providing insights for today’s hvac system designer Engineers Newsletter volume 47–3 impacts of Chilled-Water System Design Decisions

Primary-secondary system bypass This Engineers Newsletter walks through a number of design decisions, with Bypass line sizing sizing. The premise of a primary- secondary system is to hydraulically discussion and examples to explain how This seemingly simple decision can have and why those decisions are made. separate (decouple) the primary (chiller) significant consequences if not done flow from the secondary (system) flow. correctly. This decoupling prevents the operating • In a primary-secondary system, the While designing a chilled-water pressures of chiller pumps from impacting bypass pipe should be the same system, a myriad of decisions must the operating pressures of the system diameter as the pipe going into the be made. Experienced engineers pumps. This is accomplished by installing largest chiller. Its length should be often make these decisions a bypass line with a small pressure drop. about 8-10 pipe diameters long or "automatically" as they have in the The bypass allows water to flow in either have an equivalent pressure drop. past, based on what they have direction and is used to indicate chiller and learned from experience. Today, all • In a variable primary flow (VPF) primary pump sequencing in this system engineers likely use an internet system, the bypass line should be arrangement (Figure 1). search engine to get direction, but sized for the largest minimum flow Chillers and constant flow primary pumps when there are differing—or even rate and it will have a control valve. are enabled in pairs, making the primary conflicting—recommendations a The reasons behind this guidance follow. flow rate a step function. As system load decision must be made. and flow increase, the excess flow rate Common decisions regarding (from supply to return) in the bypass line chilled-water system designs decreases. At the point that system flow include: rate exceeds chiller flow rate, deficit flow

• bypass line sizing in variable Figure 1. Primary-secondary system flow systems • dynamically varying condenser water flow primary pumps • number of chilled-water pumps to operate secondary pumps with • series chillers and power VSDs consumption chilled-water coling coils • whether to use pressure- with two-way independent control valves valves

bypass line, free of restrictions, sized the same as the largest chiller’s pipe, aim for a pressure drop equivalent to a pipe that is 8-10 pipe diameters long.

©2018 Trane. All rights reserved. 1

(from return to supply) in the bypass Figure 2. Oversized bypass can allow simultaneous flow in both directions line occurs. This warm, deficit water flow increases the system supply- chiller return chiller supply water temperature and at some point water water an additional chiller and primary pump are enabled. short-circuited chiller A chiller and pump are disabled when supply water excess flow rate in the bypass line is high enough to still have excess flow after a primary pump is turned off. Therefore the maximum flow rate the bypass line ever experiences is a little short-circuited system higher than the design flow rate of the return water largest chiller. Often designers wait for 10-15 percent excess flow, to ensure system that chillers are not cycled on and off system return water supply water rapidly. Therefore the bypass pipe should be sized for 110 to115 percent • Much of that water continues to the of the largest chiller's flow rate—which Add pressure drop. A better option chillers but some of the warm water most designers simplify to the same may be to place an orifice in the line. This (due to low pressure drop in the as the pipe size going into the largest can simply be a plate with a hole in it. bypass) flows in the deficit direction chiller. This imposes a pressure drop, but allows in the bypass line (from return to the water to flow freely in either the surplus supply). or deficit direction. In a few extreme What happens if bypass sizing is cases it has been neccessary to not right? Too little or too much As a result there is simultaneous flow in completely block off the oversized/short pressure can cause issues... opposite directions because the bypass bypass pipe and install a properly sized line has become so large it functions like pipe of sufficient length to eliminate the Too small, pressure too high? An a tank! mixing condition. undersized bypass line can result in high fluid velocity, which in extreme The diluted (higher) system supply-water The best way to avoid issues in a conditions may cause pipe erosion, temperature results in more pumping primary-secondary system is to size the vibration and acoustic issues. The high energy. The short-circuited supply water bypass line properly during the design fluid velocity can also result in high mixing with return water results in process. Simply check the drawings and enough pressure drop so that flow is reduced chiller return-water temperature. ensure that the bypass line is smaller restricted through the bypass, and This can restrict the chillers’ ability to than the manifold, and the same size as may actually cause the primary and fully load. the pipe going into the largest chiller and secondary pump pressures and flow 8 to 10 equivalent pipe diameters long. If rates to affect each other. If the pipes are already installed how the pipe is less than 8 to10 pipe can the situation be improved? A diameters long, using elbows to form a Too large, pressure too low? A resolution for too low of a pressure drop "U" adds an appropriate pressure drop. benefit of an oversized bypass pipe is a is to impose a modest restriction to keep very low pressure drop, however if it's water from short-circuiting. Aim for the too low, the water may not flow as Variable-primary-flow system bypass equivalent pressure drop of a pipe that is intended. Figure 2 illustrates the issue: line sizing. Conceptually a variable- 8 to 10 pipe diameters long. primary-flow system is simpler to get • Cold water leaves the chillers and right. The valve in the bypass line only Often the first solution considered is to flows into the chiller supply opens when the system flow rate add a valve. Given that the bypass line is manifold. approaches the minimum flow rate of the oversized, it’s likely the additional valve operating chiller(s). So the bypass pipe • A portion of that cold water flows will be big and expensive. It’s also likely and valve only need to be sized for the toward the secondary pump. that at some point a well-meaning but largest minimum flow rate. Usually that's However due to the very low uninformed operator will close the valve the largest chiller's minimum flow rate. pressure drop in the bypass, some too much. Or, they might open it all the However, depending on chiller of the cold water seeks the path of way and defeat this solution. Recall that selections, the largest minimum flow least resistance and flows through the bypass line in a primary-secondary rate might not be for the largest chiller in the bypass line in the excess system should allow water to flow freely, the plant. Also consider the combined direction. in either direction, as needed. minimum flow required when two • Similarily in the return-water side, chillers are operating at part load, just warm return water flows in the before sequencing off one chiller. system return manifold.

