Journal of Cleaner Production 183 (2018) 1228e1240

Contents lists available at ScienceDirect

Journal of Cleaner Production

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

New classification of chemical hazardous liquid waste for the

estimation of its energy recovery potential based on existing
measurements

n-García b, *, Oliver Weder a, Konrad Hungerbühler a
Vasco Bolis a, Elisabet Capo
a
Institute for Chemical and Bioengineering, ETH Zürich, Vladimir-Prelog-Weg 1, 8093, Zürich Switzerland
b
ABB Switzerland Ltd., Segelhofstrasse 1K, 5405 Baden-Da €ttwil, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history: Waste solvents generated by the process industry are frequently incinerated as a part of their treatment,
Received 7 August 2017 thus allowing for a partial recovery of their combustion heat. While large chemical sites often have in-
Received in revised form house waste incineration facilities, smaller enterprises have to outsource the treatment of their residuals,
29 December 2017
leading to significant shipments of hazardous substances. In this sense, there is a significant optimization
Accepted 5 February 2018
Available online 12 February 2018
potential to reduce both shipments and the consumption of auxiliary fuels, thus leading to an overall
reduction of the primary energy required to produce and deliver chemical products and intermediates.
However, inconsistent data systems and incomplete information represent main barriers for investi-
Keywords:
Waste-to-energy
gating the potential of such environmental benefits. As information about waste properties does not
Hazardous waste management need to be reported, it is consequently not possible to rigorously estimate the energetic potential of the
Energy potential hazardous waste produced in a specific area, which would be a crucial first step in the optimization of
both design and management of a supply chain network of industrial waste. This work proposes a novel
approach for the estimation of hazardous waste properties, combining existing information stemming
from industrial and institutional partners. Such framework creates a new waste classification system
based on water and pollutant content that links inconsistent data sources with different degrees of detail,
and allows for estimating average properties, including the energy content, of different types of primary
produced hazardous liquid waste. The validity of the developed methodology for a broad application
range is tested with a case study about Switzerland, which, because of its sparse chemical and phar-
maceutical industry and an extreme variety in terms of generated waste residues, represents one of the
most challenging cases. Such case study investigates the evolution of Swiss industrial waste from 2010 to
2014, determines average properties for 12 different waste classes and their total energy recovery
potential.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction amount of energy is devoted to process heating, and 25% to other

direct energy requirements for operating the process (Swiss Federal
The industrial sector worldwide is responsible for 27% of the Office of Energy, 2015a), such as electricity for pumps and com-
total energy use, of which 29% is consumed by the chemical and pressors. Approximately 50% and 25% of the gross energy con-
petrochemical industry (Banerjee et al., 2012), representing about sumption in Switzerland stem from fossil and nuclear fuels,
7.8% of the global energy consumption. In Switzerland, the chem- respectively (Swiss Federal Office of Energy, 2015b). Therefore,
ical and process industry consume approximately 3.5% of the total increasing pressure to decrease the dependence on non-renewable
energy (Swiss Federal Office of Energy, 2014). About 55% of that energy sources poses a non-trivial challenge to these and other
energy-intensive industrial sectors. This situation results in an
excellent opportunity to explore alternative energy sources that
* Corresponding author. were previously considered of secondary importance. In particular,

n-García).
E-mail address: elisabet.capon@ch.abb.com (E. Capo

https://doi.org/10.1016/j.jclepro.2018.02.050
0959-6526/© 2018 Elsevier Ltd. All rights reserved.
 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240 1229

Terminology

Chemical liquid hazardous waste General expression for the liquid waste produced by the chemical,
pharmaceutical and process industry; in this publication, the words
residual, residue, (industrial) waste and waste streams are frequently
used as synonyms for this term.
Incineration site/plant Integrated chemical complex with own incineration facilities.
Cement plant Plant for the production of cement that can use certain waste streams as substitution fuel.
External site Chemical site without own incineration facilities.
Primary waste Waste produced in the production processes present in the chemical site before being processed
externally by third parties.
Secondary waste Waste produced in the production processes present in the chemical site and externally processed
by third parties.
External waste Primary waste produced by external sites.
Internal waste Primary waste produced within incineration sites.
Produced waste Total amount of generated primary waste that can be treated by incineration.
Incinerated waste Waste incinerated in the aforementioned incineration and cement plants.
Potentially incinerable waste Primary waste that could have been incinerated (in Switzerland) but has been treated differently or
exported. In this work, it is estimated by subtracting the amount of incinerated waste to the
produced one.

