Guia de Projeto de Sistemas de Amostragem para Analisadores de Processo (2026)

Sampling System Design Guide: How to Build a Reliable Closed-Loop Sampling System for Process Analyzers

A well-designed sampling system is the difference between an analyzer that delivers trustworthy data and one that produces false alarms, premature plugging, or — worse — sample misrepresentation that leads to bad process decisions. Whether you are sizing a new analyzer shelter, upgrading from a slip-stream to a closed-loop configuration, or troubleshooting a system that keeps failing, this guide walks through the engineering decisions that determine whether your sample arrives at the analyzer fast, clean, and representative.

What Is a Sampling System?

A sampling system is the network of components that extracts a process fluid from a pipe or vessel, conditions it to a state the analyzer can measure, and returns it (in closed-loop systems) or vents it (in slip-stream / open-loop). The system typically includes:

• Probe — retractable or fixed insertion into the process line
• Primary filter / coarse filter — removes particulates that would foul downstream components
• Pressure regulation — reduces process pressure to analyzer-friendly levels (typically 5-30 psig)
• Heat tracing / cooling — maintains sample above dew point or below vaporization temperature for the species of interest
• Phase separation — knock-out pot or membrane to remove liquids from gas samples (or vice versa)
• Fine filtration — sub-micron filters to protect analyzer cells
• Flow measurement and control — rotameter or mass flow controller with bypass
• Sample return or vent — back to process at lower pressure (closed-loop) or to safe location (open-loop)
• Fast-loop / bypass loop — high-flow circulation to reduce transport lag, with analyzer tapping off a side stream

For most online analyzer applications in oil & gas, petrochemical, and refining, the goal is to deliver the sample to the analyzer within a defined transport lag time (typically 30 seconds to 5 minutes) at a stable flow rate, pressure, e temperature — without altering the chemical composition.

Closed-Loop vs. Open-Loop vs. Slip-Stream

The three common architectures each have distinct tradeoffs. Choose based on process value, safety, and environmental constraints.

Practical recommendation: If you are sampling hydrocarbons, hydrogen sulfide, or any other material with safety, environmental, or economic value, default to closed-loop. It is more complex to design, but it pays back the moment a regulator asks about vented emissions or a process engineer wants faster analyzer response.

Design Step 1 — Define the Measurement Objective

Before specifying components, answer four questions:

1. What is the analyzer measuring? (H₂S, O₂, moisture, dew point, composition, pH, conductivity — each drives different conditioning)
2. What is the required accuracy and response time? (ppb-level H₂S in natural gas requires more conditioning than percent-level O₂ in air)
3. What is the phase of the sample at the sample point? (Single-phase gas, single-phase liquid, or two-phase that requires conditioning)
4. What is the upset condition? (Slugging, hydrate formation, particulate storms — define the worst case the system must survive)

These answers drive every subsequent decision: filter micron rating, regulator type, heat-tracing requirement, and whether you need a liquid pump or can rely on process pressure alone.

Design Step 2 — Probe Selection and Sample Point Location

The probe is where sampling problems begin or end. A good sample point is:

• Located in a section of pipe with well-mixed, single-phase flow (avoid dead legs, the bottom of horizontal pipes, or just downstream of tees)
• Installed on the top or side of horizontal pipe for gas service, and on the side or bottom for liquid service
• Oriented so the probe faces upstream (typically 45° angle) so the sample does not carry debris into the probe
• Equipped with a retractable probe assembly with isolation valve if the process cannot be shut down for maintenance

For services with heavy particulates, consider a filter probe (5-20 micron sintered element) at the sample point. This dramatically extends the life of downstream filters and protects regulators from erosion.

Design Step 3 — Transport Lag and Fast-Loop Sizing

Transport lag is the time between a process change and the analyzer detecting it. For most process control loops, total lag should be under 60 seconds. For safety interlocks, under 10 seconds.

Formula: Transport lag (seconds) = (Volume of sample line in cc) / (Volumetric flow rate in cc/sec)

Two ways to reduce lag:

• Smaller diameter tubing — 1/8″ OD (3 mm) instead of 1/4″ OD (6 mm) cuts volume by 75%
• Faster flow rate (fast-loop) — a 1-2 L/min fast-loop with a 100-200 cc/min analyzer side stream typically achieves 10-20 second transport lag over 50-100 meters

Rule of thumb: For runs over 50 m, always use a fast-loop. The investment in a small circulation pump and bypass flowmeter pays back in faster analyzer response and better process control.

Design Step 4 — Pressure Regulation and Relief

Pressure regulation has two jobs: reduce process pressure to analyzer inlet specification, and protect the analyzer from over-pressure during a process upset. Best practice uses a two-stage regulator with a relief valve between stages:

1. First-stage regulator — reduces from process pressure to intermediate (e.g., 50-100 psig)
2. Relief valve — vents to safe location if first-stage fails open
3. Second-stage regulator — reduces to analyzer inlet (typically 15-30 psig for gas chromatographs, 5-10 psig for some IR cells)

Use metal-to-metal diaphragm regulators for flammable or toxic service — soft-seated regulators can leak small amounts that are unacceptable for H₂S or benzene service.

Design Step 5 — Heat Management and Trace Heating

Two heat management scenarios are common:

For natural gas service with moisture, the standard is heat-traced and insulated sample lines at 50-80°C. Without it, hydrate formation can plug the line within hours. For liquid hydrocarbon service with dissolved H₂S or O₂, maintaining sample temperature above the bubble point prevents outgassing and false analyzer readings.

