DE19527506A1 - Appts. for biologically treating solvent-contg. waste air - has buffer system to cope with solvent concn. spikes, allowing size of bio=reactor to be reduced - Google Patents

Abstract

The appts. for biological treatment of solvent-contg. waste air has an integrated buffer system that can be switched in as required, allowing the concn. spikes to be balanced out and permitting a long-term redn. in pollutant load during operating periods. The waste air can be supplied to the reactor directly and via a pre-separator in infinitely variable proportions, where the proportion supplied directly to the bioreactor is such that it can be handled [by a bioreactor] with the desired degradation capacity, this value being determined beforehand in a series of tests with direct gas injection (100% waste air direct into the bioreactor). During periods of high pollutant concn., a side stream is introduced into the buffer system, from which the adsorbed solvent can be continuously desorbed with compressed air during periods when operations are at a standstill, where temp./time programs can be established to meet the needs of the particular waste air problem.

Description

1 Introduction

Although the application of biofilters originally for the elimination of odor was designed to be an important alternative to the developed physical and chemical methods of emission control. It could be shown that biofilters are capable of solvent on an industrial basis To remove exhaust gases. One can expect that in the future for special application can be used in industry "tailor made" high performance filters.

The objective of this research is to develop a powerful, reproducible approachable and interpretable biofilter system through the use of special, Adapted microorganisms, as well as by the application of process technology For a suitable biofilter planning, the insights in the Processes that take place in a biofilter indispensable. While with pilot plants in situ the macro-kinetics is determined, the micro-kinetics should be reduced to laboratory scale to be looked for. By means of modeling these examination results can be processed and be used for the dimensioning of a large plant.

By a biological cleaning of the exhaust air of a paint producing company should below the legally prescribed emission limit values. The focus is on developing pollution - related criteria for the Use of the bioreactor, such. B. the behavior when concentration peaks occur and their leveling by using a buffer substance, the first time in this Context, as well as the naming of pollutant and biomass-related Criteria for the choice of filter material. Particular importance is attached to the use a solvent mixture whose composition corresponds to the real exhaust air, and this together with the reactor filling in a laboratory test facility information about the Behavior of predominantly occurring in practice mixed gases should give. Especially in the In the paint / varnish area, multi-component mixtures with over 30 components are not uncommon.

Taking into account the relevant specialist literature (inter alia, VDI Guidelines 3477 ¹ and 3478 ²) and current publications, had to be checked by examination of the operational Be conditions and by emission measurements the necessary questions for the choice of ge suitable method (bioscrubber, biofilter or trickling filter reactor). It turned out that the purification of the exhaust gases by means of a bio-scrubber for Exhaust air of the paint factory is out of the question. In addition to the absorption of pollutants in one  Liquid, d. H. Inclusion and even distribution of gas molecules, becomes heavy water-soluble components adsorption to a solid surface as additional Sorption process necessary. These adsorption effects resulting from physisorption and Chemisorption are not present in a bioscrubber.

The next step was to design a bioreactor for the present one Cleaning problem represents the method of choice. Finally, the designs were through realized the construction of a laboratory test facility, which in extensive test series much promising results on the biodegradation of the exhaust air supplied by a paint factory Has. In addition, the different adsorption and desorption behavior could the individual solvent components are tested with respect to the buffer substance, which in this context was used for the first time. Statements about the regeneration of the Buffer material, the service life of the reactor, as well as the degradation behavior in a prolonged downtime complete the test results.  

2. Detection of the exhaust air problem in the paint factory 2.1. General

In the production hall II of the paint factory, mainly varnishes for the wood ver working industry. In addition, the company produces to a lesser extent also sieve cleaner. The operating time of the company operating in a single-shift system is 8 hours a day or 40 hours a week. The emission takes place during this period is discontinuous on a regular basis, whereby the operator has an emission period of 8 hours a day. The resulting solvent vapors are by means of a fan passed through the roof to the outside air. Other facilities for Emission reductions are not available so far.

The emission measurements for the quantitative determination of the exhaust air problem were in Zu co-operation with a local laboratory for pollutant analysis and environmental measuring technology, carried out. The results are in section 2.2. listed.

2.2. Emission measurement (FID)

Site of the facility: Hall II
Type of measurement: emission measurement; Measurement of the hydrocarbon concentration by means of FID; organic carbon.

Description of emission sources

At the time of the measurements, the following were used when filling the mixing tank materials used in Hall II:

Solvent:

Solvent naphtha
ethyl acetate
butyl
toluene
1-methoxypropyl acetate-2
1-methoxy-2
diacetone
Special fuel 80/100
ORM regenerate (composition see 2.3.).

In addition, a possible addition of:

Binders such. B .:

nitrocellulose
acrylic resins
polyester resins
alkyd resins
urea resins

Plasticizers such. B .:

dibutyl phthalate
dioctylphthalate
Biethylsteorat

Matting agents such. B .:

PE waxes
silicas

or other additives.

The TA-Luft sets the following limit values:

analyzer:

- FID Ratfisch, type R555
- Calibration gas: propane in synthetic air 992 ppm
- set measuring range was 0-1000 ppm
- suitability tested
- DRAEGER tubes for determining the humidity; DRAEGER water vapor 1a: 0.1.

dates

Emissions (FID)

mass flow

Evaluation of the measurement results

TA-Luft, point 3.2.2.2 .: was fulfilled.
TA-Luft, point 3. 1.7 .: in the case of exhaust air consisting of organic substances of several classes, in a mass flow of 3 kg / h or more in addition to the requirements of sentence 1, the mass concentration in the exhaust gas shall not exceed 0.15 g / m³.

- The mass flow of 3 kg / h is not achieved.
- the mass concentration of 0.15 g / m³ is also not reached by any of the mean values; However, the concentration peaks of both measuring points exceed this benchmark by more than double.

2.3. Measurements to determine the concentrations of the individual Solvent components in different types of production (GC)

Through these emission measurements should be the concentrations of each solvent broken down into components, d. H. compared with the FID measurements in addition to the quantitative and a qualitative characterization of the exhaust gas. To the Product range of this production hall and thus the used solvents representative The emissions measurements were taken within a period of about six Weeks performed. During this time were three different types of paint and one Screen cleaner made. The adsorbed by activated carbon samples were in the laboratory with a suitable solvent (n-nonane, o-xylene) and eluted by gas chromatography evaluates. The aim of these measurements was to break down the exhaust gas components to produce representative solvent mixture, which then as "synthetic Exhaust mixture "should serve to feed the bioreactor test facility.  

Results of the exhaust air measurements (GC)

Location: Hall II
Type of measurement: emission measurement; Gas chromatography after enrichment with activated charcoal.
Exhaust air velocity measurement with pitot tube.

Emission sources:
Measuring point 1: rectangular channel, western chimney (see also 2.2.)

- flow velocity: 8.5 m / s;
- Sampling about 70 cm under exhaust air outlet.

Measuring point 2: round channel, east chimney (see 2.2)

Flow rate: 9.0 m / s;
- Sampling about 40 cm under exhaust air outlet.

analyzer:
Activated charcoal tubes: (company: KARL OHG) for adsorption; GC: VARIAN Aerograph; Series 2700;
Capillary column: SPB1 / 30 m; ID 0.32 mm, 0.25 μm (SUPELCO)
Eluate: o-xylene and n-nonane from FLUKA (standards for GC);
different GC solvent standards.

At the time of the measurements, when filling the mixing tank in Hall II, the following became Solvent used:

acetone
ethyl acetate
n-BAC
n-butanol
diacetone
Dowanol PM (1-methoxy-2-propanol)
ethoxypropylacetate
MIBK (methyl isobutyl ketone)
ORM regenerated *
Solvent naphtha
toluene
technical xylene (mixture of ethylbenzene, ortho-, meta- and para-xylene).

The following solvents or mixtures did not become available during the period of emission measurements used, but must also be mentioned as frequently used components:

MEK (methyl ethyl ketone)
Special fuel SB 80/110.

