CS252452B2 - Process for cleaning industrial gas - Google Patents

Abstract

A method for purifying industrial gas such as hydrogen, carbon monoxide, air, oxygen, helium, argon, monoolefins with 2 to 5 carbon atoms in the molecule, diolefins with 4 to 5 carbon atoms in the molecule, paraffins or acetylenes, by removing trace amounts of one or more impurities from the group comprising water, hydrogen sulfide, carbon dioxide, carbonyl sulfide, sulfur dioxide, hydrogen chloride, hydrogen cyanide, nitric acid and mercaptans, in which the purified gas is brought into contact at a temperature of 15 to 100 °C with a liquid, anhydrous solution containing 0.5 to 15% by weight of an alkali metal or alkaline earth metal, based on its hydroxide or weak acid salt, in a liquid aliphatic polyhydric alcohol with a carbon to oxygen ratio of 1 to 5 and with at least two oxygen atoms separated by at most two subsequent carbon atoms.

Description

(54) Industrial gas purification method

A method for purifying industrial gas such as hydrogen, carbon monoxide, air, oxygen, helium, argon, monoolefins with 2 to 5 carbon atoms in the molecule, diolefins with 4 to 5 carbon atoms in the molecule, paraffins or acetylenes, by removing trace amounts of one or more impurities from the group comprising water, hydrogen sulfide, carbon dioxide, carbonyl sulfide, sulfur dioxide, hydrogen chloride, hydrogen cyanide, nitric acid and mercaptans, in which the purified gas is brought into contact at a temperature of up to 100 °C with a liquid, anhydrous solution containing 0.5 to 15% by weight of an alkali metal or alkaline earth metal, based on its hydroxide or weak acid salt, in a liquid aliphatic polyhydric alcohol with a carbon to oxygen ratio of 1 to 5 and with at least two oxygen atoms separated by at most two subsequent carbon atoms.

The invention relates to a method for purifying industrial gases and in particular to removing minute amounts of impurities from the gases being purified.

The reaction of basic alkali metals or alkaline earth metals as such or in the form of hydroxides or salts of weak acids is known in the art of chemistry. The chemistry of these reactions is the basis of important industrial processes. For example, in the field of purification of industrial gases such as hydrogen, carbon monoxide, air, oxygen, nitrogen, argon, helium, lower olefins, lower diolefins and similar gases, layers or suspensions of solid particles of sodium or potassium hydroxide are used to remove large quantities of acid gases. However, these processes are relatively ineffective when it comes to removing trace amounts of impurities such as hydrogen sulfide, carbon dioxide, carbonyl sulfide, sulfur dioxide and mercaptans and similar impurities from the gases. Alkali metal hydroxides have been considered suitable absorbents for the separation of acetylene from ethylene, as is evident from U.S. Patent No. 2,742,517.

U.S. Patent No. 2,185,332 (issued January 2, 1940) describes a method for dehydrating and cleaning refrigeration systems, based on the introduction of an alkali metal alcoholate, particularly sodium methylate, into the refrigeration system. The alcoholate is introduced into the system to remove water by double decomposition, which produces an alkali metal hydroxide, such as sodium hydroxide. The sodium hydroxide then remains in the system. The consequence is obvious. Caustic soda is a corrosive medium and its presence in the refrigerant gas cannot be favorable for long-term operation. In addition, the patent describes the removal of acids as such.

For example, hydrogen chloride, probably in the form of hydrochloric acid, is converted to sodium chloride, but it is certainly not advantageous to leave the sodium chloride in the cooling gas. Thus, according to the patent, a known chemical reaction is used to solve one problem, but another problem is created.

U.S. Patent No. 2,177,068, issued October 24, 1939, describes a method for treating natural gas to remove water and acid gases. It uses polyhydric alcohols together with amines, such as alkanolamine, to remove water and acid gases. This process is based on the old knowledge of using amines to remove acid gases and polyols to remove water.

According to the above-mentioned patents, the gases to be purified are saturated hydrocarbons, the inertness of which is well known. The impurities to be removed do not adversely affect the function of these gases, but they do adversely affect the environment.

The use of amines for the removal of acid gases according to the above-mentioned patent specification has the disadvantage that the amine systems are corrosive.

The development of industrial gas purification is related to the improvement of methods for determining impurities in gases. Improved analytical methods for evaluating industrial gases have created the need for more advanced processes for removing newly detectable impurities.

The term non-reactive and inert gases used means industrial gases that can be processed by the method according to the invention because they do not react with selected solutions of alkali metal and alkaline earth metal compounds and polyhydric alcohol defined below.

It is generally recognized that olefins, due to their production, storage and/or handling, may contain trace amounts of water vapor, hydrogen sulfide, carbon dioxide, carbonyl sulfide, sulfur dioxide and mercaptans in the order of ppm. Many syntheses require an olefin having below 0.5 ppm of each of these impurities.

For example, to produce ethylene of suitable quality for the production of certain polyethylenes, the small concentration of carbon dioxide, about 10 to 25 ppm, in the olefin has to be reduced to less than 1 ppm. At present, sodium hydroxide pellets are used for this purpose. The main problems with this method, however, are that only about 3% of the sodium hydroxide contained in the bed is converted to sodium carbonate according to the equation:

NaOH + CO _2? = Na ₂ CO ₃ + H ₂ 0 and that after the surface of the sodium hydroxide pellets is converted to sodium carbonate, the released water causes agglomeration of particles, forming channels, and the layer solidifies.

The fact that the entire layer solidifies is a particularly difficult problem, as at present this layer has to be removed manually. A sodium hydroxide solution has been proposed as an alternative solution. However, the disadvantage is that water vapour from the solution is introduced into the ethylene and therefore zeolite layers must be used to absorb the water: however, this layer is expensive to install and its regeneration is costly.

The process according to the invention uses a liquid that is essentially free of water, which therefore does not increase the water content of the gas being treated. The treatment liquid has a high capacity to accept alkali metals or alkaline earth metals, a very low vapor pressure, forms a chemically stable system with the alkali metals or alkaline earth metals, and has a sufficiently low viscosity to allow good wetting of the gas being treated.

According to the invention, an alkali metal or alkaline earth metal compound and a liquid polyhydric alcohol in solution form are used to remove the above-mentioned impurities contained in trace amounts and in particular to simultaneously remove any impurities contained in trace amounts. The method according to the invention is therefore not aimed at removing one specific impurity, as has been customary up to now.

