density

0.85 W/sqft (9.18W/m2)

Equipment power density

1.3 W/sqft (14.04W/ m2)

Baseline RTU Capacity

EER
Gas furnace efficiency
Reheat

12.5 ton
9.7
81% AFUE
VAV box with elec. reheat

FRP with VRF System

The VRF system was designed based on the load calculation from Manual N (Rutkowski 2008), and the
corresponding indoor units and outdoor unit were recommended by the manufacturer. Figure 2 shows the location
the indoor units and the assigned rooms in the first and second floor of the FRP. The outdoor unit is a type of heat
pump (i.e., no heat recovery operation) and equipped with two scroll compressors, and charged with R-410a. Total
cooling and heating capacities of the outdoor unit are about 144 MBH (42.2 kW) and 162 MBH (47.5 kW),
respectively, and 10 indoor units (6 wall mounted and 4 ceiling mounted) cooling and heating capacities are ranged
from 7.5 (2.2 kW) to 18 MBH (5.3 kW), and 8.5 (2.5 kW) to 20 MBH (5.9 kW), respectively. The rated EER and
IEER for the outdoor condensing unit are 11.2 and 22.7, respectively. Since the VRF system did not have any feature
for providing ventilation for current installation, a separate intake vent was installed for each floor to introduce the
outdoor air in conjunction with the existing exhaust fan in order to provide adequate ventilation. For pair comparison
between the RTU and VRF systems, the original outdoor air intake for RTU was completely closed, and the
ventilation air from this separate intake vent was used for both systems. The OA requirements for the building was
estimated according to the the ASHRAE Standard 62.1-2013 given the 225 sqft/person (20.9 m2/person) of occupant
density and 3,200 sqft (297 m2) of floor area.
The heating and cooling setpoints for the VRF system were the same as that of the baseline RTU.

Staircase
101

Room 102

Room 104

Room 103

High pressure
ref. line
Low pressure
ref. line

Room 105

Wall mounted
indoor unit

Room 106

4 way indoor

Condensing

Thermostat

First Floor

High pressure
ref. line
Low pressure
ref. line
Wall mounted
indoor unit
4 way indoor

From Condensing Unit

Thermostat

Second Floor

Figure 2

Layout of the VRF systems in the first and the second floors on the FRP

DMS-Bnet

SUMMER DATA COMPARISON ANALYSIS

Data from the summer of 2014 was analyzed by plotting the hourly energy consumption of each system during
occupied hours against the corresponding hourly average outdoor air temperature.
The results of this analysis are shown in Figure 3 and Figure 4. The VRF system has very high energy savings at
mild outdoor conditions when the unit is operating at part load, which is expected of a unit with variable speed
operation. In Figure 4, the increased energy use of the RTU between 50F (10C) and 68F (20C) is due to the VAV
boxes using the electric resistance elements to reheat the air to maintain a proper supply temperature for each zone as
seen. At high outdoor air temperatures, the energy use of the two systems converges. This is an expected trend since
the EER ratings of the units are fairly similar. Using the curve fits from Figure 3, the energy use for the period of
7/7/2014 to 9/30/2014 for the hours between 8 am and 6 pm was calculated for both systems. This data showed a
12.1% savings for the VRF system, Table 2.
In addition to the savings during regular operation, there was significant energy savings during periods of
standby operation, such as the unoccupied hours when the thermostat was setup. The RTU unit used, on average,
900 W with most of this being attributed to the fan for circulating the air. In contrast, the VRF unit only used 250 W
on average. If this standby energy use is included for the unoccupied hours, the VRF energy savings increase to
19.8% for the same period, Table 3.

12000
10000
JZ

O)
to
=)
>
tifl
O)
c

8000

6000
O RTU

LU

J>~

VRF

4000

2000

10

15

20

25

30

35

Hourly Ambient Air Temperature (C)

Figure 3 Hourly HVAC cooling energy use as a function of outdoor air temperature for the baseline roof top unit
and the VRF system.

12000

10000

OJ

8000
9

in

Z>
2?
<u

o\

6000

\o
o\

LU

3
O

4000

2000

o'^ 8
COD O

0
0

10

0CfGD Sfco "OCDOO'

15

20

25

30

35

Hourly Ambient Air Temperature (C)

Figure 4

Hourly reheat energy use as a function of outdoor air temperature for the baseline roof top unit

Table 2.

RTU and VRF Energy Use Estimates for the Period between 7/7/2014 and
9/30/2014 for the Hours of 8 am to 6 pm on Weekdays

Total Energy Use (Wh)

% Difference (vs. RTU)

RTU

VRF

4,909,974
-

4,315,795
12.1%

Table 3. RTU and VRF Energy Use Estimates for the Period between 7/7/2014 and
9/30/2014 for the Hours of 8 am to 6 pm on Weekdays and including Standby Power Use
for the Unoccupied Hours

Total Energy Use (Wh)

% Difference (vs. RTU)

RTU

VRF

5,635,374
-

4,517,295
19.8%

SUMMARY
The study to investigate a multi-split VRF systems energy performance compared to typical RTU with VAV
reheat was introduced in this paper. The systems were installed in ORNLs unique national user facility, aka FRP,
which was operated with simulated occupancy and controlled internal loads in natural weather exposure. The baseline
RTU and VRF system were operated and monitored every week alternately, and the system performances were
compared. The VRF system showed a 19.8% energy savings over the baseline RTU for the period of July 7th to
September 30th in 2014.

ACKNOWLEDGMENTS
Oak Ridge National Laboratory is managed by UT-Battelle, LLC, for the U.S. Dept. of Energy under contract DEAC05-00OR22725. This manuscript has been authored by UT-Battelle, LLC, under Contract Number DE-AC0500OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by
accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paidup, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to
do so, for United States Government purposes.

DISCLAIMERS
This effort was supported by Samsung Electronics and U.S. Department of Energy. Neither the United States
Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes
any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus,
product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to
any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not
necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or
any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the
United States Government or any agency thereof.

REFERENCES
Amarnath A., and M. Blatt. 2008. Variable Refrigerant Flow: An Emerging Air Conditioner and Heat Pump Technology,
2008 ACEEE Summer Study on Energy Efficiency in Buildings 3, 1-13.
ASHARE. 1989, ASHRAE Standard 90.1-1989 - Energy Standard for Buildings except Low-Rise Residential Buildings. Atlanta:
ASHRAE.
ASHRAE 2013, ANSI/ASHRAE Standard 62.1-2013-Ventilationfor Acceptable Indoor AirQuality. Atlanta: ASHRAE.
Aynur, T., Y. Hwang, and R. Radermacher. 2009. Simulation comparison of VAV and VRF air conditioning systems in an
existing building for the cooling season, Energy and Buildings 41, 1143-1150.
DOE. 2009. EnergyPlus Engineering Reference: The Reference to EnergyPlus Calculations".
Energy Information Administration (EIA). 2003. 2003 Commercial Building Energy Consumption and Expenditures (CBECS),
Public
Use
Data,
Micro-data
files
on
EIA
website:
http://www.eia.doe.gov/emeu/cbecs/cbecs2003/public use 2003/cbecs pudata2003.html.
Goetzler W. 2007. Variable Refrigerant flow systems, ASHRAE Journal 49 (4) 24-31.
Huang, J., H. Akbari, L. Rainer, and R. Ritschard. 1991. 481 Prototypical Commercial Buildings for 20 Urban Market Areas,
Technical Report, GRI-90/2036, LBL-29798.
Rutkowski, H. 2008, Manual N Commercial Load Calculation for Small Commercial Buildings, Air Conditioning Contractors of
America; 5th Edition.