Energy Management
Content
A S H RA E
JOURNAL
The following article was published in ASHRAE Journal, December 1998. © Copyright 1998 American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. It is presented for educational purposes only. This article may not be copied and/or distributed electronically or in paper form without permission of ASHRAE.
At left is the office building that used an automated energy management system for 4.5 years. The top photo shows the building’s two constant volume packaged rooftop units that provide heating and cooling.
Energy Management Impact In a Small Office Building
By Jack S. Wolpert, Ph.D.
Member ASHRAE and
Gary F. Schroeder
Associate Member ASHRAE
M
ost studies on the effectiveness of energy conservation measures focus on past energy consumption and savings achieved after implementation. This article presents an unusual case where the effectiveness of an energy management system (EMS) was observed both before installation and after the system was removed. The subject of this study is a small office building. It covers a period of twelve years beginning with a feasibility study in 1986. After an initial demonstration in 1987 in which proposed EMS functions were performed manually, the EMS was installed About the Authors
Jack S. Wolpert, Ph.D., is president of E-Cube in Boulder, Colo. He is a member of ASHRAE Standards Guideline Project Committee 14P, Measurement of Energy and Demand Savings. Gary F. Schroeder was working as an energy analysis and commissioning engineer with E-Cube when this study was carried out. Currently, he works at VERIS, a manufacturer of flow measuring devices in Boulder, Colo.
with expectations of improved occupant comfort and reduced utility bills. The EMS was then operated at various levels of supervision for a period of 4.5 years from 1988 through the fall of 1992. After a failure of one of the field panels, the decision was made to remove the system and install standard electric controls consisting of time clocks, 24 Vac auto-changeover master thermostats and economizer controls. The whole-facility energy use for a four-year period (1993– 97) after the removal of the EMS is compared with the EMS operation period. The results show that the energy consumption and peak electrical demand increased significantly after the removal of the EMS.
Background The facility is a three-story office building in Boulder, Colo. totaling 12,750 ft2 (1251 m2). Built in 1972, it is made of brick and glass with some insulation and large areas of single-pane glass with mirrored finish to reduce solar gains. The typical climate in Boulder consists of an average of 6,406 heating degree days, base 65°F (18.3°C) (HDD65) and 954 cooling degree days, base 65°F (18.3°C) (CDD65). The ASHRAE 1% design conditions are 2°F (–16.7°C) dry-bulb temperature in winter, and 93°F (33.9°C) dry-bulb temperature, 65°F (18.3°C) wet-bulb temperature in summer. Extreme weather conditions of –15°F (–26.1°C) dry-bulb temperature in the winter and 99°F (37.2°C) dry-bulb temperature in the summer are not uncommon. Typically, low humidity levels and cool nights make for
ASHRAE Journal 41
December 1998
Advertisement in the print edition formerly in this space.
ENERGY MANAGEMENT
ideal economizer operation. In addition, at an elevation of 5,445 ft (1660 m), in a climatic zone with low cloud cover, the solar gains can be significant, especially in the spring and fall when the sun angles are lower. The building is used almost entirely for offices with the majority of occupancy occurring between 8 a.m. and 6 p.m., Monday through Friday. Two constant-volume packaged rooftop units (RTUs) provide heating and cooling, each equipped with gas heat (high/low firing capability), dual R-22 hermetic compressors, and airside economizers. While the RTUs provide gaspowered heat, additional heating is provided by electric resistance heat in the terminal boxes of each zone. The building is zoned so that the east RTU serves all three floors on the east side and the west RTU serves all three floors on the west side. The east and west RTUs were each controlled (pre- and post-EMS) by a single thermostat. These two main zones are further subdivided to serve the offices. Approximately 30 room thermostats control the electric re-heat boxes (see next section for RTU operation). The design allows for simultaneous cooling (RTU) and heating (electric re-heats). This condition can occur when occupants are allowed access to the RTU and zone thermostats, especially if the facilities staff is not alert to changes in the settings. Occupant changes can result in a potentially large increase in electric utility usage and costs.
Existing Operation Prior to the installation of the EMS, the two RTUs were controlled with zone thermostats during occupied periods and a timer to cycle the units off during unoccupied periods. During the initial audit, it was discovered that the heating setpoints for the thermostats that controlled the gas-fired heating equipment were kept relatively low. This was a direct result of the following conditions: 1) The cooling setpoints were set low to meet the cooling requirements of one or two warm zones; 2) The occupants preferred cool temperatures in the spaces that had the two controlling thermostats; and 3) Occupants turned down thermostats on warm afternoons and forgot to re-adjust them (in general, comfort could be maintained through use of the re-heat). Because of the use of auto-changeover thermostats, this low cooling setpoint automatically reduced the heating setpoint. The effect of the combination of these conditions during the heating season was that the RTU gas heat would raise the room temperature to no more than 68°F to 70°F (20°C to 21°C) at which time the electric heat in most zones would engage to raise the temperature to their 73°F to 75°F (22°C to 23°C) setpoints. These conditions led to poor occupant comfort and high utility bills. When operating in cooling mode, this low cooling setpoint similarly resulted in extensive use of electric re-heat. Both the mechanical system and the existing operation are discussed in de-
Advertisement in the print edition formerly in this space.
