PID Control Basics: Proportional, Integral and Derivative Control Explained-Instrumentation Cable, Thermocouple, RTD Sensor & Pressure Transmitter Manufacturer

— A Practical Guide for Engineers, EPCs, and Project Teams

PID control is the most widely used control algorithm in industrial process control. From flow, pressure, temperature, and level loops in refineries to chemical reactors, power plants, water treatment, and food processing, the PID controller is the workhorse of automatic regulation. Despite its simplicity, PID control can be difficult to tune well, and a poorly tuned loop can cause instability, product quality problems, energy waste, and equipment wear.

This guide explains the fundamentals of PID control—what proportional, integral, and derivative actions do, how they work together, and how to apply them effectively in industrial applications.

1. What Is PID Control?

A PID controller continuously calculates an error value as the difference between a measured process variable (PV) and a desired setpoint (SP). It then applies a correction to the final control element (such as a control valve, VFD, or heater) based on three terms:

Proportional (P) — responds to the present error
Integral (I) — responds to the accumulation of past errors
Derivative (D) — anticipates future error based on the rate of change

The PID equation (ideal form):

$u (t) = K_{p} [e (t) + \frac{1}{T_{i}} \int_{0}^{t} e (t) d t + T_{d} \frac{d e (t)}{d t}]$

Where:

Symbol	Meaning
$u (t)$	Controller output
$e (t)$	Error = SP − PV
$K_{p}$	Proportional gain
$T_{i}$	Integral time (reset)
$T_{d}$	Derivative time (rate)

In practice, most industrial controllers use the standard or industrial form:

$u (t) = K_{p} e (t) + K_{i} \int_{0}^{t} e (t) d t + K_{d} \frac{d e (t)}{d t}$

Where $K_{i} = K_{p} / T_{i}$ and $K_{d} = K_{p} T_{d}$ .

2. Open-Loop vs Closed-Loop Control

Control Type	Description	Example
Open-loop	Controller output does not use feedback from the process	Manual valve position, timer-based dosing
Closed-loop	Controller output is adjusted based on feedback from the process	PID control of flow, pressure, temperature

PID control is a closed-loop strategy. It continuously compares the measured value with the setpoint and adjusts the output to reduce the error to zero.

3. Proportional Action (P)

What it does: The proportional term produces an output proportional to the current error. The larger the error, the larger the corrective action.

Equation:

$u_{P} = K_{p} e (t)$

Key characteristics:

Characteristic	Description
Speed	Fast response to changes in error
Offset	Cannot eliminate steady-state error alone
Stability	Too much gain causes oscillation and instability
Proportional band	The range of error over which the output changes from 0% to 100%

Proportional band (PB) is often used instead of gain:

$P B = \frac{100 %}{K_{p}}$

A small proportional band means high gain (aggressive control). A large proportional band means low gain (sluggish control).

Example: A temperature controller with a 20% proportional band will produce a full output change over a 20% change in error. If the setpoint is 200°C and the range is 0–400°C, the proportional band is 80°C.

Limitation: Proportional-only control always leaves a steady-state offset—the process variable settles at a value different from the setpoint.

4. Integral Action (I)

What it does: The integral term integrates the error over time. It continues to adjust the output as long as any error remains, eliminating steady-state offset.

Equation:

$u_{I} = K_{i} \int_{0}^{t} e (t) d t$

Key characteristics:

Characteristic	Description
Eliminates offset	Drives steady-state error to zero
Slow response	Integral action is slower than proportional action
Windup	Output can saturate during large or prolonged errors
Reset time	The time required for the integral term to repeat the proportional action

Integral time (reset) $T_{i}$ is the time it takes for the integral term to produce an output change equal to the proportional term for a constant error. A smaller $T_{i}$ means stronger integral action.

Example: If $T_{i} = 60$ seconds, the integral term will add an amount equal to the proportional output every 60 seconds of constant error.

Limitation: Integral action can cause overshoot and oscillation if too aggressive, and it can cause integral windup when the output saturates.

5. Derivative Action (D)

What it does: The derivative term responds to the rate of change of the error. It provides a corrective action that anticipates where the error is going, rather than where it is now.

