Future-proofing a chilled-water system
Chilled-water system design often focuses on the proper initial operation of the system. However, understanding—and planning for—changes that occur as the system ages is critical to designing a chilled-water system that will operate properly for years to come.
- Understand what will happen to chilled-water coil performance as a facility evolves and the chilled-water system gets older.
- Be aware of issues that will occur if long-term changes to the chilled-water system performance are not accounted for in the initial design of chillers, pumps, and cooling coils.
- Learn some recommended practices for chilled-water system component design to prevent long-term issues from occurring.
A properly functioning chilled-water system is critical to the comfort of commercial facilities. Often, much design consideration is given to the initial operation of these systems. However, chilled-water systems are expected to operate effectively for decades, even as the facilities they serve constantly evolve and change.
Chilled-water systems are a balance between the chillers and the various chilled-water coils throughout the system. Much attention is given to how chiller performance impacts chilled-water system operation. However, just as important, cooling coil performance also impacts chilled-water system operation.
To evaluate best design practices for proper long-term operation of chilled-water systems, it’s first necessary to understand what happens to the performance of chilled-water cooling coils as a system ages. To do this, we’ll analyze some common changes that occur during the life of a chilled-water system and the impacts these changes have on chilled-water coil performance.
This analysis will be performed using a commercial coil-selection software. By modifying the coil-selection parameters, it’s possible to simulate changing conditions at chilled-water coils. For simplicity, this analysis will focus on a single chilled-water coil in a central air handling unit (AHU). This will be used to represent what occurs at all coils throughout a system.
First, we’ll start with a baseline cooling coil, which represents the coil’s performance at system start-up. The required operating conditions of the baseline coil (summarized in Table 1) are used to select the baseline coil. (The waterside performance of this baseline coil selection is summarized in Table 2.) For the rest of this analysis, the physical properties of this baseline coil selection (number of rows, fin spacing, fin type, tube type, etc.) will be held constant. This will simulate changing conditions at this specific coil.
The first cooling coil change we’ll analyze is fouling. As water and air flow through and around the coil, particulates and scale accumulate on the airside and waterside surfaces of the coil, a phenomenon known as fouling. Fouling impedes the heat transfer at the coil and affects coil performance (see 2016 ASHRAE Handbook—HVAC Systems and Equipment). Specifically, chilled-water coils can develop significant fouling (see 2017 ASHRAE Handbook—Fundamentals). Coil-selection software simulates the effects of fouling with the fouling-factor parameter. This factor represents an additional thermal resistance in the coil heat-transfer calculations.
From this author’s experience, dirty cooling coils can experience fouling factors on the range of 0.001 sq ft-°F-hr/Btu. This fouling-factor value is at the top of the range considered by the Air-Conditioning, Heating, & Refrigeration Institute’s (AHRI) Standard 550/590, so it represents a realistically dirty coil. Adding this fouling factor to the baseline coil results in the modified waterside performance summarized by Change A in Table 2.
The next cooling coil change we’ll analyze is increased airflow. Renovations are an inevitable reality as facilities evolve. Often, renovations result in additional airflow requirements due to newer codes, increased heat loads, or additional square footage. Typically, this additional airflow is provided by simply increasing the speed of the fans within the AHU serving the space. The cooling coils in the AHUs normally remain untouched, due to cost limitations. While increasing the fan speed may satisfy the needs of the renovation, it does negatively affect the chilled-water coil performance. Over time, numerous renovation projects can lead to significantly higher airflows through a cooling coil than originally designed.
It’s common to come across AHUs in existing facilities that are being run harder than they were designed to run. Increasing the airflow through the baseline coil by 15% while leaving all other operating conditions the same (entering air conditions, leaving air temperature, entering water temperature) results in the modified waterside performance summarized by Change B in Table 2.
