Vous êtes sur la page 1sur 11

Journal of Cleaner Production 12 (2004) 685695 www.elsevier.

com/locate/jclepro

End-of-pipe versus process-integrated water conservation solutions A comparison of planning, implementation and operating phases
Karl A. Zotter
Institute for Innovation and Environmental Management, Karl-Franzens-University of Graz, Universita tsstrasse 15, A-8010 Graz, Austria Received 13 December 2002; accepted 2 July 2003

Abstract The technical and economic advantages and disadvantages of numerous approaches for fullling environmental protection aims are discussed in the literature. This article compares water conservation solutions that are applied to the wastewater from the production processes (end-of-pipe solutions) and solutions that intervene in the production processes (process-integrated solutions). Three arguments that are commonly used in the literature to compare the technical and economic advantage of end-of-pipe versus process-integrated solutions are unilateral decoupling, technological maturity and legal security. A unilateral decoupling exists when the production system inuences the operating status of the end-of-pipe solution but not vice versa. This prevents the endof-pipe solution having an unwanted effect on the production system. As markets have developed for end-of-pipe solutions, specialized companies have continuously developed them. The same is not true for process-integrated solutions. Therefore, compared to process-integrated solutions, end-of-pipe solutions can be regarded as old, mature technologies. The authorities being aware of the state-of-the art of water conservation technology favor end-of-pipe solutions. Using end-of-pipe solutions therefore, brings a certain level of legal security. These arguments and their economic importance were used as the basis for a survey on water conservation in Austrian industry. The survey compared the estimated risks of planned water conservation solutions and problems, which actually occurred with end-of-pipe and process-integrated solutions. The main result of the survey is that, for the most part, the expected economic advantages of end-of-pipe solutions as compared to integrated solutions could not be proven. Explanations are given for these deviations between the literature and the survey results. 2003 Elsevier Ltd. All rights reserved.
Keywords: Water conservation; End-of-pipe solutions; Process-integrated solutions

1. Introduction Nowadays, environmental conservation is an indispensable part of the goals of most Austrian companies. The reason for this is usually external pressure (laws, the demands of customers, etc.), although, many companies are now realizing that environmental performance improvement can lead to an increase in prots and a reduction in costs [2,4]. Independent of the reasons for the integration of environmental goals into the company philosophy, each rm must decide which approaches to use to achieve

these goals. The choice of approach is of special importance for production companies. For these companies in particular, the decision to implement a concrete problemsolving concept requires a large investment. The wrong decision can therefore have considerable negative consequences. The technical and economic advantages and disadvantages of numerous approaches for fullling environmental protection aims are discussed in literature (e.g. [9,12,19]). The approaches are often divided into two categories: end-of-pipe solutions and process-integrated solutions [4,12,13].

Tel.: +43-316-380-3233; fax: +43-316-380-9585. E-mail address: karl.zotter@kfunigraz.ac.at (K.A. Zotter).

0959-6526/$ - see front matter 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0959-6526(03)00115-X

686

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

2. The denition of end-of-pipe versus processintegrated approaches End-of-pipe solutions are understood to be control technologies, which are more or less stuck onto the end of the production system [7,22]. From a systems point of view, end-of-pipe solutions are clearly (physically) separate from production processes. The purpose of endof-pipe solutions is to modify the residual products of production processes (Fig. 1, water stream A) so that they are less damaging to the natural environment than untreated residual products [13] (Fig. 1, water stream C). These outputs are usually unwanted from an economic point of view because the cost of disposing of the resulting wastewater sludge and wastewater has a negative inuence on company success. Prot is made only in exceptional cases. Furthermore, end-of-pipe solutions require energy and raw materials, which also cost money. End-of-pipe solutions should also be viewed critically from an ecological point of view. Although the outputs are changed qualitatively so that they can be disposed of according to the prevailing law, end-of-pipe solutions do require additional energy and materials. This increases the companys total consumption of materials and energy. Process-integrated solutions should be differentiated from end-of-pipe solutions [3,7]. Process-integrated solutions include recycling solutions and process moderniz-

ation and process improvement measures. The aim of process-integrated solutions is to prevent wastewater being produced in the production process. This usually covers not only the reduction of the wastewater produced but also the reduction in the amount of water used in the company [23]. As with end-of-pipe solutions, recycling solutions use plants or parts of plants, which are clearly and physically separated from production processes. In contrast to endof-pipe solutions, however, this approach aims to create desired goods in the form of impulse factors for production processes. This approach should therefore be seen as an aid to the production system (Fig. 1). After the conditioning of residual products, it is often the case that part of the output from the treatment system is released into the natural environment and the remaining part is re-used in production processes. Here, the categorization of the treatment system as either an end-ofpipe or recycling solution is not possible. This problem of denitions is dealt with pragmatically in the survey carried out in Austrian industry. Whether a solution is seen as an end-of-pipe or recycling solution is dened in this survey by the primary objective of the solution. With the above-mentioned end-of-pipe and recycling solutions, the solutions for improving environmental performance were clearly and physically distinct from the production processes. The term process modernization measures [6] is

Fig. 1.

