Study of catalyzed ozonation for advanced treatment of pulp and paper mill effluents PDF

Preview

Summary

ARTICLE IN PRESS WAT E R R E S E A R C H 40 (2006) 303 – 310 Available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/watres Study of catalyzed ozonation for advanced treatment of pulp and paper mill effluents Virginie Fontaniera, Vincent Farinesa, Joe¨l Albeta,, Sylvie Baigb, J...

Description

ARTICLE IN PRESS WAT E R R E S E A R C H

40 (2006) 303 – 310

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/watres

Study of catalyzed ozonation for advanced treatment of pulp and paper mill effluents Virginie Fontaniera, Vincent Farinesa, Joe¨l Albeta,, Sylvie Baigb, Jacques Moliniera a

ENSIACET, Laboratoire de Chimie Agro-Industrielle, UMR INRA 31A1010, Equipe Ge´nie Chimique, 118 Route de Narbonne, 31077 Toulouse Cedex 04, France b DEGREMONT SA, 183 Avenue du 18 Juin 1940, 92508 Rueil-Malmaison Cedex, France

art i cle info

A B S T R A C T

Article history:

Ozonation and catalytic ozonation (TOCCATAs process) were used as tertiary treatments of

Received 17 September 2003

wastewaters from three different pulp and paper mills. Laboratory batch experiments were

Received in revised form

conducted to assess the efficiency of each oxidation system for removal of organic matter.

12 October 2004

The investigations measured ozone consumption rate, variations in chemical oxygen

Accepted 8 November 2005

demand (COD), total organic carbon (TOC), suspended solids (SS) and molecular weight distribution with contact time. For conventional ozonation, ozone consumption rate was

Keywords:

dependent on the nature of the effluent. Organic matter elimination occurred both by

Catalytic ozonation

oxidation and precipitation. Precipitation played a major role on TOC removal varying with

Advanced oxidation

the effluent, and was responsible for production of high final SS concentrations. However,

Pulp mill effluents

the effluent type did not affect the ozone consumption rate for TOCCATAs-catalyzed

Organic matter

reactions. Using TOCCATAs, it was shown that organic matter was removed through

Mineralization

steady conversion of organic carbon to carbon dioxide. Finally the two oxidation systems

Precipitation

were compared with respect to their impact on molecular weight distribution. A total removal of the two initial fractions of compounds (high and low molecular weights) was observed with two effluents. With the third effluent, only the initial fraction of low molecular weight compounds was removed by the two oxidizing systems. The results showed that ozonation and TOCCATAs-catalyzed ozonation could achieve removals of COD of 36–76%. Depending on the effluent type, the amount of ozone consumed per gram of COD removed was lower for conventional or for catalytic ozonation. & 2005 Elsevier Ltd. All rights reserved.

1.

Introduction

The pulp and paper industry is one of the most important water consumers: 15–60 m3 of fresh water are used to produce 1 t of paper, depending on the type of production process involved (Thompson et al., 2001). A high degree of water reuse within modern mills means that the water leaves with a heavy burden of dissolved and particulate organics, usually measured as chemical oxygen demand (COD), BOD and suspended solids (SS). The levels of these emissions are Corresponding author. Tel.: +33 5 62 88 56 97; fax: +33 5 62 88 56 00.

E-mail address: [email protected] (J. Albet). 0043-1354/$ - see front matter & 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2005.11.007

strictly regulated. The integrated pollution prevention and control (IPPC) advises the pulp and paper mills to use the Best Available Techniques in order to achieve COD emission levels in the range of 8–23 kg of COD per ton of pulp for Kraft mills and in the range of 2–4 kg/t of paper for integrated recovered paper mills (IPPC, 2001). The two basic processes used to produce pulp, namely mechanical and chemical, generate very different wastewaters. The main organic pollutants responsible for the total organic carbon (TOC) and COD present in those effluents can