2 Trane Engineers Newsletter volume 47-3 providing insights for today’s HVAC system designer Should we dynamically vary Table 1. Plant annualized kW/ton and percent savings compared to two-chiller base for three alternatives Condenser Condenser Tower Plant the condenser water flow? Chiller Cooling water flow Alternative water flow control annualized Savings type tower fan rate type method* kW/ton Yes, savings are available in existing (gpm/ton) systems designed between 2.5 to 3.2 gpm/ton (12 to 9.4ºF ΔT) condenser water Base VS VS 3 CF Opt 0.5462 NA flow rate. Controls complexity should be accounted for and the system needs to be Variable CW flow VS VS 3 VF Opt 0.5260 3.7% properly commissioned. Reduced design VS VS 2 CF Opt 0.5255 3.8% No, in new systems designed for 1.8 to 2.2 CW flow gpm/ton (16.6 to 13.6ºF ΔT) condenser water flow rate since they already achieve Reduced design VS VS 2 VF Opt 0.5252 3.8% almost all the savings and reduce system CW flow and variable CW flow complexity—a lot. Systems designed at these flow rates do not require varying the *Near optimal control (Opt) is minimum the sum of chiller + cooling tower fan kW at each operating point during condenser water flow rate. the year

Why? Varying condenser water flow rate Example. Let’s compare the condenser • Pumps are smaller with significantly can be complicated. There are several limit water flow options. lower power conditions and setpoints to manage: It's imperative that the sum of chiller, • Chiller power rises marginally • The flow rate has to stay above the condenser water pump and cooling tower minimum flow rate as defined by the • Cooling towers become more effective fan power is considered. Less-than-optimal highest of minimum tower flow, as heat exchangers, since warmer results can happen if decisions are based minimum chiller flow, or to produce the water is sent to the cooling tower, on only one component. static lift in the open portion of the resulting in condenser water system. When any of The measure of efficiency is the annual – Reduced tower fan power these are close to the system design kWh of chiller, cooling tower fans and – Reduced tower cost flow, variable flow should not be condenser water pumps divided by the

