Process Design

CSTR vs Plug Flow Residence Time

The same nominal residence time τ = V/Q can mean very different performance in a well-mixed (CSTR), plug-flow, or non-ideal vessel. Learn why the residence-time distribution — not just the average — governs conversion, and why short-circuiting and dead zones break the nominal number.

TypeEngineering guide — concept explainer

Definition

Nominal residence time is τ = V/Q — the working volume divided by the volumetric flow — and it is the same number whether a vessel is perfectly mixed, in plug flow, or somewhere in between. But that single average hides how material actually moves through the vessel. In an ideal continuous stirred-tank reactor (CSTR) the contents are uniform, so a parcel of fluid can leave almost immediately or linger far longer than τ — a broad residence-time distribution (RTD). In ideal plug flow every parcel takes exactly τ to traverse the vessel — a narrow RTD, effectively a spike at τ. Real vessels sit between these ideals, and short-circuiting or dead zones can move the actual distribution a long way from either.

Why it matters

Two vessels with identical τ can deliver very different performance, because conversion, contact, and treatment depend on the distribution of times, not just the mean. For a reaction faster than first order, plug flow generally reaches higher conversion than a single CSTR at the same τ, because no fluid escapes early. A single well-mixed tank, by contrast, always lets some fluid leave almost at once — which is why leach and contact circuits are built as several tanks in series, approaching plug-flow behaviour while keeping the mixing each stage needs. If you size only on nominal τ and assume ideal behaviour, you can badly over- or under-estimate what the vessel achieves, and you will miss the short-circuiting and dead-zone problems that a tracer test would expose.

Formula

Nominal residence time

τ = V / Q

Mean residence time (from RTD)

t_mean = ∫ t·C dt / ∫ C dt

Tanks in series (concept)

N equal CSTRs → plug flow as N → ∞

Spread (variance)

σ² = ∫ (t − t_mean)²·C dt / ∫ C dt

Units involved

•V — working volume in m³, litres, or gallons
•Q — volumetric flow in m³/h, L/s, or gpm
•τ, t_mean — time in min, h, or s (keep volume and flow time units consistent)
•σ² — variance in time²; the dimensionless variance σ²/t_mean² is unitless
•N — number of equal tanks in series (dimensionless)

Concept diagram

Worked example

A contact duty needs τ = 60 min of nominal residence time at Q = 30 m³/h. Compare a single 30 m³ tank with four 7.5 m³ tanks in series — same total volume, same nominal τ.

01Single tank: V = Q·τ = 30 × 1 h = 30 m³, τ = 30 / 30 = 1 h = 60 min
02Four tanks in series: total V = 4 × 7.5 = 30 m³ → same nominal τ = 60 min
03Single CSTR: broad RTD — some fluid leaves in minutes, lowering effective contact
04Four-in-series: narrower RTD, closer to plug flow — less short-circuiting
05Both read τ = 60 min, but the four-tank train treats the fluid more uniformly

Result

Identical nominal residence time, different residence-time distribution: the tanks-in-series train behaves closer to plug flow and short-circuits less than a single equal-volume CSTR.

Common mistakes

•Treating nominal τ = V/Q as a guarantee that every parcel of fluid stays for τ — it is only an average.
•Assuming a single stirred tank and a plug-flow vessel of equal volume perform the same at equal τ.
•Ignoring short-circuiting and dead zones, which move the actual RTD far from the ideal.
•Using one big tank where a train of smaller tanks in series would give the contact time the process needs.
•Reading residence time as reaction completion — kinetics, not just time, set conversion.

When to use the calculator

Use the residence-time calculator for nominal τ = V/Q, and the RTD tracer test calculator to estimate the mean residence time and variance from an actual tracer response — the measured spread is what tells you how close a vessel is to CSTR or plug-flow behaviour.

Residence Time Calculator →RTD Tracer Test Calculator →Tank Volume Calculator →

FAQ

Why do leach and contact circuits use several tanks in series?

A single stirred tank has a broad residence-time distribution, so some fluid short-circuits and leaves before it has had enough contact. Splitting the same total volume into several tanks in series narrows the distribution toward plug flow — each parcel must pass through every stage — while still giving each tank the mixing it needs. Four or more stages is a common rule of thumb to limit short-circuiting.

Does plug flow always beat a CSTR?

For most reactions faster than zero order, plug flow reaches higher conversion than a single CSTR at the same nominal residence time, because no fluid escapes early. The advantage depends on the kinetics, and a tanks-in-series arrangement captures most of it. This guide is conceptual — the actual comparison for a given duty needs the reaction kinetics and a model.

What is the dimensionless variance for?

The dimensionless variance σ²/t_mean² is a compact measure of how spread out the RTD is. Near zero suggests plug-flow behaviour; near one suggests a single well-mixed tank. It is often used to back out an equivalent number of tanks in series or a dispersion number — but those are modelling steps beyond this concept guide.

How do I tell which behaviour my real tank has?

Run a tracer test and look at the response curve. A sharp pulse near τ points to plug flow; a broad decaying curve points to a well-mixed tank; an early spike or a long tail points to short-circuiting or dead volume. The RTD tracer test calculator turns the response into a mean and variance you can compare against the ideals.