Formula sheet

Chemical Process Control and Plant Operations Formula Sheet

Process control formulas for operating envelopes, balance closure, residence time, dynamics, sensor lag, valve authority, feedforward, alarms, interlocks, and validation.

Branch: Chemical Engineering
Content: Formula sheet
Updated: Jun 23, 2026
Revision: v1.0.0 · reviewed

This formula sheet collects first-pass calculations used in chemical process control and plant operations. Use it to review operating envelopes, balance closure, residence time, first-order response, sensor lag, control-valve authority, feedforward utility demand, alarm response margin, interlock proof testing, capacity margin, and validation evidence.

The formulas are operational screening tools, not substitutes for a process hazard analysis, safety-instrumented-function verification, dynamic simulation, control-loop test, utility study, equipment data sheet, or operating procedure. They are useful when the process boundary, units, measurement basis, response time, alarm or trip limit, and validation record are clearly stated.

Symbols and Units

Symbol	Meaning	Common unit
$\dot{m}$	mass flow rate	kg/h or kg/s
$Q$	volumetric flow rate	m^3/h
$\rho$	density	kg/m^3
$V$	process volume or inventory	m^3
$\tau$	residence time or time constant	s, min, h
$\theta$	dead time	s or min
$K$	process gain	output unit/input unit
$C_v$	valve coefficient in vendor units	depends on convention
$\Delta P$	pressure drop	kPa, bar, psi
$U$	utility capacity or controller output	project-specific
$t$	time	s, min, h
$\lambda_{DU}$	dangerous undetected failure rate	1/time
$T_I$	proof-test interval	time
$PFD_{avg}$	average probability of failure on demand	dimensionless

Always state whether a flow is mass, volumetric, or molar. A stable trend is not automatically a valid measurement if density, calibration, sensor location, lag, bypass state, or phase condition is wrong.

Operating Envelope Margin

For an upper operating limit:

M_{upper}=x_{limit}-x_{current}

For a lower operating limit:

M_{lower}=x_{current}-x_{limit}

A normalized margin can be written as:

\displaystyle m=\frac{x_{limit}-x_{current}}{x_{limit}-x_{normal}}

for an upper limit, using a defined normal operating value. The sign convention must be stated. A positive margin means the current state is inside the reviewed envelope.

Worked Example: Reactor Temperature Margin

A reactor normally runs at $82\ \text{deg C}$ . The high alarm is at $92\ \text{deg C}$ , and the interlock trip is at $98\ \text{deg C}$ . The current temperature is $89\ \text{deg C}$ .

Alarm margin:

M_{alarm}=92-89=3\ \text{deg C}

Trip margin:

M_{trip}=98-89=9\ \text{deg C}

Normalized trip margin relative to normal operation:

\displaystyle m_{trip}=\frac{98-89}{98-82}=\frac{9}{16}=0.56

The process is still inside the envelope, but the alarm margin is small. Operators should not treat the state as normal only because the trip has not occurred.

Balance Closure Residual

For a total mass balance:

\displaystyle R_m=\sum \dot{m}_{in}-\sum \dot{m}_{out}-\frac{dM}{dt}

At steady state:

\displaystyle R_m=\sum \dot{m}_{in}-\sum \dot{m}_{out}

Relative residual:

\displaystyle r_m=\frac{R_m}{\sum \dot{m}_{in}}

Use absolute value when checking closure tolerance:

|r_m|\le r_{allowed}

Balance closure is limited by meter uncertainty, inventory change, sampling alignment, unmeasured vents, leaks, recycle holdup, and non-steady operation.

Worked Example: Closure with Inventory Change

A vessel receives $5100\ \text{kg/h}$ and discharges $4980\ \text{kg/h}$ . Level trend shows inventory increasing at $95\ \text{kg/h}$ .

Residual:

R_m=5100-4980-95=25\ \text{kg/h}

Relative residual:

\displaystyle r_m=\frac{25}{5100}=0.0049=0.49\%

If the operations tolerance is $1.0\%$ , the balance closes for this review. Without the inventory term, the apparent residual would be $120\ \text{kg/h}$ and the diagnostic conclusion would be wrong.

Residence Time and Inventory Turnover

Nominal residence time for a liquid process volume is:

\displaystyle \tau=\frac{V}{Q}

Inventory turnover rate is:

\displaystyle N=\frac{Q}{V}=\frac{1}{\tau}

For a density-based mass inventory:

M=\rho V

and:

\displaystyle \tau=\frac{M}{\dot{m}}

Residence time is a nominal hydraulic value. Dead zones, bypassing, mixing quality, gas holdup, foaming, solids, viscosity, and level control can make actual residence-time distribution different from $V/Q$ .

