Enhanced Operations & Maintenance Procedures for Small Packaged Rooftop HVAC Systems

Protocol Development, Field Review, and Measure Assessment

Final Report April 2002

Prepared for:

Eugene Water and Electric Board Eugene, OR

Robert Davis Paul Francisco Mike Kennedy David Baylon

Bruce Manclark, Delta-T

i

TABLE OF CONTENTS Executive Summary ........................................................................................................... iv 1. Introduction..................................................................................................................... 1

1.1. Background .............................................................................................................. 1 1.2. Project Overview ..................................................................................................... 2 1.3. Field Measurement Overview.................................................................................. 3

2. Protocol Development .................................................................................................... 5 2.1. Evaporator Airflow .................................................................................................. 5 2.2. Refrigerant Charge................................................................................................... 8 2.3. Economizers........................................................................................................... 10

2.3.1. Economizer Characterization.......................................................................... 12 2.3.2. Generalized Sequence of Operation................................................................ 16 2.3.3. Economizer Functional Testing ...................................................................... 17 2.3.4. Sensor Operation............................................................................................. 19 2.3.5. Minimum Outside Air..................................................................................... 19

2.4. Duct Losses............................................................................................................ 20 2.5. Thermostats............................................................................................................ 20

3. Results........................................................................................................................... 21 3.1. Evaporator Airflow ................................................................................................ 21 3.2. Refrigerant Charge................................................................................................. 23

3.2.1. Coil Cleaning .................................................................................................. 25 3.3. Economizers........................................................................................................... 27

3.3.1. Field Results.................................................................................................... 27 3.4. Duct Losses............................................................................................................ 34

4. Energy Savings Impacts and Conservation Measures .................................................. 35 4.1. Coil Cleaning ......................................................................................................... 35 4.2. Airflow Improvement ............................................................................................ 36 4.3. Refrigerant Charge................................................................................................. 37 4.4. Economizers........................................................................................................... 38

4.4.1. Economizer Simulations ................................................................................. 38 4.4.2. Economizer Repair Measures ......................................................................... 40

4.5. Energy Savings and O&M Impacts ....................................................................... 42 4.6. Cost/Benefit Analysis ............................................................................................ 43

4.6.1. Operations and Commissioning Packages ...................................................... 44 4.6.2. Maintenance and Repair ................................................................................. 46 4.6.3. Enhanced Maintenance Procedures ................................................................ 48

5. Contractor Relations ..................................................................................................... 48 6. Conclusions................................................................................................................... 52 7. References..................................................................................................................... 55 Appendix A: Rooftop Packaged Unit Protocols ............................................................. A-1 Appendix B: Duct Loss Discussion ................................................................................ B-1

i

TABLE OF TABLES

Table 3-1. System Fan Flow Results ............................................................................... 22 Table 3-2. Refrigerant Test Results ................................................................................. 25 Table 3-3. Summary of Diagnosed Economizer Operational Issues ............................... 27 Table 3-5. Outdoor Sensor Characterization ................................................................... 29 Table 3-6. Economizer Airflow Results .......................................................................... 33 Table 4-1. Damper Adjustment Impacts .......................................................................... 37 Table 4-2. Ratio of Actual Economizer Savings to Ideal Economizer ............................ 40 Table 4-3. Savings (Percent of Annual Cooling Energy) ................................................ 42 Table 4-4. Cost/Benefit Analysis (7-ton Prototype Unit) ................................................. 44

ii

TABLE OF FIGURES Figure 1-1. Economizers on Eugene Rooftop……………………………………………3 Figure 2-1. TrueFlow™ Meters in Unit Slot……………………………………………. 8 Figure 2-2. TrueFlow™ Meters on Economizer…………………………………………8 Figure 2-3. Fouled Evaporator Coil.……………………………………………………...9 Figure 2-4. Water Drainage from Cleaning Fouled Evaporator Coil…………………….9 Figure 2-5. Fouled Condenser Coil…………………………………………………….. 10 Figure 2-6. Cleaning Fouled Condenser Coil…………………………………………...10 Figure 2-7. Fouled Evaporator Coil……………………………………………………..13 Figure 2-8. Water Drainage from Fouled Condenser Coil……………………………...13 Figure 2-9. Solid State Dry Bulb Sensor………………………………………………..13 Figure 2-10. Electromagnetic Snap Disk………………………………………………..13 Figure 3-1. Outdoor Temperature vs. Compressor Amp Draw…………………………26 Figure 3-2. Outdoor Air Fraction………………………………………………………..34

iii

Executive Summary Introduction This report details the development of a field protocol used to evaluate the performance of packaged heating and cooling equipment installed on commercial buildings served by the Eugene Water and Electric Board (EWEB). The protocol is intended to identify energy savings opportunities and document changes in equipment operation with the potential to exploit such opportunities. Work was carried out during the summer and early fall of 2001. Rooftop packaged heat pumps and air conditioners are commonly used for space conditioning in light commercial buildings. This equipment is generally characterized by a constant-volume, single-speed fan with a gas-fired heating element or heat pump to provide heating; a cooling coil and compressor for air conditioning; and series of dampers and plenums to manage airflow and ventilation to the space served. The systems can be installed with economizers, which are meant to provide cooling to spaces using outdoor air when air temperature is “appropriate.” Since the use of the economizers in the Pacific Northwest should be able to offset half the electricity associated with cooling requirements, this technology has been mandated by energy codes in most units since the mid-1980s. Furthermore, the economizer has been a key factor in utility sponsored conservation programs for the commercial sector. By 1996, the economizer was mandated for all packaged rooftop systems with more than 5 tons of nominal cooling capacity (60,000 BTUh). As a result, an increasing number of these units are getting installed with economizers, which are meant to use outdoor air for cooling whenever possible. In most cases, HVAC companies do relatively non-invasive preventive maintenance on rooftop units, including changing system filters, performing routine checks on system operation, and occasionally performing more involved assessments of refrigerant charge. Little if any work is done to assure proper airflow or proper operation of economizers, or to assess and mitigate the possible impacts of duct losses. This level of maintenance is designed to provide an annual check of the equipment, but it seldom extends beyond a visual inspection and new air filters. This work is done under a maintenance contract that pays for a small amount of service and an on-call technician in the event of a unit failure. There are only limited studies showing the potential energy savings from more aggressive maintenance in packaged units. Hewett et al. (1992) found savings on the order of 1,900 kWh/yr for 17 Midwest rooftop sites that received refrigerant, airflow, and duct repairs. Vick et al. (1991) reported modest savings from tuning up the gas furnace portion of commercial package units; duct repairs were also undertaken but were not monitored, and economizer function was not evaluated (though it was identified as a potential source of substantial energy savings). Houghton (1997) reported savings of about 25% attributable to maintenance of air filters, coils, compressors and outdoor air dampers (economizers).

iv

Delp et al. (1998) largely disproved the assumption that duct losses, because they were thought to be within the air and thermal barriers of the building, are not significant. Literature on economizers consists of the few studies in which economizer review was conducted. These studies have been based on a limited sample in a various localities. In general, reports of broken economizers range from about 50% to 80% of all units (e.g. Lunneberg, 1999; Davis Energy Group, 2001). Other efforts (e.g. Breuker et al. 2000; Pratt et al. 2000) have focused on monitoring and optimizing economizer function in the context of automated diagnostic systems. All these efforts are based on data collected in other parts of the country, generally using relatively older equipment. The relationship to the EWEB service territory or any other Pacific Northwest locality is very debatable. Furthermore, none of these papers distinguish between the various control options and configurations available for economizers and the distinct energy and maintenance issues raised by these strategies. Very little work has been done to develop and deliver hands-on protocols that can be used by HVAC technicians to characterize these systems and document energy efficiency improvements. It was hoped that, if a protocol was developed and tested in the field, technicians could incorporate it into a more comprehensive approach to troubleshooting and, in some cases, repairing rooftop units. Project Overview In early summer of 2001, a project to develop and deliver such a protocol for rooftop units in the Eugene Water and Electric Board (EWEB) service territory was undertaken by Ecotope, Inc., of Seattle. Technical and coordination support was provided by Delta T, Inc., and Eugene area HVAC contractors were hired to work on the systems and participate in learning the evaluation protocol. The primary goals of this project were to:

1. Develop a protocol that would allow contractors to evaluate the performance of the units, and to train contractors in using the protocol. The protocol included evaluation of refrigerant charge, airflow, economizer operation, and duct losses.

2. Develop an approach to the maintenance and repair of the units using information gathered during the evaluation to improve unit efficiency.

3. Use HVAC technicians to assist in the use of the protocol and to deliver this enhanced maintenance.

4. Evaluate the savings available from these systems and the cost-effectiveness of measures that might be applied to these units.

Because a comprehensive protocol did not exist prior to the project, this process required field testing and evaluation of the performance of a number of units, which has the added advantage of establishing a database of results found for tested systems. Refrigerant

v

charge, airflow, and economizer operation were the primary focuses of the fieldwork, with duct losses of secondary interest. Also of interest was system scheduling (thermostat settings and air handler operation). Since evaluation of refrigerant charge was of primary interest, it was necessary to perform the fieldwork during the cooling season. Conditioning equipment tested in this project ranged in size from 2.5 to 15 nominal tons of cooling, with the average being about 7 tons. The units ranged in age from the mid-1980s to very new (installed in the year 2001). All units but one were unitary packaged systems. One split-system unit was included as part of the study: a 5-ton unit that used R-410A as its refrigerant. Six of the units had dual compressors; the rest had single compressors. Some units used natural gas for heat, some were heat pumps, and others did not include heating. Most systems were constant volume; a few cases used variable-volume controls or zone dampers. In most cases, the evaporator fan was set to run on constant speed, though we did encounter some variable speed air handler systems. Testing was only done on units that could be maintained at a constant flow rate for the duration of the tests; if a system was constantly changing airflow and mixed air temperature (such as is often the case with VVT systems) most tests were not performed. Economizers were evaluated for a variety of issues. These included type of ambient sensor (dry-bulb vs. enthalpy), sensor setting, damper type, damper position controller, ability to move between positions, and airflow with the dampers in different positions. Economizers were also evaluated when possible on whether the dampers changed position at the sensor setpoint. Duct systems and thermostat type were also reviewed. Ducts were evaluated on whether they ran outside the thermal and air boundary envelopes of the building, ran completely inside, or ran in some sort of buffer space which might be influenced by both inside and outside conditions. Thermostats were inspected to determine settings during different periods, such as occupied vs. unoccupied periods and weekends vs. weekdays. They were also checked for constant fan operation. In a few cases it was determined whether they were single-stage or two-stage thermostats. Field Results and Energy Savings Estimates Various measurements of equipment performance were collected as part of the work. Of primary interest were refrigerant charge, evaporator airflow, economizer airflow (minimum/maximum air flow), and economizer function. Energy savings and measure economics are summarized in the table found near the end of this summary. Refrigerant charge evaluation was carried out with the CheckMe!™ program, developed by Proctor Engineering. This program uses the Carrier method (actual superheat or subcooling versus target superheat or subcooling). The CheckMe!™ procedure could be used on 14 units, with the following results:

vi

• Four units were undercharged, with two showing significant undercharge (needing

more than an 8 oz adjustment). • Five units were overcharged; the average amount of overcharge was about 10% of the

factory charge. In two cases, an apparent overcharge was due to a dirty condenser coil. When the coil was cleaned, the evaluation changed to correct charge. Depending on the size of the charge adjustment, annual cooling energy savings can be substantial. Evaporator airflow has significant bearing on system performance. Airflow was measured on 27 units with the new TrueFlow™ device, which enables direct assessment of airflow (unlike the indirect assessment offered by the Carrier method).

• Average evaporator airflow was 304 CFM/ton, with a range from 99-420 CFM/ton.

• Manufacturers recommend evaporator airflow of 400 CFM/ton of nominal cooling; the measured lower average represents an energy savings potential of about 10%.

Economizer airflow has direct bearing on indoor air quality and upon the amount of outside air available for “free” cooling (when outdoor air is used for cooling rather than the compressor).

• Minimum airflow averaged 20% of full system airflow. On average, this is close to the level of outside air recommended for ventilation by ASHRAE. The amount of minimum outdoor air fraction ranged from 0 to 92% of system airflow.

• Maximum outdoor airflow averaged 65% of full system airflow. This is an interesting finding, since it suggests that full system airflow cannot be supplied by outdoor air alone and therefore “free” cooling is less than supposed.

A review of economizer function found a variety of problems with sensor operation, system changeover settings, damper operation, and related issues. Because the economizer section of the protocol was in continuous development during the project, it is difficult to generalize problems with economizers. However, the following tendencies are noted:

• Changeover settings are usually set to non-aggressive levels (55° F or cooler) and can be changed to more aggressive levels (60° F). This represents a potential savings of about 10% of cooling energy, based on DOE-2.1e simulations, and is estimated to occur in about 35% of EWEB’s service territory.

• Where economizers are non-operational (estimated at 30% of all cases in the EWEB service territory) and are upgraded to a dry-bulb changeover setting of 60° F, simulated annual cooling energy savings are on the order of 25%.

vii

In general, there are considerable opportunities for saving energy in the EWEB service territory from a combination of coil cleaning, air handler fan adjustment, and economizer adjustment. A systematic protocol will identify these opportunities and monitor progress in exploiting them. Cost/Benefit Analysis To evaluate the costs and benefits of an enhanced maintenance package for rooftop units, the individual measures had to be combined to reflect the manner in which these costs and benefits interact in specific cases. While the nature of these combinations is somewhat arbitrary, they do allow costs and benefits to be assigned to packages that are composites of enhanced maintenance and repair measures. This evaluation is based on an average unit observed in the EWEB service territory attached to a retail building characterized by the Bonneville Power Administration (BPA) small retail prototype. The average unit size that we observed in this sample was between 6.5 and 7 tons of cooling capacity, or approximately 80,000 BTU/hr of total cooling output. For purposes of this analysis, we have assumed that an individual packaged unit serves a retail zone of 2,700 square feet, with a median lighting power density that corresponds to an overall predicted cooling load of 11,000 kilowatt hours per year in the Eugene climate. A cost/benefit analysis is presented in Table 1. It has been derived from the combination of individual measures that were applied or were thought to be applicable to the rooftop units evaluated in Section 3 of this report. Table 1. Cost-Benefit Analysis (7-ton Prototype Unit)

Savings Cost ($) (kWh/Year) ($/kWh)

Payback(Years)

Commissioning Measures 1. Change-Over 275 1,100 .031 3.3 2. Change-Over with T-Stat 675 2,500 .033 3.6 3. Control Board 775 3,200 .030 4.8 4. Control Board with T-Stat 1,125 4,100 .034 3.6 5. New Economizer 1,600 4,600 .043 4.6 Repair Measures 1. Refrigerant Charge* 100 350 .064 3.8 2a. Damper Repair (Reset)* 150 450 .075 4.4 2b. Damper Repair (Flow Adjustment) 250 820 .069 4.1 3. Gas Combustion Test* 100 - - - *Gas savings and/or heating impacts not calculated. For purposes of analysis, five separate retrofit packages were assessed, ranging in complexity and expense. These packages were designed to address the diagnosis and repair of economizers and airflow as a result of the direct review of the equipment using a

viii

systematic protocol. In all cases, these repairs and procedures were designed to have a fairly long life and to correct difficulties that have resulted in deferred maintenance and related wear and tear, as well as design and installation that resulted in less-than-optimal operation of the equipment. Package 1: Coil Cleaning and recalibration of economizer sensors and minimum air setting: This package includes coil-cleaning of both the evaporator and condenser coils together with a review of the economizer sensors and economizer changeover settings, etc. This package of measures could be considered routine commissioning, in which a fairly straightforward diagnostic results in the technician understanding and repairing the economizer. Package 2: Reset change-over with addition of two-stage thermostat control : In approximately 25% of the units reviewed under this program, the control provided by the thermostat was a single-stage cooling set-point. In order for the economizer to function optimally, a two-stage cooling set-point was included. This package is essentially identical to Package 1, except that the thermostat would be placed in the zone and the system would be re-wired to accept two-stage control. Package 3: New economizer control board: This package would be based on an initial review of the economizer controls to determine the level of diagnostic effort required to repair or commission the economizer. In the event that these procedures did not yield sufficient information, a new control board would be installed. This control board would be assumed to handle two stages of cooling and provide integrated control for the economizer. This measure would allow the economizer to be optimized using a set of known controls and sensors, and a documented calibration to facilitate future maintenance. Package 4: Package 3 with additional two-stage thermostat: This option is included on the assumption that under some circumstances, a two stage thermostat would not be present and would have to be included in the economizer repair package in order to get both optimized control and effective temperature management in the zone. Package 5: New economizer: In spite of extensive diagnostics, there are some circumstances in which it may not be possible to even install a new control system. In these cases, the most obvious maintenance strategy would be to remove the existing economizer and install a modern model. In many cases (and especially in older equipment) this would obviate the need for detailed diagnostics. Work With Contractors The purpose of this effort was to try and establish the veracity of an enhanced HVAC contractor-based maintenance procedure in developing both energy savings and the services available to EWEB customers. Notwithstanding for a very limited sample, this effort showed the importance of enhanced maintenance in the overall functionality of rooftop units.

ix

In this sense, the mere act of having a systematic protocol that requires certain things to be checked, and especially a protocol that requires a review of refrigerant charge and damper function, would harness a significant fraction of the savings and benefits that were identified in our review. The more sophisticated issues with economizer controls and complex issues with dampers would require a larger amount of effort. In many cases, this larger amount of effort would also yield noticeably greater savings, although in most cases these savings would also include functional compromising of the equipment to the point where a catastrophic failure might result in the absence of the measure. Contractors did express some hesitation to using the advanced diagnostic techniques in this project. Regarding airflow, there was some feeling that since nothing could be done to change the flow in many systems, one should not bother to measure it. For economizers, the process is seen as lengthy and nebulous, since it can be difficult to identify the source of a problem and since there are so many different types of economizers. It is quite apparent from the nature of the business we observed in this project and from the time allotted to this operation that the existing maintenance contracts are unlikely to result in any of the potential savings measures being identified, let alone attempted. Only if the contractors are able to sell an enhanced service, paid for either directly by the utility or by the customer, would any of these reviews even be possible. However, it must also be pointed out that, even in the presence of such a marketing effort, it is not entirely clear that the contractor services as currently observed would be able to deliver a sophisticated system review on any sort of regular basis. Conclusions This effort showed, albeit for a very limited sample, the importance of enhanced maintenance in the overall functionality of rooftop units. Even where enhanced controls would not be considered a viable measure, about 60% of the units reviewed required some maintenance that would result in improved function and efficiency. In about 20% of the cases, the review identified catastrophic or near-catastrophic failures that had not been recognized in previous maintenance visits. Contractors generally sell the routine maintenance check as a loss-leader. Seldom is the budget or time available to review even the refrigerant charge. Most technicians we spoke with believe this to be a waste of time and, in fact, in many cases it is of marginal significance. However, in a few cases (15% of those reviewed in our protocol), the result of applying the CheckMe!™ program led to the identification of serious failures or serious difficulties with the air coils that could not have been easily identified any other way. With economizers, there are really two layers of review, but at this stage the ability to understand a wide variety of economizers and the potential benefits from a detailed approach is limited.

x

It is quite clear that even economizers that are set up properly and operate correctly may not be delivering a very high fraction of the potential economizer benefits because of non-aggressive changeover settings. Technicians often have neither the time nor the tools to address these issues. Only if the contractors are able to sell an enhanced service, paid for either directly by the utility or by the customer, would any of the reviews even be possible. It is quite likely that commissioning these small rooftop units so that the controls and dampers are properly set up at installation might achieve these performance benefits without enhanced maintenance services. In this scale of equipment, there is generally no quality control on the installation, let alone commissioning. It is obvious from the evaluation of the measure packages that a part of a cost-effective program would have to include a long term and extensive review of the operation, installation, and maintenance of the entire packaged unit. At the outset, this would be an extensive and expensive addition to the existing maintenance agreements offered by the service companies. If this program offered the opportunity to supply a diagnostic service and replace non-functional parts profitably, contractor resistance to this level of involvement would be reduced. An enhanced diagnostic procedure would include full operational review, repair, and replacement of economizer controls, reset of change-over controls, and evaluation and reset of airflow charge and damper settings in the initial year of operation. This would be followed by annual or semi-annual visits to ensure that the repairs continue to function and that any deficiencies in charge are quickly identified and repaired. This would also include periodic cleaning of coils and changing of air filters, thus replacing existing maintenance contracts. While we believe that this protocol remains somewhat incomplete and certainly has not addressed all the potential savings available from economizers, it is clearly evident that an enhanced service of this type can yield great benefits and energy savings to the utility and substantial benefits in reduced maintenance and energy costs to the building operators. As such, considerable effort should be expended to establish this protocol inside of the contractor community in Eugene, and to ensure that owners and operators of this type are aware of and demand the level of maintenance and review that this effort suggests.

