Building Control Systems

The Original Systems

To understand where we are today, it is worth reviewing where we have been.

As buildings grew in sophistication, a need developed to control the environment. Early buildings were small, were not cooled in summer, and were heated in winter with stoves or fireplaces. Early control systems were not automatic and were designed to limit the rate at which coal or wood burned in the stove. As the Industrial Revolution progressed, buildings grew larger, and individual fireplaces and stoves became unworkable. To overcome these shortcomings, designers put in steam heat as a means to heat spaces, and individual windows for daylight, cool air, and ventilation. Now a fire could be stoked in one location, and the heat from the fire could be distributed. The distribution of heat created the need for the first control system.

A first system involved the use of a bell. When an occupant in a room was too cold, he or she would give the bell a loud ring. The fire stoker would increase the amount of heat flowing to the room by manually adjusting a valve. In 1885, Professor Warren Johnson of Milwaukee, Wisconsin, discovered that certain metals bend in a peculiar way in the presence of heat. Professor Johnson saw that these metals could provide for automatic temperature sensing. When he combined a metal temperature sensor with an electrical switch, he created the first automatic temperature control: the thermostat. When he connected this new sensing device to a controlled device, an electric bell, he developed the first automatic control system. Thus, when a room was cold, the thermostat could automatically ring the bell.

As building size and sophistication increased, new systems were included in designs. Early buildings allowed each person to be near a window. However, the widespread use of electricity in buildings soon brought numerous changes. Larger buildings were created with interior spaces without windows, and thus building designers needed to provide fans for ventilation. Electric motors were coupled to valves that had previously been operated manually. Thus, the electric thermostat could operate the valve directly and automatically, eliminating both the bell and the fire stoker. Control systems were no longer a luxury but were becoming a real need in buildings.

In these early systems, the controls operated controlled devices in an “on-and-off” manner (two-position control). When a room was cold, the heat was on to the maximum amount. When the desired temperature was reached, the heat was entirely off. This resulted in widely fluctuating temperatures that would be unacceptable by current standards. However, it was a significant advance in the early years of the 20th century.

These widely fluctuating temperature conditions, particularly in mild weather, resulted in a need for better control. There was a need for systems that could put a small amount of heat in a room when there was a small need for heat. To respond to this need, pneumatic controls were introduced, and the concept of proportional control (which gradually opens or closes the controlled device in response to the gradual change in the sensed variable) was developed for heating and cooling controls.

Pneumatic controls were a response to the need for proportional control, as well as the need for lower costs. Although electric actuation was possible, its cost was prohibitive. Air-operated pneumatic controls could be made inexpensively and could offer proportional control. Proportional control was truly a breakthrough in control systems. New types of controlled devices were needed to accomplish proportional control. It was found that certain metals could withstand repeated flexing without breakage. These metals were formed into designs similar to fireplace air bellows and were known as sylphons. Sylphons could utilize a low pressure air supply, typically 15 lbs. per square inch gauge (psig.), and incrementally position a controlled medium (e.g., the flow of water in a pipe or the flow of air through a duct). With the advent of sylphons, the need for pneumatic thermostats became evident. Pneumatic thermostats were developed around 1910. These devices also utilized metals that bent when heated, but instead of snapping an electrical switch, they allowed compressed air to gradually escape. The escape of compressed air was linearized to correspond to sensed changes in temperature. The result was an air pressure output from the thermostat that changed gradually in proportion to the sensed variable. These devices were the beginning of pneumatic proportional controls, which are still in use today.

The Pneumatic Controls Period

The relatively low cost of pneumatic controls allowed a much wider use of controls in buildings. Temperature control became a standard in buildings, and controls for other environmental needs were developed. For example, it was becoming apparent in northern climates that the introduction of humidity into a space would increase comfort, as discussed in the BOK’s Building Mechanical Systems chapter. Sylphons were available that could gradually control the introduction of steam or water into the air. However, there was a need for a sensor and controller that could regulate moisture in the air. It was found that there were materials that could change their physical dimensions when subjected to moisture in the air. These materials—human hair, animal bone, animal horn, and wood—were the foundation materials in the development of the hygrostat, the product currently known as the humidistat.

An additional major development in pneumatic controls was the development of capillary sensing. As buildings grew in size, space heating, humidification, and ventilation were accomplished with centralized systems. Room-type thermostats were inadequate to sense air conditions in ducts and certainly could not be mounted in pipes to sense water temperatures. The capillary sensor was a means of using small-diameter metal tubing, attaching a small-bellows sylphon to one end, and filling the tubing with a liquid or gas that would expand or contract in relation to the sensed temperature. This change in volume as a result of expansion caused the bellows sylphon to move. The motion could then be used to control the leakage rate of air from a controller to achieve proportional pneumatic signals. Similarly, controllers with external humidity sensors were developed to sense duct humidity. Remote sensor controllers were developed for sensing the temperature of water in pipes. Additional sensor controllers were developed for pipe and duct pressure sensing. There was now a complete array of tools to sense and control building environmental conditions.

Armed with this wide array of control tools, building systems grew in complexity. Although early buildings may have had one thermostat for the entire building, modern buildings of the pneumatic period had individual thermostats for each room and thermostats and humidistats for central air and water systems. Major companies such as Johnson Service Company, Minneapolis Honeywell Regulator Company, Powers Regulator Company, and Fulton Sylphon Company grew and became major suppliers to the building industry.

