Latest from Air Conditioning

Smith & Wesson: The Story of a Chilled-Water Retrofit

March 1, 2005
Editor's note: This is Part 1 of a two-part series. Smith & Wesson manufactures handguns, law-enforcement products, and firearm-safety and security products

Editor's note: This is Part 1 of a two-part series.

Smith & Wesson manufactures handguns, law-enforcement products, and firearm-safety and security products in multiple buildings totaling approximately 600,000 sq ft in Springfield, Mass. For approximately 20 years, air conditioning for two of those buildings, which total approximately 107,000 sq ft, had been provided via 10 rooftop units and six split systems with direct-expansion (DX) cooling coils utilizing R-22 refrigerant. Those single-zone rooftop units and split systems provided constant-volume air distribution, with a cumulative design cooling capacity of 555 tons and a design heating capacity of 4,600 MBH.

Although the enclosures of the rooftop units and split systems, as well as the air-distribution fans and heating section, had held up well over the years, the cooling compressors were prone to refrigerant leaks and often in need of repair. Failed compressors had been replaced with rebuilt models. Additionally, the rooftop units' and split systems' dry-bulb-temperature economizer control system operated inefficiently. Some actuators were broken, leading to fixed damper positions, while the sensors were in need of calibration.

With a limited budget to satisfy stringent payback requirements, Smith & Wesson sought an alternative to the existing system. To evaluate potential energy savings and implementation costs, an energy-conservation feasibility study was conducted.

ENERGY-CONSERVATION FEASIBILITY STUDY

Three options were considered in the energy-conservation feasibility study:

  • Install new rooftop units and split systems with DX-cooling and steam-heating coils.

  • Install new rooftop units and split systems with chilled-water and steam-heating coils.

  • Replace the DX coils with chilled-water coils and the dry-bulb-temperature economizers with integrated enthalpy economizers and optimize operation of the entire system.

After the first option was deemed too inefficient and the second too costly, the third was chosen for further consideration and more-detailed analysis.

Electrical- and thermal-energy-conservation measures for the third option included:

  • Variable-frequency-drive (VFD) controls for the rooftop units' and split systems' air-distribution fans.

  • A high-efficiency chilled-water system with two 300-ton, water-cooled chillers equipped with VFD control and served by a two-cell evaporative cooling tower with on/off fan control.

  • Chilled-water primary-loop and VFD secondary-loop control.

  • Integrated-enthalpy-economizer control.

  • Chilled-water-temperature reset control.

  • Space-heating-setback air-temperature control.

Because of the facility's operations and corporate policies regarding return on investment, other measures for conserving energy, such as thermal storage and energy-recovery wheels, were not recommended.

Two-stage cooling and heating system with low-limit airflow control. In the original system, the air-distribution fans maintained a constant airflow throughout the three shifts of operation, even though manufacturing activity varied and there always were fewer people in the facility during the second and third shifts than there were during the first.

The total horsepower of the motors serving the air-distribution fans was 220 (Table 1). A customized VFD control strategy was suggested to reduce the fans' electrical-energy consumption. A two-stage control system giving priority to airflow-rate control was to be utilized. The first stage of control, or quantitative/qualitative control, was to be implemented with a variable airflow rate and a constant chilled-water/steam-flow rate. The steam coils did not have face and bypass dampers.

The system's minimum discharge-air temperature during cooling mode was set at about 50 F. This was acceptable because the supply duct had enough insulation to prevent condensation on its outside surface.

Upon receiving a return-air-temperature sensor's call for a reduction in cooling/heating output, the control system would slow the air-distribution-fan motors. If reducing the airflow rate to its lowest designated value failed to satisfy the required load, the second stage of control would be initiated.

The second stage of control, or qualitative/quantitative control, was to be implemented with a constant airflow rate and variable chilled-water or steam flow. Upon receiving a call for a further reduction in cooling/heating output, the building-automation system (BAS) would modulate a two-way chilled-water/steam control valve to its closed position to satisfy the return-air-temperature set point. The system did not have direct control of the humidity in the space.

Upon a call for an increase in cooling/heating output, the reverse would occur: First, the BAS would modulate the two-way chilled-water/steam control valve to its open position to satisfy the load. If the required return-air-temperature set point could not be satisfied at the minimum airflow rate, then the control system would allow the airflow rate to be increased. Only upon another call for a higher cooling/heating output would the BAS gradually increase the airflow rate to its maximum value to satisfy the return-air-temperature set point.

When the reduction of a system's cooling/heating capacity is necessary, this sequential-control strategy is represented by the following initial and end modes of operation:

  • Minimum airflow rate at maximum chilled-water/steam flow rate.