The subject of the invention is therefore a method for purifying industrial gas such as hydrogen, carbon monoxide, air, oxygen, helium, argon, monoolefins with 2 to 5 carbon atoms in the molecule, diolefins with 4 to 5 carbon atoms in the molecule, paraffins or acetylenes by removing trace amounts of one or more impurities from the group comprising water, hydrogen sulfide, carbon dioxide, carbonyl sulfide, sulfur dioxide, hydrogen chloride, hydrogen cyanide, nitric acid and mercaptans, which is characterized in that the purified gas is brought into contact at a temperature of 15 to 100 °C with a liquid, anhydrous solution containing by weight 0.5 to 15% of an alkali metal or alkaline earth metal, based on its hydroxide or salt of a weak acid, in a liquid aliphatic polyhydric alcohol with a ratio of carbon atoms to oxygen atoms of 1 to 5 and with at least two oxygen atoms separated by no more than two the following carbon atoms.

In the process according to the invention, the industrial gas, as defined above, is continuously introduced into the described solution of the alkali metal or alkaline earth metal compound and polyhydric alcohol, maintained at a temperature of about 15 to 100 ° C and the gas is removed from the solution into an environment whose partial pressure prevents an increase in the moisture content in the gas being treated. In a preferred embodiment of the process according to the invention, the water concentration in the gas removed from the solution and in the environment into which the gas is introduced is less than the water concentration in the gas introduced into the solution.

The term "alkali metal and alkaline earth metal" here always refers to elements of groups IA and IIA of the periodic table of elements; in particular, it refers to lithium, sodium, potassium, rubidium, cesium, magnesium, calcium, strontium and barium. Preferred alkali metals are sodium and potassium, with potassium being particularly preferred for the process according to the invention.

For simplicity, the method according to the invention is described using the example of removing trace amounts of impurities from ethylene gas.

The alkaline reaction of a polyhydric alcohol solution is formed by hydroxides or salts of weak inorganic acids and alkali metals or alkaline earth metals.

Polyhydric alcohols include glycerin, alkylene glycols, such as alkylene glycols with up to 12 carbon atoms, for example, ethylene glycol, diethylene glycol, triethylene glycol, tetraethylene glycol, propylene glycol, dipropylene glycol, 1,2-butylene glycol, di-1,2-butylene glycol, and alkylene oxide adducts prepared by reacting an alkylene oxide with a polyol, such as the above-mentioned glycols, glycerin, and other triols, such as 1,2,6-hexanetriol, erythritol monomethyl ether, and pentaerythritol monomethyl ether. Ethylene oxide and propylene oxide adducts of polyols are preferred, with ethylene oxide adducts alone or reacted with propylene oxide being most preferred.

The method according to the invention is suitable for purifying a gas which contains as its main component an inert gas which therefore does not react with a liquid solution of an alkali metal or alkaline earth metal compound and a polyhydric alcohol selected for this purpose, the purified gas containing in trace amounts another component which chemically reacts with this liquid solution of an alkali metal or alkaline earth metal compound and a polyhydric alcohol. Examples of non-reactive components of the purified gas are: hydrogen, carbon monoxide, air, oxygen, nitrogen, helium, argon, lower monoolefins, lower diolefins and their derivatives, paraffins such as methane, ethane and similar paraffins and acetylenes. Olefins here always mean olefins selected from the group comprising monoolefins with 2 to 5 carbon atoms, diolefins with 4 or 5 carbon atoms and their derivatives. Such typical olefins are ethylene, propylene, butylenes, pentylenes, 1,4-pentadiene, 1,3-butadiene and their derivatives.

The reactive components of the gases to be purified are primarily those which form acids in aqueous solution. These substances react with the alkali metal or alkaline earth metal compound in the solvent to form inorganic or organic salts which can be removed from the gases.

Examples of acidic or reactive components of the gases being purified include hydrogen sulfide, carbon dioxide, carbonyl sulfide, sulfur dioxide, mercaptans, hydrogen chloride, hydrogen cyanide, nitric oxide, and similar impurities.

When carrying out the method according to the invention, it was found that water vapor can be simultaneously removed from the purified gas, even if it is contained in a larger amount than what could be described as trace amounts. The water vapor is removed by the action of a solution of alkali metal or alkaline earth metal compounds and a polyhydric alcohol, probably according to a completely different mechanism than the trace amounts of reactive components of the purified gas. The mechanism of water vapor removal is probably based on the ability of polyhydric alcohols to absorb moisture, while the ability of the solution based on double decomposition can also be additionally applied:

RONa + _H20 ' ROH + NaOH and

Na ₂ CO ₃ + H ₂ 0 _s ·- * 2 NaOH + C0 ₂

Additionally, it has been found that water can be removed from industrial gases by a solution of an alkali metal compound and an alkaline earth metal and a polyhydric alcohol, wherein the water can be contained up to the saturation of the respective gas (up to the dew point). The hydroscopic ability of the polyhydric alcohol, which induces the absorption of water by the hydrogen bonding mechanism, does not reduce the capacity of the solution of the alkali metal compound or alkaline earth metal and a polyhydric alcohol to react with other impurities mentioned above, contained in trace amounts.

Therefore, any amount of water vapor contained in the non-reactive purified gas, introduced into the non-reactive gas before or during its production or processing, is removed by the polyhydric alcohol and the alkali metal or alkaline earth metal compound of the absorption solution. Similarly, any water of reaction formed in situ is also removed by the components of the alkaline solution.

Therefore, the absorption solution is referred to herein as a substantially non-aqueous solution which, according to the invention, removes trace amounts of reactive impurities from the non-reactive purified gas in the purification of gases. It is recalled that the partial pressure of water maintained above the absorption solution is lower than that of the non-reactive purified gas, so that water vapor is not entrained into the non-reactive gas.

Typically, the non-aqueous absorption solutions of the invention are prepared by adding an alkali metal or alkaline earth metal compound to a solution of a selected liquid polyhydric alcohol, such as glycerin and the glycols mentioned above. This addition to the polyhydric alcohol in solution form can be carried out by directly adding the alkali metal or alkaline earth metal in the form of their hydroxides, carbonates or bicarbonates or other salts of similar inorganic weak acids. The metal or hydroxide or salt can react to form a polyhydric alcoholate of the alkali metal or alkaline earth metal. It is possible that the carbonate form is first hydrolyzed with water to form the hydroxide before the alcoholate is formed.

