MEA Storage & Shelf Life: Colour, Water Pickup and Nitrogen Blanketing

Monoethanolamine is usually received and stored as a water-white liquid intermediate for gas treating, detergent builders, corrosion inhibitors, and polyurethane catalyst precursors. Bulk storage must control three interacting shelf-life variables: hygroscopic water pickup, oxidative colour development, and carbon dioxide absorption into carbamate species. These variables do not operate independently. Water ingress can increase iron dissolution from carbon steel, dissolved iron accelerates oxygen-induced colour body formation, and oxygen or CO2 entry through the same tank vent can simultaneously raise Pt-Co colour and reduce free amine value. A certificate of analysis for fresh technical-grade MEA typically specifies an assay of at least 99.0 wt%, water content not exceeding 0.5 wt%, and colour not exceeding 15 Pt-Co units by ASTM D1209. For low-water derivative applications, the water ceiling is reduced to 0.3 wt%, measured by Karl Fischer titration to ASTM E203. Because MEA is fully miscible with water and strongly hygroscopic, water ingress does not form a separate phase and cannot be detected by visual phase separation. It is detectable only by coulometric or volumetric titration. Storage temperature is also relevant because neat MEA freezes near 10.5 °C, while elevated temperature accelerates oxidative colour formation. A partially filled fixed-roof tank in an unheated tank farm can therefore cycle between freeze-risk and oxidation-risk conditions if the liquid is not maintained between 20 °C and 35 °C. Fixed-roof tanks, horizontal pressure vessels, and indoor day tanks are all used for MEA. Fixed-roof tanks are typically equipped with pressure/vacuum conservation vents or inert-gas blanketing valves. Horizontal vessels may be pad-blanketed and fitted with pressure regulators. Day tanks in production buildings often rely on desiccant breathers, but these only address water vapour and not oxygen or CO2. The headspace composition above the liquid is therefore a critical process parameter, not a passive storage detail.

What Causes Colour Shift in Stored MEA and Which Specification Limits Apply?

Colour development in MEA is primarily oxidative rather than purely thermal degradation. MEA exposed to oxygen forms aldehydes, ammonia, and coloured oligomeric amine condensation products. The reaction is slow at ambient temperature but accelerated by dissolved metals, particularly soluble iron. Carbon steel tanks and unlined steel pipelines can release ferrous ions into the slightly alkaline MEA. The amine then complexes the iron, producing green-to-brown colour bodies. Copper, zinc, aluminium, and galvanized components are generally more aggressively attacked by MEA, so they are excluded from colour-sensitive storage and transfer metallurgy. At 40 °C and above, oxidation rates increase rapidly, and stored liquid can move from 5 Pt-Co to 25 Pt-Co before a significant change in assay is observed by gas chromatography. This makes Pt-Co colour a leading indicator of oxidative stress rather than a simple aesthetic parameter. The release limit for technical-grade MEA is commonly 15 Pt-Co under ASTM D1209. End users in polyurethane and textile intermediate applications often establish internal reject limits of 25 to 50 Pt-Co because coloured MEA can carry into final products or indicate metal contamination. Colour shift does not require bulk oxygen replacement; a tank with an open conservation vent can admit sufficient oxygen through normal diurnal thermal breathing. Field observations on MEA bulk terminals indicate that carbon steel tanks without inerting exhibit visible colour drift within 90 to 180 days, though published colour-stability data for all tank configurations is limited. Oxidative degradation of alkanolamines proceeds through hydroperoxide and imine intermediates, and the coloured products include aldol condensation products and higher-molecular-weight amine species. Because the reaction consumes dissolved oxygen, nitrogen blanketing is more effective than antioxidant addition for bulk MEA storage.

Water ingress in MEA storage is not a simple dilution event. The strong water affinity of MEA means that open or inadequately blanketed tanks can absorb atmospheric moisture until the solution vapour pressure approaches the water partial pressure in the headspace. A fixed-roof tank breathing through a conservation vent exchanges vapour-space volume with ambient air as temperature rises and falls. A 10 K diurnal swing produces a thermal volume change of about 3% of the vapour space; over many cycles this acts as a slow gas pump. In coastal or humid locations, the water content of MEA can rise from 0.3 wt% to 0.6 wt% within several weeks if the tank is not blanketed and the MEA surface is exposed to humid air. Because water is fully miscible with MEA, the absorbed water changes the amine equivalent weight and alters the molar balance in downstream syntheses. In polyurethane foam production, excess water reacts with isocyanate to generate carbon dioxide and polyurea; a water excursion from 0.3 wt% to 0.8 wt% can shift the apparent blowing ratio and change foam density. Karl Fischer titration to ASTM E203 is the accepted method for detecting such changes. Refractive index or density alone is insufficient for low-level water pickup because both also respond to carbamate and colour-forming impurities. Water also participates in corrosion. Dissolved water increases the tendency for iron to pass into solution, and the resulting iron-amine complex catalyses further oxygen uptake. Water control therefore protects both stoichiometry and colour. In addition, carbon dioxide ingress is a separate shelf-life concern because MEA reacts reversibly with CO2 to form carbamate and bicarbonate species. This reaction consumes active amine and increases liquid viscosity. In gas processing, MEA acid-gas loading is expressed as mol CO2 per mol MEA. In storage, even ambient air containing approximately 400 ppm CO2 can cause measurable amine-value loss if air ingress is continuous. Nitrogen blanketing prevents CO2 ingress, while dry-air blanketing would not.