First-Order Process Response

A stable first-order process response after a step change can be approximated by:

y(t)=y_0+K\Delta u\left(1-e^{-t/\tau}\right)

where $K$ is process gain, $\Delta u$ is manipulated-variable change, and $\tau$ is process time constant.

With dead time:

y(t)=y_0\quad \text{for}\quad t<\theta

and:

y(t)=y_0+K\Delta u\left(1-e^{-(t-\theta)/\tau}\right)\quad \text{for}\quad t\ge\theta

Worked Example: Cooling Response Delay

A heat exchanger outlet temperature responds to a cooling-valve step with process gain:

K=-0.40\ \text{deg C}/\%

The valve is opened by $12\%$ , dead time is $1.5\ \text{min}$ , and time constant is $4.0\ \text{min}$ . Estimate the temperature change after $7.5\ \text{min}$ .

Effective response time:

t-\theta=7.5-1.5=6.0\ \text{min}

Final temperature change from the step:

K\Delta u=-0.40(12)=-4.8\ \text{deg C}

Fraction completed:

1-e^{-6.0/4.0}=1-e^{-1.5}=0.777

Temperature change:

\Delta T=-4.8(0.777)=-3.7\ \text{deg C}

The operator should not expect the full $4.8\ \text{deg C}$ reduction after $7.5\ \text{min}$ . Dead time and time constant define how fast the action becomes visible.

Sensor Lag and Measurement Validity

For a first-order sensor responding to a step in the true process variable:

y_m(t)=y_{true}+\left(y_{m,0}-y_{true}\right)e^{-t/\tau_s}

The sensor reaches a fraction $f$ of the final change at:

t=-\tau_s\ln(1-f)

Common values:

Fraction of final change	Time
$63.2\%$	$1\tau_s$
$95.0\%$	$3\tau_s$
$99.3\%$	$5\tau_s$

Sensor lag is operationally important when alarm response time is short, batch transitions are fast, or the sensor is installed in a thermowell, sample loop, dead leg, or fouling service.

Valve Flow and Valve Authority

Valve coefficient relationships depend on vendor convention and units. For liquid service in a common US-style convention:

\displaystyle Q=C_v\sqrt{\frac{\Delta P}{SG}}

where $Q$ is flow in gpm, $\Delta P$ is psi, and $SG$ is specific gravity. Do not use this equation with SI units unless the coefficient convention is converted.

Valve authority can be screened as:

\displaystyle A_v=\frac{\Delta P_{valve}}{\Delta P_{system}}

where $\Delta P_{system}$ is the total controllable pressure drop in the flow path at the operating case. Very low valve authority makes the loop insensitive to valve movement; very high valve pressure drop can waste pumping energy or create noise, flashing, cavitation, or erosion.

Worked Example: Valve Authority

A cooling-water valve has $80\ \text{kPa}$ pressure drop at the design flow. The total circuit pressure drop, including valve, exchanger and piping, is $260\ \text{kPa}$ .

Valve authority:

\displaystyle A_v=\frac{80}{260}=0.31

This is a usable screening value for many control applications. The result does not prove valve stability, cavitation margin, actuator sizing, installed characteristic, or controllability over turndown. It simply says the valve has meaningful authority in the installed hydraulic circuit.

Feedforward Utility Demand

For a heat duty:

\dot{Q}=\dot{m}c_p\Delta T

For a cooling utility:

\displaystyle \dot{m}_{coolant}=\frac{\dot{Q}}{c_{p,c}\Delta T_c}

A simple feedforward ratio for a measured feed-rate disturbance can be:

u_{ff}=K_{ff}\dot{m}_{feed}

where $K_{ff}$ is determined from heat balance, stoichiometry, or validated plant test data.

Worked Example: Cooling Flow Feedforward

A reactor feed increase adds estimated heat duty:

\dot{Q}=420\ \text{kW}

Cooling water is allowed to rise by:

\Delta T_c=8.0\ \text{K}

Use:

c_p=4.18\ \text{kJ/(kg K)}

Required cooling-water mass flow:

\displaystyle \dot{m}_{cw}=\frac{420\ \text{kJ/s}}{4.18(8.0)}=12.6\ \text{kg/s}

If the existing validated cooling flow is $9.0\ \text{kg/s}$ , the feedforward action must add about:

12.6-9.0=3.6\ \text{kg/s}

This calculation supports a feedforward setpoint only if cooling-water temperature, exchanger fouling, valve authority, and reactor heat-release assumptions are still valid.