Integration-levels of water-protection-approaches.

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

687

taken to mean improvements to existing processes and the control of these processes and also the replacement of existing processes with more environmentally friendly processes. With process modernization measures, the separation of production system and the improvement of environmental performance is not possible. The production process and solution to improve environmental performance form one physical, process and control unit. A distinguishing characteristic of process modernization measures is therefore the absence of explicitly distinct system elements with the purpose of improving environmental performance. Therefore, every production process, independent of its actual usefulness in the improvement of environmental performance, could be called as a process-integrated environmental conservation solution. This approach is not suitable for use in a survey on environmental conservation. The distinguishing characteristic of process modernization measures with integrated solutions was therefore dened as changes to processes (technologies), which have a positive effect on the companys environmental performance. 3. Arguments that are used to substantiate the advantages of end-of-pipe solutions This section explains in detail the properties of end-ofpipe solutions that are said to substantiate an economic advantage when compared to integrated solutions. When discussing the advantages of end-of-pipe solutions in literature, the following arguments are used: 1. End-of-pipe solutions do not (or hardly) affect production because they are implemented after the production process and can be operated independently of the production process [12]; 2. End-of-pipe technologies are mature technologies that have been used often in the past and have been continuously improved, and 3. End-of-pipe solutions have a high degree of legal security because they are suggested and recognized by the ofcial authorities [1,5]. From these basic arguments, it is assumed that endof-pipe solutions cause fewer and less serious problems in the planning, implementation and operating phases than process-integrated solutions. This gives end-of-pipe solutions economic advantages. Using an empirical study of water conservation measures in Austrian industry, it is investigated whether these arguments found in literature and the resulting economic advantages can be proven. 3.1. End-of-pipe solutions can be seen as decoupled from the production systems The physical separation of the end-of-pipe solution from the actual production process is one of the dis-

tinguishing characteristics of such solutions [12,22]. A further specic characteristic is that it is assumed that the output of the end-of-pipe solution is not used in production or consumption processes but released in its entirety into the natural environment. There is, therefore, merely a material and information technological relationship between the production system and the end-of-pipe solution: The unwanted output of the production system is transferred via an interface and becomes an input for the end-of-pipe solution. Information for the smooth operation of the end-of-pipe solution is communicated similarly. The separation of the production process from the end-of-pipe solution could be considered a unilateral decoupling. This means that the production system inuences the operating conditions of the end-of-pipe solution, but not vice versa. If the conditions for a unilateral decoupling are fullled (as is often assumed in the literature), this results in technical and economic advantages in the planning, implementation and operating phases of end-of-pipe solutions [11,12,15]. 3.1.1. Advantages of end-of-pipe solutions in the planning phase If the conditions for unilateral decoupling are met, technical and economic advantages of end-of-pipe solutions compared to process-integrated solutions can be seen as early as in the planning stage [13]. These advantages arise because the conditions allow simpler construction and general solutions can be implemented for effective and efcient system design. 3.1.1.1. General conditions which determine the construction of water conservation solutions Due to the unilateral decoupling of the end-of-pipe solution from the production system, the end-of-pipe solution can be regarded as self-sufcient in its operation. Decisive for the operation of the end-of-pipe solution are: as an input, data on the residual products from the production system and, as an output, the permitted emissions into the natural environment. For the constructive design of end-ofpipe solutions, the following time-dependent variables are decisive: Quality; Quantity of the end-of-pipe-input, and Qualitative and quantitative emission limits, which must be adhered to. In contrast to end-of-pipe solutions, process-integrated solutions usually have several input and output interfaces with different production processes. These interface data must be coordinated. Decisive variables for the constructive design of process-integrated solutions are: Quality;

688

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

Quantity of all input ows, and Quality and quantity of all required output ows [3]. It must be taken into consideration that all materials/energy ows depend on the current operating point of the systems involved. Furthermore, the effects on production processes, which are not directly connected with the process-integrated environmental performance improvement solution, may also need to be considered. The central condition of these coordination measures is that the desired production output should not be negatively affected, either qualitatively or quantitatively. In summary, it can be said that in most cases, the number of processes involved which require coordinating is higher with process, integrated solutions than with end-of-pipe solutions. The amount of data required for the design of the water conservation solution and the complexity of these data is also greater with processintegrated solutions than with end-of-pipe solutions. Therefore, the planning of process-integrated solutions is often more complex compared to end-of-pipe solutions [3,14,15]. The planning risk of not fullling cost objectives and performance targets is also correspondingly higher with process-integrated solutions. 3.1.1.2. Principles of construction for the design of solutions When designing solutions, it seems useful to adhere to the following principles [10]: Minimization of complexity; Lowest prejudgment, and Modular design of systems [4,10]. The principle of the minimization of complexity states that the system should be kept as simple as possible so that systems are comprehensible and controllable. Additional processes and process relationships in production systems lead to increasing complexity. End-ofpipe solutions fulll the principle of minimization of complexity to a large extent due to unilateral decoupling. As end-of-pipe solutions do not affect the production system, the complexity of the production system remains unchanged. Process-integrated measures very often increase the complexity of the production systems. With processintegrated solutions, additional or new processes are implemented to create new relationships within the production system. It is impossible to make a generalized statement on the effect process-integrated approaches have on the complexity of the production system. They can increase or decrease production system complexity. The principle of lowest prejudgment states in this context that the solution that least dictates the further development of the production system should be chosen. Generally, end-of-pipe solutions only inuence the

