Synergies and trade-offs between ecological and productivity-enhancing measures in industrial production – a systematic review

Purpose – Improving productivity and efficiency has always been crucial for industrial companies to remain competitive. In recent years, the topic of environmental impact has become increasingly important. Published researchindicatesthatenvironmentalandeconomicgoalscanenforceorrivaleachother.However,fewpapershavebeenpublishedthataddresstheinteractionandintegrationofthesetwogoals. Design/methodology/approach – In this paper, we identify both, synergies and trade-offs based on a systematic review incorporating 66 publications issued between 1992 and 2021. We analyze, quantify and clusterexamplesofconjunctionsofecologicalandeconomicmeasuresandtherebydevelopaframeworkforthecombinedimprovementofperformanceandenvironmentalcompatibility. Findings – Our findings indicate an increased significance of a combined consideration of these two dimensionsofsustainability. We foundthat caseswhereenforcing synergiesbetweeneconomicand ecological effects were identified are by far more frequent than reports on trade-offs. For the individual categories, cost savings are uniformly considered as the most important economic aspect while, energy savings appear to be marginally more relevant than waste reduction in terms of environmental aspects. Originality/value – No previous literature review provides a comparable graphical treatment of synergies andtrade-offsbetweencostsavingsandecologicaleffects.Forthefirsttime,identifiedmeasureswereclassifiedina3 3 3 table considering type and principle.


Introduction
More than 30 years ago, the so-called Brundtland Report concluded that industry is both a cause of environmental problems and an important enabler for change through economic growth.Even then the need to reconcile environmental protection and economic growth was recognized as possible and desirable (World Commission on Environment and Development, 1987).
Meanwhile, the manufacturing industry must not only meet customer demand efficiently, but also adhere to the social and environmental requirements of a wider set of stakeholders.Therefore, companies' strategy should simultaneously consider and balance financial, ecological and social goals.Spreckley (1981) was presumably the first who criticized that the Eco-efficiency in manufacturing appraisal of industrial and commercial performance ignores social and environmental costs of the production process.This led to the well-established triple bottom line perspective of sustainability: economic prosperity, environmental quality and social justice (Elkington, 1994, 1997, Parkin et al., 2003).The shear zone between the economic and environmental agendas is called eco-efficiency (Elkington, 1997) and has been examined in a more differentiated manner over time (Kleine and von Hauff, 2009;McDonough and Braungart, 2002a) as shown in Figure 1.
In literature one can often find the blanket assertion that improving the company's environmental performance is linked to long-term cost reduction (e.g.Florida, 1996;Despeisse et al., 2012;Hart, 1995).Higher profits of a more ecological production could be attributed to higher productivity, but they can just as well be the result of avoided penalties (Tan et al., 2022) or increased sales due to a better corporate image (Mart ın-de Castro, 2021).Baines et al. (2012) have already shown in a detailed review that a higher resource productivity can offset the cost of environmental improvements.However, this evaluation was conducted a decade ago and should now be repeated to include state-of-the-art measures to increase productivity and/or environmental performance.
According to Baah et al. (2021) environmentally sensitive production processes have a negative influence on financial performance.It is also conceivable that ecological aspects (and related monetary and non-monetary goals) outweigh the exploitation of all productivity potentials.
The target of this paper is to screen publications with cases where quantitative results of symbiosis, synergies, conflicts or rivalries between ecological and productivity-enhancing actions have been reported.Similar to Abolhassani et al. (2018) we use the term "productivity enhancement" synonym to productivity improvement and other expressions for increased production efficiency or production performance.The examples should also fit into a descriptive framework which consists out of a starting condition, followed by an event with one or more actions and resulting in an outcome where the impact caused by the implementation is evaluated (used in a similar way by Baldassarre et al., 2019).Afterwards, statistical characteristics as well as correlations and patterns are recognized and revealed.
