The control device shown in Figure 4 can be positioned only to a maximum or minimum state (i.e., on or off). Because two-position control is simple and inexpensive, it is used extensively for both industrial and commercial control. A typical home thermostat that starts and stops a furnace is an example.

Controller differential, as it applies to two-position control action, is the difference between a setting at which the controller operates to one position and a setting at which it operates to the other. Thermostat ratings usually refer to the differential (in degrees) that becomes apparent by raising and lowering the dial setting. This differential is known as the manual differential of the thermostat. When the same thermostat is applied to an operating system, the total change in temperature that occurs between a “turn-on” state and a “turn-off” state is usually different from the mechanical differential. The operating differential may be greater because of thermostat lag or hysteresis, or less because of heating or cooling anticipators built into the thermostat.

Anticipation Applied to Two-Position Action. This common variation of strictly two-position action is often used on room thermostats to reduce the operating differential. In heating thermostats, a heater element in the thermostat is energized during on periods, thus shortening the on time because the heater warms the thermostat (heat anticipation). The same anticipation action can be obtained in cooling thermostats by energizing a heater thermostat at off periods. In both cases, the percentage of on time is varied in proportion to the load, and the total cycle time remains relatively constant.

With modulating control, the controller’s output can vary over its entire range. The following terms are used to describe this type of control:

In each of the following examples of modulating control, there is a set of parameters that quantifies the controller’s response. The values of these parameters affect the control loop’s speed, stability, and accuracy. In every case, control loop performance depends on matching (or tuning) the parameter values to the characteristics of the system under control.

Proportional Control. In proportional control, the controlled device is positioned proportionally in response to changes in the controlled variable (Figure 5). A proportional controller can be described mathematically by

where

Vp	=	controller output
Kp	=	proportional gain parameter (inversely proportional to throttling range)
e	=	error signal or offset
Vo	=	offset adjustment parameter

The controller output is proportional to the difference between the sensed value, the controlled variable, and its set point. The controlled device is normally adjusted to be in the middle of its control range at set point by using an offset adjustment. This control is similar to that shown in Figure 5.

Proportional plus Integral (PI) Control. PI control improves on simple proportional control by adding another component to the control action that eliminates the offset typical of proportional control (Figure 6). Reset action may be described by

where

Ki	=	integral gain parameter
θ	=	time

The second term in Equation (2) implies that the longer error e exists, the more the controller output changes in attempting to eliminate the error. Proper selection of proportional and integral gain constants increases stability and eliminates offset, giving greater control accuracy.

Proportional-Integral-Derivative (PID) Control. This is PI control with a derivative term added to the controller. It varies with the value of the derivative of the error. The equation for PID control is

where

Kd	=	derivative gain parameter of controller
de/dθ	=	time derivative of error

Adding the derivative term gives some anticipatory action to the controller, which results in a faster response and greater stability. However, the derivative term also makes the controller more sensitive to noisy signals and harder to tune than a PI controller. Most HVAC control loops perform satisfactorily with PI control alone.

Adaptive Control. An adaptive controller adjusts the parameters that define its response as the dynamic characteristics of the process change. If the controller is PID based, then it adjusts feedback gains. An adaptive controller may be based on other feedback rules. The key is that it adjusts its parameters to match the characteristics of the process. When the process changes, the tuning parameters change to match it. Adaptive control is applied in HVAC systems because normal variations in operating conditions affect the characteristics relevant to tuning. For instance, the extent to which zone dampers are open or closed in a VAV system affects the way duct pressure responds to fan speed, and entering fluid temperatures at a coil affect the way the leaving temperature responds to the valve position. Neural networks and self-learning performance-predictive controllers are sophisticated adaptive controllers.

Fuzzy Logic. This type of control offers an alternative to traditional control algorithms. A fuzzy logic controller uses a series of “if-then” rules that emulates the way a human operator might control the process. Examples of fuzzy logic might include

The designer of a fuzzy logic controller must first define the rules and then define terms such as high, increasing, decreasing, a lot, and a little. Room temperature, for instance, might be mapped into a series of functions that include very low, low, OK, high, and very high. The “fuzzy” element is introduced when the functions overlap and the room temperature is, for example, 70% high and 30% OK. In this case, multiple rules are combined to determine the appropriate control action.

Some control loops include two-position components in a system that exhibits nearly modulating response.

Timed Two-Position Control. This cycles a two-position heating or cooling element on and off quickly enough that the effect on the controlled temperature approximates a modulating device. In this case, a controller may adjust the duty cycle (“on-time” as a percentage of “cycle-time”) as a modulating controlled variable. For example, an element may be turned on for two minutes and off for one minute when the deviation from set point is 3°F. Timed two-position action combines a modulating controller with a two-position controlled device.

Floating Control. This combines a modulating controlled device with a pair of two-position outputs. The controlled device has a continuous operating range, but the actuators that move it only turn on and off. The controller selects one of three operations: moving the controlled device toward its open position, moving it toward its closed position, or leaving the device in its current position. Control is accomplished by applying a pair of two-position contacts with a selected gap between their set points (Figure 7). Generally, a neutral zone between the two positions allows the controlled device to stop at any position when the controlled variable is within the differential of the controller. When the controlled variable falls outside the differential of the controller, the controller moves the controlled device in the proper direction. To function properly, the sensing element must react faster than the actuator drive time. If not, the control functions the same as a two-position control. When applied with a digital controller, floating-point control is also referred to as tri-state control.

Incremental Control. This variation of floating control varies the pulse action to open or close an actuator, depending on how close the controlled variable is to the set point. As the controlled variable comes close to the set point, the pulses become shorter. This allows closer control using floating motor actuators. When applied with a digital controller, incremental control is also referred to as pulse-width-modulation (PWM) control.

Computers perform the control functions in direct digital control (DDC) systems. Uses range from personal computers used as operator interfaces for DDC systems to embedded program microprocessors used to control variable-air-volume boxes, fan-coil units, heat pumps, and other terminal HVAC equipment. Other uses include primary HVAC equipment programmable controllers, distributed network controllers, and servers used to store DDC system trend data. Chapter 40 of the 2019 ASHRAE Handbook—HVAC Applications covers computer components and HVAC computer applications more extensively.

An automatic valve is designed to control the flow of steam, water, gas, or other fluids. It can be considered a variable orifice positioned by an actuator in response to impulses or signals from the controller. It may be equipped with a throttling plug, V-port, or rotating ball specially designed to provide a desired flow characteristic.

Types of automatic valves include the following:

A three-way mixing valve (Figure 8A) has two inlet connections and one outlet connection and a double-faced disk operating between two seats. It is used to mix two fluids entering through the two inlet connections and leaving through the common outlet, according to the position of the valve stem and disk.

A three-way diverting valve (Figure 8B) has one inlet connection and two outlet connections, and two separate disks and seats. It is used to divert flow to either of the outlets or to proportion the flow to both outlets. Three-way diverting valves are more expensive and have more complex applications, and generally are not used in typical HVAC systems.