The solvents used during the sampling were the respective Arbeitsvor taken from employees. For additives already in dissolved templates, could from the corresponding DIN safety data sheets of Specialist companies, the solvent used can be looked up. Butyl acetate (BAC) in different concentrations z. B. in the starting materials SYNTHALATE A 1633 (50% BAC) or in the solution STANCLERE TL (10% BAC). The composition of the often used ORM Regenerats should also be listed here:

* ORM regenerated composition:

BAC 21.9 Vol .-% ethyl acetate 12.4 vol.% toluene 12.0 vol.% technical xylene 12.0 vol.% acetone 10.5 vol.% Methyl acetate 85 (15% MeOH) 4.5 vol.% isopropanol 2.7% by volume Methanol, H₂O, etc. each <2.4% by volume also: 8 vol.% Solvesso 100 (bp. 170 ° C)

Emission measurement 1 (GC)

Date: 09. 06. 93
Production method: DS paint for automatic spray gun

Ingredients (solvents only, alphabetically):
acetone
ethyl acetate
BAC (butyl acetate)
Dowanol PM (1-methoxy-2-propanol)
MIBK (methyl isobutyl ketone)
ORM regenerated *
Silicon PL solution 1% in xylene (see DIN data sheet)
toluene

Sample volume: 6.5 l (flow rate: 0.65 l / min, measuring time: 10 min)
Extractant: 1 ml of o-xylene

Emissions [mg / m³]

Emission measurement 2 (GC)

Date: 29. 06. 93
Production method: SH-varnish, matt

ingredients
(only solvents, alphabetically):
ethyl acetate
BAC 98/100 (butyl acetate)
butanol
Dowanol PM (1-methoxy-2-propanol)
ORM regenerated *
toluene
technical xylene

Sample volume: 6.0 l (flow rate: 0.60 l / min, measurement time: 10 min)
Extractant: 1 ml of n-nonane

Emissions [mg / m³]

Emission measurement 3 (GC)

Date: 13. 07. 93
Production type: NC coating varnish

ingredients
(solvent only, alphabetically):
acetone
ethyl acetate
butanol
Dowanol PM (1-methoxy-2-propanol)
ORM regenerated *
toluene

Sample volume: 6.0 l (flow rate: 0.60 l / min, measurement time: 10 min)
Extractant: 1 ml of n-nonane

Emissions [mg / m³]

Emission measurement 4 (GC)

Date: 26. 07. 93
Production Type: Screen Cleaner (REMCOLIN)

ingredients
(only solvents, alphabetically):
Aethylacatat
diacetone
Dowanol PM (1-methoxy-2-propanol)
ethoxypropylacetate
Solvent naphtha

Sample volume: 30.0 l (flow rate: 1.50 l / min, measuring time: 20 min)
Extractant: 1 ml of o-xylene

Emissions [mg / m³]

comment

No paint production took place on the day of sampling. The filling works for the Production of the sieve cleaner were not directly under the suction hoods, but in the middle of the production hall. This explains the inadequate registration of the Solvent in the chimney exhaust.  

2.4. Summary of emission measurements (FID and GC)

Both the results of the FID and the GC emission measurements, as well as the comparison These two types of determination clearly reflect the discontinuous pollutant emission in the paint-producing operation again. While at the averages of the FID measurements the mass concentrations are still below the officially required values the maximum values already exceed the double. The impact of However, peak concentrations occur only in the GC emission measurements drastically down. Here, at the values of just four measurements, there is already a span in reaches the solvent concentrations of about 0.01 to about 0.31 g / m³ (TA-Luft calls 0.15 g Corg./m³). It should be noted that the GC emission measurements parallel to the filling and mixing work was carried out and thus the concentration peaks very well reflected. To realistic test conditions and thus meaningful To achieve results, it is expedient, from the beginning with fluctuating Crude gas concentrations to work. This will be an adaptation of biology to a prevents constant pollutant concentration.

The average composition of the Exhaust air is listed in Table 1. According to these values that was in the laboratory tried used synthetic exhaust gas mixture prepared. For ethoxypropyl acetate and the mixture solvent naphtha, the proportion in the exhaust gas mixture was increased by about 50%, since must be assumed that these two components in the manufacture of the Screen Cleaner REMCOLIN with optimal suction in higher concentrations occur could be found in the GC emission measurement 4. In addition, the Investigation of the biodegradation of the aromatic hydrocarbon mixture Solvent Naphta especially interesting. By a higher dosage of only in small Concentration (<1% in the solvent mixture) occurring solvent is in the Series of tests also allows a better quantitative evaluation.

Already it can be seen that in the multi-fuel mixture both slightly water-soluble Compounds and sparingly soluble compounds (xylenes, toluene, MIBK, Solvent Naphta) are present. Due to the lower availability in the biofilm of the moist Reactor material must be expected with problems in the degradation of these compounds become. In addition, it is known that in xylene loaded biofilters immediately about 80% the originally existing organisms stop growing.³ That's why it will be necessary at higher load with BTX aromatics a "dry" buffer that keeps the concentrations of these hardly degradable compounds constant and as small as possible (see sections 3.3.4 and 4.3.6 to 4.3.8.).

3. Planning and construction of the bioreactor plant with integrated Buffer system on a laboratory scale 3.1. General to the method choice

Both the composition of the raw gas and the operational conditions Need to be in the selection of a suitable process for the decomposition of organic solvents be taken into account by microorganisms. The following points must be to be checked:

• 1. nature, solubility, toxicity and concentration (usually <2 g / m³) of the pollutants,
• 2. Biodegradability of the main components; all pollutants have one be accessible to aerobic degradation.
• 3. Type of exhaust gas source (point, lines, areas, single stream or electricity from different those sources),
• 4. raw gas stream (exhaust gas volume flow),
• 5. Raw gas concentration (fluctuations in the raw gas?)
• 6. Concentration peaks in mg / m³
• 7. What is the required clean gas concentration?
• 8. Rohgasfeuchte,
• 9. raw gas,
• 10. Corrosivity of the raw gas,
• 11. dust or fat content in the raw gas,
• 12. Self-ignitability, explosion limits,
• 13. Plant operating times (emission cycle); Are there any shutdowns except for weekends ?; Does the operation work continuously?
• 14. What is the size of the area on the premises that is the location of a filter system? can be used?

Many of the questions listed here were able to take into account the results of the Emission measurements in Section 2.2. and 2.3. get answered.

The composition of the exhaust air, as well as the main characteristics of each Components are listed in Table 1 (chapter 2.4.). As volume flow under Betriebsbe conditions (point 4) were measured about 20000 m³ / h. As regards point 2, it should be noted that one between readily biodegradable, so-called biogenetic compounds and anthropogenetic compounds ("xenobiotics"). The latter possess sometimes structures that are not found in nature, which makes their mining difficult or even impossible.  

In principle, biodegradability of all components occurring here is to be reckoned with.¹ To this problem, reference should be made at this point to a few selected investigations which have already been described in the literature. ^3-7 Some of the research work focuses specifically on the degradation of toluene as a moderately degradable component ^8-10 , or xylene as a poorly degradable solvent¹¹. With diffi culties but must be expected in the microbial degradation of the aromatic hydrocarbon mixture solvent naphtha, which consists mainly of mesitylene and alkylated Benzolabkömmlingen (the exact composition can be found in the corresponding DIN data sheets). Literature for the degradation of this mixture, as in the case of Dowanol PM and ethoxypropyl acetate, is not yet available.

The required raw gas concentrations are described in TA-Luft. In the case of this Work examined exhaust air is a mixture of organic matter several classes. Here, at a mass flow of 3 kg / h and more, the masses may concentration in the exhaust gas does not exceed 0.15 g Corg / m³. The exhaust gas collection in paint-producing operation via two suction devices, d. H. it will merged various exhaust sources and discharged through chimney to the outside air. The difficulty with the single-shift operation is the discontinuous production and shutdowns on weekends and Holidays. By the occurrence of concentration peaks within the daily Working time of eight hours will be an adaptation of the microorganisms to the pollutants also difficult. During downtime without adding nutrient solution better with can be bridged by a biofilter, one by the occurrence of Concentration peaks required buffer effect better by a bioscrubber too achieve. Buffering is the reduction of the influence of Concentration fluctuations in the raw gas on the clean gas concentration. The decision between the two emission control method is in the present case both by the Poor solubility of some components, some of which occur in higher concentrations, as well as by an extremely low humidity of the exhaust air in favor of the biofilter Humidification system decided. As a disadvantage must be mentioned here that no optimal pH control and no product removal is possible. One possibly occurring pH reduction caused by the microbial degradation of ethyl acetate (ethyl acetate) is to be expected, but can by addition of basic additives such as CaCO₃ be counteracted. Against the use of a classic biofilter, d. H. of a surface or floor filters on the relatively narrow company premises of the company, speaks the large area Design of these filter types. In addition, by the strongly fluctuating raw gas Concentrations, the upstream of a buffer system are considered inevitable. To remove the particulate exhaust components must in addition either a Dry separator (eg fabric filter or deposition chamber) or a wet scrubber  (Venturi scrubber or packed column) upstream of the actual exhaust gas purification become.