the process and chemical industrial sectors generate considerable As shown by recent research activities in the field of municipal
amounts of high-calorific waste solvents that are frequently solid waste, information about the chemical composition of the
incinerated as a part of their disposal (Seyler et al., 2006), thus residues leads to a more accurate quantification of the environ-
allowing for partial recovery of the combustion heat in the from of mental impact of different management strategies, helping to
steam (Capo
n-García et al., 2014). determine the best available options (Burnley, 2007) and assess the
Recent studies have shown that an appropriate management of environmental risk of hazardous residuals (Slack et al., 2007). In
waste incineration is indeed an efficient tool to decrease primary this sense, a broad knowledge about the different residues is an
energy consumption of integrated chemical sites, reducing both important aspect for supporting goal-oriented decision processes
costs and environmental load (Abaecherli et al., 2017), especially in waste management networks (Allesch and Brunner, 2017), and is
when considering plant-wide (Chakraborty et al., 2003) or a necessary condition for the development of dedicated optimiza-
enterprise-wide aspects (Wassick, 2009). However, little to no tion tools (Levis et al., 2014). Reliable information about chemical
attention has been devoted to even broader network perspectives. composition and energy content of green and municipal waste is
This is of particular interest considering that smaller enterprises particularly important for planning design and operation of ther-
have usually neither the technical nor the financial means to operate mochemical conversion systems, for which estimations of the en-
own waste incineration facilities or other treatment processes, and ergy content based on empirical models are often insufficient (Hla
have to outsource the disposal of their waste to third companies. and Roberts, 2015). As the application of similar optimization tools
Such treatment facilities are in many cases in-house incineration could potentially lead to important benefits also in the case of
plants that belong to large chemical sites, since they have the hazardous waste stemming from the chemical and process in-
necessary waste volumes to operate in a cost-effective way, as well dustry, especially considering its large volumes, it is crucial to
as the possibility to directly utilize the recovered energy in other obtain more detailed information also for such residues.
production processes on the site. These large sites analyse the Despite the fact that transportation of industrial and hazardous
composition and heating value of the different generated waste waste has nowadays reached considerable amounts, the informa-
streams. In contrast, many of the industrial waste shipments to such tion about production and management is still extremely scarce,
incinerators are often organized unsystematically based on short- even in OECD countries (OECD, 2013). In general, data is restricted
term waste generation forecasts and both intermediate storage to authorizations of transboundary movements requested by the
and treatment capacity at incineration sites, and information on Basel Convention (Basel Convention, 1989), but many European
waste quality and heating value is poor. As a consequence, there is a states, such as EU members (European Parliament and Council,
potential benefit for systematically optimizing the management of 2006) and Switzerland (Il Consiglio federale svizzero, 2005), also
both waste transportation and treatment of the whole network. For require notification and authorization for domestic shipments. In
instance, cost and environmental savings may be achieved by both cases, however, there is no information about the heating
decreasing the total transportation distance, and by reducing the use value of residuals, since the legislation only obliges to notify
of auxiliary fuels in the incineration process by using combustible amounts according to classification systems based either on the
waste solvents to burn non-combustible residuals, which would type of process generating the waste or roughly on pollutant spe-
otherwise require additional oil or natural gas. Such measures to cies. Whereas these systems are undoubtedly very useful for
reduce consumption of fossil fuels and environmental load by re- guaranteeing transportation safety, both in terms of health and
utilizing waste as a resource are definitely in line with the goals of environment perspectives, they do not provide enough data to
many recent political initiatives aimed at environmental protection estimate the energetic potential of these residuals in a specific area.
and resource efficiency, such as the Energy Strategy 2050 in Although there is a broad set of available characterization tech-
Switzerland (Swiss Federal Office of Energy, 2017) and the General niques, also in terms of energy content, the few research activities
Union Environment Action Programme of the European Union that tackled the composition of industrial residuals have been
(European Parliament and Council, 2013). limited to single plants (Seyler et al., 2005) or very specific waste
1230 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240

streams (Ghimpusan et al., 2017), presumably because of the large previously discussed, the related information focuses on safe
experimental effort required to obtain such information. For these transport conditions and does not involve specific data about the
reasons, it has not yet been possible to apply systematic optimi- energy content of the transported residuals. In contrast, most Swiss
zation tools for both design and management of a treatment incineration plants keep records of the heating value and other
network aimed at maximizing the contribution of industrial waste properties of their waste streams, but they often refer to own
incineration to a more sustainable energy system. classification systems for waste streams treated internally. There-
The Swiss chemical and pharmaceutical industry is strongly fore, Switzerland represents an ideal case study for investigating
based on high-revenue specialty chemicals, which represent the potential of a more systematic management of industrial waste
nowadays more than 90% of its portfolio, and is extremely differ- incineration, with many sparse industries producing a highly
entiated, with more than 30,000 different products ranging from differentiated set of waste streams and centralized treatment sites.
fragrances to dyes (Papadokonstantakis et al., 2013). In this country, Consequently, a rigorous estimation of the energy content of liquid
thermal treatment is compulsory whenever material recycle or hazardous residuals is a crucial first step for the design of sus-
regeneration are not technically feasible, economically viable or tainable industrial waste incineration networks, as well as for their
make no sense from an environmental perspective, i.e. when proactive management.
recycling leads to higher environmental load than ex novo pro- The aims of this work are i) to define a general classification
duction (Il Consiglio federale svizzero, 2015). Hence, many liquid system for chemical hazardous liquid waste; ii) to create a meth-
hazardous residuals undergo thermal treatment in one of the five odology linking such new classification to inconsistent information
incineration (waste-to-energy) plants located in five large chemical from waste incinerators and compulsory transport notification,
sites (VBSA, 2017), while a smaller and less polluted fraction is used characterizing several physicochemical properties for all the newly
as substitution fuel in cement plants (Cemsuisse, 2015) or other defined waste classes; and, iii) to show possible applications of such
industrial furnaces (DATEC, 2005). Furthermore, since the Swiss methodology for assessing the energetic potential of hazardous
chemical industrial sector is composed of many small-medium waste incineration with an illustrative case study. This investigates
enterprises (Swiss Federal Statistical Office, 2017b) spreading the evolution of the chemical hazardous liquid waste generated in
across all regions of the country (Swiss Federal Statistical Office, Switzerland for the period from 2010 to 2014, based on historical
2017a), waste shipments to treatment sites are required (Fig. 1). data, and estimates the total energy content of such residues in the
Shipments of all kinds of hazardous waste have to be notified (Il different regions of the country. Since it only relies on existing in-
Consiglio federale svizzero, 2005) under a specific classification formation sources, the methodology can be constantly updated
(DATEC, 2005), which is substantially equivalent to the one used in with the latest information concerning waste composition and
the European Union (European Parliament and Council, 2006). As amounts, without requiring additional time-consuming experi-

Fig. 1. The seven Grand Regions of Switzerland used in statistical studies (Swiss Federal Statistical Office, 1999), together with the five integrated chemical sites with facilities for
industrial waste incineration (VBSA, 2017) and the six Swiss cement plants (Cemsuisse, 2015).
 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240 1231

mental measurements. The generated knowledge can be used by of waste generated in production sites without incineration ca-
the incinerators to better exploit the energy content of the waste, pacity, the amounts of each class produced in a specific area can be
thus reducing the total consumption of primary energy for the estimated from transport notification data.
different production processes on the site.