development of production systems in that future changes should not exceed the capacity limits of the endof-pipe solution [14]. These limits can be adapted to requirements by setting up additional capacities. The investments should be taken into consideration in the planning of changes to the production system. These systems are therefore prejudged from an economic point of view. Process-integrated solutions have prejudging effects similar to the implementation of new production processes. Process-integrated solutions restrict the scope for later changes to the production system. Prejudging characteristics can therefore be found for both approaches. However, measures can always be applied which reduce (or remove) the prejudging effects from a technical point of view. From an economic point of view, no generalized statement can be made whether the further development of production systems are more restricted by end-of-pipe or by process-integrated solutions. The principle of modular design states that solution systems should be designed so that they can be used in a variety of situations. End-of-pipe solutions usually fulll these requirements very well. The reason for this is that different companies often have similar qualitative and quantitative ows of residual products. These can be treated in technically similar end-of-pipe solutions. This is especially often the case within the same industrial branch [17,18]. End-of-pipe solutions can therefore be standardized relatively well. With relatively few adjustments, they can usually be used in a number of comparable applications. This large number of similar applications forms a relatively large market for end-of-pipe solutions. System providers can cover this demand using only a comparatively small amount of required technological knowledge. Due to the number of similar applications, system providers can build up a wealth of experience in the planning, implementation and operation of end-of-pipe solutions. This experience can then be applied when designing new plants. The construction of new plants therefore involves little that is truly new for the system providers. This suggests that the planning risk for endof-pipe solutions is low [4]. The requirements for modular system design are seldom fullled by process-integrated solutions. These systems have to be adapted to the specic characteristics of each individual production system [14] and there is usually no way of developing modules, which can be used in many different production systems. One exception is treatment technologies, which are used in recycling solutions. The requirement for modular design is therefore rarely fullled by process-integrated solutions. System providers as described above therefore nd it hard to become established. A broad range of technological know-how is required for relatively low market volume.

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

689

This effect was clearly revealed in the survey results. 75.4% of the end-of-pipe solutions were planned and completed by a third party, whereas 69.8% of the process-integrated approaches were planned and completed by the companies themselves. One of most stated reasons for the planning and realization of process-integrated solutions by the company itself was the lack of suitable consultants and providers (77.2% of all the process-integrated solutions carried out by the company itself). A further consequence of the need to design processintegrated solutions to meet the specic requirements of individual production systems is that experience gained from completed projects is not of use (or is only of limited use) for other projects. The novelty of each project is therefore higher and the risk of planning errors must therefore be rated higher than with end-of-pipe solutions. An additional advantage of end-of-pipe solutions, which results from the availability of system providers, is that companies requiring environmental performance improvement solutions can keep the costs for searching for and choosing a suitable system provider low with end-of-pipe solutions. Companies usually know of suitable system providers or else system providers actively approach companies. Usually, a system provider can provide information on reference solutions [4]. For process-integrated solutions, the providers (if there are providers) are often not known to companies. There are usually no reference solutions available. This suggests that the search and evaluation costs of processintegrated solutions are therefore quite high in comparison to end-of-pipe solutions. The search for a provider of a process-integrated solution often fails because no suitable partner can be found [13]. In summary, it can be said that the general conditions determining construction are easier with end-of-pipe solutions and the adherence to principles for effective and efcient system design is possible to a larger degree than is the case with process-integrated solutions. The complexity of the planning and therefore the risk of planning errors is therefore greater with process-integrated solutions than with end-of-pipe solutions. 3.1.2. Advantages in the implementation phase of endof-pipe solutions End-of-pipe solutions offer considerable advantages in the plant implementation phase due to unilateral decoupling. The central argument here is that end-of-pipe solutions can be built separate from production systems and can be utilized without disturbing the production processes. With unilateral decoupling, the production system and end-of-pipe solution need not be coordinated. Process and control intervention in production systems are therefore unnecessary, idle capacity costs. Loss of revenue as a result of planned or unplanned plant shutdown in the implementation phase are therefore kept to