xi

1. Introduction

1.1. Background Rooftop packaged heat pumps and air conditioners are commonly used for space conditioning in light commercial buildings. This equipment is generally characterized by a constant volume single speed fan with a gas fired heating element or heat pump to provide heating; a cooling coil and compressor for air conditioning; and series of dampers and plenums to manage air flow and ventilation to the space served. Several important variations were observed which include variable volume/temperature (VVT) systems designed to modulate air delivery based on zone temperature. We also observed several systems that were designed to provide cooling only. In those cases the heating was supplied by a secondary system in the space (usually electric zone heating). The systems can be installed with economizers, which are meant to provide cooling to spaces using outdoor air when air temperature is “appropriate.” Since the use of the economizers in the Pacific Northwest should be able to offset half the electricity associated with cooling requirements, this technology has been mandated by energy codes in most units since the mid 1980s. Furthermore, the economizer has been a key factor in utility sponsored conservation programs for the commercial sector. By 1996 the economizer was mandated for all packaged rooftop systems larger than 5 tons of nominal cooling capacity (65,000 BTUh) As a result an increasing number of these units are getting installed with economizers. While these code requirements have been part of the Oregon and Washington energy codes for over 20 years the nature of the economizer and economizer controls in these small package systems have evolved with increasingly complex and effective controls which operate the economizer and control the heating and cooling. The installation and maintenance of these economizers is critical to the overall efficiency of the system but the capacity and controls are designed to function without any of the energy savings features operating at all. As a result, the effectiveness of the unit at providing heating and cooling is very often maintained while the economizer and other control and operating features are compromised. In most cases, HVAC companies do relatively non-invasive preventive maintenance on rooftop units, including changing system filters, performing routine checks on system operation, and occasionally performing more involved assessments of refrigerant charge. Little if any work is done to assure proper airflow or proper operation of economizers, or to assess and mitigate the possible impacts of duct losses. This level of maintenance is designed to provide an annual check of the equipment but seldom extends beyond a visual inspection and new air filters. This work is done under a maintenance contract that pays for a small amount of service and an on-call technician in the event of a unit failure. There are only limited studies showing the potential energy savings from more aggressive maintenance in packaged units. Hewett et al. (1992) found savings on the order of 1900 kWh/yr for 17 Midwest rooftop sites that received refrigerant, airflow, and duct repairs. Vick et al. (1991) reported modest savings from tuning up the gas furnace portion of

1

commercial package units; duct repairs were also undertaken but not monitored, and economizer function was not evaluated (but was identified as a potential source of substantial energy savings). Houghton (1997) reported savings of about 25% attributable to maintenance of air filters, coils, compressors and outdoor air dampers (economizers). Delp et al. (1998) showed that, despite the assumption that duct losses are not important because the ducts are thought to be within the air and thermal barriers of the building, these assumptions are often incorrect. Literature on the economizer has focused on a few studies where economizer review was conducted. These studies have been based on a limited sample in a various localities. In general reports of broken economizers range from about 50% to about 80% of all units (e.g. Lunneberg, 1999; Davis Energy Group, 2001). Other efforts (e.g. Breuker et al. 2000; Pratt et al. 2000) have focused on monitoring and optimizing economizer function in the context of automated diagnostic systems. All these efforts are based on data collected in other parts of the country usually using equipment that is relatively older. The relationship to the EWEB service territory or any other Pacific Northwest locality is very debatable. Furthermore, none of these papers distinguish between the various control options and configurations available for economizers and the distinct energy and maintenance issues raised by these strategies. Very little work has been done to develop and deliver hands-on protocols that can be used by HVAC technicians to characterize these systems and document energy efficiency improvements. It is hoped that, if a protocol was developed, technicians could incorporate it into a more comprehensive approach to troubleshooting and, in some cases, repairing rooftop units.

1.2. Project Overview In early summer of 2001, a project to develop and deliver such a protocol for rooftop units in the Eugene Water and Electric Board (EWEB) service territory was undertaken by Ecotope, Inc., of Seattle. Technical and coordination support was provided by Delta T, Inc., and Eugene area HVAC contractors were hired to work on the systems and participate in learning the evaluation protocol. See Fig. 1-1 for an example of a typical rooftop. The primary goals of this project were to 5. Develop a protocol that would allow contractors to evaluate the performance of the

units, and to train contractors in using the protocol. The protocol included evaluation of refrigerant charge, airflow, economizer operation, and duct losses.

6. Develop an approach to the maintenance and repair of the unit using information gathered during the evaluation to improve the efficiency of the unit.

7. Use HVAC technicians to assist in the use of the protocol and to deliver this enhanced maintenance.

8. Evaluate the savings available from these systems and assess the cost-effectiveness of various measure operations and maintenance packages.

2

Figure 1-1. Roof of pet store in Eugene with four package units with economizers visible. Economizers are the triangular hoods at the ends of the units.

1.3. Field Measurement Overview Because a comprehensive protocol did not exist prior to the project, this process required field testing and evaluation of the performance of a number of units, which has the added advantage of establishing a database of results found for tested systems. Refrigerant charge, airflow, and economizer operation were the primary focuses of the fieldwork, with duct losses of secondary interest. Also of interest was system scheduling (thermostat settings and air handler operation). Since evaluation of refrigerant charge was of primary interest, it was necessary to perform the fieldwork during the cooling season. As much of the work as possible would be carried out using standard service tools (refrigerant gauges, voltage test meters, hand tools, etc.). Refrigerant charge would be conclusively evaluated by the “Carrier method”, as run in semi-automated fashion by Proctor Engineering Group (PEG) in their CheckMe!™ Program, which allows contractors to call a toll free number, relay their measurements, and receive results and any recommendation for changing the charge. Airflow would be evaluated using the new TrueFlow™ Air Handler Flow Meters. Since evaluation of economizer operation is more a qualitative assessment of mechanical operation and controls than a quantitative measurement, this portion of the protocol would be essentially answering a set of questions and diagnosing any failures by whatever means necessary.

3

Some units would be monitored with data-logging equipment to assess changes in energy uses as a result of repairs. EWEB currently keeps track of energy usage on over 100 commercial buildings with the automated meter reading (AMR) process. It was hoped that this process could be utilized in assessing energy savings from units in buildings involved in the project. To further estimate the potential savings from performing additional maintenance and repairs on these types of units simulations were performed using DOE-2.1e. In developing the protocol, the intention was to test 30 buildings. The definition of “building” is somewhat fluid but, given the time required to complete the protocol, 1 to 2 units could be visited in an 8 hour period and so a total of 30 to 45 units were expected to be included in the project. Because of various factors, only 30 units were evaluated in a total of 19 separate businesses, including a small group of buildings in the Puget Sound area. Some units were in complexes serving a number of businesses, such as Willamette Square in Eugene and Kingsgate in the Puget Sound area. Not all parts of the final protocol were completed for each unit, since the protocol was under development throughout the project and because, in some cases, various problems prevented one or more major categories of the protocol from being completed. Also, toward the end of the field testing, it was felt that time was best spent getting the best understanding possible on economizers, so the flow and charge protocols, which were both well in hand, were sacrificed. Conditioning equipment tested in this project ranged in size from 2.5 to 15 nominal tons of cooling with the average being about 7 tons. The units ranged in age from the mid-1980s to very recent (i.e., installed in the year 2001). All units but one were unitary packaged systems. One split-system unit was included as part of the study: a 5-ton unit that used R-410A as its refrigerant. Six of the units had dual compressors; the rest had single compressors. Some units used natural gas for heat, some were heat pumps, and others did not include heating. Most systems were constant volume single zone; a few cases used variable volume controls or zone dampers. In most cases, the evaporator fan was set to run on constant speed, though we did encounter some variable speed air handler systems. Testing was only done on units that could be maintained at a constant flow rate for the duration of the tests; if a system was constantly changing airflow and mixed air temperature, such as is often the case with VVT systems, most tests were not performed. Economizers were evaluated on a variety of issues. These included type of ambient sensor (dry-bulb vs. enthalpy); control logic; orientation of dampers; style of dampers; damper position controller; ability to move between positions accurately; and airflow with the dampers in different positions. Economizers were also functionally tested, when possible, to determine that the unit responded correctly to ambient conditions. To understand the impact of various economizer operation strategies and control approaches, we developed several detailed DOE-2.1e simulations. This process used the prototype commercial buildings used to evaluate regional conservation programs. The

4

effect of these runs was to provide guidance as to the control algorithms and other operational approaches to diagnosing the economizers. Duct systems were reviewed qualitatively in this project, primarily by characterizing the type of air space in which the majority of the ducts run. The goal was to judge whether the ducts were running outside the thermal and air boundary envelopes of the building, running completely inside, or running in some sort of buffer space which may be influenced by both inside and outside conditions. The effect of duct losses depends greatly on the location of the ducts. Thermostats were inspected to determine settings during different periods, such as occupied vs. unoccupied periods and weekends vs. weekdays. They were also checked for constant fan operation. In a few cases it was determined whether they were single-stage or two-stage thermostats. In addition to the collection of technical information and the development of the protocol, it was also of great interest to work with HVAC technicians in the field and try to facilitate independent contractor utilization of this protocol. In part because of scheduling difficulties and contractor availability, this part of the project was not successful; that is, contractors did not get to a point where they could use the procedure on their own. It is hoped that follow-up work will achieve this goal, since even the best protocol will be of no use to the utility if it is never used. The following section, section 2, of the report discusses protocol development. Section 3 includes the results from the field tests, which is broken down into sections on evaporator/economizer airflow; refrigerant charge; economizer function; and ducts. Section 4 discusses estimates of energy savings from improvements, including data from tested units as well as simulation results. Section 5 discusses contractor participation and the possibility of developing a utility-based program for this sector. Section 6 includes the conclusions and recommendations that result from this study. The final protocol is included as Appendix A. This Appendix includes a protocol checklist, as well as detailed descriptions of how to perform various tests. 2. Protocol Development It was understood from the beginning of the project that the field protocol would be under constant development. The elements of the protocol would be: measuring evaporator airflow, measuring system refrigerant charge, assessing the type and function of the economizer, and reviewing the performance of the duct system. During the project, several items were added to the protocol (including more specific information on compressor and economizer function).

2.1. Evaporator Airflow One of the main points of concentration in this study was the direct measurement of airflow across the system evaporator. Several researchers (e.g. Parker et al. 1997; Proctor et al. 1997; Proctor 1991) have described the heat delivery efficiency of heating and

5

cooling systems as a function of the airflow across the evaporator. The mass flow of air across the heat exchanger will have a direct bearing on the ability of the system to meet the cooling or heating load. Air conditioning systems depend on a certain amount of air moving across the evaporator at any given time to effectively transfer heat from the conditioned space to the refrigerant, after which the refrigerant being circulated by the compressor gives up the heat at the condenser. Manufacturers of air conditioning equipment have typically suggested that an airflow rate of 400 ft3/min (CFM) per nominal ton of cooling is the desired airflow. Although manufacturers vary somewhat about this number, 400 CFM/ton is the generally agreed-upon benchmark for assessing whether airflow is adequate. Various laboratory and field researchers have found that airflows of 350 and even 325 CFM/ton do not significantly degrade either system capacity or efficiency, though reduction of the flow rate does cause a greater portion of the capacity to be directed toward latent cooling (moisture removal is largely irrelevant in the Pacific Northwest “cooling” climates). Low airflow is often due to inadequately-sized ducts or dirty filters or coils. Other causes can be inappropriate selection of fan speed on a multi-speed fan. High airflow is usually due to an incorrect fan speed setting. There are various methods that have been used in the past to measure system airflow. The industry standard method is that described by Carrier Corporation in their refrigerant-charging/airflow measurement procedures. In these procedures, the cooling load on the air conditioner is used along with the sensible temperature split across the evaporator (the temperature difference between air upstream and air downstream of the evaporator coil) to impute whether the airflow is adequate, too low, or too high. The cooling load is determined using the condensing air entering temperature (usually fairly close to the ambient—i.e., outdoor—air temperature) and the wet bulb temperature of the air stream entering the evaporator. The condenser air entering temperature describes the heat sink to which heat gathered by the evaporator is rejected. The entering air wet bulb temperature describes the amount of moisture that must be removed as part of the cooling. If air is very moist then more of the capacity will be used to condense moisture out of the air, and less will be available for sensible cooling, which will result in a lower expected temperature split. Carrier suggests that, if the sensible temperature split is within three degrees of the target level, the airflow is about of 400 CFM/ton and adequate. In practice this is untrue. Consider that a common temperature split is about 20 F. A three degree discrepancy is 15%, which translates directly into a 15% difference in flow. This shows that, rather than being “on the order of 400 CFM/ton”, a split that is 3 F high would translate to a flow of 340 CFM/ton, assuming that the charge was correct and that the coils were clean. The temperature split method also has the problem that it is not directly measuring flow. Rather, it is using temperature measurements in conjunction with charts to impute airflow. The temperature split is essentially the result of three things: the amount of refrigerant in the evaporator coil, the airflow across the coil, and the ability of the coil to transfer heat.

6

The charts do also evaluate charge, but the results are not used in assessing airflow. Instead, if the unit is either over- or undercharged, but the airflow is said to be adequate, it is likely that correcting the charge will result in the estimated airflow no longer being adequate. In effect, the airflow assessment is relative to the charge. The charts also assume a clean evaporator,, which is often a poor assumption in these units since coil cleaning is not part of routine maintenance in the EWEB service territory. Dirt reduces the ability of the coil to transfer heat, resulting in a temperature split that is smaller than it should be for the given charge and airflow. Clearly, since the temperature split method relies on assumptions that may easily be incorrect, it cannot be depended on to provide a good quantitative assessment of airflow. This also means that no quantitative estimate of potential savings due to flow changes can be made using this method. This method can be useful at a qualitative level, however, as it will typically catch the large problems. A further problem with simply using the temperature split to evaluate airflow is that we also want to know how much air is coming in through the economizer under various situations. The temperature split does not evaluate this flow, and even if it did, the magnitude of the uncertainty in the result is such that nothing could be inferred as to the fraction of the air that was from outdoors. For this project, Ecotope utilized a new airflow measurement device, the TrueFlow™ Air Handler Flow Meter. This device was developed by Ecotope in concert with The Energy Conservatory. This device is a calibrated perforated plate that correlates a measured pressure drop across the plate with an airflow, and has been shown to be a very accurate means of directly measuring airflow (Palmiter and Francisco 2000). This device was designed for residential-size filter slots. For the larger filter slots of commercial units, several plates were often required. In these cases, the airflow through each plate is measured and the results are added together. The TrueFlow™ meter is currently being absorbed into the HVAC community and should become a relatively common piece of technology within the next several years since it substantially reduces the uncertainty of the temperature split methods and is fast and simple to use. At this point, the use of the device represents a significant addition to the testing usually done by contractors, but the reliability of the estimates warrant its use in this type of program. The protocol involves measuring both the economizer flow and air handler fan flow simultaneously (see Figs. 2-1 and 2-2 for examples of TrueFlow™ installation), to get an estimate of the percentage of outdoor air under various economizer positions. Flow meters are inserted in place of the system filters and at the inlet of the economizer. The protocol sheets provide places to indicate which readings are for economizer flows and which are for system flows, as well as what position the economizer is in. The supply and return pressures are also measured and entered, both with the flow meters installed

7

and with the filters installed, which allows for a simple correction to be made to account for any change in flow caused by the flow meters. This adjustment is typically small. Once the data has been collected, the total flow is calculated for both the economizer and the system fan, and the adjustment is applied to each to get the corrected flow. There are also spaces on the data sheets to enter the percent of outdoor air and the CFM/ton of cooling. A separate sheet is used for each economizer position tested.

Figure 2-1. TrueFlow™ Air Handler Flow Meters Figure 2-2. TrueFlow™ Air Handler Flow Meters installed in a rooftop unit filter slot. installed on an economizer.

2.2. Refrigerant Charge Refrigerant charge is a major focus of any work with air conditioning systems. The amount of refrigerant in the system being circulated by the compressor has a direct bearing on capacity and efficiency of air conditioning. Many reports have been written on the effects of too much or too little refrigerant charge on the cooling capacity and efficiency of these systems. The refrigerant charge portion of the protocol required the least development, because standard methods for checking charge are essentially the best available methods. The primary change over what contractors currently do is the use of the CheckMe!™ standardized form that requires certain information, which is called in to a toll free number. In order for contractors to use the CheckMe!™ program, which has its foundation in the Carrier method, they must go through a one-day certification class. An additional feature of this program is that it not only provides the results but also gives recommendations for how much refrigerant to add or remove, if appropriate. For refrigerant charge, this approach uses the outdoor dry bulb temperature and the return air wet bulb temperature to evaluate the sensible/latent cooling split. The results can be used with lookup tables to determine the expected increase in temperature in the refrigerant vapor line (the “superheat”) or decrease in temperature in the refrigerant liquid line (the “subcooling”, only used for systems with thermostatic expansion valves (TXVs)). The target can be compared to measured superheat or subcooling, based on pressure and temperature measurements from the refrigerant lines. A deviation from the target indicates that there is either too little or too much refrigerant.

8

There are several important qualifications for using this protocol. Major system faults (e.g., bad electrical components, very dirty indoor/outdoor coils) can invalidate the charge assessment. These qualifications were not fully understood when this project began, primarily regarding coil cleanliness. In fact, after observing the results on a few sites with fouled coils, we consulted Proctor Engineering and realized that we should have been more careful in ensuring that coils were clean before proceeding with the CheckMe!™ process. However, once this was cleared up, some confusing results started to make more sense. For example, dirty coils are less effective heat exchangers and therefore, CheckMe!™ might identify an apparent overcharge where one does not exist. In package units, unless there is an initial mistake to factory charging, there are leaky components, or the system has been "adjusted" at some time in the past, the system should be fairly close to the factory charge. The corollary of this fact is that, in our observation, technicians are very reluctant to make refrigerant charge changes to commercial rooftop units unless the change is large. During this project, in fact, even when CheckMe!™ suggested that adjustments should be made, the recommended adjustments were usually small and technicians typically did not change the charge in the system. Proctor Engineering has made improvements to CheckMe!™ since this project that incorporate the factory charge and unit size and type (commercial/residential) into the CheckMe!™ database. This information will eventually be used to refine recommended charge adjustments for packaged commercial units. An important addition to the refrigeration protocol has to do with ensuring the cleanliness of refrigeration coils before performing refrigeration charge assessment, as the Carrier method assumes clean coils. If one plunges ahead in evaluating refrigerant charge and evaporator airflow without ensuring that both the condenser and evaporator coils are relatively clean, the conclusions reached can be faulty. This is because dirty coils compromise the ability of the coils to transfer heat, which can show up as a superheat that is too low, implying too high of a charge. Especially where condenser coils are very dirty and discharge pressures are very high (over 300 lbs. psig), the compressor works much harder than needed to deliver a given amount of cooling. See Figs. 2-3 through 2-8 for examples of dirty coils and coil cleaning.

Figure 2-3. Fouled evaporator coil at bakery. Figure 2-4. Water drainage from cleaning fouled evaporator coil at bakery.

9

Figure 2-5. Fouled condenser coil at bakery. Figure 2-6. Cleaning fouled condenser coil at bakery.

Figure 2-7. Fouled evaporator coil at pet Figure 2-8. Water drainage from cleaning fouled grooming facility. condenser coil at pet grooming facility. Recording the as-found compressor running amps and after cleaning compressor running amps can be very useful in showing the value of coil cleaning, and as a result these measurements were also added to the protocol. Long-term data collection of this information before and after coil cleaning provided a first order estimate of the savings from coil cleaning in one of the units. The outdoor coil must be dry when the post-cleaning readings are taken so that the true effect of the cleaning can be assessed. (If the coil is still wet, the head pressure will be much reduced and the apparent compressor running amps will be artificially low.)