attempted. annual ton-hours of cooling. This results in – Possible reduction of tower size • At each operating point during the year, annualized performance in terms of (which further reduces tower cost). determine the optimal condenser kW/ton for those components. water pump and cooling tower speed. Observations. In all cases the operating Table 1 illustrates a comparison of four savings (Table 1) are very similar, so what • Make sure not to reduce the alternative two-chiller systems: guidance can be provided? condenser flow at conditions and operating points that cause the chiller • Base design of 3 gpm/ton and constant • Many existing systems were designed to surge. condenser water flow rate (CF) using historical practices and a condenser flow rate of 3 gpm/ton. In • Ensure the sequence is documented • Design of 3 gpm/ton and varying condenser water flow rate (VF) existing systems, annualized plant and properly commissioned. energy can be reduced by varying the • Design of 2 gpm/ton with constant condenser flow rate. Be sure to This complexity requires that the system condenser water flow rate be commissioned, the controls remain properly commission the system and ensure control is performed correctly operational and future changes are • Design of 2 gpm/ton and varying accommodated. Careful consideration condenser water flow rate long-term. should be made for system changes such • For new systems or when all existing as chiller, pump, and/or tower Assumptions. These alternatives assume chillers are being replaced, using low replacements. the pipe size remains unchanged. And that design condenser flow rates and near optimal tower control resets the constant flow reduces installed costs, tower water setpoint to achieve the saves the same amount of energy (or Table 2. Industry recommendations for condenser water design flow rates minimum sum of the chiller plus cooling more) as if condenser flow rate is varied tower fan kW at each operating point and and keeps system control both Flow rate(s) during the year. Two gpm/ton is shown in simple and understandable. Source ∆T (ºF) (gpm/ton) our example but projects have varying Historical 9.4 3.0 optimal design flow rates. practice Today’s Industry recommendations Effect of flow rate on equipment. With ASHRAE 12-18 2.3 - 1.7 reduced design condenser water flow rate GreenGuide1 based on present industry Kelly and Chan2 15 2.0 recommendations (Table 2): Taylor3 15 1.9

providing insights for today’s HVAC system designer Trane Engineers Newsletter volume 47–3 3

Figure 3. Pressure drops in a VPF system Does operating two pumps chiller flow rate and at lower speed save more pressure drops identical energy than operating one pump and fitting pump? flow rate and pressure drops change Perhaps a little, but not nearly as much as many people think. system flow rate and pressure drops A B identical Why? This came up during an ASHRAE conference conversation with coils and respect to the affinity laws, and a wise valves consulting engineer raised his eyebrows and said one word: "Think."

As it turns out, the reason the question pipes, elbows is asked is because of a and fittings misunderstanding, or perhaps lack of thinking—about the pump, or affinity, laws. The pump laws describe the relationship between flow rate and power and, in a straight pipe that relationship is cubic. So the thought Example. Consider the system in So, is operating an additional pump process is, if an additional pump is Figure 3. At a given point in time, the beneficial? enabled, the flow rate per pump will go system flow rate is identical—no matter how many pumps operate—because it's • The system flow rate does not down and the kW per pump will go change, down with the cube of that flow rate. dependent on the required coil flow rates. The flow rate is the same through the coils • Almost the entire system pressure This just isn't true - and is an improper and valves, as well as the pipes, elbows drop is identical. and fittings. Meaning their pressure drop understanding of the pump laws for • There is only a small change in is the same. The flow rate through the this situation. What did we miss? system pressure drop - between the chillers also remains identical, as do their pump return and supply manifold. As the pump power equation below pressure drops. The only change in the shows, pump power is dependent on system is that there are now two paths for • And without study, we don't know if the flow rate and system pressure drop water flow across the pump manifold. the pump, motor and drive irrespective of the number of pumps efficiencies are better or worse at that are operating. The 0.746 and 3960 When the flow rate through the operating reduced pump speed. Some pump terms are conversion factors. In the pump is reduced, so is the pressure drop manufacturer's selection programs denominator are the pump, motor and through its fittings - and the pump itself. can provide manifolded pump/drive drive efficiencies. They may improve or So there is a pressure drop reduction - but efficiency ratings with various worsen at different pump operating it is only the pressure drop from the return numbers of pumps in operation. conditions. manifold (A) to the supply manifold (B), and through the pump. As long as combined efficiency is the same or better, the result is pump power How about pump, motor and drive that is a little lower - due to the system efficiencies with two pumps operating? pressure drop being a little lower. But it is Burt Rishel's ASHRAE Journal article4 nowhere near the "cubic" pump savings titled, "Wire-to-Water Efficiency of that some assume. Pumping Systems," explains that there are some operating conditions that allow the combined pump, motor and drive efficiency to rise when more pumps operate. While it takes some calculation time, if the combined efficiency is higher, pump power is lower.