a minimum. If problems occur, the end-of-pipe solution can be temporarily disconnected from the production system and the problem solved without disrupting production. The economic risks with end-of-pipe solutions in the implementation phase are therefore the costs of improvements and possible payment of nes. Process-integrated solutions constantly intervene in the operation of production processes. Process and control coordination of the two systems is necessary. The risk of breakdown of one of the integrated solutions is therefore considerably higher than with end-of-pipe solutions. If the process-integrated solution fails, other production processes can be expected to shutdown. With process-integrated solutions, the possible costs of plant improvement and payment of possible nes must be considered as well as the economic risk of loss of revenue and idle capacity costs [8]. 3.1.3. Advantages in the operating phase of end-ofpipe solutions After the end-of-pipe solution has been implemented and there has been a short test phase, the end-of-pipe solution can begin normal operation. In normal operation, further advantages due to the unilateral decoupling come into play [12]: As there is no intervention in the production system, the existing production program can remain unchanged. The raw material structure remains unchanged. Only additional energy/raw materials are required to operate the end-of-pipe solution. The operating cost structure of the production system therefore remains unchanged. The economic risk of unexpected operating costs is therefore avoided. The only extra costs to take into consideration are the operating costs of the end-of-pipe solution. Processintegrated solutions usually intervene in the raw material structure of the production process. This changes the operating costs. If the process-integrated solution does not meet the required performance targets, this can lead to an unspecied increase in additional operating costs. The cost risk for the normal operation of process-integrated solutions is therefore correspondingly higher. If the end-of-pipe solution fails, this does not inevitably lead to a shutdown in production. The residual products can be temporarily released directly into the natural environment, bypassing the end-of-pipe solution. This exceeds emission limits temporarily but is often legally allowed with the appropriate ofcial permission. The economic risks of the normal operation of end-of-pipe solutions are therefore repair costs and possible payment of nes. If a process-integrated solution fails, loss of revenue due to plant shutdown has to be considered as well as repair costs and payment of nes.

690

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

Reduction of unit cost as a consequence of past learning processes in production is still possible when using end-of-pipe solutions. The efcient use of the end-of-pipe solution must be learned. This suggests that the qualications required of the employees working in production must not be changed. In contrast, the implementation of process-integrated solutions may require the employees to have new qualications. This must be achieved by offering training courses, making redundancies and hiring new employees [21]. 3.2. End-of-pipe solution are mature technologies As explained in Section 3.1.1, there are relatively large markets for end-of-pipe solutions. These can be served by system providers. These providers have an economic interest in the further development of the applied technologies. The R&D investments in end-ofpipe solutions are therefore relatively high. As end-ofpipe solutions are plants, which fulll the communitys aim of environmental conservation and represent nancial burdens for the company, the government sponsors them. The further development of end-of-pipe solutions is therefore supported nancially not only by investments from companies but also from the government. It was therefore possible to further develop and distribute end-of-pipe technologies rapidly. For this reason, end-of-pipe technologies can be regarded as a basic technology, which is mastered by specialized providers. It can therefore be assumed that the applied technology is often tested and improved by specialist providers in the planning and design stages and that it is therefore controllable. This suggests that the performance risks in planning and implementing end-of-pipe solutions can therefore be considered low [4]. A characteristic of process-integrated solutions is that they have to be designed to suit the individual production system. Each production system therefore requires an individual process-integrated solution. This means that each project has a relatively high degree of novelty (compared to end-of-pipe solutions). Experience gained from past projects is only of limited usefulness. The risk of not fullling the cost and performance targets must therefore be estimated higher for the planning and realization of integrated solutions [20]. The individual nature of process-integrated solutions also means that a provider with a standard product, which could be used by many companies, has little chance of establishing itself. Due to the lack of providers, the company itself must often develop the process-integrated solution. In such cases, knowledge about a specic process-integrated solution is gained during the development of the solution. However, this knowhow is usually not exploited (internally or externally)

after the project has been nished. A continuous technological development can therefore not occur. With process-integrated solutions, as compared to end-of-pipe solutions, there are no clear boundaries between components, which serve production, and components, which serve environmental conservation (see Section 3.1.1). Financial support for the investment in environmental conservation is therefore often not possible. These circumstances also make it difcult for providers of process-integrated solutions to establish themselves. In summary, it can be concluded that, in the past, an end-of-pipe solution could be further developed to make standard solutions for numerous applications. Due their company-specic requirements, process-integrated solutions can never be used as standard products which are repeatedly applied in the same way. A continuous learning process cannot occur with integrated solutions. The main reason for this is that the technological knowledge gained from completed systems does not lie with the system provider but exists as know-how potential within the producing company (see Section 3.1.1). As this special potential is only rarely built on further, no continuous build-up of technological know-how occurs. This suggests that the performance risk in the planning and implementation of process-integrated solutions must therefore be rated correspondingly higher. 3.3. End-of-pipe solutions have a high level of legal security Environmental performance improvement requirements are dened by working groups based on existing technology. The working groups gather information about available environmental performance improvement technologies from users and providers in this eld. The basis for dening environmental performance improvement requirements can therefore only be technology, which is widely applicable and not individual solutions for company-specic problems [9]. These requirements are best met by end-of-pipe solutions (see Section 3.1.1) and end-of-pipe solutions are therefore a considerable inuencing factor in the establishment of environmental performance improvement requirements. Furthermore, in some countries technological information is transferred to companies via consultation with licensing authorities. As most of the information gathered concerns completed end-of-pipe solutions and the authorities cannot know about every specialized solution for every eld of application, most of the consultations will focus on available end-of-pipe solutions. By consulting with the ofcial authorities and using the available legal information, companies obtain a certain degree of legal security (by using ofcially recognized technology) [4,16]. If companies follow the