Advertisement in the print edition formerly in this space.
December 1998
ASHRAE Journal
43
tail in a previous publication.1 In addition, it was discovered during the initial audit that the HVAC time clocks were no longer in service, causing the units to operate continuously seven days per week. The time clocks had been bypassed to eliminate problems associated with fan airflow switches that incorrectly restricted the electric reheat from operating.
EMS Operation Following the feasibility study in late 1986, a new control strategy that was expected to improve comfort and reduce energy costs was tested as part of an aggressive Planned Maintenance (PM) program. In this manual EMS phase, the heating and cooling setpoints on the two thermostats controlling the RTUs were raised to the warm day setting of approximately 75°F to 76°F (23.9°C to 24.4°C) cooling (from a setting of 70°F to 72°F [21°C to 22°C] or lower, depending on the last setting) and 73°F to 74°F (22.8°C to 23.3°C) heating (from 68°F to 70°F [20°C to 21.1°C]). Thermostat covers were locked and settings were checked weekly by the PM contractor (a necessary step, as the occupants proved resourceful at bypassing the thermostat locks). The time clocks were also returned to service for this period. The manual EMS phase was continued through approximately Sept. 1987 when reduced utility bills proved the efficacy of this control strategy. By this time, it was decided to install the EMS to provide more refined and reliable control, higher occupant comfort and lower utility bills. In the latter part of 1987 and early 1988 the electronic EMS was requisitioned and installed. Significantly less attention was paid to the manual EMS measures during this transition to the new electronic EMS. The EMS was de-bugged and tuned throughout most of 1988. A few of the more significant control strategies performed by the EMS were as follows:1 • Performing seven day/holiday time clock functions. • Using input from a number of zones (as opposed to the two zones used previously) to control the RTUs. • Duty-cycling the electric heat, electric domestic hot water and electric sump heaters for the hydraulic elevator. • Preventing AC compressors (two per
44 ASHRAE Journal
Figure 1: Comparison of kWh and cooling degree days base 65°F (18.3°C).
Figure 2: Annual average and peak electrical demand.
RTU) from operating when economizer was sufficient. • Locking out the electric re-heat when the second-stage AC compressor was enabled. • Using a winter morning warm-up routine that prevented electric heat operation and raised the gas heating setpoint 1°F above normal to further minimize the need for re-heat when the system went to occupied mode. During 1989, the EMS was closely supervised and thus its performance was near optimal. Through 1990 and the first half of 1991, the EMS continued to operate with moderate supervision. Through the end of 1991 and into 1992 there was
little supervision of the EMS. In the fall of 1992, one of the microprocessor boards failed. Although the proposed repair cost was only $1,500, the decision was made by the owner to remove the EMS and install standard electric controls. Since then, the building has been running on standard controls consisting of the previously mentioned time clocks, 24 Vac auto-changeover master thermostats and economizer controls. During this period, facility operation and other loads were fairly constant. A small computer training lab (approximately 20 to 25 units) was added, which slightly increased plug loads, but other usage remained relatively constant.
December 1998
ENERGY MANAGEMENT
Analysis The analysis was done using monthly utility billing data. The EMS period was defined as the six-year period from 1987– 92. Note that although 1987 was the manual EMS year, it is included because many of the functions performed manually were later performed by the EMS. The non-EMS period included 1986 and 1993– 97 (one year before installation of the EMS and five years after its removal). To establish a relationship with which to compare energy use, electric kilowatt hours (kWh) and natural gas hundred cubic feet (CCF) were compared to annual cooling degree days (CDD65) and heating degree days (HDD65) respectively (using historical weather data for Denver). For cooling, the ratio of kWh/CDD was found to be 42% less during the EMS period compared to the non-EMS period. The ratio of CCF/HDD was found to be 21% less during the EMS period. These EMS and non-EMS ratios were used to estimate average energy use for the two conditions. By taking the average CDDs or HDDs and multiplying them by the respective ratio, the average energy can be estimated. To calculate the change in demand, the average annual demand was estimated for each of the two periods by dividing the annual demand for each year (sum of peak demand values for 12 months) by the number of years. The difference between these two averages resulted in the estimated annual demand savings. The estimated annual electric energy and demand and gas energy was used to calculate the approximate annual cost savings attributable to the EMS. This was calculated using the 1986 through mid1996 rates of $0.03468/kWh, $10.537/kW, and $0.3382/CCF ($0.1194/m3). These costs represent base utility charges with riders and a franchise fee, but no sales tax. Note that although a rate change occurred in mid-1996, the same rates were used throughout the analysis in order to have consistent data (actual utility costs in 1996 and 1997 were higher under the new rates). Results Figures 1–4 show the annual electric and gas consumption, heating and cooling degree days and total utility costs for the period from 1986–97. The period of
December 1998
Figure 3: Comparison of gas hundred cubic feet (CCF) and heating degree days (HDD) base 65°F (18.3°C).