2.3. Economizers Economizers have been installed on much of the small package equipment installed over the last 15 years. Anecdotal information from many sources indicates that there are significant numbers of units that no longer operate correctly, or were never installed correctly. A study by New England Power Services Company evaluated 52 2-year-old package cooling units and found that 56% were not operating correctly. With high failure rates it is essential that procedures be developed that can quickly determine economizer functionality. With better feedback, system installation should improve and failure rates fall. The economizer portion of this project aimed broadly at five goals:

10

• determine the range of economizer types installed; • develop a generalize protocol for assessing economizer operation; • apply the protocol to assess the operational capabilities of installed economizers; • determine optimal configuration of economizer logic and setpoints; • estimate the energy impacts and savings impacts of economizer assessment and

repair. The hope was that a field checklist could be developed that typical service technicians could follow to determine the health of the economizer system. The checklist was to include characterization as well as functional tests. The functional operation verification aspect of this checklist proved to be the most challenging part of the project. The range of equipment found and the difficulty in performing some tests made the development of a generalized protocol elusive. However, significant progress has been made towards a generalized protocol. In the early stages, the economizer protocol was fairly loose and included a lot of space for descriptive detail of the system in use. The terminology that would be used by a field technician was also somewhat fluid. Some of the terminology is consistent between manufacturers, but there can be some specialized meanings and, especially in the case of the different control boards and their strategies, the possibilities for confusion can multiply. After field visits were complete (in late September), considerable effort was made to systematize references to economizer parts and function. In general, sufficient understanding had been gained by that point to tell the difference between the various types of sensors (dry bulb and enthalpy sensors) and various types of actuators (direct drive or systems controlled by microswitches). Furthermore, some progress had been made in determining how diagnostics could be performed on different systems. Some systems could be diagnosed using specified procedures from the manufacturer, while others required development of our own methodologies. The protocol still includes spaces for descriptive detail about economizers, but there are many items that have been called out for specific information. These include sensor type (dry bulb or enthalpy); sensor setting; sensor electronics (electromechanical or solid state); control logic (changeover or differential); whether it is integrated with cooling; damper orientation (vertical or horizontal); damper type (louvers or slider); and how the damper position is controlled. There are also spaces for information about the minimum air setting, including what the setting is, whether it can be adjusted, and whether it shuts down in unoccupied mode. Details about checking damper and sensor operation are also requested on the protocol. Finally, there are spaces for manufacturer and model number for the economizer itself, the controller board, the actuator, and the sensor. The non-uniformity of economizers remains a problem with the economizer protocol; there are some tests that will simply not be practical on some systems. In such cases it will simply not be known whether the economizer is working properly. There may also be cases where the economizer seems to pass the checks, but it is not operating as one

11

might expect. The time required to determine the cause may well be prohibitive in many such cases. Going to automated, higher level monitoring systems and procedures (e. g. Breuker et al. 2000) is probably not practical on many smaller package units because of costs. The protocol presented in the appendix attempts to deal with these issues but the number of units reviewed and the time constraints prevented a generalized approach. In some cases, the original equipment troubleshooting manual is available to the technician, or at least it is on the shelf at his shop, but this information does not typically make it into the field. This may be in part due to the large volume of material involved. The following subsections more fully describe many of the system components that are in the protocol, and some of the strategies for diagnosing problems.

2.3.1. Economizer Characterization Identifying the various components is key to understanding any given economizer system. The economizer system is composed of several major components: sensors, logic controller (board or otherwise), dampers, interface between the economizer and the unit controller, and the indoor thermostat (single- or dual-stage) or control system. Sensors A large variety of outdoor air sensors are used with modern economizers. Sensors will either monitor the outdoor dry bulb temperature or enthalpy, depending on the sensor selected. Solid-state sensors feed a variable resistance or voltage to the control board and usually take the form of small probes or, in the case of Honeywell, box-shaped devices. These can be either dry bulb temperature or enthalpy. Mechanical sensors include simple thermostats, often with dial controls to set the temperature at which they switch closed or open, and small snap disks, calibrated to a certain temperature. Figures 2-9 and 2-10 show examples of dry bulb sensors. Typically the outside air sensor is mounted on the economizer hood, though some units (e.g. Trane) can have the sensor in the corner post of the cabinet near the main unit control logic board. Sensors can also be placed in the return air to provide differential control, which is discussed in the following section. When this is done, the return air sensor is of the same type as the outdoor air sensor.

12

Figure 2-9. Solid state dry bulb sensor. Figure 2-10. Electromechanical snap disk. It is important that technicians be able to identify the basic sensor types in order to correctly identify the control logic being used and to aid in troubleshooting the system. The main distinctions are whether the sensor is a dry bulb or enthalpy sensor, and whether it is electrical/mechanical (snap disk, thermostat) or solid state. Control Logic Modern economizers utilize one of three decision control logics for determining whether they should operate during a cooling cycle. These are dry bulb changeover, enthalpy changeover, and differential enthalpy. All three strategies involve an outdoor air sensor that detects either dry bulb temperature or enthalpy. The differential strategy requires an additional return air sensor to determine the enthalpy of the return air. The changeover point is the temperature or enthalpy above which the economizer will no longer operate. Most economizers allow a range of setting options. An aggressive setting allows the economizer to operate more often but can lead to comfort issues depending upon how thermostat and compressor staging is handled. Conservative settings only run the economizer when outdoor conditions are cooler. When outdoor conditions are cool enough, the economizer can provide the equivalent to what the compressor would produce, and thus occupants will sense little or no difference. In northwest climates this setting has large impacts on energy savings from economizers. Older units typically have a mechanical dial thermostat on the economizer hood as part of the outdoor temperature sensor. The settings are numeric for dry bulb sensors and from A-D (A is aggressive, D conservative) for enthalpy sensors. Enthalpy changeover set points that correspond roughly to the traditional A, B, C, and D settings are, at 50% relative humidity: A—28 BTU/lb., B—25 BTU/lb., C—22 BTU/lb., D—20 BTU/lb. Most Honeywell controllers have an ABCD setting on a small dial potentiometer on the logic board no matter the sensor type. In higher-end Trane models with Trane economizer controls, the changeover point is set with a small switch block on the economizer control board. This switch block consists of two switches, providing four setting options. The options are analogous to the dial settings. For dry bulb changeover, the switch positions reflect the numeric temperature

13

setting, whereas for enthalpy changeover the positions correspond to the A-D settings. In a confusing twist in the default dry bulb mode, the relative conservativeness or aggressiveness of these settings is different than in the enthalpy mode. As a result, the technician needs to be alert to the type of outdoor sensor when setting the switches. For differential enthalpy, the decision of whether or not to use the economizer for cooling is based on comparing the return-air and outdoor enthalpies, and using whichever one is most beneficial to the extent possible. In this case, there is not an adjustable setting, since what is most important is the relative benefit of outdoor air vs. return air, not the actual outdoor conditions. Any of these control logics are typically combined with a mixed air sensor, which requires that the temperature at the system fan not be below a specified level. This prevents the unit from bringing in extremely cold outdoor air through the economizer. Many businesses need some amount of cooling even when it is cold outside, and it would be undesirable to bring in 30 F air simply because the outdoor air would make the occupants uncomfortable. In this case, a mixed air sensor would override the economizer and cause the system to use return air. Determining which logic is being used requires some familiarity with either sensors or control logic boards. Most logic boards in use (i.e., the standard Honeywell and Trane boards) come with the ability to utilize any of the three strategies identified (dry bulb changeover, enthalpy changeover, differential enthalpy). Switching between the strategies is simply a matter of connecting the appropriate sensors and jumpers. In these cases, familiarity with the sensor contacts on the logic modules allows technicians to determine which strategy is being utilized. If return air sensors are connected and functional then differential logic is being used, otherwise a changeover control is being used. The Trane Voyager units may allow dry bulb differential control, as they allow a return-air dry bulb sensor input into the board. Neither this control strategy nor the return air dry bulb input are mentioned anywhere in the installation, troubleshooting, or micro-wiring handbooks. In Honeywell controllers, the logic board used the same input channel for either enthalpy or dry bulb sensors. As such, as long as the correct sensor is used, it would be possible to do dry bulb or enthalpy differential, although literature discusses only enthalpy differential. Older economizers, which utilize electrical/mechanical relays instead of logic boards, generally use only changeover controls. One problem that is found in some economizers is that an electromechanical outdoor sensor, such as a snap disk, is used in combination with a controller that allows an adjustable setting. In some cases the electromechanical sensor is installed by contractors when the existing sensor fails, and sometimes the systems come from the manufacturer this way. The low cost of these sensors is likely the primary cause of their installation. While the adjustable changeover setting is frequently set to a specific setting, which does not then behave differently than a snap disk, it does defeat the possibility of selecting a

14

different setting to improve the performance of the economizer. In these cases the economizer provides little savings but is not malfunctioning. Dampers Economizers use one or two dampers to control outdoor airflow, and one to control return airflow. The “economizer” damper generally refers to the intake outdoor air damper. It works in concert with the return air damper to vary the mix of outdoor and return air. When the economizer damper is fully open the return damper closes off the return. In most systems there is also a relief air damper that allows indoor air to escape when the economizer introduces large amounts of outdoor air. Economizer and return dampers are either louvers or sliders. Louvers are more common, though many Carrier units from the early-mid 1990’s have sliders. Louvers are typically oriented horizontally, whereas sliders can be either horizontal or vertical. The dampers are usually controlled by the same linkage. In some cases, it is the same damper that controls both air flows, such that by opening up the economizer it is covering over the return. The key to economizer operation is the ability to operate the damper from minimum outdoor air position to fully open. In most modern units, the damper actuator itself has a feedback potentiometer with a variable resistance so that the unit can sense the position of the damper. The sensing device is integral to the actuator in these cases. The standard Honeywell actuator is an example of this. Older style units utilized various methods of mechanical switching to determine fully open, fully closed, and minimum air stops. Some Carrier and Micrometl® units use a set of four microswitches along a track to sense damper position. Thermostat Staging Thermostat staging has important consequences to successful economizer operation. In small equipment it is not uncommon to set thermostats up to provide only a single stage. Without a second stage, economizer operation will lock out the compressor. To ensure adequate cooling, a conservative changeover setting (e.g., 55 F or the lowest enthalpy setting) must be used in these cases. Determining the number of cooling stages in a thermostat requires specific knowledge of the thermostat being used, or inspection of the wires coming from the thermostat to the unit. Many thermostats with two or three cooling stages can be setup for one or two stage cooling. This setup must be examined to verify the actual number of stages available. Multistage cooling has implications for how much savings can be garnered from integration of economizer operation with compressor operation. For equipment in the size range investigated in this study (5-20 nominal cooling tons), it is unlikely modulating (variable speed) compressors would be found. Instead, the amount of cooling capacity is determined by the number of compressors in the unit and how they are staged based on cooling requirements in the space and the type of thermostat installed. The scope of this

15

study was such that this issue was not investigated in detail in the field. However, it has been addressed in the DOE-2.1e modeling presented in Section 4. Fully integrated operation is modeled; this configuration allows varying combinations of outside air and compressor cooling to achieve the most efficient combination.

2.3.2. Generalized Sequence of Operation Most economizer logic modules built in the last decade allow economizer operation during cooling. Three generalized sequences of operations describe a majority of small economizer installations. Single-compressor units controlled by single-stage cooling thermostats operate simply. A signal for cooling is passed to the economizer module. The economizer module determines whether conditions are appropriate for economizer operation. If so then the economizer operates, otherwise the compressor operates. At no time will the economizer and compressor operate together. In addition, the compressor will not start up when outdoor temperature satisfies the economizer logic but the cooling delivered by the economizer is inadequate for the space. Thus it is important to use a conservative changeover setpoint in cases with a single stage thermostat. For single-compressor units controlled by thermostat allowing two stages of cooling, the sequence of operation is as follows: when the thermostat calls for stage-one cooling, the call is passed to the economizer control, which decides whether the conditions are appropriate for economizer operation. If so, the economizer runs until either stage-one cooling is satisfied or there is a signal for stage-two cooling. A stage two cooling signal is generated by the thermostat. Various thermostats use temperature rise in the space, cycle time with stage one, or rate of change of air temperature to determine the need for stage two cooling. In the event that there is a stage two cooling signal, the compressor is activated. Most units are set up in a configuration where the economizer can remain open during compressor operation. However, economizer position is usually modulated to maintain a discharge air temperature, typically somewhere between 50 – 55 F. With the compressor operating, discharge air is typically close to this range and the economizer dampers close. Thus, with a single-stage compressor, economizers and compressors have limited abilities to operate simultaneously. The sequence of operation for units with two compressors and two-stage thermostats is different. The stage one cooling signal is sent to the economizer; it operates if it can, otherwise the compressor operates. If stage two cooling is requested and the economizer is stage one, then the economizer and the first compressor are operated. If stage two is requested and the first compressor is stage one, then the first and second compressors are operated. With this sequence, in conditions where the economizer is operating, the two compressors do not operate together even with a stage two signal. Because only one of the compressors is on, the economizer can successfully operate at the same as the compressor because the discharge air temperature is elevated, since only half of the cooling is being delivered by the compressor.

16

2.3.3. Economizer Functional Testing

A complete checkout of the economizer must verify the following items: • Damper closes • Damper opens 100% • Minimum air setting is reasonable • Settings for control strategy are appropriate • Damper opens and modulates to maintain discharge air temperature when conditions

warrant; this includes checking proper performance of sensors • Damper stays at minimum air when conditions are not appropriate. Economizer checkout can be tricky due to the variation in economizer types and to the number of modes in which the economizer can and must operate. When the unit is in heating or fan-only mode with no calls for cooling, the economizer should be at the setting for minimum outdoor air when the fan is running. This setting is for providing adequate ventilation. When there is a call for cooling, and conditions are acceptable to the economizer logic, the economizer should be partially or fully open. Most economizers will close the dampers completely when the fan is off. Most rooftop equipment economizers come with instructions for economizer checkout. This is generally far beyond what field technicians check, though often it does not check all possible error conditions. We acquired the troubleshooting protocols for two common economizers and found that none of the technicians had seen or used this literature. This is interesting since, for one type of unit, the same troubleshooting steps were included in the installation instructions. When technicians do assess economizer operation, they typically only check whether the dampers are mechanically functional (i.e., whether the dampers open and close), but not whether the economizer dampers change position at the right outdoor air conditions or whether the damper position is correct. To fully diagnose the operation of an economizer requires evaluation of the sensor operation, logic controller operation, mechanical operation, and the minimum air setting. In addition, there are some units that open fully whenever the air handler fan is on, which means that the unit is using a lot of outdoor air regardless of the outdoor conditions. This may happen for a variety of reasons, such as incorrect minimum air setting or dampers that are installed incorrectly and are not able to close down. It is best to use the manufacturer’s checkout procedure to evaluate the functionality of the dampers, if possible. There are a number of other methods used for assessing whether the dampers are mechanically functional. Some units have a minimum air potentiometer on the control board, which can be turned with a small screwdriver. Some units have a “test mode”, whereby the function of each operational mode of the HVAC unit can be checked sequentially regardless of outdoor conditions. Another method is to adjust the changeover setting to cause the logic controller to decide economizer operation is appropriate.

17

There are drawbacks to each of these methods. Adjusting the minimum air potentiometer to determine whether the economizer damper works does not evaluate the operation of the sensors, or the control logic. If the unit is in economizer mode, it will not respond to the minimum potentiometer. Also, depending on how the unit is setup, this may only demonstrate part of the economizer motor’s range. Finally, in some units with more elaborate digital control systems, minimum outside air may be determined from the main computer system rather than at the unit. For units with a test mode, utilizing this feature does enable the technician to run the economizer through minimum and full open positions, without needing to worry about whether the economizer is already being used for cooling. This does not, however, check the operation of the sensors or control logic. Adjusting the changeover setting can both check damper operation and, to some extent, sensor operation. However, it requires that the outdoor conditions be within the range of available changeover settings, and that the thermostat be calling for cooling. Also, if a snap disk has been installed, adjusting the changeover setting will have no effect. Mixed air sensors can also defeat this method if the outdoor air is too cold. Units that have differential control also will not respond to this type of adjustment. Using a jumper to send a signal to the economizer to open only checks the mechanical operation, not the sensor functionality. It also requires that the economizer not be providing cooling at the time. There are methods that can evaluate economizer operation much more thoroughly. Honeywell economizers can be tested with a resistance test kit. This kit comes with a number of resistors that can be plugged onto the control board to artificially send signals telling the economizer to open or close. These resistors do this by emulating specific temperatures. Using these kits not only check mechanical operation of the damper, but also whether the control board interprets the temperature signal appropriately. The sensors, however, are not evaluated with this technique, and this method may not catch all possible failures. It is, however, significantly more in depth than what technicians currently do. The Trane Voyager systems have a checkout procedure that can reliably check damper operability and, unlike the Honeywell systems, sensor operation. However, the procedure does not necessarily determine whether the control board interprets the sensor signal correctly. Obviously, a more systematic approach is needed to fully assess economizer operations. Ensuring damper operation only verifies two of the six items listed at the beginning of this section. A goal of this project was to develop a protocol that addressed all of these areas for a large range of economizer types. This turned out to be unrealistic, due to the wide variation in economizers, their components, and the diagnostics required at each type to evaluate the economizer function completely. However, the protocol does enable more detailed evaluation to be made of each type of economizer encountered than was

18

previously done, and for some types the protocol can be used to provide a fairly complete analysis of the economizer performance.

2.3.4. Sensor Operation It is important that the outdoor and discharge air sensors be checked for correct readings. Temperature and enthalpy sensors can be checked in a variety of ways depending on the sensor type. Solid-state sensors can be checked if one has access to the resistance chart for the sensor. The Trane Voyager includes this information for the outdoor air and the mixed air dry bulb sensors. With the unit operating, the voltage across the terminal inputs on the main unit board can be checked for the outdoor air temperature, and the mixed air sensor can be checked by measuring the voltage across its input to the economizer board. This value can then be compared with temperature or dry bulb measurements made by the technician. Trane was unable to provide similar values for the enthalpy sensor resistance tables. The installation guides for several Honeywell units include resistance charts for dry bulb and enthalpy sensors, which can be checked in a similar way. Snap disk sensors are a little more difficult. If the outdoor air temperature is cooler, warming the temperature should cause the snap disk to snap. This can be felt or checked with an ohmmeter. In warmer conditions, the sensor would need to be cooled down, perhaps with ice. Reportedly, failures of these switches occur not so much as changes in calibration but simple failures to operate at all. Mixed air sensors are more difficult to check, and just as important to successful economizer operation. Again, these can be resistance or on-off devices. We have had very little success in finding resistance charts for these sensors. Some units use two snap disks set at 52 F and 57 F to control mixed air temperature. These can be checked as above. Electrical/mechanical switches and outdoor sensors with a dial for setting the changeover temperature can be checked by changing the position of the dial and feeling when the sensor clicks; if it clicks close to the correct temperature, the sensor is within the bounds of checkable functionality. This requires, of course, that the outdoor conditions be within the range of settings available for the sensor.

2.3.5. Minimum Outside Air

Minimum outside air is used to establish the fresh air intake to the space. Requirements for outside air are determined based on the use of the space and the number of people occupying it. It appears that standard practices in units of this size do not take this information into account, but merely set the unit to some approximate position for a minimal setting. In some situations, however, this position may be set based on some detailed engineering and test-and-balance procedures.

19

In evaluating the airflow through the economizer, it is probably best to integrate the economizer and TrueFlow™ tests. Flows with the economizer damper in minimum position are of primary interest. This is easily done when the unit is not in cooling mode. In conditions where cooling signals are coming from the thermostat, the unit must be put into minimum air position: either the call for cooling needs to be removed, the changeover set point reduced to a point that the economizer does not work, or more elaborate procedures employed to defeat the economizer. The economizer can be reliably put into its minimum air setting in several different ways depending on the unit. Trane units with the test board in the unit controls can be put into minimum position in its test mode. For other units, disconnecting the thermostat wires that call for cooling should put the unit into a minimum outside air position. An alternate approach is to override or adjust the change over setting so that the economizer thinks the outside air is unsuitable. This has the advantage of making the thermostat calls unimportant. Adjusting the change over requires that the outdoor conditions be within the range of the available changeover settings. It also does not work for snap disks or for the differential enthalpy control logic. Overriding the signal by disconnecting the outdoor air sensor is a common solution. Units utilizing the standard Honeywell logic boards can be put in the minimum air position by disconnecting the mixed air sensor from the board. This essentially tells the unit that it has a very low temperature, which shuts the damper. Other Honeywell controllers may be slightly different. In general, the installation instructions have a description of setting the minimum position adjustment reliably.

2.4. Duct Losses Duct systems in commercial buildings were to be evaluated as part of the protocol. Although tools are available to measure commercial-size flows, in this set of buildings it was generally impractical to measure flows because of ceiling heights, diffuser types, or occupancy issues. In some cases measuring flows was unimportant because ducts were running more or less entirely inside the heated space, and any duct losses would be confined to the conditioned space. As a result, the protocol for duct losses became a description of where the ducts were located (e.g., inside conditioned space, in buffer spaces such as dropped ceilings, or completely outside). For ducts located in buffer spaces, the protocol contains spaces for information relating to the placement of insulation with respect to the ducts and the degree of venting of the buffer space to outdoors. There is also a section of the protocol for entering measured register and grille flows, though this section was not completed in any of the tests performed to date.

2.5. Thermostats

20

The type of thermostat installed was primarily of interest to characterize the situation under which the rooftop unit was required to operate (scheduling). In addition, it was of interest to see how frequently thermostats were of the two-stage variety rather than having only a single stage. Thermostat setpoints and schedules are requested in the protocol, including what hours and days are occupied vs. unoccupied, and whether the system fan was set to run continuously during occupied hours. 3. Results This section discusses what was learned about the systems tested in this study. There is also a discussion in Section 4 of the results of computer simulations of the effects of different strategies for the economizer.