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

691

advice given by the authorities, it can be assumed that no considerable legal problems will occur in normal operation. As the effects of integrated solutions are not known (or only known to a lesser degree), it can be assumed that legal problems are more likely to occur in the implementation and operating phases.

4. Results of the survey This section compares the statements made in the literature with the results of the above-mentioned survey. The advantages given for end-of-pipe solutions in the literature are compared to the results of a survey of Austrian industry. The survey covered completed and planned water conservation solutions in 87 Austrian companies. All the surveyed companies had experience with process-integrated measures; 69 companies had installed end-of-pipe solutions in the past. One hundred and twenty-two (48.6%) of the projects were processintegrated measures and 60 (23.9%) solutions were recycling measures. Before sending out the standardized survey, the person responsible for the project was contacted by telephone. This ensured that the person who answered the survey was responsible for the planning, realization and operation of the environmental conservation measures and therefore ensured sufcient expert knowledge. Comparisons between end-of-pipe and integrated solutions were only analyzed when these came from companies with experience and expectations of both solutions. The survey is divided into three areas: The subjective judgment of end-of-pipe solutions and process-integrated solutions as regards the technical, economic and legal risks; The identication of problems in the implementation phase of environmental performance improvement solutions, and The identication of problems in normal operation. 4.1. Estimated risks in the application of solutions to improve environmental performance At the beginning of the survey, an analysis of the estimated risks in the planning and implementation of environmental performance improvement solutions is given in order to discover whether the view of the risk found in the literature corresponds with the view of those working in industry. Table 1 shows the results of this questioning. The contents of the table is explained in subsequent sections. In summary, it can be reported that, as regards the judgment of the risk of water performance improvement projects, the statements given by the people surveyed were more or less identical to the statements made in

the literature. The increase in complexity caused by the application of process-integrated solutions is given as the main problem (Question 1) [13]. The advantages of endof-pipe solutions given by the people surveyed were the same as those given in the literature. Minor (not signicant) advantages were reported for end-of-pipe solutions in answer to the questions about possible plant shutdown (Question 5) and quantitative reductions (Question 4) in the production output. The people surveyed estimated the risk of breakdown of the end-of-pipe solution during operation as being lower than the same risk with process-integrated solutions. Decreased production output is also feared less with endof-pipe solutions than with process-integrated solutions. End-of-pipe solutions were rated signicantly lower than process-integrated solutions as regards the inuence on product quality (Question 2). The respondents obviously assumed a very high degree of feedback from the end-of-pipe solution to the production system. As the respondents have experience in the implementation and operation of end-of-pipe solutions, it can be assumed that end-of-pipe solutions have had negative effects on production systems in the past (see Section 4). The respondents also rated the likelihood of exceeding the legal emission limits as signicantly higher when using end-of-pipe solutions than when using processintegrated solutions (Question 3). Less trust is placed in end-of-pipe solutions to fulll environmental goals than is the case for process-integrated solutions. This risk estimation can be attributed to the surveyed peoples experience with end-of-pipe and process-integrated solutions. As regards the economic risks of water conservation solutions, the importance of the following factors was surveyed: The costs of implementing the environmental performance improvement solution (Question 7); Operating costs for normal operation (Question 8), and Expected time required for amortization (Question 6). The interviewees only judged the signicance of the time required for amortization differently for the two different solutions. For end-of-pipe solutions, the amortization time was judged to be considerably less signicant for decision-making than was the case for process-integrated solutions. This may be due to the fact that endof-pipe solutions must be installed to ensure lawful production processes and therefore the question of amortization time does not arise [13,14]. The main concerns are keeping the investment and operating costs as low as possible [11]. Process-integrated solutions, however, are judged by their cost-effectiveness. The required time for amortization is therefore assigned more importance [13,16].