  • Minimum airflow rate at minimum chilled-water/steam flow rate.

When the increase of a system's cooling/heating capacity is necessary, this sequential-control strategy is represented by the following initial and end modes of operation:

  • Minimum airflow rate at maximum chilled-water/steam flow rate.

  • Maximum airflow rate at maximum chilled-water/steam flow rate.

This two-stage control is different when an enthalpy economizer is used. In such cases, outside airflow is allowed to vary from its minimum set-point value (about 10 percent) to its maximum rate (100 percent) as necessary to satisfy system cooling loads. Figure 1 shows a typical chilled-water-airflow configuration for an air-handling unit (AHU).

For Smith & Wesson, the importance of the two-stage control strategy was underlined by the fact that the power demand of the air-distribution fans (adjusted for the actual cooling load of 400 tons) was close to 142 kw, or 31.6 percent of the cumulative power demand of all of the elements of the major equipment serving the chilled-water plant, as well as the rooftop units and split systems (Table 2). Table 3 shows the power demand of the original DX cooling system.

With the existing air-conditioning and space-heating systems serving interconnected open areas within the two buildings (Figure 2), the customized control strategy was selected. Air-distribution-fan motor speed would be modulated as necessary to maintain air-temperature set point in return ductwork and, thus, control system capacity. A traditional variable-air-volume (VAV) arrangement with VAV boxes at the terminals was deemed cost-prohibitive, as it would have required substantial ductwork modification.

The served manufacturing areas have about 15-ft-high ceilings. Air discharged from the distribution grilles is well-mixed with room air prior to entering areas in which machine operators are working. The original air change per hour (ACH) — the ratio of the volume of air moved by all AHUs to the volume of the entire area — was nearly 7.7.

During the first shift of operation (8 a.m. to 4 p.m.), ACH would be reduced from 7.7 to 5.8 (low-limit airflow rate) to satisfy loads. During the second and third shifts of operation, it would be set to fluctuate from 7.7 down to 3.9 (low-limit airflow rate).

The cooling load imposed by the manufacturing equipment changes over the course of a day, usually peaking during the first shift. Approximately 57 percent of the design cooling load represents constant load (i.e., manufacturing equipment, lights, etc.). The remainder depends on outdoor-air temperature (i.e., infiltration, heat transmission via building exposures, etc.).

With the exception of a few units, the cooling system would not run when the buildings were unoccupied. Therefore, during the cooling season, most of the air-distribution fans would be turned off when the buildings were unoccupied.

For the heating system, the original control strategy was to maintain the same air-temperature set points during occupied and non-occupied hours. Airflow rate was maintained at its maximum design value. Upon a call from the area thermostat to reduce/increase system heating output, the two-way steam control valves modulated to their open/close position.

For non-occupied hours, two AHU-air-distribution-fan control strategies were considered:

  • Run the fans intermittently to satisfy the heating load and maintain the setback air temperature.

  • Vary the fans' speeds with variable-speed controls to satisfy the heating load.

The second option was chosen to maximize electrical-energy savings. Only if fan-motor speed was reduced to approximately 25 percent of its design value and there was a call for a further reduction of system heating output would an air-distribution fan be turned off.

The thermostats that controlled space temperature would be replaced with temperature sensors located at the return ducts of each AHU. The return grilles were located about 8 ft above the finished floor.

Integrated enthalpy economizers

The existing dry-bulb-temperature economizers were to be replaced with integrated enthalpy economizers, which were to be controlled with a new direct-digital-control automation system that would replace the original pneumatic control system. To optimize the cost of the project, the AHUs would be strategically organized into four groups of four.

The enthalpy of the air near each of the AHUs in a group would be compared to the enthalpy of the outside air. Based on the comparison, the BAS would decide when to use the economizer control. The enthalpies of the inside and outside air would be determined via readings from dry-bulb (DB) and wet-bulb (WB) temperature sensors (Figure 1).

Ventilation air and exhaust

The cumulative intake of ventilation air through outside-air dampers would be close to 20,600 scfm. An additional 10,300 scfm of ventilation air would be introduced via separate ventilation-air-intake openings. This air would be exhausted in the facility by the process equipment.

Annual energy savings with air-distribution-fan VFD control

Table 4 shows projected annual electrical-energy savings with variable-airflow-rate control during occupied space-cooling hours for the 16 AHUs. The calculations were performed using Springfield's bin outside-air temperatures.

Based on the AHUs' installed cooling capacity, the design cumulative cooling load (at 92-F DB, 73-F WB) was close to 555 tons. The actual design cooling load was found, based on trend-data analysis, to be approximately 400 tons, about one-third of the total cooling load in the manufacturing part of the entire facility.