Figure 4: Annual utility costs during the EMS period and after its removal.
EMS operation is shown in each figure. Note that the cost data have not been normalized for differences in heating and cooling degree day values (refer to the Dicussion section). The effects of the different phases of the project can be seen in each of the graphs. Of note in Figure 1, is the reduction in electrical energy use during the period from 1987 through 1992 and the dramatic increase in energy usage in 1993 and beyond. In Figure 2, the average annual demand is relatively flat from 1986–92 and increases significantly in 1993, 1994 and 1997. The drop-off in 1995 and 1996 pre-
sumably represents better adjustment of the new controls (although the average is still above the EMS period and electric and gas energy remain significantly above the earlier values as well). The effect on peak annual demand is even more dramatic where high demand is seen in the periods before installation as well as in three of the five years after removal of the EMS. The natural gas usage shown in Figure 3, shows a slight increase through 1992 (especially in 1991 and 1992), for unknown reasons. This is followed by a sharp increase in 1993 with standard controls back in operation. Figure 4 shows a significant increase in utility costs after removal of the
ASHRAE Journal 45
EMS in 1992. The values used in Figures 1–4 are summarized in Table 1. Table 1 shows the average utility use and change for periods with and without the EMS (as described previously). Note that while electric energy consumption without the EMS increased 52%, the increase per degree day was 71%. For natural gas the increase in consumption without the EMS was 26% while the increase per degree day was also 26%. Finally, the overall increase in cost was 38%.
A nnua l A v e r a g e Co m m o d i t y Electric Energy Electric Demand Natural Gas Un i t s kWh kW CCF ( m3 ) Cost kWh/CDD CCF/HDD ( m3 / H D D ) w i t h EM S 2 0 7 ,3 0 7 9 2 .8 8 ,6 6 7 (2 4 5 4 5 ) $ 1 7 ,8 3 5 2 0 8 .2 1 .3 9 3 .9 4 w i t hout EM S 3 1 4 ,5 2 0 1 1 2 .4 1 0 ,9 3 7 (3 0 9 7 4 ) $ 2 4 ,5 3 0 3 5 6 .8 1 .7 6 4 .9 7 I n cr e a s e w i t h o u t EM S 1 0 7 ,2 1 3 1 9 .6 2 ,2 7 0 (6 4 2 9 ) $ 6 ,6 9 5 1 4 8 .6 0 .3 6 1 .0 3 38% 71% 26% 52% 21% 26%
Discussion As shown by the data in the previous section, a significant difference exists in the building’s utility cost with and without the EMS. Of note are the relative changes in all three indices (gas and electricity usage and electric demand). As mentioned in the introduction, originally (1986 and previous) there was heavy reliance on electric reheat. Although significant, this reliance appears to have increased even more in the post-EMS period. One of the functions of the EMS was to reduce the amount of electric reheat by increasing the amount of gas heating. The utility bills from the pre-EMS period (1986) to the EMS period (1987–92) show this change with a reduction in electrical usage and an increase in gas usage. One would expect this relationship to reverse once the EMS was disconnected (reverting to 1986 operation). Surprisingly, electric and gas usage increased in the post-EMS period which added to the detrimental effects of removing the system. After the removal of the EMS in 1992, the jump in utility use to a level much higher than the baseline year of 1986 may be explained by the following. While pre-EMS controls operation (1986 and earlier) was not as efficient as the EMS-based strategies, there was meaningful interest in conserving where possible. This awareness led to the discovery of the malfunctioning time clocks, provided the motivation and follow through for installation of an EMS and other efforts (e.g., utility usage tracking/reporting) during that time. However, the operation and control package installed postEMS appears to have focused significantly less on efficient operation than did the pre-EMS system. Perhaps this was due to a loss of understanding of the energy efficiency issues during the years the EMS “automatically” maintained control. Based on similar types of patterns in 1996, the net result is probably causing very significant over cooling and electric re-heating in summer. Advantages such as staging of the two-stage compressors and efficient airside economizer operation may also be lost. The period from 1995–97 shows a general trend of gas usage tracking degree days, but it also shows that electric usage was quite unpredictable. This seems to indicate that overall costs increase without the EMS and demonstrates the system’s vulnerability to changes in control adjustments. As a result, utility usage forecasts are more likely to show wide fluctuations. Improved occupant comfort also was a goal for the EMS. While this was not surveyed and can be quite hard to accurately assess, a few comments can be offered. First, occupant comfort complaints were significantly reduced in the first two years of EMS usage (when this was tracked to some degree).