692

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

Table 1 Estimated risks with the implementation of water conservation measures Question No. Risks which the person surveyed estimated to be especially important when carrying out water conservation measures 1 Problems occur in the production system due to increased complexity arising from the integration of water conservation measures The water conservation measures have a negative inuence on the product quality Lawful emission limits are exceeded due to the water conservation measures The production output decreases due to the water conservation measures The water conservation measures fail The planned amortization period of the water conservation solution is considerably exceeded The costs of introducing the water conservation solution are considerably exceeded The operating costs of the water conservation solution are considerably exceeded There are problems with the ofcial authorization of the water conservation plant is not signicant (p 0.05). is signicant at the level 0.01 (high signicance, p 0.01). is signicant at the level 0.001 (very high signicance, p 0.001).
a

The type of water conservation solution which Signicance level was estimated to have the lowest risk End-of-pipe solutions
a

2 3 4 5 6 7 8 9

Process-integrated solutions Process-integrated solutions End-of-pipe solutions End-of-pipe solutions Process-integrated solutions End-of-pipe solutions End-of-pipe solutions Process-integrated solutions

It is interesting to consider the general conditions in which the decision is made to use end-of-pipe solutions. In a survey undertaken in 1995, it was discovered that end-of-pipe solutions are implemented (signicantly) more often just before a tightening-up in the law than is the case with process-integrated solutions. Process-integrated solutions are usually not implemented when there is pressure due to lack of time. Signicant differences were also discovered in the importance of problems with the ofcial approval of environmental performance improvement solutions (Question 9). The interviewees rated the risk to be considerably higher with end-of-pipe solutions than with process-integrated solutions. The reason for this might be that end-of-pipe measures usually require ofcial approval, whereas this is often not the case for integrated solutions. 4.2. Problems in the implementation and normal operation of water performance improvement solutions When planning projects, companies try to obtain and analyze information so they can reduce or exclude any risks in implementation and operation. As explained in Section 3.1.1, the complexity of the planning and therefore the planning risk are higher with process-integrated solutions than end-of-pipe solutions. If we assume a qualitatively identical planning phase, independent of the type of solution, it is expected that end-of-pipe solutions cause considerably less problems in the implementation and operation phases than process-integrated solutions.

Contrary to these expectations, it was discovered that more problems occur in the implementation phase of end-of-pipe solutions than occur in the implementation phase of process-integrated solutions. Companies experienced considerable problems in the implementation stage with 72.5% of the end-of-pipe solutions and only 61% of the process-integrated solutions (Table 2). Independent of the type of solution, technical problems were named as the most important type of problem. 76.1% of the problems with process-integrated solutions were problems in the technical implementation of the system. Considerably fewer of the problems with endof-pipe solutions (62%) were problems in the technical implementation of the system. The most common technical problem was the coordination with other processes. This formed 58.8% of the technical problems with endof-pipe solutions and 42.4% of the technical problems with process-integrated solutions. A further central problem was the increased occurrence of plant shutdown with end-of-pipe solutions, with repercussions for production (end-of-pipe: 27.5%; process-integrated solutions: 33.3%) [16]. Less important for both end-of-pipe and integrated solutions were the qualitative decrease in production output (end-of-pipe: 3.9%; process-integrated solutions: 9.1%) and the quantitative decrease in production output (end-of-pipe: 2%; process-integrated solutions: 12%) and incorrect operation by employees (endof-pipe: 7.4%; process-integrated solutions: 3%). Even after the end-of-pipe solutions began normal operation there were repercussions for production systems. Plant shutdown causes repercussions for the pro-

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

693

Table 2 Problems in the implementation of water conservation measures Problems in the implementation phase of completed water conservation measures End-of-pipe solutions (%) 72.5 62.0 58.8 27.5 3.9 2.0 7.4 28.0 8.0 Process-integrated solutions (%) 61.0 76.1 42.4 33.3 9.1 12.0 3 21.1 0.0

Considerable problems in the implementation phase Of which the following % were technical problems Of which the following % were Problems coordinating with other processes High breakdown rate of water conservation solutions Temporary decrease in the quality of the production output Temporary reduction in the amount of production output Operating error caused by an employee Of which the following % were economic problems Of which the following % were legal problems

duction process in 7.2% of the end-of-pipe solutions. With 21.7% of the end-of-pipe solutions, qualitative and/or quantitative decreases in the production output were reported. This suggests that the central argument for the economic advantages of end-of-pipe solutionsthe unilateral decouplingcarries less weight than was previously assumed. 4.3. Why end-of-pipe solutions should not be seen as decoupled from the production system To completely rule out the repercussions for the production system caused by end-of-pipe solutions, the process and control effects of end-of-pipe solutions on the production system must be prevented. 4.3.1. Process decoupling When establishing the operating point of a transformation process, the interface data are xed both for the input and the output. Interface data are the qualitative composition and the quantities of materials/energy, which are transferred from one process to another dependent on time. The operating point of a transformation process is the vector of process variables (speed, pressure, temperature, etc.), which, in combination with the process input, produce a certain output. The selection of the operating point also determines the physical properties (pressure, temperature, specic volume, etc.) of the input and output ows at the interfaces of the transformation process. With the unilateral decoupling of systems (according to the above denition), the interfaces between the production and the end-of-pipe solution are set autonomously by the mode of operation of the production system. If permanent legal conformity is demanded, the end-of-pipe solution must also be able to handle sudden uctuations in the input volume (independent of its composition and state) and process this according to the prescribed emission levels. This requirement does not

