SRI/USEPA-GHG-VR-12
September 2001
Environmental
Technology
Verification Report
KMC Controls, Inc.
SLE-1001 Sight Glass Monitor
Prepared by:
Greenhouse Gas Technology Center
Southern Research Institute
Under a Cooperative Agreement With
U.S. Environmental Protection Agency
-------
EPA REVIEW NOTICE
This report has been peer and administratively reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
SRI/USEP A-GHG-VR-12
September 2001
Greenhouse Gas Technology Center
A U.S. EPA Sponsored Environmental Technology Verification (£|-^ ) Organization
Environmental Technology Verification Report
KMC Controls, Inc.
SLE-1001 Sight Glass Monitor
Prepared By:
Greenhouse Gas Technology Center
Southern Research Institute
PO Box 13825
Research Triangle Park, NC 27709 USA
Telephone: 919/806-3456
Under EPA Cooperative Agreement CR 826311-01-0
U.S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711 USA
EPA Project Officer: David A. Kirchgessner
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES iii
LIST OF TABLES iii
ACKNOWLEDGMENTS iv
ACRONYMS/ABBREVIATIONS v
1.0 INTRODUCTION 1-1
1.1 BACKGROUND 1-1
1.2 SLE-1001 SIGHT GLASS MONITOR DESCRIPTION 1-2
1.3 TEST FACILITY DESCRIPTION, MODIFICATION, AND CHECKOUT 1-6
1.4 OVERVIEW OF VERIFICATION PARAMETERS AND EVALUATION
STRATEGIES 1-8
1.4.1 SGM Cost and Installation Requirements 1-9
1.4.2 Refrigerant Leak Detection Sensitivity 1-9
1.4.3 Estimated Potential Refrigerant Savings and Potential Cost Savings 1-15
2.0 VERIFICATION RESULTS 2-1
2.1 OVERVIEW 2-1
2.2 SGM COST AND INSTALLATION REQUIREMENTS 2-1
2.3 REFRIGERANT LEAK DETECTION SENSITIVITY 2-2
2.3.1 System Refrigerant Evacuation and Leak Checking 2-2
2.3.2 SGM Leak Detection Sensitivity 2-3
2.4 ESTIMATED POTENTIAL REFRIGERANT AND COST SAVINGS 2-7
3.0 DATA QUALITY ASSESSMENT 3-1
3.1 DATA QUALITY OBJECTIVES 3-1
3.1.1 Leak Detection Sensitivity DQO Reconciliation 3-2
3.1.2 Data Completeness DQO Reconciliation 3-5
3.2 QA/QC CHECKS FOR NON-CRITICAL MEASUREMENTS 3-6
3.3 POTENTIAL REFRIGERANT COST SAVINGS 3-7
4.0 TECHNICAL AND PERFORMANCE DATA SUPPLIED BY KMC 4-1
5.0 REFERENCES 5-1
-------
LIST OF FIGURES
Page
Figure 1-1 Simplified Diagram of SGM Installation 1-3
Figure 1-2 KMC Sight Glass Monitor 1-5
Figure 1-3 Photographs of Test Systems 1-7
Figure 1-4 Refrigerant Leak Detection Sensitivity Testing Procedures 1-10
Figure 1-5 Refrigeration Gauge Manifold and Hoses 1-11
Figure 1-6 Simplified Diagram of Refrigeration Manifold System 1-12
LIST OF TABLES
Page
Table 1-1 Profiles of Test Systems 1-7
Table 2-1 SGM Costs 2-2
Table 2-2a Rooftop HVAC Full Charge Determinations 2-4
Table 2-2b Reciprocating Chiller Full Charge Determinations 2-4
Table 2-3a Test Unit Operating Conditions - Rooftop HV AC Unit 2-5
Table 2-3b Test Unit Operating Conditions - Reciprocating Chiller 2-5
Table 2-4a Leak Detection Sensitivity at Manufacturer Specified Full Refrigerant Charge 2-6
Table 2-4b Leak Detection Sensitivity with KMC Specified Full Refrigerant Charge 2-6
Table 2-5a Rooftop HV AC Unit Potential Annual Refrigerant Savings 2-8
Table 2-5b Reciprocating Chiller Potential Annual Refrigerant Savings 2-8
Table 3-1 Measurement Instrument Specifications and Data Quality Indicator Goals 3-1
Table 3-2a Scale Calibration and Precision Data- Rooftop HVAC Unit 3-3
Table 3-2b Scale Calibration and Precision Data - Reciprocating Chiller 3-3
Table 3-3 Summary of Measurement and Leak Detection Sensitivity Errors 3-4
Table 3-4a Rooftop HVAC Data Completeness Goals 3-5
Table 3-4b Chiller Data Completeness Goals 3-6
Table 3-5 Summary of QA/QC Checks 3-6
111
-------
ACKNOWLEDGMENTS
The Greenhouse Gas Technology Center wishes to thank the following staff of North Carolina State
University for their activities in hosting this Verification: Wayne Friedrich, J.C. Boykin, and M.
Gonzalez. C. Warren Williams, with Jones Management Services, provided valuable operations support
to the test as a certified HVAC technician. The University provided the cooling and refrigeration units
that were used during the verification, allowed operations to be altered to support testing, and provided
key unit operating information. Thanks are also extended to the members of the GHG Center's Technical
Panel Reviewers including Cynthia Gage of U.S. EPA (NRMRL-RTP) and Richard Wells of Brady
Services for reviewing and providing input on the testing strategy and this Verification Report.
IV
-------
ACRONYMS/ABBREVIATIONS
CFC
DQI
DQO
EPA
ETV
°F
FID
ft2
g
GHGs
GHG Center
HCFC
HFC
hr
in.
KMC
Ib
N2
NCSU
NIST
ORD
PFC
psia
psig
QA/QC
QMP
RH
SGM
SRI
Test Plan
VA
chlorofluorocarbon
data quality indicator
data quality objective
Environmental Protection Agency
Environmental Technology Verification program
degrees Fahrenheit
flame ionization detector
square feet
grams
greenhouse gases
Greenhouse Gas Technology Center
hydrochlorofluorocarbon
hydrofluorocarbon
hours
inches
KMC Controls, Inc.
pounds
nitrogen
North Carolina State University
National Institute for Standards and Technology
Office of Research and Development
perfluorocarbon
pounds per square inch absolute
pounds per square inch gauge
Quality Assurance/Quality Control
Quality Management Plan
relative humidity
Sight Glass Monitor
Southern Research Institute
Test and Quality Assurance Plan
volt amperes
-------
-------
1.0 INTRODUCTION
1.1 BACKGROUND
The U.S. Environmental Protection Agency's Office of Research and Development (EPA-ORD) operates
a program to facilitate the deployment of innovative technologies through performance verification and
information dissemination. The goal of the Environmental Technology Verification (ETV) program is to
further environmental protection by substantially accelerating the acceptance and use of improved and
innovative environmental technologies. ETV is funded by Congress in response to the belief that there
are many viable environmental technologies that are not being used for the lack of credible third-party
performance data. With performance data developed under ETV, technology buyers, financiers, and
permitters in the United States and abroad will be better equipped to make informed decisions regarding
environmental technology purchase and use.
The Greenhouse Gas Technology Center (GHG Center) is one of several verification organizations
operating under ETV. The GHG Center is managed by the U.S. EPA's partner verification organization,
Southern Research Institute (SRI), which conducts verification testing of promising GHG mitigation and
monitoring technologies. The GHG Center's verification process consists of developing verification
protocols, conducting field tests, collecting and interpreting field and other test data, obtaining
independent peer review input, and reporting findings. Performance evaluations are conducted according
to externally reviewed Verification Test and Quality Assurance Test Plans (Test Plans) and established
protocols for quality assurance.
The GHG Center is guided by volunteer groups of stakeholders. These stakeholders offer advice on
specific technologies most appropriate for testing, help disseminate results, and review Test Plans and
Verification Reports. The GHG Center's stakeholder groups consist of national and international experts
in the areas of climate science and environmental policy, technology, and regulation. Members include
industry trade organizations, technology purchasers, environmental technology finance groups,
governmental organizations, and other interested groups. In certain cases, industry specific stakeholder
groups and technical panels are assembled for technology areas where specific expertise is needed.
Technical panel members assist in selecting verification factors and provide guidance to ensure that the
performance evaluation is based on recognized and reliable field measurement and data analysis
procedures. Also, selected members peer review key documents prepared by the GHG Center.
Among the most potent GHGs emitted to the atmosphere through anthropogenic activities are hydro-
chlorofluorocarbons (HCFCs) and hydrofluorocarbons (HFCs) currently used in most refrigeration and
air-conditioning systems worldwide. These chemicals have high global warming potentials and
extremely long atmospheric lifetimes, resulting in their essentially irreversible accumulation in the
atmosphere (EIA 1997, 1999). In the upper atmosphere, HFCs and HCFCs contribute to the destruction
of Earth's protective ozone layer. HCFC-22 (R-22) and HFC-134a are most often used in air-
conditioning and refrigeration equipment. Although these refrigerants are maintained in closed systems,
some of the refrigerant escapes to the atmosphere during routine installation, operation, and servicing of
the equipment. In addition, fugitive emissions escape into the atmosphere from leaky components,
resulting in further refrigerant loss. These releases to the atmosphere vary among different types and
sizes of equipment and operating practices, and directly contribute to greenhouse gas emissions.
EPA promulgated leak-repair requirements for systems containing CFCs and HCFCs (60 FR 40420)
under Section 608 of the Clean Air Act Amendments of 1990. More recently, EPA has proposed another
1-1
-------
rule (63 FR 32044) to include substitute refrigerants such as HFCs and perfluorocarbons (PFCs). Under
both rules, when refrigerant has leaked in a quantity that exceeds a specified trigger amount from an
appliance that normally contains a refrigerant charge of more than 50 Ib, the owner or operator of the
appliance must take corrective action.
In response to these EPA regulations, manufacturers have made improvements to reduce refrigerant loss
through design changes, and new equipment for measuring and detecting leaks has entered the market.
KMC Controls, Inc. (KMC), of New Paris, Indiana, and Future Controls, Inc. of Fort Myers, Florida,
have jointly developed a leak-monitoring device which allows refrigeration and air-conditioning
equipment operators to provide early detection of refrigerant loss. The device, titled the KMC SLE-1001
Sight Glass Monitor (SGM), identifies when a system's refrigerant charge is low and is in need of
maintenance and possible repair (including leak repair). This is accomplished using an infrared radiation
detector that continuously monitors the presence of flash gas through an existing refrigerant sight glass.
The ability of the SGM to detect relatively small levels of refrigerant loss is of significant interest to most
users, particularly those facing EPA regulations.
KMC requested that the GHG Center perform an independent third-party performance verification of the
SGM on commercial- and industrial-scale refrigeration and air-conditioning systems. SGM performance
was verified using air-conditioning and refrigeration systems at North Carolina State University (NCSU).
Details on the verification test design, measurement test procedures, and Quality Assurance/Quality
Control (QA/QC) procedures can be found in the Test Plan titled Test and Quality Assurance Plan for the
KMC Controls, Inc. SLE-1001 Sight Glass Monitor (SRI 2001). It can be downloaded from the GHG
Center's Web site (www.sri-rtp.com) or from the U.S. EPA Web site (www.epa.gov/etv). The Test Plan
describes the rationale for the experimental design, the testing and instrument calibration procedures
planned for use, and the specific QA/QC goals and procedures. The Test Plan was reviewed and revised
based on comments received from KMC, selected members of the GHG Center's stakeholder groups, and
the EPA Quality Assurance Team. The Test Plan meets the requirements of the GHG Center's Quality
Management Plan (QMP), and thereby satisfies ETV QMP requirements. In some cases, deviations from
the Test Plan were required. These deviations, and the alternative procedures selected for use, are
discussed in this report.
