US5416237A - Process for the production of acetic acid
USA - Process for the production of acetic acid
This application is a continuation of application Ser. No. 08/066,724, filed May 24, now abandoned.
If you are looking for more details, kindly visit SL Tec.
This invention relates to a process for the production of acetic acid and in particular to a process for the production of acetic acid by carbonylation of methanol.
Processes for the production of acetic acid by carbonylation of methanol are well known and are operated industrially.
Thus, UK patent GB 1,233,121 describes a process for the production of an organic acid or its corresponding ester by carbonylation using a rhodium catalyst.
When it is desired to obtain pure carboxylic acid from such processes it is necessary to remove contaminants such as water, iodide compounds and higher boiling contaminants such as propionic acid as in the case of carbonylation processes for the production of acetic acid.
UK patent GB 1,350,726 describes a process for the purification of monocarboxylic acid streams containing water and alkyl halide and/or hydrogen iodide contaminants, which process comprises (a) introducing a monocarboxylic acid stream containing water and alkyl halide and/or hydrogen halide contaminants into the upper half of a distillation zone, (b) removing an overhead fraction containing a major proportion of the water and any alkyl halide charged to the zone, (c) removing a stream from the middle portion of the zone and below the point of introduction in (a) containing a major proportion of any hydrogen halide present in the zone, and (d) removing a product monocarboxylic acid stream from the lower part of the zone, the product acid stream being substantially dry and substantially free of any alkyl halide or hydrogen halide contaminants charged to the zone. Although in the examples a product acid stream containing 87 to 132 ppm water was obtained from a process stream containing from 17.86 to 19.16 percent by weight water in a single, 40-plate distillation column, it is our belief that such a system would require significantly large column feed and overhead process stream flow rates and therefore large energy and column diameter requirements.
UK patent GB 1,343,855 describes a similar system for the purification of carboxylic acids but uses two separation stages.
In a process for the production of acetic acid by carbonylation of methanol, described by R. T. Eby and T. C. Singleton in Applied Industrial Catalysis, Vol 1, p275-296, , crude acetic acid product is purified in three separate distillation stages (a) a light ends column in which a heads light ends stream and a base heavy ends stream are separated for recycle to the reactor from a wet acetic acid side stream, (b) a drying stage in which the wet acetic acid is dried by distillation, the separated water being recycled to the reactor, and (c) a heavy ends column in which propionic acid by-product is separated from the dry acetic acid. In such a process the water concentration in the carbonylation reaction medium is relatively high, for example up to about 14-15 weight percent and removal of this water represents a significant cost to the process for producing pure, dry acetic acid.
European published patent application number EP-A- describes a reaction system by means of which an alcohol, as exemplified by methanol, can be carbonylated to a carboxylic acid derivative such as acetic acid while using a liquid reaction medium having a low water content. This is achieved by the use of defined concentrations of an iodide salt, alkyl iodide and ester in the liquid reaction medium to maintain rhodium catalyst stability and reactor productivity. EP-A- recognises that water is an undesirable component of the crude acetic acid and that the more water there is in the stream the greater will be the operating costs and required capital investment in the product recovery-purification system. However, in example 1 of EP-A-, although the water concentration of the liquid reaction medium is reduced to 4 to 5 weight percent from a prior art value of approximately 15 weight percent, the water concentration of the crude acetic acid from the methyl iodide-acetic acid splitter column is only reduced to approximately 4 to 7 weight percent and would therefore require further purification to remove the remaining water. According to European patent application number EP-A- which describes the same process as in EP-A-, one such method of drying the acid is a drying column.
We have now found that by operating a liquid-phase carbonylation reaction with a defined liquid reaction medium composition it is possible to use an improved product recovery system which uses a single distillation zone.
Thus according to the present invention there is provided a process for the production of acetic acid which process comprises:
(a) feeding methanol and carbon monoxide to a carbonylation zone in which there is maintained during the course of the process a liquid reaction composition comprising:
(i) a rhodium carbonylation catalyst;
(ii) methyl iodide;
(iii) a carbonylation catalyst stabiliser comprising an iodide salt which is soluble in the reaction composition;
(iv) a finite amount of water at a concentration of up to about 10% by weight preferably up to about 8% by weight;
(v) methyl acetate at a concentration of at least 2% by weight; and
(vi) acetic acid,
(b) withdrawing liquid reaction composition from the reactor and introducing it, with or without the addition of heat, to a flash zone to form a vapour fraction comprising water up to about 8% by weight preferably up to about 6% by weight, acetic acid product, propionic acid by-product and the majority of the methyl acetate and methyl iodide from the flash zone feed, and a liquid fraction comprising involatile rhodium catalyst, involatile catalyst stabiliser, acetic acid, water and the remainder of the methyl acetate, methyl iodide and propionic acid by-product from the flash zone feed,
(c) recycling the liquid fraction from the flash zone to the reaction zone, and recovering acetic acid product from the flash zone vapour fraction by use of a single distillation zone by:
(d) introducing the vapour fraction (as a vapour and/or liquid) from the flash zone into the distillation zone,
(e) removing from the head of the distillation zone a light ends recycle stream comprising water, methyl acetate, methyl iodide and acetic acid, and
(f) removing from the distillation zone at a point below the introduction point of the flash zone vapour fraction, an acid product stream comprising acetic acid having a water concentration of less than ppm, preferably less than 500 ppm and a propionic acid concentration of less than 500 ppm, preferably less than 200 ppm.
In the present invention the use of a defined reaction liquid reaction composition and limited water concentration in the flash zone vapour fraction allows the product acetic acid to be purified by use of a single distillation zone.
The process of the present invention may be performed as a batch or continuous process, preferably as a continuous process. The methanol feed to the carbonylation zone may be essentially pure as prepared by known industrial processes.