Alarm Response Margin

Alarm response margin can be screened as:

t_{margin}=t_{consequence}-t_{detect}-t_{operator}-t_{action}

where:

$t_{consequence}$ is time from initiating condition to unacceptable consequence;
$t_{detect}$ is measurement and alarm delay;
$t_{operator}$ is operator diagnosis and decision time;
$t_{action}$ is time for the corrective action to become effective.

A positive margin means the alarm response can be credible if procedures, staffing, alarm priority, and training support the assumed times.

Worked Example: High-Temperature Alarm Margin

A runaway screening study estimates $18\ \text{min}$ from cooling loss to a defined high-temperature consequence. The temperature alarm has $2\ \text{min}$ measurement and alarm delay. Operator response is credited at $6\ \text{min}$ . Opening emergency cooling takes $4\ \text{min}$ to produce effective heat removal.

Alarm margin:

t_{margin}=18-2-6-4=6\ \text{min}

The alarm has a positive margin in the simplified timing model. The credit should not be accepted unless alarm priority, procedure clarity, control-room staffing, emergency cooling availability, and drill evidence support the assumed response time.

Interlock Proof-Test Screening

For a low-demand protective function with dangerous undetected failure rate $\lambda_{DU}$ and proof-test interval $T_I$ , a simplified average probability of failure on demand is:

\displaystyle PFD_{avg}\approx\frac{\lambda_{DU}T_I}{2}

This approximation assumes constant dangerous undetected failure rate, effective proof testing, low demand rate, and no major common-cause or systematic failure contribution.

Worked Example: Proof-Test Interval Effect

An interlock sensor and logic path have estimated:

\lambda_{DU}=2.0\times10^{-6}\ \text{h}^{-1}

If the proof-test interval is one year:

T_I=8760\ \text{h}

then:

\displaystyle PFD_{avg}\approx\frac{(2.0\times10^{-6})(8760)}{2}=0.00876

If the interval is reduced to six months:

T_I=4380\ \text{h}

then:

PFD_{avg}\approx0.00438

The formula shows why proof-test interval matters. It does not replace a full functional-safety calculation, proof-test coverage review, bypass management, common-cause analysis, or validation of the final element.

Capacity and Utility Margin

For a capacity-limited utility:

M_U=U_{available}-U_{required}

Relative margin:

\displaystyle m_U=\frac{U_{available}-U_{required}}{U_{required}}

For multiple users on a shared header:

\displaystyle U_{available,net}=U_{source}-\sum U_{other\ users}

The relevant available capacity may be lower than nameplate because of fouling, ambient conditions, standby equipment, header pressure, control-valve authority, maintenance state, or contingency requirements.

Validation Table

Formula area	Validation evidence
operating envelope margin	approved operating limit table, alarm/trip setpoints, current trend
balance closure	calibrated flowmeters, inventory trend, sampling alignment
residence time	level, flow, tracer test or product transition evidence
first-order dynamics	bump test or historical step-response data
sensor lag	calibration record, thermowell/sample-line review, response test
valve authority	hydraulic calculation, valve data sheet, installed flow test
feedforward utility	measured duty, coolant flow, inlet/outlet temperatures
alarm margin	alarm rationalization, operator drill, response-time evidence
interlock proof test	proof-test procedure, bypass log, demand history, final-element test
utility margin	header trend, equipment availability, fouling and ambient basis

Good plant operations preserve not only the number, but the evidence behind the number: instrument tag, calibration state, operating mode, process boundary, timestamp, units, uncertainty, bypass state, and operator action.

Common Mistakes

Common mistakes include closing a mass balance without inventory change, using volumetric flow when the decision requires mass flow, applying steady-state formulas during startup, trusting a slow analyzer during a fast transition, copying control-theory tuning rules without process dead-time review, and treating alarm setpoints as safeguards without response-time evidence.

Other recurring mistakes are operational: crediting an interlock without proof-test coverage, changing utility lineups without recalculating available capacity, ignoring valve authority after a pump or exchanger change, and using a historical operating range as if it were an approved operating envelope. A formula is useful only when it is tied to a measurement, a limit, and an auditable operating decision.

REF

Disciplines