In section 2, relevant terms are described since a clarification is needed for the clustering of the result of the literature review.Section 3 is about the methodology, design and search strategy of the literature review.The listing and examination of the data from 66 publications

Background
Increasing efficiency in production has always been in focus; prominent examples around the middle of the 18th century are the concept of division of labor and the beginning of the Industrial Revolution.In the past decades, ecological sustainability has become increasingly important in addition to pure business activities: stringent environmental policies and regulations, natural resource preservation and the public image affect the competitive advantage of a company (Despeisse et al., 2012).Back in the 1980 and 90s, environmentally conscious thinking and acting was recognized and researched as an unavoidable success factor for companies (i.a.Hart, 1995).
A typical scenario where environmental protection and profitability are not compatible, are more ecological productions whose additional costs do not amortize (e.g.due to elasticity of demand).A vivid example is the dilemma of surface treatment mentioned in Luttropp and Karlsson (2001): just omitting coating would be more ecological due to the avoidance of harmful substances, but could have a negative impact on aesthetics and customer acceptance and in the end on sales and narrow profitability goals.
There are countless names for the various approaches and different possibilities for joining efficient industrial production with environmentalism.However, there is also no consistent terminological definition with strict distinctions.This complicates a literature review in that the search strategy cannot be narrowed to just a few keywords without running the risk of overlooking relevant publications.In their review in the field of "industrial sustainability", Smart et al. (2017) distilled 114 related keywords within multiple iterations and the final set of search strings still consisted of 90 keywords.
In general, it can be stated that terms that contain the word "sustainability" (e.g.sustainability manufacturing, industrial sustainability) consider all three dimensions mentioned in Section 1.Most publications also adhere to this concept (e.g.Paramanathan et al., 2004;Smart et al., 2017).In case of deviations, this is indicated by corresponding designation (e.g.environmental sustainability in Sarkis and Zhu, 2018).Davis and Costa (1995) coined the term "environmentally conscious manufacturing"; it improves the environmental attributes of product manufacturing, ideally without sacrificing quality, cost or performance.The focus is on materials processing and manufacturing operation steps in order to reduce their environmental impact independently of the product.Smart et al. (2017) have decided to use the theme "material utilization and process optimization", which includes concepts like "sustainable manufacturing", "eco efficient processing" or "eco materials".
It is also Smart et al. (2017) who distinguish here between product and process/production.Of course, many environmental aspects such as the substitution of toxic materials by nontoxic ones, selection of suitable recyclates, source reduction and durability, reparability and dismantling of products are determined in the product development process.In this context, the terms "design for environment" (DfE, Davis and Costa, 1995) and "eco-design" (e.g.Luttropp and Lagerstedt, 2006) are commonly used, sometimes even as synonyms (e.g.Despeisse et al., 2012).According to Graedel and Allenby (1995), DfE deals with products and processes before they are introduced and integrates environmental aspects over the entire lifespan of a product starting with the early stages of product design.It has become an important constraint for other considerations (e.g.design for manufacturing, design for (dis) assembly).Despeisse et al. (2012) criticized that DfE has a strong focus on products and encompass more than what a manufacturer has immediate control over.Anyway, an earlier integration of environmental thinking leads to better results in decreasing more effectively Eco-efficiency in manufacturing the environmental impacts of a product can be reduced (Keoleian and Menerey, 1994;Luttropp and Lagerstedt, 2006).
For this paper, the distinction between eco-efficiency and eco-effectiveness is of major importance.Eco-efficiency is about production processes with lower ecological impact.Most of the related strategies cover reduction of pollution (up to zero emissions), minimization of throughput and toxicity of materials used and energy-saving measures.Recycling capabilities only lead to downgraded material quality with limited usability and are not altering the linear progression of material flows (also known as downcycling; Ellen MacArthur Foundation, 2013).Eco-effectiveness is a concept where the production of goods goes beyond the reduction of negative consequences still existing in eco-efficiency.It is characterized by upcycling and a cradle-to-cradle design: products and processes are changed to be supportive for the environment (Burchart-Korol et al., 2013).A self-sufficient closed-loop circulation of resources where waste from one component of the system represents input to another resembles a biological ecosystem and is seen as the highest level of ecological sustainability (Graedel, 1994;McDonough and Braungart, 2002b).According to Despeisse et al. (2012), few industries have considered their manufacturing as such ecosystems where material, energy and waste are used not only in an efficient way, but also in an effective way.By tracking process flows with a holistic view, compatible outputs and demands of processes can be identified.Ideally, virgin inputs can be substituted by wasteful or unwanted outputs generated elsewhere in the production system.Thus, for instance, efforts for imports and exports of resources are reduced or completely avoided, and thereby the environmental impacts are reduced while achieving economic savings.Baldassarre et al. (2019) summarizes industrial symbiosis as a synonym for a cooperative network of separate industries to exchange materials, energy, water and/or by-products.Related areas and aspects are circular economy and industrial ecology; Bruel et al. (2018) lists concepts, principles and tools both have in common, but also recommend to extend the focus by socioeconomic aspects.