The last cooling coil change we’ll analyze is the decreased leaving-air temperature setpoint. Again, due to the constant evolution of spaces within a facility, it is common for spaces to get too warm. If airflow cannot easily be increased in those spaces, a common tool employed by maintenance personnel is to lower the discharge-air temperature leaving the AHU. While this may satisfy the immediate needs of the warm spaces, it also negatively impacts the chilled-water coil performance. Decreasing the leaving-air temperature from the baseline coil by 2°F while leaving all other operating conditions the same results in the modified waterside performance summarized by Change C in Table 2. A
review of the performances in Table 2 shows how drastically these changes can affect cooling coil performance. Each change results in lower waterside delta T and increased chilled-water flow at the cooling coil. In practice, cooling coils often deal with a combination of the three changes explored. As the coil performances degrade, the chilled-water flow will ultimately have to increase to maintain setpoints, either by increased chilled-water pump differential pressure or by staff readjusting balancing valves.
It’s no wonder that “low-delta T syndrome” is so common in chilled-water systems. As this analysis shows, there are several factors working in unison to decrease the waterside delta T at chilled-water coils throughout the system. Lower waterside delta T is almost a natural part of a chilled-water system’s aging process.
The perils of short-sighted design
As a cautionary example, let’s explore the long-term implications of a chilled-water system design approach that only focuses on the initial operation of the system. For simplicity’s sake, this example will focus on a basic chilled-water system with two chillers and a primary/secondary pumping arrangement. (refer to Figure 2 for an overall schematic of the chilled-water system used in this example.)
Going back to the baseline cooling coil selection from Tables 1 and 2, assume that all cooling coils in the chilled-water system were selected in the same manner. As a result, the average cooling coil design’s waterside delta T would be 12°F, and the design chilled-water supply temperature to the cooling coils would be 45°F. Chiller selection is generally dictated by the required capacity of the chiller.
A careful analysis of all the connected cooling equipment and building loads should be performed to determine the maximum cooling capacity required by the chillers. After completing this analysis, assume that the chillers are selected with this required capacity at a 12°F waterside delta T and 45°F chilled-water supply temperature, which matches the cooling coils. In this example, assume that each chiller is sized at 50% of the required cooling capacity.
For the primary and secondary pumps, assume these are also sized at the same 12°F delta T as the chillers and cooling coils. This results in the primary and secondary loops having the same design flow rate. (This system arrangement and the selection criteria used for the system components is summarized in Figure 2.)
Now let’s analyze how this chilled-water system design will be controlled and operated throughout the life of the system. When the chilled-water system first begins operation, everything should perform as designed if properly installed and commissioned. First, the cooling coils will achieve their design performance including the design waterside flow rate and delta T. The secondary pump will modulate its speed to satisfy the needs of the coils. The chillers, and their associated primary pumps, will be staged as required to maintain the chilled-water supply temperature setpoint. At low-load conditions, only a single chiller and primary pump will operate.
Once the cooling load of the facility exceeds 50% of the design capacity of the system, the second chiller will turn on. When the system is at its design capacity, the flow through the primary loop will equal the flow in the secondary loop; there will be zero flow through the decoupler. Everything appears to operate as intended and designed.
However, let’s now explore how this system will be controlled and operated several years in the future. From the cooling coil analysis, we know the delta T being achieved at the cooling coils will have decreased. This means the chilled-water flow rate required by the cooling coils with respect to cooling load will be higher than design. From the cooling coil simulation results, it’s reasonable to envision the required cooling coil flow rate at 50% of design when the cooling load is only at 40% of design.
In this situation, the secondary pump will still be able to modulate its speed to satisfy the needs of the chilled-water coils. A single chiller will be operating at 80% of its capacity to maintain the chilled-water supply temperature setpoint. However, the flow in the secondary loop is at 50% of its design, which is equal to the flow in the primary loop with one chiller operating. As the cooling load increases, the secondary loop will now have a higher flow rate than the primary loop. When this happens, there is “reverse” flow in the decoupler. Warm secondary return water will mix with primary supply water.
Although the chiller is generating 45°F water, the secondary chilled-water supply temperature will be higher than 45°F, which won’t satisfy the cooling coils in the system. As a result, the chilled-water system needs to turn on the second chiller, even though the first chiller is only operating at 80% of its capacity. This is not ideal, as two chillers are now running when only one chiller is required.