Taking into account the selection criteria referred to in this section, a Designed laboratory laboratory whose core is a bioreactor by a Humidification circuit has a constant moisture content of over 95% relative humidity in the Filter allowed. The planning and design of this bioreactor plant is the following Dedicated chapter.  

3.2. Planning of the bioreactor test facility 3.2.1. The bioreactor

The bioreactor consists of a biologically active material, which with a organic carrier substance is mixed. The moisture content of the filter medium is kept constant, regardless of fluctuations in the ambient temperature, the Humidity and the volume flow. This is done by a sprinkler circuit reached, the daily only the evaporated liquid content must be supplied. The System is thus almost wastewater free. The supply of microorganisms with nutrient salts and the sampling for the verification of the circulating water can also on the Sprinkler equipment done. Special attention comes to the regulated, periodic humidification of the filter material. It must be prevented that the Filter bed dries when dried and breaks or that by overhydration Filter zones arise that are no longer adequately ventilated and therefore anaerobic. The bioreactor is better for less soluble than a scrubber Pollutants that occur in higher and fluctuating concentrations suitable. in the Compared to the classic biofilter, the advantages of the reactor lie in the better ones Suitability for dry exhaust gases as well as in the space-saving design. In some Investigations in bioreactors with higher construction a better Stollübergang for poorly soluble substances, which are due to the higher velocity in the Reactor with smaller cross-section and proportionally increasing height recycled can be. This structural design also allows an uncomplicated expansion of Cleaning system for two-stage reactor. In this way z. B. a more successful microbial recovery heavy or slow degradable compounds such. B. Xylenes or solvent naphtha.

Since the exhaust air of the paint factory at a temperature of 18 ° C is very dry, must at their Humidification to a relative humidity of at least 90% with an adiabatic Cooling to about 6 ° C can be expected. This fact and the fact that the Optimum biological activity in the filter is about 30-40 ° C, makes an increase in the Exhaust gas temperature necessary. You can with a temperature increase by about 10 ° C with expect a doubling of activity.

The developed laboratory test facility and the bioreactor are shown schematically in Figs .

3.2.2. The filter medium

The following criteria must be considered when choosing the filter medium, or The following properties are a prerequisite for optimal microbial degradation:

• - species-rich microorganism colonization
• - Structure (eg a uniform structure ensures a constant filter burden.)
• - void volume
• - good sorptive properties; the sorption capacity in case of bad water Soluble substances can be prepared by admixing adsorbents (eg coal) are increased.
• - Large specific surface area (high growth area) of the carrier material
• Low resistivity; this results in a low pressure loss (good: about 2 mbar / m)
• - good water retention capacity; Humidity at least 30%; optimal water content about 50%.
• - lifetime
• - low odor, d. H. Prevention of secondary emissions
• - good nutrition with nitrogen, phosphorus and trace elements.
• - low maintenance costs
• - low cost.

For a sufficient growth of heterotrophic bacteria are still the Absence of inhibitors, oxygenation, the right pH and a constant temperature important parameters.

So far, the following filter materials in the degradation of organic solvents with Success used:

• - Compost / Rindenmischung⁵
• - bark humus / root wood ^{12, 13}
• - peat / compost ^{4, 14}
• - Peat / Rindenmulch⁸
• - bark humus / Rindenmulch⁹.

As an active biological material compost from biotope waste and the Green waste from the city of Heidelberg used. The life of the compost, me provided by the waste composting plant Heidelberg-Wieblingen, was approx. 5.5 weeks (~ rump 2). The chopped, fresh compost consisted of different large units and originally had a strong spicy to cheesy odor. For better retention of water and to improve the structure was in the first Test series as support material small granular lava (particle size approx. 3-5 mm) mixed in. In  In the later test series, the lava was replaced by ceramic rings (EHEIM-EHFIMECH; about 1 cm) replaced. In addition, raw gas inlet and biological filter material has been added an approximately 5 cm thick layer of coarse-grained lava (grain size about 10 mm) introduced. The Use of natural support materials in addition to the obvious advantages, such. B. adequate nutrient supply and microorganism colonization, but the disadvantage of Material decomposition (see 4.1 and 4.3.1.). In this case, the pressure drop in the reactor so high that a replacement of the filter medium is recommended.

In order to maintain the pollutant metabolism of the microorganisms must in addition to the supply of carbonaceous substrates and oxygen and the supply be ensured with inorganic nutrients. Because of the strong nitrogen deficiency that in the degradation of BTX aromatics was observed alongside the N supplier (NH₄) ₂SO₄ additionally (NH₄) added NO₃. The added inorganic nutrient solution has the following composition:

OTTENGRAF ET AL¹⁴ add about 200 cm³ to a similar solution at 0.2-1.0 mmol solvent a composite nutrient solution. In this first test series, the Moisturizing water is added twice a month, 100 ml each of the solution. In the In later series of tests, it was found that the addition of weekly 1 × 50 ml Advantage of a lower back pressure in the reactor has. Also, the degradation of the hard soluble components by adding smaller amounts of liquid in general less impaired.

On an inoculation of the biologically active material with specially cultivated micro Organisms was initially omitted. It can be assumed that the biological Colonization of fresh compost with a mixed population of microorganisms is sufficient and the nutrient addition can be kept low. Nevertheless, became For comparison in a later test series the filter material with sewage sludge from a Septic tank of BASF AG in Ludwigshafen inoculated (see 4.3.2.).  

3.3. The bioreactor plant in the laboratory scale 3.3.1. Supply lines and conditioning of the compressed air

For the separation of oil components, the compressed air is purified via an activated carbon column. A downstream blue gel column also serves as a dust filter. The air lines to the filter should be smooth and large enough to keep the pressure drop (Δp ~ r ^-4 ) low. The pipelines of the reactor plant consist of brass tubes with 6 mm diameter. For humidification up to the saturation limit, the air flows through three fry washing bottles filled with water, which are in a bath tempered to 28 ° C. The device to compensate for evaporation losses in the filter, as well as the removal of any resulting excess water is not made in this moistening unit, but directly on the separate humidifying the Bio reactor. At the wash bottles themselves, only by the air saturation resulting daily H₂O loss is compensated (see also 4.2.). Because of the expected adiabatic cooling of the air during their humidification on the one hand and the optimum exhaust air temperature of 30 to max. 50 ° C in the reactor on the other hand, the moist air must be heated. This is done in a downstream and wrapped with a heating tape tube in which the air is brought to a temperature of about 50 ° C before the exhaust gas metering. Furthermore, all subsequent pipes to the reactor are wrapped against cooling with insulating material.

3.3.2. Dosing of the exhaust gas

For all test series, a solvent mixture of defined composition used. It has been shown in various studies that of the Results obtained by microbial degradation of individual components, not on the degradation behavior in the mixture can be concluded. For this reason, a Gas mixture produced, which in its composition of the exhaust air of the paint factory equivalent. The dosage of the by in section 2.3. listed emission measurements determined exhaust gas mixture can be by carburetor technology or via injection with a Dosing pump (plunger pump) with subsequent evaporation in the past flowing air take place. In the here planned plant u. a. for cost reasons one Metering device used, which consists of a motor which drives a coil forward and a gas-tight syringe, which through the coil their contents in a defined manner in the Pipe system issues. The temperature at the solvent inlet is approx. 50 ° C. The Injection volume varied between 100 and 1050 μl / h. As the pollutant concentrations in the real exhaust air of the paint shop constantly fluctuate, was in the test series no strict injected constant amounts, but there were deliberately small fluctuations (eg  selective concentration peaks). To get reliable statements, was but the gas inlet concentration between the daily analyzes of raw and clean gas always kept constant.

3.3.3. interpretation

With a standard volume flow of approx. 20000 m³ / h, the measured FID maximum values correspond to 400 mg _org. / m³ of a pollutant load of 8 kg / h. Correspondingly, the mean mass concentrations were 130 mg C _org. / m³ mass flows between 2 and 3 kg / h. The maximum GC emission value of 310 mg C _org. / m³ corresponds to a load of 6.2 kg / h (mean mass concentration of 100 mg _org. / m³).

Biofiltertechniksanlagen (semi-technical pilot plants) work usually with a volume flow of the gas phase of 1 to 20 m³ / h. Pilot plants, with one part stream of the entire volume of gas (large-scale pilot plants), can have currents of several thousand m³ / h. The laboratory test facility was opened a volume flow of 0.1 to max. 1.5 m³ / h designed. The previously performed test rows were carried out with a raw gas volume flow of 0.20-0.24 m³ / h. The the highest delivered load was 1 mg / h and thus the maximum raw gas load 4 mg / m³.

The actual bioreactor consists of a glass cylinder, which is gas-tightly connected to the aluminum reactor pot via a flange (see Fig. 2).