2.2. Waste classification system

2. Methodology
Since internal classification systems used in treatment sites vary
2.1. Sources of information for each site and differ from the one used for transport notification,
it has been decided to define a new classification in collaboration
This work focuses on the analysis of liquid hazardous waste with the involved industrial and institutional partners (Expert
stemming from chemical, pharmaceutical and process industries, Panel 2017). Therefore, this work proposes a new classification
without considering used oils. The overall analysis relies on system for waste streams based on incineration related features
different data sources for the system description. Specifically, in- and estimates average properties for the newly defined waste
formation is given about: (i) waste generation in chemical sites that categories. These waste classes group several types of waste
possess an incineration plant; (ii) measurements of lower heating streams of similar composition and energetic properties.
value and other physicochemical properties of incinerated streams; This classification system combines aspects that are relevant for
and, (iii) waste generation in municipalities from the compulsory waste treatment in the plants with requirements of future studies
notification system for hazardous waste shipments. regarding the optimization of waste incineration management and
On the one hand, although incineration is a versatile treatment energy recovery. First, for thermal valorization purposes, it is
type, capable of simultaneously handling very different waste important to distinguish self-combustible waste streams from non-
streams, a good management is always necessary to run the incin- combustible ones that necessitate auxiliary fuels to destroy the
eration process at safe conditions while fulfilling emission limits. pollutant species contained in them. Specifically, streams with both
Therefore, incineration enterprises often perform detailed analysis a heating value HV higher than 20 MJ/kg and a weight fraction of
of the residuals to be incinerated based on specific plant re- water xwater less than 15% are hereinafter defined as combustibles
quirements and internal waste characteristics (Expert Panel 2017). (CO), while the remaining ones are referred as mother liquors (ML).
Such measurements normally include the lower heating value, the Secondly, incompatibilities might occur during mixing and incin-
amount of air required for the complete combustion of organic eration depending on the presence of certain chemical elements,
species, elementary weight composition, as well as the water mass which can also cause problems with the recycling of some material
fraction. They are used as decision support in waste management species contained in the combustion fumes. Hence, it has been
when preparing schedules for both incineration and temporary
storage units. However, these measurements usually differ widely
from site to site, and thus some data is not available for all plants.
On the other hand, the extremely differentiated chemical pro-
duction leads to an even higher variety of waste streams of dispa-
rate composition and physicochemical properties, whose
shipments have to be notified to authorities (European Parliament
and Council, 2013). The information for residues generated in
external chemical sites is also rather poor (OECD, 2013), especially
considering that it is mostly limited to shipment notification. Based
on such considerations, it has been assumed that average figures
obtained from available data (treatment sites) are representative
for the rest of the system, and so they are used for estimating
physicochemical properties and composition of uncharacterized
residuals. Hence, available data from the incineration sites has been
used to link a set of newly defined waste classes to the classification
system used for notifying waste shipments, as well as for deter-
mining average properties for each class and their evolution in
time. By assuming that such average values are also representative Fig. 2. Procedure to assign a waste residue to its corresponding class.

Table 1
Proposed hazardous waste classification system consisting of twelve categories.

Waste class Type Water content Heating value Pollutant species p

COCl Combustible < 15% and  20 MJ/kg Chlorine (xCl  2 %)

MLCl Mother liquor  15% or < 20 MJ/kg Chlorine (xCl  2 %)
COBr Combustible < 15% and  20 MJ/kg Bromine (xBr  2 %)
MLBr Mother liquor  15% or < 20 MJ/kg Bromine (xBr  2 %)
COF Combustible < 15% and  20 MJ/kg Fluorine (xF  2 %)
MLF Mother liquor  15% or < 20 MJ/kg Fluorine (xF  2 %)
COP Combustible < 15% and  20 MJ/kg Phosphorous (xP  2 %)
MLP Mother liquor  15% or < 20 MJ/kg Phosphorous (xP  2 %)
COS Combustible < 15% and  20 MJ/kg Sulfur (xS  2 %)
MLS Mother liquor  15% or < 20 MJ/kg Sulfur (xS  2 %)
CONO Combustible < 15% and  20 MJ/kg None of the above (NO)
MLNO Mother liquor  15% or < 20 MJ/kg None of the above (NO)
1232 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240