A two-way globe valve may be either single or double seated. A single-seated valve (Figure 9A) is designed for tight shutoff. Appropriate disk materials for various pressures and media are used. A double-seated or balanced valve (Figure 9B) is designed so that the media pressure acting against the valve disk is essentially balanced, reducing the actuator force required. It is widely used where fluid pressure is too high to allow a single-seated valve to close or to modulate properly. It is not usually used where tight shutoff is required.

A butterfly valve consists of a heavy ring enclosing a disk that rotates on an axis at or near its center and is similar to a round single-blade damper. In principle, the disk seats against a ring machined within the body or a resilient liner in the body. Two butterfly valves can be used together to act like a three-way valve for mixing or diverting. In this arrangement, one valve is set up as normally open and the other is normally closed. In applications with pipe sizes 4 in. and above, butterfly valves are either two position or modulating, and they are less expensive than globe-style valves (see Chapter 47 of the 2020 ASHRAE Handbook—HVAC Systems and Equipment for information on globe valves).

A ball valve consists of a ball with a hole drilled through it, rotating in a valve body. Ball valves are increasingly popular because of their low cost and high close-off ratings. Features that provide flow characteristics similar to or better than globe valves are available.

Pressure-independent valves are control valves with integral pressure regulators. This allows the valve to maintain a constant flow at a given shaft position because the integral pressure regulator maintains a constant differential pressure across the valve’s orifice, regardless of system pressure fluctuations.

Flow Characteristics. Valve performance is expressed in terms of its flow characteristics as it operates through its stroke, based on a constant pressure drop. Three common characteristics are shown in Figure 10 and are defined as follows:

On a pressure-dependent valve, because pressure drop across the valve’s orifice seldom remains constant as its opening changes, actual performance usually deviates from the published characteristic curve. The magnitude of deviation is determined by the overall design. For example, in a system arranged so that control valves or dampers can shut off all flow, pressure drop across a controlled device increases from a minimum at design conditions to total pressure drop at no flow. Figure 11 shows the extent of resulting deviations for a valve or damper designed with a linear characteristic, when selection is based on various percentages of total system pressure drop. To allow for adequate control by the valve, design pressure drop should be a reasonably large percentage of total system pressure drop at the valve (i.e., adequate valve authority) and the system should be designed and controlled so that pressure drop remains relatively constant. Hydronic piping circuits and valve authority are discussed in Chapters 13 and 47 of the 2020 ASHRAE Handbook—HVAC Systems and Equipment, respectively.

Selection and Sizing. Higher pressure drops for control devices are obtained by using smaller sizes, with a possible increase in size of other equipment in the system. Sizing control valves is discussed in Chapter 47 of the 2020 ASHRAE Handbook—HVAC Systems and Equipment.

Steam Valves. Steam-to-water and steam-to-air heat exchangers are typically controlled by regulating steam flow using a two-way throttling valve. One-pipe steam systems require a line-size, two-position valve for proper condensate drainage and steam flow; two-pipe steam systems can be controlled by two-position or modulating (throttling) valves. Maximum pressure drop for steam valves is a function of operating pressure and cannot be exceeded.

Water Valves. Valves for water service may be two- or three-way and two-position or proportional. Proportional valves are used most often, but two-position valves are not unusual and are sometimes essential. Variable-flow systems are designed to keep the pressure differential constant from supply to return. For valve selection, it is safer to assume that the pressure drop across the valve increases as it modulates from fully open to fully closed.

Equal-percentage valves provide better control at part load, particularly in hot-water coils because the coil’s heat output is not linearly related to flow. As flow reduces, more heat is transferred with each unit of water, counteracting the reduction in flow. Equal-percentage valves can provide linear heat transfer from the coil with respect to the control signal.

Actuators. Valve actuators include the following general types:

Automatic dampers are used in air conditioning and ventilation to control airflow. They may be used (1) in a modulating application to maintain a controlled variable, such as mixed air temperature or supply air duct static pressure; or (2) for two-position control to initiate operation, such as opening minimum outdoor air dampers when a fan is started.

Multiblade dampers are typically available in two arrangements: parallel-blade and opposed-blade (Figure 12), although combinations of the two are manufactured. They are used to control flow through large openings typical of those in air handlers. Both types are adequate for two-position control.

When dampers are applied in modulating control loops, a nonlinear relationship between flow and stroke can lead to difficulties in tuning a control loop for performance. Nonlinearity is expressed as variation in the slope of the flow versus stroke curve. Perfect linearity is not required: if slope varies throughout the range of required flow by less than a factor of 2 from the slope at the point where the loop is tuned, nonlinearity is not likely to disrupt performance.

Parallel blades are used for modulating control when the design condition pressure drop of the damper is about 25% or more of the pressure in a subsystem (Figure 13A). Opposed-blade dampers are preferable for modulating control when the damper is about 15% or less of the pressure drop in a subsystem (Figure 13B). A subsystem is defined as a portion of the duct system between two relatively constant pressure points (e.g., the return air section between the mixed air and return plenum tee). A combination may be considered between 15 and 25% damper drop. Single-blade dampers are typically used for flow control in small terminal units.

In Figure 13, A is authority, which is the ratio of pressure drop across the fully open damper at design flow to total subsystem pressure drop, including fully open control damper pressure drop. The curves here are typical for ducted applications. The Air Movement and Control Association (AMCA Standard 500) defined a number of geometric arrangements of dampers for testing pressure losses. The curves in Figure 13 are those of an AMCA 5.3 geometry, which is a fully ducted arrangement with long sections of duct before and after a damper. Other geometric applications, such as plenum or wall-mounted dampers, exhibit different response curves (Felker and Felker 2009; van Becelaere et al. 2004).

Figure 14 shows two parallel-blade (PB) applications, two opposed-blade (OB) applications, and an “anti-PB” (aPB) arrangement. The response curves are not like those of the AMCA 5.3 ducted application. These are “inherent” curves, where pressure drop across a damper is held constant as the damper rotates (so it has 100% authority). In real applications, the authority is lower (higher losses of other system components besides the damper). As system pressure losses increase, the curves move up. Note that PB dampers are significantly above linear in most cases.

Figure 15 shows three more applications with PB and OB dampers. The ducted damper has some disturbance and pressure loss ahead of it, to simulate a more realistic situation than those of AMCA 5.3. Nevertheless, the response curves are similar to AMCA 5.3. The plenum entry dampers show irregular results. Again, these are inherent curves, and lower authority causes the curves to move up toward more flow at smaller angles.

The curves shown here are typical, but do not represent every scenario. Thousands of installation variations exist, and slight variations in response always occur. For additional application examples and greater detail, see ASHRAE research project RP-1157 (van Becelaere et al. 2004).