Equation:

$u_{D} = K_{d} \frac{d e (t)}{d t}$

Key characteristics:

Characteristic	Description
Anticipatory	Responds to the rate of change, not the magnitude
Damping	Reduces overshoot and improves stability
Noise sensitivity	Differentiates noise, causing erratic output
Rarely used alone	Almost always combined with P and I

Derivative time (rate) $T_{d}$ is the time by which the derivative action anticipates the error. A larger $T_{d}$ means stronger derivative action.

Limitation: Derivative action amplifies high-frequency noise. In most process control applications, derivative action is either not used or used with a derivative filter.

6. How P, I, and D Work Together

Controller Type	Response	Offset	Stability	Typical Use
P only	Fast	Yes	Good at low gain	Level control, less critical loops
P + I	Fast + offset elimination	No	Moderate	Flow, pressure, most process loops
P + I + D	Fast + offset elimination + damping	No	Best for slow processes	Temperature, composition, slow loops

Practical guide:

Flow loops: PI control (fast, noisy; derivative not useful)
Pressure loops: PI control (fast, derivative often unnecessary)
Level loops: P or PI control (averaging level; avoid derivative)
Temperature loops: PID control (slow, derivative improves response)
Composition loops: PID control (very slow; derivative helps)

7. PID Tuning Methods

Tuning is the process of selecting $K_{p}$ , $T_{i}$ , and $T_{d}$ to achieve the desired response. Common methods:

Method	Description	Best For
Trial and error	Adjust parameters based on observed response	Experienced engineers; simple loops
Ziegler-Nichols (open-loop)	Use process reaction curve to calculate parameters	Processes that can be stepped
Ziegler-Nichols (closed-loop)	Increase gain until oscillation; use ultimate gain and period	Processes that can tolerate oscillation
Cohen-Coon	Uses process reaction curve with more complex formulas	Processes with dead time
IMC (Internal Model Control)	Uses process model to calculate parameters	Model-based tuning
Lambda tuning	Targets a specific closed-loop time constant	Self-regulating processes
AMIGO	Approximate M-constrained Integral Gain Optimization	Robust tuning

Ziegler-Nichols open-loop formulas (based on process reaction curve):

Controller	$K_{p}$	$T_{i}$	$T_{d}$
P	$1 / a$	—	—
PI	$0.9 / a$	$3 L$	—
PID	$1.2 / a$	$2 L$	$0.5 L$

Where $a = K \times L / T$ , $K$ is process gain, $L$ is dead time, and $T$ is time constant.

Important: These formulas provide a starting point. Final tuning should always be based on actual process response.

8. Digital PID Implementation

Most modern PID controllers are implemented in DCS, PLC, or single-loop controllers. Key considerations:

Consideration	Description
Sample time	Must be fast enough to capture process dynamics; typically 0.1–1 s for flow, 1–10 s for temperature
Integral windup	Output saturates during large errors; use anti-windup protection
Derivative filter	Low-pass filter to reduce noise amplification
Bumpless transfer	Smooth transition between manual and automatic modes
Output limits	Prevent valve over-travel or actuator saturation
Setpoint filtering	Prevents sudden setpoint changes from causing output spikes
Feedforward	Adds a corrective signal based on a measured disturbance

9. Common PID Problems and Solutions

Problem	Likely Cause	Solution
Oscillation	Too much gain; too little integral time; too much derivative	Reduce $K_{p}$ ; increase $T_{i}$ ; reduce $T_{d}$
Slow response	Too little gain; too much integral time	Increase $K_{p}$ ; decrease $T_{i}$
Steady-state offset	No integral action or too little	Add integral action; decrease $T_{i}$
Overshoot	Too much integral action; too little derivative	Increase $T_{i}$ ; add derivative
Noisy output	Derivative action amplifying noise	Add derivative filter; reduce $T_{d}$
Integral windup	Output saturates during large errors	Enable anti-windup; limit integral term
Valve cycling	Overly aggressive tuning; valve deadband	Reduce gain; check valve deadband; add deadband compensation
Loop instability	Excessive gain; process nonlinearity	Reduce gain; linearise process; re-tune