4 Trane Engineers Newsletter volume 47-3 providing insights for today’s HVAC system designer Figure 4. Chiller lift How much power can be saved by piping chillers in a "series counterflow" kW a load x lift lift Pcnd – Pevp arrangement compared to lift Tlvg cnd – Tlvg evp piping chillers in parallel? 98.9ºF – 37ºF 98.9ºF

lift (ΔT) 85ºF Almost 13 percent chiller power 61.9ºF reduction is available by piping in a series-counterflow configuration. 55ºF • Using Trane Duplex chillers, more 37ºF that 19 percent can be saved. Increased pumping power reduces temperatures are based on a system So the respective lifts are: these savings a little, but can be installed in a convention center and • Downstream chiller: 54.3°F mitigated. detailed in an ASHRAE Journal5 article. Each chiller is designed to produce 37°F • Upstream chiller: 53.8°F • Series evaporators and condensers chilled water and return water back to the should be considered when system • Average: 54.05°F cooling tower at 98.9°F. So the lift of each temperature differences are chiller is 61.9°F. 14°F or larger. The average lift reduction compared to chillers piped in parallel is 7.85°F. Since The chillers could also be piped with both power is proportional to lift, the power Why? Before we examine the system evaporators and condensers in series production is reduced by 12.7 percent. configuration, recall that the two major (Figure 5). To simplify chiller selection, impacts on chiller performance are: the chiller making the coldest chilled When designing a chilled-water system, water receives the coolest tower water. • The cooling load the chiller must optimized system performance is the goal. This is referred to as "counterflow" since satisfy. the condenser water flows counter to the While the chillers become more efficient, • The refrigerant "lift" the compressor chilled water. By installing the chillers in there is additional pump power due to the must develop. series, two levels of thermal staging are higher pressure drop of pumping all the created—reducing each chiller's lift. water through both chillers. In order to Chiller power is proportional to load minimize pumping energy, it’s common to multiplied by lift. So when either load or The upstream chiller receives 55ºF return design series systems with ΔTs of 14°F or lift are reduced, so is chiller power. water and satisfies about half of the load, larger. ANSI/ASHRAE/IESNA 90.1-2016 cooling the water to 45.1°F. The requires a 15°F chilled-water ΔT, so series Chiller power (kW) Load x Lift downstream chiller then cools the water chillers are expected to become more to the desired 37°F. common. One may also consider utilizing So, what is "lift"? . Refrigerant is single-pass evaporators and condensers to On the condenser side, the flow goes in compressed from its evaporator reduce pressure drop and pump power. pressure to its condenser pressure. This the opposite (counter) direction. The difference is referred to as the lift. So downstream chiller receives 85°F water measuring these pressures allows lift to and raises it to 91.3°F. The upstream be determined. But measuring further increases the condenser water temperatures is simpler. temperature to 98.9°F.

Since refrigerant is saturated in both Figure 5. Chiller lift reduction savings series-series (or series counterflow) vessels, at a specific refrigerant condenser water pressure the refrigerant has a specific temperature. The colder the evaporator refrigerant and the warmer the condenser refrigerant, the higher the lift. To simplify, lift is often approximated to the difference between leaving upstream downstream condenser water temperature and leaving evaporator water temperature.

With this background, Figure 4 shows the lift calculation for a chiller used in a plant where both the evaporators and condensers are piped in parallel. These

providing insights for today’s HVAC system designer Trane Engineers Newsletter volume 47–3 5

Now let's take the same design using Figure 6. Chiller lift reduction savings series-series-series (or series-series counterflow) two Trane Duplex® chillers (Figure 6). Each Duplex chiller is essentially a condenser water packaged series-counterflow chiller. Installing two of these chillers in series results in four levels of thermal staging. The lift is reduced to an average of 50.1°F. This staging results in over 19 percent reduction in lift compared to the base parallel configuration. Parallel lift = 61.9ºF Series lift = 50.1ºF 48.9ºF To summarize, chiller power is directly 50.0ºF 50.3ºF proportional to compressor lift. A simple Lift reduction = 51.1ºF way to reduce chiller power by almost 13 (61.9 - 50.1) = 11.8ºF percent is to pipe the chillers in a series- counterflow arrangement. Two Duplex chillers in a series-counterflow arrangement save over 19 percent chiller power (Table 3).