that specific notification code, resulting in a characteristic distri-

bution. An illustrative example of this concept can be found in
Fig. 3, whereas a list with the Swiss distribution for each considered
waste code is provided in Appendix A (Table A.5).
Therefore, the total amount of chemical waste is computed from
two data sources, namely i) the companies without in-house
incineration facilities, and ii) the production sites with in-house
facilities (integrated chemical sites). On the one hand, all sites
without in-house incineration facilities are grouped according to
the region in which they are located, and consequently waste
Fig. 3. Example of the Swiss waste classes distribution corresponding to notification
code 07 01 08 “Wastes from the manufacture, formulation, supply and use of basic availability corresponds to shipments. On the other hand, inte-
organic chemicals: other still bottoms and reaction residues” (European Parliament grated chemical sites are considered separately, since they are both
and Council, 2006). waste producers and waste treatment companies and thus it is not
possible to distinguish i) shipments of primary waste, from ii)
shipments outsourcing secondary residuals, which have previously
decided to further differentiate both combustibles and mother li- been accepted from external sources (either other incineration sites
quors according to their elemental composition concerning the or enterprises without treatment facilities). Hence, the information
following species p: chlorine, bromine, fluorine, phosphorous and about such integrated sites has been directly retrieved from the
sulfur. Precisely, streams with a weight composition in these companies in order to avoid double counting of certain residues. In
components xp higher than 2% are considered being part of a cor- the case of integrated chemical sites, an appropriate waste class can
responding category, while the remaining streams are grouped in a usually be assigned to each internal residue based on available
separate one. Possible conflicts concerning streams potentially measurements with the procedure explained in Fig. 2, leading to
belonging to multiple categories (i.e., with more of the aforemen- more precise results than the estimation from characteristic dis-
tioned components exceeding 2% in weight composition) have tributions (used for residues originated in companies without in-
been addressed by relying on definitions from industrial partners, if house incineration facilities).
available, and based on the most abundant component, otherwise. Although such waste class distributions are presumably
The combination of the combustibility criterion and the six country-specific, as depending on the nature of the local chemical
composition categories results in the twelve waste classes listed in industry, the connecting procedure and the newly defined waste
Table 1, whereas a more detailed description of the classification classification can be applied to most European countries thanks to
procedure can be found in Fig. 2. Such classification has also been the common notification system (European Parliament and
approved by the involved research partners for the scope of this Council, 2006) and to the general nature of the newly defined
work (Expert Panel, 2017). In this sense, it is noteworthy to mention waste classification, based on standard requirements of incinera-
that cement plants can partially substitute conventional fuels with tion processes.
some combustible residuals considered in this study. In particular,
they can be approximated a priori as a fraction of the CONO class 3. Results
because of higher purity requirements. Such fraction has been
determined with the waste analysis provided by several industrial 3.1. Characterization of the chemical hazardous liquid waste
partners.
After having defined the twelve waste categories, representative The representative properties of the twelve categories of the
average values have been determined for each class concerning proposed classification system have been obtained by averaging
heating value, required air demand for a complete combustion, and experimental measurements provided by the incineration com-
density. In a first step, each waste stream with available analytic panies as explained in the previous section. The representative
measurements has been easily assigned to the corresponding waste figures have been determined for each waste class with a weighted
class according to its characteristics. In a second step, yearly mean of all waste streams with available analytic measurements.
average values for several waste properties have been calculated for Concerning industrial waste incineration, the following properties
each class with a weighted mean based on the relative abundance are particularly relevant for modelling purposes:
of each stream within its corresponding class. Finally, an arithmetic
mean of the average yearly values has been used to obtain repre-  Heating value. It is used for determining which residues can be
sentative values over the whole period of study. simultaneously incinerated to achieve required operating tem-
peratures in the furnaces, as well as their energetic potential.
2.3. Relation between notification and waste classification systems  Combustion air (stoichiometric requirement for a complete
combustion of the residues, without considering any excess
In order to estimate properties of the waste residuals generated factor). It is related to the amount of fumes that will be produced
in industries lacking treatment facilities, it has been necessary to from incineration, and since there is a limited fumes cleaning
link the classification used in transport notification to the newly capacity, it also restricts the amount of treated waste.
defined waste classes. Therefore, analysis data from industrial  Density. It determines the maximum mass inventory levels for
partners have been used to determine the characteristic waste class given storage volumes.
distribution corresponding to each relevant notification code. In
practice, all liquid waste streams with available analysis bearing the The evolution along time of these parameters has been carefully
same notification code have been assigned to a single waste class monitored for the period 2010 to 2014 in order to unveil significant
according to the procedure presented in Fig. 2. Then, based on the changes in the waste composition. Therefore, average values have
relative abundance of all waste classes within the same notification been calculated for each of the five years in the 2010e2014 horizon.
code, it has been possible to determine its fraction with respect for The results are presented in Fig. 4. Further information about the
 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240 1233

Fig. 4. Standard box plot of the evolution of yearly average properties in period 2010e2014 for each waste class. The red line represents the median, whereas outliers and arithmetic
mean values of the yearly averages are marked with symbols “x” and “”, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the
Web version of this article.)
1234 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240

Table 2
Arithmetic mean of yearly averaged values for selected waste properties of the different categories, together with complimentary information about the sample (number of
streams and their total amount for each waste class). This consists of 824384 t of waste divided in 4176 different streams with available measurements.

Waste class Lower heating value [MJ/kg] Combustion air [kg/kg] Density [kg/m3] Number of streams [-] Mass [kt]

COCl 24.9 8.3 970 231 70.0

MLCl 3.0 2.6 1065 914 192.7
COBr 31.1 9.8 958 15 9.4
MLBr 3.7 1.8 1300 1 9.1
COF 27.5 9.4 871 21 17.3
MLF 8.1 3.4 929 19 5.0
COP 32.8 11.3 1000 13 2.2
MLP 3.1 1.4 1004 9 1.1
COS 28.7 6.6 1000 31 0.7
MLS 9.5 3.2 1000 38 4.8
CONO 29.3 10.4 982 1425 140.6
MLNO 3.0 2.9 837 1459 371.4

arithmetic mean of the yearly average values and the considered  Regarding the required amount of air for a complete combus-
sample can be found in Table 2. tion, combustible classes necessitate from 5.7 to 11.8 kg of air
Fig. 4 represents the average yearly distribution of the waste per kg of waste, while mother liquors only require between 1.7
properties for the different waste categories. From this figure, the and 3.4 kg. Such figures have indeed been expected because of
following conclusions can be drawn: the higher organic content of waste solvents in comparison to
aqueous residuals, which represent the largest fractions of
 Average heating values and required combustion air of combustible and mother liquor residuals, respectively.
combustible classes are much higher compared to mother liquor  Waste density does not change much from class to class and
ones. More precisely, average heating values of combustible between combustibles and mother liquors as it could be ex-
waste classes are comprised between 23.5 and 33.4 MJ/kg, pected from the densities of common industrial solvents and
whereas the values of mother liquor only assume values ranging water. This is presumably due to the fact that liquid waste re-
from 0.4 to 3.5 MJ/kg. siduals are actually mixtures of different solvents, water, solved