The carbon monoxide fed to the reactor may be essentially pure or may contain inert impurities such as carbon dioxide, methane, nitrogen, noble gases, water and C1 to C4 paraffinic hydrocarbons such as are known in the art. The partial pressure of carbon monoxide in the reactor is suitably maintained at 2.5 to 100 bara, preferably at 3 to 20 bara. Hydrogen present in the reactor as a result of the water gas shift reaction and optionally as part of the gas feed is preferably maintained at a partial pressure of at least 2 psi, preferably up to a maximum partial pressure of about 150 psi.
The carbonylation zone is preferably maintained at a pressure in the range 17 to 100 bara, preferably in the range 20 to 40 bara.
The carbonylation zone is preferably maintained at a temperature in the range 150° to 250° C., most preferably in the range 170° to 220° C.
The rhodium carbonylation catalyst concentration in the liquid reaction composition is preferably maintained at a concentration in the range 100 to ppm rhodium, most preferably in the range 150 to ppm. The rhodium carbonylation catalyst may be introduced to the carbonylation zone in any suitable form known in the art.
The carbonylation catalyst stabiliser is preferably an iodide salt of an alkali or alkaline earth metal or is a quarternary ammonium iodide or a quaternary phosphonium iodide. The alkali metals are lithium, sodium, potassium, rubidium and cesium. The alkaline earth metals are beryllium, magnesium, calcium, strontium, barium and radium. Preferably, the catalyst stabiliser is an iodide salt of lithium, sodium, potassium or calcium, most preferably an iodide salt of lithium. The catalyst stabiliser may also be a quarternary ammonium iodide salt such a quarternised amine, pyridine, pyrrolidine or imidazole, e.g. N,N' dimethyl imidazolium iodide or other heterocyclic nitrogen containing compound. Suitable heterocyclic iodide catalyst stabilisers are described in EP-A- which describes the use of catalyst stabilisers selected from the group consisting of quarternary ammonium iodides having the formula: ##STR1## wherein the R and R1 groups are independently selected from hydrogen or C1 to C20 alkyl groups with the proviso that at least one R1 group is other than hydrogen.
According to EP-A- it is preferred that at least one of the R groups is the same as the R2 group comprising the organic moiety of the alcohol, iodide derivative and carboxylic acid. The R1 groups on the other hand are suitably hydrogen or C1 to C8 alkyl, preferably hydrogen or C1 to C6 alkyl with the proviso defined above. Examples of preferred catalyst stabilisers in each of classes (1) and (2) are those where the R1 groups are selected from hydrogen, methyl, ethyl, n-propyl, iso-propyl, n-butyl, sec-butyl and t-butyl.
One particularly preferred class of catalyst stabilisers are iodide salts of the cation: ##STR2## where (i) R1 and R2 are methyl
(ii) R5 is hydrogen
(iii) R3 is C1 to C20 alkyl or hydrogen, and
(iv) R4 is C1 to C20 alkyl
Most preferred examples of this class are where (1) R3 =C2 H5, R1, R2 and R4 =CH3 and R5 =H or (2) R3 and R5 =H and R1, R2 and R4 =CH3.
Another particularly important class of catalyst stabiliser is comprised of iodide salts of the cation ##STR3## where R6 is either hydrogen or methyl, R7 is C1 to C4 alkyl and R1 is methyl. Preferred examples are where (1) R6 =H and R7 =C2 H5, (2) R6 =H and R7 =t-C4 H9 and (3) R6 and R7 =CH3.
Suitable quaternary phosphinium iodides include methyl tributyl phosphonium iodide, tetrabutyl phosphonium iodide, methyl triphenyl phosphonium iodide and the like.
The iodide catalyst stabiliser may be introduced directly to the carbonylation reactor. Alternatively, the iodide salt may be generated in-situ since under the operating conditions of the carbonylation reactor a wide range of precursors will react with methyl iodide to generate the iodide stabiliser. Metal iodide stabilisers may be generated from C1 to C6 carboxylate salts, e.g. acetate or propionate salts and quarternary iodides may be generated from the corresponding amine or phosphine.
The concentration of carbonylation catalyst stabiliser in the liquid reaction composition should be sufficient to maintain at a suitable level, the activity and stability of the rhodium carbonylation catalyst at the water concentrations in the liquid reaction composition. Preferably the catalyst stabiliser is present at a concentration of at least 0.4 mol per liter of liquid reaction medium measured at cold degassed conditions and up to the solubility limit of the stabiliser, most preferably 0.8 mol/l to 1.8 mol/l.
It has been found that certain selected iodide salts suppress the volatility of water relative to that of acetic acid and so their presence in the liquid reaction composition reduces the concentration of water relative to acetic acid in the vapour fraction produced when the liquid reaction composition is introduced into the flash zone. These iodide salts are iodides of alkali or alkaline earth metals or of hydrogen or aluminium, preferably iodides of lithium, sodium or potassium. Such iodide salts may be present in addition to iodide catalyst stabilisers or where the relative volatility suppressing iodide salt can also act as a catalyst stabiliser such a salt may be present for both functions of relative volatility suppression and carbonylation catalyst stabilisation. The presence of these relative volatility suppression iodide salts can allow the concentration of water in the flash zone vapour fraction to be controlled to the level necessary to achieve acetic acid product purification with a single distillation column in situations where in the absence of the relative volatility suppressant the water concentration in the liquid reactor composition would produce an unacceptably high water concentration in the flash zone vapour fraction. Thus, in the absence of such water relative volatility suppressants the water concentration in the liquid reaction composition should be maintained at no more than about 8% by weight in order that a water concentration of up to 8% can be achieved in the vapour fraction from the flash zone or no more than about 6% by weight if a water concentration of up to 6% is to be achieved in the flash zone vapour fraction. With an effective amount of water relative volatility suppressant the water concentration in the liquid reaction composition may be increased up to 10% by weight to achieve up to about 8% by weight water in the flash zone vapour fraction or up to about 8% by weight to achieve up to about 6% by weight water in the flash zone vapour fraction. The use of such iodide salts is described in our European Patent Application Publication number EP, Application No. .3. Suitable concentrations of relative volatility suppressing iodide salts are in the range 0.1% to 50% by weight subject to the limit of solubility of the salt. It will be appreciated that the relative volatility suppression effect may be as a result of a reduction in the volatility of the water or an increase in the volatility of the acetic acid or a combination of both effects.