3.1. Evaporator Airflow Of the 30 units examined, airflow was measured in 27 with the TrueFlow™ meter. The results are shown in Table 3-1. The average flow was 304 CFM/ton. The measurement was done with the economizer in either a minimum air position or closed. Further discussion of economizer airflow is found in Section 3.3. The total air flow ranged from 99 to 429 CFM/ton. About two-thirds of the units had airflow less than 350 CFM/ton. It is interesting to note that the average was very close to what other researchers have found looking at larger sets of randomly selected data, mostly from residential studies. The level at which low airflow begins to have a significant effect on performance varies from climate to climate, but 320 CFM/ton is commonly considered low enough to warrant modification. Repairs were attempted on several systems, with partial success depending on the evaporator fan control algorithms. In some cases, DIP switches were changed in an effort to increase airflow, though the flow usually only increased by about 50 CFM when this was done (10 CFM/ton or less). In other cases, notes were made with the HVAC contractor to change air handler flow rates by adjusting the pulley sheaves on belt drive fans. These can be relatively straightforward repairs, but they are often neglected at the time of installation; factory settings are assumed to be accurate and careful review is not undertaken. This is in large part because of the lack of direct measurement equipment available until very recently.

21

Table 3-1. System Fan Flow Results Flow Business/Unit Nominal Tons1 (CFM) (CFM/ton) CheckMe!™Shopping common area 5/5 1985 198 -- Restaurant 5 2008 402 Low Bakery 7.5 1915 255 Low Florist 5 1288 258 Low Pizza Shop 4 1047 262 -- Tanning salon - unit 1 5 1675 335 OK Tanning salon - unit 2 3 584 195 Low Tanning salon - unit 3 5 959 192 Low Research facility - unit B-8 6 1990 332 OK Drug store - unit C 4 1715 429 OK Drug store - unit D 5 2010 402 OK Pet grooming facility 6 2037 340 OK Pet veterinarian 4.5/3 2639 352 -- Pet store - unit 8 3 1057 352 OK Eugene Airport lower roof 10/5 1489 99 -- Eugene Airport upper roof 7.5 2712 362 -- Eugene School District 2.5 777 311 -- Office building A - 2.5 ton 2.5 767 307 -- Office building A - 4 ton 4 1630 408 OK Office building A - 8.5 ton 4.5/4 2685 316 -- Office building B - 7.5 ton 7.5 2327 310 -- Office building B - 15 ton 15 5480 365 -- Travel agency - unit 1 7.5 1978 264 -- Travel agency - unit 2 7.5 1770 236 -- Travel agency - unit 3 7.5 1850 247 -- Athletic club 12.5 4492 359 -- Office building C 5 1610 322 OK Average 6.6 1943 304

1. Two numbers indicate sizes of first and second stages in dual compressor systems.

For these cases, flow expressed in CFM/ton is based on the total available cooling. None of the units tested had a higher speed for the second stage.In some cases repairs were not possible because the apparent cause of the low flow was undersized duct work. Two of the lowest air flows (pizza shop and airport lower roof) occurred in units with the highest external static pressure across the fan. An additional unit with low airflow (the bakery) had a very high supply pressure, indicating a significant restriction on the supply side of the system. Two others with very low flow (tanning salon unit 2 and florist) did not have exceptionally high external static pressures, but had very low return pressures. The unit at the tanning salon had only 9 Pa on the return side but 64 Pa on the supply side, suggesting that there was a blockage on the supply side that the small fan was not able to

22

deal with. The unit at the florist had neither high supply nor high return pressures, and may have simply been the wrong fan for the system. There were 12 cases in which the Carrier and TrueFlow™ methods were both used to measure evaporator airflow. In the seven cases where the Carrier method suggests flow is adequate (i.e., the measured sensible temperature split is within plus or minus three degrees of the target), the average TrueFlow™ measurement is 371 CFM/ton. Six of these cases are initial tests before any changes have been made to the system of either charge or airflow. In another case, the condenser and evaporator coils were cleaned thoroughly before the TrueFlow™ measurement was taken. In four cases, the Carrier method called out the system as having low evaporator airflow before any adjustments were made to charge or the exchanger coils. In these cases, the average airflow as measured by TrueFlow™ was 261 CFM. This is distorted by the restaurant unit, which had a flow of about 400 CFM/ton even though the Carrier method said the flow was low. The next highest was 258 CFM/ton, with two others below 200 CFM/ton. The final case, at the bakery, was said to have adequate airflow by the Carrier method in the initial run. However, the evaporator was extremely dirty, and after cleaning, the temperature split went up sufficiently that the Carrier method then correctly stated that the flow was low. This shows how a dirty evaporator can cause the Carrier method to return an incorrect qualitative assessment of charge. These results do suggest that the Carrier method is helpful in a qualitative way, though it is not exact. Four of the cases where the Carrier method suggested flow that was adequate actually had flows of only 330-352 CFM/ton, which agrees with some of Ecotope’s earlier measurements, where the Carrier method gave an indication of adequate airflow where flows were actually only within 300 – 350 CFM/ton range.

3.2. Refrigerant Charge As shown in Table 3-2, of the 30 units in the study, 14 had successful CheckMe!™ runs, in which a conclusion was reached on whether the system was properly charged or needed an adjustment. It should be noted that the recommendations for charge modifications are only recommendations. Although they are based on the results of other recommendations and changes (as detailed in the CheckMe!™ database), the individual cases of recommendations can sometimes be inaccurate. Therefore, the technicians are urged to use their best judgment in how they adjust these systems. This is especially true with rooftops units, which often have a relatively small amount of charge, and subtle alterations in refrigerant charge can result in relatively large swings in measured superheat. Eleven of the 14 units used fixed metering devices and therefore system superheat was used to evaluate charge; the other three used TXVs and therefore were evaluated on system subcooling. Of the 14 successful CheckMe!™ runs, 7 initially returned results of overcharge, with the average recommendation being removal of 8 ounces of refrigerant. In two of these cases,

23

the bakery and the pet grooming facility, the apparent overcharge was due to dirty coils; when the coils were cleaned, CheckMe!™ returned a result of proper charge. The table reflects the results after cleaning; prior to cleaning the recommendations were to remove 8-10 ounces. This illustrates clearly the importance of cleaning coils prior to performing CheckMe!™. In the other five cases of overcharge, the recommended adjustment was at or near the minimum adjustment that CheckMe!™ will return, and the technicians were reluctant to make any changes in the system charge, so the system was left as found. In four cases (29%), the system was said to be undercharged by CheckMe!™. In one of these cases, the system had a very severe leak in an after-market pressure controller. The controller was replaced and three pounds of refrigerant were added. After this change was made, the system was still undercharged (based on a measured superheat significantly higher than the target superheat), but it was left at this point since the measured airflow was about 200 CFM/ton (because of an extremely undersized return) and the technician was concerned about making further charge adjustments. Refrigerant was also added in unit D at the drug store. This unit was found to be about 20 ounces low, and 19 ounces were added. This shortage is fairly significant, and the contractor agreed with the CheckMe!™ assessment. After the refrigerant was added CheckMe!™ said that the charge was fine CheckMe!™ returned a result of proper charge on the initial test for the remaining three systems which, when added to the two systems that were said to be fine after coil cleaning, provided a total of five systems out of fourteen (36%) which were said to be properly charged. The occurrence in this very small group of units of proper charge (36%), undercharge (29%) and overcharge (36%) is very similar to results found in much larger studies of residential air conditioning systems. A dataset of approximately 3800 CheckMe runs on commercial units (Proctor Engineering Group 2002) found frequencies of 39% (proper charge), 38% (undercharge), and 23% (overcharge).

24

Table 3-2. Refrigerant Test Results Superheat (Subcooling)1 (F) Adjustment (oz.)2

Business/Unit Target Measured CheckMe!™ Made Restaurant 19 4 -14 0 Bakery 18 20 0 0 Florist (15) (22) -6 0 Tanning salon - unit 1 (15) (10) 6 0 Tanning salon - unit 2 10 3 -7 0 Tanning salon - unit 3 10 68 --3 48 Research facility - unit B-8 20 19 0 0 Drug store - unit C 13 6 -7 0 Drug store - unit D 10 29 20 19 Pet grooming facility 13 11 0 0 Pet store - unit 8 17 13 0 0 Office building A - 4 ton 3 7 6 0 Consulting firm (12) (15) 0 0 Office Building C 14 8 -6 0 1. Italicized numbers in parentheses are subcooling, only used for TXV systems. 2. Negative numbers indicate removal of refrigerant, positive numbers indicate addition of refrigerant. 3. Beyond some point, CheckMe! does not give specific recommendations, only that a major change is

needed.

3.2.1. Coil Cleaning Dirty condenser and evaporator coils were found to be a major issue in both the diagnosing of charge and as a potential for savings. It is also a major factor if airflow is assessed using the Carrier method. Dirty coils inhibit heat transfer at the coils, which both reduces the temperature change of the air across the evaporator and causes the compressor to work harder to reject heat at the condenser, reducing both capacity (overall cooling) and efficiency. In two units, CheckMe!™ initially returned an assessment of too much refrigerant and correct airflow. After doing nothing more than cleaning the coils at these two units, CheckMe!™ returned an assessment of correct charge and low airflow. The airflow had appeared to be correct given that the charge was too high, but once the charge was found to be correct the flow was no longer appropriate. Changing the charge would have been a mistake, since if the coils had subsequently been cleaned the charge would have been too low. Coils will continue to get dirtier, which will continue to alter the assessment of charge and airflow evaluation if cleaning is never performed. Evaporator coils usually do not get fouled very quickly unless the return air is highly contaminated, as was the case at these two units (bakery and the pet grooming facility). The bread baking process caused a large amount of yeast and other baking products to collect on the evaporator coil, and the evaporator coil at the pet grooming facility was covered with pet hair (see Figs. 2-3 and 2-7).

25

Condenser coils can get fouled much more easily, though the fouling can be more difficult to spot. In the Eugene area, a lot of this is due to the large number of cottonwood trees. The cotton can collect in the condenser coils and, especially in units with two coil layers, can get inside such that it is not obvious without further inspection. We monitored compressor amp draw at the pet grooming facility both before and after both coils were cleaned. The data is shown in Fig. 3-1. This graph shows the relationship of outdoor temperature and compressor amp draw before and after cleaning. Both coils were fully dry when the “after” readings were taken.

Compressor amp draw

4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

Out

door

Tem

pera

ture

, F

60

65

70

75

80

85

90

Before coil cleaningAfter coil cleaning

Figure 3-1. Relationship between outdoor temperature and compressor amp draw before and after coil cleaning. An analysis of the data suggests that simply cleaning the coils caused a reduction of about 6% in amp draw for this unit at typical cooling conditions. Other than fan energy, this corresponds directly to a 6% energy savings from doing nothing more than coil cleaning. Another way to interpret the graph is that the compressor can handle about 4-5 F greater at the same amp draw. The fact that service contracts do not tend to include coil cleaning would appear to result in a significant lost opportunity for energy savings. It is not fair to say that this magnitude of savings would result from coil cleaning at every unit, and it is not known how long it takes for coils to get sufficiently fouled to provide this opportunity. It is fair to say that most coils will get fouled over time unless cleaning becomes commonplace.

26

3.3. Economizers Evaluation of the unit economizers was a major portion of this project. This included characterization of the components as well as whether and how they worked. The wide variety of components, methods of operation, and mechanisms of failure preclude many general conclusions to be made about economizers, but much useful information was gathered regarding what types of systems and failures are found in these systems.

3.3.1. Field Results Table 3-3 summarizes the operational issues found with the economizers. Since sensor calibrations were not checked this potential fault is not included in the table. The table also does not include changeover that occurs under the wrong conditions; while it is true that no problems were found of this nature, many units were not checked for this problem. Table 3-3. Summary of Diagnosed Economizer Operational Issues Business/Unit Broken

damper control

Damper installed

open

Off/Low change-

over

Damper opens to full

with compressor operation.

Remote damper control

Shopping common area Restaurant √ Bakery Florist Tanning salon - unit 1 √ Pizza shop √ Research facility - unit B-8 Drug store - unit C Drug store - unit D Pet grooming facility √ Pet veterinarian √ Pet store - unit 8 √ Eugene Airport lower roof Eugene School District Office building A - 2.5 ton √ Office building A - 8.5 ton √ Office building B - 7.5 ton Office building B - 15 ton Travel agency - unit 1 √ Travel agency - unit 2 √ Travel agency - unit 3 √ Shopping mall – unit 27 √ Shopping mall – unit 10 √ Total 3 2 3 2 3

27

Damper and sensor information was collected on 22 economizers. These results are tabulated in Tables 3-4 for dampers and 3-5 for sensors. Table 3-4. Damper Configuration Business/Unit Type Position Control Shopping common area Louvers Actuator Restaurant Louvers Actuator Bakery Louvers Actuator Florist Louvers Actuator Tanning salon - unit 1 Louvers Actuator Research facility - unit B-8 Louvers Actuator Drug store - unit C Louvers Actuator Drug store - unit D Louvers Actuator Pet grooming facility Louvers Actuator Pet veterinarian Louvers Actuator Pet store - unit 8 Louvers Actuator Eugene Airport lower roof Louvers Actuator Eugene School District Louvers Actuator Office building A - 2.5 ton Louvers Actuator Office building A - 8.5 ton Louvers Actuator Office building B - 7.5 ton Louvers Actuator Office building B - 15 ton Louvers Actuator Travel agency - unit 1 Slider Microswitches Travel agency - unit 2 Slider Microswitches Travel agency - unit 3 Slider Microswitches Shopping mall - unit 27 Slider Microswitches Shopping mall - unit 10 Louvers Actuator

28

Table 3-5. Outdoor Sensor Characterization Business/Unit Type1 Technology2 Orig. Setting Final Setting Shopping common area DB SS 75 F 60 F Restaurant DB SS Off 60 F Bakery DB SS 60 F 60 F Florist DB SS Tanning salon - unit 1 h SS Research facility - unit B-8 DB SS 55 F 55 F Drug store - unit C h SS C C Drug store - unit D h SS C/D C/D Pet grooming facility DB SS 60 F 60 F Pet veterinarian DB SS Pet store - unit 8 DB SS 60 F 60 F Eugene Airport lower roof h EM B/C B/C Eugene School District h EM C C Office building A - 2.5 ton DB EM Office building A - 8.5 ton h SS C/D B/C Office building B - 7.5 ton DB SS Office building B - 15 ton h SS Travel agency - unit 1 DB EM Travel agency - unit 2 DB EM Travel agency - unit 3 DB EM Shopping mall – unit 27 DB EM 40 F 57 F Shopping mall – unit 10 DB EM Off ~70 F 1. Dry bulb (DB) or enthalpy (h) 2. Solid state (SS) or electromechanical (EM) Table 3-4 shows that the vast majority of the systems tested used louvers and actuators. Only four units had slider dampers. All of these units were of the Carrier Weathermaker type, and used microswitches for position control. Three of these were found to have broken microswitches that caused the dampers to be jammed. Of these three, which were all at the travel agency in the Puget Sound region, one had the damper fixed in a position that would allow about 20% of the air to come from outside to provide ventilation, while the other two were fixed closed. One of the units with louvers, tanning salon unit 1, was found to have been installed with the dampers fixed in a fully open position. This had the effect of having about ¾ of the system flow come from outdoors in all cases, regardless of outdoor conditions. The contractor fixed this problem. The pizza shop had this problem as well, though this unit is not listed in the table above since it did not have the information that goes into the table. However, flow testing was done on the unit and it was found to have the dampers set to provide 50% outdoor air as a minimum; this was changed by the contractor. No other problems were identified as being specifically a damper or actuator problem in the units tested.

29

Table 3-5 shows that, of the 22 economizers, 15 used dry bulb changeover controls. The remaining 7 used enthalpy changeover controls. None used differential enthalpy. Some of the units that have dry bulb controls had been originally installed with enthalpy controls, but when the sensor failed it was replaced with a dry bulb sensor. This is common practice, in part because the low latent cooling loads in the Eugene climate do not offer much benefit from using enthalpy sensors. On eight of the dry bulb units, changeover setpoint data was recorded. In two of these eight units the economizer was set to off, meaning it would never be used for cooling. In another the setpoint was set to 40 F. While this does not preclude using the economizer for cooling, the number of hours in which it may be used for cooling will be exceedingly small. The setpoint at these units was changed to be 60 F (for one of the two that had been off), 57 F (for the one that had been set to 40 F), or about 70 F (for unit 10 at the shopping mall). The two set to 60 F had switch blocks that allowed only 55 F, 60 F, or 65 F. The one set to about 70 F had an ABCD dial that was set to A. It should be noted that at this unit the 70 F setting is from the snap disk, and it is unknown whether the snap disk setting is actually 70 F. Setting the dial to A would have different interpretations if the sensor was a solid state device sending a resistance to the control board. Of the remaining four units with recorded dry bulb changeover settings, two were at the pet grooming facility and unit 8 of the pet store, which were at the same building. At this building the changeover did not appear to have any effect at any of the units tested, although the dampers all were found to be functional. Thermostat settings for this building were set at a central headquarters in Phoenix, AZ, and it is likely that the economizer settings are also programmed from this remote location. It appears that the settings at the unit are irrelevant and are being bypassed remotely. The remaining three units with recorded dry bulb changeover setpoints, the bakery, shopping common area, and unit B-8 of the research facility, all appeared to be working. This does not mean that we were able to verify that there were no problems, but none of the diagnostics that were performed revealed any fault. It is possible that the sensor calibrations were off. Of the seven units with dry bulb changeover but no recorded setpoints, three were the travel agency units that had the broken microswitches, causing the economizers to be inoperable. One of the remaining four was the pet veterinarian, which was at the same pet store with the apparent programming from Phoenix. The 2.5 ton unit at office building A did not have a damper problem, in that the damper was fully functional, but the damper went to fully open when the compressor came on. Replacement of the control board did not fix the problem. The cause of this problem has not been determined. That leaves two units without recorded dry bulb changeover settings that did not show any specific problem (the florist and the 7.5 ton unit at office building B). Of the seven economizers with enthalpy control, five did not show any specific problems. One of the remaining two was the unit with the louvers installed fully open. The final unit was the Eugene Airport administration building, which did not have an obvious way

30

to assess the functionality of the economizer when the outdoor conditions were outside of its range, which was the case during the site visit. Only one of the enthalpy control units had a setting changed. This was the 8.5 ton unit at office building A, which had been set between C and D and was changed to be between B and C. Economizer Airflow Airflow measurements through the economizer were made at 17 units in up to 3 different economizer configurations: economizer closed, economizer at minimum position, and economizer fully open. Measurements were made with the TrueFlow™ Air Handler Flow Meter. Measurements with the economizer closed were intended primarily to determine whether there was significant bypass around the dampers. In general, when economizers were closed there was no measurable flow through the economizer. Of much more interest was the fraction of outdoor air with the economizer at minimum position and with the economizer fully open (i.e., economizer being used for cooling). In addition to measuring the fraction of outdoor air in each of these situations, we wanted to investigate whether there was a significant change in system flow when the economizer was fully open. In cases where the flow rate does change significantly, a likely cause is that the duct resistance is substantially different from the resistance of the economizer. Table 3-6 shows the results of this testing. The flow rates under each configuration are total system flow, with economizer flow represented as a percentage of the system flow. The final column shows the ratio of the system flow with the economizer fully open to the flow with the economizer at minimum position. The table shows that, on average, minimum position for 16 units provided about 20% of the total airflow. This corresponds to an average minimum economizer flow of about 320 CFM. However, while there is cluster at the 10-25% range, there is a wide variation among the units. At four units, minimum position was closed. At two units, both at office building A, the economizer allowed in a very large amount of outdoor air during air handler operation. At the 2.5 ton unit at office building A, the damper went to fully open whenever the fan was on, regardless of any need for cooling. Replacing the control board did not fix the problem, and no fault was found with the dampers themselves, so it remains unknown as to why this unit exhibited this behavior. At the 8.5 ton unit at office building A, the unit is set to deliver a minimum of about 54%. The building manager stated that the flow was set this way intentionally, so despite the high minimum air flow rate this cannot be considered a malfunction. Two units, tanning salon unit 1 and the pizza shop, were found upon arrival to have dampers that were open about halfway when the fan was on. These incorrect damper positions were changed prior to testing, however, so no quantitative results of the as-found minimum setting are available.