An eco-effective transformation of resources can only be achieved via complete cyclicity.However, in industrial companies, individual components of the ecosystem (production processes, factories, etc.) are still examined in isolation.Thus, rather linear or quasicyclic flows are redesigned in the sense of (eco)efficiency (Graedel, 1994).At inter-enterprise level, the reduction of net resource input as well as pollutant and waste outputs remain essential.Nevertheless, a company that reuses its waste internally is more efficient than one which only focuses on the ratio of output over input of individual processes and fails to consider waste as a resource (Despeisse et al., 2012); the optimized environmental impact through resource depletion is a pleasant side effect.With all the above information, further topic narrowing is possible.The Venn diagram related to the triple bottom line (see Figure 1) is extended by two sets for product and production (illustrated by triangles, Figure 2).Since a product comprises more than just the processes of its creation, while production is meaningless without an associated product, the area of the product triangle is larger than that one for production and encloses it.The concentricity indicates, that for all dimensions and their shear zones, there are aspects that relate to the product but not to its fabrication; Sole exception is the center where all three agendas meet, as goods should only be considered fully sustainable when this also applies for their production.The production related economic/ecological shear zone marks the scope of this literature review and depicted by the grey area in Figure 2.
Consequently, product design and development, the utilization of the product and its endof-life management are out of scope of this literature review.Also, the selection of (virgin) materials which are directly incorporated in the product (raw and auxiliary material, semifinished from suppliers) are not considered.However, operating supplies and their environmental footprint as well as the utilization of by-products are relevant, as they are MEQ allocated to production.What also stays in focus are all kind of production systems.Viewed at the most abstract level, a production system converts resources into valuable output, typically via machining and assembly (Hitomi, 1996;Riggs, 1970).These systems are depending upon technology, equipment and industrial engineering techniques, and aiming at maximum productivity (Hitomi, 1996).Our further consideration includes individual processes and production steps, single production lines, factories, up to production networks (sometimes also called industrial parks).In individual cases, manufacturing technologies are also taken into account, as long as the increase in productivity has a significant environmental contribution (i.e. it exceeds the related energy savings) and vice versa.

Methodology
This systematic literature review is based on the PRISMA statement (preferred reporting items for systematic reviews and meta-analyses) developed by Moher et al. (2009).This guidance assists researchers in conducting reviews and reporting the results to ensure quality, clarity and transparency.For this purpose, the review process is subdivided into four phases: identification, screening, eligibility and finalizing the list of included studies.
The first phase is used to identify the objective of the research and the relevant keywords.Moher et al. (2009) as well as the latest update of the PRISMA statement (Page et al., 2021) recommend to provide an explicit statement of all objectives or questions the review addresses.The corresponding checklist (latest version PRISMA, 2020) explicitly refers to the PICO framework.PICO is an acronym standing for population, intervention, comparison, outcome.This methodology was initially developed to test the effectiveness of interventions in medical practice (Richardson et al., 1995) and is widely used in health research.Although the PICO framework has been used almost exclusively in evidence-based medicine so far, there are a few applications in other research areas recently (i.a.Baashar et al., 2020;Burton et al., 2018;Dong et al., 2021).For this paper, a modified form of the PICO framework with its four criteria is used.In the present case, intervention (5 ecological measures) and comparison (5productivityenhancing measures) may but do not have to counteract each other since the targets of both measures can be combined.PICO stays suitable because it structures the literature search where we want to identify the relationship of two different measures (within industrial production (5population).The effects on the performance (e.g.improvements due to symbiosis and synergies, or deteriorations due to conflicts and rivalries) represent the outcome.A definition of the PICO criteria and their counterparts in this systematic review are listed in Table 1.