46 ASHRAE Journal
Table 1: Annual average summary with and without EMS.
Much of the improvement came from eliminating large changes in setpoint, and thus temperature “swings,” for the RTU controlling thermostats. Even with reheat, the system could not always respond adequately. A second area of improvement came from proper morning warm-up/pre-cooling resulting in occupants arriving to a comfortable building. Finally, while the higher cooling mode setpoints (typically 75°F to 76°F [23.9°C to 24.4°C]) did generate some complaints in localized areas, the overall balance between zones and the ability of the EMS to observe all zones and reset temperatures accordingly did much to offset this issue. For this study, only a simple form of weather adjustment is applied to the data. As shown in Table 1, the usage per HDD or CDD is reported. While more sophisticated forms of weather correction could be applied, this study does not need this level of analysis. The overall pre- and post-EMS trends are clear and the annual weather values did not vary all that much. While a more detailed analysis (including monthly data, adjustments for non-HVAC usage, etc.), would yield interesting information and probably would explain small variations between some years, the overall conclusions likely would not change.
Conclusions The removal of the EMS caused an average increase of 38% in annual utility costs (from $1.40 to $1.92 per square foot). Viewed in reverse, the “savings” of 27% that were realized in going from standard controls to EMS operation are on the upper end, but are within reason based upon the authors’ experience. This is understandable when considering that the system is a constant volume system with electric reheat and that the $1.92 per square foot utility costs are on the high end of the expected range for this type of facility in this region. These findings are also consistent with published reports where these types of measures were implemented through better utilization of the existing EMS.3, 4 The annual increase of approximately $6,695/year, associated with the removal of the EMS, would have paid for the $1,500 repair needed in 1992 in less than three months. Had the EMS been repaired in 1992, the agency would have saved about
December 1998
ENERGY MANAGEMENT
$33,000 in utility costs (with the repair factored in, but without any differential O&M costs) from 1992–97. The authors suspect that the decision to remove the EMS was made for reasons other than technical or financial merit. These reasons may have included: 1) Ease of maintenance with existing staff; 2) Inability or unwillingness to support the EMS with a service contract; 3) Lack of knowledge of the magnitude of the financial impact; and 4) Lack of control by occupants and facility staff. A similar lack of understanding and utilization of EMS have been reported by others.5 The results of this study point to an important conclusion. Understanding the HVAC system and its proper operation often offers the potential for large utility savings. This was shown in 1987, when the manual EMS concept was tested. Much of the utility savings obtained from the installation of the EMS resulted from these measures, although the labor costs, effort, required expertise and dedication were quite high and unlikely to be sustained for long. Additionally, once the EMS was installed, investing in occasional supervision and infrequent repairs would likely have extended its life and led to prolonged utility savings. It should be noted that an active role by maintenance personnel in optimizing the EMS was responsible, in large part, for the magnitude of savings obtained from it.
Acknowledgments We would like to thank Mark Bowman of E-Cube for his contributions in the analysis and review of this article. References
1. Wolpert, J., S. Wolpert and G. Martin. 1992. Optimization of a dual-fuel heating system utilizing an EMS to maintain persistence of measures. 15th World Energy Engineering Congress, Atlanta. 3. Wolpert, J. and C. Robbins. 1991. Improving an older building through field audits and operational changes. Proceedings, Fifth Forum on Practical Applications of Building Energy Efficiency. Dallas. 4. Liu, M., Y. Zhu and D. Claridge. 1997. “Use of EMCS recorded data to identify potential savings due to improved HVAC operations and maintenance.” ASHRAE Transactions 103(2):4063. 5. Yago, J. 1992. “Survey shows users often fail to make the most of automation.” Energy User News, (4):25. Please circle the appropriate number on the Reader Service Card at the back of the publication. Extremely Helpful ....................................................... 458 Helpful ..................................................................... 459 Somewhat Helpful ..................................................... 460 Not Helpful ............................................................... 461
Advertisement in the print edition formerly in this space.