The remaining discussion in this section describes the SGM technology, describes the test facility,
discusses the verification approach, and lists the performance verification parameters that were quantified.
Section 2 presents the verification test results, and Section 3 assesses the quality of the data obtained.
KMC provided Section 4 containing additional information regarding the SGM, which has not been
independently verified by the GHG Center.
1.2 SLE-1001 SIGHT GLASS MONITOR DESCRIPTION
Heat in refrigeration and air-conditioning systems is transferred by a refrigerant operating in a closed
system. Refrigerated systems are primarily designed to cool products, whereas air-conditioning systems
cool spaces. Figure 1-1 illustrates a typical air-conditioning system. It consists of four basic components:
(1) compressor, (2) condenser, (3) expansion valve or flow controller, and (4) evaporator.
1-2
-------
Figure 1-1. Simplified Diagram of SGM Installation
ttttttttt
x°
8
KMC Sight /
Glass V
%•
3
ttttttttt
cooling fluid
(water or air)
umuu
^%
1 High Pressure
1
""' 1
a
Compressor
1
xo
8
-30 % vapor
Expansion
Valve
umuu
Evaporator
The compressor pressurizes the low-pressure refrigerant vapor, forming hot, high-pressure, superheated
vapor. The compressor also provides the motive force needed to circulate the refrigerant through the
other basic components and interconnecting piping network of the refrigeration system. The high-
pressure vapor discharged from the compressor enters a condenser that cools the refrigerant vapor to a
warm, high-pressure, sub-cooled liquid state. The condenser transfers the heat that was contained in the
vapor to the external cooling fluid (e.g., water or outdoor air).
The flow controller, often located immediately upstream of the evaporator coils, controls the flow of
refrigerant from the condenser to the evaporator. This device acts as a restriction to reduce the pressure
of the liquid refrigerant. Several types of flow controllers are used in the industry, with a thermostatic
expansion valve being one of the most commonly used controllers. The valve position is pre-adjusted to
maintain the optimum amount of refrigerant flow into the evaporator under varying indoor and outdoor
temperatures.
The evaporator serves to remove heat from the heat transfer fluid (indoor air or chiller water) passing over
it. Inside the evaporator, liquid refrigerant exiting the expansion valve boils and is converted into a vapor
as it absorbs heat from the indoor air or water. This cools the refrigerant, the evaporator coils, and the
indoor air or water. The cool vapor then returns to the compressor to be recompressed and recirculated.
If the incompressible liquid refrigerant reaches the compresssor, it can seriously damage the compressor
1-3
-------
(slugging). For this reason, all of the liquid refrigerant must be returned to a vapor state prior to leaving
the evaporator.
In order for a thermostatic expansion valve to operate properly, it must receive a continuous stream of
sub-cooled liquid (10 to 20 °F subcooling) at the proper pressure. To determine if the condenser is
supplying liquid refrigerant that meets these requirements, a sight glass is installed in the liquid line to
allow visual inspection of the refrigerant condition. Most commercial and industrial equipment is
manufactured with a sight glass near the condenser outlet, but ideally, the sight glass should be located as
near as possible to the thermostatic expansion valve (Moravek2001).
A clear liquid in the sight glass indicates that there is adequate refrigerant charge in the system to ensure
proper feed through the expansion valve. Bubbles in the sight glass, however, can indicate the presence
of refrigerant vapor or noncondensables in the liquid line (Moravek 2001). A continuous presence of
refrigerant vapor or bubbles during compressor operation can indicate that the system is short of
refrigerant charge. Noncondensables such as air, nitrogen, or other types of refrigerants not compatible
with the system design can also cause bubbles. The presence of these noncondensables can be related to
poor refrigerant evacuation activities that result from the operator's failure to completely evacuate air
from the system prior to charging. A major restriction in the liquid line, such as a clogged filter, can also
result in bubbles in the sight glass due to excessive pressure drop in the line. This restriction causes the
refrigerant to boil or flash off to a vapor. Drastic load changes and excessive compressor cycling may
also cause bubbles to form. Some sight glasses are equipped with a moisture element inside the sight
glass, which can indicate the presence of water.
Despite its intended purpose, the utility of the sight glass as a reliable indicator of low refrigerant charge
and moisture levels is often hampered by the relative inaccessibility of the sight glass, and the inability of
FfVAC technicians to properly interpret sight glass conditions. System operators do not routinely monitor
sight glasses during normal daily operations, and may not be aware of bubbles even though they may be
present.
The KMC SLE-1001 Sight Glass Monitor (SGM), shown in Figure 1-2, is designed to automatically
interpret the condition of the refrigerant and provide operators with audible alarms or remote feedback of
actual conditions. The SGM is an external device that is installed on an existing factory-installed sight
glass. It is specifically designed to be used with sight glasses marketed by the Sporlan® Valve Company,
which provides about 90 percent of sight glasses currently in operation. The SGM monitors two
conditions through the sight glass window: bubbles and moisture content. The device emits infrared (IR)
radiation which is reflected by bubbles in the refrigerant. When bubbles or flash gases of noncondensed
refrigerant are detected in the sight glass, a red light emitting diode (LED) on the monitor housing flashes.
The pulse frequency of the red LED increases with increased frequency of bubble detection. The
moisture LED changes from green to yellow when moisture is detected in the system. As moisture levels
increase, the LED glows brighter in proportion to the degree of moisture detected. In both cases, the
SGM provides a 0- to 5-V output that increases proportionally to the red LED flash frequency and the
yellow LED intensity. Because the SGM draws its operating power from the existing 24 VAC
refrigeration system control circuits, it does not normally require an external power supply.
1-4
-------
Figure 1-2. KMC Sight Glass Monitor
Figure 1-2. KMC Sight Glass Monitor
Flash Gas or
Refrigerant
Signal
Moisture
Signal
Existing
Sight Glass
0- to 5-V
Output to
Alarm
The SGM can be installed on existing systems with 0.25 in. or more clearance surrounding the exterior of
the sight glass window frame. The sight glass window must be clear and the side in contact with the
refrigerant should not be dark or discolored. The light sensor is placed flush over the sight glass window,
and the assembly is held firmly in place with two stainless steel springs that loop around the sight glass'
inlet and outlet pipes. Installed in this manner, it is non-invasive and does not require interrupting the
HVAC system operation. KMC recommends installing the sight glass and the SGM in a vertical position,
with the flow of refrigerant upward through the sight glass. According to KMC, installation in horizontal
positions can cause the SGM to be exposed to bubbles that are not associated with low refrigerant charge.
To address bubbles formed from restrictions in the line, poor evacuation, or clogged filters, KMC
installation procedures specify that operators should maintain clean filters, use manufacturer
recommended operational procedures, and follow industry standard evacuation and charging procedures
prior to use of the SGM. Also, KMC recommends that operators allow the compressor to run at least 10
minutes to allow the system to equilibrate before interpreting the SGM output.
The system used in this verification consisted of a SGM equipped with a voltage-controlled, relay-
actuated timer and annunciator. When the SGM records flash vapor conditions resulting in a greater than
4.0 VDC signal for more than 60 seconds, the annunciator produces a visible and audible alarm. During
each test run, the alarm served to notify the verification testing crew that bubbles occurred at the sight
glass (i.e., a possible leak indication existed).
The SGM can be applied to a variety of installations which were not used or evaluated as part of this
verification. The 0- to 5-V output signal can be wired to the optional KMDigital Controller, which allows
1-5
-------
real-time monitoring, logging, and trend analysis of sight glass conditions. The KMDigital Controller is a
programmable logic controller intended for integration with building automation systems. It allows
inputs from multiple sensors such as temperature probes, thermostats, air velocity, and pressure sensors,
and contains additional input channels for signals produced from the SGM. The KMDigital Controller
can also deliver data directly or via modem to the optional KMDigital Facilities Management System at a
central computer. These options can be set up to page an operator, engage a hardwired relay which
sounds an onsite alarm, etc. The Management System allows a certain amount of customization. For
example, the backside of the sight glasses may have different shades of reflectivity due to age or
overheating during installation. By using engineering adjustments in software, the Management System
can reportedly compensate for the different shades of reflectivity in a sight glass per KMC's instruction
procedures.
1.3 TEST FACILITY DESCRIPTION, MODIFICATION, AND CHECKOUT
The SGM was verified on a commercial-scale rooftop air-conditioning system and a reciprocating chiller.
Both units are owned and operated by North Carolina State University's (NCSU) Centennial Campus in
Raleigh, North Carolina. The test systems are representative of the types and sizes of commercial-scale
systems to which KMC plans to market the device. KMC has indicated that users can install the SGM
and apply the technology to a wide range of sizes and types of equipment. The verification team made
reasonable efforts to identify and select representative commercial scale test systems, particularly in sizes
which fall under existing EPA regulations. Nevertheless, the test results are limited to the types of
systems tested, and may or may not be applicable to other systems (e.g., equipment with centralized
receiver tanks).
The Test Plan also specified SGM verification on a supermarket type refrigeration unit. During testing
this system was found to be equipped with a head-pressure, flooding control valve in the refrigerant
circuit. This valve's location in the circuit prevents bubbles from reaching the sight glass at specific
ambient temperatures. KMC indicated that, in its current stage of development, the SGM is not capable
of functioning with such a control valve. As a result, the SGM verification on this unit did not occur.
Section 2.1 of this report provides more detail regarding the inability to conduct verification testing with
this refrigeration unit.
Figure 1-3 presents photographs of the two systems that were used in the verification and Table 1-1
summarizes their key features. A brief description of each system follows.
1-6
-------
Figure 1-3. Photographs of Test Systems
•i
Commercial Roof-Top Air-Conditioning System
Reciprocating Chiller
Table 1-1. Profiles of Test Systems
Manufacturer
Model
Cooling Capacity (nominal)
Number of Compressors
Size of Compressors
Refrigerant Charge (nominal)
Refrigerant Type
Refrigerant Operating Pressures
(maximum)
High
Low
Nameplate Voltage
Compressor Electrical Data
RLAa (maximum)
LRAb (maximum)
Condenser Electrical Data
Number of Fans
Horsepower
FLAC
LRAb
Commercial-Scale
Rooftop HVAC System
Carrier
50DKB074DAA600FM
75 tons
2 (parallel systems)
10 hp each
Compressor System A: 73.5 Ib
Compressor System B: 64.5 Ib*
R-22
410psig
ISOpsig
460 volts
65.4 amps
345.0 amps
5
1 hp each
13.5 amps
n/a
Reciprocating Chiller
Carrier
30GT-070-500ka
70 tons
2 (parallel systems)
7.5 hp each
Compressor System A: 70 Ib
Compressor System B: 69 Ib*
R-22
450 psig
278 psig
208/230 volts
147.7 amps
690 amps
6
1 hp each
37.8 amps
186.4 amps
* Test compressor where the SGM was installed and verified
a Rated load amps
b Locked rotor amps
0 Full load amps
1-7
-------
Carrier manufactured the rooftop HVAC system selected for testing. This air-to-air exchange unit is a
moderately large (75 tons) commercial unit, providing comfort cooling for the tenants of the Research 4
building of the NCSU Centennial Campus. It is one of four identical systems that meet the cooling loads
of the approximately 38,000 ft2 office building. Records are available listing repairs conducted and
refrigerant additions. Each of the two compressors and its associated condenser, evaporator, and valving
operates independently from the other. The SGM was installed on the "B" compressor, which is rated for
a nominal charge of 64.5 Ib R-22 refrigerant.