Solutions of the alkali compound and the polyhydric alcohol are prepared by simply mixing the alkali component with the polyhydric alcohol component until a solution of both components is formed. This mixing can be carried out at room temperature or at an elevated temperature, for example at the boiling point of the polyhydric alcohol used. As a rule, the two components are mixed at a temperature of about 30 to 250 ° C, preferably at a temperature of about 40 to 220 ° C.

The alkaline component can be introduced into the polyhydric alcohol in the form of hydroxide and in the form of a salt of weak inorganic acids, for example carbonic acid. Preferably, the alkali metal or alkaline earth metal is introduced in the form of hydroxide, carbonate and/or bicarbonate.

As will be discussed below, the greatest efficiency in removing impurities is achieved when the alkali metal or alkaline earth metal is used in the form of a hydroxide, but the most stable system is achieved when a carbonate is used. Examples of suitable alkali metal or alkaline earth metal compounds include lithium hydroxide, lithium carbonate or bicarbonate, sodium hydroxide, carbonate or bicarbonate, potassium hydroxide, carbonate or bicarbonate, rubidium hydroxide, carbonate or bicarbonate, cesium hydroxide, carbonate or bicarbonate, magnesium hydroxide, calcium hydroxide, strontium hydroxide or bicarbonate, and magnesium hydroxide, calcium hydroxide, strontium hydroxide or barium hydroxide, carbonate or bicarbonate.

According to the invention, 0.5 to 15% by weight of an alkali metal or alkaline earth metal is used to form an absorption solution with a sufficient concentration of alcoholate after reaction with a small proportion of a polyhydric alcohol, for example glycerin or a selected glycol.

To determine the technological values and to study the operating characteristics of the system according to the invention, a high-pressure pilot plant is constructed. For this purpose, a glass column with an internal diameter of 28 mm and 20 trays is used. The glass column is incorporated into a pressure unit for high-pressure operations. The introduced gas - ethylene is preheated to the desired column temperature in a shell and tube heat exchanger. Another heat exchanger is used between the circulation pump and the column to control the temperature of the liquid introduced in the column cover.

At a temperature of 50 to 75 °C, sodium hydroxide pellets are stirred in glycol while purging with nitrogen: 3 hours are sufficient to completely dissolve the pellets. The solution is then transferred to the reservoir part of a glass column for water stripping. The solution is allowed to circulate at a temperature of 50 °C and at an overpressure of 345 kPa with a countercurrent flow of nitrogen. Stripping is carried out until the water concentration in the washing nitrogen drops below 10 ppm.

The glass column used has three simple taps for analyzing the gas introduced into the column and discharged from the column. The gas samples are continuously determined for carbon dioxide and water vapor. Liquid samples are also taken periodically for viscosity determination.

The column is tested under various technological conditions:

Solvents ethylene glycol

Reactants ⁺ triethylene glycol (TEG) feed gas

2% or 5% sodium hydroxide 2% pyrogallol

2% manganese chloride nitrogen or ethylene

Overpressure in the column, kPa Temperature in the column, °C 690 to 2 38 to 60 518.5 Weight ratio of liquid to gas 0.023 to 0.167 Gas velocity, m/s Carbon dioxide concentration in the 0.045 to 0.09 input gas, ppm Water concentration in the input gas- 1 to 110 nu , ppm Oxygen concentration in the inlet 1 to 1,000 gas, ppm Hydrogen sulfide concentration in the introduced 0 to 50 gas, ppm 0 to 20

% by weight of the reactant with respect to the weight of the entire solution

For good gas-liquid contact and for stable operation, the gas velocity should not exceed the column fouling velocity. The column fouling velocity is calculated for 5% sodium hydroxide and triethylene glycol solutions according to Table I:

Table I

Column packing rate using 5% sodium hydroxide solution and triethylene glycol, ethylene as gas and a column diameter of 28 mm Weight ratio Overpressure in the column Temperature The rate at which a liquid cools fluid flow in a column with a column against the rising to gas gas ⁺ kPa °c m/s 0.1 2,070 20 0.324 0.2 2,070 20 0.291 0.1 2,760 20 0.339 0.2 2,760 20 0.315 0.1 103.5 75 1,500 0.2 103.5 75 1,350 0.1 793.5 40 0.540 ⁺ Based on the entire cross-sectional area columns For pressure column 1,725 kPa calculated overflow speed about 0.3 m/s.

The value given represents the most common industrial operations with aqueous solutions and therefore does not include the viscosity of the absorption liquid. However, the viscosity of the absorption liquid when using caustic soda and glycol must be taken into account.

The results of the carbon dioxide removal tests are shown in Table II.

O

XD

P α

with

XD and

xn e Ό Ci xd CM >

+ + + + k© m or><NCNOOOO + m ^fOrNOmfNCNi—(CNf)KO<OOi—

OťNrHrHOOOOOOOOrHfOiHO

X o

Φ ϋ

«3

P

P (3

Φ

At α

oh my

G G XD řd Ό rH Xd CM >

o o in in Tť CO Tť co in in cm cn •m* rH o

IT)

Φ >ϋ o

CM

P

Φ

P cd

CM

O O O O O CN CN <N CN CN o o rH i—f

0 'řd 'řd 'řd >< >1 'id 'řd '>Ί P X 44 44 44 44 X X 44 44 'řd 'řd 'řd 'id 'řd 'řd 0 0 0 0 0 0 0 0 0 44 44 44 44 44 44 > « W ω w W W d d d N N N N N N X rd id id id rd rd >1 >1 rd M M M M M M >1 0 > > ř> > > > > > > (3 G G G G G

Carbon dioxide removal using glycol and sodium hydroxide solution β

>1 day

p c + ><1) -d +

S H o d CM CM

CMO

Φ

Ch

Xd

r—I d

P CM Φ X >P CM

Ό •H ·

X P

O 'řd Q MCE Ό Ό «5 >i 0

W W * >Φ

P

Xfl

CM •cp CO 00 cn ^poococofo^^mk©

M3 cn cn kO kO k£> k© o o o σ> ro in cn m co m o o o in in m in in in in k© k© 00 CO σ\σϊσ>σ»σισ>σϊσ>σ\οοα)ΐηιηΗιΗ kOkOkOkOkOkCkOkOkOCOfOkDkOiniO unincNCNinininininininininininininin

about about about about about about about about about about about About About About about w « w w w w w w w w w W W w ω E-t E-t E-t EH E-t E-t E-t EH Ed E-t E-t

