Temperature Controls Design

2006 American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Journal (Vol. 48, June 2006). For personal use only. Additional distribution in either paper or digital form is not permitted without ASHRAE s permission. Temperature Controls Design By Ira Goldschmidt, P.E., ember ASHRAE Although temperature control is vital to the operation of an HVAC system, it seems that controls design is often an afterthought to an HVAC design. A variety of perceptions may contribute to this, including: Temperature controls for an HVAC system are dictated by the system s design and can be dealt with at the end of the HVAC design process; Each controls contractor/manufacturer implements controls for a given HVAC system differently, so no need exists for a detailed operation sequence or a point list; Temperature controls contractors provide a design/build system based mainly on the engineer s HVAC system design; and odern temperature controls with their confusing product names, proprietary programming languages, sophisticated algorithms, open/standard protocols, and the challenging trend of equipment-mounted controls have become so complicated that current engineering fees cannot support the necessary design effort. This article debunks these perceptions and promotes the importance of a formalized HVAC controls design process. It provides an overview of the basic principles of controls design and the process of a formalized HVAC controls design process. The Good Old Days Engineers working during the era of pneumatic controls often link the emergence of modern computerized temperature controls to the diminished role of controls design in the overall HVAC design process. This is based on the fact that, because pneumatic controls all worked about the same and were limited in what they could do, almost any engineer could write an operation sequence that would work well and provide competitive bidding. Instead, for these same reasons, a pneumatic control system often was designed for the engineer by a controls contractor as the basis of design. Either way, the process worked fairly well until HVAC systems (and energyefficient controls strategies) became too complicated for pneumatic controls. Nevertheless, pneumatics did offer better opportunities for the engineer to learn controls design. Due to the high level of standardization, controls manufacturers provided design guides with operation sequences and controls designs for a variety of typical systems. Independent controls contractors offered training on pneumatic controls that applied to almost any manu- About the Author Ira Goldschmidt, P.E., is the owner of Goldschmidt Engineering Solutions in Denver. He served on the committees for ASHRAE s BACnet standard, and ASHRAE Guideline 13, Specifying Direct Digital Controls. 3 2 A S H R A E J o u r n a l a s h r a e. o r g J u n e 2 0 0 6

Controls Design Process Examples A List All Systems to be Controlled 1. VAV Air-Handling Unit 2. VAV Terminal Box 3. Boiler Plant 4. Chiller Plant 5. Etcetera B Boiler Starter ixing Valve Secondary Pump VFD HWR Create a diagram of each system (e.g., boiler plant) HWS C List Components & odes Example: Boiler Plant Boiler Occupied ode Start/Stop Unoccupied ode Start/Stop Pumps Constant Volume Primary Pump Variable Volume Secondary Pump ixing Valve Failure D Write Each Sequence Element Boiler Plant Boiler Occupied ode Start/Stop - Start when OAT is under 65 F Unoccupied ode Start/Stop - Start when OAT is under 45 F Pumps Constant Volume Primary Pump - Interlock to boiler Variable Volume Secondary Pump - Interlock to boiler and modulate VFD to maintain loop differential pressure ixing Valve When boiler is on modulate to maintain the HWST to the reset schedule... Failure Alarm Issue alarm when boiler failure is detected Note that italics indicate points identifi ed by the sequence. E Add Points to the Diagram AI Secondary HW Supply Temp. AO ixing Valve AI Primary HW Supply Temp. BI HW Pump Status BO Boiler Enable BI Boiler Status Boiler Starter BI Circ. Pump Status Interlock Start/Stop To Boiler Interlock Start/Stop To Boiler AO Sec. Pump Speed VFD AI Sec. Loop Differential Pressure (Locate Across Furthest Coil) Differential Pressure Transducer F Final Point List Format Choices The control schematics (with points shown) List within each sequence; e.g., points one space temperature sensor AI, one airfl ow AI, one damper modulation AO, one reheat valve AO, etc. Table (see at right) Point Name Hardware Points Software Points AI AO BI BO AV BV Sched. Trend Alarm Show On Graphic Zone Temp. X X X Airfl ow X X X Zone Damper X X Reheat Valve X X X Fan Start/Stop X X