December 1998
ASHRAE Journal
47
JOURNAL
The following article was published in ASHRAE Journal, December 1998. © Copyright 1998 American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. It is presented for educational purposes only. This article may not be copied and/or distributed electronically or in paper form without permission of ASHRAE.
At left is the office building that used an automated energy management system for 4.5 years. The top photo shows the building’s two constant volume packaged rooftop units that provide heating and cooling.
Energy Management Impact In a Small Office Building
By Jack S. Wolpert, Ph.D.
Member ASHRAE and
Gary F. Schroeder
Associate Member ASHRAE
M
ost studies on the effectiveness of energy conservation measures focus on past energy consumption and savings achieved after implementation. This article presents an unusual case where the effectiveness of an energy management system (EMS) was observed both before installation and after the system was removed. The subject of this study is a small office building. It covers a period of twelve years beginning with a feasibility study in 1986. After an initial demonstration in 1987 in which proposed EMS functions were performed manually, the EMS was installed About the Authors
Jack S. Wolpert, Ph.D., is president of E-Cube in Boulder, Colo. He is a member of ASHRAE Standards Guideline Project Committee 14P, Measurement of Energy and Demand Savings. Gary F. Schroeder was working as an energy analysis and commissioning engineer with E-Cube when this study was carried out. Currently, he works at VERIS, a manufacturer of flow measuring devices in Boulder, Colo.
with expectations of improved occupant comfort and reduced utility bills. The EMS was then operated at various levels of supervision for a period of 4.5 years from 1988 through the fall of 1992. After a failure of one of the field panels, the decision was made to remove the system and install standard electric controls consisting of time clocks, 24 Vac auto-changeover master thermostats and economizer controls. The whole-facility energy use for a four-year period (1993– 97) after the removal of the EMS is compared with the EMS operation period. The results show that the energy consumption and peak electrical demand increased significantly after the removal of the EMS.
Background The facility is a three-story office building in Boulder, Colo. totaling 12,750 ft2 (1251 m2). Built in 1972, it is made of brick and glass with some insulation and large areas of single-pane glass with mirrored finish to reduce solar gains. The typical climate in Boulder consists of an average of 6,406 heating degree days, base 65°F (18.3°C) (HDD65) and 954 cooling degree days, base 65°F (18.3°C) (CDD65). The ASHRAE 1% design conditions are 2°F (–16.7°C) dry-bulb temperature in winter, and 93°F (33.9°C) dry-bulb temperature, 65°F (18.3°C) wet-bulb temperature in summer. Extreme weather conditions of –15°F (–26.1°C) dry-bulb temperature in the winter and 99°F (37.2°C) dry-bulb temperature in the summer are not uncommon. Typically, low humidity levels and cool nights make for
ASHRAE Journal 41
December 1998
Advertisement in the print edition formerly in this space.
ENERGY MANAGEMENT
ideal economizer operation. In addition, at an elevation of 5,445 ft (1660 m), in a climatic zone with low cloud cover, the solar gains can be significant, especially in the spring and fall when the sun angles are lower. The building is used almost entirely for offices with the majority of occupancy occurring between 8 a.m. and 6 p.m., Monday through Friday. Two constant-volume packaged rooftop units (RTUs) provide heating and cooling, each equipped with gas heat (high/low firing capability), dual R-22 hermetic compressors, and airside economizers. While the RTUs provide gaspowered heat, additional heating is provided by electric resistance heat in the terminal boxes of each zone. The building is zoned so that the east RTU serves all three floors on the east side and the west RTU serves all three floors on the west side. The east and west RTUs were each controlled (pre- and post-EMS) by a single thermostat. These two main zones are further subdivided to serve the offices. Approximately 30 room thermostats control the electric re-heat boxes (see next section for RTU operation). The design allows for simultaneous cooling (RTU) and heating (electric re-heats). This condition can occur when occupants are allowed access to the RTU and zone thermostats, especially if the facilities staff is not alert to changes in the settings. Occupant changes can result in a potentially large increase in electric utility usage and costs.
Existing Operation Prior to the installation of the EMS, the two RTUs were controlled with zone thermostats during occupied periods and a timer to cycle the units off during unoccupied periods. During the initial audit, it was discovered that the heating setpoints for the thermostats that controlled the gas-fired heating equipment were kept relatively low. This was a direct result of the following conditions: 1) The cooling setpoints were set low to meet the cooling requirements of one or two warm zones; 2) The occupants preferred cool temperatures in the spaces that had the two controlling thermostats; and 3) Occupants turned down thermostats on warm afternoons and forgot to re-adjust them (in general, comfort could be maintained through use of the re-heat). Because of the use of auto-changeover thermostats, this low cooling setpoint automatically reduced the heating setpoint. The effect of the combination of these conditions during the heating season was that the RTU gas heat would raise the room temperature to no more than 68°F to 70°F (20°C to 21°C) at which time the electric heat in most zones would engage to raise the temperature to their 73°F to 75°F (22°C to 23°C) setpoints. These conditions led to poor occupant comfort and high utility bills. When operating in cooling mode, this low cooling setpoint similarly resulted in extensive use of electric re-heat. Both the mechanical system and the existing operation are discussed in de-
Advertisement in the print edition formerly in this space.