•3 G M >P P 3 Λ 0 Ό About CM XD G XD τ—l G xd 0 3 rH G 0 •H 43 44 <H P 0 0 G xd > 44 0 44 rd 44 P 3 rH 0 G rH Cn rd G rd 0 G rH Ό rH 44 Φ Φ Φ CM >1 rH >N •r*i rH rd >Φ P Cn 43 43 43 d G P ϋ 0 Φ Φ rd M <H rH •H 44 G 43 rd Φ > P □ 43 P Xd d rd P P Ό 0 P 0 G xd XD P xd xd G P 0 G G Φ Φ with <D Φ g P ,2 g With d X d d G >Φ φ G G N with > N N About + + CM about H + CC w Ed

257452

Based on the results shown in the table, it can be stated: Both the mixture of ethylene glycol and sodium hydroxide and the mixture of triethylene glycol and sodium hydroxide are very effective in removing traces of carbon dioxide: triethylene glycol is, however, preferable due to its lower vapor pressure. The concentration of carbon dioxide in ethylene can be reduced to less than 0.5 ppm in a 10-tray column provided that good gas-liquid contact is maintained. Even a low liquid to gas ratio, for example 0.05, is sufficient for effective removal in triethylene glycol, which has good fluidity at 40 to 50°C with very good plate efficiency. The viscosity of this solution at 40°C is 110 mPa. After three weeks of continuous operation with a single dose of 5% sodium hydroxide solution in triethylene glycol, no increase in viscosity is observed, even though the solution contains sodium carbonate as a by-product. No solid sodium carbonate deposition is evident in the vessel, but some solids remain on the filters after the liquid has been filtered. The rate of sodium carbonate deposition is very slow. The removal of carbon dioxide is equally effective both before and after the water has been removed by stripping the solution.

A single batch of 5% solution (percentages are still by weight unless otherwise stated) of sodium hydroxide in triethylene glycol will last for three weeks until the concentration of carbon dioxide in the column exhaust gas increases due to the decrease in sodium hydroxide content. To determine whether the decrease in carbon dioxide removal efficiency is due to a lack of free water in the system or to a decrease in sodium hydroxide content, 0.75 liters of triethylene glycol containing 0.1% water is added to the liquid reservoir and the column is restarted. No improvement in carbon dioxide removal is observed, indicating that the loss in column efficiency is due to the decrease in sodium hydroxide content.

Three samples of the liquid were taken for viscosity measurement during the continuous operation. No changes in viscosity were observed, indicating that dimerization and oligomerization reactions were insignificant.

The solution is filtered to remove the solid sodium carbonate, although no significant settling occurs in the tank. The observed settling is consistent with the calculation of settling using the Stokes equation. For example, these calculations give a settling rate of 0.0017 m/day for sodium carbonate particles of one micrometer diameter in a 5% sodium hydroxide solution in triethylene glycol at 40°C. Scanning electron microscopy measurements show that the sodium carbonate particles are 1 to 10 micrometers in size, with most particles being in the lower end of this range. Therefore, filtration may be necessary to remove solids.

In addition to using sodium and lithium as alkali metals for in situ reaction with a selected glycol to form a non-aqueous absorption solution, tests were conducted to determine the effectiveness of various alkaline earth hydroxides for the preparation of glycolates. Table III shows the addition of equivalent percentages of strontium, barium, and calcium hydroxides to diethylene glycol and the carbon dioxide reduction and theoretical percentage removal of carbon dioxide during a nine-cycle regeneration using these alkaline earth metal hydroxides (Table IV).

Figure 3 shows the efficiency graphs of non-aqueous reversible carbon dioxide removal for barium, calcium, potassium and sodium based systems over nine regeneration cycles. The sodium based system represents a highly efficient process over all nine cycles.

Table III

Initial carbon dioxide content in various glycolates (% are always by weight)

Solvent Solvent g CO 9 binding 5% solution of SrlOHJj in diethylene glycol 125 1.95 5% solution of Ba(OH) ₂ in diethylene glycol 125 2.80 5% solution of Ca(OH)2 in diethylene glycol 125 4.50

Conditions: room temperature, carbon dioxide pressure equal to ambient pressure

Table IV

Regeneration of alkaline earth metal glycolates (% by weight)

Solvent Regeneration cycle

Decrease in C0% theoretical removal of CO5% solution of Sr(OH) ₂ in diethylene glycol

5% solution of Ba(OH)2 in diethylene glycol

5% solution of Ca(OH)2 in diethylene glycol

g % 0.8 44 0.3 17 0.5 28 0.9 50 0.4 22 0.4 22 0.4 22 0.3 17 0.35 19 0.5 39 0/5 39 0.35 27 0.45 35 0.4 31 0.5 39 0.5 39 0.45 35 0.5 39 1.5 51 1.1 37 1.2 40 1.05 35 1.0 34 1.2 40 1.0 34 0.9 30 0.95 32 at 0 pressure, .1 MPa and for

Conditions: carbon dioxide during the preparation of the test solution was introduced at a pressure of 0.1 MPa and at room temperature, regeneration at a temperature of 200 °C, solvent weight 125 g.

The method according to the invention is illustrated by the attached drawings, which are described in detail below:

Fig. 1 is a simplified flow diagram of a device for carrying out gas purification according to the invention, Fig. 2 is a flow diagram of a similar device to Fig. 1 having additional means for regenerating the non-aqueous absorption solution and for recycling the regenerated solution to the liquid-gas contactor for further use;

Fig. 3 curve of carbon dioxide removal efficiency for the number of thermal regeneration cycles, with the thermal regeneration cycles plotted on the x-axis and the carbon dioxide removal efficiency on the y-axis, and curve 1 applies to the 7% by weight potassium carbonate/triethylene glycol system, curve 2 to the 5% by weight sodium hydroxide/diethylene glycol system, curve 3 to the 5% by weight calcium hydroxide/diethylene glycol system and curve 4 to the 5% by weight barium hydroxide/diethylene glycol system;

Fig. 4 simplified flow diagram of a pilot plant with continuous gas purification and regeneration of the absorbent, where no metal carbonate precipitate is formed, Fig. 5 simplified flow diagram of a pilot plant with continuous gas purification and cyclical regeneration of the absorbent, where an insoluble metal carbonate precipitate is formed.