The chiller system selected for testing uses a reciprocating compressor, and is also manufactured by
Carrier. This water chiller is a moderately large (70 tons) system. Similar to the rooftop HVAC unit, it
consists of two separate compressors with their associated condensers, evaporators, and valving.
Operational and maintenance records are available for this unit. The reciprocating chiller is specified to
operate at ambient temperatures from 0 to 125 °F. The maximum water temperature entering the cooler is
specified to be 95 °F and the minimum discharge temperature is 40 °F. The SGM was installed on the
"B" compressor, which is rated for a nominal charge of 70 Ib R-22 refrigerant.
All test systems were previously equipped with factory installed sight glasses. The rooftop air-
conditioning unit was factory equipped with a vertically oriented sight glass. No change in glass
orientation was required for that unit, but the glass was quite dirty. A certified HVAC technician replaced
the sight glass and the inline refrigerant filter/dryer. The technician relocated the reciprocating chiller
sight glass from horizontal to vertical position and replaced the inline refrigerant filter/dryer core.
Prior to performance testing of the SGM, each test system was verified to be operating according to the
original equipment manufacturers' specifications. This was done to ensure the refrigeration systems were
operating representatively, and to prevent potential malfunctions in the test systems which would affect
the performance results of the SGM. An independent contractor (Brady Services, Inc.), certified by
Carrier to service both Carrier systems, was retained to assess the systems. The certified HVAC
technician recorded the following operational parameters and compared them with the manufacturer's
specifications:
• Suction and discharge pressures and temperatures
• Liquid line and evaporator temperatures
• Condenser inlet and outlet air temperatures
• Superheat conditions
• Voltages and current draws
• Vacuum leak test results
The assessments occurred during June 2001 under normal operating conditions for both units. The
technician confirmed that both units were operating normally within the Carrier specifications, and with
no malfunctions. GHG Center representatives were present to observe and document these activities.
1.4 OVERVIEW OF VERIFICATION PARAMETERS AND EVALUATION STRATEGIES
The verification test focused on assessing performance parameters of significant interest to potential
future customers of the SGM. The verification addressed the following parameters:
• SGM Installed Cost
• Refrigerant Leak Detection Sensitivity
• Potential Refrigerant Savings and Cost Savings
-------
The following subsections discuss the verification approach, evaluation strategies, and measurement
procedures used to conduct the verification. Detailed descriptions of the field measurement
instrumentation and procedures are available in the Test Plan and are not entirely repeated here.
1.4.1 SGM Cost and Installation Requirements
Installation and capital costs of the SGM were verified for each test system. Installation of an SGM
system includes the following tasks:
1. Installation of a new sight glass (where required per KMC specifications)
2. Installation and configuration of the SGM
The first task requires a certified HVAC technician. An in-house maintenance technician could perform
the second task which includes SGM installation and wiring a simple annunciator as was used during the
verification testing. The GHG Center obtained or computed actual costs associated for this work which
includes the SGM capital cost, parts and supplies for new sight glasses, and equipment costs.
Sight glass installation costs included replacement of the sight glass and filter/drier at the rooftop HVAC
unit and relocation/installation of a new sight glass and drier core at the chiller. The GHG Center
obtained the actual costs billed for this work from Brady Services, Inc.
Capital costs were verified by obtaining price data from KMC for the SGM, relay actuated timer, and
annunciator. KMC also supplied cost data for the optional KMDigital Controller, even though it was not
tested or evaluated in this verification. The SGM and timer/annunciator were temporarily installed near
the test unit. Permanent installation would require an electrician to run conduit and low voltage (24
VAC) wiring from the unit's low voltage control power supply to the SGM and timer/annunciator. A cost
estimate for this activity was obtained from Brady Services, Inc., and is used as the installation cost for
the SGM.
1.4.2 Refrigerant Leak Detection Sensitivity
Refrigerant leak detection sensitivity is defined as the percentage of full charge at which, when leaked or
removed, the SGM will detect low refrigerant levels, sound an alarm, and provide a visual alarm. To
verify this parameter, the GHG Center measured the full charge of each test system, and systematically
drew out incremental quantities of refrigerant until a low charge alarm was indicated. The weight of
refrigerant withdrawn at the point of monitor alarm, divided by the weight of full charge, times 100
represents the leak detection sensitivity of the monitor.
The charge capacity of each system was quantified by fully evacuating the entire system, and charging the
system using the manufacturer's and industry-standard procedures. An HVAC technician, certified as
required by the EPA under the Clean Air Act of 1990, §608, as amended, and Title 40 CFR 82, Subpart F,
performed all refrigerant handling procedures with the proper equipment.
The following discussion presents the approach used to verify this parameter and a brief description of
procedures used. Details regarding refrigerant evacuation and charging on each of the systems can be
found in the Test Plan. Figure 1-4 presents a schematic of the key procedures that were followed.
1-9
-------
Figure 1-4. Refrigerant Leak Detection Sensitivity Testing Procedures
Figure 1-5. Refrigerant Leak Detection Sensitivity Testing Procedures
Evacuate System
1. Screen for leaks and fix any found
2. Remove refrigerant using recovery
system
Charge System
1. Remove air and moisture using
vacuum pump
2. Inject refrigerant
3. Measure cumulative refrigerant added
Verify System is Fully
Charged
1. Verify pressure settings
2. Verify current draws
Repeat Test Run
Withdraw Refrigerant
1. Remove -0.20 % of full charge
2. Measure weight of refrigerant removed
3. Document compressor operating
parameters
Calculate Cumulative
Weight of Refrigerant
Removed
Calculate Refrigerant Loss
Detection Sensitivity
1. Recharge with amount withdrawn
2. Repeat until the 90 % confidence interval for mean
leak detection sensitivity is less than 0.50 %
Step 1. Initial System Refrigerant Evacuation
The first step was to evacuate the refrigeration systems after identifying and fixing potential leaks present
in the system. Screening for existing leaks in piping, fittings, valves, and other accessories was performed
according to industry-accepted methods with a hand-held electronic leak detector. The TIP electronic
leak detector complied with ASHRAE Standard 15-1994, which requires the use of an instrument of this
type where air-conditioning and refrigeration systems are installed. This detector produces an audible
signal (beep), about once/second. The beeps occur faster in the presence of trace amounts of refrigerant.
The detector's threshold is approximately 0.5 oz (0.03 Ib) of refrigerant per year. Once a refrigerant leak
was isolated, NCSU operators fixed the leaks, and verified their repair during normal operation. The leak
detector was used as a screening device only, and did not require field calibration.
The systems were then completely evacuated using an EPA-certified refrigerant recovery system and
manifold with gauges. The recovery unit is a compact, heavy-duty oilless compressor unit equipped with
the appropriate valves and self-sealing quick connects. The recovery unit is connected to a refrigerant
1-10
-------
evacuation cylinder and is capable of moving vapor or liquid from either the high or low-pressure side of
a refrigeration system into the cylinder. Refrigerant was collected in pre-weighed EPA certified
evacuation cylinders (50-lb capacity), and the final weight of the refrigerant-filled cylinders was
measured and recorded.
The technician used a vacuum pump to remove air and moisture present in the refrigeration system after
the recovery unit had removed the refrigerant. The vacuum pump removed moisture by lowering the
pressure within the system and vaporizing the moisture, then exhausting it along with air. Mounted on
the vacuum pump is a Thermal Engineering vacuum micron gauge which allows vacuum measurements
to be made in microns of Hg. The micron is industry standard nomenclature to record absolute pressures
below 29.5 in. Hg. Standard atmospheric pressure is approximately 760,000 microns, or 29.92 in. Hg.
The pump can achieve approximately 100 microns maximum vacuum; the micron gauge can indicate to
approximately 5 microns vacuum.
After removing moisture and air, the technician performed a vacuum leak check according to standard
industry practice. The Test Plan incorrectly stated that the system would be evacuated and held at about
50 microns Hg for this leak check. Standard practice, as outlined in instructions accompanying the
vacuum gauge, is to evacuate the system to 500 to 1,000 microns {Thermal Engineering Company). The
system is then sealed and monitored to verify leak tightness. If the vacuum level remains below 1000
microns for at least 15 minutes, the system is considered to be free of leaks.
Step 2. System Refrigerant Charging
The second step was to recharge each system with refrigerant and determine the full charge capacities.
This was done with an industry standard design gauge manifold system. The manifold system is
universally recognized as the instrument for testing air-conditioning equipment. It is used for checking
operating pressures, adding or removing refrigerant, adding oil, and performing other necessary
operations such as leak testing. Figure 1-5 illustrates a gauge manifold system: the HVAC technician
used a newly purchased TIP manifold system during the verification tests. Figure 1-6 provides a
simplified diagram of the manifold system installed on a refrigeration unit.
Figure 1-5. Refrigeration Gauge Manifold and Hoses
Gauge Manifold
(Source: Imperial Eastman)
Refrigeration Hoses
(Source: Robinaire)
1-11
-------
The technician attaches the manifold system hoses to the refrigeration system at factory-installed service
valves on the suction and discharge sides of the compressor. Opening and closing the needle valves on
the gauge manifold can produce different refrigerant flow patterns and service activities. One indication
of proper system function is that the pressure gauge readings are consistent with manufacturer-specified
values. Normal pressure readings on an air-conditioning R-22 unit range between 65 and 80 psig pressure
on the low side, also called the compound gauge, and 175 to 350 psig on the high side. Actual operating
pressures vary depending on the ambient conditions and the load on the refrigeration system. The gauges
are constructed such that both pressure and temperature readings can be made simultaneously. Each
gauge displays the condensing and evaporating temperatures on their inner rings and pressures on the
outer rings. The gauge manifold valves are also manipulated to evacuate and charge the refrigeration
system.
Figure 1-6. Simplified Diagram of Refrigeration Manifold System
Condenser
Evaporator
f Quick Conn.
Fittings
JTyp.)
Expansion
Valve
To Recovery Unit
or
Vacuum Pump
or
Refrigerant Cylinder
Two definitions of the system full refrigerant charge were addressed during the verification. The Test
Plan defined full charge according to industry standard and manufacturer specifications as the amount of
refrigerant needed to achieve a visually clear sight glass. The following paragraph discusses this
procedure. In addition, the Test Plan described KMC's definition of full charge as the amount of
refrigerant needed to achieve SGM flash signals greater than or equal to 4.0 VDC for no more than 15
seconds in any 5-minute period. This report refers to the KMC definition as the voltage/time charging
method. The verification results are based on the clear sight glass definition because of its widespread
acceptance in the industry and the specifications cited by the test units' manufacturer. It is also used in
this verification report to estimate potential cost savings, because users of this technology are likely to be
interested in net savings relative to current industry practices. Results based on the voltage/time method
are included as additional performance data for readers interested in employing an alternate charging
method.
1-12
-------
Per Carrier's specifications and industry practice, the technician started charging the system while it was
under vacuum. The refrigerant's pressure in the charging cylinder moved the refrigerant into the
evacuated system through the gauge manifold until the system and cylinder pressures equalized. The
technician then started the unit and continued charging refrigerant while observing the sight glass. As full
charge was approached, bubbles would begin to disappear from the sight glass. When the sight glass was
running clear, the technician would cease adding refrigerant and observe the sight glass for 3 to 5
minutes. If any bubbles appeared, he would observe the sight glass for an additional 3 to 5 minutes. If
any bubbles appeared, he would then add a small amount (approximately 0.2 to 0.5 Ib) of refrigerant to
the system and repeat the process. When he observed that the sight glass was clear, he declared the
system to be charged, and GHG Center personnel recorded the weight of refrigerant that had been
supplied to the system. The total charge injected into the system was computed as the difference between
the initial and final weight(s) of the charging cylinder(s).