Advertisement in the print edition formerly in this space.
December 1998
ASHRAE Journal
43
tail in a previous publication.1 In addition, it was discovered during the initial audit that the HVAC time clocks were no longer in service, causing the units to operate continuously seven days per week. The time clocks had been bypassed to eliminate problems associated with fan airflow switches that incorrectly restricted the electric reheat from operating.
EMS Operation Following the feasibility study in late 1986, a new control strategy that was expected to improve comfort and reduce energy costs was tested as part of an aggressive Planned Maintenance (PM) program. In this manual EMS phase, the heating and cooling setpoints on the two thermostats controlling the RTUs were raised to the warm day setting of approximately 75°F to 76°F (23.9°C to 24.4°C) cooling (from a setting of 70°F to 72°F [21°C to 22°C] or lower, depending on the last setting) and 73°F to 74°F (22.8°C to 23.3°C) heating (from 68°F to 70°F [20°C to 21.1°C]). Thermostat covers were locked and settings were checked weekly by the PM contractor (a necessary step, as the occupants proved resourceful at bypassing the thermostat locks). The time clocks were also returned to service for this period. The manual EMS phase was continued through approximately Sept. 1987 when reduced utility bills proved the efficacy of this control strategy. By this time, it was decided to install the EMS to provide more refined and reliable control, higher occupant comfort and lower utility bills. In the latter part of 1987 and early 1988 the electronic EMS was requisitioned and installed. Significantly less attention was paid to the manual EMS measures during this transition to the new electronic EMS. The EMS was de-bugged and tuned throughout most of 1988. A few of the more significant control strategies performed by the EMS were as follows:1 • Performing seven day/holiday time clock functions. • Using input from a number of zones (as opposed to the two zones used previously) to control the RTUs. • Duty-cycling the electric heat, electric domestic hot water and electric sump heaters for the hydraulic elevator. • Preventing AC compressors (two per
44 ASHRAE Journal
Figure 1: Comparison of kWh and cooling degree days base 65°F (18.3°C).
Figure 2: Annual average and peak electrical demand.
RTU) from operating when economizer was sufficient. • Locking out the electric re-heat when the second-stage AC compressor was enabled. • Using a winter morning warm-up routine that prevented electric heat operation and raised the gas heating setpoint 1°F above normal to further minimize the need for re-heat when the system went to occupied mode. During 1989, the EMS was closely supervised and thus its performance was near optimal. Through 1990 and the first half of 1991, the EMS continued to operate with moderate supervision. Through the end of 1991 and into 1992 there was
little supervision of the EMS. In the fall of 1992, one of the microprocessor boards failed. Although the proposed repair cost was only $1,500, the decision was made by the owner to remove the EMS and install standard electric controls. Since then, the building has been running on standard controls consisting of the previously mentioned time clocks, 24 Vac auto-changeover master thermostats and economizer controls. During this period, facility operation and other loads were fairly constant. A small computer training lab (approximately 20 to 25 units) was added, which slightly increased plug loads, but other usage remained relatively constant.
December 1998
ENERGY MANAGEMENT
Analysis The analysis was done using monthly utility billing data. The EMS period was defined as the six-year period from 1987– 92. Note that although 1987 was the manual EMS year, it is included because many of the functions performed manually were later performed by the EMS. The non-EMS period included 1986 and 1993– 97 (one year before installation of the EMS and five years after its removal). To establish a relationship with which to compare energy use, electric kilowatt hours (kWh) and natural gas hundred cubic feet (CCF) were compared to annual cooling degree days (CDD65) and heating degree days (HDD65) respectively (using historical weather data for Denver). For cooling, the ratio of kWh/CDD was found to be 42% less during the EMS period compared to the non-EMS period. The ratio of CCF/HDD was found to be 21% less during the EMS period. These EMS and non-EMS ratios were used to estimate average energy use for the two conditions. By taking the average CDDs or HDDs and multiplying them by the respective ratio, the average energy can be estimated. To calculate the change in demand, the average annual demand was estimated for each of the two periods by dividing the annual demand for each year (sum of peak demand values for 12 months) by the number of years. The difference between these two averages resulted in the estimated annual demand savings. The estimated annual electric energy and demand and gas energy was used to calculate the approximate annual cost savings attributable to the EMS. This was calculated using the 1986 through mid1996 rates of $0.03468/kWh, $10.537/kW, and $0.3382/CCF ($0.1194/m3). These costs represent base utility charges with riders and a franchise fee, but no sales tax. Note that although a rate change occurred in mid-1996, the same rates were used throughout the analysis in order to have consistent data (actual utility costs in 1996 and 1997 were higher under the new rates). Results Figures 1–4 show the annual electric and gas consumption, heating and cooling degree days and total utility costs for the period from 1986–97. The period of
December 1998
Figure 3: Comparison of gas hundred cubic feet (CCF) and heating degree days (HDD) base 65°F (18.3°C).