ARTICLE IN PRESS 304

WA T E R R E S E A R C H

be divided into five categories (Lenes and Hoel, 1999): carbohydrates (glucose, galactose, arabinose, xylose, mannose), extractives (resin and fatty acids, sterols, triglycerides), lignans (low molecular weight lignin polymers), lignin and its phenolic derivatives, and low molecular weight compounds such as acetic, formic, or oxalic acids. These organic pollutants correspond to a COD of 2 or 3 g/L in raw wastewaters from both chemical and mechanical mills (Ristolainen and Alen, 1998; Wang et al., 1999). In addition, the wastewaters from the paper machine white-water loop contain a SS concentration that can vary from 400 to 3000 mg/L (Ryoso and Manner, 1999). Such effluents have to be treated to meet the standards for discharge. The primary treatment usually implemented consists of clarification based on sedimentation or flotation. It results in 80–90% removal of SS (Thompson et al., 2001). This step has very little effect on TOC, COD or BOD concentrations, and is generally followed by a biological treatment, usually aerobic. This secondary treatment achieves large COD and BOD abatements. A review of the results for pulp mills in Finland shows that such techniques remove 73–99% of BOD and 50–92% of COD (Saunama¨ki, 1997). However, secondary treatment effluents still contain biorefractory organic matter whose further reduction requires a tertiary treatment. The major types of full scale tertiary treatment used at full scale are sand filtration, coagulation and ultrafiltration (Thompson et al., 2001). Ultrafiltration ensures partial removal of COD and color but is associated with serious disadvantages such as generation of concentrated streams needing disposal and need for pre-treatments to avoid membrane fouling and ageing. Other potential tertiary treatments are reverse osmosis, adsorption, precipitation and oxidation using ozone or other strong oxidants. Ozonation represents an interesting treatment alternative. Once dissolved in water, ozone can react with many organic compounds via either of two pathways: direct reaction of molecular ozone or indirect reaction through formation of secondary oxidants, especially free radical species (Staehelin and Hoigne´, 1982). Numerous applications of ozone for treatment of paper mill wastewaters highlight that ozone oxidation is able to reduce color and toxicity, targeting organic compounds like organochlorines (AOX), extractives, chlorophenolic compounds, lignins, resin and fatty acids (Alvares et al., 2001; Chen and Horan, 1998; Freire et al., 2001; Korhonen and Tuhkanen, 2000; Laari et al., 2000; Marco et al., 1997; Mohammed and Smith, 1992; Roy-Arcand and Archibald, 1996). However, the action of ozone is often limited to a partial oxidation of organic matter resulting in poor TOC and COD removals. Particularly simple compounds like formic, acetic or oxalic acids are very resistant to ozone mineralization. Catalyzed ozonation may be an effective way to improve the rate extent of oxidation (Fontanier et al., 2003). The benefits provided by the use of an ozone-based catalytic system have been demonstrated for the removal of organic pollutants such as atrazine, chlorobenzenes, halocarbons (Ma and Graham, 1999; Volk et al., 1997; Cortes et al., 1998; Cooper and Burch, 1999). Use of heterogeneous catalytic ozonation for the treatment of different industrial wastewaters have also been reported by Baig and Petitpain (2003), Kaptjin et al.

40 (2006) 303– 310

(1995) and Logemann and Annee (1997). Development of an effective catalytic process is complex. Surface catalysis (heterogeneous catalysis) involves five consecutive steps which influence on the overall rate of the chemical conversion: (1) diffusion of the reagents toward the catalyst, (2) interaction of the reagents with the catalyst (adsorption), (3) reaction between the adsorbed reagents to give the products, (4) desorption of the products from the surface of the catalyst towards the medium, and (5) diffusion of the products away from the catalyst. Whereas the first and last steps correspond to physical processes of matter transfer, the intermediate steps consist in chemical phenomena, the reaction taking place between the chemically adsorbed species by arrangement of surface complexes. Consequently, it is obvious that adsorption only constitutes a first step in the chemical reaction. The catalyst cannot therefore be selected only for its adsorbent capacity, all the more so since in order to interest a broad range of compounds, the adsorption must be limited to a physical adsorption through Van der Waals interactions. This physical adsorption allows the real chemical reaction initiated by the formation of covalent or ionic bonds between the surface of the catalyst and the adsorbed pollutant molecule. If too strong, this interaction can lead to catalyst poisoning. Similarly, irreversible adsorption of the reaction products will prevent subsequent reactions. In contrast with physical adsorption, catalytic reaction is more specific and consequently determines the overall catalyst efficiency. The TOCCATAs process was developed to enhance the oxidation by ozone of organic pollutants to carbon dioxide and water. TOCCATAs promotes a catalytic process by forming active species of very high oxidizing potential in the presence of ozone. The TOCCATAs heterogeneous catalyst uses metal oxides and is usable in bulk or supported and shaped according to the reactor design (Degre´mont, 1999). The aim of this study was to assess the suitability of the TOCCATAs ozonation system for the tertiary treatment of three pulp or paper mill effluents compared to conventional ozonation. The experiments were conducted at laboratory scale in batch mode and investigated organic matter removal through time. Parameters such as COD, TOC, SS and molecular weight distribution were followed. The catalyst used was in powder form added to a stirred reactor. This was to avoid kinetic limitation due to diffusion phenomenon.