According to Figs. 1, 2, 4 and 5, the purified gas containing impurities is introduced into the gas-liquid contactor 2 via a line 2, from which the purified gas is discharged via a line 2 to a device 4 for removing traces of the non-aqueous solution, from which the purified gas is discharged via a line 2. According to Figs. 2 and 4, the non-aqueous solution removed ^from the gas in the device 2 via a line 9 is introduced into a line 6,_. by which, according to Fig. 1 and 2, a non-aqueous treatment solution is introduced into the contactor 2. For the purpose of regeneration, the used solution can be discharged from the contactor 2 according to Fig. 1 and 2 through a pipe 7_ via a pump 2 into a stripping column 10, into which nitrogen is optionally introduced according to Fig. 2 through a pipe 11, whereby, according to Fig. 1, absorbed solid impurities are removed from the stripping column 10 through a pipe 12, the waste solution through a pipe 13, according to Fig. 2, gaseous impurities through a pipe 14, and according to Fig. 1 and 2, the purified solution is returned via a pump 2 to a pipe 6, through which, according to Fig. 1 and 2, a non-aqueous treatment solution is introduced into the contactor 2. According to Fig. 2, this purified solution is introduced into a cooler 16 through a pipe 2.

The individual reference numerals in Fig. 4 and 5 have the same meaning as in Fig. 1 and 2. According to Fig. 4, the used solution from the contactor 2 is introduced via the pump 2 through the pipe 17 through the valve 16 into the stripping column 12, into which hot steam is introduced via the pipe 18 or cooling water via the pipe 19, as required, while both media, both steam and water, are discharged via the pipe 21.

Carbon dioxide and water vapor are removed from the stripping column 10 through line 22, while organic liquid, for example triethylene glycol, is added through line 23 and alkali, for example potassium carbonate, is added through line 24. The purified and supplemented non-aqueous treatment solution is then fed through pumps 25 through line 26 to a reservoir 27 of regenerated non-aqueous treatment solution and from there it is returned through line 28 through valve 29 after mixing with the non-aqueous solution from the device 2 to remove traces of non-aqueous solution from the purified gas from the contactor 2_,

According to Fig. 5, the spent solution is discharged from the contactor 2 through the line T_ and is passed through a filter 30, from which the metal carbonate precipitate is discharged through the line 31 and the spent solution freed from solids is discharged through the line 32 through the valve 43 into the stripping column 10 equipped with a stirrer 34, through which steam is passed through the line 18, nitrogen is introduced through the line 35 and sodium hydroxide and triethylene glycol are added through the line 36. The gaseous products are discharged from the stripping column 10 through the line 37, while the purified non-aqueous conditioning solution is introduced through the line 38 through the pump 39 through the line 40 equipped with the valve 21 through the pump 33 through the line 42 into the gas-liquid contactor 2.

As is apparent, the contactor 2 may be a vessel filled with sodium hydroxide in the form of an extraction solution into which the gas to be purified is introduced. In the process according to the invention, the ethylene introduced, for example, contains as impurities carbon dioxide, water, sulfur dioxide, hydrogen sulfide, carbonyl sulfide and mercaptans: it is mixed with a non-aqueous solution in a gas-liquid contactor 2 containing at least one reaction zone. The impurities are removed from the gas by reaction with the components of the absorption solution or by dissolution in the solution. Optionally, the purified gas leaving the gas-liquid contactor 2 may be further purified, for example by passage through a selective absorption layer in the device 2 N ° <1removal of traces of the non-aqueous solution.

Depending on the concentration of impurities in the feed gas and the amount of gas to be purified, the solution may be regenerated intermittently or continuously by removing solid reaction products, purifying the spent solution, and/or separating the spent solution and absorbed impurities from the active solution.

To compensate for losses of solution: its entrainment or evaporation into the purified gas, a conditioning non-aqueous solution is added to the process.

A liquid droplet trap may be located at the top of the column. The ethylene gas leaves this device and passes through a filter to remove submicron glycol particles. The retained liquid may be returned to the stripping column or, at low flow rates, may be collected and periodically discharged. Downstream of the filter, an activated carbon trap may be provided to remove the last traces of triethylene glycol from the ethylene. At a temperature of 50 °C and a pressure of 3105 kPa, the concentration of triethylene glycol in the ethylene will be less than 0.1 ppm.

If removal of triethylene glycol is desired below this limit, adsorption on activated carbon is used. Activated carbon is expected to accept more than 20% of the triethylene glycol. With this adsorption capacity, 362 kg of activated carbon is sufficient to completely remove all traces of triethylene glycol from ethylene from a column processing 18,120 kg of ethylene per year (8,000 hours) of operation. Therefore, the used activated carbon can simply be removed from the process and does not need to be regenerated for economic reasons.

Two filters can be used to remove solid sodium carbonate from the solution. The sodium hydroxide solution in glycol leaving the filters is divided into two streams. Approximately 20% of the solution is pressure-reduced to 130 kPa and fed to the stripping column, and the remaining 80% is returned upward to the absorber using a circulation pump. For a column processing 18,120 kg.h of ethylene, at least 37.8 l of liquid should be used per minute (liquid to gas ratio 0.125).

The stripping column operates at a pressure of 68.7 to 137.4 kPa, at a temperature of 100 °C, and the column has two main tasks: to reduce the moisture content in fresh sodium hydroxide in glycol solutions to less than 10 ppm in the gas phase: this is achieved by purging with nitrogen at 100 °C intermittently as sodium hydroxide is added to the system. The second task is to remove moisture adsorbed by the circulating solution by continuously purging with nitrogen at 100 °C. In doing so, the sodium hydroxide in glycol solution removes carbon dioxide by chemical reaction and water by physical adsorption.

With a low moisture content in the ethylene feed (approximately 5 ppm), only a small portion of the circulating solution needs to be desorbed. The distillation unit should be sized for 456.6 liters of 10% sodium hydroxide solution in triethylene glycol.

Is the stripping column equipped with a turbine-type stirrer to dissolve solid sodium hydroxide in the glycol solution? Once the sodium hydroxide has been removed from the system by reaction with carbon dioxide, sodium hydroxide must be added to the system. To add sodium hydroxide, the column is equalized to ambient pressure, the column is flushed with nitrogen, opened, and sodium hydroxide flakes are added to provide a solution containing, for example, 15% sodium hydroxide. When the moisture content of the purge nitrogen is below 10 ppm, the concentrated product is gradually mixed with the circulating solution. During the addition of solid sodium hydroxide and the removal of moisture, all liquid from the filters is returned to the column and no side streams are discharged to the stripper.

The heat source for the stripping column can be low-pressure steam conducted through a jacket. It must be capable of heating the solution to 150 °C, although 100 °C is sufficient for moisture removal.