facturer s line of products. However, rather than wish for the return of pneumatics, it is important to note that regardless of the control system technology used, the basic principles of control theory still apply. Therefore, the control system design process should focus on applying sound temperature control principles (through a fully engineered operation sequence), not on the technology. Treating Controls Design Seriously A controls design cannot be properly performed at the end of the HVAC design process. An HVAC system s mechanical and controls designs are interdependent. Therefore, temperature controls design is an iterative process that must be performed as part of the overall HVAC design process. Leaving the controls design to the end of the project will greatly increase the chance that neither the HVAC system nor the controls will work well. A detailed operation sequence and point list are the most important components to a temperature control design. It should be obvious that a controls contractor cannot reasonably be expected to extrapolate an HVAC system s intended operation sequence from the mechanical plans and specifications alone. The importance of a point list is less obvious but can be equally justified by: The cost of a modern, computerized temperature control system is primarily determined by the point quantity and types. While a controls contractor can reasonably extrapolate the point list from a well-written operation sequence, making this effort part of the bid process increases risk that the winning contractor will not provide the intended system; and The process of developing a point list provides the engineer with an excellent check of the operation sequence and the rest of the control system design (more on this later). Controls contractors do not provide a design/build system. In the strictest sense, design/build contracting is based on performance requirements only. In the case of a temperature control system, performance requirements describe the end result of the system s operation. This end result only covers items such as the differing space temperature setpoints for occupied and unoccupied hours, the acceptable deviation from setpoint in terms of quantity and duration, end-to-end accuracy in measuring sensed conditions (not just the accuracy of the associated temperature and other sensors), etc. This is a clear departure from how our industry specifies controls (nor would it work well). Instead, controls specifications actually are prescriptive because they provide extensive details about the means and methods of constructing the system. Unless a prescriptive design is well-engineered, the contractor cannot reasonably be expected to provide what is intended. It is true that temperature control systems are more complicated than in the past, but so are many other HVAC systems. In fact, across the entire A/E design industry, fees are decreasing while services and design complexity are increasing. For the HVAC design industry to thrive, we need to obtain sufficient fees for our services. The alternative is the gradual loss of our leadership role in the industry. One other possible rationalization for the de-emphasis of controls design is that it is viewed as more art than science, and, therefore, not worthy of the same level of design rigor as for other HVAC system components. It is true that controls design does not involve the same types of engineering calculations used on, say, fan/pump sizing. Instead, the mathematical system of Boolean logic is what drives temperature controls design. This is just as engineering-oriented as the design process for any other HVAC system component (if not more so given today s computerized load calculations and equipment selection software programs). The Basic Principles of Controls Design The following is a list and brief discussion of the most important controls principles that need to be understood to successfully design a temperature control system: odern commercial temperature control is provided by a building automation system (BAS), with some lower priority HVAC systems controlled by conventional thermostats (e.g., a unit heater). A BAS is built-up from a variety of computerized controllers typically linked together by a communications system (e.g., BACnet using Ethernet and S/TP, or proprietary protocols, etc.). A BAS control system s logic (the program) involves a series of schedules, control loops, and interlocks all bound together by Boolean logic (e.g., IF/THEN/ELSE) with associated math statements. A BAS uses four types of physical point connections to the building and HVAC system: an AO (or analog output that controls a modulating device like a valve), BO (or binary output that controls a two-state device like a motor start/ stop), AI (or analog input which senses a varying signal such as temperature), and a BI (binary input that senses a two-state process such as the on/off status of a motor). Control loops are at the heart of a BAS system s control logic. This is where DDC (direct digital control) is implemented. A control loop must have an output (typically an AO or BO point), an input (typically an AI point) and a setpoint. Without these three items you do not have a control loop. The two types of control loops are open and closed. They are important to and equally used in the BAS s control logic (though closed loop control generally is preferred where applicable). In an open control loop, the output has no effect on the input. For example, if a boiler (the output) is turned on when the outside air temperature (the input) is below 55ºF (the setpoint), the boiler s operation clearly has no effect on the outside air temperature. An open control loop often can use a BI point or even a software variable as its input. An example is AHU start/stop (the output) based on a schedule (where the time clock is the input and the start time is the setpoint). 3 4 A S H R A E J o u r n a l a s h r a e. o r g J u n e 2 0 0 6