The buildings are occupied 24 hr a day, 5.5 days a week. The electrical-energy consumption of the original system was calculated assuming a constant airflow rate. The electrical-energy consumption of the new system was calculated assuming the minimum daily airflow rate (low-limit airflow rate) was 0.58 of the maximum (design) airflow rate. This assumption was based on the control strategy, which allowed the air-distribution fans' minimum airflow to vary from 0.75 of the design value during the first shift of operation to 0.50 during the second and third. The average daily minimum airflow rate during the occupied period, then, was 0.58 ([0.75 + 0.5 + 0.5] ÷ 3).

The electrical-energy consumption of the air-distribution-fan motors was assumed to vary in proportion to the ratio of the current flow rate to the design flow rate in power 2.75. This is lower than the power factor 3, which is recommended for centrifugal fans and pumps. The calculated electrical-energy savings were adjusted for the actual hours of operation.

When the outdoor dry-bulb air temperature ranged from 57 F to 62 F, “free cooling” through utilization of integrated economizer control was assumed.

The projected annual electrical-energy savings were close to 205,687 kwh per year, or about 52 percent of the electrical-energy consumption of the original system. It should be noted that the actual savings exceeded the calculated ones in Table 4 because four AHUs continued running during non-occupied periods.

The design cumulative heating load at -5 F for the 16 AHUs was close to 4,500,000 Btuh.

The heat gains from the manufacturing equipment, lights, etc. were substantial. For outdoor-air temperatures from 27 F to 52 F, no space heating would be necessary during most of the occupied hours. Instead of heating, “free cooling” would be provided by modulating the return- and outdoor-air dampers at each AHU to maintain temperature set point.

The projected electrical-energy savings were 424,630 kwh per year, or about 77 percent of the electrical-energy consumption of the original system. These savings were adjusted for the actual number of occupied hours.

Electrical-energy savings with water-cooled centrifugal chillers

Projected electrical-energy savings attributed to the use of water-cooled chillers, as opposed to DX cooling coils, are shown in Table 5. The chillers were assumed to be used when the outdoor dry-bulb air temperature was above 62 F.

The original DX cooling coils' power demand and electrical-energy consumption at design dry-bulb outdoor-air temperatures of 92 F and above were assumed to be approximately 1.807 kw per ton, based on electric-utility historical data for similar DX cooling coils 15 years of age and older. The power demand of the DX cooling coils at dry-bulb outdoor-air temperatures below 92 F was assumed based on changes in power demand of a typical DX cooling coil.

The chillers' power demand would vary, depending on the load and entering-condenser-water temperature.

Included in the chillers' power demand was the power demand of the cooling towers, condenser-water pumps, and primary- and secondary-loop pumps, which totaled about 75 kw. According to the data in Table 5, utilization of the chillers would reduce annual electrical-energy consumption by 375,594 kwh, or 40 percent of the original system's electrical-energy consumption. The two chillers would run together when the cooling load was higher than 280 tons per hour; otherwise, the capacity of one of the chillers would be enough to carry the load.

With the chillers, the facilities' power demand would be reduced by approximately 415 kw (Table 5), or 57 percent of the original system's power demand. (Because it considers the approximately 4-kw increase in power demand of the air-distribution fans, Table 2 states the reduction as 411 kw.)

Combined energy savings

Electrical-energy savings from all energy-conservation measures were projected to total approximately 1,363,000 kwh per year, or about 68 percent of the original system's electrical-energy consumption.

Based on results of the energy-conservation feasibility study, the decision to design and install a new chilled-water system and retrofit the existing AHUs with chilled-water coils and associated controls was made. The decision was supported with incentives from an energy-conservation program mandated by the state and administered by the electric utility.

Alexander L. Burd, PhD, PE, is president of, and Galina S. Burd is a project manager for, Advanced Research Technology, an engineering and research consulting firm with offices in Suffield, Conn., and Green Bay, Wis. Alexander ( [email protected]) has 30 years of experience in the design, research, and optimization of HVAC and district energy systems, which includes publication of more than 30 research and technical papers in American and European journals, while Galina ([email protected]) has more than 20 years of design and research experience in the HVAC and architectural-engineering fields. Mauro De Maio, CPM, is facilities manager for Smith & Wesson Corp. De Maio ([email protected]) has worked in various capacities for Smith & Wesson for more than 30 years.

In July, this series will conclude with a discussion of the design and implementation of the new chilled-water system, the retrofit of the AHUs, the analysis of trend data, the reduction of power demand, and more.

For HPAC Engineering feature articles dating back to January 1992, visit www.hpac.com.