The following Table V shows typical examples of the practical implementation of the process according to the invention using various feed gases, sodium and potassium based absorption systems, using various flow rates and operating pressures, x

From the examples given, it is clear that in the method according to the invention, a glycol and glycolate solution, as the main part of the absorbent, is used to purify gas, nitrogen or ethylene, by reducing the content of impurities such as carbon dioxide, carbonyl sulfide, and hydrogen sulfide from a relatively high concentration to a concentration of a few ppm.

242452 +J

O +J tn •H >O

0) and

cl

CL

O

About ϋ

X4 (0 μ

O

4J

0 >%rH rH I 43 CL 43 υ cn >1 3 e 2 34 2

and 3 3 3 3 3 about 0 0 0 About 0 34 44 44 £ 44 34 34 >N XM >N W >N >N >N XQ xd xd and and and and and xd Xd xd CL >1 CL >1 CL ί>1 and 43 43 43 43 43 CL >1 CL &H CL φ cj φ th Φ and 43 10 that that that 10 Φ and Φ and Φ >M r—| >μ rH >μ 1—1 » μ μ μ μ μ >μ ι—1 >μ rH >μ & > CL > CL > Φ + Φ + Φ Φ + Φ + <u + CL > CL fc> CL Ό + Ό + Ό Ό + Ό + Ό + tn xu tn XI) tn Xl, r-1 + Η -Ι- rH rH + rH + rH -Ι- tn XI) <n xp tn > > > About M O J H O μ About μ About μ The > t> Φ about Φ About Φ 0 Φ < At Φ φ φ φ Φ 0 Φ 0 Φ at rH υ i—1 at rH and -P and μ and +» and 44 and 4J and -Μ at rH ϋ rH υ *rl in •r| 0 •rl here and and and and and and and and and and and and •rl Φ •rl Φ •r| jq at 43 ϋ 43 at About CL About CL CL about CL about Λ About CL 43 About 43 About 43 about 3 0 fl 0 r—1 rH rH rH rH rH 3 0 3 0 3 μ μ μ 0 o 0 About 0 about About about 0 about About μ μ μ 44 N 44 N 44 N 34 CM rH 44 rH 34 CM 34 CM 34 rH •μ N P N 44

oh oh

>N (0

CM

O

O >N and

CM

About

>N <0 (*)

About rH rH ι-Η fO

In in in tn tn

CM rH rH

In « in

Γ* >1 and

XU >

about RH

Φ

O

About

vo tn o in γμ vo rtn rtn rtn tn r- rtn tn in

CM CM CM

CM VO r> tn vo >

about about

XT vo vo tn cn C' o o x* vo vo in n t·· o o

>1 rH ** xr xr CL CM CM CM CM CM CM CM CM 2 2 2 2 2 2 2 2 2 2 2 CM CM CM '3ί 3 •«In \ •p. ***. ie At About 3 >1 CM CM CM CM CM CM CM CM -*» ***. p. >Φ rH About About About About About About About About CM tn tn Ό CL at at at About υ About υ At About about CM xd \ About about 2 > g xí« xr about about about about about cn rn cn and 2j about about about about about about about about about about about N CL rH rH rH tn tn tn m m rH rH rH

rH 0 34 >x rH 0» t—1 and 0 rH rH φ 34 0 0 rH >1 34 34 >1 rH >1 ►i 43 θ' rH rH 44 and 0 and 0> Cn Φ φ > rH •r| and and •r| rH •H *3 μ - φ φ μ >1 X XI) φ rH rH 44 43 0 H4 υ >t >X ‘'-, 44 and μ XO rn 43 43 + Φ *α 3 H •P P +’ •H 34 >1 0 3» Φ Φ 2 μ 43 CL About 44 rH N 2 2 2 and ’, 0 About About About 2 2 3 and μ and and and About 6 -p 2 2 2 <*> and 4) XI, 0 S>i 2 p i and <* OP OP xr and tn λ »3 about OP >» about about about about EH tn <#> and rH rH CM tn tn

rH i-tI rH 0 0 0 34 34 34 >1 >% >, rH rH rH rH rH rH 0 0 0 tn tp θ' 34 34 34 and and and >1 >1 >1 φ Φ φ rH rH iH rH H rH 3» Reply 0» >1 >1 >1 • and and and 43 43 43 φ φ φ 44 44 44 rH rH rH Φ Φ Φ >1 >1 >1 •rl •H •rl 43 43 43 μ μ μ 44 44 44 44 44 44 Φ Φ Φ \ 'V. 2 2 2 2 2 2 About About about About About About and and and 2 w 2 2 2 0P OP OP OP OP OP about about about tn CM CM CM CM CM

gas introduced at ambient pressure; ⁺⁺⁺ column flow rate 100 gh ⁺⁺ added as 2.9% sodium metal

Another example of carrying out the cleaning method according to the invention is illustrated on the pilot plant according to Fig. 5.

Pre-purified ethylene from an ethylene production plant is passed through a non-aqueous absorption solution containing 6% sodium hydroxide and 94% triethylene glycol for 56 days. The temperature in the scrubber is maintained at 30°C and the overpressure of the ethylene feed is 3.036 kPa. The carbon dioxide content in the feed gas is 1 to 6 ppm and in the exhaust gas 0.05 to 0.15 ppm. 13.6 kg of ethylene is fed per hour and the gas velocity in the scrubber is 0.18 m/min. On the last day of operation, the carbon dioxide content in the feed gas is reduced in the scrubber from 5.35 ppm to 0.10 ppm.

In the first trials of the process of the invention, a sodium-based system was investigated, whereby the absorption solution was prepared by adding sodium hydroxide or carbonate or sodium metal to an excess of glycol or glycerin solution to form the desired sodium salt of glycol or glycerin in the unreacted glycol or glycerin solvent.

This sodium-based system may not be trouble-free in operation. When using sodium hydroxide and liquid polyhydric alcohol based absorption solutions, filtration of the spent solution may be necessary to remove very fine sodium carbonate particles, 1 to 40 microns in size, which are formed in the presence of moisture. The sodium-based system may operate intermittently rather than continuously.

This deficiency is not present in the potassium-based absorption system. The non-aqueous absorption solution is a potassium glycolate or glycerate formed by reacting either elemental potassium or potassium hydroxide or potassium carbonate with a polyol, such as glycol and/or glycerin. If the preferred potassium carbonate is used as the potassium source, it is necessary to convert it to the hydroxide at a temperature of about 200°C to optimize the formation of the alcoholate.