Fig. 5. Total amounts of liquid industrial waste produced in the Swiss chemical sites without treatment facilities, grouped in the Grand Regions (Fig. 1), for the period 2010e2014,
and the corresponding estimated energy content.
 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240 1235

salts, and can also contain variable amounts of reactant and 3.2. Overall waste evolution in Switzerland in period 2010e2014
product species, resulting in similar values for all waste classes.
Also average densities of waste classes do not change signifi- This section presents the time evolution of liquid hazardous
cantly in time, even for the ones presenting the highest differ- waste in Switzerland, which is estimated by combining transport
ences, namely COF, MLF and MLNO classes. data with the available historical performance of incineration fa-
 Concerning lower heating value and required air for a complete cilities. After having processed the available transport notification
waste combustion, in general, mother liquor waste classes data into their corresponding waste classes according to the pro-
present higher variations between average yearly values of the cedure presented in section 2, the primary amounts of each waste
same class. This can be straightforwardly explained with the class produced in chemical sites without own incineration facilities
much broader range for water content that can be assumed by has been determined. Next, the energy content of the produced
waste residuals belonging to mother liquor waste classes residues has been estimated by multiplying their amounts by the
(xwater  15%) compared to the possible values for combustible corresponding representative heating values shown in Table 2.
ones (xwater < 15%). Therefore, the total mass and the energy content of hazardous
waste produced in the seven Swiss regions for the period
The most abundant waste categories present average properties 2010e2014 are shown in Fig. 5. From this, it is observed that both
that are stable along time. Only the waste classes MLF, MLS, MLP amounts and energy content of the waste are rather constant in
and COS, composed of less different waste streams, have small time for every region, with the only exception of Zentralschweiz
fluctuations in yearly amounts in their streams, which lead to which has considerably higher quantities in the last three years of
considerable changes in average physicochemical properties, such the investigated period, although definitely not important
as heating value and required combustion air. Nevertheless, such compared to the total Swiss production. The major amount of waste
variations can be considered as negligible for the purpose of this is generated in the two regions with the highest presence of the
study, since the produced amounts of such waste are very low chemical and pharmaceutical sector in Switzerland, namely Re
gion
compared to the other classes. Hence, the mean properties sum-
manique and Nordwestschweiz (Swiss Federal Statistical Office,
le
marized in Table 2 can be considered as a reasonable approxima- 2017a). In these two regions, many incineration and cement
tion for the estimation of the total energy potential of the liquid plants are also located (Fig. 1). Thus, most of the generated residuals
industrial waste produced in Switzerland, as well as for future could potentially be treated close to its source sites, decreasing both
studies regarding the optimization of industrial waste-to-energy environmental load and safety concerns arising from trans-
networks. portation of hazardous waste. Nevertheless, this would require a

Fig. 6. Total amounts of liquid industrial waste incinerated in the Swiss incineration and cement plants (Fig. 1) in period 2010e2014, divided in the defined classes (Table 1), and the
corresponding estimated energy content.
1236 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240

systematic planning of both shipments and incineration facilities, different regions is quite similar to the mass distribution.
which has still to be applied in practice as the system is currently The evolution of different waste classes incinerated in the fa-
managed based on short-term local incineration and temporary cilities in Fig. 1 during the period 2010e2014, both in terms of
storage capacities. Likewise, the energy content share of the amounts and energy content, is presented in Fig. 6. Similarly to the

Fig. 7. Estimation of exploited and unexploited energetic potential of liquid industrial waste in Switzerland, on an average yearly basis. The produced waste includes both external
sites and integrated complexes with incineration capability. The data used to draw these graphs can be found in appendix Appendix A (Table A.3).

Fig. 8. Comparison between incinerated and energetically unexploited liquid industrial waste per year in Switzerland, stemming from external chemical sites only. Both integrated
chemical sites and cement plants are considered for the incineration of external waste. Each step of the graph represents a single waste class, with characteristic heating value and
average yearly production, whereas the area beneath the line equals its total energy content. More detailed data about the waste amounts depicted in this figure is available in
appendix Appendix A (Table A.4), whereas the lower heating values for each class are provided in Table 2.
 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240 1237

