Continuous hydrolysis of propylene oxide to propylene glycols proceeds through a network of consecutive oxirane ring-opening reactions in which the initially formed monopropylene glycol remains nucleophilically active and can add a second propylene oxide molecule to yield dipropylene glycol, which can add another to yield tripropylene glycol and higher oligomers. In the non-catalytic industrial route, the reaction is conducted with a large molar excess of water, typically 15:1 to 25:1 water to propylene oxide at the reactor inlet, at temperatures of 190 °C to 220 °C and pressures of 15 bar to 25 bar sufficient to retain a liquid phase. Under these conditions propylene oxide conversion exceeds 99%, and the product distribution can be steered toward monopropylene glycol by maintaining a high local water concentration relative to the concentration of glycol hydroxyl groups. The selectivity dependence is not governed by a single equilibrium constant but by the relative rates of water attack on the epoxide and the competing attack of mono-, di-, and tripropylene glycol on the epoxide. Since the rate constants for water and alcohol hydroxyl attack are of the same order of magnitude, the local molar ratio of water to total glycol hydroxyl equivalents becomes the dominant process variable.

In a conventional adiabatic tubular reactor, fresh propylene oxide, demineralized water, and recovered water are preheated to approximately 150 °C before entering a static mixer; the mixture then flows through a shell-and-tube reactor with internal diameters in the 25 mm to 50 mm range and length-to-internal-diameter ratios commonly above 1000:1 to maintain plug-flow behavior. A tube-side pressure drop of 2 bar to 6 bar is typical, and the reactor effluent is cooled against incoming feed before entering the first distillation column. Local water-to-propylene oxide ratio is not constant along the tube because reaction consumes water and accumulates glycols; therefore, even a feed ratio of 15:1 may decline in the final reactor zone to an effective water-to-glycol hydroxyl ratio below 8:1 unless the reactor is designed with sufficient water holdup. The resulting axial selectivity drift is measurable by gas chromatography according to ASTM E202-18 and by water determination per ASTM E203-16, and it explains why reactor conversions near completion can produce more dipropylene glycol than predicted from feed composition alone.

How Water-to-Propylene Oxide Molar Ratio Governs Homolog Distribution in Non-Catalytic Hydrolysis

The water-to-propylene oxide molar ratio at the reactor inlet exerts a pronounced nonlinear influence on monopropylene glycol selectivity. Representative steady-state distributions reported in open process literature for uncatalyzed propylene oxide hydrolysis at 190 °C to 200 °C and 15 bar to 20 bar, with total propylene oxide conversion above 99%, are summarized below. The tabulated ranges are consistent with kinetic models and patent disclosures for non-catalytic hydrolysis, although published plant-specific data are treated as proprietary and vary with residence time, temperature profile, and recycle stream purity.

At 20:1, monopropylene glycol molar selectivity generally falls between 92% and 96%; at 15:1, the range remains above 89% but dipropylene glycol increases. At 5:1, the monopropylene glycol selectivity drops below 80%, and the combined selectivity to dipropylene glycol and higher homologs exceeds 20%. At 3:1, dipropylene glycol formation becomes sufficiently competitive that monopropylene glycol selectivity can fall into the 52% to 65% range, and below 2:1 the product distribution shifts toward a broad oligomer mixture. These ratios refer to the total feed and assume no prior reaction; recovered water containing residual glycols reduces the effective ratio further, a point that must be incorporated into recycle-based process control.

Kinetic Consequences of a Shifting Nucleophile-to-Epoxide Population Balance

For a pseudo-homogeneous liquid-phase system, the rate of propylene oxide consumption by water is first-order in propylene oxide and first-order in water, while the rate of consumption by monopropylene glycol is first-order in propylene oxide and first-order in the terminal hydroxyl group concentration of monopropylene glycol. The selectivity at any point along the reactor is approximated by the ratio of the water-initiated rate to the sum of all hydroxyl-initiated rates. Because the nucleophilicities of water and glycol hydroxyl groups are similar, the rate constants for water, monopropylene glycol, and dipropylene glycol attack lie within roughly one order of magnitude, so selectivity follows the concentration ratio rather than a strong intrinsic rate preference. The activation energy for aqueous propylene oxide hydrolysis is often reported in the 70 kJ/mol to 85 kJ/mol range, which couples the ratio effect with a strong temperature sensitivity: a temperature increase of 10 °C at constant water-to-propylene oxide ratio raises the overall consumption rate but does not alter the concentration ratio, whereas a reduction in water-to-propylene oxide ratio at constant temperature directly increases the probability that an epoxide molecule encounters a glycol hydroxyl rather than a water molecule. Stoichiometrically, one mole of propylene oxide plus one mole of water yields one mole of monopropylene glycol; one additional mole of propylene oxide reacts with monopropylene glycol to yield one mole of dipropylene glycol. The molecular weights of monopropylene glycol, dipropylene glycol, and tripropylene glycol are 76.10 g/mol, 134.18 g/mol, and 192.26 g/mol, respectively, which are needed to convert molar yields to mass-based product distribution. This kinetic feature explains why feed ratio and reactor temperature must be controlled independently, and why the ratio set point must be verified at the reactor inlet after all recycle streams are combined.