A continuous system is characterized by the thermal regeneration of a solution containing potassium glycolate or glycerate, which is used to remove carbon dioxide from various gases. Potassium carbonate, soluble in glycol solutions, has been found to be readily soluble up to a concentration of 10% at ambient temperature or at a slightly elevated temperature of 20 to 25 °C. It has also been found that by heating these potassium carbonate-containing solutions to a temperature of about 160 to 200 °C, the potassium carbonate decomposes to release carbon dioxide with the simultaneous formation of the potassium glycolate salt. The reaction probably proceeds as follows:

N ₂

K,CO, + ROH -=-> 2 ROK + H,0 + CO, (A) ^J '(180 to 200 °C) where ROH means any glycol.

The potassium glycolate salt of the general formula ROK is therefore of the same type as when produced from potassium hydroxide according to the equation:

N,

ROH + KOH -=-\ ROK + H,0 (B) (125 °C)' ²

If potassium glycolate of formula ROK is prepared in the presence of water, to use potassium hydroxide as the alkali metal source (see equation B), the water is displaced from the system in the regenerator using nitrogen gas.

When displacing carbon dioxide in the absence of water, the reaction probably proceeds according to the following equation C:

II

ROK + CO, -RO - C - OK (25 to 100 °C)

Produced substance of the general formula

242452

O

II

RO - C- OK is also soluble in excess glycol as a solvent and also decomposes at a temperature of 180 to 200 ° C to carbon dioxide and a salt of the general formula ROK for reuse. If a trace amount of water is introduced into the scrubber together with the introduced gas, organic potassium carbonate is likely to be formed, but unlike sodium carbonate, which is contained in solid form, potassium carbonate remains in solution and can be regenerated again thermally. The advantage of this implementation of the method according to the invention is that under the given conditions no precipitate is formed and any carbonate formed can be decomposed thermally, whereby a solution for removing carbon dioxide is again formed after cooling.

In this embodiment, the method according to the invention therefore represents a heat-regenerable carbon dioxide removal system that does not require any filtration device.

The potassium hydroxide or carbonate and glycol process for carbon dioxide removal is based on the use of a stable non-aqueous solvent, such as triethylene glycol, with low vapor pressure and high dissolving power.

The chemistry of potassium hydroxide and carbonate solutions and glycol is not yet fully understood. It is not assumed that potassium carbonate is simply physically dissolved in glycols. Most likely, some reaction occurs between the alkali and, for example, triethylene glycol; partial alcoholation probably occurs.

Commercially available triethylene glycol (TEG) contains about 0.1% by weight of water. In addition, water is formed by the reaction of potassium hydroxide and triethylene glycol. Therefore, when preparing a glycolate or glycerate, the mixture after preparation contains some water. This water is usually removed by stripping before using the solution to remove the impurities mentioned above. During stripping, the equilibrium continuously shifts in favor of the formation of the alcoholate as water is removed from the system. When the majority of the water is removed from the system, the solution contains potassium glycolate, unreacted glycol, and small amounts of free, unreacted potassium hydroxide and/or potassium carbonate (if used as starting materials).

Potassium hydroxide and glycol solutions are extremely hygroscopic and always contain some condensed water and bound with excess glycol, which water remains in the system even after drying. This water is removed from the system by heating and flushing with inert gas or the like, as it must not accumulate in the system. However, such water removal is not the essence of the invention. Potassium hydroxide and glycol solutions can therefore also serve to remove water.

In Table V above, typical examples of the practical implementation of the process according to the invention using different feed gases, different sodium hydroxide and potassium hydroxide systems and solvent, different flow rates and operating temperatures are given.

From these examples it is clear that the process of the invention uses a large amount of absorbent - a glycol and glycolate solution system to purify gases such as nitrogen or ethylene and reduce the content of impurities such as carbon dioxide, carbonyl sulfide and hydrogen sulfide from a relatively high concentration to a concentration corresponding to a few ppm.

The use of a potassium-based system improves the system's ability to simultaneously remove water while removing carbon dioxide without forming solids. In the case of ethylene gas, which also contains traces of water, this system eliminates the need for absorbents to remove water, or at least significantly reduces their volume. The method of the invention is ideally suited for removing atmospheric carbon dioxide from air used for the production of nitrogen and oxygen by cryogenic air separation. As noted above, the spent non-aqueous absorbent solution is regenerated in a cyclic process as shown in Fig. 2.

To regenerate the solution, nitrogen purging is used along with heating to a temperature of 180 to 200 °C to decompose the carbonate formed in the solution to form carbon dioxide, with the carbon dioxide being removed along with water vapor and other gaseous impurities that are released and removed from the system during the process.

As mentioned above, Fig. 4 is a simplified flow diagram of a pilot plant for continuous ethylene purification using a potassium-based system or any other system in which a solid carbonate precipitate does not form in the absorption solution. As shown, valves VI and V2 are opened only to interrupt the daily continuous operation. The cycle allows for intermittent or continuous regeneration operation. In the case of continuous operation, a cooler must be included in the line between the conditioning and regeneration and storage vessels. It is also recalled that the solution in the regeneration vessel is cooled before being introduced into the storage vessel. The embodiment according to Fig. 4 is intended for ethylene purification using a 6% by weight solution of potassium carbonate in a solution of triethylene glycol. The use of nitrogen flushing to remove carbon dioxide is an optional measure in the case of a nitrogen-based system.

The use of a non-aqueous sodium-based absorption system for a prolonged period of time can result in the formation of a fine precipitate of sodium carbonate particles with a diameter of 1 to 40 microns, and removal of such precipitate is difficult even using a filtration step. It has been found that such fine particles can agglomerate into larger particles with a diameter of 200 to 400 microns by various means. Such larger particles can then be removed from the solution using simple direct filtration, as shown in the embodiment of Fig. 5.

Such agglomeration can be achieved in one of a number of ways, such as: adding up to 15% by weight of water prior to heat treatment at a temperature of about 150°C; and adding up to 2% by weight of a common ionic additive, such as sodium chloride for chloride ions, with or without the simultaneous addition of about 5% by weight of water. Such additives to the cycle affect agglomeration and are illustrated in the simplified flow diagram of FIG. 5, but are used in conjunction with the use of a burner in the line immediately upstream of the filter press.

