Saving Energy with Small Chillers
Engineers in virtually all industries are interested in finding cooling solutions for their manufacturing operations, plants and buildings that will allow them to be as competitive and profitable as possible. Because every application, industry and building has specific needs, there is no silver bullet choice when it comes to chiller selection. Instead, engineers must conduct energy audits to determine energy consumption patterns and use energy-modeling tools to make the optimal decision for their application. In many cases, the ideal solution uses chilled water to provide high-quality, cost-effective cooling and will include one or more optimization strategies such as variable flow or temperature reset.
The term “chilled water” is most commonly associated with a “central chilled water plant” -- a phrase that might conjure up visions of voluminous cooling towers; rows of large, high-capacity chillers; and a series of variable-speed pumps sending chilled water throughout a plant, hospital or campus. Despite this common perception, however, chilled water cooling is not limited to large installations or water-cooled chillers.
“Small” chilled water systems can be used in installations of all sizes with the use of low flow or variable flow, and/or with chilled water reset when variable flow is not an option. With a combination of simplicity, reliability and a low first cost, small chilled water solutions have all of the qualities necessary to optimize cooling systems and fit applications as small as 10 tons.
Small vs. LargeLarge applications such as a manufacturing complex, hospital or large campus are well suited to centrifugal chillers because of the compressor's variable-volume load characteristic. Centrifugal chillers are selected for specific operating temperatures and capacity at the time of manufacture, yet they efficiently regulate compressor capacity in response to changes in load and season.
Large applications can be optimized to ensure both energy and cost efficiency. Strategies to optimize pumping, plant partitioning and cooling tower operation all share the same design goal: to reduce the cost of owning and operating any process heating or cooling equipment as well as the heating, ventilation and air-conditioning (HVAC) system. These strategies favor large chilled water plants and centrifugal compressors, where chiller efficiency provides operating cost savings that help offset the cost of additional pumps and controls. But, in smaller applications where chilled water can be used, it might not be possible to rationalize the higher price tag for a centrifugal chiller or an elaborate automation system.
Small chilled water applications commonly employ chillers with screw, scroll or reciprocating compressors. These positive-displacement compressors respond differently than centrifugal compressors to changes in leaving chilled water temperature and leaving condenser water temperature. In this case, “small” also can refer to an application where variable-flow pumping is difficult to justify or implement such as in a 30-ton system (360,000 BTU/hr of cooling).
The massive flow rates of large chilled water plants make it easy for an engineer to justify variable-flow chilled water distribution. But small chilled water systems rarely consume enough pump energy to warrant the added cost and complexity of variable-flow pumping schemes. Small-capacity chillers also are seldom equipped to tolerate variable water flow through the evaporator.
Parallel or SeriesFortunately, several of the design and control concepts that work well in large chilled water plants are suitable for smaller systems. Engineers will find that reduced chilled water flow, variable chilled water flow, and series evaporators -- as well as chilled water reset when variable flow isn't an option -- can be beneficial when used in a smaller system.
A simple example is a conventional “no frills' chilled water system that uses air-cooled chillers to provide 400 tons (4,800,000 BTU/hr) of cooling capacity. The design of this system is based on a 10°F (5.6°C) rise in chilled water temperature ( ΔT) and constant water flow. Three-way valves regulate flow through the cooling coils. These criteria result in the chiller and pump selections shown in figure 1.
At $0.10/kWh, the cost of running the constant-volume pump amounts to $1.39/hr. If the pump operates 2,000 hours annually, the annual pumping cost is about $2,800.
Colder Water, Lower Flow. By using optimization strategies such as colder water and lower flow, the expense can be reduced significantly. Producing colder water (figure 2) would allow engineers to circulate less chilled water without compromising coil capacity. By circulating less water, the pressure drop is substantially lowered through the chiller, cooling coils and distribution piping. Dropping the chilled water temperature, by 3°F (1.6°C) in this case, reduces the required pump power by more than half. Now the annual cost of moving chilled water is less than $1,240 -- more than $1,500 less than the annual pumping cost in the original design.
Basing the design on constant flow means that it is compatible with the unit controllers on most chillers while avoiding the expense of a variable-speed drive on the chilled water pump. Unfortunately, however, this low-flow solution is not without trade-offs: Chiller efficiency drops by 2 percent, and chiller capacity decreases by 20 tons. Given the operating characteristics of positive-displacement compressors, the simple switch from one compressor type to another does not avoid these losses. (See the “Compressor Axioms” sidebar.)
