# Residence Time Distribution in Dryers: Tracer Tests and Tanks in Series

> Residence time distribution in dryers with a worked tracer example: 9.5 tanks in series versus 2.0 changes the early-leaving share from about 14% to 30%.

Canonical: https://engitech.in/residence-time-distribution-dryers-tracer-test/ | Published: 2026-11-17 | Updated: 2026-11-17 | Author: Jignesh S Panchal

**Short answer:** Residence time distribution (RTD) describes how long different particles actually stay in a dryer or reactor, not just the average. A tracer test (inject a pulse of a harmless marker at the inlet and record its concentration at the outlet) gives the mean time and the spread. The spread tells you whether the dryer behaves like plug flow (narrow, all particles dry for the same time) or like a mixed tank (wide, some particles leave early and wet). It explains why average moisture is fine but some product is too wet.

Engineers often size dryers on the average residence time and then wonder why a small part of the product leaves wet or why the dryer cannot hold tight moisture limits. The cause is usually the spread of residence times. In my conversations with process engineers, a tracer test is the single most useful low-cost test that few plants have run. This guide explains the idea with a simple worked example. It builds on indirect dryer sizing, rotary dryer shell sizing and fluid bed dryer channeling.

## Plug flow and mixed flow

| Pattern | Description | Residence times | Effect on drying |
| --- | --- | --- | --- |
| Plug flow | Material moves as a block with no mixing along the flow | All equal to the mean | Uniform drying; ideal for tight moisture limits |
| Perfectly mixed | Feed is instantly mixed through the volume | Wide spread, some leave almost immediately | Uniform inside but part of the output is barely dried |
| Real dryers | Somewhere between: axial mixing, short-circuiting, dead zones | Moderate spread | Wetter tail and over-dried portion |

## How to run a tracer test

1. **Choose a tracer:** a harmless material that behaves like the product and is easy to detect: a coloured version of the product, salt (conductivity), a dye, a fluorescent powder or a different-sized particle. It should not change the process.
2. **Stabilise the dryer** at normal feed rate, temperature and speed.
3. **Inject a pulse** quickly at the feed point, noting the time, in a small amount.
4. **Sample the outlet frequently** (every 1 to 5 minutes for a dryer with a mean time of an hour) until the tracer is gone.
5. **Analyse each sample** for tracer concentration (by colour, conductivity or weight).
6. **Calculate** the mean residence time and variance from the concentration–time data.
7. **Safety and product:** make sure the tracer does not contaminate the product, or divert the output during the test.

## Calculating from the data

The mean residence time τ is the concentration-weighted average of time; the variance σ² measures how spread out the times are. If the dryer has a hold-up of M kg and the throughput is F kg/h, the nominal mean time is M / F. The number of ideal stirred tanks in series that would give the same spread is N = τ² / σ². N = 1 means fully mixed; large N approaches plug flow.

## Worked example

A dryer holds 800 kg of material and handles 400 kg/h, so the nominal mean time is 800 / 400 = 2.0 hours. A pulse tracer test gives a measured mean time of 1.9 hours (a little shorter, which hints at some dead volume or short-circuiting) and a variance of 0.38 h².

- Number of tanks in series N = τ² / σ² = 1.9² / 0.38 = 3.61 / 0.38 = 9.5, close to plug flow.
- Standard deviation of residence time = √0.38 = 0.62 hours (37 minutes), about 32 per cent of the mean.

Compare a second dryer with the same mean time and a variance of 1.8 h²: N = 3.61 / 1.8 = 2.0, closer to a mixed system, with a standard deviation of √1.8 = 1.34 hours. Suppose the product needs at least 1.2 hours of drying to reach target moisture. In the first dryer, 1.2 hours is (1.9 − 1.2) / 0.62 = 1.1 standard deviations below the mean, so roughly 14 per cent of the particles (for a normal approximation) would leave early; in the second, (1.9 − 1.2) / 1.34 = 0.52 standard deviations, about 30 per cent leave with less than the required time. The average moisture may look fine, but the share of wet product differs by a factor of two. The normal approximation is rough, since real RTD curves are skewed, but it shows why narrowing the RTD matters. Options include baffles or weirs, changing the paddle pitch, controlling the fill level, or using two stages in series.

## Using the results

- **Mean time lower than hold-up divided by throughput:** dead zones or short-circuiting; inspect internals, fill levels and feed distribution.
- **Long tail:** some material gets stuck and recirculates; look for build-up, bridging or poor discharge.
- **Two peaks:** parallel paths, such as bypassing around a baffle; fix the flow pattern.
- **Very wide spread:** too much axial mixing; use staged compartments or baffles.

## Applications beyond dryers

The same test is used on blenders, screw conveyors, extruders, reactors and kilns. For mixers, see blend uniformity testing. For plant-wide quality control, sampling and measurement must be sound; see measurement uncertainty.

## What I see in plants

Plants that run a tracer test usually learn something unexpected, such as 20 per cent of the dryer volume being dead, or a feed arrangement that sends half the feed straight through. They then correct it for little cost. Plants that do not test often replace the dryer or increase the temperature to cover the problem, causing quality loss. Tests can be done in a day with simple materials and a stopwatch.

## Frequently asked questions

### What is residence time distribution?

The spread of times that different particles spend in a vessel, as opposed to the average.

### Why is a narrow RTD good for drying?

Because all the product gets close to the same drying time, so there are fewer wet or over-dried particles.

### What tracer should I use?

One that flows like the product and is easily detected: a dyed portion of the product, salt or a fluorescent additive.

### How many samples do I need?

Enough to describe the curve: often 20 to 40 samples spaced across at least two to three times the mean residence time.

### Can I fix a wide RTD?

Often, by adding baffles or weirs, changing the feed point, adjusting speed or fill, or staging the dryer.

## Sources and further reading

For standards, data and independent guidance on this topic, see [US Department of Energy process heating resources](https://www.energy.gov/eere/amo/process-heating) (industrial heating and drying efficiency), [The Engineering ToolBox](https://www.engineeringtoolbox.com/) (engineering reference data and calculators), [Bureau of Energy Efficiency](https://beeindia.gov.in/) (energy audit and efficiency programmes in India) and [NABL](https://nabl-india.org/) (accredited calibration and testing laboratories). Always check the current edition and your site conditions before applying any figure.

**Related reading:** [Indirect Dryer Sizing: Heat Transfer Area from a Mass and Energy Balance](https://engitech.in/indirect-dryer-heat-transfer-area-sizing/); [Vacuum Drying Basics: Pressure, Boiling Point and Drying Time](https://engitech.in/vacuum-drying-basics-boiling-point-time/); [Hot Air Generation for Dryers: Direct-Fired, Indirect and Electric Compared](https://engitech.in/dryer-hot-air-generation-direct-indirect/); [Rotary Dryer Shell Sizing: Hold-Up, Retention Time and Speed](https://engitech.in/rotary-dryer-shell-sizing-holdup-retention/).

*This article is general information from my experience with dryers and process equipment. The test data are invented for illustration; the normal approximation is rough. Plan real tests with safety and product quality in mind.*
