Properly Position Sampling Nozzles

Their location and orientation can significantly affect analyzer performance.

By Dean Slejko, Swagelok Company

Share Print Related RSS

In an analytical sampling system, a nozzle is used at the tap location to supply the sample to the analyzer. Proper placement of this nozzle is critical. It must be located in the right position and oriented correctly to ensure timely, accurate analytical measurements. Poor nozzle placement may lead to analysis delays, sample contamination and inaccurate results.

Ideally, a sampling system engineer will dictate piping layouts and even process vessel design to make certain the nozzle is properly situated. However, it’s more likely the engineer must work with existing schematics. And it’s possible the prescribed nozzle location isn’t in the right place or doesn’t have the correct orientation to guarantee a timely, uncompromised sample.

So, here, we’ll review some important considerations for locating and constructing nozzles for both gas and liquid stream analysis. Follow these guidelines for your facility. In addition, involve qualified analyzer engineers, process engineers and chemists, and component suppliers to ensure you consider every variable.

Nozzles typically are short and have a smaller diameter than the main process line from which they branch off. They often house a probe, which in its simplest form is a metal, glass or ceramic proboscis that pokes into the process fluid and withdraws a continuous flow for analysis.

When determining where to position the nozzle in a pipeline or vessel, select a location where the process fluid has thoroughly mixed so your sample accurately reflects process conditions. If possible, install an inline static mixer in the process line. If that isn’t an option, locate the tap downstream from a point of induced turbulence such as a pump discharge, flow orifice or piping elbow. The turbulence will help mix the process fluid before sampling.

Placing a sampling tap immediately after a source of turbulence isn’t good practice. The turbulence will cause pressure fluctuations and eddy currents — either of which may affect the analytical measurements.

You are better served by placing the tap at least two pipe diameters downstream of the last flow disturbance. The U.S. Environmental Protection Agency (EPA) recommends this practice. It permits two locations for manual stack gas sampling: 1) at least eight stack or duct diameters downstream and two diameters upstream of any flow disturbance; or 2) at least two stack or duct diameters downstream and a half diameter upstream from any flow disturbance. The EPA considers the first location to be ideal. If you can’t meet the first rule, EPA guidelines require additional sampling points to guard against the possibility of stratification.

For pipeline sampling, locate your tap at least two pipe diameters downstream of the last flow disturbance, anywhere it doesn’t interfere with a flow-metering element. However, if the process fluid is a liquid that’s close to its bubble point, it’s wise to be more conservative. To avoid getting bubbles in your sample, locate the tap where there’re at least five pipe diameters of clear straight pipe upstream and two diameters downstream (Figure 1).

When the stream being sampled is a vapor at or near its dew point temperature, the tap location becomes more critical. Some condensation may occur at pressure points near flow disturbances; you don’t want your gas sample to contain liquid condensate. To minimize this potential, one European standard (ISO 10715 1997, 13) for measuring natural gas requires the sampling tap to be at least 20 pipe diameters downstream of the last flow disturbance; the relevant American standard (API MPMS 14.1 2006, 15) requires at least five pipe diameters. If the pipe contains another probe, such as a thermowell, the sampling probe should be located at least five thermowell diameters away from the thermowell.

Even greater separation is necessary for isokinetic sampling, which calls for the velocity in the probe to match that in the process line. For example, for saturated steam, the relevant American standard (ASTM D1066) recommends the sampling tap be at least 35 pipe diameters downstream and four pipe diameters upstream of a flow disturbance. Because this separation may be difficult to achieve, ASTM suggests that noncompliant locations maintain a 9:1 ratio of upstream to downstream distances.

As you move into the design stage, you’ll notice that nozzle design, including orientation, highly depends upon the needs of the probe. Nozzle, and therefore probe, location and orientation can influence how particles within a process stream affect analysis. The right nozzle and probe placement can minimize the potential for particles to enter the probe where they may contaminate the sample or collect and eventually block the probe.

The simplest way to construct a nozzle is to weld a female threaded boss to a pipeline and drill it through (Figure 2). First, weld a reinforcing plate to the pipeline, unless its wall is adequately thick. Next, select a boss size to accommodate the threaded probe or valve being used. Another approach is to weld a male threaded pipe nipple onto the line (Figure 2).

Of course, neither of these simple methods will satisfy a process piping specification that prohibits threaded joints at the process envelope. Facing these constraints, you can weld a valve onto the nipple or use a flanged pipe nozzle instead.