Application. Dampers require engineering to achieve defined goals. A common application is a flow control damper, which modulates airflow. The curves in Figure 13 can be used to pick a damper with a pressure drop and authority that provides near-linear response. Another common application is economizer outdoor air, return air, and exhaust air dampers. Selection of these dampers depends on the system design, as discussed in ASHRAE Guideline 16-2014.

Damper leakage is a concern, particularly where tight shutoff is necessary to significantly reduce energy consumption. Also, outdoor air dampers in cold climates must close tightly to prevent coils and pipes from freezing. Low-leakage dampers cost more and require larger actuators because of friction of the seals in the closed position; however, the energy savings offset the extra cost.

Actuators. Either electricity or compressed air is used to actuate dampers.

Pneumatic damper actuators are similar to pneumatic valve actuators, except that they have a longer stroke or the stroke is increased by a multiplying lever. Increasing air pressure produces a linear shaft motion, which, through a linkage, moves the crank arm to open or close the dampers. Releasing air pressure allows a spring to return the actuator. Double-acting actuators without springs are available also.

Electric damper actuators can be proportional or two-position. They can be spring return or nonspring. The simplest form of control is a floating three-point controller, in which contact closures drive the motor clockwise or counterclockwise. In addition, a variety of standard electronic signals from electronic controllers or DDC systems, such as 4 to 20 mA, 2 to 10 V (dc), or 0 to 10 V (dc), can be used to control proportionally actuated dampers.

Most modern actuators are electronic and use more sophisticated methods of control and operation. They are inherently positive positioning and may have communications capabilities similar to industrial-process-quality actuators.

A two-position spring-return actuator moves in one direction when power is applied to its internal windings. When no power is present, the actuator returns (via spring force) to its fail-safe (normal) position. Depending on how the actuator is connected, this action opens or closes the dampers. A proportional actuator may also have spring-return action.

Mounting. Damper actuators are mounted in different ways, depending on the size and accessibility of the damper, and the power required to move the damper. The most common method of mounting electric actuators is directly over the damper shaft (direct coupling) with no external linkage. Actuators can also be mounted in the airflow on the damper frame and be linked directly to a damper blade (though this may deform blades and make access for repairs difficult), or mounted outside the duct and connected to a crank arm attached to a shaft extension of one of the blades. Mounting methods that link the actuator to a single blade are not recommended because of the potential to bend the single drive blade or to have the blade-to-blade linkages get out of adjustment.

Large dampers may require two or more actuators, which are usually mounted at separate points on the damper. An alternative is to install the damper in two or more sections, each section being controlled by a single damper actuator. Positive positioners may be required for proper sequencing, as with a small damper controlled to maintain healthy indoor air quality, with a large damper being independently controlled for economy-cycle free cooling.

Where accurate positioning of a modulating pneumatic damper or valve is required, use positive positioners. A positive positioner provides up to full supply control air pressure to the actuator for any change in position required by the controller. A pneumatic actuatorwithout a positioner may not respond quickly or accurately enough to small changes in control pressure caused by friction in the actuator or load, or to changing load conditions such as wind acting on a damper blade. A positive positioner provides finite and repeatable positioning change and allows adjustment of the control range (spring range of the actuator) to provide a proper sequencing control of two or more control devices.

Temperature-sensing elements generally detect changes in either (1) relative dimension (caused by differences in thermal expansion), (2) the state of a vapor or liquid, or (3) some electrical property. Within each category, there are various sensing elements to measure room, duct, water, and surface temperatures. Temperature-sensing technologies commonly used in HVAC applications are as follows:

Humidity sensors, or hygrometers, measure relative humidity, dew point, or absolute humidity of ambient or moving air. Transmitters convert the signal from the humidity sensor to an industry-standard output such as 0 to 10 V or 4 to 20 mA. Some transmitters compensate for temperature variations. Two types of sensors that detect relative humidity are mechanical hygrometers and electronic hygrometers.

Electronic hygrometers can use either resistance or capacitance sensing elements. The resistance element is a conductive grid coated with a hygroscopic substance. The grid’s conductivity varies with the water retained; thus, resistance varies with relative humidity. The conductive element is arranged in an AC-excited Wheatstone bridge and responds rapidly to humidity changes.

The capacitance element is a stretched membrane of nonconductive film, coated on both sides with metal electrodes and mounted in a perforated plastic capsule. The response of the sensor’s capacity to rising relative humidity is nonlinear. The signal is linearized and temperature is compensated in the amplifier circuit to provide an output signal as relative humidity changes from 0 to 100%.

The chilled-mirror humidity sensor determines dew point rather than relative humidity. Air flows across a small mirror in the sensor. A thermoelectric cooler lowers the surface temperature of the mirror until it reaches the dew point of the air. Condensation on the surface reduces the amount of light reflected from the mirror compared to a reference light level.

Dispersive infrared (DIR) technology can be used to sense absolute humidity or dew point. It is similar to technology used to sense carbon dioxide or other gases. Infrared water vapor sensors are optical sensors that detect the amount of water vapor in air based on the infrared light absorption characteristics of water molecules. Light absorption is proportional to the number of molecules present. An infrared hygrometer typically provides a value of absolute humidity or dew point, and can operate in diffusion or flow-through sample mode. This type of humidity sensor is unique in that the sensing element (a light detector and an infrared filter) is behind a transparent window and is never exposed directly to the sample environment. As a result, this sensor has excellent long-term stability, long life, and fast response time; is not subject to saturation; and operates equally well in very high or low humidity. Previously used solely for high-end applications, infrared hygrometers are now commonly used in HVAC applications because they cost about the same as mid-range-accuracy (1 to 3%) humidity sensors.

A pneumatic pressure transmitter converts a change in absolute, gage, or differential pressure to a mechanical motion using a bellows, diaphragm, or Bourdon tube mechanism. When connected through appropriate links, this mechanical motion produces a change in air pressure to a controller. In some instances, sensing and control functions are combined in a single component, a pressure controller.

An electronic pressure transducer may use mechanical actuation of a diaphragm or Bourdon tube to operate a potentiometer or differential transformer. Another type uses a strain gage bonded to a diaphragm. The strain gage detects displacement resulting from the force applied to the diaphragm. Capacitance transducers are most often used for measurements below 1 in. of water because of their high sensitivity and repeatability. Electronic circuits provide temperature compensation and amplification to produce a standard output signal.

Orifice plate, pitot-static tube, venturi, turbine, magnetic flow, thermal dispersion, vortex shedding, and ultrasonic meters are some of the technologies used to sense fluid flow. In general, pressure differential devices (orifice plates, venturi, and pitot tubes) are less expensive and simpler to use, but have limited range; thus, their accuracy depends on how they are applied and where in a system they are located.

More sophisticated flow devices, such as turbine, magnetic, and vortex shedding meters, usually have better range and are more accurate over a wide range. If an existing piping system is being considered for retrofit with a flow device, the expense of shutting down the system and cutting into a pipe must be considered. In this case, a noninvasive meter, such as an ultrasonic flow meter, may be cost effective.