When Nitrogen Blanketing Pressure Control Is Inadequate

A nitrogen pad over MEA is not simply an inert-gas supply; it is a pressure-regulated barrier that must hold positive pressure across the entire tank breathing cycle. If the blanketing valve is undersized for the withdrawal rate, or if the regulator setpoint is too close to the vacuum breaker setting, pump-out can draw the vapour space below atmospheric pressure. The conservation vent then opens on vacuum and admits moist, oxygen-bearing air directly into the headspace. Typical low-pressure blanketing systems for MEA operate at 0.5 to 2.5 kPa gauge, with vacuum protection set at -0.2 kPa and pressure relief at 2.8 kPa to 3.0 kPa. API 2000 7th edition provides the normal and emergency venting calculations for such fixed-roof tanks. The nitrogen blanket can reduce normal venting but does not replace the emergency vent. Supply nitrogen of 98 vol% to 99.9 vol% is common. Membrane nitrogen at the lower end may leave 1–2 vol% oxygen in the pad, often sufficient to sustain slow colour-body formation over months. For colour-sensitive MEA, residual oxygen in the blanket gas should be below 0.5 vol% and measured at the tank inlet, not at the nitrogen plant. When MEA is withdrawn at 25 m³/h from a 100 m³ tank, the nitrogen supply must replace the displaced liquid volume at the same rate. If the blanketing regulator is sized only for thermal breathing, the pressure will fall below the vacuum breaker setpoint. Thermal in-breathing can add roughly 0.3 m³/h per 100 m³ of vapour space under a 10 K diurnal cycle, but pump-out displacement can be 25–50 m³/h. Process historians on bulk storage tanks show a characteristic sawtooth pressure trace when vacuum breakers lift repeatedly during pump-out. This pattern indicates that the blanketing valve is not maintaining the pad, and every vacuum event is a water and oxygen ingress event. Blocked flame arrestors, liquid-logged conservation vents, or regulator droop can all produce this condition. In another common failure mode, a tank cooled overnight contracts the vapour space faster than the blanketing regulator can compensate, especially when the nitrogen supply header pressure is throttled. The resulting vacuum opening is brief but repeating. Carbon dioxide is excluded along with oxygen. If CO2 enters, it reacts with MEA to form carbamate, consuming amine equivalent and increasing viscosity. A nitrogen pad therefore protects free amine value and colour simultaneously.

Shelf-Life Monitoring and Re-Test Intervals Under Nitrogen

Supplier technical literature commonly lists a shelf life of 12 months for unopened, nitrogen-blanketed MEA containers stored at 15–30 °C. Bulk storage tanks under continuous nitrogen pad are usually re-tested every 3 months, while indoor drum storage may be re-tested every 6 months unless water or colour trends require shorter intervals. The re-test panel should include water and colour as minimum parameters. Assay and iron are added when colour exceeds the operational trigger. ASTM D1209 is used for colour, ASTM E203 for water, and ASTM D2074 for total amine value. Total amine value can identify carbamate accumulation from CO2 ingress because carbamate formation consumes free amine. A shift of more than 2 mg KOH/g from the initial certificate-of-analysis value is typically considered significant. The storage temperature should be maintained above 15 °C to avoid freezing, because neat MEA freezes near 10.5 °C, and below 35 °C to slow colour formation. External heating coils should use tempered water or low-pressure steam to keep coil skin temperatures below 60 °C. Local hot spots accelerate oxidation and iron pickup. Filtration through 0.5–1.0 µm filters removes particulate iron oxides but does not remove chelated iron; filtration cannot reverse colour already formed. Batch-to-batch variation in initial colour and water is usually small for producer-distilled MEA, but repackaged or drummed material can vary. Procurement specifications should therefore include not only the certificate of analysis but also lot-specific water, colour, assay, and iron data. Tanks that have been hydrostatically tested or washed must be dried before MEA charging because residual water films can raise the first-fill water content and accelerate iron dissolution.

Metallurgy and elastomer choices determine whether a nitrogen blanket can preserve MEA colour over a full shelf life. Carbon steel is acceptable for short-cycle storage where a nitrogen pad is continuously maintained and where colour tolerance is wider. Stainless steel grades 304L and 316L are preferred for low-water, low-colour MEA because they reduce iron contamination and resist the alkaline amine environment. Copper, zinc, galvanized components, and aluminium are incompatible with MEA and contribute coloured metal complexes. Elastomer seals should be selected for amine service; EPDM and butyl rubber are common materials, while natural rubber and some fluoropolymer seals may be unsuitable depending on temperature and amine concentration. Drums and intermediate bulk containers that are repeatedly opened and sampled show cumulative water and CO2 pickup; each headspace exchange with humid plant air adds a measurable increment of water. Opened containers should be re-blanketed and re-tested if held beyond the site’s validated open-container interval. Published universal kinetic data for MEA colour formation across all tank materials and climates is limited; therefore a defensible storage programme must be built on initial certificate-of-analysis values, measured water and colour trend data, and the specific nitrogen pad performance of the tank.