How to Verify Sample-System Lag Time
When an analyzer reading lags the process it is measuring, the delay is almost always transport time in the sample system, and knowing that number matters for control tuning, custody timing, and troubleshooting alike. This procedure is for the analyzer or controls engineer who needs to confirm how long it actually takes a change at the sample tap to reach the analyzer cell. It shows the symbolic calculation from tubing volume and flow, then how to prove that figure with a real gas-switch step test rather than trusting the arithmetic alone. The worked example gives you a number you can defend.
Verify Sample-System Lag Time in one line: To verify sample-system lag time, first calculate the transport delay symbolically as the internal volume of the sample tubing and conditioning components divided by the volumetric flow rate through them, which gives the residence time. Then confirm that figure empirically with a step test: switch the sample stream from process to a known gas at the probe, start a timer, and record how long until the analyzer reading begins to move and until it reaches its new steady value. The measured delay should agree with the calculated residence time.
Gather the System Data You Need
The calculation needs the physical dimensions of the sample path and the flow through it. Collect the length and internal diameter of every tube run from the sample tap to the analyzer, the internal volume of any filters, coalescers, knockouts, or a fast loop that the sample passes through, and the actual flow rate at the analyzer, read from its rotameter or flow indicator rather than assumed from a design figure. The distinction between the flow that races past in a fast loop and the slower flow that actually reaches the cell matters, so if the system uses one, understand it through the guide to a fast loop sample system.
Be honest about where the volume hides. A large-bore filter housing or a knockout pot can hold far more volume than the tubing, and that stagnant volume dominates the lag. The concept of how long sample sits in the loop is covered in the note on sample loop residence time, which is the same quantity you are about to compute. Note whether the analyzer draws a continuous flow or takes a discrete grab, because a grab-based cycle adds the analysis time on top of the transport delay.
Calculate the Transport Lag Symbolically
Transport lag is residence time: the volume the sample must displace divided by the rate it is displaced. Symbolically, for a tube of internal diameter d and length L, the internal volume is V equals pi times (d divided by 2) squared, times L. Sum that tube volume with the internal volumes of the filters and conditioning components in the path to get total system volume V_total. The transport lag t_lag then equals V_total divided by Q, the volumetric flow rate through the path in the same volume units per unit time.
A worked example makes the arithmetic concrete. Take a sample line of 6 mm internal diameter and 30 m length. The cross-sectional area is pi times (0.003 m) squared, which is about 2.83 times ten to the minus five square metres. Multiply by 30 m and the tube volume is about 8.48 times ten to the minus four cubic metres, or roughly 0.85 litres. Suppose a filter and knockout add another 0.4 litres, giving V_total of about 1.25 litres. If the analyzer draws 1 litre per minute, the transport lag is 1.25 divided by 1, which is about 1.25 minutes, or 75 seconds, before a change at the tap even begins to reach the cell. Double the flow to 2 litres per minute and the lag halves to about 38 seconds, which is exactly why flow rate is the lever you have over sample lag. This is the transport component of the total delay discussed in the note on analyzer sample lag time in gas quality.
Run a Gas-Switch Step Test to Confirm It
The calculation assumes plug flow and no mixing, which is never perfectly true, so confirm it with a step test. At the sample probe or the nearest upstream point, switch the analyzer feed from the process stream to a cylinder of gas with a clearly different composition - a zero gas or a span gas works well because the analyzer response will be unmistakable. Start a timer at the instant you make the switch.
Record two times. The first is the dead time: how long until the analyzer reading first begins to move, which is the pure transport delay you calculated. The second is the time to reach the new steady value, which adds the mixing and washout of the volumes to the pure transport time. The dead time should land close to your calculated t_lag; if the measured dead time is far longer, you have more volume in the path than you accounted for, a partially plugged line, or a lower flow than the rotameter suggests. If it is shorter, you may have a bypass or a leak drawing gas in downstream of where you switched. The total analyzer lag, transport plus mixing plus any analysis cycle, is what control loops and custody timing actually experience, as described in the note on analyzer response time and lag.
Verify the Result and Reconcile the Two Numbers
A verified lag time is one where the calculated residence time and the measured step-test dead time agree within reason, and where any discrepancy has an explanation you can name. If they match, you have a defensible transport lag you can use to offset the analyzer against the process in trends, to set control loop expectations, or to explain why a custody value reflects gas that entered the tap over a minute ago. If they do not match, the step test wins, because it measures reality, and the mismatch points you at the wrong flow reading, an unaccounted volume, or a restriction.
Recording the verified lag against the sample system gives you a baseline that a monitoring platform such as Merobix can help you defend over time. If the analyzer feeds a trended tag and its apparent lag grows, a line is slowly plugging or a filter is loading, and the lag you established at commissioning is the number that later behavior is compared against. The calculation gives you the design figure; the step test proves it; the trend shows it drifting when the sample path degrades.
Frequently Asked Questions
How do I calculate analyzer sample lag from tubing dimensions?
Compute the internal volume of the sample path and divide by the flow through it. For a tube, volume equals pi times the internal radius squared times the length, and you add the internal volumes of filters, knockouts, and conditioning components. Dividing that total volume by the actual volumetric flow at the analyzer gives the transport lag in time. Doubling the flow halves the lag, which is the main lever you have over sample delay.
Why measure lag with a step test if I can calculate it?
The calculation assumes ideal plug flow with no mixing and the correct flow and volume figures, none of which is guaranteed. A gas-switch step test measures the real dead time and the real time to steady state, catching unaccounted volume, partial line plugging, or a flow that differs from the rotameter reading. When the calculated and measured values disagree, the step test is the one to trust because it reflects the system as actually built and flowing.
What is the difference between dead time and time to steady state in a step test?
Dead time is how long after the gas switch the analyzer reading first begins to move, and it equals the pure transport delay of pushing the new gas through the path. Time to steady state is longer because it also includes the mixing and washout of the sample volumes as the old gas is displaced. Control loops care most about the dead time, while the full time to steady value tells you how fast a real process change fully registers.
Automation services
Need help turning this into a working system?
Merobix integrates SCADA, programs Allen-Bradley and Siemens PLCs, and designs and fabricates industrial control panels.
Meeting requests are reviewed before confirmation.