There are two types of ultrasonic flow meters. One uses the Doppler effect, which is better for monitoring flow in slurry applications. The other type uses the transit time effect, which is more accurate at lower flows and with standard nonslurry fluids.

For air velocity metering, pitot-static tubes provide a naturally larger signal change at high velocities, with limitations on their application below 500 to 600 fpm. Vortex shedding for airflow applications has similar low-velocity limitations. Thermal dispersion sensors provide a naturally larger signal change at lower velocities without appreciable losses through velocities common in most ventilation systems, which makes them more suitable for applications below 1000 fpm.

Indoor air quality control can be divided into two categories: ventilation control and contamination protection. In spaces with dense populations and intermittent or highly variable occupancy, ventilation can be more efficiently applied by detecting changes in population or ventilation requirements [demand-controlled ventilation (DCV)]. This involves using time schedules and population counters, and measuring the indoor/outdoor differential levels of carbon dioxide (CO₂) or other contaminants in a space. Changes in differential CO₂ mirror changes in space population; thus, the amount of outdoor air introduced into the occupied space can then be controlled. Demand control helps maintain proper ventilation rates at all levels of occupancy. Control set-point levels for carbon dioxide are determined by the specific relationships between differential CO₂, rate of CO₂ production by occupants, the variable airflow rate required by the changing population, and a fixed amount of ventilation required to dilute building-generated contaminants unrelated to CO₂ production. ASHRAE Standard 62.1 and its user’s manual (ASHRAE 2013) provide further information on ventilation for acceptable indoor air quality and DCV for single-zone systems.

Contamination protection sensors monitor levels of hazardous or toxic substances and issue warning signals and/or initiate corrective actions through the building automation system (BAS). Sensors are available for many different gases. The carbon monoxide (CO) sensor is one of the most common, and is often used in buildings wherever combustion occurs (e.g., parking garages). Refrigerant-specific sensors are used to measure, alarm, and initiate ventilation purging in enclosed spaces that house refrigeration equipment, to prevent occupant suffocation upon a refrigerant leak (see ASHRAE Standard 15 for more information). The type selected, substances monitored, and action taken in an alarm condition all depend on where these sensors are applied.

Analog lighting level transmitters packaged in various configurations allow control of ambient lighting levels using building automation strategies for energy conservation. Examples include ceiling-mounted indoor light sensors used to measure room lighting levels; outdoor ambient lighting sensors used to control parking, general exterior, security, and sign lighting; and interior skylight sensors used to monitor and control light levels in skylight wells and other atrium spaces.

Passive electronic devices that sense the magnetic field around a conductor carrying current allow low-cost instrumentation of power circuits. A wire in the sensor forms an inductive coupling that powers the internal function and senses the level of the power signal. These devices can provide an analog output signal to monitor current flow or operate a switch at a user-set level to turn on an alarm or other device.

Digital controls use microprocessors to execute software programs that are customized for use in commercial buildings. Controllers use sensors to measure values such as temperature and humidity, perform control routines in software programs, and exert control using output signals to actuators such as valves and electric or pneumatic actuators connected to dampers. The operator may enter parameters such as set points, proportional or integral gains, minimum on and off times, or high and low limits, but the control algorithms make the control decisions. The controller scans input devices, executes control algorithms, and then positions the output device(s), in a stepwise scheme. The controller calculates proper control signals digitally rather than using an analog circuit or mechanical change, as in electric/electronic and pneumatic controllers. Use of digital controls in building automation is referred to as direct digital control (DDC).

Digital controls can be used as stand-alones or can be integrated into building management systems (BMSs) through network communications. Simple controls may have a single control loop that can perform a single control function (e.g., temperature control of a unit ventilator); larger versions can control a larger number of loops or even perform complex strategies such as managing energy allocation to multiple processes and open loop controls, and balancing multiple objectives (e.g., comfort, energy consumption, equipment wear).

Advantages of digital controls include the following:

A single control that is fixed in functionality with flexibility to change set points and small configurations is called an application-specific controller. Many manufacturers include application-specific controls with their HVAC equipment, such as air-handling units and chillers.

Firmware and Software. Preprogrammed control routines, known as firmware, are sometimes stored in permanent memory such as programmable read-only memory (PROM) or electrically programmable read-only memory (EPROM), and the application or set points are stored in changeable memory such as electrically erasable programmable read-only memory (EEPROM). The operator can modify parameters such as set points, limits, and minimum off times within the control routines, but the primary program logic cannot be changed without replacing the memory chips.

User-programmable controllers allow the algorithms to be changed by the user. The programming language provided with the controller can vary from a derivation of a standard language to custom language developed by the controller’s manufacturer, to graphically based programming. Preprogrammed routines for proportional, proportional plus integral, Boolean logic, timers, etc., are typically included in the language. Standard energy management routines may also be preprogrammed and may interact with other control loops where appropriate.

Digital controllers can have both preprogrammed firmware and user-programmed routines. These routines can automatically modify the firmware’s parameters according to user-defined conditions to accomplish the control sequence designed by the control engineer.

Operator Interface. Some digital controllers (e.g., a programmable room thermostat) are designed for dedicated purposes and are adjustable only through manual switches and potentiometers mounted on the controller. This type of controller cannot be networked with other controllers. A direct digital controller can have manually adjustable features, but it is more typically adjusted either through a built-in LED or LCD display, a hand-held device, or a terminal or computer. The direct digital controller’s digital communication allows remote connection to other controllers and to higher-level computing devices and host operating stations.

A terminal allows the user to communicate with the controller and, where applicable, to modify the program in the controller. Terminals can range from hand-held units with an LCD display and several buttons to a full-sized console with a video monitor and keyboard. The terminal can be limited in function to allow only display of sensor and parameter values, or powerful enough to allow changing or reprogramming the control strategies. In some instances, a terminal can communicate remotely with one or more controllers, thus allowing central displays, alarms, and commands. Usually, hand-held terminals are used by technicians for troubleshooting, and full-sized, fully functional terminals are used at a fixed location to monitor the entire digital control system. Standard Internet browsers can sometimes be used to access system information, complemented with online help and application libraries, plus cloud-based troubleshooting and graphic user interface design tools.

For two-position control, the controller output may be a simple electrical contact that starts a fan or pump, or one that actuates a spring-return valve or damper actuator. Electrical contacts can be normally open (NO), normally closed (NC), or have one of each (double throw). An output that has three terminals (common, NO, and NC) is called single-pole, double-throw (SPDT). SPDTs simplify designs, because the same device can be used as NO or NC depending on the requirements of the application. In some cases, both contacts are used (e.g., to send a signal to one device or another, depending on whether the system is in cooling or heating mode). Both single-pole, single-throw (SPST) and SPDT circuits can be used for timed two-position action.

Floating control uses two NO output contacts with a neutral zone where neither contact is made. This control is used with reversible motors; it has a slow response and a wide throttling range.

