BTL Mark: Resolve interoperability issues & increase buyer confidence
Back to Basics – Part Two
A crash course in the concepts of HVAC control
Last month I kicked off a three-part summer-long series, getting back
to the basics of control systems theory and design, conceptual and
practical. Part One covered the fundamentals. Part Two, presented here,
picks up where we left off. The three main topics of discussion in this
installment are as follows: methods of control, controls components,
and control system architecture.
Methods of Control
In HVAC, we typically run across three types of control methods. Two-position (two-state) control, staged control, and proportional control are all methods of controlling mechanical equipment in HVAC systems, each with its benefits when applied properly, and with its potential drawbacks when misapplied.
To illustrate, in simple concepts, the difference between the three above-mentioned methods of control, consider an example of a light bulb. The light bulb is the controlled device, and you are the controller. Being the controller, you have the ability to establish a preference, or setpoint. You can prefer to have the light on, to have it off, or even perhaps have it glow at some intermediate level. Also as the controller, you have the power to control the bulb to your preference. The closer the status of the bulb is to your preference, and better the overall control.
With two-position control of the light bulb, only two positions, or “states”, are achievable. The bulb can be on or off, with no in-between. If your preference, as the controller, is only for either one of these two states, then two-position control is adequate. However, if your preference is more demanding, and you require intermediate levels of light from the bulb, then two-position control doesn’t cut it.
Staged control breaks up the control of the “end device” into parts, from zero percent to one hundred percent. In the case of our light bulb, zero percent is the light being off, and one hundred percent is the light being on. If we break up control of the light bulb into, let’s say, three stages, then we can basically have the bulb assume any one of three levels of light. With no stages called for by the controller, the light is off. If the controller calls for one stage, then the light illuminates at approximately 33 percent. A call for two stages will result in the light burning at 66 percent, and a call for three stages will result in the light burning at full power. As the controller, you have three levels of light at your disposal, to satisfy your preference.
What if your preference is in the middle of two stages? In other words, what if your preference is, say, 50 percent? If you had more stages of control at your disposal, you could at least get closer to your preference. You might conclude from this that the more stages of control at its disposal, the better chance the controller has at achieving and maintaining its setpoint. If we could “extrapolate” the number of stages to infinity, we would no longer be limited to “discreet” stages, or levels, of light. We would be able to assume any level we want, as driven by our preference, in the whole range of zero to one hundred percent. This type of control, of which the controller can command the end device to be in any position throughout its whole range of operation, is the fundamental concept of proportional control.
In last month’s column we touched upon the concept of sensors and controllers, so that’s where we’ll start with here, as far as control system components go. As discussed, a controller has three main functions, to process information gathered by a sensor, to establish a setpoint, and to act upon an end device in a manner beneficial to the controlled process. A controller can be nothing more than a device with a single input and output. Sometimes the sensor is an integral part of the controller, as is the case of the typical room thermostat. Controllers run the gamut, from simple two-state operation, to fully modulating or proportional control.
Two-state “single purpose” controllers are typically electrical in nature. An on-board sensing apparatus and setpoint dial handle the first two functions of the controller. The output is nothing more than an automatic switch, opening or closing based on sensor info and setpoint, to affect the operation of a two-state end device. Examples of two-state controllers include thermostats, humidistats, flow and pressure switches, current sensing switches, and occupancy sensing switches.
Staging controllers or sequencers may be electrical or electronic, the latter becoming more the norm these days, especially when you consider that many staging or sequencing applications are handled via digital control. Such a controller will have multiple outputs, one for each stage of control required, and will be set up such that, as the controlled variable strays further from setpoint, more stages of control are engaged, in an attempt to bring the controlled variable back to setpoint. There is more to it than that, and the intricacies extend beyond the scope of this lesson, but hopefully you get the concept. Examples of staging controllers would be those controllers that would be applied to a multistage process, for instance an electric duct heater or a hot water boiler plant with multiple boilers. In the case of the duct heater, the sensor may be in the space served, and the controller may be local to the heater. The further the space temperature is from setpoint, the more stages of electric heat are on. For the boiler plant, the sensor may be installed in the hot water supply pipe, wired back to the controller, with each stage of the controller operating an individual boiler. In a four-boiler system, the controller will stage through boilers 1-4 in order to achieve and maintain the hot water supply temperature setpoint.