Eco-efficiency in manufacturing
Each criterion is used to define a related search strings by identifying relevant keywords that serve as inclusion and exclusion criteria.This approach is in line with PRISMA (2020) according to which the PICO framework can be used to specify characteristics used to decide whether a publication is eligible for inclusion in the review or not.
Keywords for the first criterion "population" should ensure that search results deal exclusively with industrial production (second sector of the economy; economic activities that fall under section C in ISIC, 2008).Outside of the scope have been papers that refer to the primary sector of the economy where raw materials are extracted or produced such as mining and agriculture (including farming, logging, forestry and fishing).Also, construction industry and utility companies have been excluded since they do not fabricate conventional chattels.First queries showed that the keyword "production" resulted in many hits related to the agricultural sector; as a consequence, this term was replaced by more specific synonyms to narrow the search.
Second, the publications must explicitly refer to ecological measures (interventions) that directly affect the production.The explanations in Section 2 give an indication which key words are suitable for this search component and which are not.For example, the obvious word "sustainability" was not used because it also includes the society dimension (referring to social equity, health, safety, fairness, etc.)."Ecology" and "environment" were more appropriate words.Keywords which are referenced in the publications mentioned in Section 2 were adopted.Since each additional key word within a search string increases the number of hits, ambiguous and misleading terms such as "carbon" were not used.Due to the focus at the level of the production system rather than corporate governance, articles dealing with such topics as policy, carbon offsetting or renewable energies have been excluded in our paper.
For the search component about productivity-enhancing measures, the relevant keywords first had to be determined in a separate search step: in December 2021 the search string (productivity AND improvement AND industry AND production) was entered in Scopus database.This query led to 3,275 document results and was then limited to documents associated to the subareas "Engineering" (1,443 results), "Business, Management and Accounting" (511), "Material Science" (428) and "Chemical Engineering" (323).In Scopus it is possible to generate an overview of the keywords declared in the found documents and their frequency of occurrence.Among others, the terms "Lean Production" (134 results), "Lean Manufacturing" (83), "Industry 4.0" (54) and "Automation" (47) were noticeably often linked to the search results and consequently included in the search string.While management techniques (namely total quality management, just-in-time and lean production) have already Effects on performance and competitiveness (outcome) can be measured via common (financial) ratios such as market share or return on investment (ROI).Moreover, a company's reputation or customer satisfaction can be used as an indicator.Figge et al. (2002) but especially Geyi et al. (2020) summarized financial and operational performance measures which are partly adopted for this review.
The formulation and selection of keywords followed the principles stated in McGowan et al. (2016).Keywords created in this way were then combined into search strings by using Boolean operators and inserted as a search filter in the command lines of databases (Table 2).Where necessary for individual databases, the filters were adjusted.
These search filters were specified to be either in the title, abstract or keywords.Neither the year range nor the subject areas were specified in the search.There was no limitation to publication type but the search excluded those sources unavailable in the English language.

Results of literature review
The searches were run in January 2022 in Scopus and Web of Science databases and yielded 1,766 publications.In addition, there were 15 already known articles, some of them were even presented in Section 2. Altogether, 1,781 publications, including 340 duplicates, were identified.After screening the titles of all results from initial search, in a first iteration, only 546 potential publications remained.However, after abstract screening, 51 papers could be excluded.
Critically reading the identified 495 publications entirely in a second iteration, determined 43 sources relevant to this review.Together with additional 23 eligible papers found at the reference lists of the remaining publications (so-called backward snowballing; Wohlin, 2014), a total of 66 publications were included in the final analysis.
The above-mentioned phases of the review process are illustrated by the flow diagram in Figure 3.