The charging cylinder rested on a digital scale, manufactured by Digimatex, which was used to measure
the total charge of each test system and refrigerant withdrawal amounts during leak detection sensitivity
testing. The maximum rated capacity of the DI 28 S-SL model is 100 Ib, and the rated accuracy is ± 0.02
percent of reading and 0.005 Ib display error. The manufacturer's precision (repeatability) specification is
± 0.02 Ib. Its platform size is 13 x 17 x 3 in. (length/width/height), large enough to allow the 50- or 30-lb
capacity refrigerant cylinders to remain in an upright position. It is battery powered and meets or exceeds
Class III and OIML standards. The scale was calibrated with NIST-traceable standard masses less than 1
month prior to the test campaign. GHG Center personnel verified the scale's accuracy immediately
before and precision immediately after each test run with NIST-traceable standard masses.
After the system was recharged with refrigerant and was operating normally, the technician and GHG
Center test personnel conducted a survey of the gauge manifold, compressor area, condenser area, and
evaporator area with the hand-held electronic leak detector. This procedure ensured that the system,
gauge manifold, hoses, and test cylinder had no leaks during the test runs.
After achieving the clear sight glass full charge, test withdrawals then commenced, and KMC personnel
notified the Field Team Leader when KMC's voltage/time full charge requirement was achieved. For
every test run, the amount of refrigerant required to achieve a visually clear sight glass (system
manufacturer definition) was larger than the amount needed to achieve KMC's voltage/time definition of
full charge. GHG Center personnel recorded the weight of refrigerant that had been withdrawn to
determine the voltage/time full charge. The net refrigerant remaining in the system was the full charge as
determined by KMC's voltage/time method.
Step 3. Leak Detection Sensitivity Testing
After charging the system, verification of normal system operation was conducted before SGM
verification testing was initiated. This process ensured that the refrigeration system operating conditions
were consistent between successive leak detection sensitivity test runs, and that the system was fully
charged per the manufacturer's recommendations. The current draw of the compressor was measured and
recorded, and verified to be operating within manufacturer-specified levels (Table 1-1). High and low
manifold pressure gauge readings were monitored and recorded to ensure they did not exceed the
manufacturer's specified maximum levels and were within the values expected for the ambient, liquid
line, and suction line temperatures observed during the tests. Each test unit's user manual was reviewed
to ensure that the measurements were representative and within the expected values for the conditions
encountered during testing.
Other parameters monitored during the leak detection sensitivity tests were ambient temperature and
relative humidity, and refrigerant suction and liquid line temperatures. Operational parameters were
1-13
-------
recorded at the beginning, during, and at the end of each test run. Outdoor temperature and humidity
were not critical measurements, but were collected for possible post-test trend analysis and to ensure that
the test units were operating representatively. The instrument used was an integrated
temperature/humidity unit (Vaisala Model HMP 35A) located in close proximity to the air intake of the
condenser. This unit uses a platinum 100-ohm, 1/3 DIN RTD (resistance temperature detector) for
temperature measurement. As the temperature changes, the resistance of the RTD changes. The
integrated unit uses a thin film capacitive sensor for humidity measurement. The dielectric polymer
capacitive element varies in capacitance as the relative humidity (RH) varies, and this change in
capacitance is detected. The response time of the temperature and humidity sensors is 0.25 seconds. Its
rated accuracy is ± 2 °F for temperature and ± 3 percent for RH. It was wired to a Campbell data logger
that was downloaded daily to a laptop computer.
Two type K thermocouples were mounted on the refrigerant suction and liquid lines. The technician
inserted the thermocouple probes under the insulation so the sensing element contacted the copper
refrigerant line and then taped the probes in place. Test operators plugged the thermocouples into a Fluke
two-channel digital thermocouple meter and recorded the resulting temperature readings on the test field
data forms. The thermocouples and meter had been calibrated at the GHG Center prior to the start of
testing.
After each system was verified to be charged and operating according to manufacturer's
recommendations, leak detection sensitivity testing was initiated. To perform these tests it was necessary
to have the compressors running continuously while the refrigerant withdrawals occurred. This was
accomplished by physically disabling the automatic thermostatic controller which determines whether the
compressor turns on or off, and overriding this with manual control. NCSU operators then manually
controlled the compressor system as testing proceeded.
Leak detection sensitivity tests were conducted by withdrawing refrigerant from the system in small
increments until the sensor alarmed. A pre-weighed, small test cylinder (30-lb capacity) resting on the
digital scale was used to collect and quantify the refrigerant withdrawn. The technician connected the
liquid port of the test cylinder to the center hose on the gauge manifold. The needle valves on the gauge
manifold allowed precise control of small vapor withdrawals (0.10 to 0.30 Ib increments) from the high-
pressure side of the system under test.
The position of the manifold hose connected to the test cylinder (the yellow hose) and the refrigerant
contained in that hose could affect the refrigerant weights determined by the scale. Test operators
verified that the manifold hose was undisturbed during each test run. The certified technician ensured
that vapor, not liquid, at system suction pressure remained in all hoses at the end of each withdrawal.
Under these conditions, the maximum weight of R-22 that could have been contained in the yellow hose
(70 inches long x 11/32 inch inside diameter) was 0.004 Ib. This is well below the scale's 0.01 Ib display
resolution and is therefore negligible.
The technician withdrew refrigerant at target increments of about 0.20 percent of the full charge into the
pre-weighed test cylinder. The weight of the test cylinder containing the refrigerant was measured and
recorded at the end of each withdrawal, using the digital scale. The refrigeration system was allowed to
operate for 5 minutes so bubbles generated from removal of the refrigerant were given sufficient time to
reach the sight glass area.
The GHG Center personnel waited to determine if an audible alarm occurred. The timer/annunciator was
set to produce an audible and visible alarm in response to flash vapor conditions that produced a greater
than 4.0 VDC SGM signal for more than 60 seconds. KMC specifies this voltage and duration as an
unequivocal signal that the sight glass is filled with flash vapor. When an alarm occurred, the GHG
1-14
-------
Center stopped the run, determined the total weight of refrigerant withdrawn, and computed leak
detection sensitivity. If the SGM did not alarm to indicate a low charge, another withdrawal (equivalent
to the target weight) was made. At the conclusion of each withdrawal, the weight of the test cylinder was
measured, and the system was allowed to stabilize for another 5 minutes. During each test, the
withdrawal process was repeated until the alarm level was reached. Leak detection sensitivity was
computed using Equation 1.
T , .. . refrigerant lost at the point of monitor alarm (Ib)
Leak Detection Sensitivit y (%) = --^- x 100 (Eqn. 1)
full charge of system (Ib)
At the conclusion of each test, the refrigerant was injected back into the system using the equipment and
procedures described earlier. The system full charge amount was again determined and recorded. Total
full charge for the next run was recorded as the full charge of the previous run, minus refrigerant
withdrawn during the previous run's leak detection sensitivity test, plus the refrigerant added to again
achieve a clear sight glass.
The Test Plan specified that individual test results must fall within a range of values (confidence interval)
around the mean of all test results. Confidence intervals include an estimate of the proportion of test
results expected to fall within the given interval (e.g., "90 percent of the individual results are within 0.30
times the mean test result"). The confidence interval depends on the sample standard deviation divided
by the square root of the number of samples. For a given standard deviation, a larger number of test
results generally tends to reduce the size of the confidence interval. For a data set with a large standard
deviation, however, even a large number of tests cannot reduce the size of the confidence interval below
certain limits. The Test Plan stated that it was reasonable to expect 90 percent of the observed test results
to fall within 0.30 times the mean leak detection sensitivity. This range, or confidence interval, was used
to determine the number of tests to conduct on each unit. Test runs were repeated until the confidence
interval was less than 0.30 times the mean leak detection sensitivity, or until a maximum of five valid
runs had been completed.
During each test, procedures were followed that allowed a minimum of five refrigerant withdrawals
before the SGM reached alarm level. The Test Plan specified a target withdrawal rate of 0.2 percent of
full charge during each withdrawal, and this value was used as a starting point. As testing progressed,
Center personnel were able to fine tune the withdrawal process by increasing the volume of the first few
withdrawals and, after nearing the mean alarm point, adjusting the remaining withdrawals to
approximately 0.10 to 0.20 pounds each.
The Test Plan required that, for at least one test run at each unit, the wait period between refrigerant
withdrawals was to increase to 30 minutes as the alarm point was approached. This was to allow the unit
to reach equilibrium and to ensure that the SGM produced a stable alarm condition. Early in the test
campaign, the HVAC technician, KMC, and GHG Center personnel concluded that this wait period was
excessive. Unit operating parameters and voltage signals from the SGM stabilized in less than 2 minutes
after each withdrawal; the consensus was that 5- to 8-minute wait periods between each pair of
withdrawals were sufficient.
1.4.3 Estimated Potential Refrigerant Savings and Potential Cost Savings
Operators of refrigeration equipment rely on different inputs to warn of excessive refrigerant loss and,
thus, the presence of refrigerant leaks. In extreme cases, catastrophic equipment failure or product loss is
the first indication that refrigeration systems require maintenance. More commonly, operators rely on
1-15
-------
maintenance records and/or regularly scheduled inspections to indicate when systems require more
refrigerant, and when excessive charge loss is occurring. To support wise purchase decisions, operators
of refrigeration systems will likely want to know if the SGM can provide potential financial or other
benefits compared to currently used methods for detecting refrigerant loss. Both the operator and
environment would benefit if the SGM can help operators reduce the amount of refrigerant leaking into
the atmosphere. This could occur if the SGM warns of losses more rapidly than currently used detection
methods and if system operators respond to SGM alarm conditions (i.e., immediately perform needed
repairs).
To assess this, the GHG Center estimated potential refrigerant savings associated with the use of the
SGM. This was accomplished by comparing the minimum refrigerant loss detectable by the SGM
(determined as described above) with refrigerant losses and outcomes occurring under routine inspection
programs. To avoid the cost and time required to determine the sensitivity of routine inspection
programs, the GHG Center used historical data maintained by NCSU. The GHG Center obtained
operational records for the two test systems and other similar equipment installed at the NCSU campus.
System operators maintain these records as a normal industry practice and they serve as a basis for
potential refrigerant savings calculations.
The NCSU facilities operators have maintained records since the responsibility for the systems was
transferred to them on December 1, 1997. They include:
• Inspection or service/repair date
• Unit ID
• Refrigerant type and fill amounts
For each service record analyzed, the amount of refrigerant replaced since the last full charge (if any) was
noted. Analysts reviewed the records for indications of catastrophic losses. Refrigerant recharge after a
catasrophic loss would be equal to the unit's capacity; none were noted in the data.
If a properly installed SGM and alarm/annunciator had existed at a given unit, operators would have been
alerted to a leak after a certain amount of refrigerant had been lost. This amount depends on the SGM
leak detection sensitivity and the unit's refrigerant capacity. For example, if full refrigerant charge at a
reciprocating chiller was 46.68 Ib of R-22, a 3.56 percent leak detection sensitivity implies that a
threshold quantity of 1.66 Ib of refrigerant must be lost before an alarm occurs. This approach assumes
that the amount of charge lost per unit time (i.e., the leak rate) is constant.