Output for floating control is a SPDT switching circuit with a neutral zone where neither contact is made. This control is used with reversible motors; it has a slow response and a wide throttling range.

Pulse modulation control is an improvement over floating control. It provides closer control by varying the duration of the contact closure. As the actual condition moves closer to the set point, the pulse duration shortens for closer control. As the actual condition moves farther from the set point, the pulse duration lengthens.

Proportional control gives continuous or incremental changes in output signal to position an electrical actuator or controlled device.

Pneumatic receiver-controllers are normally combined with pneumatic elements that use a mechanical force or position reaction to the sensed variable to obtain a variable-output air pressure. Control is usually proportional, but other modes (e.g., proportional-plus-integral) can be used. These controllers are generally classified as nonrelay or relay, and as direct-acting or reverse-acting.

The nonrelay (single-pipe) pneumatic controller uses low-volume output. A relay (two-pipe) pneumatic controller actuates a relay device that amplifies the air volume available for control. The relay provides quicker response to a variable change.

Direct-acting controllers increase the output signal as the controlled variable increases. Reverse-acting controllers increase the output signal as the controlled variable decreases. For example, a reverse-acting pneumatic thermostat increases output pressure when the measured temperature drops and decreases output pressure when the measured temperature rises.

Thermostats combine sensing and control functions in a single device. Microprocessor-based thermostats have many of the following features.

Relays provide a means for one electrical source to switch to a different electrical circuit. The switched voltage may be the same or different. Relay configurations include several variations:

Electric relays can be used to start and stop electric heaters, burners, compressors, fans, pumps, or other apparatus for which the electrical load is too large to be handled directly by the controller. Other uses include time-delay and circuit-interlocking safety applications. Form letters are part of an industry consensus that defines the arrangement of contacts for relays and switches. The three most common types in HVAC are forms A, B, and C. Form A relays are single pole, single throw, normally open (SPST-NO). Form B relays are single pole, single throw, normally closed (SPST-NC). Form C relays are single pole, double throw (SPDT) and can be either normally open or normally closed, so control panels can be built using all form C (SPDT) relays. In small sizes, the added cost of the second contact is insignificant. As current ratings go up, forms A and B become more cost effective.

Time-delay relays, similar to control relays, include an adjustable time delay that is set using dipswitches and/or a knob. The device is either a delay-on-make (on delay) or delay-on-break (off delay). The time delay is either in seconds to minutes or minutes to hours. Uses include preventing multiple pieces of equipment from starting simultaneously and overloading the electrical supply, ensuring equipment completes its warm-up cycle, preventing triggering of nuisance alarms during short transitions (e.g., before equipment has reached enough speed for its status sensors to activate), and temporarily changing the mode of operation (e.g., thermostat override timer during unoccupied hours).

Power relays handle high-power switching of electrical loads in motor control centers and lighting control applications. They may be of open-frame construction or enclosed, single or multiple pole, panel mounted or field mounted, with or without auxiliary contacts.

Solid-state relays are optically isolated relays used to switch voltages up to 600 V with high amperage loads using a form A contact. They have a high surge dielectric strength, and reverse voltage protection. They operate using an input voltage of 4 to 32 V DC. These devices are easily affected by high temperatures and induced currents, but their life expectancy, measured in number of operations, is much greater than that of electromechanical switches, so they are used wherever switching every few minutes or seconds is required, as in pulse-width modulation (PWM) controls.

Control relay sockets are used with control relays to terminate all the wires that need to be connected to the relay. They may have either blade- or pin-type terminals. Sockets make wiring easier (without the relay getting in the way), provide a way to ensure equipment is not started until the system is commissioned, and allow relay replacement without redoing the wiring terminations.

Auxiliary contacts are sometimes used to determine equipment status. The auxiliary contacts close when the starter’s contactor closes, so they do not account for broken belts, couplings, or wires; locked rotors; flow obstructions; or missing power.

Differential pressure switches are used for status indication for air filters, fans, and pumps. Additionally, they can be used to provide flow and level status and in safety circuits to protect system components.

Current switches are current-sensing relays used to monitor the status of electrical devices. Typically, they have one or more adjustable current set points. They should be adjusted at start-up so that if the fan or pump motor coupling breaks, the current switch does not indicate on status.

Paddle switches indicate flow of water or air and are used to sense the status of pumps and fans.

Equipment with embedded controls such as boilers, chillers, terminal units, and variable-speed pumps and fans usually provide their status through contacts and over serial communication.

Limit switches convert mechanical motion into a switching action. Common applications include valve and damper position and proximity feedback.

Manual switches, either two-position or multiple-position with single or multiple poles, are used to switch equipment from one state to another.

Auxiliary switches on electric or electronic equipment, such as valve and damper actuators, sensors, and variable-speed controls are used to select a sequence or mode of operation.

Moisture switches are used to detect moisture in drain pains, under raised floors, and in containment areas to shut down equipment and/or alert operators before flooding or damage occurs.

Level switches (float, conductive probe, or ultrasonic) are used to start/stop equipment, either in normal operation, such as supply or drain pumps, or in protection circuits, as in boilers or chillers.

Time switches (mechanical or electronic) turn electrical loads on and off, based on a 24 h/7 day or 24 h/365 day schedule.

Analog time switches are set manually to control an electrical load. They have either normally open or normally closed contacts. Timing is either in minutes or hours, and may have an override hold feature to keep the load on or off continuously.

Digital time switches are set manually to control an electrical load and have either normally open or normally closed contacts. The timing function is either in minutes or hours, and sometimes combines a regular weekly calendar with a yearly exceptions calendar used for holidays and other special events. They may have an override hold feature to keep the load on or off continuously, and a flash or beeper option to notify the operator that the load will be turning off shortly.

Transducers are devices that convert one signal (sensor reading or command) into a signal of another type that conveys some or all of the original information. Examples include pressure into voltage, voltage into current, voltage into contact closure, and rotational speed into pulse frequency. Transducers may convert proportional input to either proportional or two-position output.

The electronic-to-pneumatic transducer (EPT is used in many applications. It converts a proportional electronic output signal into a proportional pneumatic signal and can be used to combine electronic and pneumatic control components to form a control loop (Figure 17). Electronic components are used for sensing and signal conditioning, whereas pneumatic components are used for actuation. The electronic controller can be either analog or digital.

The EPT presents a special option for retrofit applications. An existing HVAC system with pneumatic controls can be retrofitted with electronic sensors and controllers while retaining the existing pneumatic actuators (Figure 18).

Signal transducers to change one standard signal into another. Variables usually transformed include voltage [0 to 10, 0 to 5, 2 to 10 V (DC)], current (4 to 20 mA), resistance (0 to 135 Ω), pressure (3 to 15, 0 to 20 psig), phase cut voltage [0 to 20 V (DC)], pulse-width modulation, and time duration pulse. Signal transducers allow use of an existing control device in a retrofit application.