residues generated in external sites (i.e., without any treatment corresponding heating value, resulting in the graph shown in Fig. 8.
facility), there are only marginal differences during the investigated As the energy content of the residuals can be obtained by multi-
period concerning quantities and composition of the incinerated plying these two values, the area beneath every step represents the
waste. Precisely, the total waste amount slightly increased from total energy of one specific waste class.
2010 to 2013, before declining a little in 2014. The classes COCl, From Fig. 8, it can be observed that, in Switzerland, about 66% of
CONO, MLCl and MLNO represent the largest fraction of the re- the 111,288 t of combustible waste produced yearly in external sites
siduals, but the share of high-calorific CONO and COCl residues is incinerated in both integrated chemical sites and cement plants.
increases when considering the overall energy content instead of Although incineration of combustible waste is undoubtedly the
using the material basis. most interesting treatment for energy recovery purposes, the
Such facts highlight the importance of both quantity and heat- remaining part, equivalent to about 38,240 t of combustible waste,
ing value for planning industrial waste incineration. While amounts has still to be exploited, for a corresponding energy amount of
of the different residuals strongly affect transportation and storage, 1038 TJ.
their heating values determine which residuals can be incinerated In contrast, only a very small fraction of mother liquor waste
simultaneously and eventual requirements of additional fuels to (i.e., with a heating value lower than 20 MJ/kg) is currently burned.
ensure a proper thermal treatment. Since mother liquor residuals cannot ensure the flame stability
required for a smooth and continuous combustion because of low
3.3. Estimation of unexploited energetic potential heating value and high water content, it is not surprising that
incineration is not preferred when other options are viable, such as
The overall energy potential of industrial waste incineration in waste water treatment, and especially in the case of companies
Switzerland is assessed based on the results presented in the pre- without in-site incineration facilities. Nonetheless, the simulta-
vious sections. neous incineration of combustible and mother liquor waste is an
Since waste composition and amounts are quite stable con- effective way to i) diminish the requirement of auxiliary fuels to
cerning both generation and incineration, it has been decided to sustain the combustion of mother liquors by taking advantage of
estimate the energy potential of the liquid hazardous waste pro- the higher heating value of combustibles, and ii) control the oven
duced in Switzerland based on average figures from the studied temperature by substituting water with mother liquors. Further-
five-year period. Specifically, the total produced waste of each more, the combination of low heating value of mother liquors and
category has been divided in the three following fractions: the huge volumes of generated residues still sums up to a decent
unexploited potential of 1038 TJ, which is in practice equivalent to
1. Incinerated waste stemming from internal production in inte- the one of combustible waste. Thus, from an energetic perspective,
grated chemical complexes with incineration facilities. In this alternative treatment processes for mother liquors are only
case, data has been directly provided by these sites. meaningful if the energy benefits of material recycle are higher
2. Waste produced in external sites without treatment capability than the ones provided by incineration. Anyway, material recovery
and incinerated in Switzerland (Fig. 1). This has been calculated from both ashes and fumes resulting from the combustion process
with information from both incineration sites and cement is also possible for some chemical species. For instance, the
plants (Cemsuisse, 2015). Kubierschky process can be implemented for producing elemen-
3. Potentially incinerable waste that has not been treated in the tary bromine from the dissolved Br present in the waters used to
considered incineration facilities (Fig. 1). As the available infor- wash the incineration fumes (Mills et al., 2000).
mation is limited to exports, it is assumed that the remaining
part, obtained by subtracting the incinerated waste from the 4. Conclusion
total production, has been treated differently or burned in other
small industrial furnaces. In the latter case, it is about a minor This work proposes a new waste classification system based on
share of calorific solvents with relatively high-purity that can be lower heating value and on both water and pollutant content of the
used as substitution fuel in other industrial furnaces (e.g. in residues. For every waste category, average energy content and
place of heating oil), as mentioned in the introduction. Since the other physicochemical properties can be obtained from measure-
available data does not include recycled or regenerated waste, it ments provided by the enterprises managing the Swiss facilities for
is reasonable to consider the complete amount of this fraction as industrial waste incineration. Such information has also been used
potentially incinerable for the purpose of this study. to establish a link between the proposed classification and the
system used for the notification of hazardous waste shipments in
Mass and energy content of all these fractions are then graphi- many European countries, thus allowing for the first time to esti-
cally compared for all waste classes in Fig. 7. As previously dis- mate the energy content of the residuals produced in a specific area,
cussed, classes COCl, CONO, MLCl and MLNO represent the majority as well as to assess its unexploited potential. This significantly ex-
of the liquid hazardous waste produced in Switzerland. Addition- pands the spatial dimension of previous studies concerning the
ally, the same waste classes also present the highest unexploited characterization of industrial waste, which, unlike works related to
energetic potential. In fact, the energy content of unexploited re- different fractions of municipal residues, have been limited to much
siduals from these classes sums up to 2.2 PJ per year in average, more local perspectives. Therefore, the comparison with smaller
while the total unexploited potential is 2.5 PJ. Both values are realities might be misleading. Thanks to the common notification
considerable compared to the energy content of the incinerated system and the generality of the new waste classification, the
waste of 4.0 PJ per year, implying that about 38% of the total liquid proposed methodology can be applied to a broad set European
industrial waste produced in Switzerland is not used for energy countries for estimating several properties and the energy recovery
recovery purposes. potential from existing and constantly updated information
Next, the energy content of the residuals produced in sites without further measurements. Given the sparse structure of the
without incineration facilities are further analysed. These residuals Swiss process and pharmaceutical industry and its extreme variety
compose almost the total part of non-incinerated primary waste. in terms of waste residues, the presented case study shows that
Therefore, the mass of each class has been plotted against its such methodology can be used also for the most complex systems.
1238 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240

The obtained results allow waste incinerators to better exploit the AG, Infrapark Baselland AG, Lonza AG, Valorec Services AG and the
calorific power of the treated waste, decreasing the consumption of Swiss Federal Office for the Environment for supplying data and
primary energy sources and thus contributing to a cleaner chemical their support in technical questions. Sincere gratitude is also
production. The generated knowledge is also a crucial first step in dedicated to Marta Roca Puigro s and Aline Gazzola for their
the development of optimization tools for the design and operation precious help concerning data processing. This research project is
of sustainable industrial waste incineration networks, aimed at part of the National Research Programme ”Energy Turnaround”
maximizing the energy efficiency while fulfilling technical and (NRP 70) of the Swiss National Science Foundation (SNSF). Further
regulatory constraints. Such tools can be used to assess the po- information on the National Research Programme can be found at
tential benefits of a more systematic management of industrial www.nrp70.ch.
waste-to-energy networks, as well as to investigate the implica-
tions of different future conditions concerning both waste
composition and amounts.
Appendix A. Supporting information

Acknowledgements

The authors thank the companies Cimo Compagnie industrielle

de Monthey SA, Dottikon Exclusive Synthesis AG, Holcim Schweiz
Table A.3
Average yearly amounts of incinerated liquid industrial waste in Switzerland, stemming from both integrated sites with incineration facilities and external companies, and the
mass of potentially incinerable residuals, together with their corresponding estimated energy content.

Waste class Internal incinerated External incinerated Potentially incinerable

Mass [t] Energy [TJ] Mass [t] Energy [TJ] Mass [t] Energy [TJ]

COCl 15,118 376 1239 31 19,221 478

MLCl 32,744 181 3393 19 55,230 305
COBr 1857 58 558 17 2100 65
MLBr 2115 8 42 0 0 0
COF 1826 50 38 1 734 20
MLF 1398 11 39 0 5337 43
COP 607 17 162 5 1506 43
MLP 237 1 3 0 492 2
COS 245 8 179 6 631 21
MLS 1074 10 19 0 1839 17
CONO 36,099 1057 70,872 2076 14,047 411
MLNO 76,765 436 5840 33 118,236 671

Total 170,085 2213 82,384 2188 219,373 2076

Table A.4
Average yearly amounts of produced and incinerated liquid industrial waste in Switzerland, stemming from external companies without incineration facilities. 53,051 t of
waste belonging to the CONO class have been incinerated in cement plants, corresponding to an energy amount of 1554 TJ.