For a given service event, analysts subtracted the threshold quantity from the amount of refrigerant added
to the unit. For example, the reciprocating chiller test unit maintenance log recorded 27 Ib refrigerant
added on September 2, 1998. After subtracting the threshold quantity from the recorded loss, a 25.3 Ib
savings in refrigerant would have been realized if the leak had been repaired immediately after the SGM
alarmed. The price for this refrigerant is currently approximately $3.75/lb. Using this price, a 25.3 Ib
savings in refrigerant would equate to a $95 cost savings for the service event. The potential annual
savings is the sum of the potential savings from each service event divided by the number of years of
available data for that unit.
1-16
-------
2.0 VERIFICATION RESULTS
2.1 OVERVIEW
Verification of the SGM was conducted at NCSU on July 25 through 28, 2001. KMC supplied the SGM
with a voltage-controlled, relay-actuated timer/annunciator. KMC personnel installed it on the two test
units immediately before testing commenced.
The Center had also planned to test the SGM on a third unit: a supermarket-type refrigeration system.
The unit's manufacturer, however, (Larkin) noted that this unit was equipped with a head pressure control
flooding valve. This valve reroutes the refrigerant at moderate ambient temperatures (less than
approximately 78 °F) and prevents bubbles from reaching the sight glass until virtually all the refrigerant
is removed from the unit. An initial test run confirmed this. The SGM technology as it stands is not
designed to operate when the valve is actuated, preventing bubbles from reaching the sight glass. Based
on this consideration, KMC and the GHG Center concluded not to test the supermarket unit as
configured.
Results for the primary verification parameters are discussed in the following subsections:
Section 2.2 - SGM Installed Costs
Section 2.3 - Refrigerant Leak Detection Sensitivity
Section 2.4 - Potential Refrigerant Savings and Cost Savings
2.2 SGM COST AND INSTALLATION REQUIREMENTS
Table 2-1 presents a summary of as-tested SGM capital and installation costs. KMC provided capital
costs for the SGM and the timer/annunciator, as used during the verification. The total capital cost for the
entire system is $360.00.
KMC installed the SGM and timer/annunciator on the test units temporarily to allow easy relocation.
Potential purchasers of the device can elect to install the timer/annunciator at the SGM (i.e., near the
refrigeration system compressor) or at a remote location. In both cases, a permanent installation would
require low voltage wiring for the SGM and timer/annunciator, conduit, connection with the unit's control
circuit power supply, and the associated labor charges. For a permanent timer/annunciator installation at
the SGM location, Brady Services, Inc. estimated the cost as approximately $170.00. Total installed cost
for the SGM and the timer/annuncitator is $530.00 for both units.
2-1
-------
Table 2-1. SGM Costs
SGM Capital Costs
Description
SGM
Voltage-controlled relay and
timer/annunciator cost estimate
SGM Subtotal
Rooftop HVAC
$210.00
$150.00B
$360.00
Chiller
$210.00 a
150.00 a'b
$360.00
SGM Permanent Installation Cost Estimate
Parts and labor: wire, conduit, connection
to existing low-voltage power supply in
unit
Total SGM Installed Cost
(with existing sight glass)
$170.00
$530.00
$170.00
$530.00
Capital and Installation Costs for Modifying Existing Sight Glass
Sight glass
Misc. piping and supplies
Inline filter/dryer
Welding supplies
Refrigerant recovery equip, charge
Service call
Certified HVAC contract labor
Sight Glass Subtotal
Total SGM Installed Cost
(with new sight glass)
$18.00
$25.00
$26.00
$12.50
$12.50
$12.50
$132.00
$238.50
$768.50
$18.00
$25.00
$20.00
$12.50
$12.50
$12.50
$132.00
$232.50
$762.50
a One SGM and timer/annunciator was used for all verification tests. KMC relocated the equipment to the
chiller unit immediately prior to the start of testing at that unit
b KMC's cost for the as-tested shop-built timer/annunciator was $225.00. In normal production, KMC
estimates this device would cost $150.00.
As stated in Section 1, the sight glass and inline drier on the rooftop HVAC unit were replaced. For the
chiller, the sight glass was relocated to a vertical position and the drier core was replaced. This work
brought the sight glasses into conformance with KMC specifications on the test units. These
modifications may or may not be required at other installations, depending on the condition and
orientation of existing sight glasses. As shown in Table 2-1, the total installed costs for an SGM with a
new sight glass increase by $238.50 for the HVAC unit and $232.50 for the chiller.
KMC also supplied the price for the KMDigital controller, even though it was not involved with the
verification tests. The price for the optional KMDigital controller is $160.00.
2.3
REFRIGERANT LEAK DETECTION SENSITIVITY
2.3.1 System Refrigerant Evacuation and Leak Checking
Refrigerant leak detection checking was conducted on the rooftop air-conditioning system and the
reciprocating chiller after installation of the SGM. Before initiating leak detection checks, both systems
2-2
-------
were fully evacuated of refrigerant and checked for leaks using the procedures specified in the Test Plan
and Section 1.4.2 of this report.
After removing as much of the refrigerant as was possible from the rooftop system with the recovery unit,
the technician evacuated the system to 505 microns vacuum with a high vacuum pump. After a 15-
minute hold time, the vacuum gauge indicated 490 microns. These results were satisfactory and indicated
a leak-tight system. The slight increase in vacuum was normal and due to pressure equalization
throughout the system {Thermal Engineering Company).
After the system was recharged with refrigerant and was operating normally, the technician and GHG
Center test personnel conducted a survey of the gauge manifold, compressor area, condenser area, and
evaporator area with the electronic leak detector. No leaks were found.
At the reciprocating chiller, GHG personnel had surveyed the evaporator and condenser areas with the
electronic leak detector while the system was operating normally. Then, after shutting the system down
and removing as much of the refrigerant as was possible with the recovery unit, the technician installed a
high vacuum pump onto the system and let the pump operate overnight. In the morning, vacuum level
was at about 800 microns. When the technician shut off the vacuum pump, the gauge indication
immediately began to rise towards 1,000 microns. The technician suspected that the indicated leak may
be in the vacuum pump, and not in the chiller. He closed the gauge manifold valves, thus isolating the
manifold gauges and the system from the vacuum pump and its micron gauge. Over the next 60 minutes,
the vacuum pump and micron gauge leaked back to atmospheric pressure while the reciprocating chiller
unit remained at 30 in. Hg vacuum (as indicated by the manifold gauges). This standard practice
indicated that the leak was in the vacuum pump and/or its micron gauge, not the test unit.
The technician recharged the system, brought the reciprocating chiller online under normal operating
conditions, and surveyed the compressor area, gauge manifold, and test cylinder areas with the electronic
leak detector. Three small leaks were found at gauge pressure ports and were corrected by tightening the
fittings approximately one-quarter turn.
2.3.2 SGM Leak Detection Sensitivity
System Refrigerant Charging
As discussed in Section 1.4.2, two definitions of system full refrigerant charge were addressed during the
verification. The Test Plan defined full charge according to manufacturer (Carrier) specifications as the
amount of refrigerant needed to achieve a visually clear sight glass. In addition, the Test Plan includes
KMC's voltage/time definition of system full refrigerant charge. The field testing revealed that these two
conditions did not result in the same full charge determination for these systems. More refrigerant was
needed to achieve a visually clear sight glass (system manufacturer definition) than the amount needed to
achieve the KMC definition of full charge.
Since system full charge is the denominator in the equation used to calculate leak detection sensitivity
(Equation 1), the verified full charge value for each of the systems has a direct impact on the verification
results. To address this, the Center calculated leak detection sensitivity using full charge capacities
determined using both definitions. The full charge determinations based on the manufacturer's clear sight
glass definition were used to calculate leak detection sensitivity as discussed in Section 1.4.2. These data
form the basis to estimate potential annual refrigerant and cost savings because potential users of this
technology are likely to be interested in potential savings based on the manufacturer's standard full
charge methodology.
2-3
-------
Test withdrawals commenced after full charge was achieved according to the clear sight glass
specification. KMC personnel notified the Field Team Leader when the KMC definition of full charge
was achieved. The full charge determinations using the KMC definition are presented here to provide
potential SGM users with data on the differences in full charge capacity that may occur if they choose to
employ the KMC procedures. Tables 2-2a and 2-2b present the verified full refrigerant charge amounts
for both systems tested using the two different definitions of full-refrigerant charge.
Table 2-2a. Rooftop HVAC Full Charge Determinations
Test Number
1
2
3
4
5
Clear Sight Glass
Method (Ib)
51.74
49.72
52.35
52.35
52.37
KMC Voltage/
Time Method (Ib)
n/aa
n/aa
51.21
50.78
50.76
a Data for these two runs were not recorded because the full charge determination strategy
was under discussion and had not vet been approved bv KMC and the GHG Center.
Table 2-2b. Reciprocating Chiller Full Charge Determinations
Test Number
1
2
3
4
5
Clear Sight Glass
Method (Ib)
47.10
47.12
46.25
46.25
46.68
KMC Voltage/
Time Method (Ib)
45.40
45.20
45.25
45.27
45.20
Test Conditions
During each test run, system operational conditions were documented including current draw, compressor
suction and discharge pressures, and refrigerant liquid and suction temperatures. These data were
recorded at the beginning and end of each test conducted. Ambient temperature and humidity were
recorded at 1-minute intervals during the test periods. System operational data and average ambient
conditions during each test run are summarized in Tables 2-3a and 2-3b.
These data are presented to document that system operations were representative of normal operations
throughout the test periods and that conclusions presented in this report are based on withdrawal of
refrigerant rather than system operational changes or upsets.
2-4
-------
Table 2-3a. Test Unit Operating Conditions- Rooftop HVAC Unit
Run
1
2
3
4
5
Average Current
Draw (amperes)
Start
62.93
60.07
57.27
58.13
58.67
End
61.30
58.90
57.67
57.93
59.03
Gauge Pressure (psig)
High Side
Start
215.0
205.0
180.0
180.0
185.0
End
205.0
199.0
180.0
180.0
190.0
Low Side
Start
54.0
50.0
41.0
43.5
45.0
End
50.0
48.0
41.0
42.5
44.0
Refrigerant Temperature (°F)
Liquid Line
Start
90.8
91.8
73.0
74.6
74.8
End
90.4
88.6
72.4
73.8
76.2
Suction Line
Start
80.6
80.8
56.8
59.4
60.6
End
80.0
80.2
58.1
62.4
62.8
Ambient
Temp
(°F)
89.9
86.9
70.1
70.5
72.8
Relative
Humidity
(%)
55.2
60.9
89.9
87.9
78.9
Table 2-3b. Test Unit Operating Conditions - Reciprocating Chiller
Run
1
2
3
4
5
Average Current
Draw (amperes)
Start
125.50
124.33
122.63
124.07
130.07
End
122.63
123.50
122.33
122.93
122.77
Gauge Pressure (psig)
High Side
Start
223.0
225.0
228.0
224.5
224.5
End
223.0
220.0
224.0
220.5
218.0
Low Side
Start
55.0
54.5
55.0
54.0
54.0
End
55.0
52.5
53.5
54.0
54.0
Refrigerant Temperature (°F)
Liquid Line
Start
99.4
97.8
101.4
100.2
98.6
End
99.4
100.0
100.6
98.6
98.6
Suction Line
Start
53.2
51.4
50.8
51.0
50.4
End
52.8
51.2
51.0
50.6
50.8
Ambient
Temp
(°F)
83.1
83.0
83.0
84.7
82.5
Relative
Humidity
(%)
45.8
40.1
43.5
42.9
46.1
Review of the field data and manufacturer's specifications showed that the systems were operating
normally at the conditions encountered during testing. For example, Figures 58 and 59 in Carrier's
"Installation, Start-up and Service Instructions" manual (No. 564-818) present the typical range of suction
and discharge pressures as a function of ambient temperatures for the rooftop unit (Carrier 1996). The
manifold gauge readings shown in Table 2-3a are consistent with these data. For example, higher
pressures are expected while operating during warmer temperatures. The manuals for the reciprocating
chiller do not contain these types of guidelines for pressures. Based on general operational parameters
outlined in the book Refrigeration and Air-Conditioning (ARI 1987), the pressures observed are
indicative of normal unit operation.