When the Feed Ratio Falls Below Four to One in Continuous Tubular Reactor Operation

In continuous tubular reactors, a feed water-to-propylene oxide ratio below 4:1 introduces a measurable oligomerization drift and increases the viscosity of the liquid reaction mixture. The viscosity of dipropylene glycol at 25 °C is approximately 75 mPa·s to 90 mPa·s, roughly twice that of monopropylene glycol, and higher homologs raise the viscosity further. An increase in liquid viscosity reduces the Reynolds number and weakens radial mixing, which can create near-wall zones where water is depleted and propylene oxide is incorporated into heavier glycols. In a shell-and-tube reactor with 25 mm internal diameter at a bulk Reynolds number near 2000, the transition from turbulent to transitional flow caused by a moderate increase in viscosity can shift the residence time distribution and amplify the axial selectivity gradient. The operational consequence is not simply a lower monopropylene glycol yield; it is an increased risk of fouling and pressure-drop rise in the later tube sections because tetrapropylene glycol and higher homologs have lower water solubility and higher boiling points. The maximum practical tube-wall temperature must also be held below about 230 °C to limit dehydration of propylene glycol to acetol and subsequent aldehyde formation. Because the reaction is exothermic, the adiabatic temperature rise depends on feed ratio: lower water ratios have lower water sensible heat capacity and therefore can produce a higher temperature rise per percent propylene oxide conversion. This coupling requires a feed-effluent heat exchanger with a bypass or trim cooler to avoid exceeding the allowable reactor skin temperature.

The economic threshold is typically quoted at water-to-propylene oxide molar ratios between 8:1 and 12:1 for non-catalytic plants, because below that range the increased purification load for dipropylene glycol and tripropylene glycol offsets the reduced steam requirement. The steam consumption depends on the ratio of water to propylene oxide; for a feed ratio of 15:1, approximately 15 kmol of water must be separated and reheated per kilomole of propylene oxide converted, while at 8:1 the water evaporation load is only about 8 kmol per kilomole of propylene oxide converted. The trade-off is governed by the relative volatility of monopropylene glycol and dipropylene glycol, which is moderate and requires vacuum operation below 50 kPa absolute to avoid thermal degradation. Published data for specific plant energy consumption at ratios below 6:1 are limited, but process simulations indicate that the distillation load for glycol separation increases because the dipropylene glycol/monopropylene glycol cut requires more theoretical stages.

Alkaline and Acidic Catalysts: Trade-Off Between Lower Water Ratios and Corrosion or By-Product Load

Acid-catalyzed hydration of propylene oxide can achieve high propylene glycol selectivity at lower water-to-propylene oxide ratios, often in the 2:1 to 6:1 range, because the protonated epoxide is more reactive toward water and the reaction can proceed at temperatures below 100 °C. However, the use of sulfuric acid or sulfonic acid ion-exchange resins introduces sulfate residues into the crude glycol, which must be removed by neutralization or ion exchange to prevent downstream corrosion and to satisfy USP monograph and Food Chemicals Codex limits. Base-catalyzed systems exhibit faster propylene oxide consumption but may promote propylene oxide isomerization to allyl alcohol and subsequent carbonyl formation; the selectivity pattern is altered, and the water-to-propylene oxide ratio may no longer be the sole determinant of homolog distribution. In alkaline media, the pH must be maintained within a narrow range, often 8.5 to 10.5, because higher alkalinity accelerates the formation of acetol and heavy condensation products, while lower pH reduces catalyst activity. These catalytic routes have lower water evaporation loads, but they introduce operational boundaries: the feed water must have hardness below 1 mg/L as CaCO₃ to avoid scale deposition on heat-exchanger surfaces, and dissolved carbon dioxide must be removed to prevent carbonate salt formation. The corrosion rate on carbon steel in dilute sulfuric acid at 80 °C is unacceptable; therefore, acid-catalyzed reactor sections are lined with fluoropolymer or constructed from corrosion-resistant alloys such as Hastelloy C-276. Because the presence of anions in the final propylene glycol can shift the acidity value measured by ASTM D1613, neutralization and deionization are not optional.