Table 3. Lift versus chiller power Configuration Lift Reduction (%) Parallel 61.9ºF baseline Series counterflow 54.05ºF 12.7 Series counterflow 50.1ºF 19.1 duplex

Additional questions about series Figure 7. Series arrangement of evaporators and condensers chillers. evaporators condensers What about redundancy? To allow either chiller to operate alone if the other chiller circuit 1 circuit 2 circuit 1 circuit 2 is being serviced, manual bypass lines and valves are encouraged. When more circuit 1 circuit 2 circuit 1 circuit 2 than two chillers are installed, chillers can be piped in upstream and downstream "pods" with a manifold between them circuit 1 circuit 2 circuit 1 circuit 2 (Figure 7). This allows any upstream chiller to operate with any downstream chiller and provides the same redundancy as if 3 chiller modules 3 chiller modules all chillers were piped in parallel. Trane Applications Engineering can provide circuit 1 circuit 2 circuit 1 circuit 2 options for series-counter flow system configurations and their operating characteristics. circuit 1 circuit 2 circuit 1 circuit 2

Chiller selection capacity. A system circuit 1 circuit 2 circuit 1 circuit 2 can be designed with each chiller handling 50 percent of the load, but it can be more efficient if the capacity split is optimized. A typical optimized (efficiency and cost) split has the upstream chiller designed to meet 53 percent of the load and the downstream the remaining 47 percent. Another benefit is that both chillers can make the system supply-water temperature in the event the downstream chiller is down or being serviced.

6 Trane Engineers Newsletter volume 47-3 providing insights for today’s HVAC system designer to these system pressure changes. Thus, flow measurement and valve Should pressure- the "controllability" of a conventional characteristics. independent control valves control valve is significantly affected by variations in pressure. The mechanical variety basically be used? combines two valve types into one: What if there were a separate device that • First, a pressure regulating section. Pressure-independent control valves absorbed the system pressure variations (PICVs) have come down in price, are As system pressures change, the so that the control valve always pressure regulating section readily available and help maintain experienced its selected pressure Δ automatically adjusts to keep a system T, but are higher priced than differential? Or what if the valve control pressure-dependent valves. Is the constant pressure across the control gains could be dynamically recalibrated to section of the valve. additional cost worth it? compensate for the changes in system pressure? In other words, what if we • Next, the control section of the valve Definition. Let’s begin by explaining could make the control valve pressure is modulated by the control system to what a pressure-independent control independent? The valve would always adjust the flow through the valve as valve is. Those who have sized a have its selected control authority, it space conditions change. conventional control valve are probably would always pass only its design flow Mechanical pressure-independent valve familiar with the flow coefficient when wide open, and it would offer advantages include: calculation shown below. control stability. • Compact size when compared to a To select the appropriate valve, one must A pressure-independent control valve conventional control valve plus flow first determine the required flow yields a number of advantages: limiting valve package, or to their coefficient; given a flow rate through the • Valve control becomes much more electronic counterparts, valve and a desired pressure drop across stable. To maintain space temperature, the valve. The valve is selected to have an • Depending on the manufacturer, the actuator no longer has to adjust for "authority" that will provide accurate and capability to be paired with any rotary varying system pressures. This will stable control. actuator, extend valve and actuator life. Also, stability positively impacts system • Easy and straightforward to select, efficiency and coil performance. • No additional controls or • Accuracy improves because the valve programming required, Where, is now controlling flow directly. • Near instantaneous response to Cv = valve flow coefficient • Selection is easier. Simply choose a changes in system pressure which Q = flow rate (gpm) valve that provides the needed flow provides optimal control stability. SG = Fluid specific gravity (water = 1.0) rate (gpm)—a flow coefficient On the other hand, an electronic ΔP = Valve pressure drop (psi) calculation is no longer needed pressure-independent valve doesn't upfront. Rearranging this equation, we see that actually maintain a constant differential flow is dependent on the pressure • Installation is easier because pressure- pressure across the control valve surface. differential for a given valve. independent valves automatically Instead, it achieves independent control balance the system. No separate similar to a pressure-independent VAV air balancing valves are required. valve: it includes a flow meter in series with a standard control valve. The Stability and accuracy are particularly electronics calculate the instantaneous important. Anything that causes a valve to flow and pressure across the valve and lose accurate and stable flow control, continuously recalibrates the control Valve performance would be fine if the under any operating condition, can result coefficients to provide stable and pressure differential across the valve was in lower than desired average waterside accurate control. Δ always at the selection condition. But T. Due to a subsequent increase in Δ during operation, other parts of the required flow rate, lower T results in an Likewise, the electronic variety has a system are constantly changing. This increase in pump power and a decrease in number of advantages: chiller plant efficiency. Sometimes the causes the pressure at the valve to • Potential for lower hardware costs, change also. With a fixed coefficient, the impacts are very significant. • Provision for actual load flow therefore has no option but to also There are two different technologies used measurement, vary. to implement pressure-independent • Programmable for alternative control; mechanical pressure regulation Even if nothing has changed in a operation methods. For example, the across the control valve and electronic particular space, the control valve must valve could be set up to limit delta T, gain modulation for the control valve using now modulate to adjust flow in response not flow.

