Control Loops and Process Control Systems

Control Loops and Process Control Systems

Introduction to Control Loops

The control loop is the fundamental element of the instrumentation trade. It constitutes a dynamic system that measures a process variable, compares it to a setpoint, and then acts on a final control element to maintain the measured variable at the desired value. For the Red Seal exam, you must master the four essential elements of any loop: the measuring element (transmitter), the controller, the final control element (valve or variable speed drive), and the process itself.

Standardized terminology is crucial. The measured variable (PV) is the actual value of the controlled quantity. The setpoint (SP) is the desired value. The error (e) is the difference between the setpoint and the measured variable: e = SP − PV. The controller output (CO) is the signal sent to the final element. These terms are universal and appear in exam questions in various formulations.

Loops are classified into two main categories: open loops and closed loops. An open loop does not use the measurement to correct the output — it applies a predetermined command. A closed loop (or feedback loop) continuously compares the measurement to the setpoint and adjusts the output accordingly. The vast majority of industrial processes use closed loops.

Signal Types and Transmission

Pneumatic Signals

The standard pneumatic signal is 3 to 15 psi (20 to 100 kPa). The live zero of 3 psi allows for detection of a line break: if the pressure drops to 0 psi, it is a failure. The span is 12 psi. To convert a process value into a pneumatic signal, use direct proportionality.

Typical calculation: For a level transmitter 0 to 200 cm, what pneumatic signal corresponds to 150 cm?

The relationship is: Signal (psi) = 3 + (Measured value / Span) × 12

Signal = 3 + (150 / 200) × 12 = 3 + 9 = 12 psi

Analog Electronic Signals

The electronic standard is 4 to 20 mA. The live zero of 4 mA allows for detection of a wire break (0 mA = fault). The current loop is preferred over the voltage loop because it is insensitive to wire resistance and voltage drops over long distances. The typical maximum load is 600 Ω for most transmitters.

Typical calculation: For a pressure transmitter 0 to 100 kPa, what current corresponds to 65 kPa?

Current (mA) = 4 + (Measured value / Span) × 16

Current = 4 + (65 / 100) × 16 = 4 + 10.4 = 14.4 mA

Digital Signals and Protocols

Digital protocols superimpose information onto the analog signal or use a fully digital medium. HART (Highway Addressable Remote Transducer) superimposes a digital FSK (Frequency Shift Keying) signal onto the 4-20 mA loop, allowing bidirectional communication without interrupting the analog signal. Foundation Fieldbus and PROFIBUS PA are fully digital protocols that allow multipoint communication on a single wire pair.

Control Action: Direct and Reverse

Determining controller action is an essential skill. A direct-acting controller increases its output when the measured variable increases. A reverse-acting controller decreases its output when the measured variable increases.

The rule of thumb: choose the action that makes the loop negative feedback. To determine the correct action, follow the complete path around the loop:

Example: A heating process where the steam valve is air-to-open. If the temperature rises above the setpoint, the valve must close to reduce heat. The controller must therefore decrease its output when the temperature increases → reverse action.

Reverse example: A cooling process where the chilled water valve is air-to-open. If the temperature rises, the valve must open further to cool. The controller must increase its output when the temperature increases → direct action.

Control Modes

On-Off Control

On-off control is the simplest: the output is either 100% or 0%. It creates a natural oscillation around the setpoint. The differential gap (or hysteresis) is the difference between the cut-in point and the cut-out point. A wider differential gap reduces the frequency of switching but increases the oscillation of the variable. This mode is used for simple applications such as level tanks or domestic furnaces.

Proportional Control (P)

Proportional control produces an output proportional to the error: CO = Kp × e + bias. The proportional gain (Kp) is the ratio between the change in output and the change in error. The proportional band (PB) is the inverse of the gain, expressed as a percentage of the span: PB = (100 / Kp).

Proportional control always leaves a residual error (offset) at steady state. This error is unavoidable because a non-zero output requires a non-zero error. The offset decreases as the gain increases, but a gain that is too high causes instability.

Typical calculation: A proportional controller has a gain of 2. The setpoint is 50% and the measured variable is 45%. The bias is 50%. What is the output?

CO = 2 × (50 − 45) + 50 = 2 × 5 + 50 = 60%

Proportional-Integral Control (PI)

Integral action eliminates the offset by integrating the error over time. The output is: CO = Kp × e + (Kp / Ti) × ∫e dt. The integral time (Ti) , expressed in minutes per repeat, is the time required for the integral action to reproduce the proportional action. A smaller value of Ti gives a more aggressive integral action.

Integral windup occurs when the integral action continues to integrate the error while the output is saturated at 100% or 0%. When the error reverses, the output remains saturated longer, causing excessive overshoot. Modern controllers include anti-windup mechanisms.

Proportional-Integral-Derivative Control (PID)

Derivative action anticipates the error by reacting to its rate of change: CO = Kp × e + (Kp / Ti) × ∫e dt + Kp × Td × (de/dt). The derivative time (Td) , expressed in minutes, determines the amount of anticipatory action.

Derivative action must never be used on a noisy measured variable (such as level with waves or turbulent flow) because it amplifies the noise. It is particularly useful for temperature loops where the dynamics are slow.

Controller Tuning

Tuning consists of determining the optimal values of Kp, Ti, and Td. The main methods are:

Ziegler-Nichols Open-Loop Method (Reaction Curve Method)

Ziegler-Nichols Closed-Loop Method (Ultimate Gain Method)

Exam trap: The Ziegler-Nichols formulas give a starting point, not a final tuning. They often produce a response with 25 to 40% overshoot. For a more damped response, use lower gains.

Advanced Loop Configurations

Cascade Control

Cascade control uses two nested loops: the master loop (outer) generates the setpoint for the slave loop (inner). The slave loop must be faster than the master loop (typically 3 to 5 times faster). This configuration rejects disturbances in the inner loop before they affect the primary variable.

Example: Temperature control of a heat exchanger where the master loop measures the outlet temperature and adjusts the steam flow setpoint, and the slave loop measures and controls the steam flow.

Ratio Control

Ratio control maintains a constant ratio between two flows. The wild flow is uncontrolled; the controlled flow is adjusted to maintain the ratio. The ratio is calculated as: R = Controlled flow / Wild flow.

Typical calculation: A process requires a 3:1 ratio between flow A (controlled) and flow B (wild). Flow B is 40 m³/h. What is the setpoint for flow A?

Setpoint A = 3 × 40 = 120 m³/h

Three-Element Level Control

Three-element control (or feedforward-compensated level control) uses three measurements: level, inlet flow, and outlet flow. The outlet flow is controlled based on the inlet flow (feedforward action) with correction by the level (feedback action). This configuration is standard in boilers and separators where flow variations are rapid.

Feedforward Control

Feedforward control measures the disturbance and acts before it affects the controlled variable. It is always combined with a feedback loop to correct residual errors. The process model must be accurate; otherwise, feedforward control can degrade performance.

Override Control and Selection

Override control uses multiple controllers whose outputs pass through a selector (high or low) to choose the active output. A high selector chooses the largest output; a low selector chooses the smallest. This configuration protects the process against dangerous conditions.

Example: A low selector between a flow controller and a pressure controller protects a pipeline against overpressure: if the pressure exceeds the setpoint, the pressure controller takes control and reduces the flow.

Safety Instrumentation and Loop Analysis

Safety Instrumented Functions (SIF)

Safety instrumented functions are loops designed to bring the process to a safe state in the event of a dangerous condition. They are distinct from normal control loops (BPCS — Basic Process Control System). The safety integrity level (SIL) is a reliability indicator: SIL 1 (risk reduction ×10 to ×100), SIL 2 (×100 to ×1000), SIL 3 (×1000 to ×10000).

The separation requirements between BPCS and SIS are defined in the standard CSA Z767 (safety instrumented systems for the process industry). Safety valves must be tested periodically to verify their operation.

Failure Analysis

A dangerous failure is a failure that prevents the safety function from executing. A safe failure is a failure that causes a transition to the safe state (for example, an air-to-close valve that closes on loss of air). The dangerous failure rate (λdu) is used to calculate the probability of failure on demand (PFD).

Loop Documentation

Standard documentation includes:

Process Analysis and Dynamics

Process Characteristics

Industrial processes have dynamic characteristics that influence the choice of controller:

Loop Stability

A loop is stable if, after a disturbance, the variable returns to a finite value. It is unstable if the oscillations increase in amplitude. The Bode stability criterion states that a loop is unstable if the gain is greater than 1 (0 dB) when the phase shift reaches −180°. The gain margin is the factor by which the gain must be increased to reach instability. The phase margin is the additional phase shift required to reach −180°. A phase margin of 30° to 60° and a gain margin of 2 to 5 are generally recommended.

Control Valves and Final Elements

Flow Characteristics

The flow characteristic of a valve describes the relationship between opening and flow:

Flow Coefficient (Cv)

The Cv is the flow of water in US gallons per minute that passes through the valve with a pressure drop of 1 psi. For liquids: Cv = Q × √(SG / ΔP), where Q is the flow in GPM, SG is the specific gravity, and ΔP is the pressure drop in psi.

Typical calculation: What Cv is required for a flow of 150 GPM of water (SG = 1.0) with a pressure drop of 25 psi?

Cv = 150 × √(1.0 / 25) = 150 × √0.04 = 150 × 0.2 = 30

Valve Positioners

The positioner compares the actual position of the valve stem to the position demanded by the controller signal and adjusts the air pressure to correct the deviation. It improves accuracy, increases response speed, and overcomes friction forces and pressure imbalances. Positioners are essential for equal percentage valves and for applications where accuracy is critical.

Communication and System Integration

Distributed Control Systems (DCS)

The distributed control system (DCS) is a distributed architecture where controllers are geographically distributed near the process and communicate via a redundant network. The DCS integrates control, sequencing, alarm, and historical functions.

Programmable Logic Controllers (PLC)

Programmable logic controllers (PLC) are industrial programmable controllers used primarily for sequential logic and safety functions. They are increasingly used for continuous control thanks to built-in PID function blocks.

Industrial Networks

Industrial networks follow the OSI model and use specific protocols. Modbus TCP/IP is the most widespread protocol for communication between DCS, PLCs, and instruments. OPC UA (Open Platform Communications Unified Architecture) is a vendor-independent interoperability standard.

Applicable Standards and Codes

The Canadian Electrical Code, Part I (C22.1) applies to electrical installations in classified areas. Rule 18-002 defines classified areas according to the nature of hazardous materials. Rule 18-102 specifies the requirements for wiring in classified areas. Instruments installed in these areas must be certified for the appropriate class, division, and group.

The standard CSA B149.1 (Natural gas and propane installation code) applies to gas installations, including burner systems and safety valves. The standard CSA Z276 applies to liquefied natural gas installations. The standard CSA B51 covers boilers, pressure vessels, and pressure piping.

For safety instrumented systems, the standard CSA Z767 is the Canadian reference. It harmonizes with IEC 61511 and specifies the requirements for the complete safety lifecycle.

Pitfalls to Avoid

Summary

To pass the exam, practice signal conversion calculations, Cv calculations, PID tuning, and flow ratio calculations until they become automatic. Review P&IDs and loop diagrams to quickly recognize standard configurations. Finally, memorize the key numerical values: 3-15 psi, 4-20 mA, 63.2% for the time constant, and the Ziegler-Nichols factors.