FireCADCalc

Feedforward Controller: Gain, Lead–Lag and Dead Time

Static feedforward gain, lead and lag time constants and dead time from first-order-plus-dead-time models of the process and the disturbance path.

Thermal power plantProcess steam and utilitiesSugar mill and cogenerationDistillery and ethanolRefinery and petrochemicalChemical and fertiliser

Inputs

Examples: Process 120 s, disturbance 200 sDisturbance faster than process

Results

G_ff = −(K_d/K_p)(τ_ps + 1)/(τ_ds + 1)e^(−(θ_d − θ_p)s). If the disturbance acts faster than the manipulated variable (θ_d < θ_p), perfect compensation is impossible; use the static gain and tune the lead–lag by test. Keep feedback trim in service.

Feedforward gain-0.75000
Lead time120.00s
Lag time200.00s
Feedforward dead time (0 if not realisable)10.000s

Download PDF CSV Link to this result Compare another case

Log in to save calculations to your dashboard.

Worked example

Example: Process 120 s, disturbance 200 s. The inputs, in the units selected above:

InputValue
Process gain (output per % controller output)2.0000 -
Process time constant120.00 s
Process dead time20.000 s
Disturbance gain1.5000 -
Disturbance time constant200.00 s
Disturbance dead time30.000 s

Applying the formula below gives:

ResultValue
Feedforward gain-0.75000
Lead time120.00 s
Lag time200.00 s
Feedforward dead time (0 if not realisable)10.000 s

Open this example in the calculator and change any input.

Formula

\(G_{ff}(s) = -\frac{K_d}{K_p}\frac{\tau_ps + 1}{\tau_ds + 1}e^{-(\theta_d - \theta_p)s}\)

Variables

Symbol Meaning SI unit
\(K_p\) Process gain (output per % controller output) -
\(\tau_p\) Process time constant s
\(\theta_p\) Process dead time s
\(K_d\) Disturbance gain -
\(\tau_d\) Disturbance time constant s
\(\theta_d\) Disturbance dead time s
\(K_{ff}\) Feedforward gain -
\(T_{lead}\) Lead time s
\(T_{lag}\) Lag time s
\(\theta_{ff}\) Feedforward dead time (0 if not realisable) s

Method and standards

Ideal feedforward from FOPDT models.

Where it is used

  • Boiler drum level, deaerator and heater loops.
  • Blending and neutralisation.

Common mistakes

  • Wrong sign on the gain.
  • Removing feedback trim.

Assumptions and limits

  • Linear FOPDT models.
  • When θ_d < θ_p the feedforward is not realisable; the dead time is set to zero, with a note.

References

  • Seborg
  • Edgar
  • Mellichamp and Doyle
  • Process Dynamics and Control
  • 4th ed. (Wiley, 2016)
  • ch. 15 (feedforward control design from FOPDT models)

Questions and answers

When is feedforward worth it?

When a measurable disturbance (steam flow, feed rate, ambient) upsets the loop faster than feedback can correct.

Why the lead–lag?

It matches the different speeds of the disturbance and manipulated-variable paths; the static gain alone gets the final value right.

What do the defaults represent?

A loop with a 120 s process lag and a slower 200 s disturbance path.

Method and article reviewed by FireCAD engineering on 13 November 2026. Every result is checked against an independent calculation on each release: how we verify our calculators.

Related calculators

Free engineering calculators: every page computes in SI and converts to your units. Results are for engineering estimation; verify critical designs against the referenced standards.

Rejoining the server...

Rejoin failed... trying again in seconds.

Failed to rejoin.
Please retry or reload the page.

The session has been paused by the server.

Failed to resume the session.
Please retry or reload the page.