The methyl acetate concentration in the liquid reaction composition is preferably in the range 2% to 15% by weight, preferably in the range 3% to 10% by weight. As the methyl acetate concentration in the liquid reaction composition is increased, the amount of propionic by-product decreases. By operating with at least 2% methyl acetate the concentration of propionic acid by-product in the liquid reaction composition is sufficiently low that its concentration in the acid product is below that level at which further purification is required, that is less than about 500 ppm.
The water concentration in the reactor is up to about 10% by weight preferably up to about 8% by weight, preferably up to 6% by weight, most preferably about 1 to 5% by weight.
Metal-iodides may be present in the liquid reaction composition as a result of corrosion of the reaction zone and associated equipment as well as being recycled to the reaction zone with returned process streams. Corrosion metals may comprise one or more of iron, chromium, manganese, nickel, molybdenum and the like. According to EP-A-, chromium and molybdenum may be beneficial to the carbonylation reaction. However, as the concentration of corrosion metals such as iron and nickel is increased, the concentration of by-product propionic acid in the liquid reaction composition increases. Therefore, whilst the total amount of corrosion metals should be maintained at a low level, it is preferable to selectively maintain particular corrosion metals such as iron, manganese and nickel at the lowest values possible. The concentration of corrosion metals which have an adverse effect on the process should be maintained as low as possible, for example typically less than ppm, preferably less than 500 ppm, most preferably less than 200 ppm in total. The adverse effect of iron and nickel in increasing by-product propionic acid production may be off-set to some extent by increasing the methyl acetate concentration. Methods for removing metal iodides from carbonylation catalysts are known in the art, for example methods which might be used are described in U.S. Pat. No. 4,007,130; U.S. Pat. No. 4,628,041 and EP-A-.
Acetic acid may comprise the balance of the liquid reaction composition together with other minor components including by-product propionic acid.
The flash zone is preferably maintained at a pressure below that of the reaction zone typically at a pressure of 1 to 6 bara. The flash zone is preferably maintained at a temperature of 100 to 160° C.
The vapour fraction may be introduced to the distillation zone as a vapour or the condensible components therein may be partially or fully condensed and the vapour fraction may be introduced as a mixed vapour/liquid or as a liquid with non-condensibles.
The distillation zone preferably has up to 25 theoretical stages. Since distillation zones may have differing efficiencies this may be equivalent to 35 actual stage with an efficiency of about 0.7 or 50 actual stages with an efficiency of about 0.5. Most preferably the distillation zone has up to about 18 theoretical stages. Preferably, the distillation zone has about 4 to 15 theoretical rectifying stages above the feed point. Preferably, the distillation zone has up to about 14 theoretical stripping stages below the feed point and above the base of the zone, most preferably about 3 to 14 theoretical stages. A suitable distillation zone, may have 18 theoretical stages with feed at theoretical stage 3 to 8 from the base and a crude acetic acid product take off as a liquid from the base of the zone. Another suitable distillation zone may have 20 theoretical stages with feed at theorectical stage 8 to 15 from the base and a crude acetic acid product take-off as a liquid from the base of the zone.
Preferably, the product acid stream may be removed at the base of the distillation zone or at a point 2 actual stages above the base of the distillation zone. The acid product may be withdrawn as a liquid or a vapour. When the acid product is withdrawn as a vapour preferably a small liquid bleed is also taken from the base of the zone.
Suitably the distillation zone may be operated at a heads pressure of about 1.3 bare and a base pressure of about 1.4 bare but higher or lower pressures may be used. The operating temperatures of the distillation zone will depend upon the composition of the feed, heads and base streams and the operating pressure. Typical base temperatures are 147° to 149° C. and heads temperatures are 115° to 118° C. The distillation zone may be operated with suitable return of reflux to the head of the distillation zone, for example at a rate of 1.5 times the heads product take-off rate.
It is expected that hydrogen iodide may be present in the feed to the distillation zone. The build up of this component may be prevented by introducing a small feed of methanol to the distillation zone, preferably below the feed point, to convert the hydrogen iodide to methyl iodide which is removed in the light ends recycle stream. It is expected that up to ppm hydrogen iodide in the feed may be treated in this way. Alternatively, or in addition, by operating the distillation zone at sufficiently elevated pressure, the operating temperatures in the distillation zone may be sufficient for the relatively high concentration of methyl acetate in the distillation zone to convert the hydrogen iodide to methyl iodide which is removed in the light ends recycle stream.
The company is the world’s best Acetic Acid Production Process supplier. We are your one-stop shop for all needs. Our staff are highly-specialized and will help you find the product you need.
It will often be the case that the vapour stream passing overhead from the distillation zone will be two phase when it is cooled. When the overhead stream is two phase it is preferred that reflux to the distillation zone be provided by separating the phases and using only the light, aqueous phase; the heavy, methyl iodide-rich phase being recycled to the reactor with the remaining light, aqueous phase as light ends recycle. The light ends recycle may comprise less than 5% by weight acetic acid.