Figure 4: Annual utility costs during the EMS period and after its removal.
EMS operation is shown in each figure. Note that the cost data have not been normalized for differences in heating and cooling degree day values (refer to the Dicussion section). The effects of the different phases of the project can be seen in each of the graphs. Of note in Figure 1, is the reduction in electrical energy use during the period from 1987 through 1992 and the dramatic increase in energy usage in 1993 and beyond. In Figure 2, the average annual demand is relatively flat from 1986–92 and increases significantly in 1993, 1994 and 1997. The drop-off in 1995 and 1996 pre-
sumably represents better adjustment of the new controls (although the average is still above the EMS period and electric and gas energy remain significantly above the earlier values as well). The effect on peak annual demand is even more dramatic where high demand is seen in the periods before installation as well as in three of the five years after removal of the EMS. The natural gas usage shown in Figure 3, shows a slight increase through 1992 (especially in 1991 and 1992), for unknown reasons. This is followed by a sharp increase in 1993 with standard controls back in operation. Figure 4 shows a significant increase in utility costs after removal of the
ASHRAE Journal 45
EMS in 1992. The values used in Figures 1–4 are summarized in Table 1. Table 1 shows the average utility use and change for periods with and without the EMS (as described previously). Note that while electric energy consumption without the EMS increased 52%, the increase per degree day was 71%. For natural gas the increase in consumption without the EMS was 26% while the increase per degree day was also 26%. Finally, the overall increase in cost was 38%.
A nnua l A v e r a g e Co m m o d i t y Electric Energy Electric Demand Natural Gas Un i t s kWh kW CCF ( m3 ) Cost kWh/CDD CCF/HDD ( m3 / H D D ) w i t h EM S 2 0 7 ,3 0 7 9 2 .8 8 ,6 6 7 (2 4 5 4 5 ) $ 1 7 ,8 3 5 2 0 8 .2 1 .3 9 3 .9 4 w i t hout EM S 3 1 4 ,5 2 0 1 1 2 .4 1 0 ,9 3 7 (3 0 9 7 4 ) $ 2 4 ,5 3 0 3 5 6 .8 1 .7 6 4 .9 7 I n cr e a s e w i t h o u t EM S 1 0 7 ,2 1 3 1 9 .6 2 ,2 7 0 (6 4 2 9 ) $ 6 ,6 9 5 1 4 8 .6 0 .3 6 1 .0 3 38% 71% 26% 52% 21% 26%
Discussion As shown by the data in the previous section, a significant difference exists in the building’s utility cost with and without the EMS. Of note are the relative changes in all three indices (gas and electricity usage and electric demand). As mentioned in the introduction, originally (1986 and previous) there was heavy reliance on electric reheat. Although significant, this reliance appears to have increased even more in the post-EMS period. One of the functions of the EMS was to reduce the amount of electric reheat by increasing the amount of gas heating. The utility bills from the pre-EMS period (1986) to the EMS period (1987–92) show this change with a reduction in electrical usage and an increase in gas usage. One would expect this relationship to reverse once the EMS was disconnected (reverting to 1986 operation). Surprisingly, electric and gas usage increased in the post-EMS period which added to the detrimental effects of removing the system. After the removal of the EMS in 1992, the jump in utility use to a level much higher than the baseline year of 1986 may be explained by the following. While pre-EMS controls operation (1986 and earlier) was not as efficient as the EMS-based strategies, there was meaningful interest in conserving where possible. This awareness led to the discovery of the malfunctioning time clocks, provided the motivation and follow through for installation of an EMS and other efforts (e.g., utility usage tracking/reporting) during that time. However, the operation and control package installed postEMS appears to have focused significantly less on efficient operation than did the pre-EMS system. Perhaps this was due to a loss of understanding of the energy efficiency issues during the years the EMS “automatically” maintained control. Based on similar types of patterns in 1996, the net result is probably causing very significant over cooling and electric re-heating in summer. Advantages such as staging of the two-stage compressors and efficient airside economizer operation may also be lost. The period from 1995–97 shows a general trend of gas usage tracking degree days, but it also shows that electric usage was quite unpredictable. This seems to indicate that overall costs increase without the EMS and demonstrates the system’s vulnerability to changes in control adjustments. As a result, utility usage forecasts are more likely to show wide fluctuations. Improved occupant comfort also was a goal for the EMS. While this was not surveyed and can be quite hard to accurately assess, a few comments can be offered. First, occupant comfort complaints were significantly reduced in the first two years of EMS usage (when this was tracked to some degree).