Increasing the cooling load
Continuing with this example, let’s explore what happens as the cooling load increases. When the cooling load is at 80% of design, the required cooling coil flow rate may now be at 100% of design. In this situation, the secondary pump can satisfy the needs of the chilled-water coils, but it’s out of capacity. As the cooling load increases, the secondary pump may be able to keep up with the cooling coils by increasing speed, but it is running into its safety margins.
Meanwhile, the secondary loop has a higher flow rate than the primary loop again. As described above, this will result in the secondary chilled-water supply temperature being higher than the primary chilled-water supply temperature. Now the chilled-water coils are getting warmer water than they are designed for, which means they will struggle to meet their required setpoints. While all this is happening, the chillers are only operating at 80% of their capacities.
To prevent losing space temperatures throughout the facility, the only choice maintenance staff has at this point is to lower the chilled-water setpoint at the chillers. However, this is undesirable for multiple reasons. First, the efficiency of a vapor-compression refrigeration system, such as chillers, is a function of the temperature difference between where heat is extracted and heat is rejected. For chillers, this temperature difference is often referred to as “lift.” In an air-cooled chiller, the lift is the temperature difference between the ambient-air’s dry-bulb temperature and the chilled-water supply temperature.
In a water-cooled chiller, the lift is the temperature difference between the leaving condenser-water temperature and the chilled-water supply temperature. Thus, lowering the chilled-water supply temperature below what is needed by the cooling coils unnecessarily decreases chiller efficiency. The other reason lowering chilled-water supply temperature is undesirable is due to chiller capacity. Increasing the lift of a chiller decreases its cooling capacity. As a result, the chillers will not be able to provide their design cooling capacity when operating at this lower supply-water temperature. For obvious reasons, this is a bad situation. Now the chilled-water system won’t have enough cooling capacity to handle the facility’s cooling needs at high-load conditions.
As this example demonstrates, designing a chilled-water system with only initial operation in mind can lead to a host of problems in the future. Pumps will have trouble keeping up with coil-water demand, more chillers will be required to run than necessary, and the system ultimately will struggle to satisfy the facility’s cooling requirements.
Designing to be ready for problems
So what steps can be taken during the initial design of a chilled-water system to prevent the above problems from occurring down the road? Like many problems, the first step is acceptance. In this case, it means accepting that the delta T of the chilled-water system will degrade over time. This reality should be embraced with the design of all components within the chilled-water system including the cooling coils, pumps, and chillers.
A project profile is presented to illustrate some recommended practices for designing and selecting these components. This project profile is a chilled-water system for a regional hospital in the upper Midwest. Much like the short-term design example above, this project had two chillers and a primary/secondary pumping arrangement.
Since the delta T of a chilled-water system is a function of the heat transfer at the cooling coils, its design is critical to the proper long-term operation of the chilled-water system. Since fouling in chilled-water coils is unavoidable, a fouling factor should be included in the initial cooling coil selections. A reasonably conservative fouling factor to be included in the selections is 0.0005 sq ft-°F-hr/Btu, which is in the middle of the range coil considered by AHRI Standard 550/590. This allows some fouling to occur in the coils while still being able to achieve the design performance.
Another way to combat lower delta T as the coils age is to maximize the initial cooling coil’s delta T. Because the delta T achieved at a coil is a function of its heat-transfer surface, it follows that maximizing the delta T means maximizing the surface area. A cooling coil’s surface area is based on the number of rows, fin spacing, and the face area of the cooling coil. Too many rows or too tight of fin spacing can have negative impacts. These include high airside pressure drop, high waterside pressure drop, and decreased ability to effectively clean and maintain the coil.
In this author’s experience, a reasonable balance between the pros of a larger cooling coil and the cons of too large of a cooling coil is a coil with eight rows and a fin spacing of 120 fins/ft. This cooling coil arrangement can be cleaned and maintained and does not result in exceptionally high airside or waterside pressure drops. This means that the cooling coils are selected based on size and not based on a given waterside delta T. As a result, the waterside delta T at each cooling coil is slightly different, based on the air velocity and entering/leaving air conditions at each coil.