Note: The 2016 ASHRAE Handbook - Systems and Equipment p.47 states that, "...an authority between 0.25 and 0.5 usually provides the right balance between controllability and energy performance." differential pressure of valve authority = differential pressure of valve + differential pessure of branch

providing insights for today’s HVAC system designer Trane Engineers Newsletter volume 47–3 7

Both mechanical and electrionic pressure- Figure 8. Pressure-independent versus conventional valve operation in Atlanta (3rd and 4th floor performance) independent valves have communication capability to enable data sharing and trending.

Advantages should be weighed against the additional complexity. For example, special software is needed to setup and maintain the valves for the life of the product. Also, operators need to be trained in this software and software must be kept current.

Example office building. So now that we understand what a pressure- independent control valve can do for a building system, let's look at case study.

A high rise building in Atlanta was identified as a good candidate because it had existing control problems. Once Summary. By Trane Applications Engineering. To subscribe or those problems were identified and view previous issues of the Engineers Newsletter visit • Pressure-independent valves are trane.com/EN. Send comments to [email protected]. resolved, one floor was retrofitted with a more stable and more accurate, pressure-independent control valve and which improves system ΔT. was compared to a floor that kept the existing conventional control valve • Pressure-independent valves make (Figure 8). valve selection much easier. A lot of References conventional valves are poorly [1] ASHRAE GreenGuide (third edition), Atlanta, GA: First notice how both valves are able to selected, or one valve CV is selected ASHRAE, 2010. control to a high chilled-water ΔT. In other for the whole building, even though words: a pressure independent valve isn't the pressure across valves varies [2] Kelly, D. W. and Tumin Chan, "Optimizing Chilled- required to achieve higher ΔTs. However, significantly by distance from the water Plants," Heating/Piping/Air Conditioning, January, 1999. the conventional control valve is less system pumps. It's highly unlikely a stable than the pressure independent poorly selected valve will provide [3] Taylor, S. “Optimizing Design & Control Of Chilled valve. In particular, notice the good control. Water Plants.” ASHRAE Journal, (December 2011): considerable variation in valve position for 22-34. Available at www.ashrae.org. • Pressure-independent valves are the conventional valve. easier to install since they eliminate [4] Rishel, Burt, "Wire-to-Water Efficiency of Pumping Systems," ASHRAE Journal, April, 2001. As previously discussed, stability the need for balancing valves. improves ΔT. Also, less action on the • Considering the price premium has [5] Schwedler, M. "Series-Series Counterflow for actuator should improve actuator been steadily dropping, pressure- Central Chilled-Water Plants," ASHRAE Journal. (June 2002):pp. 23-29 reliability. independent valves may be cost neutral if all costs, including balancing, are considered.

So are pressure-independent valves worth it? If you're not sure you can get high quality, properly selected, location- specific valves installed on a job; then by all means, specify quality pressure independent valves. It should be well worth the customer's investment.

Trane, the Circle Logo, and TRACE are trademarks of Trane in the United States and other countries. ASHRAE is a regis- tered trademark of the American Society of Heating, Refrigeration, and Air-Conditioning Engineers, Inc. Trane is a brand of Ingersoll Rand, a world leader in creating comfortable, sustainable and efficient environments. Ingersoll Rand's family of brands includes Club Car®, Ingersoll Rand®, Thermo King® and Trane®.

Trane, This newsletter is for informational purposes only and does not constitute legal advice. A business of Ingersoll Rand Trane believes the facts and suggestions presented here to be accurate. However, final design and application decisions are your responsibility. Trane disclaims any responsibility for actions taken on For more information, contact your local Trane the material presented. office or e-mail us at [email protected]

8 Trane Engineers Newsletter volume 47–3 ADM-APN067-EN (October 2018)