Waste class Produced waste Incinerated waste Potentially incinerable

Mass [t] Energy [TJ] Mass [t] Energy [TJ] Mass [t] Energy [TJ]

COCl 20,460 509 1239 31 19,221 478

MLCl 58,623 323 3393 19 55,230 305
COBr 2658 83 558 17 2100 65
MLBr 42 0 42 0 0 0
COF 773 21 38 1 734 20
MLF 5375 44 39 0 5337 43
COP 1668 48 162 5 1506 43
MLP 496 2 3 0 492 2
COS 810 27 179 6 631 21
MLS 1858 18 19 0 1839 17
CONO 84,919 2487 70,872 2076 14,047 411
MLNO 124,076 705 5840 33 118,236 671

Total 301,758 4267 82,384 2188 219,373 2076

 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240 1239

Table A.5 Table A.5 (continued )

The Swiss waste classes distribution for each notification codes, as used in this study.
(European Parliament and Council, 2006; DATEC, 2005) Code Class Fraction [%]

MLNO 70
Code Class Fraction [%]
MLS 1
60101 MLS 100 70403 COCl 12
60102 MLCl 100 MLCl 43
60103 MLF 100 MLNO 44
60104 MLP 100 70404 COCl 6
60105 MLNO 100 CONO 43
60106 MLNO 100 COS 0
60201 MLNO 100 MLCl 8
60203 MLNO 100 MLNO 41
60204 MLNO 100 MLS 1
60205 MLNO 100 70407 COCl 3
60704 MLCl 100 CONO 52
61301 MLNO 100 MLCl 23
70101 COCl 9 MLNO 22
CONO 1 70408 CONO 54
MLCl 26 MLCl 7
MLF 3 MLNO 39
MLNO 56 70501 COCl 1
MLS 5 COF 0
70103 COBr 2 CONO 10
COCl 2 COS 1
COF 2 MLCl 15
CONO 7 MLF 8
MLCl 35 MLNO 57
MLNO 49 MLP 2
MLS 3 MLS 4
70104 COCl 2 70503 COCl 8
CONO 40 CONO 35
COS 1 COS 0
MLCl 12 MLCl 23
MLNO 41 MLNO 34
MLP 2 70504 COCl 2
MLS 2 COF 0
70107 COCl 2 CONO 36
CONO 43 COS 0
COP 4 MLCl 14
MLCl 22 MLF 1
MLNO 29 MLNO 48
MLS 0 70507 COCl 4
70108 COCl 14 MLCl 89
CONO 29 MLNO 7
MLCl 1 70508 COCl 1
MLNO 56 COF 31
70111 MLCl 100 CONO 47
70201 COCl 100 MLCl 11
70203 COCl 50 MLNO 10
MLCl 50 70601 MLCl 100
70204 COCl 13 70603 MLNO 100
CONO 37 70604 MLNO 100
MLCl 50 70607 COCl 2
70207 COCl 2 CONO 43
CONO 43 COP 4
COP 4 MLCl 22
MLCl 22 MLNO 29
MLNO 29 MLS 0
MLS 0 70608 COCl 14
70208 COCl 14 CONO 29
CONO 29 MLCl 1
MLCl 1 MLNO 56
MLNO 56 70701 CONO 23
70301 MLCl 100 MLCl 7
70303 MLCl 100 MLNO 70
70304 CONO 62 70703 COBr 30
MLNO 38 COCl 47
70307 COCl 2 CONO 9
CONO 43 MLCl 14
COP 4 70704 COCl 15
MLCl 22 CONO 36
MLNO 29 COP 3
MLS 0 MLCl 6
70308 MLNO 100 MLF 9
70401 CONO 1 MLNO 31
COS 1 70707 MLCl 100
MLCl 27 70708 COCl 76
(continued on next page)
1240 V. Bolis et al. / Journal of Cleaner Production 183 (2018) 1228e1240