Dimensions:

Height: 800 mm
Diameter (inside): 140 mm
Void volume: 12.3 l

The residence time in the filter layer should be a maximum of 60 seconds. This time also determines the height of the filter material in the reactor column.  

3.3.4. The buffer system

The pre-separator is a switchable as needed buffer system, which serves both the compensation of occurring concentration peaks as well as a permanent reduction of pollutant load. For this purpose, an organic adsorber based on styrene-divinylbenzene was used in two series of tests (addition of solvent about 1000 μl / h). The material of the company SUPELCO (AMBERSORB XEN-572) is produced in a patented pyrolysis process and can be easily desorbed with tempered compressed air after adsorption. This desorption should take place at night during downtimes or even at weekends and should be completed at the beginning of the shift at best. These test results and the optimal "buffer program" are discussed in sections 4.3.7 and 4.3.8. reported in detail. The previously published in the literature studies on the use of buffer materials all have the disadvantage that on the one hand almost exclusively for the degradation of Lackierereiabluft unwanted activated carbon was tested, and on the other hand, that the adsorption processes described all highly sensitive to moisture - and thus for the Inappropriate solvent mixtures are unsuitable - are ^15,16 . In addition to various types of activated carbon only silica gel and Al₂O₃ have been tested, which are also ineffective under humid conditions. Although it was possible with an activated carbon, the toluene concentration of 0-1.5 g / m³ buffered to 200-800 mg / m³ (equivalent to 13- 53%), but since in this experiment toluene was used as the only pollutant component, this result has no validity for real Lackierereiabgase¹⁶.

3.3.5. materials

The bioreactor consists of a better observation of the filter material state of a Glass cylinder. The sieve plate on which the filter material rests, and the height-adjustable Reactor covers were made of Teflon. The reactor bottom consists of a Aluminum pot in which the running irrigation water collected and for Umwalz pump is forwarded. Here you can also remove the condensate. The Zuleit pipes with an internal diameter of 6 mm and all other metal parts are made Made of brass. The connection of the individual parts takes place via Swagelok fittings.  

4. Cleaning simulated paint shop air with the bioreactor experimental plant 4. 1. Filling of the reactor and adaptation of the biologically active filter materials 4.1.1. Filling for test series 1 and 2

The biologically active compost (compost 2) had one at the time of filling Temperature of about 32 ° C. In the first series of tests, the filling level in the reactor was after Mixing of approx. 1.5 l compost with fine-grained lava and moistening with 200 ml Water 52 cm (with the lowest layer of coarse lava: 57 cm). After three hours was the temperature in the reactor has dropped to 24.5 ° C, which u. a. also on a decline in indicates biological activity. The outside temperature fluctuated around 23 ° C during this time. The air flow was controlled to 0.2 m³ / h (2.8 l / min) and with the heating tape to a Temperature of about 50 ° C, which was measured before the solvent inlet, heated. Two Days after filling the excess and foul smelling water was drained and replaced by 200 ml of the inorganic nutrient solution (3.2.2.). In addition, were 300 ml of water added via the sprinkler. After optimization of the sprinkler cycle, d. H. after setting the circulation pump, which is a constant condensate guaranteed quantity in the reactor pot, the bioreactor stood still for 14 days. While During this time only the evaporated liquid components were filled up. The temperature in the reactor fluctuated in this time around 25 ° C and the filling height was at the end of Adaptation time dropped by 6 cm. The filled with filter material reactor volume was now 7.1 l. The course of the filling level during test series 1 in comparison with the filling a later test series is recorded in Table 2. The bioreactor exhaust air had after 14 Days a slightly spicy, not unpleasant odor and the initially still cloudy Circulating water was now clear. Shortly before the beginning of the first test series was tried, the To determine residence time by blowing in smoke. This was not possible because of the Smoke from the filter material was absorbed. Therefore, the residence time as well as in the later test series of subjects on the first smell perception of a water insoluble solvent (toluene) can be determined. It was about 100 seconds.

4.1.2. Filling for test series 3 to 7b

Analogous to the first filling, the biologically active material consisted of compost of the Composting plant Heidelberg / Wieblingen. This time, however, a rent that was a stand time of about 5.5 weeks (rump 2) showed with a second rent of 10.5 weeks  time mixed. As the lowest layer 3 cm coarse-grained lava were introduced, on the a 36 cm high mixture of 5.9 l compost and 2.1 l ceramic rings (EHEIM / EHFIMECH) was applied. The covering layer was compost with 0.6 l of coarse-grained lava mixed (height: 18 cm). The temperature of the compost was initially 42-43 ° C and after mixing with the carrier materials and filling in the reactor only 32 ° C. Three hours after filling, the temperature in the reactor was around 27 ° C set. As wetting and irrigation water 950 ml of activated sludge (pH: 7.0) from a retention basin of BASF AG in Ludwigshafen. Few Hours after filling the recirculating water had both in the reactor pot and in the Filter and in the hose connections of the Sprieselungskreislaufes voluminous sludge flaked off. Therefore, the liquid was drained again, the disturbing Solids separated and the clear, amber-colored filtrate back into the circulation filled. The course of the filling level in the reactor in comparison with the first filling is in Table 2 listed. It can be seen that the second filling a much cheaper Behavior with respect to the service life shows. Although the filling height falls during the first Day by almost 3 cm and within the first two weeks even by about 6 cm, then But it shows a constant small decline and finally reaches a final value, the until March '95 has not yet fallen below (about 10 cm / year). This is due to the stability of Ceramics rings, in contrast to the lava during reactor operation do not succumb to change. So here a final value of the filling height is reached, as soon as the Structural changes that the active compost material experiences are completed.

Table 2

Decrease of the filling level in the bioreactor with different filling

4.2. The measurement series at a glance

During the test series, the following parameters were checked daily and recorded:

• - filling height
• - air flow
• - Smell of the exhaust air at the clean gas outlet
• Temperature:
  the water bath in which the washing bottles are located
  of the heating tape
  in the reactor
• - Watering program of the circulation pump
• - pH, odor and appearance of the circulating water

In addition, the Spritzer head was rotated daily to prevent the Dripping water creates breakthroughs in the filter. The filter material (filter cotton wool) of the Filters for the irrigation water had to be in the first week daily, later until the end the test series can not be changed.

water consumption

The water consumption in the wash bottles (about 100 ml / day) and in the reactor pot (about 30 ml / day) had to be compensated daily. The relative moisture content within the Reactor was determined with a terrarium hygrometer of DEHNER, directly over the filling material was attached in front of the clean gas outlet. The relative humidity was over the entire series of tests over 98%. Throughout the test series averaged 700, based on a maximum air saturation at 27 ° C per week ml of water consumed. Fluctuations in the range of 650 to 760 ml per Be determined week. A maximum of 900 ml / week was reached in August, when the outside temperature had risen to over 30 ° C. By heating the buffer During the downtimes in the later series of tests, a system could produce water more consumption of about 20-40 ml / week to be registered.

The following GC measurements were performed routinely:

• - calibration measurements with the "synthetic exhaust gas mixture" in the range 0.5-5.0 ppm, or 5.0-15.0 ppm
• Circulating water before and after the end of the solvent addition, around the dissolved in the irrigation water and not broken dissolving average shares (see 4.3 a 1.)
• - raw and clean gas measurements; in the test series with buffer also raw Gas measurements before and after passing through the buffer system.

The concentrations of the individual pollutant components during the test series Overall, they varied over the following areas:

1. acetone 355-3000 [μg / d] and 185-1565 [μg / m³] 2. toluene 170-1450 [μg / d] and 89-756 [μg / m³] 3. ethyl acetate 140-1200 [μg / d] and 73-626 [μg / m³] 4. Methyl acetate 85 65-550 [μg / d] and 34-287 [μg / m³] 5. n-butyl acetate 60-500 [μg / d] and 31-261 [μg / m³] 6. n-butanol 56-470 [μg / d] and 29-245 [μg / m³] 7. Isopropanol 15-125 [μg / d] and 8-65 [μg / m³] 8. techn. xylene 14-120 [μg / d] and 7-63 [μg / m³] 9. Dowanol PM 13-110 [μg / d] and 7-57 [μg / m³] 10. MIBK 10-75 [μg / d] and 5-39 [μg / m³] 11. ethoxypropyl acetate 2-20 [μg / d] and 1-10 [μg / m³] 12. Diacetone alcohol 2-20 [μg / d] and 1-10 [μg / m³] 13. Solvent Naphta 1-5 [μg / d] and 0.5-3 [μg / m³]

The following table gives an overview of all test series:

Table 3

Overview of the series of measurements 1 to 7b

4.3. Measurements with the bioreactor test facility 4.3.1. Results: Test series 1

As expected, the water-soluble components were not at the beginning of the test series 1 detectable in the clean air, since they are immediately absorbed by the circulation water and there too were dismantled. Therefore had to regularly samples from the Sprieselungskreislauf taken and examined by gas chromatography. At the end of the first Week in circulation water after addition of daily 800 .mu.l of solvent only the amounts of pollutants listed in Table 4 are detected.

Table 4

Solvent concentrations in the sprinkler circuit and separation efficiency at the end of the first test week (test series 1)

The test results from test series 1 with regard to the degree of separation and clean gas concentrations of the not completely degraded solvent components are summarized in the following table. Fig. 3 shows the GC diagrams of the crude and pure gas from 26. 1. 94 (4th week of the experiment).

Since the values listed in the tables can not resist the actual course of the day reflect should be noted at this point that the clean gas Konzen concentrations during a day (9 ⁰⁰ -17 ⁰⁰⁾ decrease continuously, analogous to the course over the period of each of Week. This behavior becomes clearer in the pictorial representation. Fig. 4 and 5 show the crude or pure gas concentrations of toluene or xylene during test series 1. One clearly recognizes the lower separation rates on Monday of a test week and the drop in clean gas concentrations later in the week.

1. Toluene: (test series 1: without inoculation / solvent supply: approx. 100 μl / h: 380 μg / m³)

A similar behavior to acetone was for isopropanol, methyl acetate and dowanol PM found: in the clean gas, these components are ver only below the detection limit ver to step. During the first two weeks but could regularly small amounts of this Substances in circulating water are detected (see Table 4). The biological degradation These components are then finally with a separation efficiency of over 98%.

In contrast, with ethyl acetate, n-butanol, n-butyl acetate (n-BAC) and ethoxypropyl acetate (EPA) in less than a week even in the circulating water no or only minimal Traces detectable. These solvents are very fast and with a separator Degraded over 98% in the reactor.  

2. techn. Xylene, total: (test series 1: without inoculation / solvent supply: about 100 μl / h, 380 μg / m³)

From the less soluble component Solvent Naphta (SN), which is only in minor concentrations is present (<0.1%), no quantitative statements are possible. One can while test series 1 still recognizes traces of this substance in the clean gas, however a qualitative decrease recognizable. The degradation behavior of Solvent Naphta is in the Test series, which work with higher gas inlet concentrations examined.

The degradation behavior of MIBK could also be reduced due to the low concentration in the Mixture (<1%) not yet explored; however, now is a slower degradation than in the more water-soluble components, such. B. Dowanol PM, detectable.  

3. acetone: (test series 1: without inoculation / solvent supply approx. 100 μl / h: 380 μg / m³)

The pH of the irrigation water, which continuously passes through the filter material flows, was after about 10 days 7.6 and after another 14 days 8.0. In comparison to it It should be noted that the water used in the filling of the reactor, as well as the Inorganic nutrient solution and the solvent mixtures have a pH of only 5.5-6.0 has. Increasing the pH promotes the microbial activity that contributes to its optimum has a pH between 8 and 9 and decreases drastically at higher values.

The temperature in the bioreactor during test series 1 was between 25 and 28 ° C.

At the end of the 5th week of testing, the filter material was tried with a wooden stick Loosen, to counteract an increasing pressure loss. These efforts However, rum had the opposite effect, because the lava particles, which are already strong Zer had decomposed, and their initial loosening effect was lost went. The compost parts were also slightly changed in their structure; d. H. now over smaller particles weighed, which together with the decomposed lava caused a pressure increase in the Reactor caused. Due to this, the irrigation water could no longer unhindered flow through the filter material and accumulated on the filter surface. Farther it was possible to detect different breakthroughs in the material (channeling). The growing pressure loss in the reactor was not the result of a biomass increase, the so-called "clogging". To restore the optimal conditions, the Plant shut down for 8 days except for the passage of humidified and tempered air. During this time, the structure of the filter medium improved again and it was be begin to add solvent again. The effects of the attempted material However, the resulting breakthroughs had such a negative effect on relaxation Degradation behavior of the microorganisms that the test series had to be terminated.  

4.3.2. Results: Test series 2

This series of experiments was carried out analogously to the first test series; the only difference was the inoculation of the filter material with sewage sludge from a suitable clarifier of BASF AG in Ludwigshafen. The test results are shown in Figs. 6 and 7. Degradation was consistently with better degradation rates than in the test series without inoculation. However, striking differences can only be noticed in the first week. Thereafter, only a marginal improvement in biodegradation takes place. For the sparingly soluble pollutants, the degradation rate in Test Series 2 is up to 10% higher. As with the first test series, the degree of separation of the readily water-soluble solvents was close to the 100% mark. In addition, no significant amounts of solvent in the circulating water had to be detected from the beginning. A comparative presentation of the test series 1 and 2 are shown in FIGS. 8 and 9.

1. Toluene: (test series 2: with inoculation / solvent supply about 100 μl / h, 380 μg / m³)

2. techn. Xylene, total: (test series 2: with inoculation / supply approx. 100 μl / h, 380 μg / m³)

3. acetone: (test series 2: with inoculation / solvent supply approx. 100 μl / h; 380 μg / m³)

4.3.3. Results: Test series 3

In the course of this series of tests, those with greatly increased and very fluctuating concentrations (2 times 2 hours: 1000 μl / h), an increase of biomass was detected be noticed both in the filter material and in all supply lines of the circulating water. Despite the increase in biological activity with respect to easily degradable, water soluble ingredients, the breakdown efficiencies are problematic Components unsatisfactory. For both toluene and xylene, the maximum degree of separation achieved only after the fourth week of testing. With values in the Range of 30 to 40%, no effective cleaning can be achieved for these substances.

To ensure that the greatly reduced degradation performance compared to the two previous series of tests (100 .mu.l / h) alone by increasing the pollutant concentration and not caused by mechanical effects in the filter was the Reactor from the 6th Week again charged with 100 .mu.l of solvent per hour. The instant Increasing the degrees of separation on the values of the last series of tests, clearly shows the preserved biological activity. Furthermore, it can be seen that despite a Overloading with BTX aromatics, on the one hand, the water-soluble components continue and a rapid recovery of biology takes place as soon as the BTX concentration is lowered again.

The problem of equalization of concentration peaks becomes quite clear in this test series especially clear. An effective cleaning of the problematic components will only can be realized by interposing a buffer system.

During this test series could also be a delicate balance between Wasserge just be detected in the filter material and separation efficiency of the BTX aromatics. Either with addition of nutrient solution as well as with technical increase of the irrigation water, the clean gas concentrations of these components increased immediately (see clean gas values on 26. 04., 03. 05. and 16. 05. 95).

Figures 10 and 11 show the results from the clean gas measurements of toluene and xylene. The separation performance is shown in Fig. 12. As with the previous experiments could be registered by the water-soluble components in spite of He increase the addition of pollutants except traces of acetone no residual amounts, ie their degradation is consistently over 95%.

The pH of the reactor water was about 7.0 in test series 3.  

1. Toluene (test series 3: solvent feed: 2 × 2100 μl / h, in total 4200 μl / d 4000 μg / m³)

2. techn. Xylene, total: (test series 3: supply: 2 × 2100 μl / 2 h, total 4200 μl / d, 4000 μg / m³)

3. acetone: (test series 3: solvent supply: 2 × 2100 μl / 2 h, total 4200 μl / d, 4000 μg / m³)

4.3.4. Results: Test series 4

In this series of experiments, the pollutant concentration was compared to the first both test series quintupled. The solvent feed of 0.525 ml / h (3780 μg / m³) was without the strong concentration variations that characterized Test Series 3, without before switched buffer system and tested at the same filter height. In the case of water-soluble Components (acetone, isopropanol, methyl acetate, ethyl acetate, n-butanol, Dowanol PM, n-BAC and EPA) required the microorganisms even at the elevated pollutant Concentrations only approx. 24 hours to achieve separation efficiencies of <99%. Both Problematic Toluene, Xylene and Solvent Naphta will be one with the first test series comparable degree of filtration achieved only after the fourth week. After this start-up period, which is mainly due to the toxic effects of xylenes, is the sum the clean gas concentrations below 150 μg / m³. The solvent components, which are still in the Pure gases remain composed of toluene, xylene and solvent naphtha. Here would in practice to avoid the relatively long start-up phase and possibly so associated increase in the filter layer at the beginning of the Commissioning be made.