31

On average, the units obtained about two-thirds of their air from outside when the economizers were fully open. This does not necessarily mean that the units could ever go to fully open under normal operation, only that if the dampers were fully open they could only deliver the stated portion of the total air flow. The vast majority of the units delivered in excess of 50% outdoor air with the economizers fully open, with only the restaurant and the 15 ton unit at office building B being below that level. About half of the units got 60-80% of their air from outdoors when the economizer was fully open, and only three had outdoor air over 90%. This is common, as many units are not designed to actually shut off the connection to the return ducts completely. The units that had over 90% probably were designed to shut down the return completely, and the portion of the air that was not from the economizer was likely bypass around the damper. For the most part, system airflow did not change much when the economizer opened, usually less than 5%. There were only four units where measured flows changed by more than 5% between minimum and full-open economizer positions, and only one of these was greater than 8%. At the remaining unit, tanning salon unit 1, the measured flow changed by almost 20%, with the flow being significantly lower with the economizer fully open. It is unknown at this time why the flow changed so dramatically at this unit.

32

Table 3-6. Economizer Airflow Results Minimum Fully Open System

Flow System

(CFM) Outdoor

(%) System (CFM)

Outdoor (%)

Open / Min.

Shopping common area 1985 13.9 2137 62.6 1.08 Restaurant 2008 0 2098 26.2 1.04 Bakery 1915 11.2 1995 53.3 1.04 Florist 1288 18.2 1342 63.5 1.04 Tanning salon - unit 11 1675 20.4 1360 73.4 0.81 Pizza shop1 -- -- 1075 74.5 -- Resrach facility - unit B-8 1990 17.1 1996 78.9 1.00 Drug store - unit C 1715 0 1589 92.7 0.93 Drug store - unit D 2010 0 1804 93.6 0.90 Pet grooming facility 2037 22.6 2138 62.0 1.05 Pet veterinarian 2639 0 2494 61.1 0.95 Pet store - unit 8 1057 31.9 1067 55.5 1.01 Eugene School District 777 17.8 823 64.3 1.06 Office building A - 2.5 ton 767 91.9 767 91.9 -- Office building A - 8.5 ton 2685 54.2 -- -- -- Office building B - 7.5 ton 2327 9.2 2327 50.9 1 Office building B - 15 ton -- -- 5480 39.9 -- Travel agency - unit 1 1978 21.6 -- -- -- Average 1803 20.6 1839 65.3 0.99 1. These units initially had dampers that remained open all of the time. The measurements were done after this was fixed. Figure 3-2 shows how data from the pet grooming facility can be used to estimate the amount of outdoor air that comes in through the economizer in minimum position. The fraction of outdoor air was calculated using the outdoor temperature, the return temperature, and the mixed air temperature, as recorded using the MDLs. The compressor was deemed to be on when the amp draw was four amps or greater. It is unclear why there are a few points in the middle of the night when the compressor is said to have come on. The data in this graph covers about a week of time. The portions of the graph that show very large oscillations reflect times when the sir handler fan is off, and there really is not any mixed air. The periods with the compressor on can also not be used for reliable estimation of outdoor air fraction, because solar effects may influence the outdoor temperature and the operation of the coil may cause the mixed air temperature to read low; either of these could cause the outdoor air fraction to be underestimated. In order to estimate outdoor air fraction a period of time when the air handler fan operates without the compressor is needed. Such a time occurs in this data between about 3:30 a.m. and 6:30 a.m. This period suggests that the outdoor air is about 20% of the total, perhaps slightly more. This agrees very well with the estimate from the TrueFlow™ Air Handler Flow Meter,

33

which estimated 22.6% outdoor air. This is represented by the horizontal line across the graph.

Hour of Day

0 3 6 9 12 15 18 21 24

Frac

tion

of O

utdo

or A

ir

0.0

0.2

0.4

0.6

0.8

1.0

.226

Compressor on

Figure 3-2. Outdoor air fraction at the pet grooming facility, based on MDL data. The results agree very well with the outdoor air fraction measured during in-field testing. The estimated outdoor air fraction during the middle of the day, when the compressor operates, is consistent with the problems described above. It is likely that the economizer damper was actually in the same position as it was during the nighttime period during economizer operation, especially as the in-field inspection suggested that the damper stays in the same position at all times.

3.4. Duct Losses While detailed investigations of duct leakage and duct conduction losses were not done during the site visits, there is still important information that can be taken from the field study. Most of the units have little, if any, duct work actually located on the roof of the buildings. In addition, ducts on the roof are typically in good repair since large leaks in ducts on the roof are usually easily found and sealed. The major question when it comes to leakage, then, is what happens in the buffer space between the roof and the occupied space. An assumption that frequently is made about commercial buildings is that the ducts are all effectively inside the conditioned space, which implies that the ceiling is fairly well connected in terms of airflow with the buffer space, and that the roof is tight. This is not

34

always the case. In many buildings the roof is vented or has a number of penetrations. A study of 15 units in 8 buildings done by Lawrence Berkeley Laboratory (Delp et. al 1998) found that 40% of the buildings had directly vented roofs. It is not fair, however, to characterize these buildings as having ducts that are all outside, because the degree to which the ducts communicate with the outside compared to the inside on a flow basis depends on the relative porosity of the ceiling versus the roof. If the ceiling and roof are equally leaky, one would expect that about half of any duct leakage in the buffer space would go to outside and the other half to the conditioned space. Another implication of the assumption that the ducts are all inside is that the roof is the thermal barrier as well as the air barrier. This is likely to be the case when the roof is insulated, but when the dropped ceiling is insulated or there is no insulation there will be significant thermal communication with outside. This is the case even if there is insulation both at the ceiling and the roof because, as with airflow, the degree to which the buffer space is connected with outside compared to inside depends on the relative amount of insulation at the roof vs. the ceiling. If both the roof and the ceiling are insulated to about the same level, one might expect that conduction between the occupied space and the buffer space is about equal to the conduction between the buffer space and outside. Further discussion of duct losses is in Appendix B. This includes details about the model used for estimating duct losses and the magnitudes of losses that one might expect for various situations. 4. Energy Savings Impacts and Conservation Measures This project was designed to assess the effectiveness of various maintenance procedures in improving the efficiency of small rooftop packaged HVAC units. The number of units actually reviewed was small, so the issues of savings potential and, more importantly, potential number of units that might be impacted by any one measure is not well determined. It is the purpose of this section to outline the parameters that should be used in assessing energy efficiency impacts associated with rooftop packaged units and economizer O&M, and to provide at least a framework for the potential energy savings impacts of a program aimed at improving the maintenance of these pieces of equipment. The protocol was divided into three major sections: airflow, refrigerant charge, and economizer. Each of these sections is impacted to some extent by control strategies, but by and large only the economizer is directly affected by the particular control strategies that we reviewed.

4.1. Coil Cleaning

It became apparent in the early stages of the equipment review that no systematic effort is made to clean outdoor or indoor coils on these systems. This can be a fairly misleading problem, since coil cleaning may not be of any particular consequence, especially on indoor coils when the environment is relatively clean, and even on outdoor coils when outdoor conditions are not particularly dusty. Conversely, in places where a large amount

35

of debris collects on the coils because of indoor conditions (e.g., the pet grooming facility) or outdoor conditions (e.g., cottonwood, which is very common in the Eugene area near the river), coil cleaning can seriously impact the performance of the unit. Our review of energy savings was limited to a single very dirty unit where we were able to directly measure the short-term impact of coil cleaning. When the outdoor coil is fouled, the COP of the unit goes down and the compressor has to work harder to evaporate the refrigerant. This brings higher amp-draw and, although it does not necessarily reduce the overall output, it certainly reduces overall compressor life. On the indoor coil, the results of fouling are mostly to reduce heat transfer. This means that the compressor must run longer to provide a given level of cooling. Although the compressor will not draw additional amps, the run-time and, consequently, the cost of cooling will increase and compressor life will be compromised. For this unit, we estimated a 10% savings resulting from coil cleaning. About 28% of coils examined had fouling that, if cleaned, could result in energy savings; most of this percentage represents outdoor coils. Savings are likely less that 10% on average. Unlike almost all the measures examined, this is a very straightforward problem: the extra cost of doing a complete coil cleaning on a unit is approximately half an hour of labor and $10 – 20 worth of cleaning solvent. Given the relative cost of such a measure, it seems reasonable to demand coil cleaning in virtually every case. Our experience and contact with local contractors indicate that coil cleaning is never considered part of routine maintenance in the Eugene market.

4.2. Airflow Improvement Under some circumstances other kinds of flow restrictions cause airflow to be below the manufacturers’ recommendations. Undersized ducts are quite common. Fifteen percent of the units we reviewed had airflow of less than half of the recommended values. This problem was difficult to address with any specific measures. In general, fan speeds were set relatively low for reasons that were not clear. Nevertheless, indications were that a 10% overall performance penalty might be expected at these low flows. For this measure, an additional study would be required to determine procedures and conditions under which airflow corrections might be made by changing fan speed. The other major measure that would be part of this category is the diagnosis and repair of the make-up air damper. About 14% of the units reviewed had damper settings or damper motors that did not modulate with the unit’s operation. This has the effect of introducing substantial outside air into the unit, which adds to the cooling and heating load in many situations. For example, a fully open damper can increase the heating load by 300% in some parts of the Northwest. In cooling mode the effect is somewhat smaller. In addition to the catastrophic damper failure, there are many conditions where minimum air should be adjusted. The nature of the savings from minor adjustment is difficult to predict. The damper settings we observed were arbitrary and so the relationship to the occupancy would require some research to establish. Table 4.1 summarizes the savings

36

available from selected damper repair and adjustment. These impacts could represent either an energy savings or an energy cost depending on which sort of adjustment was required. Item #3 in the table is the result of resetting a damper which is now admitting 10% outside air at minimum setting to provide ten percent more make-up air.. In general, the impact of damper setting is the same on both the heating and cooling load for the retail applications that were modeled. In buildings with much less internal gains than the retail buildings used in this analysis, the heating impact would be much greater. Table 4-1. Damper Adjustment Impacts Adjustment Mode % Savings 1. Stuck - open Heating 17 Cooling 15 2. Stuck - half open Heating 6 Cooling 7 3. Reset minimum Heating -2 Cooling -3

4.3. Refrigerant Charge

Of the units we reviewed, 9 out of 14 (64%) were either over or undercharged, according to the Carrier method of evaluation (as applied in the CheckMe!™ procedure). In 2 of these cases, the units were very undercharged (needing more than a pound of refrigerant), and technicians added refrigerant. In these cases, the low charge was the result of an obvious system problem -- a leaking service valve or pressure control seal. It is important to note that these are factory-charged units with relatively few opportunities for leaks to develop in the absence of a failure of the head pressure valve, TXV, or some other component of the system. Furthermore, the total amount of charge in these systems is relatively small. Because of these factors, the field technicians were reluctant to add or subtract refrigerant unless an identifiable difficulty was noted. The results reported here are based on very few tests. An ongoing review of several hundred commerical units throughout California (Proctor Engineering Group 2002) found 48% of the units required some charge adjustment and 12% of the units required more than a 10% charge adjustment (based on factory charge). These results are roughly congruent with what was observed in this project. Energy savings associated with charge adjustments have been estimated in several field studies done by Proctor Engineering. The basis for savings estimates in this report are the 1991 and 1995 field studies carried out for Pacific Gas & Electric. The estimated annual cooling energy savings are 6% for charge adjustments of less than 10% of the factory charge and 11% for charge adjustments >= 10% of factory charge.

37

4.4. Economizers Describing the impact of economizers on the operation of packaged rooftop units is a somewhat more complicated problem. In general, economizers have been used as an energy savings device mandated by energy codes for the past twenty years. However, the operation of economizers is not specifically regulated by code and, indeed, the installation and control set-up are based on installer preference and manufacturer recommendation, not necessarily on the optimum performance of the economizer. In order to assess the potential impacts of economizer measures, Ecotope ran a series of prototype simulations aimed at describing the total potential savings within the Eugene climate of various control and operation strategies.

4.4.1. Economizer Simulations Economizers in small rooftop package equipment differ significantly from those in larger, more complex equipment. The “ideal economizer” will utilize outdoor air any time the outdoor temperature is cooler or has a lower enthalpy than the return air; the remaining cooling is then made up by the compressor. This control strategy is known as integrated cooling. Critical to this optimum cooling strategy is a device that allows the cooling capacity of the compressor to be modulated. This would then allow small, medium, and large temperature differences all to be utilized. Units of 10 tons or less total cooling capacity generally have a single compressor with no modulation capabilities. In the 10 – 20 ton range, devices have two compressors, allowing a degree of modulation. Several economizer strategies are employed by small package equipment to work around these limits. To elucidate the variability of savings due to economizer type and strategy, we modeled the various economizer strategies using DOE-2.1e. The BPA small retail and small office prototypes were utilized for this exercise. Three lighting power densities (LPDs) were used in the small retail prototype to simulate the range of lighting power found in new retail space in Oregon. TMY2 weather data for Eugene were used. This table assumes a minimum air setting of 20%. It should be noted that the retail and office prototypes differ substantially in both heating and cooling loads. For purposes of this analysis, the retail prototypes were used since this most closely represents the buildings represented in this sample. Both changeover and differential logic strategies were modeled with changeover set points varying from 55 – 65 F. In addition, differential enthalpy and enthalpy changeover were modeled along with compressor lockout versus integrated compressor (simultaneous) operation. The results of these simulations are presented in Table 4-2. Even though nearly all modern units are capable of integrated cooling, their control strategies and lack of modulation result in a reduction in energy savings from the ideal. In particular, the most common configuration (changeover at 60 F outdoor dry bulb) achieved an average savings of 50% of the ideal economizer. Using this setting can be problematic, since it involves occupant comfort or perceived comfort and, from the

38

service technician’s perspective, reliability, in terms of delivered cooling. (That is, the technician’s job is to make sure cool air is always delivered when it is needed, and a system that uses only the refrigerant cycle for cooling, rather than using some amount of outdoor air, will always provide cool air, assuming it is not far out of adjustment.) A program to directly explore occupant satisfaction with different changeover set points would be very useful. Utilization of EMS systems, smarter thermostats, or thermostats with a smaller difference between stage-one and stage-two cooling should allow the aggressive changeover setting of 65 F or differential enthalpy or dry bulb control, even if comfort issues turn out to truly be an issue. The simulations validate the assumption that enthalpy changeover offers little advantage over dry bulb changeover in this Pacific Northwest climate. Energy savings between the two sensor types are negligible for the changeover strategy and the differential strategy. It is important to note that differential logic offers savings above those achieved by changeover logic particularly when economizer and compressor function are integrated. Since most controllers only offer or discuss differential logic with enthalpy sensors, (this is the only way to insure that the economizer is not increasing energy use in times of high latent outdoor conditions) the use of enthalpy sensors should not be dismissed out of hand. The simulations also highlight the fact that integrated, simultaneous operation of economizer and compressor only leads to savings when an aggressive changeover set point or differential logic is used. At higher outdoor dry bulb changeover points, most hours of cooling are satisfied by the economizer, thus greatly reducing the need for integrated operation. In general for the purposes of this table, only units with two compressors and two- or three-stage thermostats should be considered integrated. In cases with only a single stage compressor, the impact of the integrated control logic is greatly reduced or eliminated.

39

Table 4-2. Ratio of Actual Economizer Savings to Ideal Economizer Building Type

Office Retail3

LPD-1.85 LPD-2.58 LPD-3.04 Average Cooling Load (kWh/sq.ft.-yr) 1.90 3.22 4.04 4.60 3.44 Ideal Economizer Savings 42.6% 41.9% 46.1% 48.4% 44.7%

Heating Load (kWh/sq.ft.-yr) 3.29 1.58 0.92 0.62 1.60

Savings Ratios (% of ideal economizer) Dry Bulb Control Changeover 55 0.11 0.20 0.31 0.37 0.25

Changeover 60 0.39 0.45 0.56 0.62 0.50 Changeover 65 0.71 0.69 0.72 0.73 0.72 Differential 0.75 0.70 0.72 0.73 0.73 Changeover 65, integrated2

0.76 0.78 0.84 0.86 0.81

Differential, integrated 1.00 1.00 1.00 1.00 1.00 Enthapy Control Changeover D1 0.13 0.25 0.34 0.38 0.27

Changeover C1 0.36 0.38 0.49 0.54 0.44 Changeover B1 0.70 0.45 0.56 0.62 0.58 Changeover A1 0.74 0.45 0.56 0.62 0.59 Differential 0.75 0.70 0.72 0.73 0.73 Changeover A1, integrated2

0.99 0.45 0.56 0.62 0.66

Differential, integrated 1.00 1.00 1.00 1.00 1.00 1. Enthalpy change-over set points correspond roughly to the traditional A, B, C, and D settings of

enthalpy change-over controls: A—28 BTU/lb., B—25 BTU/lb., C—22 BTU/lb., D—20 BTU/lb. 2. Integrated indicates that economizer and compressor (s) can run simultaneously. 3. Lighting Power Density (LPD) allowance varies between 2.0 w/sf and 6.0 w/sf in the Oregon Energy

Code for retail occupancies

4.4.2. Economizer Repair Measures

For the economizers reviewed in this project, there are four general classes of repair and efficiency measures identified. 1. The first were measures associated with broken or fixed-position outside air dampers

that do not allow the economizer to modulate outdoor air regardless of outdoor temperature or indoor cooling demand. In general, these dampers were fixed-position, probably at installation and usually by adjusting minimum damper setting to 100% open or by removing the jumper that normally allows the damper actuator to be called by the control system. While these fixes are often difficult to identify, 22% of the economizers reviewed had some sort of dysfunctional damper control that made the economizer fail to operate no matter what the control logic.

40

2. The second measure involves the economizer controllers themselves. In these cases, the economizer controller is either set up poorly or disabled so that the economizer will never actually operate under Eugene conditions. In general, this occurs when the changeover set point is turned to off or a very low temperature. This has the effect of disabling the economizer under virtually any condition that might occur. Other reasons for this could be poor set-up of the control boxes themselves, where jumpers are not installed in accordance with manufacturers’ recommendations, or where sensors are disabled or otherwise non-functional. Thirty-five percent of the economizers we reviewed had this condition. (This percentage does not include any of the economizers where damper or damper function were disabled.)

3. The third measure identified was a measure based on a working economizer with a

conservative changeover setting which reduced the function of the economizer. This can be seen by reviewing Table 4-2, where an economizer with a changeover temperature of 65 F can more than double the energy savings of a 55 F changeover of. The difficulty here is that these more aggressive temperature settings introduce the possibility of inadequate cooling under some circumstances, especially where the unit is operated with a single-stage cooling control, in which case the changeover can only occur when the economizer closes and the compressor operates. In these conditions, when changeover temperature is set at 65 F, the cooling capacity of the unit is reduced by a factor of about two as the unit nears the changeover temperature. If large amounts of load such as solar or interior lighting are present, this could prove inadequate. Two-stage thermostats should be installed in these cases.

A compromise could be made in which the changeover temperatures are set to 60 F dry bulb. This has the effect of substantially increasing the overall performance of the economizer and reducing overall cooling energy by about 11%. At these temperatures, cooling capacity is not appreciably reduced, and the overall energy savings benefits are fairly significant. The only difficult part of this measure is determining the DIP switch and temperature calibration so that the changeover can be reset reliably.

4. The final economizer measure s was to reinstall a control board and introduce an

integrated/differential control logic. This has the potential for large cooling savings in virtually every unit we reviewed. In no cases were differential controls used, with or without integrated controls. However, large savings from such a measure are available when the unit is controlled by single-stage changeover (either dry bulb or enthalpy). Given the results of our simulations, this represented about 30% of the economizers we reviewed. Such a measure would save between 20 – 25% of the cooling energy. The benefit of this measure assumes a dual compressors. In units larger than 10 tons this was the norm. Without the second stage the integrated control offers very little benefit over the differential control.

