Tray Levelness Testing for Cryogenic Distillation Columns: Methods, Standards, and Best Practices

Tray levelness testing is a critical quality-control activity for cryogenic distillation columns used in air separation, LNG, and ethylene plants, where even small elevation deviations across a tray can cause liquid maldistribution, weeping, and measurable losses in separation efficiency. This article explains the technical principles, tested components, measurement methods, acceptance criteria, and typical applications of tray levelness detection, with reference to recognized standards such as GB 50231 and EN 13445.

Technical Overview

Cryogenic distillation columns operate at temperatures ranging from approximately -100 °C down to -196 °C to separate close-boiling gas mixtures such as oxygen and nitrogen, methane and nitrogen, or ethane and ethylene. Because the relative volatility between these components is extremely small, the columns rely on dozens or even hundreds of mass transfer trays to achieve the required purity. Each tray must hold a uniform layer of liquid so that rising vapor contacts the liquid evenly across the entire active area. When a tray is not level, liquid depth becomes unequal: shallow zones weep or blow dry while deep zones channel vapor, and the overall tray efficiency, hydraulic capacity, and pressure drop all deteriorate.

Tray levelness refers to the deviation of the tray deck surface from a true horizontal reference plane. It is normally quantified as the peak-to-valley difference, meaning the elevation difference between the highest and lowest measured points on the tray, or as the deviation of individual points from the calculated mean plane. Because cryogenic columns and their internals are frequently fabricated from aluminum alloys with relatively low stiffness, and because repeated cool-down and warm-up cycles introduce thermal stresses, levelness must be verified at installation, after any repair, and periodically during plant turnarounds.

Acceptance criteria are typically defined by the equipment design documents and the tray manufacturer's specification, within the general framework of installation acceptance standards such as GB 50231 (general code for construction and acceptance of machinery and equipment installation), GB 150 and EN 13445 for pressure equipment integrity, and the quality management principles of ISO 9001. Commonly specified limits are a peak-to-valley tray deck deviation of approximately 3 mm for small-diameter trays and up to approximately 6 mm for large-diameter trays, with stricter limits, often around 2 mm or less, applied to outlet weirs and other liquid-depth-critical edges.

Testing Samples

Sieve Trays

Sieve trays, consisting of perforated panels supported on rings and beams, are the most common tray type in air separation and light hydrocarbon service. Their levelness directly controls liquid holdup and vapor distribution across the perforated active area, so both the deck panels and the supporting structure are measured.

Float Valve Trays

Float valve trays use movable valve units over perforated or slotted openings. Although the valves tolerate some vapor maldistribution, an unlevel deck still produces uneven liquid depth and premature weeping. Testing covers the deck plane, the valve seat areas, and the outlet weir.

Bubble Cap Trays

Bubble cap trays maintain liquid on the deck at all loads, which makes them sensitive to levelness because liquid depth governs the static seal over the cap slots. Deck levelness, riser elevation uniformity, and weir levelness are all included in the measurement program.

Associated Internal Components

Levelness verification is also applied to outlet weirs, inlet weirs, downcomer aprons and their bottom edges, support rings and support beams, tray manway panels, and the support surfaces of liquid distributors and collectors in packed sections of hybrid columns.

Testing Items

A complete tray levelness detection program for a cryogenic distillation column typically includes the following measurement items:

• Overall tray deck levelness, expressed as the peak-to-valley elevation difference across the active area of each tray.
• Outlet weir levelness, which governs the uniform liquid crest height and the liquid flow pattern toward the downcomer.
• Downcomer bottom edge levelness and clearance, ensuring equal liquid seal and preventing vapor bypass up the downcomer.
• Support ring and major beam levelness, because supporting structure errors propagate into every panel installed on it.
• Tray-to-tray spacing and cumulative elevation error over the full column stack, verifying that no progressive tilt accumulates along the column.
• Local deformation checks, including panel dishing, bowing, and warping introduced during fabrication, transport, or foot traffic during installation.
• Post-repair verification measurements confirming that corrective shimming or panel replacement restored the tray to within specification.

Testing Methods

Datum and Reference Plane Establishment

Before any measurement, a stable horizontal datum must be established inside the column. This is usually achieved by transferring a benchmark elevation from the column shell datum ring or from the foundation reference mark, using the verified column verticality record as a cross-check. All tray measurements are then referred to this common datum so that tray-to-tray comparisons and cumulative error calculations are meaningful.