NCSU's technician induced a suitable building load for the rooftop unit by commanding the heating
system to operate at the same time as the air-conditioning system. For the reciprocating chiller, existing
building loads were sufficient to allow the reciprocating chiller to operate continuously during all tests.
Average current draws were 90.5 and 84.0 percent of the rated full load specification for the rooftop
HVAC and reciprocating chiller units, respectively (Table 1-1). This indicates that both units were
operating in a representative manner during all test runs.
Leak Detection Sensitivity Results
The leak detection sensitivity tests were conducted after full refrigerant charges were achieved and the
systems were verified to be operating normally and properly. In general, precise portions of refrigerant
(0.10 to 0.30 Ib increments) were removed from the systems until the SGM alarm level was reached. The
2-5
-------
certified HVAC technician controlled refrigerant withdrawals with the needle valves on the gauge
manifold, and the amount of refrigerant withdrawn during each step was measured using the calibrated
scale and recorded.
Results of the leak detection sensitivity tests are presented in Tables 2-4a and 2-4b. Table 2-4a presents
test results for both systems with full charge procedures conducted in accordance with manufacturer
specifications (clear sight glass procedure). These data are used to estimate potential refrigerant and cost
savings in the following section. The results presented in Table 2-4b represent leak detection sensitivities
on both units as determined using the KMC voltage/time full charge procedures.
Table 2-4a. Leak Detection Sensitivity at Manufacturer Specified Full Refrigerant Charge
(clear sight glass)
Rooftop HVAC Unit
Run
1
2
o
J
4
5
Total Refrigerant
Withdrawn (Ib)
3.43
1.86
2.42
2.78
2.70
Leak Detection
Sensitivity (%)
6.63
3.74
4.62
5.31
5.16
Average and 90%
Confidence Interval
5.09 ± 1.01 %
Reciprocating Chiller
1
2
o
J
4
5
2.13
2.10
1.27
1.17
1.65
4.52
4.46
2.75
2.53
3.53
3. 56 ±0.88%
Table 2-4b. Leak Detection Sensitivity with KMC Specified Full Refrigerant Charge
(voltage/time method)
Packaged Rooftop HVAC Unit
Run
3
4
5
Total Refrigerant
Withdrawn (Ib)
1.86
1.21
1.09
KMC Adjusted
Leak Detection
Sensitivity (%)
3.63
2.38
2.15
Average and 90%
Confidence Interval
2.72 ± 1.34%
Reciprocating Chiller
1
2
3
4
5
0.43
0.18
0.27
0.19
0.17
0.95
0.40
0.60
0.42
0.38
0.55 ± 0.23 %
Following manufacturer specifications for system charging procedures, the average leak detection
sensitivity performance of the SGM on the rooftop HVAC and reciprocating chiller systems were 5.09
2-6
-------
and 3.56 percent of full charge, respectively. This corresponds to average refrigerant losses of 2.64 and
1.66 Ib on the two systems.
It is instructive to relate these results to the EPA regulations which apply to commercial refrigeration
units. The current Rules (for system refrigerant capacities greater than 50 Ib), limit annual leaks to 35
percent of the unit's capacity or approximately 18.10 Ib per year for the rooftop HVAC unit. EPA has
proposed a Rule at 63 FR 32044 (June 11, 1998) which would reduce the allowable annual leaks to 10
percent of the unit's capacity, or approximately 5.17 Ib for this unit. Based on these verification results,
the SGM could be an important tool to assist facilities in complying with either Rule. This is because
system operators could respond to and repair refrigerant leaks well before approaching the regulated leak
amounts.
The leak detection sensitivities quoted here apply only to these two 70 to 75 ton (nominal) capacity
reciprocating units using R-22 refrigerant and tested under the ambient conditions found during the test
campaign. Extrapolation of the verification results to other units with different compressor designs,
capacities, refrigerants, and ambient conditions may not be valid.
2.4 ESTIMATED POTENTIAL REFRIGERANT AND COST SAVINGS
The SGM leak detection sensitivity results obtained during the field verification testing were used to
estimate potential refrigerant savings (reduction of refrigerant losses through leaks). This analysis was
conducted by determining refrigerant losses via current industry operating and maintenance practices.
This consisted of obtaining historical system maintenance data from the test units and other similar
systems at NCSU. The measured leak detection sensitivities were applied to refrigerant losses reported in
the maintenance logs to determine the savings that could have occurred with use of the SGM. The
analysis includes a total of 13 service event reports from the systems tested (11 for the reciprocating
chiller and 2 for the rooftop HVAC system) and 16 entries from HVAC and reciprocating chiller units
similar to those tested.
This analysis was conducted using the average leak detection sensitivities as determined using the
manufacturer's definition of system full charge and summarized in Table 2-4a (i.e., 5.09 percent for the
HVAC system and 3.56 percent for the reciprocating chiller). Additional savings could be realized if an
operator chose to define system full charge using the KMC voltage/time procedure. As outlined in
section 1.4.3, a threshold refrigerant loss based on the leak detection sensitivity and the unit's full charge
capacity was subtracted from each service event to yield the potential refrigerant (and cost) savings. The
sum of the potential savings divided by the years of record is the potential annual savings. Responsibility
for the equipment analyzed was transferred to the NCSU maintenance contractor on December 1, 1997.
This is taken as the starting date for the analysis, and represents the initial point at which each system
contained full refrigerant charge before the analyzed service events.
All units analyzed use R-22 refrigerant. Many units have multiple compressors and refrigerant circuits. It
was often impossible to apportion a logged refrigerant addition to a specific circuit: the log entries
mention only the unit being serviced. In these cases, the analysis treats each log entry as a separate
service event. This approach is conservative, because it applies the threshold loss to each service event
and will tend to under-predict the potential savings. Another important assumption in this analysis was
that system leaks would be repaired immediately after the threshold was reached and the SGM alarmed.
Delays in responding to SGM alarms will reduce potential refrigerant and cost savings realized by an
operator.
Tables 2-5a and 2-5b summarize the potential savings in refrigerant and costs for each of the units
examined. The tables summarize analyses of maintenance records from the two test units, as well as
2-7
-------
records from four other rooftop packaged HVAC units manufactured by McQuay, and a York
reciprocating chiller. For simplicity, the tables quote the units' nominal refrigerant capacities as the clear
sight glass full charge. Note that accurate full charge data for the McQuay and York units can only be
obtained via the clear sight glass full charge determination procedures described in the Test Plan.
Table 2-5a. Rooftop HVAC Unit Potential Annual Refrigerant Savings
Unit Description
*Camer Rooftop HVAC
McQuay Packaged Unit
McQuay Packaged Unit
McQuay Packaged Unit
McQuay Packaged Unit
R-22 Full
Charge (Ib)
51.71
56
88
56
56
Total
Number
of Service
Events
2
o
J
o
J
5
2
Years
of
Record
1.5
3.5
1.6
3.6
1.6
Total
Potential
R-22
Savings
Ob)
11.7
30.8
0
9.0
7.1
Verified SGM Leak Detection
Sensitivity (%)
5.09
Potential Annual Savings
Ibof
Refrigerant
7.8
8.8
0
2.5
4.4
Dollars
$29
$o o
JJ
$0
$9
$17
* Unit tested during this Verification
Table 2-5b. Reciprocating Chiller Potential Annual Refrigerant Savings
Unit Description
*Camer Chiller
York Chiller
R-22 Full
Charge (Ib)
46.68
70
Total
Number
of Service
Events
11
3
Years
of
Record
1.5
1.5
Total
Potential
R-22
Savings
Ob)
205.9
74
Verified SGM Leak Detection
Sensitivity (%)
3.56
Potential Annual Savings
Ibof
Refrigerant
137.3
49.3
Dollars
$515
$185
* Unit tested during this Verification. This unit shows a high number of refrigerant losses per year as compared to the York chiller
and the rooftoD HVAC units, which mav indicate svstem Droblems.
These estimates demonstrate that savings in refrigerant needed to maintain system full charge and
associated costs can vary greatly depending on the condition of the system and the number of
maintenance activities needed to maintain proper operation. The highest annual usage (or leakage) for the
HVAC systems examined was 8.8 Ib per year. Conversely, the reciprocating chiller used for this
verification required over 200 Ib of refrigerant to be added over the 1.5 years of record. Systems with a
history of leaks or other operational problems could realize substantial savings through installation of an
SGM, provided alarm responses are timely. The chiller test unit, however, is an example of a system at
which operators may not realize the full potential savings. It could be difficult for a service organization
to respond to recurring alarms (11 over 18 months) in a timely manner. For this unit, the potential
savings in Table 2-5b could be overestimated.
Proper use of and response to an SGM may provide cost savings by improving system operation and
efficiency. It is likely that a fully charged HVAC or reciprocating chiller system will operate more
-------
efficiently than an undercharged or overcharged system, although cost savings of this nature were not
analyzed during this study. It is also possible that certain scheduled refrigerant maintenance activities
could be eliminated by SGM installation. These include routine sight glass observations, electronic leak
detection surveys, and gauge manifold installations for checking refrigerant full charges. Some facilities,
however, include these maintenance checks in quarterly system inspections (i.e., technicians perform
other system diagnosis and preventive maintenance at the same time). NCSU follows this practice, and
may not realize savings from the avoided labor for sight glass observations, etc.
The GHG Center recognizes that several factors may contribute to uncertainties in the historical data and
thus, in this evaluation. Examples of confounding factors in the historical data include: (1) refrigerant
service provider rounding-off the amount of refrigerant added, (2) pressure gauge or other instruments
used to monitor charge loss and amount added could have malfunctioned, (3) gauge manifold and other
charging equipment were not completely screened for leaks, (4) data transcription errors occurred, and/or
(5) the technician's weighing scale was not calibrated.
The age of a particular unit will affect the available data. For example, a brand new unit may not have
any leaks for a long time during which an SGM could alarm. As it ages and begins to leak, however, an
SGM would then begin to track and alarm the leaks. It is also possible that system operators would not
respond to SGM alarms in a timely fashion, thereby not realizing the full potential savings, or that they
would simply recharge a system without performing repairs. Finally, the historical data contain a limited
population of systems, all managed by one operator. It is beyond the scope of this verification to quantify
all of the uncertainties associated with each factor. Thus, the potential savings reported here represent
the maximum potential savings for the systems operated by this facility.
-------
-------
3.0 DATA QUALITY ASSESSMENT
3.1
DATA QUALITY OBJECTIVES
In verifications conducted by the GHG Center and EPA-ORD, measurement methodologies and
instruments are selected to ensure that a desired level of data quality occurs in the final results. The
primary verification parameter for this verification was leak detection sensitivity. Other verification
parameters (installation costs and requirements and potential refrigerant savings) did not require physical
measurements or instrumentation. Therefore, the Test Plan specified a DQO for leak detection sensitivity
only.