Iodide impurities in the acetic acid product from the single distillation zone may conveniently be removed by passing the acid through one or more ion exchange resin beds. Alternatively, the iodide impurities may be removed by use of a silver salt scavanger as described in EP .
The acetic acid product may be purified to remove anionic iodide contaminants by passing it through an anion exchange resin bed such as described in UK patent application GB 2,112,394 the contents of which are hereby incorporated by reference. A particularly preferred resin for removing anionic iodide is a macroreticular weak base resin such as Reillex 425 (trade mark) which is a macrorecticular poly 4-vinylpyridine weak base resin having high temperature stability. Also preferred is the weakly basic resin Lewatit MP62 (trade mark), which is a macroporous anion exchange resin with tertiary amine groups (mono functional).
The acetic acid may be purified to remove organic iodide contaminants by passing it through a silver-loaded resin bed such as described in our European patent application publication EP; application number .9 which describes a process for removing iodide compounds from a carboxylic acid using an ion exchange resin having functional groups which each have at least one sulphur-donor atom and which resin has at least 1% of its sulphur-donor functional groups occupied by silver, palladium and/or mercury. Preferably the resin has thiol or substituted thiol groups. Preferred thiol groups comprise --SH groups, aliphatic thiol groups, aryl thiol groups, alicyclic thiol groups and/or thiouronium groups and preferred substituted thiol groups comprise isothiouronium groups.
Other suitable resins for removing organic iodides are described in EP-A- which describes a method of producing silver-exchanged macroreticular resins suitable for removing halides from liquid carboxylic acid contaminated with a halide impurity.
Suitable resins are also described in European patent application number EP-A- which describes a method for removing iodide compounds from a non-aqueous organic medium by contacting with a macroreticular strong acid cation exchange resin which has at least 1% of its active sites in the silver or mercury form.
Yet another class of resins suitable for removing iodide derivatives are described in our European patent application publication EP , application number EP .9 which describes a process for removing iodide derivatives from liquid acetic acid using a strong acid cation exchange resin having from about 4% to about 12% crosslinking, a surface area in the proton exchanged form of less than 10m2 g-1 after drying from the water wet state and a surface area of greater than 10m2 g-1 after drying from a wet state in which water has been replaced by methanol, said resin having at least 1% of its active sites converted to the silver form. Preferred resins are Purolite C145, Purolite CT145, Bayer K, Bayer K and Bayer K (trade marks) which have at least 1% of their active sites in the silver form.
Preferably the acetic acid is passed through one or more resin beds to remove anionic iodide contaminants before being passed through a resin bed to remove organic iodide contaminants thereby increasing the operating life of the organic iodide-removal resin beds which-would be rapidly saturated by anionic iodide contaminants which are generally present in greater quantities than organic iodides.
The operating lives of the silver-loaded resins are preferably extended by the use of strong acid cation exchange resin in the proton form to remove metal ion contaminants as described in our European patent application publication EP , application number EP .4.
The level to which the iodide concentrations are reduced will depend upon the intended use of the acetic acid product. Iodide contaminants may typically be reduced to less than several hundred ppb and preferably to less than 10 ppb.
The invention will now be illustrated by reference to the following examples.
Methanol was continuously carbonylated in the presence of a rhodium carbonylation catalyst, methyl iodide, lithium iodide catalyst stabiliser, a finite water concentration of up to about 10% by weight, methyl acetate at a concentration of at least 2% by weight and acetic acid, in a 6 liter zirconium, stirred reactor with a working mass of 4 kg (measured at ambient temperature in a bubble free state) at a pressure of 30 barg and a temperature in the range 180°-190° C. The temperature in the reactor was maintained by a hot oil jacket. Carbon monoxide was fed to the reactor on pressure demand via a sparge below the stirrer. Liquid reaction composition was continuously withdrawn from the reactor and passed to a flash tank operated at a pressure of 1.4 barg and a temperature of about 130° C. A vapour fraction comprising acetic acid product, propionic acid by-product, methyl iodide, methyl acetate and up to about 8% by weight water was passed overhead from the flash tank through an irrigated packed section and through a demister. The vapour fraction was condensed and introduced as a liquid into a distillation column operated at 1.2-1.3 barg. The liquid fraction from the flash tank comprising involatile rhodium catalyst, involatile lithium salt stabiliser, acetic acid, water and the remainder of the methyl iodide, methyl acetate and propionic acid was recycled to the reactor.
In the distillation column the acetic acid product was removed from the base. The methyl iodide, methyl acetate and water, together with some of the acetic acid passed overhead and were condensed into two phases. The light, aqueous phase was split: some was used as reflux to the column, the remainder was recycled to the reactor together with the heavy, methyl iodide rich phase as light ends recycle. Methanol could be fed into the column to react with any hydrogen iodide present; the methyl iodide and water produced being removed overhead.
The non-condensibles from the flash tank vapour and the head of the distillation column were first cooled to minimise the loss of volatiles from the process. The resulting off-gas stream was then passed to a scrubber where it was contacted countercurrently with chilled methanol. The methanol leaving the base of the scrubber was added to pure methanol and used as feed to the reactor.
The carbonylation and distillation stages of the process of the present invention were illustrated by the following two examples using the general procedure outlined above.
EXAMPLE 11.25 kgh-1 methanol and 1.36 kgh-1 carbon monoxide were fed continuously into the reactor held at an average of 186.6° C. The average composition of the reactor contents was: 2.6 wt % methyl acetate, 5.6 wt % water, 14.0 wt % methyl iodide, 61.9 wt % acetic acid, 0.55 wt % lithium (present at least in part as iodide salt) and 11.6 wt % iodide with 580 ppm rhodium, 190 ppm iron and 50 ppm chromium. The lithium iodide in the reactor composition functioned both as a carbonylation catalyst stabiliser in the reactor and a water relative volatility suppressant in the flash tank.