Precision Optical Leveling

High-precision automatic levels or digital levels are set up on a stable platform, typically at a mid-height manway or on a dedicated instrument tray, and readings are taken on a graduated staff or short ruler placed at a grid of predefined points on the tray. A grid spacing of roughly 300 to 500 mm is common for active areas, with additional points along weirs and near the downcomer. Precision levels with a double-run standard deviation of 1 mm per km or better provide ample accuracy for millimeter-level tolerances.

Rotating Laser Level Method

A rotating laser level projects a horizontal laser plane, and a receiver mounted on a graduated rod indicates the elevation of each grid point relative to the laser plane. This method is fast, requires only one operator inside the column, and is well suited to columns with many trays of identical geometry, such as air separation upper columns. Instrument self-calibration functions should be checked at each tray to guard against drift.

Hydrostatic Hose Leveling

In locations where sight lines for optical or laser instruments are obstructed, a hydrostatic leveling hose with transparent tubes at both ends provides a direct equal-elevation reference. This method is useful for verifying weir edges and downcomer bottoms inside insulated cold boxes, although the water temperature must be equalized and read quickly to avoid errors from temperature gradients and evaporation.

Digital Inclinometers and Laser Trackers

Precision digital inclinometers placed sequentially on beam flanges and deck panels provide rapid spot checks of local tilt and are convenient during installation. For large-diameter, high-value columns, portable laser trackers or laser scanners can generate full three-dimensional surface maps of each tray, allowing contour visualization, permanent digital records, and comparison between successive turnarounds.

Data Processing and Acceptance Criteria

All point elevations are corrected to the datum, plotted as a tray contour map, and evaluated against the specification. Typical evaluation steps include computing the mean plane, identifying the highest and lowest points, verifying the peak-to-valley value against the diameter-based limit, and confirming weir and downcomer edge tolerances separately. Where deviations exceed limits, corrective actions such as shimming supports, re-flattening or replacing panels, or adjusting weir height are applied, followed by re-measurement and formal documentation in the inspection record.

Applications

Tray levelness detection is applied wherever cryogenic mass transfer columns are installed, repaired, or periodically inspected, most commonly in the following situations:

• Air separation units producing oxygen, nitrogen, and argon, including high-pressure and low-pressure columns, argon side-draw columns, and pure argon rectification columns.
• Ethylene plant cold sections, such as demethanizers, deethanizers, ethylene rectifiers, and acetylene removal columns.
• LNG baseload and peak-shaving facilities, nitrogen rejection units, and end flash columns.
• Natural gas processing plants with cryogenic expander circuits for NGL recovery and dew-pointing.
• Industrial and rare gas purification trains, including krypton and xenon recovery columns.
• Turnaround inspection and post-repair verification programs in refineries, gas plants, and chemical complexes, often executed by accredited third-party laboratories together with tray integrity and fastener checks.

FAQ

Q: Why is tray levelness more critical in cryogenic columns than in conventional distillation columns? A: Cryogenic separations involve components with very close boiling points, requiring many theoretical stages and very high tray efficiency. Even a few millimeters of levelness error can cause weeping and channeling that measurably reduce product purity and increase energy consumption in refrigeration cycles.

Q: What levelness tolerance is typically acceptable? A: Design documents and tray manufacturer specifications govern, but common practice limits peak-to-valley deck deviation to approximately 3 mm for smaller trays and up to approximately 6 mm for very large trays, while outlet weirs are often held to about 2 mm or tighter.

Q: Can levelness be measured while the column is cold and in service? A: Direct tray measurement requires access to the internals, so it is performed at ambient temperature during construction or shutdown. Some plants infer hydraulic anomalies from temperature and composition profiles during operation and then confirm the cause by physical measurement at the next turnaround.

Q: What is the difference between levelness and flatness of a tray? A: Levelness describes the orientation of the tray relative to a horizontal datum over the whole tray, while flatness describes local surface deformation of an individual panel. A tray can be level overall yet fail due to local dishing, so both are evaluated.

Q: How often should tray levelness be checked? A: It is verified at initial installation, after any internal repair or panel replacement, and periodically at major turnarounds, particularly for columns that have experienced thermal cycling, flooding events, or evidence of separation performance decline.

Q: Which standards apply to the measurement? A: Measurement and acceptance generally follow the equipment design documents and tray vendor specifications, supported by installation acceptance standards such as GB 50231, pressure equipment standards such as GB 150 and EN 13445, and quality management systems aligned with ISO 9001.