46 ASHRAE Journal
Table 1: Annual average summary with and without EMS.
Much of the improvement came from eliminating large changes in setpoint, and thus temperature “swings,” for the RTU controlling thermostats. Even with reheat, the system could not always respond adequately. A second area of improvement came from proper morning warm-up/pre-cooling resulting in occupants arriving to a comfortable building. Finally, while the higher cooling mode setpoints (typically 75°F to 76°F [23.9°C to 24.4°C]) did generate some complaints in localized areas, the overall balance between zones and the ability of the EMS to observe all zones and reset temperatures accordingly did much to offset this issue. For this study, only a simple form of weather adjustment is applied to the data. As shown in Table 1, the usage per HDD or CDD is reported. While more sophisticated forms of weather correction could be applied, this study does not need this level of analysis. The overall pre- and post-EMS trends are clear and the annual weather values did not vary all that much. While a more detailed analysis (including monthly data, adjustments for non-HVAC usage, etc.), would yield interesting information and probably would explain small variations between some years, the overall conclusions likely would not change.
Conclusions The removal of the EMS caused an average increase of 38% in annual utility costs (from $1.40 to $1.92 per square foot). Viewed in reverse, the “savings” of 27% that were realized in going from standard controls to EMS operation are on the upper end, but are within reason based upon the authors’ experience. This is understandable when considering that the system is a constant volume system with electric reheat and that the $1.92 per square foot utility costs are on the high end of the expected range for this type of facility in this region. These findings are also consistent with published reports where these types of measures were implemented through better utilization of the existing EMS.3, 4 The annual increase of approximately $6,695/year, associated with the removal of the EMS, would have paid for the $1,500 repair needed in 1992 in less than three months. Had the EMS been repaired in 1992, the agency would have saved about
December 1998
ENERGY MANAGEMENT
$33,000 in utility costs (with the repair factored in, but without any differential O&M costs) from 1992–97. The authors suspect that the decision to remove the EMS was made for reasons other than technical or financial merit. These reasons may have included: 1) Ease of maintenance with existing staff; 2) Inability or unwillingness to support the EMS with a service contract; 3) Lack of knowledge of the magnitude of the financial impact; and 4) Lack of control by occupants and facility staff. A similar lack of understanding and utilization of EMS have been reported by others.5 The results of this study point to an important conclusion. Understanding the HVAC system and its proper operation often offers the potential for large utility savings. This was shown in 1987, when the manual EMS concept was tested. Much of the utility savings obtained from the installation of the EMS resulted from these measures, although the labor costs, effort, required expertise and dedication were quite high and unlikely to be sustained for long. Additionally, once the EMS was installed, investing in occasional supervision and infrequent repairs would likely have extended its life and led to prolonged utility savings. It should be noted that an active role by maintenance personnel in optimizing the EMS was responsible, in large part, for the magnitude of savings obtained from it.
Acknowledgments We would like to thank Mark Bowman of E-Cube for his contributions in the analysis and review of this article. References
1. Wolpert, J., S. Wolpert and G. Martin. 1992. Optimization of a dual-fuel heating system utilizing an EMS to maintain persistence of measures. 15th World Energy Engineering Congress, Atlanta. 3. Wolpert, J. and C. Robbins. 1991. Improving an older building through field audits and operational changes. Proceedings, Fifth Forum on Practical Applications of Building Energy Efficiency. Dallas. 4. Liu, M., Y. Zhu and D. Claridge. 1997. “Use of EMCS recorded data to identify potential savings due to improved HVAC operations and maintenance.” ASHRAE Transactions 103(2):4063. 5. Yago, J. 1992. “Survey shows users often fail to make the most of automation.” Energy User News, (4):25. Please circle the appropriate number on the Reader Service Card at the back of the publication. Extremely Helpful ....................................................... 458 Helpful ..................................................................... 459 Somewhat Helpful ..................................................... 460 Not Helpful ............................................................... 461
Advertisement in the print edition formerly in this space.
December 1998
ASHRAE Journal
47