A difference to the first two series of tests is the behavior of the degradation performance in the Determine the course of an operating week. While in test series 1 and 2 the separations just always increase from Monday to Friday, d. H. the pure gas concentrations can fall here also an increase of the clean gas values are noticed.

During this test series, it was checked whether accumulation of pollutants in the Re actor material that could distort the measured clean gas concentrations. To The reactor exhaust was a few minutes after an eight-hour addition of solvent under examined. The gas chromatogram showed no pollutants, so that assumed can be that the reactor flowing through the Lackierereiabluft within the dwell time is reduced and no enrichment takes place. This result showed all the rest Clean gas measurements taken during downtime.

Furthermore, the influence of a longer downtime was tested. After the second week of operation, the filter was charged with conditioned air for 16 days without addition of nutrients. When restarting the reactor could already after z. T. less than 24 hours the old mining rates are reached again. The degradation of acetone reaches the maximum value of over 98% with a high starting value (about 81%) within this time. Toluene is recycled as well after 24 h as before the break, albeit with a weaker starting value (about 27%). Only with xylene, the degradation rates during the first three days are slightly lower than the previously weak values before the break. The highest degree of separation of xylene is reached only after one month. A longer downtime can thus be bridged with relatively short adaptation time and without intervening losses in terms of separation efficiency. The results for the three most important components are listed in the following tables and in Figs. 13 to 15. Fig. 16 gives a comparison between the test series 3 and 4.

1. Toluene: (test series 4: solvent feed: 525 μl / h, in total 4200 μl / d, 2000 μg / m³)

2. techn. Xylene, total: (test series 4: solvent feed: 525 μl / h, total 4200 μl / d, 2000 μg / m³)

3. acetone: (test series 4: solvent feed: 525 μl / h, total 4200 μl / d, 2000 μg / m³)

4.3.5. Results: Test series 5

In this test series, the pollutant concentration was doubled again to reach the limit of the degradation capacity of the reactor (1050 μl / h). As already seen in test series 4, at higher pollutant levels, the clean gas concentrations of the BTX aromatics increase over a week and even during the daily addition over 8 hours. This is to be regarded as a clear indication of the achievement of maximum mining capacity. While the water-soluble components are still excellently degraded (3 g / m³ acetone with an average separation rate of 88%, all others <95%), the poorly water-soluble, aromatic components only offer maximum separation rates of about 30%. In this test series, the clean gas values for the individual components of the technical xylene could be split up. The most important results are listed in the tables and Figs. 17 to 19.

The delicate balance between water content in the reactor and clean gas concentrations The presence of BTX aromatics was also evident in this series of experiments. As each Addition of water to the reactor negatively affects the separation efficiency with respect to the heavy water-soluble compounds, an addition of weekly was only 50 ml Nutrient solution of the previous addition of 100 ml after two weeks preferred.

1. Toluene: (test series 5: without buffer / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³)

2. techn. Xylene, total: (test series 5: without buffer / solvent supply: 1050 μl / h, 4000 μg / m³)

3rd techn. Xylene, separated: (test series 5: without buffer / solvent feed: 1050 μl / h; 4000 μg / m³)

4. acetone: (test series 5: without buffer / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³)

5. Solvent Naphta: (test series 5: without buffer / feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³)

4.3.6. Results: Test series 6

For the same amounts of raw gas, a buffer system was switched between the actual bioreactor. On the one hand, the increase in degradation power and, on the other, the desorption behavior of the individual components with respect to the buffer material should be investigated. For this purpose, the synthetic exhaust gas mixture was passed completely over 30 g of a buffer substance. By measurements both directly after the buffer system and after passing through the filter layer - ie at the clean gas end - statements could be made both about the desorption behavior and about the degradation performance with respect to the components. The results are shown in the following tables and Figs. 20 to 22.

Already about five minutes after the buffer was switched on, no pollutants were left measurable. The measured clean gas values of the first test day (22. 09.) still stir the last series of tests, so that in the case of sparingly soluble components toluene and xylene first breakthroughs on the buffer only on 24. 09. were detected. In contrast, acetone was For the first time after about 24 hours the buffer output is detected again. Next to her outstanding performance even under very humid conditions is the good adsorption behavior with regard to the critical, poorly water-soluble constituents a great advantage the buffer substance used. While the latter components are adsorbed for a long time and Finally, "metered" are issued, the biologically easily recoverable water soluble components of biology are readily available. A detailed overview of the Desorption behavior of all solvent components are given in Table 3. In the table are both the first breakthroughs of the components at the end of the buffer (bold numbers!) as also characteristic maximum values (underlined numbers!) highlighted.

The pure gas measurements contained in this test series in the worst case high Amounts of toluene and technical grade xylene, as well as traces of acetone, MIBK and solvent Naphtha. Once the aromatic solvents have broken through the buffer it They also seem relatively fast (a few hours) at the clean gas outlet. Your removal in the Biofilter is therefore insufficient due to the sudden increase in concentration. By contrast, the sudden increase in acetone concentration in the filter has little effect Negative on the acetone separation, what a trouble-free metabolism also high amounts of this easily soluble substance speaks. Isopropanol, methyl acetate, Ethyl acetate, n-butanol, Dowanol PM, n-BAC and EPA ceased to be pure detected!

During the entire series of tests, the clean gas values were in the afternoon - after 14 Clock - measured higher than the corresponding measurements at 10 o'clock in the morning. Also was clearly the moment of saturation of each component and thus her Breakthrough visible at the end of the buffer. If this time with the solvent supply  During the operating hours, this meant a strong concentration fluctuation, their leveling in the following series of tests by regeneration measures is sought at the buffer.

Figures 20 to 22 show the degradation of the most important components. As in all test series, for components that were undetectable and therefore not listed in the tables, a separation of well over 90% applies.

In direct comparison with test series 5, the performance of the buffer system becomes particularly clear. There, the degrees of separation of both toluene and technical xylene could not exceed values of 30 to 38%. At the same inlet concentrations, on the other hand, separation rates of constant over 60% were achieved when a buffer was connected upstream. The fact that this result can be increased is shown by the excellent degradation rates (80-100%) in the first two weeks. For this purpose, however, an effective regeneration of the buffer material during the downtime must be developed, which counteracts exhaustion of the buffer capacity. The results are reported in Test Series 7. An interesting comparison between the test series 5, 6 and 7, which worked with the same gas inlet concentrations, Fig. 27 in Chapter 4.3.7.

1. Toluene: (test series 6: with buffer / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³)

2. techn. Xylene, total: (test series 6: with buffer / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³)

3rd techn. Xylene, separated: (test series 6: with buffer / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³)

4. acetone: (test series 6: with buffer / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³)

5. MIBK: (test series 6: with buffer / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³)

4.3.7 Results: Trial 7a

In this series of tests was attempted taking into account the experiences and results from the previous investigations an optimal cleaning concept with buffer set and buffer regeneration. For this the following points were optimized:

• - Buffer quantity: 35 g (corresponds to an increase of approx. 15%)
• - direct partial flow: approx. 40% of the raw air is led directly into the reactor
• - Buffer regeneration: Heating: from room temperature to max. 55 ° C (eight Hours); Heating the buffer material during the Downtime at intervals; in between 2 times 30 minutes heating break to avoid con zentrationsspitzen.

The buffer material works well under the humid ones required by biology Conditions and possesses the property, just the problem for the microorganisms Atisches solvent components and retain them in optimal regeneration dosed and thus easier to deliver degradable. In test series 6, the synthetic Exhaust gas mixture passed 100% over the buffer. This exhausted for a relative the buffering capacity quickly, and the initial lack of any solution medium in the bioreactor as a disadvantage. Therefore, about 40% of the Raw gas passed directly into the reactor. So biology was part of the Solvent amount available for metabolism. During test series 7a was one Significant increase in biomass detectable - an indication of good degradation performance.

Finally, the cleaning capacity for the same filter material and reactor volume could and keeping the maximum separation efficiency achieved in Test Series 2 on the 10 times increased! The degradation of toluene and technical xylene was too over 80%. Besides these, with the exception of acetone and solvent naphtha, which were still in Traces on the clean gas side were detectable, all other components were no longer detectable be detected by gas chromatography. These solvents can even in another Increase of their raw gas values still counted with sufficient separation capacity become. The degradation of BTX aromatics toluene and xylene must be considered determinant be considered in the filter design.

In the following tables and Figs. 24 to 27 the results are summarized. Considering that by increasing the reactor volume or the amount of buffer material or by further optimization of the heating program (see 4.3.8.), The cleaning performance is still to increase, the importance of this cleaning concept is clear.