As stated above, empirical cases providing examples and results of ecological and productivity-increasing measures are the main criteria for a publication to be considered for the final analysis.In exceptional cases, theoretical work like simulations or mathematical or

Search component
Search string

Industrial production
Manufacturing OR assembly OR fabrication AND Measures to improve the environmental sustainability "environmentally conscious manufacturing" OR "design for sustainability" OR ecodesign OR "design for the environment" OR greenhouse OR emission* OR "carbon footprint*" OR "environmental footprint" OR "circular economy" OR clean* OR ecolog* OR environmental OR "life cycle assessment" OR decarbonization AND Measures to improve productivity productivity OR lean OR "industrial economics" OR "industry 4.0" OR agile OR automation OR digitalization OR "lead time reduction" OR "technical efficiency" OR "cost reduction" AND Effects on performance and competitiveness competitive* OR profit* OR revenue OR "return on investment" OR roi OR "return on capital employed" OR roce OR "market share" OR sales OR turnover OR "customer satisfaction" OR reputation Source(s): Authors' own elaboration Table 2.

Search filter
Eco-efficiency in manufacturing experimental results and conceptual papers were also included, but not interview and questionnaire-based research when providing qualitative results only.Since the focus is primarily on industrial production, publications relating to the complete supply chain or the product itself (esp.design, material, remanufacturing; as mentioned at the end of Section 2) were excluded as ineligible.The same applies to publications related to the energy sector (including biofuel production) and services.Articles referring to governmental aspects (e.g.Porter hypothesis, emission taxation, national pollution control) as well as economics (such as region-or country-based studies) are beyond the scope and were ignored.
The risk of bias was mainly assessed by examining the disclosure statements and sponsors of the publications.In this respect, it should be noted that nine publications (Buandra, 2019;Glick and Shareef, 2019;Jarrell, 1992;Parthasarathy et al., 2005;Stoll et al., 2008;Takada et al., 2008;Tokawa et al., 2001;Vargas and Scott, 2017;Yamazaki, 2017) are coauthored by company representatives.In three additional cases, cooperation with companies is at least noteworthy (Tiwari et al., 2020;Yang and Feng, 2008;Zhu et al., 2007).The majority of publications are peer-reviewed journal papers or part of conference proceedings; exceptions are Ndikumana (2019), Pampanelli et al. (2015) and Wills (2009).Nevertheless, the authors of this paper concluded that there are no reasons for the exclusion of specific sources.All authors agreed on the final selection of publications and cases.
The high number of 66 relevant publications containing 84 cases from industry indicates the importance of the topic.A list of these cases together with a short description of measures and their economic and ecological effects can be found in the Appendix.
The company size is not documented for all of the sample cases, but at least 19 of them fall within the definition of a small and medium-sized enterprise (SME).

MEQ
The selected body of literature comprises papers that have been published in a 30-year period ranging from 1992 to 2021.As Figure 4 shows, most of the sources (52) have been published in the more recent years, starting from 2011, with the highest number of works (8) published in 2017 and 2020, thus highlighting the growing interest devoted to the topic by scholars.Furthermore, the Journal of Cleaner Production dominates in terms of number of relevant studies, with 12 articles, followed by Production Planning and Control (5 papers),  Sustainability (5) and Clean Technologies and Environmental Policy (3).
The company location is stated for 72 of the 84 cases.Figure 5 presents the information on the geographical distribution.Asian rank first by representing 38% of all analyzed cases led by India (10 cases) and China (7 cases).European countries are the source of 17 cases (24%), followed by North America (21%) and South America (16%).The lowest publication level was identified in Africa (2 cases) and Australia (no case in the final selection).With 14 cases, the USA are the country with the biggest contribution to this study.
Figure 6 portrays the industry-based distribution of the cases.The classification is in line with the divisions mentioned in ISIC ( 2008), 20 out of the 24 manufacturing-related divisions defined by ISIC are represented by the cases.Three cases do not allow conclusions about the industry.Since a company can operate in more than one single industry sector, the total number of assignments listed here is 88 and therefore still higher than the original number of cases.A considerable amount of research has been done in the manufacture of fabricated 43 of all cases refer to "costs" suggesting that cost savings are an important indicator for the economic success of a measure.