10. Practical Tuning Tips

Tip	Why
Start with P-only	Understand the process response before adding I and D
Tune with the process at normal operating conditions	PID performance depends on process gain and dynamics
Change one parameter at a time	Isolate the effect of each change
Use the fastest sample time practical	Improves control response
Add derivative only for slow processes	Temperature, composition; not flow or pressure
Use anti-windup for all integral controllers	Prevents saturation problems
Document the final tuning	Essential for maintenance and troubleshooting
Re-tune after process changes	Process gain and dynamics change with operating conditions
Use simulation where possible	Safe way to test tuning before applying to the process

11. Common Mistakes to Avoid

Mistake	Consequence	Prevention
Using derivative on flow loops	Noisy output; valve cycling	Use PI only for flow
Using derivative on level loops	Erratic level control	Use P or PI for level
Ignoring integral windup	Overshoot; instability after saturation	Enable anti-windup
Tuning with process offline	Parameters do not match operating conditions	Tune at normal operating conditions
Using default tuning parameters	Poor performance; instability	Always tune for the specific loop
Not documenting tuning	Maintenance and troubleshooting difficulties	Record final parameters
Ignoring valve deadband	Limit cycles; poor control	Check valve deadband; compensate if necessary
Using too fast a sample time	Noise amplification; unnecessary CPU load	Match sample time to process dynamics
Using too slow a sample time	Poor control; instability	Match sample time to process dynamics
Not considering process nonlinearity	Good control at one operating point, poor at another	Consider gain scheduling or adaptive control

12. Applicable Standards and References

Standard / Reference	Scope
ISA-5.1	Instrumentation symbols and identification
ISA-77.44.01	Fossil fuel power plant—steam temperature control
IEC 61131-3	PLC programming languages—PID function blocks
ISA-101	HMI design
ISA-18.2	Alarm management
Åström & Hägglund, PID Controllers	Classic reference on PID control
Shinskey, Process Control Systems	Practical process control reference

13. Why Choose Anhui Tiankang for Process Control Solutions?

Anhui Tiankang (Group) Co., Ltd. has nearly five decades of experience in industrial instrumentation and control systems. We provide instruments and engineering support for PID control loops in flow, pressure, temperature, level, and composition applications.

Products for PID control:

Product	Control Application
Pressure transmitters	Pressure control loops
Temperature transmitters	Temperature control loops
Level instruments	Level control loops
Flow meters	Flow control loops
Control valves	Final control element
Positioners	Valve positioning for PID control
Signal isolators	Signal conditioning for control loops
Instrumentation cables	Reliable signal transmission

Engineering support:

Control loop design review
PID tuning assistance
Loop troubleshooting
Control valve sizing and selection
Instrumentation selection for control loops

Core advantages:

Complete certifications: CCC Ex, ATEX, IECEx, SIL
CNAS-accredited laboratory: full performance testing
Long-term supplier to CNPC, Sinopec, CNOOC, and international EPC projects
One-stop supply: from instruments to cables to Ex accessories

14. Conclusion

PID control is the foundation of industrial process control. Understanding proportional, integral, and derivative actions—and how they work together—is essential for engineers, EPCs, and plant operators.

Key takeaways:

Term	Action	Best For
Proportional (P)	Responds to present error	Fast response, but leaves offset
Integral (I)	Responds to accumulated error	Eliminates offset, but can cause overshoot
Derivative (D)	Responds to rate of change	Damping, but sensitive to noise

Controller selection guide:

Loop Type	Recommended Controller
Flow	PI
Pressure	PI
Level	P or PI
Temperature	PID
Composition	PID

The most important rule: Tune the controller for the specific process, at normal operating conditions, and document the final parameters. A well-tuned PID loop delivers stable operation, product quality, energy efficiency, and long equipment life.

Remember: PID control is simple in concept but challenging in practice. The difference between a poorly tuned loop and a well-tuned loop is measured in product quality, energy consumption, and plant profitability.

Contact Us

For PID control support, instrument selection, or project quotations, please contact:

Yin Shuangjie
International Sales Manager
📧 Email: [email protected]
📱 WhatsApp / Zalo: +86 17856068126
🌐 Website: http://www.tiankang-global.com/

Anhui Tiankang – Your partner for reliable process control and instrumentation solutions.

NAV

Search

Close