Variable-frequency drives (VFDs) control AC motors’ speed by using electronics to vary the frequency and voltage amplitude of the electrical input power. VFDs are commonly applied to fans, heat recovery wheels, pumps, and compressors.

Occupancy sensors are used to automatically adjust controlled variables (e.g., lighting, ventilation rate, temperature) based on occupancy.

Potentiometers are used for manual positioning of proportional control devices, for remote set-point adjustment of electronic controllers, and for position feedback.

Smoke detectors sense smoke and other combustion products in air moving through HVAC ducts. Through sampling tubes, selected based on duct size, they test air moving in the duct. For duct applications, using only photoelectric sensor heads is recommended. The device typically has two alarm contacts, used to shut down the associated equipment and provide remote indication, and a trouble contact, which monitors incoming power and removal of the detector head. Where a fire alarm system is installed, the smoke detectors must be listed for use with the fire alarm system. Addressable fire alarm systems may use a separate programmable relay for fan shutdown rather than a hard-wired connection to the detector.

Transient voltage surge suppressors (TVSSs), formerly called lightning arrestors, protect communication lines and critical power lines between buildings or at building entrance vaults against high-voltage transients caused by VFDs, motors, transmitters, and lightning. To be effective, they must be grounded to a grounding rod in compliance with the National Electrical Code^® (NFPA Standard 70) or local equivalent, and the manufacturer’s recommendations.

Transformers provide current at the required voltage.

Regulated DC power supply devices convert AC voltage into a DC voltage, kept constant independently of the current draw, and usually between 12 and 24 V DC. They may be used to power temperature, pressure, and humidity transmitters.

Fuses are safety devices with a specific amp rating, used with power supplies, circuit boards, control transformers, and transducers. They may be rated for either high or low inrush currents, and are available in slow-blow or fast-acting models.

Step controllers operate several switches in sequence using a proportional electric or pneumatic input signal. They are commonly used to control several steps of heating or refrigeration capacity. They may be arranged to prevent simultaneous starting of compressors and to alternate the sequence to equalize wear.

Power controllers control electric power input to resistance heating elements. They are available with various ratings for single- or three-phase heater loads and are usually arranged to regulate power input to the heater in response to the demands of the proportional electronic or pneumatic controllers. Silicon controlled rectifiers (SCRs) are the most common form of power controller used for electric heat. Solid-state controllers in fast-switching two-position control modes are used because they are without mechanical contacts, which wear out too quickly to be of practical use for this application, and can arc when power is applied or removed.

High-temperature limits are safety devices that shut down equipment when the temperature exceeds the high limit set point. They are typically set at 125 to 150°F for heating air applications and 195 to 205°F for hot-water applications. A manual or automatic reset reactivates the device once the condition has cleared.

Low-temperature limits for ducts are used to protect chilled-water and preheat coils from freezing. They typically have a 20 ft long vapor-charged sensing element, set at 35°F, that shuts down equipment when the temperature in any 12 or 18 in. section falls below its set point. They may be manually or automatically reset. The device must be mounted parallel to the coil tubes with capillary mounting clips for proper measurement. Limits may be either SPST or double-pole, double-throw (DPDT), with the second contact used to alert operators that the device has tripped.

Three- and five-valve manifolds protect water differential pressure sensors from overpressurization during installation, start-up, shutdown, system testing, and maintenance. Three-valve manifolds are comprised of two isolation valves and a bypass valve. A five-valve manifold includes two additional valves to allow online calibration. Depending on the application, snubbers may be required, as well.

Snubbers, made of brass or stainless steel, stop shocks and pulsations caused by fluid hammering and system surges. Two types of pressure snubbers are used in HVAC applications: porous and piston. A porous snubber has no moving parts and uses a porous material to stop device damage. A piston snubber uses a moving piston inside a tube that moves up and down to stop device damage and push away any sediment or scale that may clog the system’s monitoring devices. Depending on the application, the type of piston to be used may need to be specified.

Steam pigtail siphons protect pressure transmitters from the high temperature of steam. They are typically made from steel or stainless steel of a specific length with a loop. The temperature of the medium being monitored determines the length and material. Most pressure devices have an operating temperature range of 0 to 200°F.

Thermostat guards are plastic or metal covers that protect switches, thermostat controllers, and sensors from damage, tampering, and unauthorized adjustment.

Enclosures/control panels may be used indoors or outdoors to protect equipment and people. In North America, enclosures are rated by the National Electrical Manufacturers Association (NEMA Standard 250) and Electrical and Electronic Manufacturer Association of Canada (EEMAC); elsewhere, they receive an ingress protection (IP) rating, which describes their degree of resistance to ingress of objects, dust, and water (IEC Standard 60529).

Pilot lights are replaceable incandescent or light-emitting diode (LED) lights that indicate the status or mode of operation of mechanical and electrical equipment. They are panel-mounted, and may be round or flat, of various colors, and powered using either AC or DC power. They are typically installed in an enclosure rated for the application under NEMA Standard 250, or with an appropriate IP rating.

Strobes use a high-intensity xenon flash tube to generate a high-intensity light that is visible in all directions. If this device is used in a safety application (e.g., refrigeration monitoring), in the United States, it must comply with UL Standard 1971 and the Americans with Disabilities Act (ADA), Title III; in Canada, with ULC Standard S526; elsewhere, it must meet requirements of British Standards Institution (BSI) Standard BS EN 54–23.

Horns provide an audible tone with a specified loudness rated in decibels (dB), and are mounted in a panel or junction box. The tone may be continuous, warbled, short beeps, or long beeps. The tone should be at least 10 dB above the ambient noise level in the area that the device is mounted. The operating voltage may be either DC or AC. When used in a life safety application, it must comply with NFPA Standard 72 or BSI Standard BS EN 54-3.

Often, a single BAS applies different network technologies at different points in the system. For example, a relatively low-speed, inexpensive network with relatively primitive functions may link a group of room controllers to a supervisory controller, while a faster, more sophisticated network links the supervisory controller to its peers and to one or more operator workstations. Figure 20 demonstrates the representation and interaction of a multitier BAS system architecture as described in ASHRAE Guideline 13-2015. The architecture model has three tiers. Tier 1 is the enterprise-level information technology (IT) network and provides access to user interfaces, dashboards, kiosks, and application interfaces. Tier 2 is the building network infrastructure and includes the building control network connectivity to the building IT network. Tier 3 is the control network and the associated controllers, equipment, sensors, and actuators.

The opposite extreme is completely flat network architecture, which links all devices through the same network. A flat architecture is more viable in small systems than in large ones, because economic constraints typically dictate that low-cost (and therefore low-speed) networks be used to connect field-level controllers. Because of their performance limitations, these networks do not scale well to large numbers of devices. As the cost of electronics for communication drops, flatter networks become more feasible.

Network structure can affect

The relative merits of one structure versus another depend on the communication functions required, hardware and software available for the task, and cost. For a given job, there is probably more than one suitable structure. Product capabilities change quickly. Engineers who choose to specify network structure must be aware of new technologies to take advantage of the most cost-effective solutions.