Another important result is the excellent degradation efficiency of the reactor with regard to a longer service life. These long-term results after 10 months of service life, of which a regular eight-hour solvent addition of approx. 1 ml / h took place for 6 months on weekdays, are also included in the tables. The degrees of separation of acetone, toluene and technical grade xylene still reach over 90% after this time. All other components are not detectable in the clean gas. Figure 23 shows the gas chromatograms of the crude and clean gas after five hours of solvent addition on 3/2/95. In the gas chromatograms shown, in the case of pure gas, a product peak emerged which previously could not be detected in the series of measurements which worked with lower pollutant concentrations than 1 ml / h. To elucidate this component, one must consider the following: Aliphatic hydrocarbons are first converted into primary alcohols under aerobic conditions by introducing oxygen. Through the aldehyde stage, these compounds are oxidized to the corresponding carboxylic acids and finally metabolized in the course of β-oxidation as a carbon and energy source. The main intermediates in the degradation of xylene and toluene are usually alkyl-substituted pyrocatechols, which are further degraded by so-called meta-cleavage processes ^{10, 18} . The fact that the component detected by gas chromatography is a genuine intermediate proves the fact that its concentration constantly increases with decreasing clean gas values. In addition, it appears at a retention time, which is still before the appearance of the first raw gas component. A comparison of the retention time with that of 4-methyl catechol showed complete agreement. By GC / MS investigations of the product peak at 7.14 minutes could indeed be ruled out that it is acetone, methanol or another crude gas component, however, it was not possible by mass spectroscopy to detect a substituted pyrocatechol. The same problems arose when trying to measure a sample of gaseous 4-methylcatechol. Neither the mole peak at mass M = 124, which was clearly found in the solution of 4-methyl catechol, nor any other fragment could be found in the gas sample. Here, possibly by a concentration of the substance sought in the clean gas, the mass spectroscopic detection could be made possible.

1. toluene: (test series 7a: with buffer and regeneration / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³))

2. techn. Xylene, total: (test series 7a: with buffer and regeneration / solvent feed: 1050 μl / h, total 8400 μl / d, 4000 μg / m³))

3rd techn. Xylene, separated: (test series 7a: with buffer and regeneration / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³))

4. acetone: (test series 7a: with buffer and regeneration / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³))

5. MIBK: (test series 7a: with buffer and regeneration / solvent feed: 1050 μl / h, total 8400 μl / d; 4000 μg / m³))

4.3.8. Results: Test series 7b

This series of experiments was carried out to clarify the solvent desorption during the buffer refenerierung in the downtime, which was introduced in test series 7a. The buffer system was heated to 55 ° C for eight hours during this time. It was found that even after 30 minutes, when a temperature of about 50 ° C was reached, a concentration peak is released, which has a negative effect on the degradation efficiency in the reactor (see Table 6). To counteract this, a heating break of half an hour was interposed after 30 minutes. During this time, the temperature in the buffer drops again to almost room temperature. Thereafter, it is heated again and with the temperature also increases the amount of desorbed solvent medium again. The maximum values are reached after 2.3 hours. This process is shown in Tables 6 and 8, and in FIGS. 28 and 30.

To even the concentration peak at 2.3 h, was another 30-minute Heating break between 2.0 h and 2.5 h interposed. The positive effect both on the amounts of solvent released by the buffer as well as on the degradation rate in the Reactor is shown in Tables 7 and 8, and particularly clear in the pictorial representations 29 and 30 documented. Finally, arise again with buffer heating after the Heating pauses both with the interposition of one and two cooling Interval no disturbing concentration maxima more, so that the temperature for another 6 h can be maintained at about 55 ° C. Thereafter, the buffer substance is regenerated and thus restored their capacity at the start of operation. The initially rated as negative dis Continuous accumulation of pollutant-laden exhaust air in shift operation, so can at Use of a cleaning system with pre-separator for the benefit of the regeneration of the Buffer material to be converted. Further optimization measures are conceivable. How out Tables 6 and 7 can be seen at 1.9 h almost no reduction in the solution medium concentrations with respect to the Rohgaseinlaßkonzentrationen instead. Another Improvement could z. B. the interposition of a short heating break after approx. Bring 15 minutes. This could cause the relatively high raw gas levels at the buffer output were measured, even further lowered. Likewise it would be possible the buffers regeneration from the beginning to perform in the 30-minute rhythm to Concentrations suppress peaks.  

Table 6

Desorption of the solvent components during downtime (one heating break!)

Table 7

Desorption of the solvent components during downtime (two heating breaks!)

Table 8

Clean gas concentrations of toluene and techn. Xylene after leaving the bioreactor during buffer regeneration with one or two heating pauses (test series 7b: 1050 μl / h)

6. Summary of the results and outlook

The results presented with this work form the basis for the development a powerful and specially tailored to the exhaust air problem of a paint factory biological cleaning system. In addition, a concept is presented, which has general validity for medium-sized companies where shifts are used discontinuous solvent-containing exhaust air is generated. Both the problem of mixing gases from water-soluble, easily degradable solvents and difficult-to-use BWX Aromatics and the equalization of unavoidable concentration peaks could be solved by introducing an upstream and regenerable buffer system.

In the following list, the most important results obtained with the developed laboratory test facility ( Figs . 1 and 2) are summarized.

1. filter material

Proven has a mixture of biologically active compost (Rotte 2) and ceramic rings (diameter about 6 mm) as a stabilizing component in the approximate volume ratio of 3: 1. After a 12-month service life was observed at a constant residence time in the filter no significant pressure drop (see also Table 2, page 24). The high separation rates of the last series of tests with a solvent supply of about 1 ml / h are still achieved after 10 months (see 4.3.7 and Fig . 23). The relative humidity was over 98% in all test series and the residence times in the filter were at a range of 50 to - in one extreme case - 130 seconds usually about one minute.

2. Inoculation of the filter material

Remarkable differences due to inoculation of the filter material with sewage sludge from a suitable basin of BASF AG could only be ascertained during the start-up phase of the first test week. Thereafter, only a marginal improvement in biodegradation takes place (see 4.3.2 and Fig . 8 and 9).

3. Water consumption and nutrient addition

At maximum air saturation at 27 ° C, water consumption was on average for the year 720-740 ml / week. In a hot month, a single peak of 900 ml reached per week. A weekly addition of 50 ml of inorganic nutrient solution with in chapter 3.2.2. said composition, has proven itself.

4. Standstill

A shutdown of 16 days - without the addition of nutrients - could easily be bridged. When restarting the reactor z. T. already after less than 24 hours the old mining rates again reached (see Section 4.3.4.).  

5. service life; Long-term results

The reactor showed excellent degradation even after 10 months Life, of which during the last 6 months on weekdays eight hours approx. 1 ml / h Solvent was added. To date - in March 95 - has not yet noticeable pressure loss in the filter made noticeable. Leave the first four weeks disregarded after the last filling of the bioreactor, the filter material in Lowered by only about 2 cm for 11 months (see page 24).

6. concentration fluctuations in the exhaust gas; Occurrence of concentration peaks

Without a pre-separator no effective cleaning performance could be achieved. The difficult-to-water-soluble components to be evaluated as problematic broke through the filter and appeared quickly at the clean gas outlet. In addition, if the same daily dose of solvent was added to the same daily dose of solvent for toluene and xylene, the rate of separation for toluene and xylene was only about half that for continuous addition (see Figure 16, and items 8 and 9 of this list). An interposed pre-separator must be considered as a prerequisite for satisfactory biodegradation in the present exhaust air problem.

7. Degradation behavior with respect to the individual solvent components

The water-soluble components acetone, methanol, methyl acetate, isopropanol, ethyl acetate, ethoxypropyl acetate, n-butyl acetate, n-butanol and Dowanol PM were unproblematically degraded both in very high and in fluctuating concentrations. When concentration peaks occurred, the degradation of MIBK was delayed. The problem was the degradation of the BTX aromatics toluene and xylene, as well as the aromatic mixture solvent naphtha. However, by using a buffer and applying buffer regeneration, satisfactory deposition could be achieved. However, these components must be considered as determinants of filter design. Compared with other studies ^{15, 16} , where toluene is described as a biologically highly degradable compound, the present results make it clear that results obtained by single-component degradation tests must not be transferred to the degradation behavior of mixed gases. Xylene as a bactericidal substance plays an important role in the biodegradation of the multicomponent mixture investigated here.

8. The buffer material

The granules AMBERSORB XEN-572 from the company SUPELCO has been specially formulated for blends made of light and difficult to dissolve water-soluble components. While the water Soluble components were adsorbed for a long time and finally "metered" were delivered, stand the biologically readily exploitable watersoluble components of biology are rapidly becoming available Available.  