Verification of the actual water-to-propylene oxide molar ratio in a recycle-containing process requires compositional analysis of the mixed feed rather than flow-ratio calculation alone. The feed stream typically contains fresh water, recovered water from the evaporation train, recycled monopropylene glycol, and possibly a small amount of recovered dipropylene glycol; therefore, the ratio of total water to propylene oxide can differ from the apparent ratio based on flow meters. A process gas chromatograph with a thermal conductivity detector is used to measure water, propylene oxide, and light glycols, while a flame ionization detector quantifies monopropylene glycol, dipropylene glycol, and tripropylene glycol after silylation or direct injection. The analytical method for glycol homologs is standardized as ASTM E202-18, which specifies gas chromatographic conditions and internal standard calibration for ethylene and propylene glycols, including mono-, di-, and tripropylene glycol homologs. Water content is commonly verified by volumetric Karl Fischer titration according to ASTM E203-16, and color is measured as platinum-cobalt units by ASTM D1209-19. Refractive index is a useful rapid secondary indicator, but it varies with the homolog distribution and cannot alone resolve the water-to-propylene oxide ratio. Online near-infrared analyzers are increasingly installed on the reactor feed line to predict water content and total glycol background in real time; however, their calibration must be periodically updated against gas chromatography because the absorbance of free water and hydrogen-bonded water in glycol changes with temperature and composition. A feed ratio deviation of 0.5 water-to-propylene oxide units at a set point of 15:1 is usually sufficient to move monopropylene glycol selectivity by roughly 0.3 to 0.8 percentage points depending on the baseline oligomer concentration, so analyzer repeatability is a process control constraint.

Why a Minimum Water-to-Propylene Oxide Ratio Must Be Maintained at the Reactor Inlet, Not Just in the Combined Fresh Feed

In recycle-based plants, the water-to-propylene oxide molar ratio reported by the control system is frequently calculated from the fresh water flow and fresh propylene oxide flow before the recycle water stream is mixed. This calculation becomes inaccurate when the recovered water contains residual monopropylene glycol, dipropylene glycol, or organic salts. For example, a recovered water stream containing 2 wt% monopropylene glycol and 0.5 wt% dipropylene glycol reduces the effective free water concentration and also provides additional hydroxyl groups that can react with propylene oxide. The true kinetic variable is not the total water-to-propylene oxide mass ratio but the ratio of water molar concentration to the total concentration of reactive hydroxyl groups supplied by water and glycols. If the recycle water contains 5 wt% total glycols, the effective water-to-propylene oxide ratio may be lower by several units relative to the fresh-feed ratio, and the reactor will produce a heavier homolog distribution than expected. To prevent this, high-purity distillation of recycle water is coupled with a mixed-feed analyzer that measures both water and total glycol carbon. The water-to-propylene oxide molar ratio must be measured after the recycle water injection point and after the feed preheater, because partial vaporization in the preheater can segregate water from propylene oxide and produce local ratio maldistribution. A minimum liquid-phase water-to-propylene oxide ratio of 8:1 is often specified at the reactor inlet for non-catalytic operation; below this value, the combined effects of recycle glycols and axial conversion can drive the final zone below the ratio at which oligomer formation accelerates. Pump selection also reflects this requirement: the recycle water pump and propylene oxide feed pump must be capable of maintaining the set ratio across turndown, with flow-control loops characterized by a ratio controller rather than independent flow loops, because independent loops can drift in opposite directions and produce short-term excursions.