The quality of controls began to increase with the development of a controls industry. Improved accuracy of control was needed. Early sensor controllers, typically known as regulators, were first-order response devices. First-order devices do not have an internal mechanism to produce controller feedback. This type of controller creates a difference between the desired result (the set point) and the actual result; this difference is known as offset. Also, the speed with which controllers could respond typically was limited to achieve stability. When these devices were applied to systems that frequently experienced wide variations, response was slow and unstable. The market responded with the creation of second-order mechanisms.

Second-order feedback-type controllers were developed in the early 1920s. They went on to become the industry standard for pneumatic controllers into the 1990s. Second-order controllers use internal feedback of the output signal to adjust the output signal. If there is a great need for output, the output is allowed to go to maximum. As the sensed variable approaches the set point, the feedback loop slows the output. Typically, the rapid output changes cause the sensed variable to go beyond the controller set point; this is known as overshoot. The controller then responds by rapidly diminishing the output to achieve set point. Second-order controllers gave users more accuracy by diminishing offset and provided a quicker response, which improved stability. In modern terminology, increased speed of response is known as derivative control (rate based) and the elimination of offset is known as integral control (time based). Thus, modern controllers can incorporate proportional-integral-derivative (PID) control. This is explained further later in the section.

In the early 1960s, building complexity continued to increase. Building size continued to grow, both vertically (the skyscraper) and horizontally (the sprawling campus environment). Centralized cooling of spaces became commonplace. Windows that had previously been operable became sealed. Also, systems typically ran 24 hours a day. In the early 1960s, building owners began to see value in turning systems off at opportune times. There was a need to centralize control of systems to allow oversight of operation, as well as to control system operation. The control systems that had served so well with capillary sensing could not meet the need for centralized information. In response, the control industry developed the concept known as pneumatic transmission.

Pneumatic transmission made the sensor a separate component of the sensor-to-controller-to-controlled device sequence. Pneumatic sensors still sensed the variable; however, instead of being connected directly to the controller, the two components were connected with a pneumatic air pipe. The air pipe could be as long as desired. For example, the sensor or pneumatic transmitter sensed the variable over a fixed range, such as 0°F to 100°F, and created a 3-psig signal at 0°F and a 15-psig signal at 100°F. The signal was linear between the range extremes. For each degree change in the example, the air pressure in the pipe would change by 0.12 psig. The control industry standardized pneumatic transmission and the 3- to 15-psig. pressure change for a sensed range. Now the same sensed variable could be used to control a process at a fan system on the roof and simultaneously to indicate the process condition on a master panel board in the basement. Previously, capillary type sensors were unable to indicate the sensed variable. A separate thermometer or hygrometer was needed adjacent to the duct-mounted controller to observe the variable. With transmission, the same sensor could be connected to a separate receiver controller as well as to a specially calibrated gauge that read in units of the sensed variable rather than in pressure. The receiving controllers as well as the gauges could be centralized in mechanical equipment rooms in control panels. Pneumatic transmission was widely accepted in the building industry and currently has been the most common type of control system installed in campus buildings.

Pneumatic Control Panel

The introduction of transmission and the flexibility of this concept introduced much greater complexity to buildings. A wide range of new control logic applications appeared as the sensed variable was made available for manipulation. Comparisons could be made of the lowest temperature, the highest pressure, and the humidity in various locations. Different receiver controllers, with different set points for daytime or nighttime, winter or summer, could be switched in and out of the control logic. The inclusion of time control clocks and interconnection to a wider range of electric devices, such as switches and lights, made the systems electro-pneumatic rather than pure pneumatic, as in earlier years.

As controller complexity increased, the ability to understand building system controls decreased. In years past, a typical building had six different types of devices that were used repeatedly to make up a control system. Now there were hundreds of choices of products. To address the change in technology, the building engineer who typically oversaw controls was replaced by control system specialists and technicians. As systems became more and more sophisticated, the shortcomings of pneumatic transmission became clear. Pneumatic controllers lacked long-term reliability, requiring frequent inspection and recalibration to achieve the needed reliability. Control rooms and centralized monitoring became important to oversee the proper operation of the controls.

A new need developed as control complexity grew and the need to centralize oversight became commonplace. The cost to bring hundreds of wires and air lines to one central location grew significantly. The industry began to offer a lower-cost system that could accomplish centralization without the numerous individual components. This product was the building monitoring system. Building monitoring systems started as multiplexed wiring systems. Multiplexing allowed numerous signals to be sent over a cable made up of many small wires. The central multiplexer switcher would select a pair of wires to read a value and sense another pair of wires to connect the sensor. The multiplexer was then connected to one common gauge at the central control panel. In a similar manner, an operator could select a motor or function at a remote mechanical location and cause it to start or stop, all over the same sets of wires. It became apparent that initial multiplexers were dependent on the console operator as they became more and more complex. Again, the industry responded in the late 1960s with the marriage of the general purpose digital computer and the multiplexing systems.

The combination of the computer with the building monitoring function gave building owners new flexibility. Until that point, all knowledge was contained in the personal memory of the building maintenance staff. Now memory in the computer could maintain this knowledge. The computer could also reproduce the manual actions of operators to interrogate systems and to start and stop system functions. Systems could be interrogated, conditions compared with stored ideals, and exception reports prepared. Computerized systems could free building maintenance staff to perform other duties. The early systems were quite crude by current standards; however, they offered necessary features, and further advancements created modern building monitoring and control systems.

Building monitoring and control systems that sprang from the early combinations of controls were a part of the continuing evolution of building control systems. Significant changes to systems occurred in the early 1970s. Computing power was increasing dramatically while computing costs were decreasing. The integrated circuit helped lower the cost of developing sophisticated circuitry, and improved systems software became more user friendly. Installation techniques for systems became less costly. Building monitoring and control systems were being recognized as a necessary part of managing large facilities.