The material produced by pyrolysis of highly sulfonated styrene-divinylbenzene works excellent under the filter in the prevailing extremely humid conditions and leaves to regenerate easily thermally. This newly launched polymer Adsorber granulate is able to reluctant in explosion-protected production halls replace used activated carbon. When using this buffer material could in a Solvent supply of approx. 1 ml / h the crude crude gas concentration of 1416 μg / m³ in places 100 to max. 600 μg / m³ (58-90%) are degraded. For the same pollutant In contrast, only toluene clean gas could be supplied without the interposition of a pre-separator values between 900 and 1200 μg / m³ (15-36%) can be achieved (see test series 5 in 4.3.5. with test series 7 in 4.3.7.).

9. The buffer regeneration

The effective buffering capacity is among other things also on the time that is available to desorb the contaminants depends. The buffer material was during the Downtimes by means of a heating program - in steps that the free exclude concentration peaks - thermally regenerated at approx. 50-60 ° C. Here Further optimization measures are conceivable (see Section 4.3.8.). Also a use operational waste heat during the regeneration step would be possible.

10. Adsorption and desorption behavior with respect to the buffer material

The first breakthroughs of the individual components at the buffer were registered in the following order (see Table 5):
Acetone - isopropanol / methyl acetate - ethyl acetate - toluene - n-butanol / Dowanol PM - MIBK - n-BAC - m-xylene - p-xylene / ethylbenzene - o-xylene - EPA - solvent naphtha.

It should be noted that the percentage of each component in the exhaust gas mixture the Time of Bufferattigung and thus the breakthrough on the clean gas side in one Determined insignificant dimensions. Thus, it can be explained that toluene, although it is from the buffer material is retained much better than z. B. n-BAC, but still faster at the reactor output is detected. The exact indication of the pollutant concentrations, which in the Desorption are found in Tab. 5 on page 51-52. One recognizes that at temporal coincidence of solvent addition during working hours and the Saturation of a component in the buffer strong concentration peaks, sometimes higher Values than the actual raw gas occurred. That's why it became indispensable Regeneration of the buffer to perform during the downtime. If the Regeneration can be further optimized - as in chapter 4.3.8. shown - the desorbed pollutant peaks are finally pushed back completely (see Tables 6, 7 and 8). In the thermal regeneration of the buffer material, the aromatic solvents shares, u. a. due to the higher boiling points, released last. As already under  Point 8, this fact has a positive effect on the separation capacity of the Bioreactor off.

11. Pilot plant or semi-pilot plant

At the end is the evaluation of the results with regard to the construction of a pilot plant, the works with a partial flow of the paint factory exhaust air. At a flow rate of 0.24 m³ / h and a pollutant concentration of 4.0 mg / m³ (about 1 mg / h) were satisfactory Separation results with efficiencies of 90% with a reactor volume of only 7 liters and a pouring height of 0.49 meters can be achieved. These were the actual Biofilter 35 g of the granules AMBERSORB XEN-572 supplied by SUPELCO as buffer on. This results in a filter surface load of 25 m³ / m²h. A pilot plant usually works with values that are around a factor of 10 - and with respect to the crude gas concentrations tration - up to a factor of 100 are greater. Due to the increased filter surface load during Operation of a pilot plant, it becomes necessary to either increase the reactor volume, or after consideration of the operational circumstances and the cost expenditure the quantity to increase on inserted buffer material. This scope in the conception of a Cleaning equipment can be another advantage when using an effective pre-separator be considered. In comparison with pilot tests already carried out in practice, The required reactor volume can be drastically reduced by the buffer insert. Furthermore, of course, regarding the raw materials used and a modernization the forced suction devices in-house avoidance and minimization measure did not want to be ignored in the future. Especially considering that aerobically degradable exhaust gases usually with concentrations up to a maximum of 2 g / m³ still too This aspect of the cleaning concept of the paint factory wins on Meaning.  

7. Literature

• 1. VDI 3477; Biological waste gas / waste air purification biofilter; December 1991.
• 2. VDI 3478; Biological waste air purification -Biowäscher-; July 1985.
• 3rd S.P.P. OTTENGRAF, A.H.C. VAN DEN OEVER and F.J.C.M. KEMPENAARS, Waste gas purification in a biological filter bed; Innovations in Biotechnology; Ed .: E.H. HOUWINK and R.R. VAN VER MEER; 157-167; 1984th
• 4. S.P.P. OTTENGRAF, A.H.C. VAN DEN OEVER; Kinetics of organic compound removal from waste gas with a biological filter; Biotechnology and Bioengeneering Vol.XXV; 3089-3102; 1,983th
• 5. A. WiNDSPERGER, R. BUCHNER and K. STEFAN; Cleaning of solvent-containing exhaust air with biofilters; Part One; Dust-clean air; 50; 465-470; 1990th
• 6. A. WINDSPERGER, R. BUCHNER and K. STEFAN; Cleaning of solvent-containing exhaust air with biofilters; Part II; Dust-clean air; 51; 15-19; 1,991th
• 7. M.T. CESARIO, H.H. Beeftink and J. Tramper; Biological treatment of waste gases containing poorly-water-soluble pollutants; Studies in Enviromental Science 51; 1991; "Biotechniques for air pollution abatement and odor control policies"; Pp. 135-140.
• 8. M. JUNGHANS, B. DITIRICH; Biofilter as an alternative to the activated carbon filter; WLB- Water, soil, air; 6; 42-43; 1,992th
• 9. G. BEYREITZ, R.HÜBNER, M. SAAKE; Biotechnological treatment solvent-based iger exhaust air; WLB water, soil, air; 9, 53-57; 1989th
• 10. K.H. ENGESSER; Microbiological aspects of biolog. Exhaust air purification; Studies in Enviromental Science 51; 1991; "Biotechniques for air pollution abatement and odour control policies ", pp. 33-40.
• 11. D. BARDTKE; Microbiological conditions for biological waste air cleaning; from: Biological exhaust air purification, Ed .: Klaus Fischer, contact & study Bd. 212, Expert-Verlag; 1-11; 1990th
• 12. R. HÜBNER; Use of a biofilter for the treatment of solvent-containing exhaust air Main substance styrene; WLB water, soil, air; 1-2, 54-56; 1,993th
• 13. B.-M. KANIS, J. SZYMROSCZYK; Biofiltereinsatz in waste gas purification in the art fuel production; WLB water, soil, air; 10; 60; 1,991th
• 14. S.P.P. OTTENGRAF, J.J.P. MEESTERS, A.H.C. VAN DEN OEVER and H.R. Rozema; Bioprocess engineering; 1; 61-69; 1986th  
• 15. H.H. Cox, J.H.M. HOUTMAN, H.J. DODDEMA and W. HARDER; Enrichment of funghi and degradation of styrene in biofilters; Biotechnology Letters; Vol. 15, no. 7; 737-742; 1,993th
• 16. FRANS J. WEBER, SYBE HARTMANS; Biological waste gas treatment with integrated adsorption for the treatment of fluctuating concentrations; Studies in Enviromental Science 51; 1991; "Biotechniques for air pollution abatement and odor control policies ", pp. 125-130.
• 17. RUDOLF KAISER, Gas Phase Chromatography Part IV: Quantitative Evaluation Part 2; 2nd, erw. edition; Hochschultaschenbücher Verlag; 1969th
• 18. B. HEUSER, J. LINDEMANN; Degradation of volatile pollutants in bio-scrubbers and organic filter; WLB water, soil, air; 3; 66-69; 1994th

Claims (2)

1. Plant for the biological purification of solvent-containing exhaust air, characterized in that

• a) a switchable as needed buffer system is integrated, which allows both the compensation of occurring concentration peaks as well as a permanent reduction of pollutant load during operating hours.
• b) the exhaust air can be introduced both via the pre-separator and directly into the reactor (infinitely variable). From a pollutant flow x as much percent of the exhaust air are given directly into the bioreactor, as this still coped with the desired degradation performance. This value is previously tested in a test series with direct gas inlet (100% of the exhaust air directly into the bioreactor).
• c) according to higher pollutant concentrations, a partial flow during working time is passed through the buffer system. The adsorbed solvent fractions are then desorbed again continuously in the operating downtime with tempered compressed air. For this purpose, a temperature / time program coordinated with the respective exhaust air problem can be created.

2. Plant for the biological purification of solvent-containing exhaust air, characterized in that

• d.) the buffer system mentioned under a.) consists of a dust-free organic adsorbent polymer-based granules, which works excellently even under very humid conditions (relative humidity <90%).
• e.) can be drastically reduced with the integrated buffer system mentioned under a.) the reactor volume.