The separation train downstream of the reactor is strongly influenced by the water-to-propylene oxide molar ratio selected upstream. At a feed ratio of 15:1, the reactor effluent contains approximately 85 wt% water, 14 wt% glycols, and 1 wt% or less light impurities; the first distillation column removes the bulk water at near-atmospheric or slight vacuum pressure, and the overhead water is returned to the reactor after polishing. The steam demand for this water removal step scales almost linearly with the water content of the effluent, so reducing the feed ratio from 15:1 to 10:1 removes approximately 5 kg less water per kilogram of propylene glycol produced, but the crude glycol then contains a higher dipropylene glycol fraction that requires additional separation capacity. Monopropylene glycol and dipropylene glycol are separated by vacuum distillation; the normal boiling point of monopropylene glycol is approximately 188 °C, while dipropylene glycol boils near 230 °C to 232 °C, but the difference is reduced at the high vacuums used to limit thermal degradation. A final monomer-grade monopropylene glycol column may operate at 10 kPa to 20 kPa absolute, with a reboiler temperature near 150 °C to 170 °C, to maintain product color below 10 Pt-Co. When the upstream water-to-propylene oxide ratio is lowered, the greater mass flow of dipropylene glycol to the downstream column raises the reboiler duty and may require additional column stages or a larger vacuum system. Published data for the exact stage count of each commercial unit is limited, but standard design practice for monopropylene glycol purification includes a light-ends column, a water column, a product column, and an oligomer recovery column. The selection of the water-to-propylene oxide ratio therefore determines not only reactor selectivity but also the reboiler duties, column diameters, and vacuum condenser loads across the entire purification train.

Effect of Water-to-Propylene Oxide Molar Ratio on Carbonyl By-Product Formation and Wastewater Load

At lower water-to-propylene oxide molar ratios, the increased concentration of glycol hydroxyls favors oligomerization, but the same conditions also increase the residence time of monopropylene glycol in the reactor and downstream hot equipment, which can promote thermal dehydration to acetol and oxidation to hydroxyacetone. Acetol and propionaldehyde are the main carbonyl impurities that affect propylene glycol odor and ultraviolet absorbance. The carbonyl content is typically expressed as milligrams of carbonyl per kilogram of sample and is measured by a colorimetric method after derivatization with 2,4-dinitrophenylhydrazine; specifications for the pharmaceutical-grade product may require carbonyl values below 20 mg/kg as acetaldehyde. A water-to-propylene oxide ratio below 4:1 can raise the carbonyl load because the lower water concentration reduces the dilution of acidic or oxidized species and because the heavier glycol stream must be held at higher reboiler temperatures during recovery. The wastewater load is also affected: high-ratio processes generate a large volume of clean water effluent or require extensive water recycle, while low-ratio processes produce a more concentrated organic stream that may require biological treatment with a longer hydraulic retention time. The chemical oxygen demand of the wastewater is dominated by residual glycols and acetol. For a facility operating at 15:1, the wastewater generation rate is typically quoted in the range of 0.5 m³ to 1.2 m³ per tonne of produced propylene glycol after water recycle, whereas published data for specific low-ratio facilities is limited. The exact allocation depends on cooling-tower blowdown, ion-exchange regeneration, and distillation column condensate reuse. Regardless of the ratio, the plant must ensure that recovered water streams are monitored for total organic carbon and for iron and chloride, because iron catalyzes oxidative degradation and chloride accelerates stress corrosion cracking in stainless steel distillation columns.

Ratio-Dependent Selectivity Models in Modern Distillation-Integrated Hydrolysis Units

Modern propylene glycol plants use a ratio-dependent selectivity model based on kinetic rate expressions and online gas chromatograph feedback. The model calculates the expected molar distribution of mono-, di-, and tripropylene glycol from the mixed-feed water-to-propylene oxide ratio, reactor outlet temperature, and residence time. The signal from the mixed-feed analyzer is used to manipulate the fresh water flow or recycle water flow. In practice, the ratio controller is constrained by the minimum reactor-inlet ratio and by the maximum water content that can be handled by the first distillation column. A typical operating window for non-catalytic propylene oxide hydrolysis is a water-to-propylene oxide molar ratio of 12:1 to 20:1, with a reactor outlet temperature of 200 °C to 215 °C and a pressure high enough to maintain liquid phase at the outlet temperature. Within this window, the monopropylene glycol selectivity can be held within a narrow range, and the dipropylene glycol selectivity is controlled to match downstream product demand. When lower-ratio operation is necessary to increase dipropylene glycol or tripropylene glycol production, the reactor integrity and downstream vacuum columns must be reassessed. The water-to-propylene oxide molar ratio should not be reduced below the point at which the predicted local ratio in the final reactor zone falls below 5:1, because below that local ratio the calculated net rate of tripropylene glycol formation becomes significant and the pressure drop in the reactor begins to increase as the heavier homologs accumulate. This operating boundary is based on kinetic modeling and is consistent with the observed behavior of continuous plug-flow reactors; however, published plant-specific data at local ratios below 5:1 is limited because most commercial units do not intentionally operate in that regime. The ratio set point is therefore a compromise among reactor selectivity, distillation energy, product demand for higher homologs, and the risk of fouling.