Proportional controllers will offer, as an output, a modulating control signal. This signal can control an end device that can be “positioned” anywhere throughout its entire range of operation. For instance, a proportional temperature controller set up to control a hot water coil’s modulating control valve in an effort to achieve and maintain a suitable duct temperature downstream of the coil.
Digital controllers can run the gamut, from a “one in and one out” controller to an I/O device able to accept many inputs and drive a multitude of end devices. A digital controller with both analog and binary inputs and outputs can feasibly be programmed to perform an array of functions, from simple two-state control and staging control, to proportional control and beyond.
End devices are those devices that are controlled by…controllers! A relay being energized in order to turn on a fan in an air handling system would be categorized as a simple two-state end device. A controller, configured accordingly and set up for two-state control, will command the relay to energize based upon some condition, perhaps a scheduling function.
Other end devices, such as control valves and motorized control dampers, can be either two-state or modulating, depending on the requirements of the application.
Safeties and limits, as described herein, are essentially two-state controllers, employed to monitor some condition and “trip” when the condition has become unacceptable in terms of safe and proper operation. Upon trip of the device, the system that it’s monitoring will likely be shut down, only to be allowed to operate once again after the press of a “manual reset” button on board the controller itself. Examples of safety devices are as follows: low limit temperature controllers (freezestats), high and low limit static pressure switches, high limit humidity controllers, high limit thermal cutout devices, and duct-mounted smoke detectors.
Control System Architecture
We consider here a networked Direct Digital Control (DDC) system, or Building Automation System (BAS), whereby all controllers operating all mechanical equipment are digital in nature, and are networked together, back to a network level controller that coordinates the operation of all of the various interdependent subsystems (that was a mouthful!).
Going back to last month’s column, referencing the discussions on equipment and controller levels, we start at the zone level. Here we have unit level controllers operating unitary or zone level equipment such as fan-coils and VAV boxes. Next we’ll have equipment level controllers operating air handlers and the like. Moving to the plant level, here we find equipment such as boilers and chillers, along with all of the required pumps, valves, and other miscellaneous equipment. At this level, we have a plant level controller, one with enough inputs and outputs to properly and effectively operate the “plant”, as it were.
All of these controllers are connected together with a pair of wires, a communication network essentially, that allows these digital controllers to communicate with the “main brain” of the system, the network level controller. It is duly noted here that all of the various controllers are perfectly capable of operating in stand-alone fashion, with no support from a higher entity. However, the network controller’s role is important to the BAS, for many reasons (distribution of global data such as outside air temperature, global time-of-day scheduling, trending, and alarming, to name a few.). Another important feature of the network controller is the ability to connect to it via a “front-end”, which in the old days was a personal computer more or less dedicated to the BAS, with graphics screens of all of the connected equipment. Nowadays, the front-end can be any computer connected to the internet, as long as the network controller is web-capable and connected as well. Diminishes the duty of checking the status of your BAS to nothing more than “surfing the web”!
Tip of the Month: Did you know??? PID control is proportional control with a couple more “dimensions” thrown in. The “I” stands for “integral” (a term from your calculus course…ouch!). With P+I control (without the “D”), performance is improved over simple proportional control, by measuring and minimizing the offset from setpoint over time. A well-tuned P+I control loop will operate in a narrow band close to setpoint. The “D” stands for “derivative” (another calculus term). What PID adds is a “predictive” element to the control response. Whereas the “I” in PID asks “How far am I from setpoint?”, the “D” asks “How quickly am I approaching setpoint?”. For What It’s Worth.
[Click Banner To Learn More]
[Home Page] [The Automator] [About] [Subscribe ] [Contact Us]