Table A.5 (continued ) van Es, D., Worrellpp, E., 2012. Chapter 8-energy end use: industry. In: Global
Energy Assessment - toward a Sustainable Future, pp. 3962e3969.
Code Class Fraction [%] Basel Convention on the Control of Transboundary Movements of Hazardous
Wastes and Their Disposal, 22.3.1989.
MLNO 24
Burnley, S., 2007. The use of chemical composition data in waste management
80313 MLCl 100
planning - a case study. Waste Manag. 27 (3), 327e336.
110106 MLCl 100
n-García, E., Papadokonstantakis, S., Hungerbühler, K., 2014. Multi-objective
Capo
110113 MLCl 100 optimization of industrial waste management in chemical sites coupled with
120107 MLCl 100 heat integration issues. Comput. Chem. Eng. 62, 21e36.
120301 MLNO 100 Cemsuisse, 2015. Chiffres-clefs 2014.
130104 CONO 100 Chakraborty, A., Colberg, R.D., Linninger, A.A., 2003. Plant-wide waste management.
130205 CONO 66 3. Long-term operation and investment planning under uncertainty. Ind. Eng.
MLNO 34 Chem. Res. 42 (20), 4772e4788.
130206 COS 91 European Parliament and Council, 2006. Regulation (EC) No 1013/2006 of the Eu-
MLCl 9 ropean Parliament and of the Council of 14 June 2006 on Shipments of Waste.
130308 CONO 100 Official Journal of the European Union.
130310 CONO 50 European Parliament and Council, 2013. Decision No 1386/2013/EU of the European
MLCl 50 Parliament and of the Council of 20 November 2013 on a General Union
Environment Action Programme to 2020 ‘Living Well, within the Limits of Our
140601 MLCl 50
Planet’. Official Journal of the European Union L 190, p. 1.
MLF 50
“Expert Panel of the project Optimisation of industrial waste-to-energy and
140603 CONO 1 resource recovery systems”, 2015-2017. Consisting of CIMO SA, Dottikon
MLNO 99 Exclusive Synthesis AG, Holcim Schweiz AG, Infrapark Baselland AG, Lonza AG,
150110 MLCl 100 Valorec Services AG, the Swiss Federal Office for the Environment and the Swiss
160303 MLCl 100 Federal Office of Energy.
160305 CONO 23 Ghimpusan, M., Nechifor, G., Nechifor, A.-C., Dima, S.-O., Passeri, P., 2017. Case
MLCl 77 studies on the physical-chemical parameters' variation during three different
160504 CONO 31 purification approaches destined to treat wastewaters from food industry.
MLF 38 J. Environ. Manag. 203 (Part 2), 811e816.
MLNO 31 Hla, S.S., Roberts, D., 2015. Characterisation of chemical composition and energy
160506 MLCl 100 content of green waste and municipal solid waste from Greater Brisbane,
160507 MLCl 100 Australia. Waste Manag. 41 (Suppl. C), 12e19.
Il Consiglio federale svizzero, 2005. Ordinanza sul traffico di rifiuti, RS 814.610
160508 COCl 2
(Swiss regulation, status as of 01.07.2016).
CONO 60
Il Consiglio federale svizzero, 2015. Ordinanza sulla prevenzione e lo smaltimento
MLCl 28
dei rifiuti, RS 814.600 (Swiss regulation, status as of 03.10.2017).
MLNO 11 Il Dipartimento federale dell’ambiente, dei trasporti, dell’energia e delle comuni-
160598 MLNO 100 cazioni (DATEC), 2005. Ordinanza del DATEC sulle liste per il traffico di rifiuti,
160708 MLNO 100 RS 814.610.1 (Swiss regulation, status as of 01.04.2017).
160709 MLNO 100 Levis, J.W., Barlaz, M.A., DeCarolis, J.F., Ranjithan, S.R., 2014. Systematic exploration
161001 MLCl 100 of efficient strategies to manage solid waste in U.S. Municipalities: perspectives
161003 MLCl 100 from the Solid Waste Optimization Life-cycle Framework (SWOLF). Environ. Sci.
200129 CONO 100 Technol. 48 (7), 3625e3631.
Mills, J.F., Frim, R., Ukeles, S.D., Yoffe, D., 2000. Bromine. In: Ullmann's Encyclopedia
of Industrial Chemistry.
OECD, 2013. Environment at a Glance 2013.
Papadokonstantakis, S., Hungerbühler, K., Sennhauser, M., 2013. The Success of
Switzerland's Chemicals and Pharmaceuticals Industries. Chemical Engineering
Progress.
Seyler, C., Hofstetter, T.B., Hungerbhler, K., 11 2005. Life cycle inventory for thermal
treatment of waste solvent from chemical industry: a multi-input allocation
model. J. Clean. Prod. 13, 1211e1224.
Seyler, C., Capello, C., Hellweg, S., Bruder, C., Bayne, D., Huwiler, A., Hungerbühler, K.,
2006. Waste-Solvent management as an element of Green chemistry: a
comprehensive study on the Swiss chemical industry. Ind. Eng. Chem. Res. 45
(22), 7700e7709.
Slack, R.J., Bonin, M., Gronow, J.R., Van Santen, A., Voulvoulis, N., 2007. Household
hazardous waste data for the UK by direct sampling. Environ. Sci. Technol. 41
(7), 2566e2571.
Swiss Federal Office of Energy, 2014. Energieverbrauch in der Industrie und im
Dienstleistungssektor (Resultate 2013).
Swiss Federal Office of Energy, 2015a. Analyse des schweizerischen Energiever-
brauchs 2000-2014 nach Verwendungszwecken.
Swiss Federal Office of Energy, 2015b. Schweizerische Gesamtenergiestatistik 2014.
Fig. A.9. Comparison between external and internal waste, generated per year in Swiss Federal Office of Energy, 2017. Energy Strategy 2050 after the Popular Vote.
Switzerland. Each step of the graph represents a single waste class, with characteristic Swiss Federal Statistical Office, 1999. Die Schweiz im europ€ aischen Regionalsystem
heating value and average yearly production, whereas the area beneath the line equals (Press release).
Swiss Federal Statistical Office, 2017a. Emplois (en e
quivalents plein temps) dans la
its total energy content. The lower heating values for each class are provided in Table 2.
branche chimique et pharmaceutique, en 2013. https://www.bfs.admin.ch/bfs/
en/home/statistics/regional-statistics/atlases/interactive-statistical-atlas-
References switzerland.html.
Swiss Federal Statistical Office, 2017b. Statistica Strutturale Delle Imprese. https://
Abaecherli, M.L., Capo
n-García, E., Szijjarto, A., Hungerbühler, K., 2017. Optimized www.bfs.admin.ch/bfs/it/home/statistiche/industria-servizi/rilevazioni/statent.
energy use through systematic short-term management of industrial waste html.
incineration. Comput. Chem. Eng. 104, 241e258. Verband der Betreiber Schweizerischer Abfallverwertungsanlagen (VBSA), 2017.
Allesch, A., Brunner, P.H., 2017. Material flow analysis as a tool to improve waste Sonderabfallverbrennung. http://vbsa.ch/anlagegruppen/sonderabfall/.
management systems: the case of Austria. Environ. Sci. Technol. 51 (1), Wassick, J.M., 2009. Enterprise-wide optimization in an integrated chemical com-
540e551. plex. Comput. Chem. Eng. 33 (12), 1950e1963.
Banerjee, R., Cong, Y., Gielen, D., Jannuzzi, G., Mare
chal, F., McKane, A.T., Rosen, M.A.,