41

4.5. Energy Savings and O&M Impacts Table 4-2 shows the impact of measures attempted in this project. Percentages in these cases are of the relevant units tested and reviewed. A total of 30 units were tested, but the number of units for which a measure was applicable could be substantially less, either because those characteristics were not reviewed or because (as with two of the units), no economizer was installed. The percent savings associated with these measures are, for the most part, independent of all the other measures. The savings are mutually exclusive within each category: each individual measure would be applied as an alternative to the other mentioned. Table 4-3. Savings (Percent of Annual Cooling Energy) Savings Percentage of Units Refrigerant Charge Greater than 10% Change 11 12 Less than 10% Change 6 36 Flow Coil (Cleaning) 10 21 Damper or Other Flow1 4 5 Economizer (60 F Change-Over) Not Operating (Damper)3 26 15 Not Operating (Control)4 26 30 Economizer (55 F Change-Over) Not Operating (Damper)3 14 10 Not Operating (Control)4 14 20 Economizer Changeover Setting (55 – 60 F) 11 37 (sensible) Differential/integrated Control2 23 30 (enthalpy) 1. With severe malfunctions, savings can exceed 40% of the cooling load. This measure also impacts

heating load by a similar percentage. 2. Assumes practical integrated operation for single stage compressors 3. Savings available from repair of the outdoor damper (stuck at minimum) which allows damper

operation. 4. Savings available from repair of controls allowing economizer operation with a change-over as

specified. The most significant savings are available from economizers; about 60% of the economizers reviewed were thought eligible for one or another of these energy saving measures. This is not to say that 60% of the economizers were dysfunctional, since in at least a third of the cases, measures that involved improving the control or increasing the changeover set point did not imply that the units were not functioning. In general, they were functioning properly, but by adjusting these control points, considerable savings would be available. This does, of course, require a level of sophistication on the part of the technician to judge when such control changes might be appropriate. In this review, changeover controls were set to at least 60 F in almost every case where the sensor could be accessed. Enthalpy controls were usually not reset to a more

42

aggressive level, although such a reset would generally result in a similar savings. Economizer damper operation was checked and repaired in almost all cases, but other control faults were not always fixed, given the range of time and resources available. Applying these savings estimates to the individual cases to estimate savings proved impractical. This was largely due to the fact that the measures and protocol were under development over the entire project and the importance of particular repairs may not have been apparent in the early visits. Furthermore, the billing analysis that would link the savings impact to actual energy use proved impractical since very few of the bills could be applied to a single rooftop unit. Finally, the impact of these measures is on the cooling energy requirements, which are often difficult to separate from the monthly billing records for the building as a whole let alone for a particular piece of equipment or zone. At this point, we believe that the simulations (Table 4.1) offer the best estimating procedures, since they are normalized to building area. More study and more systematic application of the measures described here would be necessary to develop useful savings estimates from enhanced maintenance service procedures.

4.6. Cost/Benefit Analysis To evaluate the costs and benefits of an enhanced maintenance package for rooftop units, the individual measures had to be combined to accommodate the manner in which these costs and benefits interact in specific cases. While the nature of these combinations is somewhat arbitrary, they do allow costs and benefits to be assigned to packages that are largely composites of diagnostic and repair measures aimed at improving the function and control of cooling in economizer packages. This evaluation is based on average size units observed in the EWEB service territory and on the normalized simulation results presented in Table 4-2. For the most part, the zone size served by single rooftop package units is a function of the design standards applied by the equipment installer. These standards are fairly uniform over a wide variety of end-uses, and particularly uniform across the retail sector. Typically, unit sizes are based on a sizing criterion of approximately 30 BTU/hr cooling capacity per square foot of conditioned space. Even in very large retail stores, this ratio is roughly consistent, with the number of units and the amount of cooling tonnage on the roof corresponding to about 30 BTU/hr per square foot. The average unit size that we observed in this sample was between 6.5 and 7 tons of cooling capacity, or approximately 80,000 BTU/hr of cooling output for each rooftop unit. Given these factors, the overall retail area associated with each rooftop unit is between 2,500 and 3,000 square feet. For purposes of this analysis, we have assumed that an individual packaged unit serves a retail zone of 2,700 square feet, with a median lighting power density that corresponds to an overall predicted cooling load of 11,000 kilowatt hours per year in the Eugene climate. This cooling load assumes that the economizer is not operating at all, so that outside make-up air is set to a fixed capacity of 20% of the total airflow at the evaporator coil. The scheduling and occupancy functions correspond to the Bonneville Power

43

Administration’s small retail prototype, which includes a set-back of 12 hours during periods of non-occupancy. It is important to note that this is a prototype analysis that corresponds to a moderately-lit retail establishment. Depending on the size of the retail space, the Oregon Code allows up to 30% more lighting power density than this. On the other hand, if the occupancy were a medical clinic or office, the cooling load would be expected to be about half of this value. Consequently, the cost/benefit ratios described here can vary substantially in particular units based on the occupancy, cooling loads, and other operational issues. The cost/benefit analysis is presented in Table 4-4. It has been derived from the combination of individual measures that were applied or were thought to be applicable to the rooftop units evaluated in Section 3 of this report. Table 4-4. Cost/Benefit Analysis (7-ton Prototype Unit)

Savings Cost ($) (kWh/Year) ($/kWh)

Payback(Years)

Commissioning Measures 1. Change-Over 275 1,100 .031 3.3 2. Change-Over with T-Stat 675 2,500 .033 3.6 3. Control Board 775 3,200 .030 4.8 4. Control Board with T-Stat 1,125 4,100 .034 3.6 5. New Economizer 1,600 4,600 .043 4.6 Repair Measures 1. Refrigerant Charge* 100 350 .064 3.8 2a. Damper Repair (Reset)* 150 450 .075 4.4 2b. Damper Repair (Flow Adjustment) 250 820 .069 4.1 3. Gas Combustion Test* 100 - - - *Gas savings and/or heating impacts not calculated.

4.6.1. Operations and Commissioning Packages For purposes of analysis, five separate retrofit packages were assessed, ranging in complexity and expense. These packages were designed to address the diagnosis and repair of economizers and airflow as a result of the direct review of the equipment by a program to enhance its operational efficiency. In all cases, these repairs and procedures were designed to have a fairly long life and to correct difficulties that have resulted in deferred maintenance and related wear and tear, as well as design and installation that resulted in less-than-optimal operation of rooftop equipment. Package 1: Coil Cleaning and recalibration of economizer sensors and minimum air setting: This package includes coil-cleaning of both the evaporator and condenser coils together with a review of the economizer sensors and economizer changeover settings, etc. This review is designed to be conducted over the course of two hours and to result in re-setting the potentiometer to a more aggressive free cooling strategy. It is also designed to ensure that sensors are properly calibrated and jumpered on the economizer control board.

44

This package of measures could be considered routine commissioning, in which a fairly straightforward diagnostic results in the technician understanding and repairing the economizer. The implication of this measure is that the technician, by application of the protocols, can understand the economizer and determine whether it is operating correctly in a relatively short period of time. No additional parts or other field diagnostics are required, and the potentiometer is easily reset. Package 2: Reset change-over with addition of two-stage thermostat control : In approximately 25% of the units reviewed under this program, the control provided by the thermostat was a single-stage cooling set-point. This is a typical level of control for virtually all residential thermostats with any level of other control. Such controls are almost always inadequate for a rooftop package unit with an economizer setting, even where a single compressor without any staging capability is present. In order for the economizer to function, a two-stage cooling set-point should be included. This package is essentially identical to Package 1, except that the thermostat would be placed in the zone and the system would be re-wired to accept two-stage control. Our experience is that this requires relatively few additional adjustments in the rooftop unit, often not even a change in dip-switch settings, though an additional wiring connection would have to be made. In commercial buildings, the control wire bundle is usually adequate to carry the extra signals involved in two-stage control. Thus, for purposes of this package, the thermostat would be replaced and the control bundles re-wired; the economizer would be commissioned and reset as in Package 1. In addition, evaporator and condenser coils should be cleaned in this package as the beginning of a routine maintenance. Package 3: New economizer control board: The integrated circuit board economizer control of these units is usually complicated. This is largely because the control boards are manufactured for a variety of OEMs with combinations of control strategies appropriate to various manufacturers and sizes of equipment. When an economizer does not function and the controls are of the older, electro-mechanical variety, it is often quite difficult for the technician to discern the repair strategy. This often results in diagnostic procedures that last hours yet fail to establish the function and efficiency of the economizer. This package, then, would be based on an initial review of the economizer controls to determine the level of diagnostic effort required to repair or commission the economizer. In the event that these procedures did not yield sufficient information, a new control board would be installed. This control board would be assumed to handle two stages of cooling and provide integrated control of the economizer. The control board is assumed to be installed with new sensors using the protocol. The technician would be expected to document the jumpering and control strategy used by the new board. This measure would allow the economizer to be optimized using a set of known controls and sensors, and a documented calibration to facilitate future maintenance to have a clear basis for reviewing and maintaining the economizer and control system. Like Packages 1 and 2, this would also include cleaning of the evaporator and condenser coils, together

45

with the full installation and review of both the economizer control and the damper actuators and control, including evaluation and reset of minimum airflow. Package 4: Package 3 with additional two-stage thermostat: This option is included on the assumption that under some circumstances, a two stage thermostat would not be present and would have to be included in the economizer repair package in order to get both optimized control and effective temperature management in the zone. Package 5: New economizer: In spite of extensive diagnostics, there are some circumstances in which it may not be possible to even install a new control system. This would be particularly true of older style economizers (from the late 1980s and early 1990s). In most of these cases, digital control boards were not included, and a detailed diagnostic would be time-consuming and probably ineffectual. Furthermore, in equipment installed before 1990, economizers were not required in smaller rooftop units and so might not be present at all. In these cases, the most obvious maintenance strategy would be to remove the existing economizer and install a modern model. There are several companies that make packaged economizer systems designed for this manner of installation, and the HVAC contractors agreed that this was a measure that they could deliver if the conditions warranted it. The cost assumption made here is that the new economizer would be optimized for integrated control, and would be wired to the central control board of the rooftop unit so that the economizer would be optimized to the operation of the equipment. In many cases (and especially in older equipment) this would obviate the need for detailed diagnostics.

4.6.2. Maintenance and Repair In addition to the packages presented above, there are several maintenance items which, if regularly performed, would provide energy savings in the event that a review showed the opportunity for repairs of the components of the equipment. Refrigerant charge: The impact of charge adjustment on the equipment varies considerably. In roughly half the cases examined in the Eugene area, some charge adjustment was dictated by the CheckMe!™ procedure. Only two of these cases were actually adjusted, and such adjustments were the result of specific failures in valves or seals that resulted in a loss of charge during the course of normal operation. However, the remaining cases were also shown to have deficiencies ranging from several ounces to a quarter or half a pound of refrigerant.

Technicians were generally hesitant to adjust the charge in these factory-sealed pieces of equipment, although both the CheckMe!™ review and literature suggest that deficiencies of even minor amounts of refrigerant can result in a reduction in system EER. It is quite clear from the reaction of technicians that the CheckMe!™ procedure needs to take into account a broader range of charge repair and to suggest charge adjustments of substantially smaller amounts in dealing with packaged units. Proctor Engineering is currently working on refining the charge adjustment recommendation portion of

46

CheckMe!™ for commercial units. Nevertheless, it seems quite apparent that charge adjustment should be made in many more cases that are not currently included in routine maintenance review. Damper adjustment and reset: This problem was typically the result of installation damper settings being either fixed in an open position or varying through a range of operations that resulted in routine introduction of substantially more outside air than was required for ventilation. While it is true that adjusting dampers to meet minimum ventilation specifications can result in increases in energy usage, in the Eugene equipment it was much more frequent that dampers were fixed open or at least partially open, resulting in increases in heating and cooling load as the excess air was introduced into the conditioned space. A third -- and perhaps more severe -- issue was that quite often the outside air damper did not allow full airflow when modulated to full open position. While this does not have an impact on ventilation air, it substantially reduces the effectiveness of the economizer. With airflow testing, dampers can be reset to modulate more fully while providing minimum ventilation during periods of equipment operation. Gas burner review and reset: The list of modern gas pack equipment includes relatively little adjustment of the burner; factory settings are usually sufficient to maintain proper heating function at rated efficiency. A combustion and line pressure test, however, should probably be conducted to establish that proper pressure and firing rates are maintained. This may or may not have an appreciable impact on the efficiency of the equipment, but in rooftop applications quite frequently some of the units are sufficiently remote and piping design is sufficiently marginal that the units are under-fired. This does not imply an adjustment of the unit itself, but it does suggest an adjustment in either gas plumbing or pressure reducers so that adequate pressure is available throughout a rooftop system. Coil cleaning: While coil cleaning was included in all of the standard packages, it should be noted that as a separate item it is relatively inexpensive and should be performed periodically. While there are certain applications in which coils may not need to be cleaned more than once every five years, most units perform markedly better with annual or biannual cleaning. We observed considerable improvement in compressor function in places where coil cleaning had been deferred for a substantial period of time. Most maintenance and operation contracts do not include coil cleaning, and technicians only occasionally propose to owners that coil cleaning be done.

47

4.6.3. Enhanced Maintenance Procedures All of the packages described in Table 4-4 and the above section would be included in an enhanced maintenance procedure in which the repairs and maintenance required by the initial protocol would be maintained over a long period so that benefits would accrue to both building owners and contractors. The entire effort with the economizer repair and commissioning would be based on the assumption that once these measures have been conducted using a protocol documenting the repair procedure, relatively small amounts of maintenance and review time would be required over a period of five to ten years in order to maintain the savings achieved from the initial packages. It is this feature which must be a part of a cost-effective program, since the price of new economizers, control boards, and thermostats cannot be justified unless these costs are amortized over longer periods. An enhanced diagnostic procedure would include full operational review, repair, and replacement of economizer controls, reset of change-over controls, and evaluation and reset of airflow charge and damper settings in the initial year of operation. This would be followed by annual or semi-annual visits to ensure that the repairs continue to function and that any deficiencies in charge or compressor seals are quickly identified and repaired. This would also include periodic cleaning of coils and changing of air filters, which are included in current maintenance contracts.

The entire effort would be based on extending the life and efficient functioning of the equipment. To our knowledge, none of the contractors we worked with on this project offered anything resembling this kind of enhanced O&M procedure. 5. Contractor Relations Much of this report deals with the technical issues involving HVAC equipment optimization. As important as these technical details are, the issues associated with the delivery of technical services are just as important. Understanding factors such as how routine maintenance services are marketed, and the fact that the party paying for maintenance is seldom the party paying the energy bills, are critical to achieving high performance of roof top units. An important goal of this study was to examine if advanced roof top diagnostics could be incorporated into the routine maintenance visits that typically are scheduled in spring and fall. What soon became obvious to the field team is that routine maintenance is definitely routine, usually only covering the most basic of maintenance tasks and having little or nothing to do with optimization of the performance of roof top units. Routine maintenance typically includes the following: • Filter change-outs • Fan belt inspection • Basic system operation (does the unit heat and cool?) • Refrigerant charge check (this is not a super heat or sub-cool test) • A physical inspection of the entire unit (dirty coil, burnt wiring, etc.)

48

These services typically take 30 minutes per unit. If a service such as coil cleaning is required, the technician prepares an estimate and submits a bid to the building owner; it is not a standard maintenance item, although one contractor informed us that it is standard in many places such as California. The technicians involved in this study related several anecdotal accounts of bids being prepared that accounted for no additional work. In cases where large ticket items were bid, the building owner would request other bids. Routine maintenance contracts are often a result of an informal bidding process where the building owner calls a few HVAC companies to get the lowest bid. The description of what constitutes a maintenance package is not defined. Without a set of minimum standards, obviously the company that spends the least time per unit is likely to be the low bidder. Another method of obtaining maintenance contracts is for the HVAC contractor to call the accounts of other contractors and ask them the price of their existing maintenance contract. Similar to long distance phone service solicitors, the caller then offers a price below what they are currently paying. Again, the lack of a package of minimum specifications allows the contractor that does the least to charge the least. Worsening this low bid environment is the fact that some contractors view these maintenance contracts as the key to selling replacement equipment when the rooftop units need replacing. This actually discourages good maintenance if the only time a profit is made is when the unit requires replacement. A low bid environment is not a place where advanced diagnostics and related services are likely to thrive. To the extent that this low bid form of maintenance works, it is probably made possible by the fact the building owner is usually not the party paying the energy bills. His or her primary concern is that the unit blows enough hot or cold air to meet the comfort needs of the tenants. An additional obstacle to the delivery of expanded services that may optimize the operation of rooftop units is that service technicians assigned to rooftop maintenance are still on call. This implies that the tech may be called away several times during the day to work on other projects that demand immediate attention. This situation obviously does not lend itself to a systematic approach to rooftop diagnostics. This on-call status is probably a direct outgrowth of the low bid environment in which routine maintenance is conducted. The maintenance visit is at a fixed price (with the exception of replacement parts). These visits essentially become filler work scheduled between more lucrative jobs that can be billed based on time and materials. If these conditions do not promote a high level of maintenance, they definitely do not produce an environment that is conducive to correcting problems that were created by improper installation. Many of the systems studied in this report had operational problems that date back to their date of installation. These problems ranged from the

49

minimum air setting on economizers being set to 50% outside air to fan flows being half of their optimized levels. It is unrealistic to assume that these types of problems will be corrected during routine maintenance visits. In essence, asking technicians to correct these types of problems is retro commissioning. Retro commissioning is a complex and time consuming task that may require research, parts orders, and follow up visits; tasks that are unlikely to occur under the conditions of a typical maintenance contract. For all of the above reasons, it is unreasonable to assume that a program that is aimed at optimizing rooftop units can be piggybacked onto existing maintenance visits. The insistent need for speed during routine maintenance visits makes it a poor vehicle for the delivery of a program that requires advanced diagnostics and follow-up visits. Designing a program that can deliver optimized rooftop units must take into account the needs of all parties involved. A brief summary of these parties and their primary goals are listed here. Party Primary Goal HVAC Company Provide profitable service that blends well with business plan Utility Save energy Building Owner Minimize tenant complaints and maximize equipment life Bill Payer Minimize energy costs while creating a good retail environment Meeting all of these goals is dependent upon a technician having adequate time and knowledge to perform all diagnostics and any follow up corrections. In order for this to occur, the HVAC company must be well compensated. At the current time, this implies a high degree of utility subsidy both in forms of actual cost and quality control. It is important to stress that the cash subsidy is not enough to assure that all parties' goals are being met. This study found several examples whereby utilities subsidized equipment that was not commissioned properly. Aside from the lack of energy savings, there is long term relational damage caused when decisions are made based on promised energy savings and none are delivered due to poor installations. A premium maintenance visit combines retro commissioning services plus other more typical maintenance services such coil cleaning and filter replacement. This implies a two-tiered delivery approach. The first tier would include the following: • Thorough economizer check out with corrective actions • Airflow measurements of air handler and economizer using TrueFlow™ Air Handler

Flow Meter, with corrective actions • Evaporator and condenser coil cleaning • After fan flow corrections are made and coil cleaning is performed then conduct a

Check Me!™ charging test and make any needed corrections. These first tier items, if done correctly, should cause the unit to be largely optimized. After that, as long as components do not break, none of these items (except for coil

50

cleanliness) should change over time, and future testing can probably be kept to once every few years just to make sure that nothing has failed. Once the tier-one work is performed, subsequent visits can focus on items that are strictly related to maintenance. This includes: • Filter change-outs • Fan belt inspection • Basic system operation (does the unit heat and cool) • Temperature drop across coils • Operational refrigerant pressures • Evaporator and condenser coil cleaning. This two-tiered approach obviously implies a data collection system that can keep track of units on which first tier tasks have been performed successfully. It also implies a level of quality control that does not currently exist. Check Me!™ provides quality control for refrigerant charging only. Quality control beyond Check Me!™ will prove more difficult. Airflow measurements using the TrueFlow™ give only a small degree of quality control. Economizer checkout procedures also provide only a minimum level of quality control. To assure that systems are being properly tuned, an independent field level verification must be made. The percentage of jobs that has to be inspected is probably about 10%. Contractors did express some hesitation to using the advanced diagnostic techniques in this project. For airflow, there was some feeling that nothing could be done to change the flow in many systems, so why should one bother to measure it. For charge, the feeling was that a problem could be determined without the use of CheckMe!™, which is probably largely true for experienced contractors. For economizers, the process is seen as lengthy and nebulous since it can be difficult to identify the source of a problem and since there are so many different types of economizers. Evaluating the units is not the only reason to perform these advanced diagnostics. For example, collecting this information allows the development of a systematic database detailing the actual performance of a large number of units, the problems found, and the corrective measures taken. This can help in determining where large savings can be found in general, what types of problems are common or can be easily avoided with more attention paid to proper installation, and what components are simply not significant problems. For example, some units were found to have economizers that had their dampers set to open fully at installation, or had their changeover set to off. These are very simple corrections that could be made, but contractors are not looking at these issues either when they install the units or when they go on maintenance visits. On the other hand, charge may not be a major issue, with only a small percentage of units needing any real adjustment. The impacts of regular coil cleaning could also be evaluated. Many lessons were learned during the course of this study. Of primary importance to the design of a successful rooftop program was the recognition on the part of field team that the semi-annual maintenance visits would make a poor vehicle to deliver a rooftop