Design Step 6 — Phase Management and Filtration

Most analyzer problems are phase or particulate problems. The conditioning train should address both:

• Coalescing filter (0.5-1 micron) — for gas services, removes entrained liquids
• Particulate filter (5-20 micron) — for liquid services, removes solids
• Membrane separator — for separating gas from liquid when phase change is undesirable
• Knock-out pot — for high-volume liquid removal in two-phase samples
• Final filter (0.1-0.5 micron) — protects the analyzer cell from any residual contamination

Filter selection rule: Change the filter when differential pressure across it exceeds the manufacturer’s recommendation (typically 15-25 psid). Mark this on the P&ID and maintenance plan.

Design Step 7 — Material Selection and Compatibility

Material selection is driven by the process fluid, temperature, and pressure. Common pairings:

Common failure mode: Choosing standard Viton seals for H₂S service. Viton swells and softens in H₂S above 100 ppm, causing leaks. Use FFKM (Kalrez, Chemraz) for sustained H₂S exposure.

Design Step 8 — Purging, Leak Testing, and Commissioning

Before bringing the system online, three things must happen:

1. Pressure-test the sample loop with nitrogen or helium at 1.5x design pressure for 30 minutes. No pressure decay allowed.
2. Leak-test all fittings with soap-bubble solution or helium sniffer, especially in H₂S service.
3. Purge the system with dry nitrogen for 3-5 volume changes to remove oxygen and moisture. Oxygen analyzers require oxygen-free start-up to avoid sensor damage.

Document all three steps in the commissioning package. Skipping the leak test is the most common cause of an analyzer shelter evacuation in the first month of operation.

Common Design Mistakes to Avoid

1. Oversized tubing — 1/4″ or 3/8″ tubing kills analyzer response time. Use 1/8″ for analyzer tap, 1/4″ for fast-loop only.
2. Long runs without fast-loop — anything over 50 m needs a fast-loop circulation pump.
3. Dead legs — analyzer side streams should tap off the fast-loop, not from a stagnant tee.
4. Wrong filter orientation — flow direction arrows are there for a reason; reverse installation halts filtration.
5. No heat tracing on wet gas service — hydrates will form and plug the line within hours.
6. Single-stage pressure regulation — a single regulator failure can destroy an analyzer cell. Always use two stages with relief.
7. Skipping the leak test — toxic gas leaks in analyzer shelters create shutdowns and OSHA / HSE reportable events.

Design Checklist

• ☐ Measurement objective and target accuracy defined
• ☐ Single-phase condition verified at sample point
• ☐ Probe orientation correct (upstream-facing, 45°)
• ☐ Sample line diameter: 1/8″ for analyzer tap, 1/4″ for fast-loop
• ☐ Total transport lag calculated: target < 60 s (control), < 10 s (safety)
• ☐ Pressure regulation: two-stage with relief valve
• ☐ Heat tracing sized for ambient extremes and dew-point margin
• ☐ Filtration train: coalescer → particulate → final
• ☐ Material compatibility verified against full process composition including trace species
• ☐ Closed-loop return pressure sufficient or pump specified
• ☐ Commissioning plan: pressure test + leak test + purge documented

Build a Reliable System with the Right Partner

Designing a sampling system that survives 5+ years of continuous service without plugging, drift, or false readings requires engineering discipline that is hard to retrofit after construction. If you are starting a new analyzer project, contact Wanan early in the design phase. We can review your P&ID, recommend component sizing, and supply a complete sample conditioning panel pre-piped, leak-tested, and documented — ready to install.

Related Resources

• See the principle in action: What Is a Closed-Loop Sampling System?
• Understand the quick-connect hardware: Sampling System Components
• Compare flow control approaches: Como funciona um PVRV? (similar logic for pressure control)

Perguntas frequentes

Q1: What is the difference between a sampling system and a probe?
The probe is the single component that penetrates the pipe or vessel. The sampling system is the entire network: probe, transport tubing, filters, regulators, heat tracing, flow control, and return path. A probe is a small piece of the larger system.

Q2: How do I size the fast-loop pump?
The fast-loop pump should deliver 1-2 L/min of circulation flow (gas service) or 5-10 L/min (liquid service). The analyzer side stream taps off 100-200 cc/min. A bypass loop with rotameter or mass flow meter on the fast-loop, and a needle valve on the analyzer tap, gives you the control you need.

Q3: How long should the sample line be?
As short as practical. Under 30 m, you can usually rely on simple transport without a fast-loop. Over 50 m, a fast-loop is mandatory. Over 200 m, consider moving the analyzer closer to the sample point, or using a field-mounted analyzer with digital communication back to the control room.

Q4: Can I use plastic tubing for analyzer sample lines?
Generally no. Plastic tubing (PFA, PTFE) is permeable to many gases, and reacts with some hydrocarbons. Use 316L SS electro-polished tubing with compression fittings (Swagelok, Gyrolok) for most analyzer service. PFA is acceptable for low-pressure aqueous service and laboratory setups, but never for process analyzer shelters with flammable service.

Q5: What is the typical maintenance interval for a sampling system?
Filter elements: 1-6 months depending on service. Regulator diaphragms: 2-5 years. Heat tracing: annual inspection. Sample line integrity: pressure-test every 2-3 years or after any process upset. A well-designed system should require less than 4 hours of routine maintenance per month.

Q6: How do I handle start-up and shutdown?
On start-up, isolate the analyzer, bring the sample loop up to pressure slowly, leak-test, then purge with nitrogen for 3-5 volume changes before slowly introducing process sample. On shutdown, the order reverses: isolate the sample, purge with nitrogen, vent the analyzer, and leave the loop pressurized with nitrogen for the next start. Documenting these sequences in the operating procedure prevents most common start-up accidents.