Leak detection sensitivity was measured by weighing incremental refrigeration losses and total unit
charges. Therefore, weight measurement errors would significantly affect the quality of the data used to
determine this verification parameter. The test plan presented the chain of calculations performed to
assess the effects of scale accuracy on leak detection sensitivity determinations. The calculations show
that, for the units tested during this verification, if the assumed weighing scale errors occur during a test
that yields a 1.00 percent leak detection sensitivity, actual leak detection sensitivity could range between
0.980 and 1.02 percent. This is a 0.02 percent deviation from the true value of 1.00 percent, and
represents a 2.00 percent error in the determination of the leak detection sensitivity. This error was the
basis for the leak sensitivity detection DQO that was specified at ±2 percent in the Test Plan
To determine if the DQO was met, data quality indicator goals (DQIs) were established for key
measurements performed during testing. In this case, the primary DQI was the accuracy of the scale used
to measure refrigerant charging and withdrawal weights. These goals, summarized in Table 3-1,
identified accuracy and precision DQIs for the scale that must be met to achieve the overall DQO. The
following section discusses the use of field calibration results to reconcile the DQO.
Table 3-1. Measurement Instrument Specifications and Data Quality Indicator Goals
Measurement
Variable
Full charge and
refrigerant
withdrawal
measurements
Instrument Type
/ Manufacturer
Digi Model DI-
28, S-SL Bench
Instrument
Range
0 to 100 Ib
Instrument Specification
Accuracy
Precision
± 0.02 % of
reading and
± 0.005 Ib
display error
± 0.02 Ib
How Verified /
Determined
Factory calibration
Pre-test field
calibrations3 - before
each run
Post-test field
calibrations'3 - replicate
weighings after each test
run
a Scale readings were compared with the following NIST-traceable standard masses: 5, 10, 15, 20, 25, 30, 50, 75, and 100 Ib
(nominal)
b Scale readings were compared with four NIST-traceable standard masses that represented weights measured during test run
3-1
-------
3.1.1 Leak Detection Sensitivity DQO Reconciliation
Two newly purchased scales were used during the verification test. The first scale malfunctioned after
the second test run at the rooftop HVAC unit. The GHG Center field team obtained a second scale for the
balance of the test runs. The distributor provided NIST-traceable calibration certificates for both scales.
These certificates are maintained at the GHG Center, and certify that both scales initially met the
accuracy and precision criteria listed in Table 3-1.
In addition to factory certification, GHG Center personnel conducted accuracy determinations before and
precision determinations after each test run using the following NIST-traceable standard masses:
Nominal Weight Ib NIST-Certified Weight Ib
5 5.000117
10 10.00025
15 15.00037
20 20.00096
25 25.00108
30 30.00145
50 50.00206
75 75.00314
100 100.0045
Prior to each test run, the test operator challenged the scale with each weight and recorded the display
reading on field data log forms. Precision was verified in the field, at the end of each test run, by
performing replicate weighings using four of the NIST-traceable standard calibration weights that were
representative of the actual weights observed during that test run. Each of these four weighings was
repeated twice at the end of each run to confirm that precision was within 0.02 Ib. Tables 3-2a and 3-2b
present the pre- and post- test field verification results. The maximum deviation in precision measured
was 0.02 Ib, so the scale met the GHG Center's 0.02 Ib precision goal.
To satisfy the accuracy goal, the Test Plan specified the scale reading be within ± 0.02 percent of standard
mass plus 0.005 Ib display error. For a 20 Ib standard mass, the calibration must result in a reading that
ranged between 19.991 and 20.009 or ± 0.009 Ib (20*0.02 % + 0.005). The digital display on the scale
was such that measurements are shown only to two decimal places, not three as needed to assess the
accuracy requirement. The two-digit display is programmed such that the weights are rounded to the
nearest 0.01 Ib. In the example above, the scale would display between 19.99 and 20.01 Ib (i.e., error of ±
0.01 Ib not ± 0.009 Ib), which (technically) would result in exceeding the accuracy goal. Based on this, it
was concluded that precise verification of scale accuracy could not be performed in a straightforward
manner. In retrospect, the accuracy goal in the Test Plan should have been made consistent with the two-
digit display capability of the scale.
Nevertheless, the pre-test and post-test calibration results shown in Tables 3-2a and 3-2b suggest the
scales performed well. The maximum difference displayed between a measured weight and any standard
mass was 0.02 Ib, and in most cases, the difference was less than 0.01 Ib. The Center has used these field
calibration results to compute potential leak detection sensitivity errors due to scale error. Specifically,
calibration results between the scale readings and NIST weights are used to compute errors in full charge
and refrigerant withdrawal measurements. These errors are then propagated to compute actual leak
detection sensitivity errors reported for each test run.
3-2
-------
Table 3-2a. Scale Calibrations and Precision Data - Rooftop HVAC Unit
NIST
Stand-
ard
Mass
5.00
10.00
15.00
20.00
25.00
30.00
50.00
75.00
100.00
Runl
Pretest
Calibra-
tion
5.00
10.01
15.00
20.00
25.00
30.00
50.02
75.00
100.02
Post-test
Precision
10.02
15.02
20.02
25.02
10.02
15.02
20.02
25.02
Run 2
Pretest
Calibra-
tion
4.99
9.99
14.99
20.00
24.99
30.00
50.01
75.01
100.02
Post-test
Precision
9.99
14.99
19.99
24.99
9.98
14.98
19.98
19.99
Run 3
Pretest
Calibra-
tion
4.99
10.00
14.99
19.99
24.99
29.99
49.99
75.01
100.01
Post-test
Precision
20.00
25.00
30.00
50.00
20.00
25.00
30.00
50.01
Run 4
Pretest
Calibra-
tion
5.00
10.00
15.00
20.00
25.00
30.00
50.01
75.01
100.02
Post-test
Precision
19.99
25.00
29.99
49.99
19.99
24.99
29.99
50.00
Run 5
Pretest
Calibra-
tion
5.00
10.00
15.00
20.01
25.00
30.00
50.00
75.01
100.02
Post-test
Precision
20.00
25.00
30.00
50.01
20.00
25.00
30.00
50.00
Table 3-2b. Scale Calibrations and Precision Data - Reciprocating Chiller
NIST
Stand-
ard
Mass
5.00
10.00
15.00
20.00
25.00
30.00
50.00
75.00
100.00
Runl
Pretest
Calibra-
tion
5.00
10.00
14.99
19.99
24.99
29.99
50.00
75.00
100.00
Post-test
Precision
19.99
25.00
30.00
50.00
20.00
25.01
29.99
50.00
Run 2
Pretest
Calibra-
tion
5.00
10.00
15.00
20.01
25.00
30.00
50.00
75.00
100.00
Post-test
Precision
19.99
25.00
30.00
50.00
20.00
25.00
30.00
50.00
Run 3
Pretest
Calibra-
tion
5.00
10.00
15.00
20.00
25.00
30.00
50.00
75.00
100.01
Post-test
Precision
20.00
25.00
30.00
50.00
20.00
25.00
30.00
50.00
Run 4
Pretest
Calibra-
tion
5.00
10.00
15.00
20.00
24.99
30.00
49.99
75.00
100.01
Post-test
Precision
19.99
24.98
29.99
49.99
19.99
24.99
29.99
50.00
Run 5
Pretest
Calibra-
tion
5.00
10.00
14.99
19.99
24.99
29.99
50.00
75.00
100.00
Post-test
Precision
19.99
25.00
29.99
49.99
19.99
24.99
29.99
49.99
3-3
-------
Full charge and refrigerant withdrawal measurement errors were determined by computing an average
difference between pre-test and post-test calibration results at the weights observed during testing:
Error (Ib) = Average [Pre-Test Difference & Post-Test Difference]
Where: Pre-Test Difference = (Scale Reading - NIST weight), Ib
Post-Test Difference = average difference of precision results, Ib
(Eqn. 2)
For example, the initial weight of the test cylinder for Run 4 on the chiller was 30.26 Ib, and the final
weight was 31.43 Ib (i.e., the point at which the SGM alarmed). Both readings are representative of the
30 Ib (nominal) NIST weight. Based on the pre- and post-test calibration results with this standard weight
(Table 3-2b), the measurement error is computed to be -0.01 Ib, per equation 2 above. When this error is
accounted for, the actual initial weight is 30.26645 Ib and the actual final weight is 31.43645 Ib. The net
difference between the initial and final weights remains unchanged, and thus, the error in refrigerant
withdrawal is 0.00 Ib. For both units, the initial and final weights were representative of a single NIST
weight, and consistent with the example shown above, the overall error in the refrigerant withdrawal
measurements is 0.00 Ib for all test runs. The same approach was used to compute errors in full charge
measurements. For both units, the initial full charge weighing was compared with the 75 Ib NIST weight,
and the final full charge weighing was compared with the 30 Ib NIST weight. The largest error in full
charge measurement was 0.02 Ib. Table 3-3 summarizes the results of refrigerant withdrawal and full
charge measurement errors for all the runs.
Using the measurement errors in full charge and refrigerant withdrawals, leak detection sensitivity errors
for each Run were computed. As shown in Table 3-3, the maximum error in leak detection sensitivity
was 0.14 percent for Run 3 at the chiller. Note that errors are shown to two decimal places in the table for
simplicity. Since all the errors are less than the ± 2 percent specified in the Test Plan, the leak detection
sensitivity DQO was met for all runs.
Table 3-3. Summary of Measurement and Leak Detection Sensitivity Errors
Run
Full Charge
Measurements
Measured
Weight
(Ib)
Errora'b
(Ib)
Refrigerant Withdrawl
Measurements
Measured
Weight
(Ib)
Errora'b
(Ib)
Leak Detection Sensitivity
(%)
Error (% of leak
detection
sensitivity)
DQO
Achieved?
Rooftop HVAC Unit
1
2
3
4
5
51.74
49.72
52.35
52.35
52.37
0.00
0.00
+ 0.01
+ 0.01
-0.02
3.43
1.86
2.42
2.78
2.70
0.00
0.00
0.00
0.00
0.00
6.63
3.74
4.62
5.31
5.16
-0.01
-0.03
-0.08
-0.02
+ 0.17
Y
Y
Y
Y
Y
Reciprocating Chiller
1
2
3
4
5
47.10
47.12
46.25
46.25
46.68
0.00
0.00
0.00
0.00
0.00
2.13
2.10
1.27
1.17
1.65
0.00
0.00
0.00
0.00
0.00
4.52
4.46
2.75
2.53
3.53
-0.06
+ 0.07
+ 0.14
0.00
-0.14
Y
Y
Y
Y
Y
a As compared to actual NIST-traceable standard masses that are representative of range observed during testing
b For simplicity, errors are shown to two decimal places
3-4
-------
3.1.2 Data Completeness DQO Reconciliation
The Test Plan discussed the expected run-to-run variability in the test results. It is reasonable to expect
that 90 percent of the observed leak detection sensitivities will fall within 0.30 times the mean. Test
personnel used this range, or confidence interval (abbreviated e below), to determine the number of tests
to conduct on each unit. The Test Plan specified a completeness DQO as follows: "Test runs must be
repeated until 90 percent of observed values are within 0.30 times the mean leak detection sensitivity or a
maximum of five valid test runs are executed."
The GHG Center conducted five test runs at each unit. The confidence interval depends on the sample
standard deviation and the number of test runs conducted as follows:
(Eqn. 3)
Where:
e = 0.30 times the mean of all test runs
toos,n-i = 90 % T distribution value ( = 2.132 for five test runs)
s = sample standard deviation
n = number of sample runs (5)
Tables 3-4a and 3-4b present the individual test run results, the mean, standard deviation, and confidence
interval for each unit. At both locations, the 90 percent confidence interval is within 0.30 times the mean
leak detection sensitivity, and therefore the completeness goal was achieved.