A feed to the distillation column of about 4.3 1 h-1 containing 3.8 wt % water, 43.3 wt % methyl iodide, 5.3 wt % methyl acetate and 48.8 wt % acetic acid was obtained. The distillation column in this example contained 35 PTFE sieve trays (stages) with the feed point at tray 20 (numbering from the bottom) with methanol fed at tray 8. A reflux ratio of 1.5 on a volume basis was employed. An acetic acid product stream of 2.26 kgh-1 was obtained from the base of the column which contained 460 ppm water, 180 ppm propionic acid and 1.7 ppm iodide.
The light ends recycle stream from the head of the distillation column was recycled to the reactor as two streams; methyl iodide rich stream containing <1 wt % water, <1 wt % acetic acid and 12.9 wt % methyl acetate and an aqueous stream containing 1.3 wt % methyl iodide, 8.8 wt % methyl acetate and <1 wt % acetic acid. Light, aqueous phase from the cooled column overhead vapour was fed to the column as reflux at a rate of 3.4 lh-1.
EXAMPLE 21.26 kgh-1 methanol and 1.34 kgh-1 carbon monoxide were fed continuously into the reactor held at an average of 183.6° C. The average composition of the reactor contents was: 5.6 wt % methyl acetate, 4.4 wt % water, 14.7 wt % methyl iodide, 55.0 wt % acetic acid, 1.18 wt % lithium (present at least in part as iodide salt and functioning as carbonylation catalyst stabiliser and relative volatility suppressant) and 15.7 wt % iodide with 425 ppm rhodium, 60 ppm iron and 15 ppm chromium.
A feed to the distillation column. of about 4.7 lh-1 containing 2.4 wt % water, 36.0 wt % methyl iodide, 16.7 wt % methyl acetate and 43.9 wt % acetic acid was obtained. The distillation column was configured as described for Example 1. An acetic acid product stream of 2.31 kgh-1 was obtained from the base of the column which contained 303 ppm water, 130 ppm propionic acid and 0.38 ppm iodide.
The light ends recycle was again recycled to the reactor as two streams. The methyl iodide rich stream contained <1 wt % water, <1 wt % acetic acid and 22.4 wt % methyl acetate. The aqueous stream contained 2.3 wt % methyl iodide, 9.6 wt % methyl acetate and 1.9 wt % acetic acid. Light, aqueous phase reflux was fed to the column at a rate of 3.3 1 h-1.
It is expected that the acetic acid products from Examples 1 and 2 may be purified of iodide contaminants by passing through suitable ion exchange resin beds. In this way pure acetic acid may be obtained using a single distillation zone without excessive recycle of acetic acid overhead.
EXAMPLES 3 to 5Further examples using the general procedure outlined above for Examples 1 and 2 were performed using the reaction conditions given in Table 1 below:
TABLE 1
______________________________________
REACTOR CONDITIONS
Example 3 4 5
______________________________________
Feeds
Methanol (kg/hr)*
1.26 1.23 1.24
Carbon Monoxide (kg/hr)
1.26 1.18 1.14
Reactor
Temperature (°C.)
188.5 189.1 186.9
Pressure (barg) 30 30 26
Liquid Reactor Composition
Methyl Acetate (%)
2.9 3.1 7.3
Water (%) 2.0 1.9 4.2
Methyl Iodide (%)
15.2 12.0 13.9
Acetic Acid (%) 63.3 67.4 53.4
Lithium (%) 0.76 0.63 0.88
Iodide (%) 12.0 13.8 15.1
Rhodium (ppm) 550 550 320
Iron (ppm) 290 270 80
Chromium (ppm) 70 80 60
______________________________________
*Including methanol fed to the distillation column.
The distillation column conditions are given in Table 2 below:
TABLE 2
______________________________________
DISTILLATION COLUMN CONDITIONS
Example 3 4 5
______________________________________
Number of Actual Trays
26 26 28
Feed at Tray* 7 15 20
Methanol Feed
Tray 6 6 8
Rate (g/hr) 55 55 50
Feed (l/hr) 3.7 4.0 5.1
Feed Composition (%)
Methyl Acetate 6.2 8.8 15.7
Water 1.0 1.9 2.6
Methyl Iodide 43.8 47.7 39.7
Acetic Acid 51.7 45.6 39.0
Reflux (l/hr) 2.3 3.4 4.6
______________________________________
*Numbered from the base.
The compositions and flow rates of the process streams from the distillation column are given in Table 3 below:
TABLE 3
______________________________________
DISTILLATION COLUMN PROCESS STREAMS
Example 3 4 5
______________________________________
Overhead Methyl Iodide
Rich Stream
Water (%) 0.6 0.6 1.3
Acetic Acid (%) 0.4 0.7 1.1
Methyl Acetate (%)
13.8 14.5 26.2
Overhead Water Rich Stream
Methyl Iodide (%) 1.4 1.5 1.7
Methyl Acetate (%)
7.0 8.3 11.2
Acetic Acid (%) 4.2 7.0 5.8
Acid Product
Rate (kg/hr) 2.33 2.19 2.33
Water (ppm) 470 670 320
Propionic Acid (ppm)
120 100 64
Iodide (ppm) 0.7 0.4 0.3
______________________________________
The acetic acid product from the distillation column in Example 5 was passed through a bed of Lewatit MP-62 resin at 80° C. at a rate of 10 resin bed volumes of acetic acid treated per hour to remove anionic iodide contaminants to less than 1 ppb. The combined overhead streams forming the light ends recycle stream were calculated to have 0.96% by weight acetic acid and 1.33% by weight acetic acid in Examples 4 and 5 respectively, that is less than 5% by weight acetic acid.