Your ideal tap and nozzle location varies depending upon whether the sample is a gas or liquid.

Whenever possible, choose a horizontal pipe for sampling process gas. The horizontal orientation allows the nozzle to be vertical. It’s also easier to find a long, straight run. Locate the nozzle on top of the line (Figure 3) so any dirt or liquid falls back into the process pipe.

If the process gas is clean and dry, you could locate a gas nozzle on the side of a horizontal pipe. However, avoid such a horizontal probe orientation for dirty samples because the gas flow in the probe may not be turbulent; solid particles could fall out and block the probe

The same concern for blockage applies to a horizontal nozzle on a vertical gas pipeline (Figure 4). The inclined nozzles also shown in Figure 4 allow any entrained liquids or solids to fall back into the process. This feature works well for liquid removal but may not for solids, which may stick in the probe

For vertical stacks, the nozzle should be horizontal or slightly inclined and much larger — sized to accommodate special probes. Here, nozzle location is based on the gas temperature: not too hot (so material costs are reasonable) but above the acid dew point temperature (so sulfuric acid doesn’t condense in the probe). For stack gas monitoring, the optimum gas temperature at the nozzle is a little more than 600°C (1,112°F) to ensure complete combustion.

When sampling from a process vapor flow that may contain droplets of condensate, such as a saturated steam system, the preferred tap location is a long, downward-flowing process pipe. However, always check the line pressure. A down line often connects to the suction side of a process pump and may be running at low pressure. A low-pressure source is good for a gas sample that will be vented to flare but not for one you want to return to the process via a fast loop. With a fast loop, you probably want to sample from the process pump discharge and return the gas to the suction side.

A vertical pipeline flowing upward often is the best location for a liquid sampling nozzle because you can be sure the pipe is full. For this setup, both the horizontal and inclined nozzles shown in Figure 5 are acceptable — but, whenever possible, opt for the horizontal nozzle as it allows a shorter probe that’s less likely to vibrate. A horizontal nozzle is fine for a liquid sample because the flow in the probe is sure to be turbulent, precluding sedimentation of solids in the probe.

The head of liquid in the piping above the tap will provide some extra pressure, which can ensure the process liquid at the tap is above its bubble point. The extra pressure also helps in transporting the sample.

Don’t sample a liquid from a vertical process pipe flowing downward — there’s no guarantee the line is full.

Sampling a liquid stream from a horizontal line is risky because the process pipe may not be full, resulting in a two-phase sample. If the process pipe turns upward after the horizontal run, you can be sure the line is full; if it turns downward, the horizontal section might house a static layer of gas trapped above the flowing layer of liquid.

Conventional wisdom says liquid taps always should be on the side of a horizontal process line; this is good advice when sampling without a probe. As illustrated in Figure 6, sampling from a nozzle on the side of the line reduces the risk of extracting entrained vapor at the top of the pipe or sludge at the bottom.

In practice, it’s usually better to use a probe to reach into the center of the line. When using a probe, orientation is less important. Usually, a vertical nozzle on top of the line is preferred because it allows heavy solids to fall back into the process pipe.

Generally, the nozzle shouldn’t be in the bottom of a horizontal pipe as solids will fall into it and increase the difficulty of sample conditioning. If you must use an existing bottom nozzle, install a probe that reaches into the center one-third of the pipe diameter.

Before settling upon your final tap location, remember that your choice of probe (or no probe) will determine the orientation and diameter of the nozzle and its end fittings. Once you select a probe, revisit your chosen location to check the physical space available for installation and maintenance work. Verify there’s enough clearance for fabricating the nozzle, maintaining the process isolation valve and withdrawing the probe. Also consider accessibility and lighting. A simple tap and pipe probe might require only a small platform and ladder but a field station will need a full platform and lighting.

This is the first article in a three-part series. The second part will discuss when to use a probe for sampling while the third will cover how to orient probes. All three articles are adapted from “Industrial Sampling Systems: Reliable Design and Maintenance for Process Analyzers,” a process sampling textbook authored by Tony Waters and published in 2013 by Swagelok Company. For more information, visit

DEAN SLEJKO is product manager, analytical products, for Swagelok Company, Solon, Ohio. E-mail him at

Share Print Reprints Permissions

What are your comments?

You cannot post comments until you have logged in. Login Here.


No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments