TOWARDS ZERO-EFFLUENT PULP
AND PAPER PRODUCTION:
The Pivotal Role of Totally Chlorine Free Bleaching
November 1996
Contents
EXECUTIVE
SUMMARY
INTRODUCTION
PULP PRODUCTION PROCESSES
1) Groundwood pulping
2) Refiner
mechanical pulping
3) Semi-chemical
pulping
4) Chemical pulps
a) Alkaline
chemical pulping
b) Sulphite
chemical pulping
c) Organosolv
pulping
WATER USE
BLEACHING AGENTS and PROCESSES FOR
CHEMICAL PULPS
FORMATION OF ORGANOCHLORINES BY CHLORINE
BLEACHING
POLLUTION CONTROL MEASURES
a) Process Internal
Measures
i) Extended
delignification
ii) Oxygen
delignification
iii) Modification
of the Bleach Process
b) Process external
measures
EFFECTS OF PROCESS MODIFICATIONS
ON ORGANOCHLORINE PRODUCTION
a) Non-dioxin
compounds and AOX
b) Chlorinated
dioxins and dibenzofurans
CHLORINE AND CHLORINE
DIOXIDE LINKAGE
BIOLOGICAL IMPACTS OF MILL DISCHARGES-RECENT
RESEARCH
ENDOCRINE DISRUPTION & PULP
AND PAPER MILLS
TOTALLY EFFLUENT FREE PROCESSES-CLOSING
THE BLEACH CIRCUITS
REFERENCES
|
Paul A. Johnston, Ruth L. Stringer, David Santillo, Angela D. Stephenson,
Irina. Ph. Labounskaia, Hannah M.A. McCartney,
Greenpeace Research Laboratories,
Earth Resources Centre,
University of Exeter,
Exeter, UK.
Technical Report 7/96, November 28, 1996
ISBN 90-73361-32-X
Greenpeace International, Amsterdam
|
EXECUTIVE
SUMMARY
The world pulp and paper industry continues to expand production and
increasingly, plants are being built in newly industrialised countries.
The dominant process is the kraft or sulphate process. Historically, substantial
pollution problems have been associated with pulp manufacturing operations.
Following the recognition of large scale environmental contamination by
organochlorines due to their formation in bleach plants, the industry implemented
a number of process internal changes and continued to develop process external
treatment processes. The bleach plant of kraft mills generates a substantial
proportion of the total process effluents.
The introduction of extended delignification and oxygen delignification
can substantially reduce the quantities of lignin entering the bleach plant.
Residues from oxygen delignification can be cycled to the recovery process.
Overall, this reduces the demand for bleaching agents and hence reduces
the generation of organochlorines. Together with substitution of elemental
chlorine (Cl2) with chlorine dioxide, these systems have substantially
reduced levels of AOX being discharged from bleach kraft mills. Organochlorines,
however, have not been eliminated from discharges, merely reduced. Swedish
research has shown detectable levels of toxicologically chlorinated dioxins
and dibenzofurans in the effluents from mills that use chlorine dioxide;
so-called elemental chlorine free (ECF) mills.
This stems from the fact that ECF processes are not free of elemental
chlorine. Commercial chlorine dioxide generators in many cases co-generate
molecular chlorine. Moreover, chemical reactions and pH dependent chemical
equilibria in pulp bleaching reactions involving chlorine dioxide liberate
molecular chlorine. This molecular chlorine then reacts with chemicals
released from the wood.
Organochlorine production in ECF processes is within a broad range of
0.1-10.0 kg per tonne of air dried pulp produced. The global average is
unknown. The improvements engendered by process internal measures are still
under investigation. Available results indicate that the greater the degree
of chlorine dioxide substitution and early delignification, the better
the effluent quality achieved. Many mills, however, have not adopted these
stages and therefore discharges of organochlorines are relatively high.
The elimination of organochlorine discharges can be achieved by the
use of TCF (Totally Chlorine Free) bleaching where agents such as hydrogen
peroxide and ozone are used. A necessary prerequisite of TCF bleaching
is a pulp with low residual lignin (kappa number) produced through extended
cooking and oxygen delignification. Conversion of existing mills favours
this technology over chlorine dioxide based processes in economic terms.
In addition the cost of new TCF plant is cheaper than ECF.
In terms of effluent toxicity, TCF systems produce a less toxic effluent
than elemental chlorine bleached processes when realistic conditions using
actual mill effluents are employed in the experiments as opposed to samples
synthesised in the laboratory. Today it is estimated that some 15-20 true
TCF mills exist. Recently, however, effects have been identified in fish
populations exposed to effluents from pulp mills producing both bleached
and unbleached pulp. The chemicals responsible are not removed by advanced
secondary treatment and are suspected to be plant sterols or their derivatives
which have a strong endocrine disruptive action.
The identification of environmental effects due to both bleached and
unbleached pulping activities has led to the concept of the totally effluent
free mill. A key impediment to fully closing mill circuits is the difficulty
of closure of the bleach lines. Although problems exist with closing both
ECF and TCF lines, those involved in closing ECF lines appear to be the
most difficult and costly to resolve.
The presence of high levels of chlorides in an acid bleach medium has
been associated with severe corrosion problems, and hence the possibility
of explosion in recovery boiler systems. Moreover, the presence of organochlorines
in both filtrates of ECF bleach liquors and in sludges from treatment plants
means that they cannot be incinerated without the emission of products
of incomplete combustion including the dioxins and furans (PCDDs and PCDFs).
Both closed circuit mills currently operating in Canada are CTMP (chemi-thermomechanical
pulp) mills which do not use chlorine chemicals in the bleach cycle.
Accordingly, in order to reach likely regulatory standards in the future,
closure of mill circuits will be necessary. This is likely to be more easily
achievable with TCF technology since there will be a lower requirement
for chloride removal from the process liquors. In addition there is much
greater potential to deal with these streams without the use of incineration
technology. In order to avoid transferring environmental impacts from one
medium to another, incineration of ECF derived sludges and other solids
will need to be eliminated. Hence, movement of the industry generally into
TCF processes rather than ECF systems promises more rapid achievement of
the zero discharge goal.
In order to promote sound environmental practice and move towards zero
effluent mills, environmental regulators need to promote the need for process
changes. Extended cooking and oxygen delignification of pulps is required
to reduce the quantities of lignin entering the bleach plant. The bleach
plant should be based upon totally chlorine free technologies to reduce
effluent toxicity as much as possible and to facilitate water and process
chemical recycling without need for incineration of sludge contaminated
with organochlorine compounds. Under these conditions, the mill circuits
can effectively be closed and operated totally effluent free (TEF).
INTRODUCTION
World wide consumption of paper products now exceeds 268 million tons
per year (PPI 1995). This is up from 238 million tons in 1990 (PPI 1991)
which, in turn, increased from 170 million tons in the late 1970s (Smook
1989). Apparent per capita consumption is greatest in the United
States at 332.6 kg of pulp products, while, by contrast the 1990 figure
for Malta was 71.2 kg and that for India was estimated at 2.7kg (PPI 1991).
The pulp and paper industry is expected generally to grow significantly
in the newly industrialised countries of Africa and Latin America. The
highest rates of growth, however, are expected in SE Asia, a premise borne
out by the latest world industry figures (PPI 1995). Elsewhere continued
growth is also forecast, although at a lower rate than in the past. World
consumption is projected to rise to over 310 million tonnes in the year
2000 (Myreen 1994). The changes within the industry are expected to have
a significant impact on the chemical industry as well. Some 16.5 million
tons of chemicals are used annually in the US and Canadian industries alone.
Approximately US$50 worth of chemicals are used in the production of every
ton of bleached softwood pulp, US$20 per ton of newsprint and US$80 per
ton of reprographic papers (Burt 1987).
The different grades are likely to show differential growth rates but
no specific product group is thought likely to disappear from the market.
The average composition of products, however, will change. In general,
the greater the requirement for fibre purity, the lower the yield of fibre
and the greater the cost. Although recycled fibre is predicted to become
a major fibre resource with some estimates that it may ultimately comprise
50% of the total fibre furnish, this is likely to be highly influenced
by considerations of energy consumption. The production of mechanical and
recycled pulps are energy intensive processes. In addition, (Bystrom &
Lonnstedt 1995), the utilisation rates for waste paper in Europe are already
regarded as high, leaving limited scope for the use of waste paper in production.
These current general trends are shown in Figure 1. The pulp market overall
will be dominated by bleached kraft (sulphate) hardwood and softwood pulps.
Historically, pulp and paper production has long been recognised as
a significant point source of pollution. Early control initiatives in the
UK, for example, concentrated upon reducing the loading of suspended solids
and biological oxygen demand to the recipient (Waldmeyer 1957). BOD control
is still problematic in areas where discharges are made to restricted watercourses
or enclosed coastal areas. A recent study of the Sone River in India (Srivastava
et al. 1988) has shown that the river effectively required 216 km
to revert to its original condition after receiving the effluent from a
mill of installed capacity of 85kt per year. Estimated BOD-7 discharges
direct to the Gulf of Bothnia, Gulf of Finland, Baltic Sea and Kattegat
totalled some 180kt per year in the late 1980s (CGB 1987).
Technologies used or under development at this time included percolating
filters, activated sludge processes, anaerobic treatment and high pressure
oxidation. These were all broadly end-of-pipe methodologies whose counterparts
may be found at many modern mills.
Aerated lagoon systems have to a large extent become the industry standard
for wastewater treatment. Process controls, then as now, included fibre
recovery systems to reduce loadings of particulates to receiving waters.
As in the UK, (Edde 1994) early pollution abatement initiatives in the
USA were based largely upon a system of recipient water quality objectives
coupled with definition of the notional "assimilative capacity" of the
recipient. Subsequently, multi-tiered frameworks of environmental regulation
have been evolved in a number of countries (see: Gifford & McFarlane
1991; Sprague 1991; Edde 1994).
Despite this preoccupation with biological oxygen demand and total suspended
solids, however, some early work (eg Ebeling 1931) had remarked on the
toxicity of pulp effluents and noted the resinous taste that they imparted
to fish flesh. Beginning in the late 1960s there was an increasing awareness
of the toxic properties of pulp and paper effluents which had previously
been obscured by the excessive BOD demand placed on the receiving waters.
The emphasis gradually shifted from unbleached kraft mill wastes to bleached
mill wastes and a number of attempts were made to isolate the toxic components.
This early work has been comprehensively reviewed by Walden (1976).
As noted by Minor (1982), the pulp and paper industry has had three
very difficult problems to control as follows: the development of suitable
methods for sulphite waste liquor recovery, elimination of malodorous and
toxic sulphur compound emissions from the kraft pulping process and the
elimination of toxic compounds, BOD, COD and colour from bleach plant effluents.
The solution of the bleach plant problems has proven to be the most difficult.
Arguably, this aspect of pulp production is the single most important determinant
of environmental impact of such operations and the single greatest impediment
to reducing and eliminating the environmental impact of pulp production
facilities.
Accordingly, this document traces the evolution of bleaching techniques
and the problems associated with pulp effluents in receiving waters. It
discusses the critical need to modify bleaching processes in order to control
more fully process effluents. In particular it examines the use of chlorine
dioxide as a substitute for chlorine as a pulp bleach. It reaches the conclusion
that if the zero effluent mill is to become a realistic prospect and the
industry standard, then all forms of chlorine based bleaching agents need
to be substituted for.
PULP
PRODUCTION PROCESSES
Although the history of papermaking dates back to the first century
BC (Minor 1982) the modern process dates back to the opening of the Frogmore
Mill in England in 1798 (Myreen 1994) producing paper from rags. The extreme
short-supply of rags was alleviated in the late 1870s when it became technically
feasible to produce pulp from wood. This was a turning point in the development
of the industry which subsequently evolved from an artisanal activity to
a fully mechanised industry (Kerski 1995). Subsequently, paper mills came
to be located by rivers with afforested catchments, facilitating transport
of the mill wood furnish and the use of the rivers for power and process
water.
The production of pulp is the single most important technique for the
chemical conversion of wood and accounts for more than one third of the
total processed mass annually: 460 million cubic metres out of a total
2600 million cubic metres harvested which includes the 1500 million cubic
metres used for fuel wood (Fengel & Wegener 1989). Another perspective
is provided by Kerski (1995) who estimates that the annual timber consumption
equates to the production of a land area of over 20,000 square kilometres,
or around half the area of Switzerland. Pulp and paper production accounts
for one per cent of the world's total economic output. The 1992, 1993 and
1994 production figures of the 10 largest pulp producing countries, accounting
for around 85% of the total world pulp production are shown in Table 1
(PPI 1992; 1993; 1994). The rapid and steady increase in demand for pulp
products is coupled with an increasing shortage of wood supplies. Hence,
short-rotation intensive culture plantations are now being actively researched
as a source of mill furnish (Sierra-Alvarez & Tjeerdsma 1995).
| Production (x1000t) |
| |
1992 |
1993 |
1994 |
| USA |
59282 |
57069 |
58724 |
| Canada |
22841 |
22897 |
24547 |
| China |
11985 |
15290 |
17054 |
| Sweden |
9589 |
9953 |
10867 |
| Japan |
11200 |
10593 |
10579 |
| Finland |
8525 |
9338 |
9962 |
| Brazil |
5368 |
5509 |
6106 |
| CIS |
6800 |
4365 |
3313 |
| France |
2609 |
2540 |
2787 |
| Norway |
2009 |
2169 |
2344 |
Table 1: Pulp production figures for top producers in 1994
as compared to previous years. Data from PPI (1993; 1994; 1995).
Wood is chemically very highly complex. The basic structural element
of the cell walls is cellulose. Lignin and hemicellulose are also distributed
throughout the cell walls, although this distribution is poorly understood
(Minor 1982; Fengel and Wegener 1989). The relative proportions vary, but
in the cells of Scots pine, the relative proportions are 28.0 weight %
lignin, 28.7 wt% hemicellulose and 40.3 weight % cellulose. In addition
wood contains 3-10% of "extractives". This fraction includes fats and esters,
terpenes and resin acids, phenolic materials and tannins. Trace quantities
of inorganic materials are also present together with heavy metals. The
lignin component binds the cellulose fibres together and in order to produce
paper the fibres must be separated from each other. This may be achieved
mechanically or by chemical dissolution of the lignin.
Pulping is the means whereby the wood is reduced to a fibrous mass for
onward processing into paper and board products. After harvesting, the
wood is debarked mechanically and converted to chips of specified size.
The chips are screened and then pulped. An extensive overview of pulping
procedures and treatments is given by Fengel & Wegener (1989) and Minor
(1982) and the reader is referred to these documents for more detailed
information. There are a variety of methods for producing pulps mechanically
and several chemical pulping methods are in use. Chlor-alkali products
are widely used in the refining and bleaching processes involved in the
manufacture of chemical pulps. The pulping method of choice will depend
upon a number of factors including the grade of paper being manufactured.
Different paper qualities utilise different types of fibres. In a copying
paper, for example, hardwood fibre is used as the bulk with softwood fibre
added to impart mechanical strength. Newsprint is manufactured predominantly
of mechanical or recycled fibres with a small component of soft wood fibre
to impart strengthening properties.
| Region |
C |
M |
O |
C |
M |
O |
| |
(1990) |
(1994) |
| E+W Europe |
28063 |
14285 |
1711 |
26120 |
13398 |
332 |
| N. America |
62346 |
16310 |
1393 |
67080 |
16189 |
2 |
| L. America |
5700 |
719 |
862 |
8433 |
559 |
626 |
| Asia |
19130 |
2904 |
2444 |
21606 |
3065 |
8089 |
| Australasia |
1270 |
1123 |
0 |
1270 |
1094 |
942 |
| Africa |
1384 |
413 |
592 |
2025 |
504 |
145 |
| TOTAL |
117893 |
35754 |
7002 |
126534 |
34809 |
10136 |
Table 2: World regional production of pulp by types for 1990 & 1994.
Apparent decline in total figures for Europe is due to closure of plants
in the former Eastern Bloc. Data from PPI (1991; 1995). C= Chemical pulp,
M= Mechanical pulp, O= Other pulp types.
| Production Method |
| |
Million Tonnes |
Percentage of Total |
| Sulphate Pulp |
72.0 |
58.2% |
| Sulphite Pulp |
11.4 |
9.2% |
| Mechanical Pulp |
26.0 |
21.1% |
| Dissolving Pulps |
4.5 |
3.7% |
| Other Pulps |
9.6 |
7.8% |
Table 3. World production of pulp according to pulping method
in 1979. Data from Fengel & Wegener (1979)
| Production (x1000t) |
| |
Sulphate |
Sulphite |
Mechanical |
| |
B |
U |
B |
U |
M |
S |
| USA |
28249 |
19870 |
1334 |
0 |
5338 |
3719 |
| Canada |
10958 |
1474 |
313 |
512 |
10851 |
440 |
| China* |
Total 893 |
Total 89 |
Total 473 |
| Sweden |
4990 |
2048 |
639 |
82 |
2858 |
250 |
| Japan |
6928 |
1638 |
9 |
0 |
1636 |
214 |
| Finland |
5157 |
687 |
0 |
0 |
4181 |
- |
| Brazil |
3606 |
1247 |
17 |
- |
333 |
37 |
| CIS |
1140 |
866 |
194 |
316 |
Total 915 |
| France |
1025 |
494 |
262 |
0 |
886 |
119 |
| Norway |
346 |
162 |
189 |
41 |
1516 |
90 |
Table 4: Pulp production figures for top producers in 1994
by category of pulp. Data from PPI (1994). B= bleached, U= unbleached,
S= semichemical, M= mechanical. Figures for China [*] derived from limited
dataset. Disparity between data in Table 1 due to omission of other pulp
types and differences in data sources.
All pulping processes are designed to separate the wood fibres present
in the wood. Mechanical processes achieve this by grinding or tearing the
fibres apart. Chemical processes dissolve the lignin present between the
cells allowing the fibres to separate with little or no mechanical action.
Techniques available vary between these two extremes. The industry balance
of pulping methods is shown in Tables 2 & 3 taken from PPI 1991;1995
& Fengel and Wegener (1989). Table 4 shows the balance of pulping methods
used by top producers. The major pulp production methods can be categorised
as follows:
1) Groundwood pulping
These processes essentially grind the raw wood to produce a pulp yield
of between 90 and 98% in modern systems. Modifications to the process include
the use of steam as a lignin softening agent and carrying out the process
under positive pressure. Uses of this pulp are generally restricted to
applications such as newsprint, toilet tissues and paperboard since it
tends to yellow on age due to the content of residual lignin.
2) Refiner mechanical pulping
This process differs from groundwood pulping principally in the use
of chips, wafers and sawmill wastes as feedstock. The dominant process
is thermo-mechanical pulp where the wood chips are preheated and steamed
before being fed to one of the three standard disc-refiner systems. The
resulting pulps are somewhat less bright for a given wood species than
groundwood pulps.
3) Semi-chemical pulping
Semi-chemical pulping processes are characterised by a chemical pretreatment
stage followed by a mechanical refining step. The process uses predominantly
hardwoods and can accommodate wood of relatively inferior quality and wood
obtained from mixed stands. The best known of these processes is the neutral
sulphite semichemical process and involves impregnation of the chips with
sodium sulphite liquor, followed by cooking at 160-190 C and a subsequent
disc refining stage. The pulps produced typically have 10-15% residual
lignin, much higher than the full chemical pulps. The pulps have high rigidity
and stiffness and are used in corrugated boards as well as printing papers,
greaseproof papers and bond papers. There is no clear demarcation between
the semi-chemical processes and the high yield chemical processes which
are generally modifications of the normal kraft and sulphite pulping methods.
4) Chemical pulps
a) Alkaline chemical pulping
The two major alkaline processes for producing chemical pulps are the
alkaline sulphate or "kraft" process and the soda process. In both these
processes wood chips are cooked with sodium hydroxide in order to dissolve
the lignin which binds the fibres together. Sodium sulphide is an additional
component of the pulping chemical mix in the kraft process. Both processes
are named according to the regeneration chemicals used to compensate for
sodium hydroxide: sodium sulphate and sodium carbonate. The kraft process
is not only the dominant chemical pulping process but the most important
overall in terms of the various production methods. The soda process is
important largely in the production of non-wood pulps. Various modifications
to the kraft and soda processes have been devised in order to attempt to
overcome low pulp yields and environmental problems. These generally involve
the addition of chemicals to the digest liquor. The most important of these
is anthraquinone (AQ). The benefits of AQ pulping include increased delignification
rates together with reduced alkali charges and improved pulp properties.
An integral and economically vital part of alkaline pulping mill operations
is the regeneration of the cooking liquors (Fengel & Wegener 1989;
Minor 1982). The recovery cycle is well defined for the kraft process and
is designed to recover pulping chemicals, reduce water pollution by combusting
organic matter in the spent liquor, generate process heat and recover by-products
of value. The main steps in the process are the evaporation of the black
liquor drained from the digester after wood chip digestion, combustion
of the concentrated liquor to produce a mineral "smelt", causticisation
of the smelt and regeneration of the lime used in the process. The energy
content of the black liquor is high. Gullichsen (1991) notes that half
of the wood is dissolved during the manufacture of chemical pulp, and this,
when combusted in the recovery boiler, provides heat for the plant systems.
The heart of the process is the recovery furnace. The black liquor is
evaporated to a solids content of between 60% and 75% using a 5-6 stage
system and this is followed by direct contact evaporation in which flue
gas from the recovery boiler is brought directly into contact with the
liquor. Tall oil soaps are recovered during the evaporation stages. Oxidation
of the liquor prior to evaporation can be carried out to reduce the emission
of malodorous compounds. When the black liquor is concentrated, sodium
sulphate and other chemicals are added to compensate for those lost in
the pulping process.
In the recovery boiler, the organic content is combusted to produce
heat. Carbon dioxide reacts with sodium hydroxide to produce sodium carbonate.
The added sodium sulphate is reduced to sodium sulphide and hence the solid
smelt produced by the boiler contains largely sodium carbonate and sodium
sulphide. This is dissolved in a tank to produce the green liquor which
is subsequently filtered and treated with calcium hydroxide (slaked lime)
to convert the sodium carbonate to sodium hydroxide. The resulting white
liquor is then returned to the digestion process. The lime is regenerated
by heating and mixing with water removed from the green liquor. This process
is, therefore, theoretically closed in relation to water use but not with
respect to atmospheric emissions, spills and condensate generation.
The pulps produced by the kraft process are characterised by good strength
properties. They are, therefore, the preferred grades in strong paper grades
such as the liner in corrugating boards or bag and wrapping papers. Hardwood
kraft pulps are used in many printing papers for bulking purposes, in mixture
with softwood pulps. The residual lignin present in the pulp is expressed
in terms of the "kappa number" which is determined by the oxidation of
lignin by potassium permanganate under acidic conditions. The lower the
kappa number of a pulp, the lower the level of residual lignin.
b) Sulphite chemical pulping
The basic processes for the manufacture of sulphite chemical pulps are
outlined by Fengel and Wegener (1989) and Minor (1982). These are the acidic
sulphite process, bisulphite process, multi-stage sulphite process, neutral
sulphite process and alkaline sulphite process. All are described in terms
of the composition of the cooking liquor which in turn defines the process
pH and the choice of basic chemicals used. The gas sulphur dioxide is used
to generate the sulphite chemicals used in digestion. Pulping conditions
vary widely according to the mill furnish and the different processes produce
pulps of differing chemical composition and papermaking application.
Sulphite pulps are a relatively small component of the pulps produced
globally in relation to kraft, semichemical and mechanical pulps as shown
by the data in Table 4 derived from PPI (1994).
c) Organosolv pulping
This pulping method has been receiving increased attention in recent
years. Solvent pulping offers a number of potential advantages over conventional
pulping techniques such as relatively low chemical and energy consumption
coupled with low capital costs and low environmental impact. Aqueous organic
solvents such as methanol and ethanol are used for delignification to produce
a bleachable pulp which can be bleached with non-chlorine chemicals. Pilot
scale tests have given high pulp yields with strength properties similar
to sulphite and kraft pulps (see: Sierra-Alvarez & Tjeerdsma 1995).
WATER USE
Water is an important consumable in the production of pulp and paper.
With time, however, there has been a progressive reduction in the use of
water at mills (Edde 1994) considered in relation to all types of mill.
Whereas in 1959 around 240 cubic metres of water were used in the production
of 1 tonne of air dried pulp, by 1988 average water usage had been reduced
to 72 cubic metres per air dried tonne. The most recent technology, installed
in South Africa, has a water demand of 16 m3 per air dried tonne. Table
5 shows the main sources of aqueous effluent produced in a bleached kraft
mill.
| Process Area |
Newer Mills |
Current designs |
Possible Future |
| Woodroom |
|
0.3 |
0.3 |
0.3 |
| Digestion |
|
1.0 |
1.0 |
0.2 |
| Washing & Screening |
1.8 |
0.5 |
0.2 |
| Bleach Plant |
- Acid |
21.1 |
10.5 |
5.3 |
| |
- Alkaline |
10.0 |
10.6 |
5.3 |
| Pulp Machine |
- Rejects |
1.3 |
0.7 |
0.3 |
| |
- General |
5.9 |
5.9 |
0.7 |
| Evaporators |
|
0.6 |
0.4 |
0.2 |
| Demineralisation |
0.1 |
0.1 |
0.1 |
| Recovery |
|
0.7 |
0.2 |
0.2 |
| Hog/Power Boiler |
3.3 |
1.0 |
0.5 |
| Recausticising |
|
1.4 |
1.1 |
0.5 |
| Chemical Preparation |
0.8 |
0.3 |
0.2 |
| Water Supply |
|
8.1 |
2.7 |
0.8 |
| Effluent Treatment |
0.0 |
1.2 |
0.2 |
| Clear water by-pass |
16.6 |
0.0 |
0.0 |
| |
|
|
|
|
| TOTAL |
|
73.0 |
36.5 |
14.8 |
TABLE 5 Effluent arisings from various mill processes with
various design concepts. All figures in cubic metres per tonne of air dried
pulp.
Various stages of the pulping process give rise to aqueous effluents
in kraft mills. Comparative water usages at given stages and states of
technology for bleached hardwood kraft pulp are given by Edde (1994) and
these figures are reproduced in Table 5. These demonstrate quite clearly
that whatever the design, the bleach plant effluents comprise a highly
significant proportion of the total process effluent arisings. In the case
of advanced future designs, the bleachery effluent will make up 71% of
the effluent flow although the total volume generated per tonne of air
dried pulp could be around a third of current levels.
Projected reductions in effluent arisings are likely to be achieved
by extensive internal collection and recycling, and the diversion of some
process effluents to feed other parts of the production cycle (Edde 1994).
One of the key factors on which this depends is the quality of bleachery
effluents and the degree to which they can be reused within the process
and minimised in volume. Since chlorine and chlorine-chemical bleaching
agents produce an effluent high in chloride they cannot be readily processed
through the recovery cycle. The chloride ion, as hydrochloric acid is highly
corrosive, giving the potential for leakage from the boiler tubes and the
risk of serious explosion if water reaches the melt of minerals in the
bottom of the boiler (McDonough 1992).
As noted by Albert (1995a) the only reason that process effluents are
generated from a bleached kraft mill is due to the need to purge dissolved
and suspended materials from the system which would otherwise inhibit production
or affect product properties. If these can be dealt with through internal
measures then closure of mill circuits and reduction of effluent volumes
becomes feasible.
BLEACHING
AGENTS and PROCESSES FOR CHEMICAL PULPS
The whiteness of produced pulp has traditionally been regarded as an
index of quality not only within the industry, but also by the consumer.
The whiteness of pulp is measured by its ability to reflect monochromatic
light in comparison to a known standard, usually of magnesium oxide. Unbleached
pulps exhibit a wide range of brightness values. Kraft pulp is generally
dark brown in colour while sulphite pulps are a light yellow-brown.
The principal aim of pulp bleaching is to increase the brightness of
the pulp. The chromophoric (light absorbing) components in pulps are predominantly
functional groups of degraded and altered residual lignin which is both
darker and more tightly bound to the fibres than the original lignin component
(McDonough 1992). This can either be converted and stabilised (lignin preserving
bleaching) or removed (lignin removing bleaching). In the production of
dissolving pulps, the bleaching stage is also regarded as part of the refining
process, helping to produce a pure pulp with high alpha-cellulose content.
In the less highly refined pulps, bleaching is regarded as removing wood
extractives and bark specks, and conferring superior strength characteristics.
Mechanical pulps are bleached using oxidative chemicals, predominantly
hydrogen peroxide.
| Stage |
Chemicals Used |
Symbol |
| Chlorine |
Cl2 |
C |
| Alkaline extraction |
NaOH |
E |
| Hypochlorite |
NaOCl+NaOH |
H |
| Chlorine Dioxide |
ClO2 |
D |
| Peroxide |
Na2O2+NaOH |
P or P/E |
| |
H2O2+NaOH |
P or P/E |
| Oxygen |
O2+NaOH |
O |
| Chlorine/ ClO2 |
Cl2/ClO2 |
CD |
| Sequential Bleach |
ClO2/ Cl2 |
D/C |
| |
Cl2/NaOCl+NaOH |
C/H |
| |
ClO2/NaOCl+NaOH |
D/H |
| Mixed Bleach |
Cl2+ClO2 |
C+D |
| Low level chlorine |
Cl2 |
(C) |
| Gas phase bleaching |
Cl2 |
Cg |
| |
ClO2 |
Dg |
| Ozone |
O3 |
Z |
| Acid |
CH3CO3H (example) |
A |
TABLE 6. Bleaching chemicals and process short hand nomenclature
from Fengel & Wegener (1989).
|
Pulp Type |
Sequence |
| Sulphite and Bisulphite Pulp |
|
| 3-Stage |
C-E-H |
| 4-Stage |
C-E-H-H |
| |
C-E-H-D |
| |
C-E-D-H |
| |
C-C-E-H |
| |
C-H-E-H |
| |
H-C-E-H |
| |
C-E-D-D/H |
| |
C+D-E-H-D |
| |
E-C-H-D |
| Five Stage |
C-E-H-D-H |
| |
C-C-E-H-H |
| Kraft Pulp |
|
| Three stage semi-bleached |
C-E-H |
| |
D/C-O-D |
| Four stage part semi-bleached |
C-E-H-D |
| |
C-E-H-P |
| |
C-E-H-H |
| |
C-H-E-H |
| |
C-D-E-D |
| |
O-C-E-H |
| |
O-C-E-D |
| |
O-D-E-D |
| |
O-D-O-D |
| Five Stages |
C-E-H-P-D |
| |
C-E-H-D-P |
| |
C-E-H-E-H |
| |
C-E-D-E-D |
| |
C-E-D-P-D |
| |
C-E-H-E-D |
| |
C-H-D-E-D |
| |
D-E-D-E-D |
| |
C-C/H-E-H-H |
| Six Stages |
C-H-E-D-E-D |
| |
C-E-H-D-E-D |
| |
C-E-H-E-H-D |
| |
C-E-H-D-P-D |
| |
C-E-H-E-D-P |
| |
C+D-E-H-D-E-D |
| |
O-C-E-D-E-D |
| |
O-C+D-E-D-E-D |
| |
O-D-E-D-E-D |
| |
O-C-D-E-H-D |
| Seven Stages |
C-H-H-D-D-D-P |
TABLE 7. Established pulp bleach sequences showing the predominant
role of Cl2 and ClO2 in the industry.
| Application |
Bleach Sequence |
| Reduced Chlorine |
(C)-P-H |
| |
(C)-P-D-H |
| |
(C)-P-H-D-H |
| |
|
| Peroxide replacement |
P-D-P |
| |
P-D-H |
| |
P-H-H |
| |
P-H-D |
| |
D-P-D |
| |
P-H-D-H |
| |
P-D-P-D |
| |
|
| Oxygen Bleaching |
O-P |
| |
O-D |
| |
O-H |
| |
O-P-D |
| |
O-D-P |
| |
O-C-P |
| |
O-H-P |
| |
O-C-P-D |
| |
O-D-P-D |
| |
O-Cg-E-Dg |
| |
O-Dg-E-Dg |
| |
|
| Ozone Bleaching |
Z-E-P |
| |
Z-E-Z |
| |
Z-E-Z-P |
| |
|
| Peracaetic acid |
P-A-P |
| |
A-E-A-E-A |
| |
|
| Others |
OQP |
| |
OQPP |
| |
OQPZP |
TABLE 8. Pulp Bleach Sequences designed to reduce or eliminate
the use of chlorine based compounds and chlorine from Fengel & Wegener
(1989).
The chemical pulps are generally bleached using a multistage process
of three to six steps, depending upon the pulp characteristics. Hardwood
pulps generally require less bleaching than softwood pulps content. Sulphite
and bisulphite pulps are more easy to bleach than kraft pulps, and can
be manufactured entirely without bleaching with chlorine or its compounds
although chlorine bleaching is used. The following text, therefore, largely
considers the bleaching of kraft pulps which is the dominant production
process on a global basis.since they have a lower lignin.
The historical use of bleaching chemicals in the pulp industry parallels
their use in textile bleaching applications. In part this is due to early
paper being manufactured from rags. The early bleaching of paper manufactured
from non-wood fibres was carried out using sunlight (Farr et al.
1992). At the end of the 19th century, hypochlorite was introduced, and
later elemental chlorine. In some cases, a combination of these chemicals
was used and generally, an intermediate alkaline lignin extraction step
was introduced (Fengel & Wegener 1989). Some commonly used bleach sequences
together with the identifying keys used, are given in Tables 6, 7 &
8.
Table 6 shows the key role of chlorine and chlorine chemicals in traditional
bleach sequences in the form of elemental chlorine or as chlorine dioxide
and hypochlorite. Common industrial bleaching sequences are shown in Table
7. Table 8 shows non-traditional sequences which are designed to reduce
or replace chlorine. Historically, in most commercial bleaching processes,
chlorine bleaching is the first step carried out at a level of 3-4% Cl2
at 20-40oC for 30-60 minutes The residual lignin is converted to water-
and alkali- soluble degradation products by the chlorine. This step is
generally followed by an alkaline extraction step to remove these components.
The final process steps involve the use of hypochlorite or chlorine dioxide.
Chlorine dioxide, despite handling difficulties, is largely replacing
elemental chlorine in the initial bleaching stages. Its perceived advantages
are: higher pulp brightness, improved fibre strength properties, lower
chemical consumption and considerable reduction in the AOX of discharged
effluents. Peroxide in combination with chlorine dioxide is often used
in the later stages of bleaching chemical pulps.
Chlorine gas became a common bleaching agent in the first thirty years
of the twentieth century (Reeve 1996). In conventional bleaching processes
around 90kg of chlorine was used per tonne of pulp produced. This figure
is stated to have now been reduced to around 25Kg per tonne and often to
around 3-10 kg/tonne (RSAES 1989) although Kringstadt & Lindstrom (1984)
state the chlorine charge to the pulp slurry to be between 60 and 70 kg
per tonne of produced pulp. Chlorine dioxide was first used on a large
scale in the 1940's and was first used in Sweden in 1947. This led to the
multi-stage CEDED bleaching sequence which allowed high brightness products
from kraft pulp without the loss of strength. The evolution of bleaching
techniques has followed pulping process developments to a very large extent.
In particular, the chemicals applied have changed as the lignin content
of the pulps entering the bleach sequence has been reduced.
The historical aspects of bleaching developments are traced by Sodra-Cell
(1996). A key factor was the development of oxygen bleaching in the process
during the 1970's followed by use of oxygen in the alkaline extraction
stages a decade or so later. Oxygen bleaching was inserted as a step after
the digestion stage and allowed the lignin content of the pulp entering
the bleach plant to be reduced from on average 5% to 2.5%. This effectively
halved the potential problem from the bleach plant effluent. In addition,
extended cooking time of the pulp in the digester also reduced the quantities
of lignin entering the bleach plant and this was commercially demonstrated
in the 1980s.
Substantial substitution of chlorine dioxide for chlorine has been known
since the 1960's but became widely accepted during the 1980's as a result
of growing concerns about the environmental impacts of organic chlorine
compounds formed during chlorine bleaching. Complete substitution led to
elemental chlorine free processes (ECF) becoming dominant first in Sweden
and then in Canada and this is widely attributed to market pressures. Even
so, cost considerations, particularly in the US appear to have inhibited
the application of chlorine dioxide. Chlorine dioxide is relatively more
expensive than chlorine with respect to the bleaching power per unit mass.
Further development led to bleaching processes that did not use chlorine
based chemicals (totally chlorine free or TCF). Development of hydrogen
peroxide and ozone bleaching technology was an essential prerequisite of
the commercial feasibility of such methods.
Overall, various bleach sequences have been employed with varying degrees
of chlorine substitution. Moreover, variation in the base technology means
that a wide range of ECF processes in particular exist. This point is discussed
further below.
FORMATION
OF ORGANOCHLORINES BY CHLORINE BLEACHING
Due to the complexity of pulp mill discharges in chemical terms, it
has become common to use various group parameters to characterise them.
In addition to COD (chemical oxygen demand) and biological oxygen demand
(BOD) the parameter of AOX is used to measure the quantities of organohalogens.
AOX denotes adsorbable organohalogen compounds and the term adsorbable
refers to the measuring technique used where the chemicals are adsorbed
onto activated carbon before quantification by a combustion technique.
Broadly, where chlorine chemicals are used, the quantity of AOX generated
in kg per tonne of air dried pulp conforms to the following approximation:
AOX = 0.1 (Cl2 charge + ClO2charge/5)
Considerable research effort has been expended on the detailed characterisation
of pulp bleach liquors, a process only partly helped by advances in analytical
techniques such as GC/MS (gas chromatography/mass spectrometry). This has
shown effluents from pulp and paper bleach plants using chlorine to be
extraordinarily complex. Suntio et al. (1988) published a compilation
of over 250 chemicals present in pulp mill effluents. 180 of those listed
are chlorinated compounds comprising phenolic, together with neutral and
acidic compounds. The chlorophenolic compounds, particularly chlorophenols,
catechols, guaiacols and syringols are important components and have, for
example, been isolated from samples taken in every sub-basin of the Baltic
Sea (Sodergren 1993). A proportion of these compounds appear to be derived
from the degradation of high molecular weight chlorinated material formed
during chlorine bleaching (Martin et al. 1995). It is only recently
that reference compounds for some of these exotic chemicals have been synthesised
for analytical purposes (Smith et al. 1994 a & b; Hyotylainen
1994).
Buser et al. (1989) identified a series of methyl-, polymethyl-
and alkyl-polychlorodibenzofurans in pulp mill sludge and sediment which
had previously been misidentified as polychloroxanthones and polychloro-xanthenes.
van Loon et al. (1990) have described the analysis of chlorolignins
in pulp mill effluent, although these were not quantified. A significant
component of the high molecular mass compounds was identified as chlorolignosulphonic
acids by van Loon (1992) in effluents discharged to the River Rhine. Chlorinated
diones and enol lactones with mutagenic properties have been identified
(McKague et al. 1988) while subsequent studies (McKague et al.
1989) have revealed the presence of trichlorothiophenes, compounds with
significant potential for bioaccumulation. McKague et al. (1990)
also identified chloroacetones at a number of mills. Low chlorinated PAHs
present in pulp effluents were isolated by Koistinen et al. (1992).
Elemental chlorine reacts primarily with residual lignin to produce
approximately 4kg of organically bound chlorine per tonne of pulp produced,
although this figure will vary according to the kappa number of the pulp
being bleached and the specific bleach sequence. Various types of reaction
mechanism can be involved and the diversity of such reactions is usefully
described by Bergman et al. (1994). About 70% of the organically
bound chlorine is present as high molecular mass material in spent liquor
from the bleach. In the extraction liquor about 95% is bound as high molecular
mass compounds with high carbon to chlorine ratios. Although these are
thought, in themselves, to be biologically and toxicologically inactive,
no data exist concerning the activity and fate of their degradation products.
In addition, their high relative molecular mass coupled with their polarity
renders them difficult to analyse by established chromatographic techniques
(Kringstadt & Lindstrom 1984).
Around 30% of the organically bound chlorine is found in the spent chlorination
liquor and 5% in the alkali extraction liquor as compounds of low molecular
mass. Quantitatively, the most important is trichloromethane which may
be produced in quantities of up to 40g per tonne of produced pulp. Trichloroethene,
pentachlorobenzene and dichlorophenol are also produced in significant
quantities together with a number of chlorinated phenolic compounds. Kallqvist
et al. (1989) reported emissions of 5.7kg/hour of dichloromethane
and 0.47 kg/hour trichloromethane from one pulp mill studied. A total of
119g/hour of various chlorophenolic compounds were also emitted. The derivatisation
of pulp and paper effluent phenolics for analytical purposes is described
by Lee et al. (1989) who considered the routine monitoring of these
compounds to be important.
The biological impact of many of these chemicals remains unknown. There
is now considerable evidence that metabolism and degradation of pulp mill
derived organochlorines can considerably complicate an already complicated
picture. For example, in a study of organochlorines accumulated by mussels
in a bleach plant recipient at levels of 2045 ug/g AOX in the lipid fraction,
only 1.1% could be related to identified organochlorines (Pellinen et
al. 1994). Other work (Wesen et al 1990) demonstrated that between
10 & 15% of EOCl in fish can be related to known compounds while in
sediments the figure is nearer 5%. It is thought that in fish, large amounts
of chlorine may be present in the form of chlorinated fatty acids whose
origin is not known but may result from biological transformations of more
complex chlorinated organic molecules.
Chlorinated phenols are known precursors of the polychlorinated dibenzo-p-dioxins
(PCDDs) and the dibenzofurans (PCDFs). Consequently, the discovery of the
chlorinated phenols in spent bleach liquor has led to pulp and paper plants
being identified as sources of the highly toxic and bioaccumulative PCDDs
and PCDFs. The industry is widely acknowledged to be one of the major sources
for the contamination of aquatic systems in N. America and Scandinavia
where most research has been carried out into the bleaching of pulp using
chlorine compounds (Swanson et al. 1988; Whittemore et al.
1990). In addition to the chlorophenolics derived from wood, oil based
defoamers have also been identified as precursor materials (see: Servos
et al. 1995). The congener of principal interest is 2,3,7,8-tetrachlorodibenzo-
-p-dioxin. This is the most toxic congener but other members of the group
are also formed by chlorine bleaching. A study of a New Zealand wood processing
facility (Campin et al. 1991) clearly demonstrates the role of chlorine
bleacheries in the overall mass balance of PCDDs and PCDFs. In addition,
contamination of the site with pentachlorophenol, originally used for sapstain
control, was also identified as contributing to the overall dioxin loading
to the receiving ecosystem.
Estimates made of organochlorine discharges from the Swedish industry
in the 1980s, before process changes to ECF and TCF were widely implemented
indicate that the quantities were extremely large. Wigilius et al.
(1988) estimated that Swedish kraft mill effluents contained around 200kt
of organochlorines corresponding to approximately 15kt of bound chlorine.
It can been estimated on this basis that 5kg organically bound chlorine
produced per tonne of pulp corresponds to a potential global discharge
of 250kt of AOX. The actual mass of compounds discharged will be somewhat
higher since empirical formulae range from C9H9O4Cl to C14H10O8Cl. Hence
the chlorine comprises a low proportion of the relative molecular mass.
On the basis of the Swedish figures, 250kt of organically bound chlorine
represents around 3000kt of organochlorine compounds. This, of course,
represents a worst case scenario. Nevertheless, the perceived improvements
in developed nations must be considered alongside the growth of pulping
in developing countries where elemental chlorine is still very widely used.
Evidence mounted throughout the 1980s that mill discharges were responsible
for increasing levels of organochlorines detected in receiving environments
and a variety of ecological effects ranging from lethal effects upon marine
macroflora to deformities and physiological abnormalities in fish. This
led to a number of comprehensive and case specific studies being carried
out (see: Sodergren & Wartiovaara 1988; Sodergren 1989; Sodergren 1993;
Lindesjoo 1992; Forlin et al. 1995). These broadly confirmed the
extent and scale of the ecological impacts of bleachery effluents from
pulp mills and became highly influential in the formulation of regulatory
policy worldwide. It has been estimated (Wulf & Rahm 1993) that around
half the organochlorines produced by the Baltic pulp industry since the
1940s are still present in the sediments of the region.
In the North American industry, analysis of pulp mill wastes showed
significant levels of chlorinated dioxins to be present. In bleached kraft
pulp samples 2,3,7,8-TCDD and TCDF were found at levels ranging between
1-51 parts per trillion and 1.2-330ppt respectively. Primary sludge content
ranged between 3.3 and 180ppt TCDD and 34-180 ppt TCDF. Levels in final
effluent ranged between 3-120 parts per quadrillion for TCDD and between
7-2,200 ppq for TCDF (see: Harriman Chemsult 1989).
Bioaccumulation of the chlorinated dioxins and dibenzofurans in commercially
important aquatic species led to the closure of several fisheries in Canada
in late 1988 (Hocking 1991). The potential importance of local discharges
upon fisheries and fish consumers may be assessed from the work of Svensson
et al. (1991). They found that subjects graded as having a high
fish consumption had blood levels of dioxins around three times higher
than average consumers. The scale of the problems encountered is reported
by Whittle et al.(1993). A case specific US study is described by
Schell et al. (1993) while bioaccumulation of dioxins in a Canadian
river system is described by Owens et al. (1994).
As a result of the complexity of bleachery effluents in relation to
their organochlorine content, there has been a trend towards monitoring
their discharge by means of the group parameter AOX. AOX (Adsorbable Organic
Halogens) measures the quantity of organically bound chlorine present in
a sample. While useful as a process control monitoring index, it has important
limitations which prevent its use as a predictor of environmental impact.
The most obvious is simply that it gives no indication of the nature of
the substances present (see eg Folke 1991; Neilson et al. 1991).
Nonetheless, this parameter is a widely used index of the quality of pulp
mill effluents.
POLLUTION
CONTROL MEASURES
In the face of growing political and public concern, the pulp and paper
industry began to research and implement ways of reducing organochlorine
formation and discharge from mills reported as AOX in addition to COD and
BOD discharges. These can be broadly categorised as process internal and
process external measures.
a) Process Internal Measures
Various process modifications have been implemented by the industry
to reduce and contain spills arising from pulping and chemical recovery
operations and these also lead to economic benefits by reducing chemical
losses (Downs 1989). Nonetheless, these measures do not address the problem
of bleachery effluents which account for a substantial proportion of total
mill effluents.
Process internal measures to reduce mill organochlorine and other polluting
outputs centre around increased removal of lignin before the pulp is sent
to the bleach plant (see: Gullichsen 1991; McDonough 1992). Enhanced lignin
removal allows modification of the bleaching process and agents and reduces
the charge of bleaching chemicals used. In turn, this positively affects
most bleach plant effluent quality indices, including BOD, COD, Colour
and AOX. In part this can be achieved by improved washing of the brown
pulp at various intermediate stages (brownstock washing). This gives a
lower loss of lignin into the bleach plant. This is particularly true of
washing after extended and oxygen delignification stages. In these cases
the lower lignin content makes lignin removal of greater relative importance.
i) Extended delignification
The degree to which lignin can be removed by extended cooking of the
woodchips alone is limited. Over cooking results in a reduction of the
strength qualities of the pulp (Minor 1982; Fengel & Wegener 1989).
While AQ pulping with anthraquinone as an additive allows cooking to lower
kappa numbers without loss of pulp quality, the additive is expensive and
this has restricted its use. Extended delignification without strength
loss can be achieved by ensuring that the alkali concentration stays as
constant as possible throughout the cook (Gullichsen 1991). Using methods
such as Modified Concurrent Countercurrent (MCC) cooking or Rapid Displacement
Heating (RDH) the lignin content of softwood kraft pulp can be reduced
from 5% to less than 4%.
ii) Oxygen delignification
The use of oxygen in combination with alkali was first practised commercially
in 1970 (McDonough 1992). It was first used in a pre-delignification stage
to remove about half of the lignin from unbleached pulp. A great advantage
of this process, is that unlike chlorine based bleach effluents, the effluent
can be recycled, concentrated and sent to the chemical recovery system
(Albert 1995a). This procedure is now widespread: most new mills and mill
expansions include it while the use of oxygen to enhance the effectiveness
of the first caustic extraction stage is regarded as almost universal.
Delignification is generally limited to between 40 & 50% since oxygen
lacks the selectivity needed to take the process further. Accordingly,
research has been directed at developing chemical pretreatments to improve
oxygen selectivity and hence avoid pulp strength loss. An example is the
PRENOX process (Gullichsen 1991), although this has not evolved to the
stage of commercial development and application. Oxygen has become one
of the main delignification technologies after steady growth during the
1980's and early 1990's (Hassi 1993) and according to industry figures
(see: Reeve 1996 a & b) by 1994 some 50% of world capacity for kraft
pulp incorporated an oxygen delignification stage. According to these figures
proportional penetration of this technology is significantly less in the
USA and Canadian industry as compared to the Scandinavian industry.
Improved delignification processes allow modification of the subsequent
bleach sequence by reducing the quantities of chemicals required to produce
pulps of the commercially requisite brightness.
iii) Modification
of the Bleach Process
The second major area of process internal measures designed to reduce
organochlorine generation is modification of the bleach sequence. Most
organochlorines are formed in the bleach plant due to the action of molecular
chlorine. Hence, elimination of its use will significantly reduce their
formation. In 1990, chlorine used in pulp and paper treatment accounted
for 5% of total chlorine use in Western Europe. Current figures suggest
that in combination with water treatment, this sector accounts for around
2% of production. Over a similar time period, elemental chlorine use has
fallen from 20% to 13% of the total chlorine market in the US reflecting
similar, although markedly less pronounced, trends (Shelley 1990; Botha
1995).
Essentially, elemental chlorine has disappeared from use in Scandinavia.
Bergman et al. 1994 state that it is no longer used in Sweden, although
they note that large quantities are still used for pulp production in the
US. Albert (1995a) notes that currently over 50 bleached kraft pulp mills
worldwide are producing pulp without the use of elemental chlorine (ECF)
and some 10 kraft plants are producing pulp without the use of chlorine
chemicals (TCF). Albert (1994) lists over sixty mills producing TCF pulp.
These figures, however, include sulphite and CTMP lines as well as kraft
mills. The latest figures suggest that some 15-20 TCF kraft mills exist
but not all are fully converted and use chlorine dioxide as a bleach agent
on occasion.
Elemental chlorine use has been substituted to varying extents by chlorine
dioxide in bleach plants. According to McDonough (1992) substitution of
elemental chlorine by between 50 & 70% chlorine dioxide has been practised
for a long time but in relatively few mills. Chlorine dioxide has also
been used widely in the final brightening stages of pulps at concentrations
of between 0.1 and 1.2%. Figures provided by Gullichsen (1991) indicate
that chlorine dioxide substitution of 90% will lead to AOX discharge levels
equal to those achieved by extended delignification, oxygen and effluent
treatment. Overall, in combination, these process modifications mean that
whereas traditional processes discharge 5-10 kg of AOX per tonne of air
dried pulp, mills may now discharge 1.5kg AOX/ADT.
The term ECF can embrace a wide range of process technologies. This
has led to considerable confusion and debate in the technical literature
as to the relative merits of ECF and TCF in environmental and marketing
terms. A good example of poor use of data is furnished by AET (1996). The
data show that ECF pulp production is rising faster than for any other
sector in the pulp industry and provides various quotes demonstrating that
market demand for TCF is limited, and that the environmental impacts are
equal to those of TCF. The counter argument is given by Gunnarsson (1995).
This turns upon the fact that as of 1994 more than 50 mills of a total
of 87 in the US were using old pulping technology without extended cooking
and oxygen delignification.
Hence, the term ECF is often applied both to pulp with a high kappa
number likely to generate large quantities of AOX and that produced with
a low kappa number, implying that much less AOX would be formed. Overall,
only 15 million tonnes of the total US ECF pulp production was produced
using technology comparable to the TCF processes in Scandinavia. Partly
as a result of the broad spectrum of ECF processes, and chlorine gas users
present in the industry, discharges of AOX may range from 0.1 to 10 kg
per air dried tonne, and the average value cannot be easily determined.
It is stated elsewhere in the literature that without oxygen delignification
and an extended cooking step, ECF bleaching alone will still result in
AOX loadings of up to 1kg per tonne of pulp produced (Axegard et al.
1993).
Albert (1995 a & b; 1996) has consistently argued that chlorine
chemical bleach based processes, including ECF, are unlikely to meet the
terms of emerging legislative frameworks. The proposed United States Effluent
Limitation Guidelines apply stringent controls to AOX, COD, Colour, BOD
and TSS. The original limits proposed in 1993 would indeed be difficult
to achieve with chlorine chemical based technology. This industrial sector,
is however, still under assessment and the proposals are being progressively
modified. The final regulations to be applied are expected to be published
in early 1997. Two options appear to be favoured for application in the
industry. The first (A) is conventional pulping coupled with 100% chlorine
dioxide bleaching while the second is oxygen delignification/ extended
cooking and chlorine dioxide bleaching. These are outlined in more detail
by USEPA (1995).
Hence, a substantial difference in AOX and COD discharge limitations
can be expected depending upon the baseline technology. Unaccountably,
despite the well documented capability of TCF processes in Scandinavia
producing full brightness pulps, TCF has apparently been excluded from
the options for defining a Best Available Technology (SPPT 1996), although
TCF mills could qualify under incentives packages proposed. The proposed
1996 regulations differ significantly from those outlined originally in
1993. In particular, option (A) would result in AOX discharges almost three
times those originally suggested in 1993.
Conversion to TCF processes has been intensively investigated and now
seems established as the future system of choice for eliminating chloro-organics
from the waste stream. In such processes, no chlorine chemicals are used,
reliance being placed on bleaching agents such as peroxides and ozone.
Research into the optimal ways of using agents such as ozone as bleach
chemicals is being actively pursued (Mielisch et al. 1995) particularly
with respect to carrying out bleaching at higher pulp consistencies thus
reducing water usage. Other work has focussed on methods of optimising
hydrogen peroxide bleaching (Axegard et al. 1996).
b) Process external measures
Process external measures largely refer to the design and construction
of various types of treatment plants. The balance of employed technologies
varies between region. and include activated sludge plants (Saunamaki et
al. 1991), aerated lagoons (Stuthridge et al. 1991; Stuthridge
& McFarlane 1994) and anaerobic plants (Habets & de Vegt 1991).
Almemark et al. (1991) suggests that in Sweden, solids from aerated
lagoon treatment tend to be discharged to water, whereas in North America,
Finland and Japan, where activated sludge systems are common, sludge handling
is an important component of the waste management cycle from mills. The
ability of treatment systems to reduce the AOX component varies markedly,
but wastewaters discharged from such plant invariably contain AOX if chlorine
based chemicals are used (Stuthridge et al. 1991) while sludges
may contain up to 50g AOX per kg of dried material.
One aspect of the US proposed "Cluster Rules" is that it will no longer
be permissible to remove contaminants from effluents and simply shift them
to air, solid wastes or into products. This is of significance given the
AOX residues likely to be found in sludges from various types of wastewater
treatment plants as exemplified by a Swedish study (Almemark et al.
1991). Biological sludge residue in aerated treatment lagoons ranged from
0 to 5 kg dry solids per tonne of pulp produced. This, in turn contained
between 2 and 30g AOX per kilogram. In activated sludge processes, 5-25
kg of sludge are produced per tonne of pulp with a content of 10-25g AOX
per kg. Chemical coagulation produces 4-80 kg of dried chemical solids
per tonne of pulp and these solids contain 10-40 g AOX per kg. These figures
were derived for the Scandinavian industry prior to widespread process
conversions in the late 1980's/early 1990's, but are likely to still be
highly relevant to the industry elsewhere.
The actual disposal methods used for such sludges vary from country
to country. In Scandinavia, incineration in the bark boiler was often used,
while in the United States, landfilling appears to be the method of choice.
Almemark et al.( 1991) note that relatively little work has been
carried out on the environmental impact of the various disposal methods.
Nonetheless, as a general phenomenon it is known that landfill leachates
can contaminate ground and surface waters. Moreover, organochlorines have
been found in the flue gas of bark boilers used to incinerate sludges and
chlorinated liquors and in the fly ash generated by these operations (Kopponen
et al. 1994). These include the chlorinated dioxins and dibenzofurans.
Sonnenberg and Nichols (1995) demonstrated that adding small amounts of
E1 concentrate containing chlorine to black liquor increased emissions
of hydrochloric acid and that PCDD and PCDF emissions increased by a factor
of ten with a 1% addition of bleach plant effluent to black liquor. The
spiked black liquor was combusted in both a laboratory scale plant and
in a small pilot incineration plant.
Treatment of bleach plant effluents can reduce toxic effects, depending
upon the bleach chemicals used and the type of treatment employed, while
the move towards chlorine dioxide as a bleach chemical has also contributed
(Rosemarin et al. 1990). It has become clear that the biological
activity of pulp effluents may persist through conventional treatment processes.
For example, a highly treated effluent in Australia was reported significantly
to induce liver MFO activity in carp. Induction was poorly correlated with
organochlorines extracted from the fish muscle, but appeared to be positively
correlated with PCDD and PCDF levels measured in muscle (Ahokas et al.
1994).
A further problem is associated with the use of chlorine dioxide. This
is the potential discharge of chlorate, a potent herbicide to which marine
brown algae are particularly sensitive (Rosemarin et al. 1990; 1994).
Discharges from the Swedish Monsteras mill in the early 1980s caused the
brown alga Fucus vesiculosus to disappear from an area of 12 km2.
While this problem appears to have largely been solved through modified
effluent treatment (Landner et al. 1994), the situation outside
of Scandinavia is unclear.
Finally, nutrient loadings to receiving waters are also of importance,
and treatment plant is being designed to minimise emissions of nitrates
and phosphates (Enell & Haglind 1994). In areas such as Austria, such
plant has operated for some years, although as in Germany, the industry
is based more or less exclusively on sulphite technology (Kroiss 1994).
EFFECTS
OF PROCESS MODIFICATIONS ON ORGANOCHLORINE PRODUCTION
a) Non-dioxin compounds
and AOX
There is little doubt that process internal modifications in particular
have sharply reduced the quantities of organochlorines (see: Table 9) discharged
from bleached kraft pulp mills. This is reflected both in studies of effluents
and of recipient waters, but even where bleach sequences have been modified,
if they involve chlorine based chemicals then organochlorines are discharged
(Palm & Lammi 1995). For example, in a study of two Finnish mills which
eliminated elemental chlorine from the bleach sequence and substituted
chlorine dioxide, sharply reduced emissions (but not elimination) of the
toxicologically important chlorinated cymenes was reported (Rantio 1995).
Further work has shown that levels of chlorinated cymenes and cymenenes
emitted from a mill employing complete chlorine substitution were higher
than from a mill where substitution was only partial (Rantio 1996a &
b). This implies that variations of within-mill systems may be highly important.
In another Finnish example, levels of chlorinated PAHs in mill wastes and
products were lower in birch pulp produced without elemental chlorine as
compared to pine pulp bleached with elemental chlorine at the initial stage
(Koistinen et al. 1994a). Koistinen et al. (1994b) also reported
chlorinated fluorenes and alkyfluorenes in birch pulp bleached with chlorine
dioxide, but at lower levels than in elemental chlorine bleached pine pulp.
A further Finnish study conducted analyses of levels of trichloracetic
acid (TCA) in pine needles situated downwind of a pulp and paper mill using
only chlorine dioxide in the bleach plant (Juuti et al. 1995). The
study found a clear correlation between levels of TCA and distance from
the mills with the highest levels found close to the mills. TCA is formed
from the chlorohydrocarbon precursors. While the mills were clearly implicated
as the source of this compound, the study was conducted only a short time
after process conversion, hence the contribution of chlorine dioxide to
the phenomenon is unclear.
| PROCESS |
Kappa
number |
BOD7
kg/t |
CODCr
kg/t |
Colour
kg Pt/t |
AOX
kg/t |
| BEFORE ACTIVATED SLUDGE TREATMENT: |
| Standard Kraft |
30 |
16 |
80 |
200 |
7.0 |
| Extended cook |
24 |
14 |
60 |
150 |
4.8 |
| Oxygen |
18 |
12 |
50 |
120 |
3.5 |
| Extended + O2 |
14 |
9 |
39 |
100 |
2.8 |
| AFTER ACTIVATED SLUDGE TREATMENT: |
| Standard Kraft |
- |
2.4 |
45 |
180 |
3.9 |
| Extended cook |
- |
2.2 |
36 |
130 |
2.2 |
| Oxygen |
- |
2.0 |
31 |
100 |
1.9 |
| Extended + O2 |
- |
1.7 |
27 |
85 |
1.3 |
TABLE 9: Effects of process internal measures designed to increase delignification
of softwood kraft pulp on the quality of the bleachplant effluent. Figures
before and after anaerobic treatment are shown. Kappa number can be converted
to percentage residual lignin estimate by dividing the figure by six. Table
reproduced from Gullichsen (1991). Bleach sequence given as CD-EO-D-E-D.
Where partial substitution of the initial chlorine charge with chlorine
dioxide has been practised, untreated effluents have been found to have
significant mutagenic activity and the ability to cause chronic liver injury
(Metcalfe et al. 1995) although carcinogenic effects in fish were
apparently absent. Mutagenic agents were also detected by Rao et al.
(1994) in a mill effluent where a high degree of chlorine dioxide substitution
was employed. It is, however, not clear in either case to what extent the
observed effects were due to organochlorines or to non-chlorinated wood
derivatives. Similarly, in studies of bleaching of eucalypt pulp 70% chlorine
dioxide substitution considerably reduced but did not eliminate the production
of chlorophenolics, chloroform, chloroacetones and chloro-dimethylsulfones
(Smith et al. 1994c).
In a US study, chlorine dioxide substitution of 60-70% (Haley et
al. 1995) resulted in only a 12% decline in AOX levels although chlorinated
resin acids, chlorinated phenols, chlorinated guaiacols and catechols were
all reduced in the effluent by much higher percentages. At a New Zealand
mill with aerated treatment lagoons and using full chlorine dioxide substitution,
elevated levels of chloro-phenolic compounds were identified in the river
and riverine sediments which received the effluent. Background levels of
chlorophenolics were achieved approximately 20km below the mill (Judd et
al. 1995).
b) Chlorinated dioxins and
dibenzofurans
A follow up to the US 104 Mill study (Gillespie 1996) has shown that
discharges of chlorinated dioxins have been substantially reduced. Reductions
are thought to be due not only bleach process changes, but the replacement
of defoaming agents made from oils with a high aromatic content and contaminated
with non-chlorinated dibenzo-p-dioxins and dibenzofurans as noted by Servos
et al. (1994). The total relevance of the US study is open to question,
however, since only the 2,3,7,8-TCDD and TCDF isomers have been considered.
Other isomers are formed in chlorine bleach operations (Campin et al.
1991; Servos et al 1994). In a study of chlorinated dioxin accumulation
in fish exposed to mill effluents and employing varying degrees of chlorine
dioxide substitution showed that in general the highest levels were found
close to mills with lower percentage substitution and lower intensity of
effluent treatment. The importance of historical inputs was not clear but
could have been a confounding factor (Servos 1994).
Rappe & Wagman (1995) analysed brownstock, ECF and TCF pulp and
identified 2,3,7,8-TCDF in chlorine dioxide bleached pulp but at levels
very much lower than previous reports in the literature. This is tentatively
attributed to the formation of free chlorine from the chlorine dioxide.
No effluents were tested in this study. Nonetheless, air sampling at a
Finnish mill converted to bleach using only chlorine dioxide detected elevated
levels of several chlorinated dioxins and furans, with the furans being
the dominant components (Rosenberg et al. 1994). In addition, levels
of 1,2,3,4,6,7,8-HpCDF, the dominant furan in the air samples, were higher
in the blood of a group of workers at the mill than in the background population
(Rosenberg et al. 1995).
Results of monitoring for chlorinated dioxins and dibenzofurans are
available for the Champion Pulp Mill, Canton, North Carolina at various
points in the effluent treatment process. According to Gleadow et al.
(1996) this mill has two bleach lines applied to pine and hardwood pulp,
using full chlorine dioxide substitution. Of the analytes determined, 2,3,4,6,7,8-heptachloro-dibenzo-p-dioxin
was found in the influent to the waste treatment plant, the sludge from
the treatment plant and in the leachate from the sludge. 2,3,4,6,7,8-HexaCDF
was present in effluent from the treatment plant. 2,3,4,6,7,8-HeptaCDF
was detected in effluent, sludge and leachate and 2,3,7,8-TetraCDF was
detected in the sludge.
| Mill & Date |
Bleach Sequence |
TEQ pg/l |
TEQ m g/tADP |
| A [89/90] |
O(C92+D8)(EO)DED |
2.8 |
0.4 |
| B [89/90] |
O(D30+C70)(EPO)DED |
0.9 |
0.03 |
| C [89/90] |
O(D30+C70)(EPO)DED |
0.6 |
0.03 |
| D [90] |
(D80+C20)(EPO)DED |
1.0 |
0.04 |
| E [90] |
(D80+C20(EPO)DED |
0.1 |
<0.01 |
| F [90] |
O(D32+C68)(EPO)D(EP)D |
1.2 |
0.07 |
| G [90] |
O(D32+C68)(EPO)D(EP)D |
0.5 |
0.03 |
| ECF-1 [92] |
OD(EO)D(EP)D |
N/D |
N/D |
| ECF-2 [92] |
OD(EO)D(EP)D |
0.2 |
0.02 |
| ECF-3 [93] |
OD(EO)D(EP)D |
0.9 |
0.07 |
| ECF-4 [93] |
OD(EO)D(EP)D |
0.1 |
<0.01 |
| ECF-5 [93] |
OD(EO)D(EP)D |
0.3 |
0.03 |
| ECF-6 [91] |
OD(EPO)D(EP)D |
0.4 |
0.01 |
| ECF-7 [91] |
OD(EPO)D(EP)D |
1.0 |
0.02 |
| ECF-8 [91] |
OD(EPO)D)EP)D |
0.9 |
0.04 |
| ECF-9 [91] |
OD(EPO)D(EP)D |
0.6 |
0.03 |
TABLE 10: Concentrations of chlorinated dioxins found in
whole mill effluent expressed as Nordic Toxic Equivalents. The bleach sequences
for each plant are shown. Effluent was as follows: A: untreated, B: untreated,
C: aerated lagoon, D: untreated, E: activated sludge, F: untreated, G:
aerated lagoon (pilot), ECF-1: untreated, ECF-2: aerated lagoon, ECF-3:
untreated, excluding rinse, ECF-4: untreated, including rinse, ECF-5: aerated
lagoon, ECF-6: bleachery, ECF-7: bleachery, aerated lagoon, ECF-8: untreated,
ECF-9: aerated lagoon.
On the basis of quarterly testing, the presence of these compounds was
found to be variable. It is not clear whether Octa-CDD and Octa-CDF were
tested for in this study. These results suggest that industrial claims
that ECF alone can fully address the problem of chlorinated dioxin generation
within the process may be over optimistic. In addition, a recent Swedish
study investigated the concentrations of chlorinated dioxins in whole mill
effluent from a number of mills using different bleach processes (Miljo93
1995). These results clearly show that chlorinated dioxins and dibenzofurans
of toxicological significance are present in the wastewaters from a number
of mills employing partial and complete chlorine dioxide substitution.
Hence, while chlorine dioxide substitution appears to reduce dioxin
discharges, they appear to be present at detectable levels, in contradiction
of the North American view that these process internal changes reduced
them below detection limits (Gillespie 1996). The potential production
of 2,3,7,8-TCDF is also recognised in the USEPA "Cluster Rule" proposals
where an effluent limit of 24.1pg/l is suggested (SPTT 1996).
It is clear from the limited number of studies carried out on mills
employing partial and full chlorine dioxide substitution, with or without
oxygen delignification, that while AOX discharges may be reduced by such
process internal measures, they are not eliminated. Neither is the production
of toxicologically significant dioxins eliminated.
CHLORINE
AND CHLORINE DIOXIDE LINKAGE
It has been suggested by Rappe and Wagemann (1995) that the explanation
for chlorinated dioxins found in brownstock was due to the presence of
elemental chlorine in the bleaching gas. Reeve et al. (1995) established
that decreasing the pulp consistency, increasing the pH or reducing the
chloride ion at low pH all tended to reduce AOX discharged with the effluent.
The Swedish study Miljo93 (1996) found chlorinated dioxins present
in ECF effluents. These findings suggest that ECF is an incorrect term
due to the inevitable presence of elemental chlorine. Hence ECF processes
cannot correctly be described as chlorine free under existing definitions.
For example, The Confederation of European Paper Industries (CEPI 1992)
has published the following definition of ECF:
ECF (Elemental Chlorine Free) refers to a pulp bleaching process
which does not use chlorine gas.
Elemental chlorine can be present in chlorine dioxide through two routes.
Production of chlorine dioxide is accompanied by the co-production of elemental
chlorine. All commercial processes are based upon the reaction of a reducing
agent with sodium chlorate. This may be hydrogen chloride, sodium chloride,
sulphur dioxide or methanol. The sulphur dioxide process (Mathieson process)
produces less elemental chlorine than other reducing agents (Clapper 1978;
McDonough 1992). A proportion of the elemental chlorine can be absorbed
in the scrubbing tower used to prepare the bleaching solution. In systems
using chloride as the reducing agent, the chlorine dioxide may have up
to 15% by weight of chlorine (Rapson & Strumila 1979). Where a sulphur
dioxide based system is used, the produced chlorine dioxide can contain
between 1 and 5% elemental chlorine (Kaczur & Caulfield 1994).
Fredette, (1996) notes that where sodium chloride is used in the R2
process, it produces around 0.6t of chlorine per tonne of chlorine dioxide.
Approximately 1g/l appears in the chlorine dioxide solution while the balance
is used to manufacture hypochlorite. Virtually all R2 units were sold in
the Southern United States. Subsequent R series and SVP series plants were
designed to reduce waste acid output, but increased levels of 2g/l of chlorine
were present in the chlorine dioxide solution. Eventually, hydrochloric
acid based systems were developed. This could be made from the by produced
chlorine. Even in plants using methanol, chlorine is not eliminated due
to the need for chloride ion to be present in the reaction mixture. In
the R8 process 0.1g/l of chlorine are present in a solution of 10g/l of
chlorine dioxide. Hydrogen peroxide processes which can eliminate the chlorine
by-production entirely are limited by the high cost of using peroxide in
these processes. CEFIC (1993) also recognise the by-production of chlorine
in chlorine dioxide generation. Reeve (1996b) notes that with the R2 &
R3/SVP producing a 10g/l chlorine dioxide solution, 10% of the oxidising
equivalents are provided by chlorine.
Even if this problem of chlorine by-production is solved, free chlorine
is also generated during bleaching with chlorine dioxide as noted by Rapson
and Strumila (1979) and Reeve et al. (1995) Hypochlorous acid is
generated as an intermediate. Chlorine dioxide reacts with the pulp to
produce bleached pulp and chlorous acid. A pH dependent equilibrium then
becomes established between the chlorous acid, chlorite ion and hydrogen
ion. The concentration of chlorous acid becomes lower with increase in
pH value. Chlorous acid is highly reactive towards lignin and in the course
of the reaction is reduced to hypochlorous acid. In the presence of chloride
ion, the hypochlorous acid enters into another pH dependent equilibrium
with free chlorine being evolved. This free chlorine is available to produce
chlorinated organic compounds measured as AOX. Hence,"elemental chlorine
free", where this means "molecular chlorine free" is not true of any bleach
sequence involving chlorine dioxide. Gaseous chlorine is inevitably evolved
either in the production of chlorine dioxide or in the pulp mixture to
which it is applied. Ultimately, therefore, it appears that only those
commercial processes entirely free of chlorine chemicals as bleach agents
merit the term of elemental chlorine free.
BIOLOGICAL
IMPACTS OF MILL DISCHARGES-RECENT RESEARCH
Recent research work has established that even with reduced AOX discharges,
mill effluents can exert significant impacts upon the receiving environment.
Some follow-up work to the original Swedish studies (Sodergren 1989; 1993)
have been conducted upon receiving environments for bleached kraft pulp
mill effluents in order to gauge the effects of conversion from elemental
chlorine to other bleaching agents. These have generally shown that following
process changes, some recovery of the environment takes place but this
is not complete. Sandstrom (1994) reported on abnormalities in fish populations
exposed to mill effluent. Following the introduction of chlorine dioxide
at the Norrsundet Mill in Sweden, some abnormalities were no longer recorded.
Nonetheless, the abundance of adult perch was low and high levels of larval
mortality and low embryo quality were recorded in the area within 1-2km
of the mill.
Kankaanpaa et al. (1995) investigated the effect of effluent
from a mill using 100% chlorine dioxide in the bleach process upon the
behaviour of an ecologically important Baltic Sea amphipod. The effluent
contained 8.3mg/l AOX. Negative impacts were observed upon swimming activity
in the test animals following exposure to the effluent at AOX levels diluted
to between 150 and 400 micrograms per litre. Kovacs et al. (1995)
found that secondarily treated effluent from a plant employing 45% chlorine
substitution for elemental chlorine suppressed reproduction in test fish.
Effluent from a mill using 70% chlorine dioxide substitution caused
induction of liver mixed function oxidase enzymes, an index of pollution
stress, in largemouth bass. Effluent concentrations of 8% and 4% effluent
were used in artificial streams. Even after oxygen delignification was
installed and full substitution of the chlorine by chlorine dioxide, enzyme
induction resulted in fish exposed to 4% and 12% effluent in artificial
streams (Bankey et al. 1995). Similar findings were made for trout
exposed to effluent from a mill employing increased chlorine dioxide substitution
(Haley et al. 1995) and a positive dose response relationship was
observed.
Suppression of immune parameters was recorded in roach exposed to effluent
from a mill employing 25% chlorine substitution in Finland (Jokinen et
al. 1995). Further subtle effects were identified in a comparative
Canadian study which revealed that fish exposed to bleached kraft mill
effluent showed reduced investment in reproduction, reduced gonad size
and reduced fecundity when compared to fish taken from an uncontaminated
site (Gagnon et al. 1995), although the precise bleach sequence
was not given in the study. More intriguingly, the short term growth of
mayfly larvae was stimulated by exposure to effluent from a bleached kraft
mill in Canada (Lowell et al. 1995) leading the authors to speculate
that chemicals acting upon the insect endocrine system could be responsible.
An apparent stimulation of fish growth in the short term is also apparent
from the work of Haley et al. (1995).
Comparison of low level, long term effluent toxicity between mills using
a variety of bleach processes has shown that effluents derived from elemental
chlorine bleaching are indeed the most toxic. In a study of structure of
artificial ecosystems exposed to mill effluents, the least structural disturbance
was recorded in the system exposed to bleachery effluent from a fluff pulp
mill using an OPD bleach sequence. Intermediate disturbances were reported
from effluent exposures where partial chlorine dioxide substitution was
used in the initial bleaching step after an oxygen delignification step,
for example, O(C85+D15)(EO)DED. The most serious disturbances were reported
in systems exposed to effluent derived from a CEHDED sequence (Tana et
al. 1994).
These observations have been generally confirmed in the industry (Axegard
et al. 1993) where lowered chlorine charge has resulted in reduced
impacts. In mills using chlorine dioxide, AOX levels have been reduced
further. There appears to be no correlation between AOX discharge levels
and environmental impact, a phenomenon noted also by Tana et al.
(1994) in their studies of specific responses of fish. In addition, other
observations (Barker et al. 1994) have documented a variety of lesions
in fish sampled adjacent to a mill using sodium hydrosulphite as a bleaching
agent, with no chlorine chemicals in use.
Grahn & Grotell (1995) reported that some environmental impacts
were significantly reduced when totally chlorine free (TCF) bleach processes
were used in comparison to ECF processes using chlorine dioxide, confirming
earlier work (Lovblad & Malmstrom 1994). Kovacs et al. (1995)
reported that the overall toxicity of effluents tested in a bioassay was
untreated ECF> untreated TCF> secondary treated ECF> secondary treated
TCF. In previously published studies noted by Kovacs et al. (1995),
the differences between TCF and ECF derived effluents were less marked.
They note that generalisations regarding a particular process and the subsequent
effects upon the mill discharge are difficult to make.
One particular problem in making comparisons stems from the wide range
of potential ECF processes discussed above. Comparisons are strictly only
valid when comparing low kappa ECF after chlorate ion removal with TCF
effluents. A further complication in comparing between studies is the fact
in some of these mill effluents were synthesised in the laboratory rather
than taken from real mill operations. These were subjected to various non-standard
treatments and were, therefore, not representative of the products of the
different and complex reactions occurring within the mill. This fundamental
criticism can be levelled at, for example, the work of O'Connor et al.
(1993) which is widely cited as evidence of ECF effluents being less toxic
than TCF effluents.
One study found a high acute toxicity of untreated TCF effluent and
this was identified as due to residual hydrogen peroxide. The toxicity
could be removed completely by biological treatment. The treatability of
the effluents produced by TCF processes in comparison to wastewaters from
chlorine gas and chlorine dioxide bleaching of softwoods was investigated
in the same study. Activated sludge treatment reduced the COD of TCF effluents
by 55-65% as against 35-45% for wastewaters containing chlorine compounds.
Similarly, chemical flocculation produced positive 85% removal of COD from
TCF wastewaters (Saunamaki 1995).
One key concern in relation to TCF mills is the use of chelating agents
to control metal build up in the processes which could affect peroxide
and oxygen bleaching. The degree to which these could affect metal mobilisation
in the wider environment is not clear but could be significant. The concentrations
of metals present in TCF wastewaters have been found to be higher than
in other bleachery effluents, probably due to the EDTA content. The EDTA
is also responsible for raised nitrogen emissions and, in common with DTPA,
does not readily degrade under biological treatment.
Overall, such studies demonstrated that while environmental improvements
could be achieved by process changes and that the elimination of chlorine
based chemicals was a key factor in such improvements, effluents from all
processes were toxic to some degree. Indeed, these and similar findings
led to increasing suspicions that chemicals other than the AOX components
present in pulp and paper effluents were at least partially responsible
for observed changes in fish populations. It has been known for some time
that acute effects of pulp effluents were attributable to resin and fatty
acids released from the processed wood. These components present in the
effluent of a thermomechanical pulp mill were found to be lethal to trout
after 2-4 weeks exposure at a two hundred fold dilution (Johnsen et
al. 1995). Analyses of resin acids in the fish showed a positive dose-response
correlation.
Resin acid contamination in sediments was found to correlate with significant
behavioural modification in benthic invertebrate species (Hickey &
Martin (1995). The resin and fatty acids from wood, however, are considered
to be largely removed by biological treatment (see: Owens 1991) while Zender
et al. reported a 96% removal of these components by secondary treatment
in a New Zealand plant. The greatest mass flow was from the foul condensate
stream and extraction bleaching stages.
An extensive study of Canadian mills using a variety of pulp production
processes revealed that physiological changes in fish exposed to effluent
were associated with significant effects upon reproductive development.
These effects were also associated with changes in liver mixed function
oxidase activity. Moreover, the effects were recorded at very low effluent
concentrations and the causative agent or agents were not removed during
normal secondary treatment. Significantly the observed effects did not
correlate with AOX levels in the effluent or levels of dioxins and furans
(Carey et al. 1992). These studies have subsequently been reported
in considerable detail. The precise relevance of MFO induction measured
as ethoxyresorufin-O-deethylase (EROD) is open to question in view of recent
research showing that temperature can strongly influence the induction
of this enzyme and the cytochrome P4501A system in marine fish species
as well as exposure to xenobiotic chemicals. (see: Eggens et al.
1995; Sleiderink et al. 1995 a, b & c). Even so, MFO enzyme
activity was induced in fish living in the recipients of most mill wastewaters
examined.
The mills studied are described, together with the receiving water characteristics,
by Robinson et al. (1994). All discharged in excess of 40,000 cubic
metres of effluent per day and included bleached kraft producers and sulphite/mechanical
pulp mills. Induction of liver MFO enzymes and depression of plasma sex
steroid levels were identified downstream of several mills including some
without chlorine bleaching and with advanced secondary treatment (Munkittrick
et al. 1994). Other responses such as reduced gonad size and enlarged
livers were also found in fish near mills using chlorine and mills using
no chlorine chemicals. The effects in feral populations could not be predicted
from the results of laboratory toxicity tests on a variety of species.
As noted above, there was no apparent correlation with PCDD/PCDF levels
and changes in physiological parameters or the biomarkers used with the
exception of EROD induction. PCDD and PCDF levels were, however, higher
near mills using elemental chlorine than those using other bleach agents
or no bleaching at all (Servos et al. 1994). 2,3,7,8-TCCD toxic
equivalence levels were strongly and negatively correlated with testosterone
levels in fish of both sexes (Heuvel et al. 1994).
ENDOCRINE
DISRUPTION & PULP AND PAPER MILLS
The findings from the Canadian research programme coupled with observations
made in the Baltic sector focussed attention on natural plant components,
particularly those capable of interfering with the endocrine systems responsible
for hormone synthesis and metabolism. Landner et al. (1994) suggested
that natural sterols present in the wood such as a
- and b -sitosterol could be a factor since
administration of these compounds to fish induced similar responses to
pulp and paper mill effluent exposure. This theme was also taken up by
Tana et al. (1994) who noted that a proportion of these chemicals
extracted from wood would not be cycled through the recovery process, but
would be passed to the bleach and effluent treatment plants.
The current situation is analogous to that prevailing after regulatory
initiatives to control BOD and COD discharges in the 1980's and before.
Removal of BOD from effluents led to improvement in recipient quality,
but allowed the identification of effects due to other agents present in
discharges, specifically organochlorines, which had previously been obscured.
Progressive reduction of organochlorine loading similarly, has allowed
the identification of yet further impacts due to other chemicals. (see:
eg. Owens 1991).
In fact, laboratory studies as early as 1941 had established that it
was possible to masculinise mosquito-fish using artificial hormones. It
was reported in 1991 that androstenedione, androstanol and spironolactone
could achieve the same result. In 1978 it had been reported that phytosterols
present in tall oils recovered from wood could be converted by microbes
to steroids. Tall oil consists of 25-25% phytosterols. Hence, the total
phytosterol production by this route alone in the US was estimated at 20,000
tons. (see: Davis & Bortone 1992). These same changes can be produced
by exposing fish to kraft mill effluent, although the authors do not specify
whether this was derived from a chlorine chemical bleach line.
As a result, there are now two distinct elements to the debate about
environmental protection from pulp and paper operations. The Canadian findings
have prompted the industry into large scale questioning of the utility
of reducing AOX emissions any further. For example, Myreen (1994) cites
the opinion of scientists in the Nordic countries and Canada that reductions
of organochlorines below 1.5 kg AOX/ADT pulp represents an inefficient
use of resources. The precise basis for this view is obscure.
Malinen et al. (1994) advance a similar viewpoint in their analysis
of the future of the industry in Finland. The views subsequently diverge
markedly. While Malinen et al. (1994) propose the development of
treatment plant to address the residual problems, Myreen (1994) considers
that total effluent free (TEF) production is now accepted by the industry
as the decisive step towards environmentally friendly pulp and paper production.
In light of the Canadian findings and the realisation that all pulp mills
can emit endocrine disruptive chemicals on a large scale, closure of the
mill circuits may be seen as an environmental imperative for the industry,
a theme taken up and discussed by Albert (1995b). In most geographical
areas there appear to be clear present and likely future environmental,
market and fiscal incentives to move towards TEF production.
TOTALLY EFFLUENT
FREE PROCESSES-CLOSING THE BLEACH CIRCUITS
Currently, zero effluent operation appears to be restricted to plants
producing bleached chemical thermal mechanical pulp and non-chlorine bleaching
agents (Edde, 1994). As noted earlier, the bleach plant is the major source
of contaminated effluent in a kraft pulp mill. Hence closure of these circuits
is an essential prerequisite for producing a zero-effluent kraft mill.
Myreen (1994) points out that this requires the simultaneous resolution
of a number of problems. These include water balance, chemicals balance,
corrosion, precipitation of salts and removal of non-product substances.
Development work is being directed at closing both ECF and TCF bleach circuits.
Closure of the bleach plant circuits implies a large scale increase in
the levels of organic and inorganic matter in the process liquors and filtrates
which need to be accommodated (Krotscheck et al. 1995).
Albert (1995a) provides a cost breakdown for a reference kraft mill
in relation to an effluent free plant based on ozone and peroxide bleaching
which has considerable capital and running cost advantages. This allows
for the re-use of all process waters in the brownstock circuits and the
chemical recovery circuits. In addition to significant chemical cost savings,
there is no need for capital investment in primary and secondary effluent
treatment plant. A closed cycle TCF plant does not require substantially
different recovery boiler design (Gleadow et al. 1995) but may involve
the treatment of purge lines to c