Colder Water, Series Evaporators. There are solutions that can efficiently recover the lost chiller capacity and still produce colder chilled water. For example, by piping the evaporators in series (figure 3), an increased pressure drop is traded for improved chiller capacity and efficiency. While the annual pumping cost for this solution climbs to $2,570, which only trims about $200 from the original design (figure 1), piping the evaporators in series improves the chiller efficiency of the design in figure 2 by 3 percent -- and adds 24 tons of capacity.
Variable Flow. The annual pumping expense can be cut in half if the chiller controls can accommodate variable water flow through the evaporator. Altering the series low-flow design (figure 3) to include variable primary flow would make an acceptable payback more likely. Otherwise, implementing a variable-flow pumping scheme will require a primary-secondary piping arrangement. It is unlikely that annual savings of only $1,400 will justify the initial investment in both a constant-flow primary pump and a variable-flow secondary pump.
The ideal solution for the foregoing example, as well as the scenarios encountered by cooling system design engineers every day, can be determined with the help of a simple energy study. Engineers can use software tools available from chiller manufacturers to perform an hour-by-hour load analysis and help identify the need for low ΔT or high ΔT, parallel or series chillers, and determine whether an investment in variable flow will provide a satisfactory return, making the guesswork methods unnecessary. Table 1 summarizes the results of such an analysis for the hypothetical system described.
Chilled Water ResetPayback is easily attainable for a series evaporator arrangement and variable flow when the chillers are large enough and sufficiently sophisticated controls will accommodate variable evaporator flow. However, several options also exist for small chillers and simple controllers that preclude variable evaporator flow.
Series evaporators and a wide ΔT remain viable concepts for even the smallest chillers. But there is another way to recoup the cost of pumping energy in lieu of a variable flow scheme.
Chilled water reset (CWR) -- that is, raising the chiller's control setpoint -- is a common practice in many chilled water plants as a means of reducing chiller energy consumption. In constant-flow applications, CWR is relatively easy to implement and can be controlled based on the drop in the return-water temperature. In variable-flow systems, however, raising the supply-water temperature increases pumping energy. While the chiller coefficient of performance (COP) ranges from 2.80 to 6.101, depending on the compressor type, the pump COP is only about 0.65. (Note: The COP range brackets Standard 90.1's minimum efficiency requirements for air- and water-cooled chillers with reciprocating, screw, scroll or centrifugal compressors.) Often, the increased pumping energy will more than offset the savings in chiller energy -- especially if the chiller typically operates at part-load conditions. Of course, circulating warmer chilled water through the cooling coils also raises the leaving-coil air temperature, perhaps to the point that it can no longer adequately dehumidify the space.
Most chiller controls readily support chilled water reset. Implementing this option is particularly easy when two small chillers are piped in series. Engineers can use a warmer setpoint to control the upstream chiller and a colder setpoint to control the downstream chiller (figure 4).
During humid weather, cold chilled water (42°F in this case) allows proper dehumidification and ensures ample cooling capacity at all coils, while the off-peak cooling season (spring and autumn) often requires even less dehumidification. For the 30-ton system shown in figure 4, supplying the coils with 48 to 50°F (9 to 10°C) chilled water might provide enough capacity to satisfy most off-peak cooling needs. Chilled water reset can be achieved simply by turning off the downstream chiller.
Many of the energy-conserving strategies developed for large chilled water plants are easily applied to small chilled water systems. The potential benefits of implementing these strategies are the same, regardless of the size of the installation.
When properly selected, a centrifugal chiller will readily provide money-saving efficiency at virtually any temperature condition. By comparison, a chiller with a positive-displacement compressor will respond to changes in water temperature with a more dramatic loss or gain in capacity and efficiency.
Sidebar: Compressor AxiomsCompressor capacity increases as suction temperature rises (assuming that the condensing temperature remains constant). Therefore, control strategies that result in lower suction temperature will also decrease the available cooling.
Compressor efficiency increases as head pressure decreases. As a result, control strategies that result in a higher head pressure will make the compressor work harder (consume more energy).
The magnitude of these effects is significantly less for chillers that use centrifugal rather than positive-displacement compression. That's because centrifugal compressors increase the temperature and pressure of refrigerant by dynamically converting flow velocity to static pressure. Positive-displacement compressors such as screw, scroll and reciprocating types “work” on the refrigerant by trapping it and shrinking its volume. For more information about compression technologies and their operating characteristics, see Chapter 34, “Compressors,” in the 2000 ASHRAE Handbook-HVAC Systems and Equipment (see www.ashrae.org).
Don Eppelheimer is an applications engineer and Brenda Bradley is an information designer on Trane's commercial systems applications engineering team, which is based in La Crosse, Wis. Trane provides indoor comfort systems and facility solutions for industrial, commercial and residential buildings. For more information about Trane's industrial chillers, call (608) 787-2000 or visit www.trane.com.