51

optimization program. Once the system is retro-commissioned, these routine maintenance visits may work as a way to keep the systems in tune. 6. Conclusions The purpose of this effort was to try and establish the veracity of an enhanced contractor-based maintenance in developing both energy savings and the services available to EWEB customers. This effort showed, albeit for a very limited sample, the importance of enhanced maintenance in the overall functionality of rooftop units. Even where enhanced controls would not be considered a viable measure, about 60% of the units reviewed required some maintenance that would result in improved function and efficiency. In about 20% of the cases, the review identified catastrophic or near-catastrophic failures that had not been recognized in previous maintenance visits. This included worn or frayed belts, coils fouled to the point that compressors were operating at significantly increased amp draw, leaking valves and seals resulting in a gradually reduced refrigerant charge, and fixed-position outside air dampers, which resulted in nearly 100% make-up air and reduced overall capacity and function of the compressor and heating equipment. While these issues may have been somewhat obvious, in all cases the units were visited regularly, and in no cases was this regular review capable of finding even the most serious and self-evident of these problems. In this sense, the mere act of having a systematic protocol that requires certain things to be checked, and especially a protocol that requires a review of refrigerant charge and damper function, would harness a significant fraction of the savings and benefits that were identified in our review. The more sophisticated issues with economizer controls and complex issues with dampers would require a larger amount of effort. In many cases, this larger amount of effort would also yield noticeably greater savings, although in most cases these savings would also include functional compromising of the equipment to the point where a catastrophic failure might result in the absence of the measure. Contractors generally sell the routine maintenance check as a loss-leader. Seldom is the budget or time available to review even the current charge. Most technicians we spoke with believe this to be a waste of time and, in fact, in many cases it is of marginal significance. However, in a few cases (15% of those reviewed in our protocol), the result of applying the CheckMe!™ program led to the identification of serious failures or difficulties with the air coils that could not have been easily identified any other way. With economizers, there are really two layers of review, but at this stage the ability to understand a wide variety of economizers and the potential benefits from a detailed approach is limited. It is quite clear that even economizers that are set up properly and operate correctly may not be delivering a very high fraction of the potential economizer benefits. Because of the conservative settings of the change-over set point and aggressive settings of the make-up air damper, much of these savings available are compromised. We know that these settings are used to make the installation more straightforward and reduce the potential for call-backs in the event that one or another complex control failure. One can

52

understand how this might be a serious problem for an installer or maintenance technician when it is apparent that about a third of the economizer controls were compromised either in installation or operation. In almost no cases were these difficulties identified without substantial review by the technicians and a review of manufacturers’ literature relating to the set-up and function of the economizer in question. In virtually no case was this information available on the roof or in the tools of the technicians assigned to the project. It is quite apparent from the nature of the business we observed in this project and from the time allotted to this operation that the existing maintenance contracts are unlikely to result in any of the potential savings measures being identified let alone attempted. Only if the contractors are able to sell an enhanced service, paid for either directly by the utility or by the customer, would any of these reviews even be possible. However, it must also be pointed out that, even in the presence of such a marketing effort, it is not entirely clear that the contractor services as currently observed would be able to deliver a sophisticated system review on any sort of regular basis. Another issue is installation. In only a few cases were we able to observe units where the installation was relatively recent and the faults observed could be directly traced to the initial set-up of the equipment. This is somewhat more problematic, since the people providing installation are not typically the field technicians who service the equipment; thus, the knowledge and detailed information that might be gleaned by a field technician cannot necessarily be applied at the level of installation, even when the same companies are involved. It is quite likely that commissioning these small rooftop units so that the controls and dampers are properly set up would be an essential feature that might be attainable without discussing the need for enhanced maintenance services. In this case, commissioning would probably take place through field technicians trained in economizer protocol. Certainly, at the time of installation, manufacturers’ literature, control diagrams, and various actuator settings are much more available, making the protocol itself more straightforward. In any case, if the economizer and damper settings are to be correctly functioning, waiting until some regular maintenance visit years in the future seems relatively short-sighted. It is obvious from the evaluation of the available measure packages that a part of a cost- effective program would have to include a long term and extensive review of the operation, installation, and maintenance of the entire packaged unit. At the outset, this would be an extensive and expensive addition to the existing maintenance agreements offered by the service companies. If this program offered the opportunity to supply a diagnostic service and replace non-functional parts profitably then the contractors resistance to this level of involvement would be reduced. An enhanced diagnostic procedure would include full operational review, repair, and replacement of economizer controls, reset of change-over controls, and evaluation and reset of airflow charge and damper settings in the initial year of operation. This would be followed by annual or semi-annual visits to ensure that the repairs continue to function and that any deficiencies in charge are quickly identified and repaired. This would also include periodic cleaning of coils and changing of air filters, thus replacing existing maintenance contracts.

53

The entire effort would be based on preserving the function of the equipment and extending its life with an enhanced maintenance procedure. To our knowledge, none of the contractors we worked with on this project offered anything resembling this kind of enhanced maintenance. While many of the services implied by the packages in Section 4.6 have been offered in an occasionally, most technicians have not been given the time or the tools to identify the function of the equipment, even when maintenance issues were identified by the owner. While we believe that this protocol remains somewhat incomplete and certainly has not addressed all the potential savings available from economizers, it is also very apparent that an enhanced service of this type can yield great benefits and energy savings to the utility and substantial benefits in reduced maintenance and energy costs to the building operators. As such, considerable effort should be expended to establish this protocol inside of the contractor community in Eugene, and to ensure that owners and operators of this type are aware of and demand the level of maintenance and review that this effort suggests.

54

7. References Breuker, M., T. Rossi, and J. Braun. 2000. “Smart Maintenance for Rooftop Units”. Printed in ASHRAE Journal. Davis Energy Group, 2001, Specification of Energy Efficient Installation Practices for Commercial Unitary HVAC Systems, Consortium for Energy Efficiency, Boston, MA Delp, W.W., N.E. Matson, E. Tschudy, M.P. Modera, and R.C. Diamond. 1998. Field Investigation of Duct System Performance in California Light Commercial Buildings. Lawrence Berkeley National Laboratory. Presented at the 1998 ASHRAE Annual Meeting. Hewett, M., D. Bohac, R. Landry, T. Dunsworth, S. Englander, and G. Peterson. 1992. Measured Energy and Demand Impacts of Efficiency Tune-ups for Small Commercial Cooling Systems. Houghton, D., 1997, Operating and Maintaining Rooftop Air-conditioning Units, E-source TU-97-2 Lunneberg, T., 1999, When Good Economizers Go Bad, E-source ER-99-34 Palmiter, L. and P.W. Francisco. 2000. Development of a Simple Device for Field Air Flow Measurement of Residential Air Handling Equipment: Phase II. Final Report. Prepared for the U.S. Department of Energy by Ecotope, Inc. Parker, D., J. Sherwin, R. Raustad, and D. Shirey. 1997. Impact of Evaporator Coil Airflow in Residential Air-Conditioning Systems. Florida Solar Energy Center. Presented at the 1997 ASHRAE Annual Meeting. Pratt, R., Katipamula, S., Brambley, M., Blanc, S., 2000, Field Results from Application of the Outdoor Air/Economizer Diagnostician for Commissioning and O&M, Presented at the 2000 National Conference on Building Commissioning. Proctor Engineering Group. 2002. Interoffice memorandum, Personal communication. Proctor, J., M. Blasnik and P. Downey. 1995. Southern California Edison Coachella Valley Duct and HVAC Retrofit Efficiency Improvement Pilot Project. Final Report. Prepared for SCE by Proctor Engineering Group. Proctor, J. 1991. Pacific Gas and Electric Appliance Doctor Pilot Project. Final Report. Prepared for PG&E by Proctor Engineering Group. Vick, A., J. Proctor, and F. Jablonski. 1991. Evaluation of a “Super Tune-Up” Pilot Program for Forced-Air Furnaces in Small Commercial Buildings. Presented at the 1991 Chicago Energy Program Evaluation Conference.

55

Appendix A:

Rooftop Packaged Unit Protocols

A-1

Rooftop Protocol Checklist

Pre-visit

Review any complaints about comfort/high bills; get an idea of big problems before hitting the roof

Gather preliminary unit info (size, mfr), contact info Indoor review

Tstat – staging/integration/scheduling If EMS, find knowledgeable individual to review settings Advise replacement if single stage Adjust scheduling as needed/allowed

Basic unit checkout (before beginning detailed procedures)

Electrical – conductors, relays, safety switches Gas – combustion efficiency, gas inlet pressure (possible piping issues) etc.

Could be combined with burner cleaning, ignition module check/replacement, etc. Economizer

Actuator/damper works (therefore econ board works) – adjust/replace as needed Sensors work?

o Electromechanical (crude tests) o Sstate (resistance/voltage tests) o Adjust/replace as needed

Min/max airflow measurements o Flowplate procedure o Adjust as needed and retest

Compressor/airflow/coils Activate compressor; measure as-found compressor/ID & OD fan amps + line and

control voltage Measure evap airflow; adjust as needed Clean coil(s) Charge check (CheckMe!); adjust as needed Re-measure compressor/ID & OD fan amps Note duct location and any strange features; check flows if possible

Additional possible protocols: Inrush/voltage sag; heat pump heating mode checks (reversing valve, defrost, outdoor thermostat, heating performance (COP), backup heat operation/staging)

A-2

Rooftop Unit Check-Out Procedure (Cooling Mode) Building Information Building Complex Utility Street Address City State Contact Name/Title Contact phone/details Unit Inspection Information ID# Type HP / AC / Gaspack Model# Serial # Nominal Cooling Tons Approx. Age Last Service Date Compressor Manufacturer Compressor Model # Compressor staging / description Factory charge Measured voltage @ contactor

Three phase one phase Approx ft2 served by unit

Nameplate indoor fan amps As-found/post-repair measured indoor fan amps

Nameplate outdoor fan amps As-found/post-repair measured outdoor fan amps

Nameplate compressor RLA As found/post-repair measured compressor RLA

Air distribution type Constant volume / VAV / bypass damper / other: Initial indoor coil condition Cleaned? Initial outdoor coil condition Cleaned? Filter type Initial filter condition Cleaned / Replaced? Economizer Inspection Information Economizer Manufacturer Economizer Model# Controller Board Manufacturer Controller Board Model # Actuator Type/Manufacturer Actuator Model # Sensor Manufacturer Sensor Model #(s) Control Logic Changeover/differential Setting Integrated w/cooling? Sensor type Dry bulb / enthalpy Sensor type Electromechanical / solid state Damper type Horiz. / vert. Louvers / Guillotine Damper position control Actuator/microswitches Minimum outside air setting (specify type of control)

Damper moves from minimum to maximum setting? Can minimum setting be adjusted? Minimum air shuts down to zero in unoccupied mode? Is there a return damper? Note any problems. Record as-found/post-repair minimum air flow Record as-found/post-repair maximum air flow Is there relief air? If yes, what kind? If no, is there a bldg. pressurization problem? How was damper operation checked? Was sensor operation checked / how? Note problems with sensors and other components and any adjustments made

If economizer failed any checks, was manufacturer’s checkout procedure performed? (Give explanation)

A-3

Rooftop Unit Check-Out Procedure Building/Unit_____________________

Thermostat Type (incl. Manufacturer) Occupied mode period(s) Occupied mode setting Unoccupied mode period(s) Unoccupied mode setting Continuous fan? # of stages Note any changes made

Ducts Note any obvious problems with ducts (undersize, constrictions, etc.)

Are there ducts on the roof? If yes, sketch below Is there a buffer space containing ducts? If yes Is it insulated? How thick is the insulation? Where is the insulation (e.g. roof / ceiling)? Is insulation between ducts and occupied zone? Is buffer space vented? Type of venting What is ceiling height? Can ducts be measured (if yes, sketch below)? Register/grille flows Flow measurement device(s) used

Supply Location Flow (cfm) 1 2 3 4 5 6 7 8 9

10 11 12

Return Location Flow (cfm) 1 2 3 4 5 6 7 8 9

10 11 12

A-4

With filters in place (Pfilter)

With flow plates in place (Pplate)

Correction factor = SQRT(Pfilter/Pplate)

Supply and return press.

Sup. = _________ Ret. = _________

Sup. = _________ Ret. = _________

From supply pressures:

Return/ Econo. (circle) Econo. Position _______

Filter Size Flow Plate size (14” or 20”) Pressure drop Raw flow

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

Return/ Econo. (circle) Econo. Position _______

Filter Size Flow Plate size (14” or 20”) Pressure drop Raw flow

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

Return/ Econo. (circle) Econo. Position _______

Filter Size Flow Plate size (14” or 20”) Pressure drop Raw flow

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

______________ ______________ ______________ ______________

Total Raw Flow, Return Total Corrected Flow, Return

______________ ______________

Total Raw Flow, Economizer Total Corrected Flow, Economizer

______________ ______________

Total Corrected System Flow Economizer Flow as a Percentage of Total Corrected System Flow

_________ _________

CFM/Ton _________

If economizer position was changed, indicate how this change was made:

A-5

A-6

Detailed RTU Economizer Information and Check Out Procedure Economizer Check-Out Economizer check-out is tricky due to the number of states or modes in which the economizer can be operating based upon the thermostat signals and ambient conditions. Of primary interest in the check-out is operation of the damper assembly in response to calls for cooling and apparent sensor readings. The following sections discuss the issues involved with economizer system operation, data recording in the field, and adjustments. Where specific procedures are described, data to be recorded, and adjustments suggested, the typeface is set in italics. Thermostat Type/Staging Before heading onto the roof, it is important to know what type of indoor thermostat is sending signals to the unit. Write down the model number of the thermostat and the number of cooling stages which appear to be connected to the subbase.. On the roof, inspect the wires coming from the thermostat to the unit. In units with a connector (transfer) board, this is often very simple: just verify that there is a wire coming from the thermostat to Y1 and Y2 for two-stage thermostats and just to Y1 for one-stage thermostats. Lower-end units without wiring boards require that the wires be examined to ensure that they are coming together with the unit wires. On units with standard Honeywell logic controllers, one can generally tell whether there’s second stage cooling from the thermostat if wires are hooked up to the Number 3 terminal. Number 1 is for first-stage cooling calls, Number 3 for second-stage. Record the thermostat wiring connections at the unit. Economizer Controller Logic Modern economizers rely on a combination of sensor readings and integrated circuit board (ICB) logic to control the economizer function. Economizers utilize one of three decision control logics for determining when they should operate during a call for cooling. These are dry bulb change-over, enthalpy change-over, and differential enthalpy. All three strategies involve an outdoor air sensor that detects outdoor dry bulb temperature or enthalpy. The differential strategy requires an additional return air sensor to determine the enthalpy of the return air. A mixed air sensor is included in most units to prevent use of outdoor air when its temperature is below a certain target. This presents introduction of outside air for cooling when it could cause occupant discomfort. The mixed air sensor, by its very name, measures the mixture of outdoor and return air.

A-7

Determining which logic is being used requires some familiarity with either sensors or control logic boards. Most logic boards in use (i.e., the standard Honeywell and Trane boards) come with the ability to utilize all three of these strategies. Switching between the strategies is simply a matter of connecting the appropriate sensors. On the standard Honeywell board/controller, the SO terminal connects to the outdoor air sensor, which may be either a dry bulb or enthalpy sensor. If there is a return air sensor, it is connected to the SR terminal. In a change-over mode, this normally just has a resistor jumper. The common Honeywell 7459 controller comes in many configurations in many units. Using the standard Honeywell sensor, continuous resistance is fed to the logic board so that it can determine the temperature. The actual cut-out point at which the economizer is enabled or disengaged is determined by the A, B, C, or D setting on the logic board. In other cases, manufacturers have implemented a simple on-off temperature device. This essentially overrides the ABCD settings. If the outside temperature device is closed, then the economizer will operate; if the circuit is open, it will not. In some units, this is a snap disk with a factory-specified trigger temperature. In others, it is an electrical/mechanical device with the ability to establish a set point. In older enthalpy units, this can be an ABCD set point or, for dry bulb controllers, a simple temperature setting. Occasionally, one will even see simple Honeywell thermostats installed in this place. The Trane controllers sense outdoor air near the main logic board. The economizer logic board has possible inputs of the return air sensor, the outdoor humidity sensor, and the return humidity sensor. Differential enthalpy requires that the outdoor humidity sensor (OHS) and the return humidity sensor (RHS) be hooked to the appropriate sensors. The default mode is dry bulb change-over using the main unit outdoor air temperature sensor. Record the model number of the controller board. Outdoor & Mixed Air Sensors A large variety of temperature sensors are used with modern economizers. Solid-state sensors feed a variable resistance or voltage to the control board and usually take the form of small probes or, in the case of Honeywell, box-shaped devices. Mechanical sensors include simple thermostats, often with dial controls to set the temperature at which they switch closed or open, and small snap disks, usually calibrated to a certain temperature. Older economizers, which utilize electrical/mechanical relays instead of logic boards, generally use either enthalpy or dry bulb change-over controls. Determining whether a sensor detects enthalpy or dry bulb can be difficult. Dry bulb sensors range from simple solid-state thermistors to more common mechanical-type snap disks (small, aluminum sensors with a half-inch diameter cam mounted on a little flange) and thermostats. Old style enthalpy sensors use either a nylon fiber sensing element or a layer of hygroscopic

A-8

salt deposited between two electrodes which senses changes in resistance based on the amount of moisture absorbance by the salt layer. Typically the outside air temperature sensor is mounted on the economizer hood, though some units (Trane for example) can have the sensor in the corner post of the cabinet near the main unit control logic board. The mixed air sensor is often mounted on or near the air handler. Record the type and model number (s) of sensor(s) on the data sheet. Change-Over Setting The most common type of economizer control utilizes outdoor air for cooling until a certain outdoor air dry bulb temperature or enthalpy is attained. At this point, the outdoor air fraction of the overall system airstream should go to minimum and the compressor comes on to provide almost all cooling. In a fully integrated system, the outdoor air fraction can modulate to a non-minimum level while a compressor or compressors are running, but this scheme is rarely seen in units with less than 20 tons of nominal cooling capacity. The point of change-over from maximum use of outdoor air for cooling is determined differently for different controllers. Most Honeywell controllers have an ABCD setting on the logic board. Of course, this is only used when a resistance outdoor temperature sensor is utilized. In higher-end Trane models, the change-over point is set with a small DIP switch block on the economizer control board: Switches 1 and 2 on the left edge of the board. The meaning of the switches depends on sensors installed and therefore what economizer logic is being used. What should be recorded is the A – D sensor and whether there is an outdoor thermostat or dial setting (if so, the setting on the dial should also be recorded). In general, for Trane units record the settings of Switch 1 and Switch 2 (whether they are on or off—generally “on” is to the right and “off” is to the left). Setting these switches can be slightly confusing. Technicians are used to four settings (A – D) on the standard dial change-over controls. In the enthalpy mode, when enthalpy sensors are hooked up to the Trane board, these can be mimicked: the D setting (the most conservative setting) can be off-off. The C-setting can be off-on, the B-setting on-off, and the A-setting (the most liberal use of the economizer) is on-on. According to the Voyager book, the off-off setting is the factory default in enthalpy mode. In a confusing twist in the default dry bulb mode, the relative conservativeness or aggressiveness of these settings is different than in the enthalpy mode. The A-setting is on-on or it’s listed as not used, which means that the default cut-out of 75° F is used. The B-setting of Switch 1 on, Switch 2 off is 65° F. For the C- and D-settings, the switches are reversed: off-off is 60°(the C-setting) and off-on is 55° (the D-setting): this is opposite of the enthalpy settings.

A-9

Ecotope recommends that dry bulb change-overs be set at 60° F west of the mountains. Record if any adjustment is made to the change-over setting. Sensor Check-Out Temperature and enthalpy sensors can be checked in a variety of ways depending on the sensor type. Solid-state sensors can be checked if one has access to the voltage vs. temperature chart for the sensor. The Trane Voyager includes this information for the outdoor air and the mixed air dry bulb sensors. Trane was unable to provide similar values for the enthalpy sensor resistance tables. With the unit operating, the resistance across the terminal inputs on the main unit board can be checked for the outdoor air temperature, and the mixed air sensor can be checked by measuring the voltage across its input to the economizer board. This value can then be compared with temperature or enthalpy measurements made by the technician. The installation guides for several Honeywell units have included resistance charts for sensors, which can be checked in a similar way. Electrical/mechanical switches and outdoor sensors with a dial for setting the change-over temperature can be checked by changing the position of the dial and feeling when the sensor clicks: if it clicks close to the correct temperature, the sensor is within the bounds of checkable functionality. Snap disk sensors are a little more difficult. If the outdoor air temperature is cooler, warming the temperature should cause the snap disk to snap. This can be felt or checked with an ohm-meter. In warmer conditions, the sensor would need to be cooled down, perhaps with ice or freon spray. Reportedly, failures of these switches occur not so much as changes in calibration but simple failures to snap. Therefore, this test should verify that the snap disk is in proper order. Mixed air sensors are perhaps more difficult to check, and are just as important or more important to successful economizer operation. Again, these can be resistance or on-off devices. Ecotope has had very little success in finding resistance charts for these sensors. Some units use two snap disks set at 52° F and 57° F to control mixed air temperature. These can be checked as above but on warmer days would need to be cooled rather than heated. Economizer Damper Position Control The key to economizer operation is being able to operate the damper from minimum air to fully open. In most modern units, the damper actuator itself has a feedback potentiometer with a variable resistance so that the unit can sense when it’s at a minimum or maximum setting. The sensing device is integral to the actuator in these cases. The standard Honeywell actuator is an example of this.