However, selecting the coils this way ensures that they can achieve the largest possible waterside delta T. (The various cooling coil selections for this example project are summarized in Table 3. As can be seen in Table 3, the average waterside delta T at each coil is on the range of 14 to 16°F.)
The chilled-water pumps can also be designed and selected to better handle an aging chilled-water system. As the previous design example illustrated, selecting the pumps at the same delta T as the coils will eventually lead to problems. Accepting a lower delta T in the long-term means sizing the primary and secondary pumps at a lower delta T than the coils from the start. In this project, both the primary pumps and secondary pumps were sized at a 12°F delta T.
This delta T value was chosen based on experience with chilled-water system degradation in other facilities. This will allow the pumps to satisfy the cooling coil needs even after considerable degradation at the coils. The secondary pump will have enough capacity to meet the increased flow-rate requirements of the cooling coils as they degrade. Also, the larger flow rate of the primary pumps will prevent reverse flow in the decoupler as the secondary pump speeds up over time. This will prevent many of the issues from the previous design example, such as operating more chillers than necessary and the secondary supply water being warmer than the primary supply water.
Lastly, the chillers should be designed to better handle long-term degradation of the system. As with any chilled-water system design, a careful analysis of all the connected cooling equipment and building loads was performed to determine the maximum cooling capacity required by the chillers. Based on discussions with the owner, each chiller was selected at 50% of the required cooling capacity of the system.
However, unlike the previous design example, the chillers were selected to provide this required cooling capacity at a chilled-water supply setpoint of 42°F. Initially, the chilled-water supply setpoint from the chillers will be 44°F, which will allow 1°F of temperature rise from the chillers to the coils. However, designing the chillers with a lower supply-water temperature gives maintenance staff the option to lower the chilled-water supply setpoint while still having enough chiller capacity to meet the cooling requirements of the facility.
A lower chilled-water supply temperature greatly benefits degraded cooling coils. This will decrease the required flow rate and increase the waterside delta T at the cooling coils. As previously discussed, doing this will decrease the efficiency of the chillers, so this should only be done when the cooling coils truly need the lower-temperature water.
The final result, when all these design approaches were incorporated in the chilled-water system on this project, is summarized in Figure 3. Note that some redundant pumps and other system components are not shown in Figure 3 for the sake of simplicity.
It should be noted that the recommendations presented here will have impacts on system cost and energy performance as compared with the “shortsighted” design. First, these recommendations will result in increased first costs, because the coils, pumps, and chillers will all be slightly larger. However, any cost increases in the initial system install will be minor compared with the costs related to needing to make equipment changes in the future. Also, designing a chilled system to operate properly over its entire lifetime should be considered a necessity and not an option.
Additionally, these recommendations will have some impact on the energy use of the chilled-water system. Overall, the pumping energy use should decrease due to the lower required flow rates of the chilled-water coils. The pump’s energy savings can be further improved by operating the primary pumps at a lower initial flow rate and only increasing their flow when the system requires. The energy use of the chillers should be very similar because the chiller lift is the same. As the system degrades, the energy use of the chillers and pumps will increase, but that is simply a necessity of the system. In general, the goal is for the system to be as energy-efficient as possible, given the age and performance of its components.
Another item to note is the importance of chilled-water system commissioning, both at system start-up and as it ages. Even a perfectly designed chilled-water system will struggle if the system isn’t installed and setup properly to begin with. Also, facility staff should strongly consider retro-commissioning prior to making chilled-water system operational changes to account for system degradation. Ensuring that the system is operating correctly and that chilled-water components are properly cleaned should always be the first step when addressing performance issues.
Chilled-water systems will naturally degrade during their lifetime and this can result in some significant long-term functionality and operational issues. However, accounting for this and using some best practices in the initial design and selection of the chilled-water system components will leave the system better prepared to handle degradation over the years to come.
Kevin Langan is a senior mechanical engineer at IMEG Corp. He serves as project manager and lead mechanical engineer on a variety of health care projects.