appear to be fullled by end-of-pipe solutions. To ensure the smooth running of the end-of-pipe solution, the endof-pipe solution-specic quantitative and qualitative restrictions placed on the input ow must be adhered to. If these restrictions are not adhered to, the following consequences may arise: Illegal emissions released into the natural environment due to an overloading of the end-of-pipe solution; Shutdown of the end-of-pipe solution (worse case scenario: damage to the plant), and Repercussions for the production system (i.e. due to increased pressure or temperature at the interface). Ignoring cases in which illegal emissions are planned, repercussions for the production system can arise due to the violation of the permitted input interface data. For example, if less than optimal operating conditions cause a heat exchanger to cool the efuent inadequately, the efuent temperature at the interface with the production system may increase. This has an effect on the production process, which produces this efuent and changes its operating conditions. From the viewpoint of process technology, a unilateral decoupling of production system and end-of-pipe solution cannot generally be assumed. 4.3.2. Control decoupling Along with pure process dependencies between the production system and end-of-pipe solution, there are usually also control dependencies. The purpose of control systems is to establish the operating point of the transformation process based on input and output conditions and requirements. Unilateral decoupling requires that the control system of the end-of-pipe solution does not inuence the control system of the production system and therefore does not inuence the mode of operation. This requirement appears to be unrealistic. If qualitative and/or quantitat-

694

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

ive uctuations occur in the output ow of the production system, it is useful if the control system of the end-of-pipe solution inuences the production system. This ensures the undisturbed operation of the end-ofpipe solution. The operating points of the production processes are thereby changed so that overloading of the end-of-pipe solution is preventively avoided. The control system of the end-of-pipe solution can also inuence the production system when there is a shutdown of the endof-pipe solution. If constant legal conformity is required, whenever the capacity limits of the end-of-pipe solution are exceeded or close to being exceeded, then the control system of the end-of-pipe solution should have a controlling inuence on the production system. Usually, then, from the viewpoint of control technology, unilateral decoupling cannot be generally assumed. To prevent process and control repercussions, buffers1 can be installed between the production and end-of-pipe solution. These buffers balance out uctuating interface data. The buffer system is then part of the end-of-pipe solution. In principle, this allows the requirements for unilateral decoupling to be fullled. However, when large quantities of residual products must be handled, this option has technical and economic limits. A temporary intermediate storage is often not possible and/or prohibitively expensive. In summary, it can be said that production systems and end-of-pipe solutions usually have a reciprocal relationship. The repercussions from the end-of-pipe solution in the production system often have a negative inuence on qualitative and quantitative production output. The advantages of unilateral decoupling found in the literature could therefore not be proven in many cases for water conservation measures. 4.4. End-of-pipe solutions are mature technologies The advantages of end-of-pipe solutions due to their technological maturity and the specialization of providers could not be veried by the survey results. Although, there were considerably fewer cases of serious technical problems in the implementation phase of endof-pipe solutions than with process-integrated solutions, the occurrence of problems (in 62% of cases) is still very high. End-of-pipe solutions also seem to pose more problems in normal operation than is the case with process-integrated solutions. Forty-two percent of the completed end-of-pipe solutions had to be improved after beginning normal operation in order to ensure continued operation. Technical problems occur regularly with 18.8% of the end-of-pipe solutions and these problems
1 Buffers are understood to be storage capacities between transformation processes. These storage capacities allow a temporary qualitative and/or quantitative equalization of the output of the producing process and the input of the following process.

lead to system breakdown. In contrast, 10.7% of the process-integrated solutions had to be improved after implementation and 3.4% suffer from regular plant shutdowns [6]. 4.5. End-of-pipe solutions have a high level of legal security This claim could not be veried by the survey results. Serious legal problems had to be overcome in the implementation phase of 8% of the completed end-ofpipe solutions. The problems occurred here in connection with the requirement for the end-of-pipe solution to be ofcially authorized and related legal regulations. No legal problems were registered in the implementation phase of process-integrated solutions. There is usually no ofcial authorization required. With 23.2% of the end-of-pipe solutions, legal problems arose after beginning normal operation. With one exception, these problems were all the result of exceeding permitted emission limits. The normal operation of process-integrated solutions posed considerably less problems of this type in comparison (2.3% of the integrated solutions) [13].