A-10

Older style units utilize various methods of mechanical switching to determine fully open, fully closed, and minimum air stops. Some Carrier and Micrometl units use a set of four microswitches along a track. If the damper assembly works at all, this means the controller board and actuator also work (if the system has a separate controller board) or the main board is sending the appropriate signal to the damper actuator. If the damper does not work, the problem can be in the controller board or the actuator (assuming damper operation is attempted by temporarily bypassing sensors). Why is Minimum Outside Air Important? Minimum outside air is used to establish the fresh air intake to the space. Requirements for outside air are determined based on the use of the space and the number of people occupying it. It appears that standard practices in units of this size do not take this information into account, but merely set the unit to some approximate position for a minimal setting. In some situations, however, this position may be set based on some detailed engineering and test-and-balance procedures. In situations where there have been no test-and-balance and there is no engineering located, minimum outside air should probably be set at 10 – 15% in terms of flow. This can be difficult to determine based on the amount of the damper that is open, or on the position of the potentiometer due to various non-linearities in the system. Either the response of the damper to the potentiometer or in the response of the air flow to the damper can be difficult to estimate. The minimum outside air can be set precisely using the True Flow™ plates (follow separate procedure). Once the minimum point is found and measured, record the potentiometer position on the data sheet, along with the minimum airflow as measured. Many technicians will simply fiddle with the minimum air potentiometer to determine whether the economizer damper works, going all the way open and closed. This is problematic in several ways. If the unit is in economizer mode, it will not respond to the minimum potentiometer. Depending on how the unit is setup, this may only demonstrate part of the economizer motor’s range and, in some units with more elaborate digital control systems, minimum outside air may be determined from the main computer system rather than at the unit. Forcing Minimum Position In conditions where cooling signals are coming from the thermostat, the unit must be put into minimum air position: either the call for cooling needs to be removed, the change-over set point reduced to a point that the economizer does not work, or more elaborate procedures employed to defeat the economizer. When the unit is in heating or fan-only mode with no calls for cooling, the economizer should be at the minimum stop when the fan is running. Most economizers will close the dampers completely when the fan is off.

A-11

For units employing the standard Honeywell controller, use the Honeywell check-out procedures directly. It is generally easier to first put the unit in minimum air mode and then to maximum position. The minimum air setting can be checked reliably in several different ways depending on the unit. Trane units with the test board in the unit controls go to minimum position in test Mode 1 (the fan mode). For other units, disconnecting the thermostat wires Y1 and Y2 to remove any calls for cooling should put the unit into a minimum outside air position. These could simply be monitored and, if the unit is in heating or fan mode, then the condition is as desired. If there is a call for cooling, removing the wires Y1 and Y2 should temporarily move the unit to minimum outside air, as long as there is a call for the air handler fan. An alternate approach is to override the outdoor temperature so that the economizer thinks the outside air is unsuitable. This has the advantage of making the thermostat calls unimportant as long as the fan is on. With units that have change-over controls there is a setting dial either on the economizer hood or in the controller. In some conditions, one should turn the dial to a low temperature, causing the unit to switch out of the economizer mode. However, this will not work if a snap disk sensor is connected to a logic board with a A-D setting. The snap disk effectively overrides the A-D setting. Units utilizing the Honeywell logic boards (most Honeywell logic boards are the standard 7459) can be put in the minimum air position by disconnecting the mixed air sensor from terminals T and T1. This essentially tells the unit that it has a very low temperature, which shuts the damper. Damper closure may take a while. Other Honeywell controllers may be slightly different. In general, the installation instructions have a description of setting the minimum position adjustment reliably. Following this procedure should guarantee that the unit is in minimum outside air position and not in economizer mode at any given time. Forcing Maximum Position Several strategies exist to force the economizer to open fully. For Trane units with a test board, putting the unit in the economizer test mode should open the economizer to the 100% outside air position. Some units open to 100% outside air on startup. This should cause it to open completely. For other units, the unit must be fooled into thinking that there is a need for cooling, that the economizer can meet the demand, and that it is having a hard time satisfying the mixed air criteria. Honeywell controllers come with a check-out procedure that simulates this progression of events and involves jumpering and removing spade controllers on their logic board. Ecotope recommends that technicians get comfortable using this procedure. It is often a little cumbersome, as the logic boards are usually in hard-to-access locations and the spade contacts are often very tight. Removal of the spades is easier done if the neck of the spade connector is grabbed with a pair of needle-nosed

A-12

pliers in such a way that the tips extend about 3/8” beyond the spade connector. The tips can then be placed against the base of the unit and the spade can be pried up. Once the minimum air position has been checked, if the ambient temperature is between 55° and 70°, the dial thermostat or potentiometer should be adjusted slightly to see whether the damper responds. Units which do not respond should be noted, and their potentiometers returned to original position. Once a reliable method has been determined to open and close the damper, the damper movement should be closely observed to note any sticking points. The linkage and dampers can be stuck or jammed. Units with Honeywell logic boards that fail in the Honeywell check-out procedure should be noted, along with how they fail in terms of the LED indicator being on or off and the motor opening or closing in the various test modes.

A-13

Airflow Measurement Procedures For measurement through return only: Prior to measurement: • Check that economizer is in desired position (minimum air, all open, all closed, etc.). • Attempt to seal up any bypass gaps downstream of filter slot, except possibly for economizer,

if it is desired to have economizer open fully or partially. These gaps can be around the filter slot, around the economizer, or at connections between the filter slot and the supply side of the blower. Gaps around the blower compartment are especially likely and need attention. Do not worry about gaps around the filter itself.

• Install a static pressure tap firmly on the supply side, with the point facing into the air stream. This tap must be fixed such that it will not move during the entire test. If there is plenum available, this is the ideal location. If it is not available, going in through the blast gate in the blower compartment is acceptable. If this is also not possible, and there is no other supply location available, place the tap as far upstream of the filter on the return side as possible.

• Prepare flow plates sufficient to fill the entire return flow area. There is a laminated card in the flow plate case to give assistance on choosing spacers. For odd sizes it may be necessary to use cardboard to fill in a gap even after spacers are attached. Usually the plates will go directly in the filter slot; however, in some systems with odd configurations of filters it may be necessary to construct a setup that covers the air stream without being in the filter slots. An example is shown below. In the unit shown, it is not possible to use the filter slots because the flow plates are too large to fit into the smaller (vertical) slot (10” x 30”; smallest flow plate requires at least 14”). To measure in this unit, a flow plate setup would have the lower edge fit into the bottom of the vertical slot and have the upper edge fit into the top of the diagonal slot, thus making a single diagonal plane.

A-14

Measurement and calculation: • With filters in place and blower on, measure supply (or return if necessary) plenum pressure

with respect to outside. Enter the result in the table provided. • Remove filters and install flow plates such that the pressure manifold is on the downstream

side of the plate. Try to place the plates in such a way that there is as little of a gap as possible around and between them. Use tape if possible and necessary. Make sure to keep pairs of flow plate hoses together and outside. Take special care not to pinch flow plate hoses off. If plates have spacers attached, and the inlet to the filter slot has a bend, preferentially put the spacers close to the inside of the bend.

• After flow plates are installed, attempt to seal off any possible bypass around plates. One common bypass is between flow plates and filter access cover. Placing tape from the flow plates to the downstream edge of the filter slots can take care of this bypass.

• Replace filter slot cover, or cover filter slot some other way (e.g. tape or cardboard). • With blower on measure pressure drop across each flow plate individually. The red hose

from the plate should go into the input side, the green hose should go into the reference side. If they are reversed the measured pressure will be negative, which is ok (just take the positive) but the only way to get a positive number is to have the hoses set up correctly, so it is a good check. Enter the results in the table provided.

• Measure the supply (or return if necessary) plenum pressure with respect to outside with the flow plates still in place. Enter the result in the table provided.

• Use the calibration equation or the laminated card to calculate the flow through the plates. Enter the results in the table provided.

• Sum all the flows through the plates, and correct the sum for any change in pressure in the plenums caused by the flow plates. The correction multiplier can be obtained either by using the reverse side of the laminated card with the flows through the plates or by calculating the square root of the ratio of the supply (or return) pressure with the filters in place to the pressure with the flow plates in place.

A-15

Airflow Measurement Procedures For simultaneous measurement of return and economizer: Prior to measurement: • Check that economizer is in desired position (minimum air, all open, all closed,

etc.). • Attempt to seal up any bypass gaps downstream of filter slot except for

economizer. These gaps can be around the filter slot, around the economizer, or at connections between the filter slot and the supply side of the blower. Gaps around the blower compartment are especially likely and need attention. Do not worry about gaps around the filter itself.

• Install a static pressure tap firmly on the supply side, with the point facing into the air stream. This tap must be fixed such that it will not move during the entire test. If there is plenum available, this is the ideal location. If it is not available, going in through the blast gate in the blower compartment is acceptable. If this is also not possible, and there is no other supply location available, place the tap as far upstream of the filter on the return side as possible.

• Prepare flow plates sufficient to fill the entire return flow area and the entire economizer area. There is a laminated card in the flow plate case to give assistance on choosing spacers. For odd sizes it may be necessary to use cardboard to fill in a gap even after spacers are attached. On the return, the plates will usually go directly in the filter slot; however, in some systems with odd configurations of filters it may be necessary to construct a setup that covers the air stream without being in the filter slots. An example is shown in the section on measuring return air only. On the economizer, the plates will usually either replace economizer filters or get taped over the entry.

Measurement and calculation: • With filters in place and blower on, measure supply (or return if necessary)

plenum pressure with respect to outside. Enter the result in the table provided. • Remove filters and install flow plates such that the pressure manifold is on the

downstream side of the plate. Try to place the plates in such a way that there is as little of a gap as possible around and between them. Use tape if possible and necessary. Make sure to keep pairs of flow plate hoses together and outside. Take special care not to pinch flow plate hoses off. If plates have spacers attached, and the inlet to the filter slot has a bend, preferentially put the spacers close to the inside of the bend.

• After return flow plates are installed, attempt to seal off any possible bypass around plates. One common bypass is between flow plates and filter access cover. Placing tape from the flow plates to the downstream edge of the filter slots can take care of this bypass.

A-16

• Replace filter slot cover, or cover filter slot some other way (e.g. tape or cardboard).

• Install economizer flow plates, either by removing economizer filters and placing flow plates in filter slot or by taping flow plates over entire entrance to economizer. Make sure to tape off all gaps around the plates.

• With blower on measure pressure drop across each flow plate individually. The red hose from the plate should go into the input side, the green hose should go into the reference side. If they are reversed the measured pressure will be negative, which is ok (just take the positive) but the only way to get a positive number is to have the hoses set up correctly, so it is a good check. Enter the results in the table provided.

• Measure the supply (or return if necessary) plenum pressure with respect to outside with the flow plates still in place. Enter the result in the table provided.

• Use the calibration equation or the laminated card to calculate the flow through the plates. Enter the results in the table provided.

• Sum all the flows through the plates (separate sums for the economizer and return), and correct the sum for any change in pressure in the plenums caused by the flow plates. The correction multiplier can be obtained either by using the reverse side of the laminated card with the flows through the plates or by calculating the square root of the ratio of the supply (or return) pressure with the filters in place to the pressure with the flow plates in place.

For measurement of economizer only: Note: It is less desirable to measure the economizer only because there is less information about the actual fraction of the total flow that is coming in through the economizer. The flow plates may change the total flow through the system, and without measurement of the return flow simultaneously it will not be known if this is the case. However, there may be cases where it is not possible to test both the return and the economizer at the same time, and this is the best that can be done. Prior to measurement: • Check that economizer is in desired position (minimum air, all open, all closed, etc.). • Install a static pressure tap firmly on the supply side, with the point facing into the air

stream. This tap must be fixed such that it will not move during the entire test. If there is plenum available, this is the ideal location. If it is not available, going in through the blast gate in the blower compartment is acceptable. If this is also not possible, and there is no other supply location available, place the tap as far away from the economizer on the return side as possible.

• Prepare flow plates sufficient to fill the entire economizer area. There is a laminated card in the flow plate case to give assistance on choosing spacers. For odd sizes it may be necessary to use cardboard to fill in a gap even after spacers are attached. The plates will usually either replace economizer filters or get taped over the entry.

A-17

A-18

Measurement and calculation: • With filters in place and blower on, measure supply (or return if necessary) plenum

pressure with respect to outside. Enter the result in the table provided. • Install economizer flow plates, either by removing economizer filters and placing flow

plates in filter slot or by taping flow plates over entire entrance to economizer. Make sure to tape off all gaps around the plates.

• With blower on measure pressure drop across each flow plate individually. The red hose from the plate should go into the input side, the green hose should go into the reference side. If they are reversed the measured pressure will be negative, which is ok (just take the positive) but the only way to get a positive number is to have the hoses set up correctly, so it is a good check. Enter the results in the table provided.

• Measure the supply (or return if necessary) plenum pressure with respect to outside with the flow plates still in place. Enter the result in the table provided.

• Use the calibration equation or the laminated card to calculate the flow through the plates. Enter the results in the table provided.

• Sum all the flows through the plates, and correct the sum for any change in pressure in the plenums caused by the flow plates. The correction multiplier can be obtained either by using the reverse side of the laminated card with the flows through the plates or by calculating the square root of the ratio of the supply (or return) pressure with the filters in place to the pressure with the flow plates in place.

Appendix B:

Duct Loss Discussion

B-1

The mathematical model for estimation of duct efficiency consists of three primary parts. The first part is the delivery efficiency, which is the fraction of the output capacity that entered the conditioned space via the registers. The second part is regain, which represents losses from the ducts that still manage to enter the conditioned space by other means, such as by conduction across the ceiling or through leaks between the buffer space and the conditioned space. The final part is an infiltration interaction term, which represents the change in conditioning energy required due to unbalanced duct leakage either pressurizing or depressurizing the space, thereby changing the amount of outside air that comes in and which must then be conditioned. In the most rigorous evaluation of regain, there is also an infiltration interaction term relating to the buffer space. This model was developed in the residential context, however, and though the delivery efficiency and regain should carry over to commercial buildings fairly well, the infiltration interaction part will likely not, either for the conditioned space or the buffer space. For strip malls, the reason is that the door will likely be opened frequently by customers. For single, large stores, the reason is both the entrance of customers and the sheer size of the building; unbalanced leakage of a single HVAC unit will not have a significant impact on the infiltration of the whole building. For office buildings, the problem is one of both size and zoning, e.g. many offices won’t have any connection to outside. For buffer spaces, the infiltration interaction will likely not apply in the commercial setting is because the buffer spaces are most often continuous, meaning that all units share the same large space. Not only will unbalanced leakage from a single unit not likely affect the infiltration of this large a space, but it is possible that a supply leak from one unit is pulled into another unit by a return leak, and there may not be much unbalanced leakage for the buffer space as a whole. As a result of these issues, infiltration interaction was ignored when considering the magnitude of duct losses in these buildings. Finally the t-bar ceilings are very porous and the connection to the main space should be very good. Because of the lack of detailed duct loss measurements at any of the buildings evaluated in the field study, it is not possible to estimate the actual duct efficiency and resulting energy penalty for these buildings. However, it is possible to look at efficiency estimates for a range of assumed loss levels, and investigate the sensitivity of these estimates to various changes in the assumptions. It is also possible to discuss the degree to which duct losses might be recovered with different buffer space characteristics, and to identify buildings that have these characteristics. The amount of conditioning energy that is lost from the ducts prior to the supply registers can be represented by the following equation:

( ) ( )e

sss

e

rrrssss T

TTT

∆∆

β−α−∆∆

βα−βα−βα=η 110

B-2

where η0 delivery efficiency (fraction of conditioning energy that enters via the supply registers) αs supply leakage efficiency (fraction of supply air that is not lost due to leakage) βs supply conduction efficiency αr return leakage efficiency (fraction of return air that is not lost due to leakage) βr return conduction efficiency ∆Ts temperature difference between the conditioned space and the supply buffer space ∆Tr temperature difference between the conditioned space and the return buffer space ∆Te temperature change across the equipment Table B-1 shows the results using a variety of inputs to this equation. These results are not intended to reflect the efficiency of any specific unit, but rather are used to show the impact of losses that are of reasonable size when combined with the other parameters. All of the results assume an indoor temperature of 75 F, a temperature change across the coil of 20 F (∆Te = -20), and that the supply and return ducts are all in the same buffer space. The fifth column shows the delivery efficiency with a buffer space temperature of 90 F (∆Ts = ∆Tr = -15), while the final column shows the delivery efficiency with a buffer space temperature of 95 F (∆Ts = ∆Tr = -20). None of these results consider solar gains due to heating of ducts on top of the roof. This table shows that, even for component losses that are not particularly large, a substantial amount of conditioning energy can be lost from ducts. This is in part due to the low temperature changes caused by coils. For larger temperature changes, as are experienced in furnaces, the last two terms of the efficiency equation become less important and the efficiency increases. It is also of interest to note that, when the temperature difference between the occupied space and the buffer space is the same as the temperature change across the coil, the return leakage has an impact comparable to the supply leakage with all other parameters held constant. The supply leakage is more important when the buffer space is closer to the temperature in the occupied space. Based on this table even relatively tight ducts with only modest leakage to outside could be expected to have duct efficiencies between 75 and 85% (15-25% losses).

B-3

Table B-1. Delivery efficiency estimates with various inputs Supply Return ∆Ts = ∆Tr = -15 ∆Ts = ∆Tr = -20

Leak. Eff. Cond. Eff. Leak. Eff. Cond. Eff. Del. Eff. Del. Eff.

1.0 0.95 0.9 0.95 0.809 0.762 0.9 0.95 0.9 0.95 0.728 0.686 0.8 0.95 0.9 0.95 0.647 0.610 0.7 0.95 0.9 0.95 0.566 0.534

0.9 1.00 0.9 0.95 0.802 0.769 0.9 0.95 0.9 0.95 0.728 0.686 0.9 0.90 0.9 0.95 0.654 0.603 0.9 0.85 0.9 0.95 0.581 0.519

0.9 0.95 1.0 0.95 0.789 0.767 0.9 0.95 0.9 0.95 0.728 0.686 0.9 0.95 0.8 0.95 0.667 0.605 0.9 0.95 0.7 0.95 0.606 0.524

0.9 0.95 0.9 1.00 0.757 0.724 0.9 0.95 0.9 0.95 0.728 0.686 0.9 0.95 0.9 0.90 0.699 0.648 0.9 0.95 0.9 0.85 0.671 0.609

One additional feature of importance is that the supply conduction loss has about twice as much of an impact as a supply leakage loss of the same size. Fortunately, many commercial systems do not have extensive supply duct runs, and the air velocity can be quite high, so the conduction losses tend to be quite small unless the ducts are completely uninsulated. It is also important to note that, if the buffer space temperature is the same as the temperature in the conditioned space, the return losses have no effect regardless of the extent to which the buffer space is connected to the conditioned space thermally or with regard to air flow, and the efficiency is simply the product of the supply leakage and conduction efficiencies. While the magnitude of the losses may appear large, it is important to remember that many of the losses may be recovered via communication between the buffer space and the occupied space. These losses will only reflect the actual losses if the ducts are outside or if the buffer space is effectively fully outside. If there is no buffer space all of the losses will be recovered. For other buildings, the amount of the losses that are recovered depends on the buffer space characteristics.

B-4

At one extreme is a buffer space that is both unvented and insulated at the roof. Unless there are large unintentional leaks to outside in these buffer spaces, the buffer space will be much better connected to the occupied space than to outside, and the losses will be mostly recovered. At the other extreme is a buffer space that is well-vented and has insulation only at the ceiling. In this case the buffer space will be almost entirely outdoors, and almost none of the losses will be recovered. In between these two extremes there are a wide range of possibilities. A buffer space that is vented to the outside but also has the insulation at the roof will be better connected with the occupied space thermally, but in terms of air flow will be much better connected to outside. In contrast, a buffer space that has the insulation at the ceiling and is unvented will probably be significantly connected to the occupied space in terms of air flow, but will be better connected to outdoors thermally. If there is insulation at both the ceiling and the roof, the buffer space will be thermally connected to the occupied space and outdoors about equally, assuming the insulation levels are about the same. Assigning a specific fraction of recovered losses to these in-between buffer spaces requires detailed investigation on a case-by-case basis. As a result, it is not possible to assign a fraction to a certain “type” of buffer space. The tables showing efficiencies with half of the losses recovered are therefore not intended to apply to a specific set of buffer space characteristics, but rather illustrate how much of an impact a partial connection to the occupied space could have.

B-5