In a sodium hydroxide scrubber, under certain operating conditions, sodium carbonate can accumulate. However, a scrubber with a non-aqueous sodium hydroxide solution is still more attractive than a sodium hydroxide pellet scrubber as an absorption system. Using a scrubber with a non-aqueous potassium carbonate solution results in a thermally regenerable system. A potassium carbonate system has two advantages over a sodium hydroxide system: no precipitate is formed during the cyclic removal of carbon dioxide and the system is thermally regenerable. However, potassium carbonate is only half as chemically effective as sodium hydroxide in removing carbon dioxide. Sodium hydroxide systems can remove approximately 42% of the theoretically contained carbon dioxide, while potassium carbonate systems remove approximately 20% of the carbon dioxide.

Table VI summarizes the results of cyclic carbon dioxide removal using solutions containing 3% lithium hydroxide in diethylene glycol, triethylene glycol, and ^dipropylene glycol. Theoretically, 5.5 g of carbon dioxide can be removed using 100 g of 3% lithium hydroxide solution in these solvents. Using 5.5 g as a numerical value for the total carbon dioxide absorption, the theoretical percentage of carbon dioxide removed during each carbon dioxide absorption/desorption cycle can be calculated.

In five carbon dioxide absorption/desorption cycles, 3% lithium hydroxide in diethylene glycol theoretically removes an average of 14.7% of the theoretical amount of carbon dioxide per cycle, while using lithium hydroxide in dipropylene glycol or triethylene glycol removes an average of 15.7 and 14.3% of the theoretical amount of carbon dioxide per cycle, respectively.

The use of a particular glycol and lithium hydroxide has little effect on carbon dioxide removal. It is also evident from Table VI that the efficiency of chemical carbon dioxide removal decreases significantly with each subsequent carbon dioxide absorption/desorption cycle regardless of the choice of glycol. For comparative purposes, Table VII shows the results for non-aqueous solutions of diethylene glycol and sodium hydroxide, potassium carbonate, and lithium hydroxide.

Table VII shows that the non-aqueous sodium hydroxide scrubber has the highest chemical efficiency compared to the non-aqueous potassium carbonate and lithium hydroxide scrubber. This conclusion is supported by the fact that after five cycles of carbon dioxide removal, the theoretical amount of carbon dioxide removed using sodium hydroxide is 1.5 times greater than that of the non-aqueous potassium carbonate and lithium hydroxide scrubber.

42.0% per cycle, while in the case of potassium carbonate or lithium hydroxide 19.9 and 17.4 * carbon dioxide per cycle.

Carbon dioxide scrubbing using a 3% by weight solution of lithium hydroxide in glycol as a solvent at 20 0 °C/

Oh my

Λ

XU

G >0)

G<0

O

Λ

XU

P

XJ

P

P •H

X o

P

Ό •Η 0 X Χί 0 XU

4->

Φ ·Η μ Κ> χθ -Η Μ Η Μ Χί Ν Ρ

G tr <u

P 0 CL

Ό 0 •h x: X xu O G

XU op G d

XU G >α υ •rM r-ί Ll) Χί ρ 2 &

P 'Ό Ο Η Χί * xU Ο G

XU OP G d

XU M Λ* -Ρ ϋ W

what IX •id about 4-> 0 » X! <U X! ’Γ 0 0 μ 0 ϋ rd XJ 0) 0 XJ <u I) '0 rd <u XU rd +j A4 Ε-» 4-1 Aaah IX

tO o O

Ο Ο Ο Ο Ο to <*> rd ΙΛ η «η © ο

flj μ Xi <U I 0 rh rd x > G rd rd cn <U >υ Ό tP G rd μ fO XU G <U μ 4-1 0) rd p XU 0 XO rd >1 G G P rd XJ XJ Ό <u About XJ 4-1 0 CL X» Φ (0 > <u N <u •H <u μ About •rl μ Ex from OJ Q EH

M M si to to o

Ο ΚΟ Γ' σ\ σ. to ιη to σ» xr ca m ••τ m to t-» oo

ΓΌΟΓχηοσίΓ^ο oo to σ ο oo σ σποοίΛωοιη f-tojm^mtor^co

O >1 i—I

Cn

G (U

I—< >1 Λ O M CL i i d 'Φ Li P P -H i W >0 IX Ό -H

H d P P do 3 O q 0 3 -úío 0 'Φ -H o o Μ -H fi 4-1 O d φλ φ Li XI) >L O fi Ρ φ >φ

Comparison of non-aqueous reversible scavenging efficiency for carbon dioxide of sodium, potassium and lithium based systems

O

Λ

KU

P <#> O ·Η rd >□ 'φ 'φ -H λ: ch υ >φ λ •h q 3 ρ d Φ Li 3 L Pd o w ·Η Q)d X H O O

T3 •H O 54 rC O 'Φ CP

P d ·η P >u d -H Li H Ρ Λ n q >□ d

Li φ ω q 3 Φ H tP44 Φ >í Pi □

IΛ O Tř

Oh Oh 04

OJ Ol 04 Ol 04 oj <n vr m

O 44 ►H r-i tp q φ q

P o

E

P

W cw O d tn Z <0 OO Ch OJ r-i oi m o

>1 rH

3» q

φ about

>1 year—I year

Φ i

Claims (6)

SUBJECT OF THE INVENTION A process for the purification of an industrial gas such as hydrogen, carbon monoxide, air, oxygen, helium, argon, monoolefins having 2 to 5 carbon atoms per molecule, diolefins having 4 to 5 carbon atoms per molecule, paraffins or acetylenes, by removing trace amounts of one or a plurality of impurities selected from water, hydrogen sulfide, carbon dioxide, carbon sulfide, sulfur dioxide, hydrogen cyanide, nitric acid and mercaptans, characterized in that the purified gas is contacted at a temperature of 15 to 100 ° C with a liquid, anhydrous solution containing 0, 5 to 15% by weight of an alkali metal or alkaline earth metal, based on its hydroxide or weak acid salt, in a liquid aliphatic polyhydric alcohol having an oxygen atom ratio of 1 to 5 and having at least two oxygen atoms separated by at most two successive carbon atoms. 2. Process according to claim 1, characterized in that 1 to 10% by weight. 3. Method according to claim 1, characterized by. in the form of hydroxide. 4. A process according to claim 3, wherein sodium hydroxide is used. 5. The process of claim 1, wherein the carbonate or hydroxide form. 6. The process of claim 1 wherein the triethylene glycol. alkali metal is used alkali metal, sodium or potassium is used alkali metal hydroxide, and alkali metal is used, and potassium in polyfunctional alcohol is used 5 drawings