Some BAS networks use other networks to connect segments of the BAS. This occurs

In each case, the link between BAS segments must be considered part of the BAS network when evaluating function, security, and performance. The link also raises new issues. The connecting segment is likely to be outside the control of the owner of the BAS, which could affect availability of service. Traffic and bandwidth issues may have to be addressed with the link administrator.

Using a dial-up connection to interact with a remote building or to serve a remote operator requires consideration of which segment may dial the other, what circumstances trigger the call, and security implications. Handling interbuilding communication through intranet or Internet connections, with the operator interface provided by a web browser communicating with a web server, has largely replaced dial-up connections.

The transmission medium is the foundation of the network. It is usually, but not always, cable. The cable may be plenum rated or non-plenum rated, depending on the installation application. Where physical cable connection is not possible or practical, devices may transfer information using wireless technologies, such as radio waves or infrared light.

RS-485 is the data link layer standard most commonly used in BAS for communications inside a building. It works over a single shielded twisted pair, 18 to 24 gage, and can reach distances of up to 1500 ft between repeaters and up to 120 devices per segment. It has no polarity (i.e., no positive and negative wires), which reduces wiring errors, and in most cases is optically isolated at the controllers, allowing for different power sources without risk of short circuits.

ANSI/TIA/EIA Standard 568-B.1 provides IT cabling specifications for commercial buildings. This cabling is used for computers and telephone networks and can also be used for low-level BAS, access, security, or fire detection systems, but close work between the BAS and IT design and execution teams is required. The potential for third-party devices halting all network traffic discourages its use for control and security applications, as does the location of outlets prescribed in the standard (in occupied spaces instead of plenums and mechanical rooms where the controllers are). Also, IT network administrators are sometimes reluctant to allow connection of unfamiliar devices they cannot configure.

Cable length. Maximum length varies with cable type, transmission speed, and protocol.

100 Ω Balanced Twisted-Pair Communications Outlet/Connector. Each four-pair cable terminates on an eight-position modular jack, and all unshielded twisted-pair (UTP) and screened twisted-pair (ScTP) telecommunications outlets must meet the requirements of IEC Standard 60603-7, as well as ANSI/TIA/EIA Standard 568-B.2 and the terminal marking and mounting requirements of ANSI/TIA/EIA Standard 570-B.

Twisted-Pair Copper Cable. A twisted-pair cable consists of multiple twisted pairs (typically 24 gage) of wire covered by an overall sheath or jacket. Varying the number of twists for each pair relative to the other pairs in the cable can greatly reduce crosstalk (interference between signals on different pairs).

Screened twisted-pair (ScTP) cable is similar in construction to UTP data cabling, except that ScTP has a foil shield between the conductors and outer jacket, as well as a drain wire used to minimize interference-related problems. ScTP cable is preferred over UTP cable in environments where high immunity and/or low emissions are critical. It also allows less crosstalk than UTP. However, ScTP requires more labor-intensive installation, and any break or improper grounding of the shield reduces its overall effectiveness.

Category 3 cable connects hardware and patch cords that are rated to a maximum frequency of 16 MHz. This cable is rated up to 16 megabits per second. This category is usually the lowest level of cable installed and is used mainly for voice and low-speed networks.

Category 5 cable connects hardware and patch cords that are rated to a maximum frequency of 100 MHz. The actual data transmission rate varies with the compression scheme used. Category 5 was defined in ANSI/TIA/EIA Standard 568-A, but is no longer recognized in the new ANSI/TIA/EIA Standard 568-B.1.

Category 5e cable connects hardware and patch cords that are rated to a maximum frequency of 100 MHz. The actual data transmission rate varies with the compression or encoding scheme used. Category 5e, the lowest category recommended for data installations, is defined by ANSI/TIA/EIA Standards 568-B.1 and B.2. This cable is rated up to 100 megabits per second.

Category 6 cable connects hardware and patch cords that are rated to a maximum frequency of 250 MHz, though the actual data transmission rate varies with the compression scheme used. UTP or STP cable is currently the most common medium. This cable is rated up to 1 gigabit per second.

Category 6e cable connects hardware and patch cords that are rated to a maximum frequency of 250 MHz, though actual data transmission rate varies with the compression scheme used. With all Category 6 systems, an eight-position jack is mandatory in the work area.

Category 7 cable connects hardware and patch cords that are rated to a maximum frequency of 600 MHz, though actual data transmission rate varies with the compression scheme used. This category is still in development and uses a braided shield surrounding all four foil-shielded pairs to reduce noise and interference. Depending on future technological developments, the current RJ-45 connector will not be used in Category 7.

Fiber-Optic Cable. Fiber-optic cable uses glass or plastic fibers to transfer data in the form of light pulses, which are typically generated by either a laser or an LED. Fiber-optic cable systems are classified as either single-mode fiber or multimode fiber systems. Table 1 compares their characteristics.

Light in a fiber-optic system loses less energy than electrical signals traveling through copper and has no capacitance. This translates into greater transmission distances and dramatically higher data transfer rates, which impose no limits on a BAS. Fiber optics also have exceptional noise immunity. However, the necessary conversions between light-based signaling and electricity-based computing make fiber optics more expensive per device, which sometimes offsets its advantages.

Structured Cabling. ANSI/TIA/EIA Standard 568 allows cable planning and installation to begin before the network engineering is finalized. It supports both voice and data. The standard was written for the telecommunications industry, but cabling is gaining recognition as building infrastructure, and the standard is being applied to BAS networks as well.

ANSI/TIA/EIA Standard 568-B specifies star topology (each device individually cabled to a hub) because connectivity is more robust and management is simpler than for busses and rings. If the wires in a leg are shorted, only that leg fails, making fault isolation easier; with a bus, all drops would fail.

The basic structure specified is a backbone, which typically runs from floor to floor in a building and possibly between buildings. Horizontal cabling runs between the distribution frames on each floor and the information outlets in the work areas.

Wireless Networks. The rapid maturity of everyday wireless technologies, now widely used for mobile phones, Internet access, and even barcode replacement, has tremendously increased the ability to collect information from the physical world. Wireless technologies offer significant opportunities in sensors and controls for building operation, especially in reducing the cost of installing data acquisition and control devices. Installation costs typically represent 20 to 80% of the total cost of a sensor and control point in any HVAC system, so reducing or eliminating the cost of installation can have a dramatic effect on the overall installed system cost. Low-cost wireless sensors and control systems also make it economical to use more sensors, thereby establishing highly energy-efficient building operations and demand responsiveness that enhance the electric grid reliability.