Production routes to bio-acetic acid: life cycle assessment
In this study the production of acetic acid via the bioconversion of poplar biomass is evaluated using life cycle analysis. Models of acetic acid production plant with an annual biomass processing of 227,000 BDT/year were simulated in ASPEN-Plus chemical engineering modeling software, producing 120,650 tonnes per year of acetic acid for EAX and ADX solvent based scenarios. Total capital expenses were estimated at 245, 197, 223 and 187 million USD for EAX OC, EAX LE, ADX OC, and ADX LE, respectively. Scenarios are assessed that measure the life cycle environmental tradeoffs between acetic acid distillation/extraction methods, and within these models looking at burning lignin onsite or using the lower capital cost approach of selling the lignin to a coal power plant. Cradle to biorefinery gate system boundaries are set for acetic acid production to include the growth and harvesting of poplar biomass, biorefinery operations, and manufacturing of all necessary inputs (i.e., process chemicals, energy). Use and disposal of acetic acid is beyond the scope of this study. Environmental impacts to be assessed include the global warming potential, and fossil fuel use.
Cradle to biorefinery exit gate system boundaries are used to evaluate acetic acid production (Fig. 6a, b). A functional unit of 1 tonne of acetic acid produced from a biorefinery system with 21 year operating time frame is used in the analysis. Environmental impacts considered are the 100 year Global Warming Potential (GWP) [18], and Fossil Fuel Use (FFU). FFU is calculated by summing all fossil fuel inputs (coal, natural gas, crude oil) per tonne of acetic acid. Guidelines for conducting a LCA are set by ISO [21] and [22] and this research follows the ISO design. LCAs in this research are developed using SimaPro v.8.0. Scenario results are compared to each other as well as to petroleum based acetic acid produced by methanol carbonylation [15]. A sensitivity analysis is conducted to investigate the effect of a decreased acetic acid yield. Additionally system expansion method for co-products is compared to both economic and mass allocation when lignin is exported to a coal burning facility.
Fig. 6a Acetic acid (AA) extracted and distilled using ethyl acetate (EA). Both lignin scenarios are represented in the system boundaries figure. Black dashed line boxes indicate lignin scenario dependent operations. Lignin can either be burned onsite in the boiler to help produce heat/steam/electricity or sold to a coal power plant and co-fired with coal to produce electricity. If lignin is burned onsite, steam is run through a turbine to produce electricity. If lignin is exported to a coal power plant, no onsite electricity is made and electricity must be purchased from the grid for biorefinery operations. Green boxes highlight product made/energy produced. b Acetic acid (AA) extracted and distilled using an alamine and diisobutyl ketone solvent (ADX). Both lignin scenarios are represented in the system boundaries figure. Black dashed line boxes indicate lignin scenario-dependent operations. Lignin can either be burned onsite in the boiler to help produce heat/steam/electricity or sold to a coal power plant and co-fired with coal to produce electricity. If lignin is burned onsite, steam is run through a turbine to produce electricity. If lignin is exported to a coal power plant, no onsite electricity is made and electricity must be purchased from the grid for biorefinery operations. Green boxes highlight product made/energy produced
Full size image
The bio-acetic acid life cycles are broken up into 4 sections: feedstock production and harvesting, ancillary chemicals, the biorefinery, and lignin use (co-product scenarios). Feedstock production and harvesting is the same for each biorefinery configuration. Biorefinery process, ancillary chemical inputs, and lignin use will vary depending on the configuration. Descriptions of each life cycle section and allocation methods for lignin use follow below. System boundary diagrams for each bioconversion pathway are displayed in Fig. 6a, b.
Feedstock
The feedstock production and harvesting model is supported by operational data from industry (GreenWood Resources, personal communication, &#;), literature [23], and LCA databases [15, 24]. It is the same feedstock model used in [4] and is discussed in more detail in that publication. A brief description is provided here. The feedstock production and harvest model is representative of a coppice harvest system, with the poplar trees being coppiced every 3 years for 6 cycles. The model includes all necessary site preparation, nursery operations, management of the poplar tree stands, harvest operations, and stump removal. Nitrogen fertilizer is applied in the spring following a harvest at a rate of 56 kg N per application. N2O emissions from fertilizer and decaying biomass are calculated using the Farm Energy Analysis Tool [25]. Storage of carbon in the harvested poplar biomass as well as in below ground biomass (stump and roots) is included. The amount of carbon stored within the below ground carbon stores is assumed to be the same as willow SRWCs and no change in soil carbon down to a depth of 45 cm is expected to occur during tree growth [26]. The equivalent amount of CO2 stored in the poplar wood is calculated using the stoichiometric relationship of CO2 to carbon of 3.66 kg kg&#;1and a carbon mass fraction of 51.7% dry wood weight [27] (Table 2a & b). Direct land use change is included using the assumption that fallow land will be used for poplar plantations. Direct land use change associated with establishing the plantation is calculated using the Forest Industry Carbon Assessment Tool v.1.3.1.1. Indirect land use change is excluded from the system boundaries due to uncertainty associated with these models [28]. A transportation distance of 100 km roundtrip is assumed to transport the harvested poplar biomass to the biorefinery gate. In total the feedstock production and harvest model covers a 21 year timespan [4].
Biorefinery
Currently no commercial facilities are using an acetogen fermentation pathway to produce biofuels and biochemicals. To assess the conversion impacts ASPEN-Plus v.8.6 chemical engineering software is used to simulate potential biorefinery process designs. The acetogen fermentation pathway ASPEN simulation is based on a combination of the NREL model [1], a proposed acetogen fermentation process [29], and laboratory work at the Biofuels and Bioproducts Laboratory at the University of Washington. The simulated biorefinery is assumed to operate on 250,000 tonnes of bone dry biomass per year.