2.

Material and methods

2.1.

Ozonation reactions

Ozonation and catalytic ozonation experiments were carried out at 20 1C using 1.5 L of wastewater in a 2 L stirred reactor. For catalytic experiments, 100 g of powdered TOCCATAs catalyst were added just before ozone introduction. The catalyst granules had an average diameter of 220 mm and a 250 m2/g surface area. Ozone was produced from pure oxygen at 40 (72%) mg/L standard temperature and pressure (STP) (Ozonia generator). The O2+O3 gas (60 (76%) L/h STP) was continuously injected at the bottom of the reactor through a

ARTICLE IN PRESS WAT E R R E S E A R C H

305

40 (20 06) 30 3 – 310

ceramic diffuser. The relative gas (O3+O2) pressure in the reactor was maintained at 1200 mbar. The mixing rate was fixed at 500 rpm. The temperature of the effluent was regulated to 20 1C. Ozone concentrations in the gas inlet and outlet to the reactor were measured using Messtechnik BMT 961 analysers (UV detection).

4000–400 cm1, using the KBr pellet technique. The carbon balance was calculated after the full time of reaction. This was calculated from change in total carbon concentration of the solution and the amount of inorganic carbon released in the gas phase as carbon dioxide trapped in a solution of sodium hydroxide.

2.2.

2.3.

Analytical assays

COD was measured according to NFT 90 101 standardized method NFT 90–101, 2001 on 0.45 mm filtered samples. In this method an excess of potassium dichromate is used as the oxidizing agent and the sample is heated with a Hach digestor for 2 h. The relative error on COD measurement was estimated to be 5%. TOC and inorganic carbon (IC) were measured with a SHIMADZU, TOC 5000A apparatus on 0.45 mm filtered samples. The relative error on TOC and IC measurements was estimated to be 5%. SS concentrations were determined according to NF EN 872 standardized method (by filtration on glass fiber filters) (NF EN 872, 1996). The percentage of carbon in the oven-dry SS was measured according to elemental microanalysis on a EA 1110 Carlo Erba apparatus. Molecular weight distribution was determined using gel permeation chromatography with detection by a UV diode array detector. The gel column used was TosoHaas TSK Gel G3000 PWXL column enabling separation of the 100–30 000 Da range (column dimensions ¼ 30 cm
7.8 mm, phase 6 mm). Infrared analyses were conducted using a Perkin–Elmer Spectrum BX spectrophotometer, in the range

Effluents

Samples of three different effluents were collected at the outlets of secondary biological treatment units at the mill sites. Mill types are given in Table 1. Effluents A and C were from processes producing, or including intermediate production of bleached sulfate (Kraft) pulp and therefore have quite similar characteristics. However, effluent B, from a corrugated cardboard mill producing mechanical unbleached pulp, had much more concentrated organic matter. All the effluents were stored at 4 1C after collection on site. Their maximum age (time between collection and ozonation) was 30 days.

3.

Results and discussion

3.1.

Ozonation

The results obtained for ozonation of the three effluents are given in Table 2 (‘‘f’’ stands for final value and ‘‘0’’ for initial value). They show that ozone consumption rates in our

Table 1 – Characteristics of the effluents Mill products

Effluent A Printing/writing papers (coated paper)

Effluent B Corrugated board (from recycled fibre)

Effluent C Bleached sulfate pulp (from softwood) for printing papers

7.1 50 20 70 270 38 — —

8.4 145 131 443 45 87 — —

7.4 23 39 119 315 27 6 84

PH IC (mg/L) TOC (mg/L) COD (mg/L) Cl (mg/L) SS (mg/L) Mg2+ (mg/L) Ca2+ (mg/L)

Table 2 – Ozonation and catalytic ozonation of effluents A, B and C Effluent

Contact time (h)

pH0

pHf

2 3 2

7.1 8.4 7.4

8.1 7.8 7.2

Catalytic ozonation A 2 B 3 C 2

7.1 8.4 7.4

5.6 7.9 6.0

Ozonation A B C

Consumed ozone (mg/L)

TOC0 (mg/L)

DTOC (%)

COD0 (mg/L)

DCOD (%)

30.3 28.4 30.9

400 1583 556

20 131 39

39 19 51

70 443 119

52 36 76

30.9 28.5 30.5

1372 1643 1355

20 131 39

74 35 73

70 443 119

67 53 72

Rate of ozone application (mg/L.min)

ARTICLE IN PRESS 306

WA T E R R E S E A R C H

40 (2006) 303– 310

1.6 1.4

TOC / TOC0

1.2 1 0.8 0.6 0.4 0.2 Effluent A

Effluent B

Effluent C

0 0

60

120

180

Contact time (min) Fig. 1 – TOC/TOC0 ratio change during ozonation of effluents A, B and C.

system are quite low, especially for effluents A and C, for which only 10–15% of the ozone introduced was consumed. In addition, as usually observed with ozonation of wastewaters (Sevimli et al., 2003), the instantaneous rate of ozone consumption fell to nearly zero after a contact time of 60 min for effluents A and C, and 90 min for effluent B. However the TOC and COD abatement yields were significant for effluents A and C (Table 2). TOC increased at the beginning of the ozonation reaction for effluents A and B (Fig. 1). This was confirmed by the results of COD determination. At the same time, the SS concentration in effluent B decreased significantly over a contact time of 60 min to increase again at the end of the ozonation reaction (Fig. 2). The evolution of TOC in effluent B results from the attack of ozone on both dissolved and particulate organics. The apparent increase of TOC (and COD) with ozonation seen in Figs. 1, 3 and 5 is very likely due to the release of soluble compounds by ozone from SS. In effluents A and C, the SS concentration after ozonation was higher than in raw samples: respectively 100 and 81 mg/L. This indicates that precipitation is also an effect of ozonation, as reported, previously, by Mo¨bius (1999). The analysis of final SS by infrared spectroscopy gave similar spectra from the three effluents. We observed the characteristic bands of carboxylic ions (COO symmetric stretch at 1325 cm1 and COO asymmetric stretch at 1642 cm1). This agrees with the decreasing ozone consumption yield in spite of a continuous decrease of the TOC/TOC0 ratio, the increase of SS and no pH evolution during ozonation reaction. SS may have been formed through precipitation of carboxylated ozonation products as calcium and magnesium complexes. Calcium and magnesium were initially present in the effluents (Table 1). The final SS from ozonation of effluent A contained 13% carbon (measured by elemental microanalysis), which confirms that precipitation was significant in the elimination of dissolved organic compounds. Such a mechanism was favored by the low reactivity of carboxylic acids towards ozone under nearly neutral pH. Indeed, kinetic constant rates of the reactions of ozone with acetic, oxalic or propionic acids are known to be

very low (Hoigne and Bader, 1983). Those results are confirmed by the carbon mass balance calculated after full reaction time from change in total carbon concentration of the solution and the amount of inorganic carbon released in the gas phase as carbon dioxide trapped in a solution of sodium hydroxide. It shows that the collected amount of CO2 corresponds to a mineralization of 0%, 28% and 59%, respectively, for effluents A, B and C.

3.2.

Catalytic ozonation

The results obtained with catalytic ozonation of the three effluents are given in Table 2. The amount of ozone consumed is similar for the three effluents, showing no effect of the wastewater quality. About 70% of TOC and COD abatement yields were obtained for effluents A and C, 35 and 53%, respectively, for TOC and COD removal yields in the case of effluent B. We also noted that pH decreased significantly during catalytic ozonation reaction as is usually the case with ozonation (Perkowski et al., 2003). Change in pH was similar for effluents A and C issued from a bleached sulfate pulp production and less significant for effluent B. Once more TOC slightly increased at the beginning of the catalytic ozonation of effluent B (Fig. 3), which may be due, as seen before, to the formation of soluble compounds from SS. As a consequence TOC removal was delayed in this case in comparison with effluent A and C, for which the final TOCs were 5 and 10 mg/L, respectively. No increase of COD was observed at the beginning of the catalytic ozonation reaction of effluent B. This suggests that the ozonation products were compounds not detected by COD measurement. This includes two types of organic compounds: compounds which cannot be fully oxidized; and volatile compounds which escape (NFT 90–101, 2001). It is worth pointing out that final TOC and COD removal rates are very similar for effluents A and C. This demonstrates that removal of organic pollutants occurred through steady mineralization, which is in agreement with the pH decrease to values less liable to the occurrence of products precipitation.

ARTICLE IN PRESS WAT E R R E S E A R C H

307

40 (20 06) 30 3 – 310

100 90 80

SS (mg/L)

70 60 50 40 30 20 10 0 120

60

0

180

Contact time (min) Fig. 2 – SS change during ozonation of effluent B.

1.4 1.2

TOC / TOC0

1.0 0.8 0.6 0.4 0.2 Effluent A

Effluent B

Effluent C

0.0 0

60
...