Wireless sensors and control networks consist of sensor and control devices that are connected to a network using radio-frequency (RF) or optical (infrared) signals. Devices can communicate bidirectionally (i.e., transmitting and receiving) or one way (transmitting only). Most RF products transmit in the industrial, scientific, or medical frequency bands, which are set aside by the Federal Communication Commission (FCC) for use without an FCC license. Wireless sensor networks have different requirements than computer networks and, thus, different network topologies, and separate communication protocols have evolved for them. The simplest is the point-to-point topology, in which two nodes communicate directly with each other. The point-to-multipoint or star topology is an extension of the point-to-point configuration in which many nodes communicate with a central receiving or gateway node. In either topology, sensor nodes might have pure transmitters, which provide one-way communication only, or transceivers, which allow two-way communication and verification of the receipt of messages. Gateways provide a means to convert and pass data between protocols (e.g., from a wireless sensor network protocol to the wired Ethernet protocol).

The communication range of the point-to-point and star topologies is limited by the maximum communication range between the sensor node (from which the measured data originates) and the receiver node. This range can be extended by using repeaters, which receive transmissions from sensor nodes and then retransmit them, usually at higher power than the original transmissions. In the mesh network topology, each sensor node includes a transceiver that can communicate directly with any other node within its communication range. These networks connect many devices to many other devices, thus forming a mesh of nodes in which signals are transmitted between distant points via multiple hops. This approach decreases the distance over which each node must communicate and reduces each node’s power use substantially, making them more compatible with onboard power sources such as batteries (Capehardt 2005).

Determining network performance requirements means identifying and quantifying the communication functions required. Ehrlich and Pittel (1999) identified the following five basic communication tasks necessary to establish network requirements.

Data Exchange. What data passes between which devices? What control and optimization data passes between controllers? What update rates are required? What data does an operator need to reach? How much delay is acceptable in retrieving values? What update rates are required on “live” data displays? (Within one system, answers may vary according to data use.) Which set points and control parameters do operators need to adjust over the network?

Alarms and Events. Where do alarms originate? Where are they logged and displayed? How much delay is acceptable? Where are they acknowledged? What information must be delivered along with the alarm? (Depending on system design, alarm messages may be passed over the network along with the alarms.) Where are alarm summary reports required? How and where do operators need to adjust alarm limits, etc.?

Schedules. For HVAC equipment that runs on schedules, where can the schedules be read? Where can they be modified?

Trends. Where does trend data originate? Where is it stored? How much will be transmitted? Where is it displayed and processed? Which user interfaces can set and modify trend collection parameters?

Network Management. What network diagnostic and maintenance functions are required at which user interfaces? Data access and security functions may be handled as network management functions.

Bushby et al. (1999) refer to the same five communication tasks as interoperability areas and list many more specific considerations in each area. ASHRAE Guideline 13 also provides more detailed information that is helpful.

Table 2 lists some applicable standard protocols that have been used in BAS. Their different characteristics make some more suited to particular tasks than others. PROFIBUS (www.profibus.com) and MODBUS (www.modbus.org) were designed for low-cost industrial process control and automated manufacturing applications, but they have been applied to BAS. LonTalk defines a LAN technology but not messages that are to be exchanged for BAS applications. BACnet^® or implementers’ agreements, such as those made by members of LonMark International, are necessary to achieve interoperability with LonTalk devices. Konnex evolved from the European Installation Bus (EIB) and several other European protocols developed for residential applications, including multifamily housing. ZigBee^® is an open communications standard for wireless devices developed by the ZigBee Alliance. Annex O of the BACnet standard (ANSI/ASHRAE Standard 135-2013) specifies using BACnet messaging with services described in the ZigBee specification. Martocci (2008) describes how a wireless ZigBee network can be integrated into a BACnet network.

BACnet is the only standard protocol developed specifically for commercial BAS applications. BACnet has been adopted as a national standard in the United States, Korea, and Japan, as a European standard, and as a world standard (EN/ISO Standard 16484-5). BACnet was designed to be used with non-BACnet networks. Principles of mapping are documented in Annex H of ANSI/ASHRAE Standard 135-2012.

Rather than conforming to a published standard, a supplier can design a specific device to exchange data with another specific device. This typically requires cooperation between two manufacturers. In some cases, it can be simpler and more cost-effective than for both manufacturers to conform to an agreed-upon standard. The device can be either custom-designed or off the shelf. In either case, the communication tasks must be carefully specified to ensure that the gateway performs as needed.

Choosing a system that supports a variety of gateways may be a way to maintain a flexible position as products and standards continue to develop.

Popular methods of determining proportional, PI, and PID controller tuning parameters include closed- and open-loop process identification methods and trial-and-error methods. For each method, carefully consider the resulting timing of system responses to avoid compromising safety or reducing the expected life of equipment. Two of the most widely used techniques for tuning these controllers are ultimate oscillation and first-order-plus-dead-time. There are many optimization calculations for these two techniques. The Ziegler-Nichols, which is given here, is well established.

Ultimate Oscillation (Closed-Loop) Method. The closed-loop method increases controller gain in proportional-only mode until the equipment continuously cycles after a set-point change (Figure 21, where K_p = 40). Proportional and integral terms are then computed from the cycle’s period of oscillation and the K_p value that caused cycling. The ultimate oscillation method is as follows:

First-Order-plus-Dead-Time (Open-Loop) Method. The open-loop method introduces a step change in input into the opened control loop. A graphical technique is used to estimate the process transfer function parameters. Proportional and integral terms are calculated from the estimated process parameters using a series of equations.

The value of the process variable must be recorded over time, and the dead time and time constant must be determined from it. This can be accomplished graphically, as seen in Figure 22. The first-order-plus-dead-time method is as follows:

Trial and Error. This method involves adjusting the gain of the proportion-only controller until the desired response to a set point is observed. Conservative tuning dictates that this response should have a small initial overshoot and quickly damp to steady-state conditions. Set-point changes should be made in the range where controller saturation, or output limit, is avoided. The integral term is then increased until changes in set point produce the same dynamic response as the controller under proportional control, but with the response now centered about the set point (Figure 23).

In tuning digital controllers, additional parameters may need to be specified. The digital controller sampling interval is critical because it can introduce harmonic distortion if not selected properly. This sampling interval is usually set at the factory and may not be adjustable. A controller sampling interval of about one-tenth of the controlled-process time constant usually provides adequate control. Many digital control algorithms include an error dead band to eliminate unnecessary control actions when the process is near set point. Hysteresis compensation is possible with digital controllers, but it must be carefully applied because overcompensation can cause continuous cycling of the control loop.

Each component of a control system can be represented by a transfer function, which is an idealized mathematical representation of the relationship between the input and output variables of the component. The transfer function must be sufficiently detailed to cover both the dynamic and static characteristics of the device. The dynamics are represented in the time domain by a differential equation. In environmental control, the transfer function of many of the components can be adequately described by a first-order differential equation, implying that the dynamic behavior is dominated by a single capacitance factor. For a solution, the differential equation is converted to its Laplace or z-transform.

For more information on computer modeling programs, see Chapter 40 of the 2019 ASHRAE Handbook—HVAC Applications.