Regardless of the product recovery method used (EAX or ADX), biorefinery operations begin with the same processes; dilute acid pretreatment, enzymatic hydrolysis, and fermentation. Pretreatment, hydrolysis, and fermentation conditions are presented in Table 4. These steps are based on National Renewable Energy Laboratory (NREL) corn stover model, but modified to use a poplar feedstock [1]. Following enzymatic hydrolysis, glucose and xylose are fermented to acetic acid using Moorella thermoacetica. The streams exiting the fermentation stage include a solid and liquid stream. Descriptions of the fate of these streams are described below. Incorporated into the biorefinery designs, and included in the LCAs is a wastewater treatment system. The WWT design is based on Humbird et al. [1]. Wastewater streams are treated in aerobic and anaerobic environments to produce clean process water, sludge, and methane. The sludge and methane are sent to the boiler. Solid waste produced from the biorefineries is comprised of ash from the boiler, which is collected and sent to a landfill for disposal.
Table 4 Process parameters for pretreatment, enzymatic hydrolysis, and fermentationFull size table
The liquid stream exiting the fermentation is 5 wt % acetic acid and water. To be marketable, acetic acid must be concentrated to 99.8 wt % (glacial acetic acid). When acetic acid concentrations are low (0.5&#;5 wt %) direct distillation of acetic acid from water is inefficient and liquid&#;liquid extraction (LLE) is the preferred acetic acid recovery method [12]. In this research two LLE methods, both achieving acetic acid yields of 532 kg per bone dry tonne of biomass, are investigated to purify acetic acid. The first method uses ethyl acetate for extraction followed by distillation to recover the ethyl acetate (EAX). The second LLE method uses an alamine/DIBK extraction (ADX). These two extraction scenarios are described below. Major inputs and outputs for the biorefinery scenarios are presented in Table 5.
Table 5 Major inputs and outputs from the biorefinery of each scenarioFull size table
Ethyl acetate extraction
An overview of the ethyl acetate extraction (EAX) process is presented in Fig. 6a. Following fermentation a mixture of water acetic acid (5% acetic acid by weight) is sent to a liquid&#;liquid extractor. Here it is mixed with ethyl acetate (EA) and the EA solubilizes the acetic acid. A mixture of acetic acid, EA, and a small amount of water are sent to a dehydration column operating at 118OC. EA and the remaining water are distilled off and glacial acetic acid (99.8% acetic acid) is produced. Water and EA are sent to stripping column to recover the EA. EA exiting the stripping column is recycled back to the extractor. Water is sent to an onsite wastewater treatment facility before being cycled back through the process. Major inputs and outputs from the biorefinery are listed in Table 5.
Alamine/Diisobutyl ketone extraction
An overview of the alamine/DIBK solvent extraction (ADX) process is presented in Fig. 6b. After fermentation, acetic acid in water (5% acetic acid by weight) is sent to an extractor. Acetic acid and water are mixed with alamine and DIBK. Acetic acid combines with alamine and DIBK and is removed from the water. This mixture is sent to a dehydration column (174 °C) to remove any residual water. Following dehydration the mixture of acetic acid, alamine, and DIBK is sent to a stripping column (190 °C). Alamine is removed and a mixture of acetic acid and DIBK is sent to a second stripping column (168 °C). DIBK is removed and glacial acetic acid is produced. DIBK and alamine are recovered and are recycled back into process. Water removed during extraction/distillation is sent to wastewater treatment before being reused. Major inputs and outputs from the biorefinery are listed in Table 5.
In both extraction methods the solid stream separated out after the fermentation stage consists of lignin and other unfermented carbohydrates. To recover these solids and remove some of the residual water, the solid streams are filter pressed to 50% solids. Following concentration of the solids two potential downstream options for the solid stream are evaluated in this study. These are described in more detail below.
Onsite lignin combustion
Option one consists of combusting lignin onsite to produce heat/steam for the biorefinery operations and producing electricity by running high pressure steam through a steam turbine; using the moderate pressure steam exiting the turbine for the process. This practice is common in proposed biofuel biorefinery designs [1, 4] and pulp mills [30]. Compared to second generation lignocellulosic ethanol production, producing glacial acetic acid requires more heat/steam and combusting lignin alone cannot meet the entire energy demand. Extraction/distillation of acetic acid requires a significant amount of steam. To meet this demand, natural gas is imported and combusted with the lignin. To reach the temperatures needed for extraction/distillation moderate pressure steam would be required. However, technoeconomic work with the ASPEN model identified an economic benefit to instead first create high pressure steam and pass it through a turbine to produce moderate and low pressure steam. The conversion of high pressure steam to lower pressure steam through the turbine generates an amount of electricity that exceeds the needs of the biorefinery. The conversion efficiency of heat to steam is assumed to be 80% for both natural gas and lignin. The excess electricity can be sold to the electrical grid to increase the revenue generated from the biomass. The production of excess electricity is greater in the ethyl acetate extraction process as this method has a greater steam demand, and therefore, more high pressure is passed through the turbine.
For the LCA of the onsite lignin combustion scenario, the electricity by-product is treated using system expansion per ISO standards [22]. The electricity by-product meets the requirements for using system expansion as it is currently produced from other sources and life cycle data for the production of electricity from these other sources can be obtained [31]. System expansion is the most common method used in biofuel LCAs to deal with an excess electricity by-product [32]. It is assumed that the electricity will be sold to the grid and displace electricity produced from natural gas, a likely candidate for the marginal electricity source [33]. An avoided production credit is generated for displacing this fossil fuel source of electricity with electricity produced from a renewable source. Fugitive emissions from process operations are estimated to be 2% of unit process flows [34].