Closed loop control is the most important aspect to a highquality temperature control system. The two types of closed loops are modulated (using an AO as the output) and twoposition (using a BO as the output). odulation generally provides better temperature control than two-position (one reason why chilled water systems provide better control than DX). odulated closed loops (or a closed loop controlling multiple stages of on/off devices) typically use a fairly complex algorithm known as PID (proportional-integral-derivative), which provides the self-correcting feedback effect. Each PID-controlled loop must be tuned (initially and as an ongoing preventive maintenance measure). If not performed properly, control quality can be severely reduced. HVAC Controls Design as a Process It is distressing that most mechanical design firms do not have a standardized temperature controls design process as part of their design standards. This lack of a process generally is justified due to controls design s perceived status as art rather than science. However, controls design can be treated as and, developed into, a process if the design firm is willing to make the investment. The following are the steps in my recommended process: 1. Review the HVAC design to initially judge if it can meet the building s/owner s requirements for controllability and function (e.g., zoning, part-load and low-load operation, areas of varying operating schedules, etc.). 2. Review the HVAC equipment specifications to determine what controls/safeties are specified with the unit. It may be obvious that some of these controls or safeties should be eliminated (and instead provided by the temperature control system). 3. Develop the operation sequence in a top-down fashion. List all unique systems to be controlled (see Box A in the sidebar Controls Design Process Examples ). Create a diagram of each system, including the various components to be controlled (e.g., fan/pump motors, valves/ dampers, etc.) (Box B). These diagrams don t need to be fancy or CAD-drafted (unless you intend to include them on the mechanical plans) and can be merely a visualized image (though don t try this on a complex central plant). Add each of the previous system s components to the systems list. List the modes of operation for each component (e.g., occupied vs. unoccupied, heating vs. cooling, etc.). See Box C as example of this and the previous step. Alternatively, if an operation mode affects multiple components directly, then these modes may be listed as separate items. Also, list the abnormal or failure modes. Finally, describe the sequence of operation for each of the component s modes listed previously (Box D). Each of these sequence elements should be short. If they aren t, this is an indication that not all of the operation modes were listed. The idea here is to create a sequence that looks more like an outline and less like a narrative. This outline approach makes the sequence easier to write (each sequence element is a small and easy to comprehend piece of the whole), and easier to implement by the contractor (the sequence essentially should be one step removed from the software code they develop). Further, be very deliberate with the use of such words as and and or since they can be directly translated into the Boolean logic statements within a BAS program. 4. Create the point list. ost points will be readily identifiable from a well-written sequence. For example, in the sequence element, Start the boiler when the building calls for heat, it is clear that there is a BO point (the boiler). However, the other points are not obvious because the control loop is vaguely described (hence, the emphasis on the phrase well-written in the previous sentence). Add the points identified by the sequence to the diagram (Box E). This step will act as a check of the sequence to ensure that it addresses all of the system s components. (Note that the final format of the point list can be shown on this diagram, or, depending on the complexity of the system or the firm s design standards, it could be a table [Box F] or even a simple list at the end of each system s sequence.) 5. Perform another review of the HVAC system design and specifications following Steps 1 and 2. This may provide greater insight into whether the HVAC system can be controlled as intended. (Determining whether to modify the sequence of operation or the HVAC design will require judgment.) This step also may reveal that further refinement to the equipment specifications is needed. 6. Use Einstein s dictum, ake things as simple as possible, but no simpler, as the final test of the HVAC system and operation sequence designs. Complex HVAC systems (or system functions) will result in complex sequences. ake sure the sequences are as simple as possible (once again judgment is required here). If this simple as possible still seems too complex, then the HVAC design or functions should probably be simplified (HVAC control and building operations do not generally perform well with overly complex operation sequences). 7. Return to Step 1 to reiterate if necessary or as the project development proceeds further. The sidebar, Controls Design Process Example Boxes A through F, shows an example of this process. Two important points about this process are first, it reflects the strong interdependence between an HVAC system s mechanical and controls designs, and second, it is a process that can dictate a change or refinement in the mechanical design a notion some may find akin to the tail wagging the dog. Editing Standard Specifications It is a bit of an oversimplification that the previous process focuses solely on the operation sequence and point list. In reality the engineer also needs to edit portions of a firm s standard temperature controls specification. The question is which portions and how much effort? Here are some guidelines to help with this question: Installation and Product Specifications. A good, standard controls specification will not need much editing because 3 6 A S H R A E J o u r n a l a s h r a e. o r g J u n e 2 0 0 6

most installation and product requirements are either essentially a given or difficult to specify (e.g., quality of workmanship). Only a few installation requirements that need to be modified from project to project. Some examples include operator interface locations, the types of color images to be provided on the operator interfaces, and where to locate the outside air sensor. Concerning products, listing DDC manufacturers generally overrules (and makes moot) most detailed specifications about the controllers. Therefore, the only editing that provides value concerns issues of system architecture (e.g., what types of controllers are preferred for the different HVAC system types [where types refer to programmable vs. application-specific controllers, and which, if any controllers communicate over Ethernet/IP]). The other product area of concern deals with field devices (e.g., sensors). Fortunately, only a few field device products offer a sufficient choice in quality/style to justify specification editing. On the one hand, temperature sensors either will be a thermistor or RTD depending on what the chosen manufacturer uses, so this issue is generally moot. On the other hand, flow and humidity sensors deserve some research and specification editing efforts. What about interoperability? Open/ standard communications protocols (e.g., BACnet) clearly warrant additional specification editing efforts. A detailed discussion of this topic is beyond the scope of this article. However, there are project types (e.g., where an owner has previously selected a sole-sourced BAS) or specification approaches (e.g., allowing only one of the open/standard protocol choices) that do not necessitate devoting considerable time to this issue. What about commissioning? It is well-known that a BAS will probably not function as well as expected without some form of commissioning. Clearly, specification editing efforts should be devoted to this issue. However, the specification of commissioning requirements should be treated as a separate task in the overall mechanical design process rather than an effort specific to the control system design. Conclusion Treating temperature controls design with the same level of rigor as that for the HVAC system s other components is important to retaining the mechanical engineer s leadership role within the building industry. For this and other reasons, it just makes good business sense. If that is not a good enough engineering justification, then consider this: without a successful temperature control system, your HVAC system design (however, perfect it might be otherwise) may be viewed as a failure. 3 8 A S H R A E J o u r n a l a s h r a e. o r g J u n e 2 0 0 6