Table 3-4a. Rooftop HVAC Data Completeness Goals
Run
1
2
3
4
5
Sample Standard
Deviation
Leak Detection
Sensitivity, %
6.63
3.74
4.62
5.31
5.16
1.057
Mean Leak Detection
Sensitivity and 90 %
Confidence Interval
5.09 ± 1.01 %
Required 90 % Confidence
Interval
± 1.53 %
Completeness DQO Achieved?
yes
3-5
-------
Table 3-4b. Chiller Data Completeness Goals
Run
1
2
3
4
5
Sample Standard
Deviation
Leak Detection
Sensitivity, %
4.52
4.46
2.75
2.53
3.53
0.929
Mean Leak Detection
Sensitivity and 90 %
Confidence Interval
3.56 ± 0.88 %
Required 90 % Confidence
Interval
± 1.07 %
Completeness DQO Achieved?
yes
3.2 QA/QC CHECKS FOR NON-CRITICAL MEASUREMENTS
The GHG Center Field Team performed QA/QC checks on additional instruments during the test
campaign. Data from these instruments did not directly contribute to leak detection sensitivity
determinations, but they allowed test personnel to verify that the test units were operating normally and
within expected parameters. Table 3-4 summarizes the results of these QA/QC checks.
Table 3-5. Summary of QA/QC Checks
Measure-
ment
Variable
Gauge
Manifold
Ambient
Temperature
and Relative
Humidity
Refrigerant
Line
Temperature
Sensors
QA/QC Check
Electronic leak check
Manifold and hose
positioning
Mfg. instrument
calibration
One-point temperature
check
Relative humidity
comparisons
Mfg. instrument
calibration
When
Performed/Frequency
Beginning of test on
each system
During testing
Within 12 months prior
to verification testing
Once per test day
Twice per test day
Prior to verification
testing
Expected or Allowable
Result
System should be leak tight
and purged of air
Hose and other accessories
connected to the cylinders
must be in identical
position during each
weighing
Temp:±0.2°F;RH±3%
± 2 °F when compared with
colocated thermocouple
±15% RH when compared
with Raleigh-Durham
International Airport
(RDU) data
Temp ±0.2 °F
Result
Acceptable?
yes
yes
yes
yes
Response to Check
Failure or Out-of-Control
Condition
Practice proper hose
purging procedures; repair
leaks as found
Restore manifold and hoses
to original position to
prevent weighing errors
Repair/replace defective
sensor or instrument;
recheck performance
Repair/replace defective
sensor or instrument;
recheck performance
3-6
-------
All QA/QC checks on non-critical instruments indicated that they performed properly throughout the
verification tests.
3.3 POTENTIAL REFRIGERANT COST SAVINGS
As indicated in the Test Plan, quantification of the accuracy and precision of potential refrigerant cost
savings is impossible because of the unknown quality of the available historical data. It was possible to
obtain handwritten logbook entries for the two units tested. Copies of these entries reside in the GHG
Center files. Section 2.4 discusses some of the interpretation limitations imposed by the unknowns in the
data.
3-7
-------
-------
4.0 TECHNICAL AND PERFORMANCE DATA SUPPLIED BY KMC CONTROLS, INC.
NOTE: This section provides an opportunity for KMC Controls, Inc. to provide additional comments
concerning the SGM and its features not addressed elsewhere in this Verification Report. The GHG
Center has not independently verified the statements made in this section.
KMC Controls, Inc. (KMC), in order to accomplish laboratory quality tests in a field environment, used
the following procedure to test this newly patented technology to conform to the requirements of the
Greenhouse Gas Technology Center (GHG Center) and Southern Research Institute (SRI). These tests
were performed for the U. S. EPA ETV program to provide quantifiable and repeatable test results for this
new technology. They were conducted for the purpose of viewing the functional and leakage related
activities of refrigerant through a Sight Glass Monitor on HVAC and other refrigeration equipment.
• In order to meet the test requirements, certain criteria were determined to be necessary to conform
to the test parameters. The charging procedures were governed by environmental conditions and,
as such, the KMC/Manufacturer's charging method is identified in the body of the report. It is
also defined in the Test and Quality Assurance Plan on Page B-5 [under the heading Procedures
for Charging the System] for the purpose of establishing a baseline methodology used during
these tests (SRI 2001).
• For the purposes of the test, the variables as encountered under the constraints of this testing
procedure may be different than those experienced under normal operating conditions. This could
deviate from some published or standard practices. KMC recommends all HVAC systems be
operated in compliance with Manufacturer's specifications and all industry and EPA guidelines.
In Addition, KMC would also like to acknowledge additional capabilities of the SLE-1001 that are not
specifically part of this testing and verification process.
• Utilizing a separate set of infrared detection electronics the KMC SGM has the ability to monitor
the moisture levels in the system. Elevated moisture levels can cause significant and catastrophic
damage or failure of a refrigeration system.
• The SGM has the ability when connected to a KMDigital system to provide continuous
monitoring and alarming features. When fully implemented, the SGM can create positive results
in operating efficiencies.
Following is the KMC SLE-1001 Sight Glass Monitor Specification Data Sheet, which contains a
functional description as well as guidelines and instructions for installation, wiring, operation, and
calibration.
4-1
-------
I
-------
White - Ground
Red - Flash Gas
Green - Moisture
Housing: Water and dust resistant, black flame retardant polymer, UL 94-5V rated
Dimensions: 3" x 2.5" x 1.5" (7.62 cm x 6.53 cm x 3.81 cm)
Ambient Limits
Operating: 32°F to 140°F (0°C to 60°C)
Shipping: -40°F to 140°F (-40°C to 60°C)
INSTALLATION |
The SLE-1001 will fit snuggly on a new or existing Sporlan Valve Company's See-All® Combination
Moisture & Liquid Indicator. Due to the mating on the sight glass window frame, only sight glasses that
have 0.25 in. clearance or more around the sight glass window frame can be used.
Existing Sight Glass Installation: The existing sight glass must be installed with the same requirements as
for a new installation. The sight glass must be bright and clear. The window must be clear and the inside
should not be dark or discolored. Some sight glasses can be reconditioned, while others may require
replacement.
New Sight Glass Installation: Install the sight glass and the SLE-1001 Sight Glass Monitor in a vertical
position with the flow of refrigerant upwards through the sight glass. If there is a pump-down solenoid in
the refrigerant system, install the sight glass upstream of the pump-down solenoid and downstream of the
drier. If the main reason for using the sight glass monitor is to increase the efficiency of the refrigerant
system, then in addition to the previous instructions, install the sight glass as close to the expansion valve
as possible. Care must be taken not to overheat the sight glass when soldering or brazing so that the sight
glass' interior body will not become discolored. Follow the sight glass manufacturer's installation
instructions for the proper installation of the sight glass.
SLE-1001 Installation: Remove the protective cap on the sight glass. The SLE-1001 fits the 1.34 in.
diameter sight glass window frame, and comes with an adapter ring for the smaller 1.13 in. diameter sight
glass window frame. Position the SLE-1001 over the sight glass window frame, using the adapter ring if
necessary. Make certain the directional arrow on the SLE-1001 label is going with the flow of the
refrigerant. The SLE-1001 mating surface must fit flush to the sight glass window. The SLE-1001 has a
set of stainless steel mounting extension springs attached to one side and a set of mounting hooks on the
other side. Take the two stainless steel mounting extension springs and pull one spring around the inlet
pipe, the other spring around the outlet pipe, and connect the loop end of each spring to a mounting hook.
The SLE-1001 should now be firmly mounted to the sight glass.
SLE-1001 Wiring: There are four wires to connect. The Black-wire connects to the phase of a 24 VAC
transformer and the White-wire connects to the ground of the transformer. The SLE-1001 has a half-wave
power supply, if the controller being used to monitor has a full-wave power supply, then a separate
transformer must be used for the SLE-1001. If a separate transformer is used, then the White-wire must
also connect to the controller's input ground. The Red-wire connects to the controller input that will be
monitoring the Flash Gas (bubbles). The Green-wire connects to the controller input that will be
4-3
-------
monitoring the moisture. If the controller being used is a KMDigital controller, remember to remove the
"pull-up resistor" from the input circuits.
OPERATION
Leak Detection: When the refrigerant system is properly charged, any flash gas detection (non-condensed
gas/bubbles) could indicate a refrigerant leak and low head pressure. The Flash LED will pulse red more
frequently the more bubbles that are detected, and the voltage output will increase proportionally the more
bubbles that are detected. KMC recommends charging per equipment manufacturer's recommendations
and until the sight glass is clear of bubbles.
Moisture Detection: When the sight glass colored indicator element changes from green to yellow,
indicating moisture in the system, the Moisture LED will begin to glow yellow and will glow brighter in
proportion to the degree of yellow of the sight glass element, and the voltage output will increase
proportionally to the degree of yellow of the sight glass element.
CALIBRATION
There is no field calibration of the SLE-1001 required. Periodic inspection of the sight glass should be
performed. The sight glass must be bright and clear. The window must be clear and the inside should not
be dark or discolored. Some sight glasses can be reconditioned, while others may require replacement.
4-4
-------
5.0 REFERENCES
Air-Conditioning and Refrigeration Institute, Refrigeration and Air-Conditioning, Second Edition,
Prentice Hall, Upper Saddle River, NJ. 1987.
ANSI/ASHRAE Standard 15. Safety Code for Mechanical Refrigeration Units. ANSI/ASHRAE 15.
American National Standards Institute/American Society of Heating, Refrigeration, and Air-Conditioning
Engineers, Atlanta, GA. 1994.
Carrier. Installation, Start-Up and Service Instructions, 48DJ, DK, NP034-074, 50DJ, DK, DY, NB,
NP034-074 Single Package Heating and Cooling Units. Carrier Corporation, Syracuse, NY. 1996.
Code of Federal Regulations, 40 CFR Part 82, Protection of Stratospheric Ozone; Refrigerant Recycling,
Vol. 58 No. 92 Friday, May 14, 1993.
Energy Information Administration, Emissions of Greenhouse Gases in the United States 1996,
DOE/EIA-0573(96), U.S. Department of Energy, Energy Information Administration, Washington, DC.
October 1997.
Energy Information Administration, Emissions of Greenhouse Gases in the United States 1999,
www.eia.doe.gov/oiaf/1605/ggrpt, DOE/EIA-0573(99), U.S. Department of Energy, Energy Information
Administration, Washington, DC. 1999.
Federal Register, 63 FR 32044, Protection of Stratospheric Ozone; Refrigerant Recycling; Substitute
Refrigerants; Proposed Rule. June 11, 1998.
Federal Register, 60 FR 40420, Protection of Stratospheric Ozone; Supplemental Rule to Amend Leak
Repair Provisions Under Section 608 of the Clean Air Act; Final Rule. August 8, 1995.
Moravek, Joseph, Air-conditioning Systems, Principles, Equipment, and Service, Air-conditioning and
Refrigeration Institute, Prentice Hall, Upper Saddle River, NJ. 2001.
SRI. Test and Quality Assurance Plan for the KMC Controls, Inc. SLE-1001 Sight Glass Monitor.
Southern Research Institute, Research Triangle Park, NC. 2001.
Thermal Engineering Company data sheet, How Long Should You Evacuate A System, Toledo, OH.
5-1
------- |