Sell lignin to power plant
The second option for lignin is to export it to a coal power plant and co-fire with coal. This has been shown to be a viable option and can economically and environmentally benefit both the biorefinery and the coal power plant [17]. In this scenario lignin is considered a co-product to acetic acid production. It is dried to about 50% lignin by weight (50% water) and shipped to a nearby coal power plant and used in place of coal. The amount of coal displaced is based on the energy content of the lignin. The moisture content of the lignin will affect the energy content and must be accounted for when calculating the amount of coal displaced (i.e., the energy required to remove water prior to combustion is included in the coal displacement calculation). To determine the coal displacement by selling the lignin as a co-product, the HHV of wet lignin (50% MC) was estimated using ASPEN. Lignin was modeled as vanillin C8H8O3 with an HHV of 25.2 MJ/kg [1], similar to the experimental value of dilute acid pretreated lignin of 21.4 MJ/kg [35]. Assumed HHVs for all combustible materials are reported in Table 6. Exporting lignin as a co-product requires that other forms of energy must be used to meet the needs of the biorefinery. Natural gas is assumed to be combusted at the biorefinery to provide heat and steam. In this scenario it is assumed that a lower pressure boiler is used and the additional expense of a turbine to generate power would not be incurred. Natural gas boilers are more commonly used in the industry due to the relatively low capital cost in the range of $8&#;$23/kW [36] and their relatively small physical size. In contrast, biomass boilers are larger in size and have high capital cost ($94&#;$125/kW) [37] due to more complex design. Consequently, natural gas boilers are typically less expensive than those that would be suitable for combusting lignin. Exporting lignin and using natural gas as the sole driving fuel represents a lower cost alternative. A high pressure boiler with turbo-generator would not be appropriate in such a biorefinery design approach. For the lignin export case the biorefinery electricity needs are assumed to be met by importing electricity from the U.S. national grid.
Table 6 High heating values (HHVs) for lignin, coal, and natural gasFull size table
Ancillary chemicals
Biorefinery operations require chemical inputs to convert the poplar biomass to acetic acid. The production of these chemicals is grouped into the ancillary chemicals section. Unit process data for the chemical inputs come from the USLCI [15], EcoInvent [24], literature, and the private sector. The electricity source in each unit process is set to come from a unit process representative of the U.S. national grid [38]. Data for enzyme production is supplied from Novozymes for their Cellic Ctec3 cellulases [39]. Transportation distances for each chemical are determined using the U.S. commodity flow survey [40].
Allocation and sensitivity analysis
System expansion is used in evaluating the base case for the four bio-acetic acid production scenarios. As discussed above, this is deemed to be the appropriate treatment of the life cycle impacts for the product and excess electricity/lignin co-product. However, the results are also evaluated using mass and economic allocation methods to determine the life cycle effect of allocating life cycle impacts between acetic acid production and the lignin co-product (in the lignin exporting scenarios). Allocating the life cycle impacts between acetic acid and lignin divides the environmental benefits (i.e., carbon sequestration) and burdens (i.e., natural gas combustion) between acetic acid and the lignin co-product. To account for the movement of carbon within the biorefining systems the carbon sequestered in the poplar biomass is allocated to either acetic acid (i.e., glucose and xylose) or lignin. Producing one tonne of acetic acid requires 1.8 tonnes of poplar biomass (dry weight). At a 51.7% carbon content [28], 1.8 tonnes of poplar contain 930 kg of carbon. Through the bioconversion process 400 kg of this carbon will go into the acetic acid. 380 kg of the carbon is contained in the lignin. The 150 kg of carbon remaining in the system (carbohydrates in liquid streams) is divided between the acetic acid product and lignin co-product. The amount of this 150 kg assigned to either the acetic acid product or lignin co-product depends on the allocation method being assessed (mass or economic). From the acetic acid product view point, allocation also removes all processes that are downstream of the biorefinery&#;and tied to the lignin co-product&#;from the life cycle production of acetic acid; including coal displacement and emissions from lignin combustion at the coal burning facility.
The mass allocation approach divides the life cycle processes amongst acetic acid and lignin according to the mass of each product. For every tonne of acetic acid produced, 394 kg of lignin (dry weight) is exported. Economic allocation divides life cycle processes amongst acetic acid and lignin based on the economic values of these two products. The minimum selling price was calculated in ASPEN and used to establish the value of the acetic acid. In the EAX LE scenario techno-economic analysis identified the minimum selling price of acetic acid to be $819 per tonne and economic value of the lignin exported to the coal facility to be $39 per tonne (assuming $4.40 per MMBTU). In the ADX LE scenario the selling price for acetic acid is $677 per tonne and the value of the exported lignin to be $39 per tonne.
A sensitivity analysis is conducted to test for model sensitivity to changes in fermentation yields. The fermentation yield of glucose to acetic acid in this research is set at 92%. Maintaining a fermentation yield of 92% may be difficult when operating at commercial scale and could likely fluctuate. If the fermentation yield decreases the amount of acetic acid produced would decrease and the amount of unfermented carbohydrates, and therefore, the amount of biomass available to burn would increase. To test the effect of a decreased fermentation yield and to evaluate model sensitivity, a simulation is performed for the EAX OC and EAX LE scenarios in which the fermentation yield is decreased by 10%. Only EAX is tested for sensitivity analysis as this system is more likely to be commercialized and it is expected that the effect of a decreased acetic acid yield would be similar for both EAX and ADX.
Contact us to discuss your requirements of Methyl Acetate Plant. Our experienced sales team can help you identify the options that best suit your needs.
Comments
0