5. Summary End-of-pipe solutions are often presented in the literature as technically and economically less problematic than process-integrated solutions. This claim is based on three assumptions: unilateral decoupling, technological maturity, and the high level of legal security of end-ofpipe solutions. A survey was carried out to investigate whether these three basic assumptions could be veried in the eld of water conservation. First, it was investigated whether the decision-makers in the production industries rated the risks of end-of-pipe and integrated solutions similarly. The risks were broken down into technical risks, economic risks and legal risks. In summary, it can be said that the complexity of process-integrated solutions was recognized as the central technical problem. Due to this complexity, interviewees feared greater problems in the implementation phase and normal operation of process-integrated solutions. Endof-pipe solutions were also rated as less risky from an economic point of view. However, the interviewees saw denite advantages with process-integrated solutions as regards the ofcial approval of the plant. In summary, it can be said that interviewees thought that process-integrated solutions involved less risk than end-of-pipe solutions. Marginal advantages of end-of-pipe solutions were recognized in the implementation phase. However, the expected advantages of end-of-pipe solutions in the implementation phase and in normal operation based on the three basic assumptions could not be veried. The

K.A. Zotter / Journal of Cleaner Production 12 (2004) 685695

695

theory that process-integrated solutions pose more problems and more serious problems than end-of-pipe solutions could not be substantiated. In fact, the exact opposite was the case here. The completed process-integrated solutions are much more stable in their operation than end-of-pipe solutions.

[10] [11] [12]

Acknowledgements I wish to thank Heinz Strebel and two anonymous referees for valuable remarks and helpful suggestions for improvement of this article. Remaining errors are mine.

[13]

[14]

[15]

References
[1] Bundesministerium fu r Land- und Forstwirtschaft. Richtlinien fu r die Begrenzung von Immissionen in Fliegewa ssern, Wien; Februar 1987. [2] Conway St, Steward F. Networks and interfaces in environmental innovation: a comparative study in the UK and Germany. J High Technol Manage Res 1998;9(2):23953. [3] Dunn RF, Bush GE. Using process integration technology for CLEANER production. J Cleaner Prod 2001;9:123. [4] Dupuy D. Technological change and evironmental policy: the diffusion of environmental technology. Growth Change 1997;28(Winter):4966. [5] Fischwasser K, Schilling H. Schwermetalleliminierung aus ssern. Wasser, Luft und Boden, H. 4; 1992. p. 346. Abwa [6] Florida R. Lean and green: the move to environmentally conscious manufacturing. Calif Manage Rev 1996;39(1):80105. [7] Fo rstner U. Integrated pollution control, Berlin; 1995. [8] Graedel Th, Horkeby I, Norberg-Bohm V. Prioritizing impacts in industrial ecology. In: Sokolow R, et al., editors. Industrial ecology and global change, Cambridge; 1994. p. 35970. [9] Green K, McMeekin A, Irwin A. Technological trajectories and

[16]

[17]

[18]

[19]

[20] [21]

[22]

[23]

R&D for environmental innovation in UK rms. Futures 1994;26(10):104759. Haberfellner R, u.a. Systems EngineeringMethodik und Praxis, 7. Au., Zu rich; 1992. Hansen J. Integrierter Umweltschutz; UmweltWirtschafts Forum, 7. Jg. H. 2; Juni 2000. p. 812. Hartje VJ. Zur Struktur des o kologisierten Kapitalstocks: Variablen und Determinanten umweltsparender technologischer Anpassungen in Unternehmungen. ed. Wissenschaftszentrum kologische Modernisierung der ProBerlin fu r Sozialforschung. O duktion, Berlin; 1990. p. 13598. Hilson G. Pollution prevention and cleaner production in the mining industry: an analysis of current issues. J Cleaner Prod 2000;8:11926. Meyer-Krahmer F. Industrial innovation strategiestoward an environmentally sustainable industrial economy. In: MeyerKrahmer F, editor. Innovation and sustainable development: lessons for innovation policies, Heidelberg; 1998. p. 6985. Michaelis P. Betriebliches Umweltmanagement: Grundlagen des Umweltmanagements in Funktionsbereichen, Berlin; 1999. Porter ME, van der Linde C. Toward a new conception of the environmentcompetitiveness relationship. J Econ Perspect 1995;9(4):97118. Sagar AD, Frosch RA. A perspective on industrial ecology and its application to a metals-industry ecosystem. J Cleaner Prod 1997;5(1-2):3945. Schwarz EJ, Steininger KW. Implementing natures lesson: the industrial recycling network enhancing regional development. J Cleaner Prod 1997;5(1-2):4756. Schwarz EJ, Ho llweger Th. Einfu hrung umweltorientierter Prozesstechnologien in Produktionsunternehmen. In: Blecker Th, Gemu nden HG, editors. Innovatives Produktions- und Technologiemanagement, Berlin; 2001. p. 34984. Thom N. Grundlagen des betrieblichen Innovationsmanagements, 2. Au. Hanstein; 1980. Vickers I, Cordey-Hayes M. Cleaner production and organizational learning. Technol Anal Strategic Manage 1999;11(1):75 94. Zimmermann K. In: Simonis UE, editor. Technologische Modernisierung der Produktion. Eine Variante pra ventiver Umweltpolitik. Frankfurt: Pra ventive Umweltpolitik; 1988. p. 20526. Zotter K. Industrielle Wasserwirtschaft, Wiesbaden; 1995.