One of the central questions is what productivity-enhancing measures were used in the cases.Of particular interest is the spread of individual techniques and trends in their utilization.For this purpose, all cases were tagged in the further course of the analysis.These tags are "Recycling/reuse/circular economy", "Automation", "VSM", "Industry 4.0/IoT", "Six Sigma" and "Lean"; depending on the measure(s), it occurred that one case corresponded none or to several of the aforementioned tags (e.g.lean Six Sigma led to one tag for lean and another for Six Sigma, if applicable also a third tag for VSM when explicitly used in the case).
The first tag also covers publications dealing with industrial symbiosis or upcycling.Although VSM is a recognized method used as part of lean methodology as well as in Six Sigma, it is occasionally also used as an independent tool; hence it is treated as a separate category.The tag "Industry 4.0/IoT" is used for all measures which are associated with digitalization of production and sensors used there.The frequent references to Six Sigma caused this term to become a distinct tag as well.Originally designed to improve manufacturing quality, it also brings structure to process improvement through a definemeasure-analyze-improve-control cycle (Pande et al., 2000).Cases tagged with "Lean" typically implemented 5S workplace organization, continuous improvement process (also known as Kaizen), cellular manufacturing and/or just-in-time production.
As can be seen in Figure 7, there "Lean" is the most brought up measure and related to 43% of all cases."Recycling/reuse/circular economy" is the second most widely used measure, thanks to the fact that it has been in focus since the 1990s and also remains relevant today.We found the theme of "Industry 4.0/IoT" represented only in 5% of the analyzed cases; those cases are from recent years and could be considered as new and promising tool.
Trends from the figure above are consistent with the findings from previous reviews (e.g.Cherrafi et al., 2016;Chugani et al., 2017) according to which lean production as well as Six Sigma (and of course combinations of both) are frequently used to achieve eco-efficiency.Direct search results and snowballing did not reveal fundamentally new techniques that were not already known when the search filter was designed.Only few detailed technical solutions (Huang et al., 2017;Yamazaki, 2017;Khan et al., 2021) as well as the use of artificial intelligence (Adeniji and Schoop, 2021) can be mentioned as extraordinary.
Since ecological aspects are multifaceted, clustering them is also a reasonable option here.It seems purposeful to loosely adhere to ISO 14044 -respectively Amrina and Yusof (2011) - MEQ and distinguish between "Waste", "Energy" and "Emissions".The first tag covers solid and liquid waste and especially hazardous substances."Energy" refers to the consumption of all kind energy sources and carriers such as fuels or electricity.Cases where air pollution are reported are attributed to "Emissions".Figure 8 illustrates the occurrence of these three ecological aspects over time; multiple tagging was possible.
Whereas waste and its reduction have been of consistent relevance constituting 53% of all cases, the issues of emissions and particularly energy have become much more important in recent years.Energy (efficiency) has a massive proportion of 57% of the cases considered.Although emissions to air and exhaust gases are comparatively seldom discussed (29% of all cases), an increase can be seen in the past few years.Whereas up to and including 2013 VOC were in focus when a case considered waste gases, since then GHG and above all CO 2 have been mentioned in almost all cases with reference to emissions.
The influence of ecological aspects on efficiency-enhancing measures in industrial production is emphasized in the introduction of most publications analyzed.Insights gained through the literature review suggest that improvements of the environmental performance Eco-efficiency in manufacturing of an industrial production have a positive impact on the productivity.For instance, Teng et al. (2014) found out that an environmental commitment proven by a certified environmental management system leads to benefits on economic performance.Especially Ben Ruben et al. (2017) states that initiatives to improve environmental performance must be aligned with traditional manufacturing strategies to improve metrics such as process efficiency, profitability or quality.Reliable figures are provided by Diaz-Elsayed et al. (2013): changes from the real state to the green state amount to 4.7% of the overall 10.8% savings in production costs.
For the vast majority of cases, the economic and ecological effects can be contrasted.Table 3 shows a rough rating of the extent of the effects.Improvements in the double-digit percentage range (for economic effects also payback periods shorter than five years) are considered substantial and represented by þþ.A smaller improvement is subsequently designated as moderate (þ).Mentioned improvements, which are given only in absolute values or without any values at all, are indicated by þ? and commented.○ is used if no changes have occurred.Declines are graded in the same way as improvements (-and -).
If both the economy dimension (i.e.positive economic effects) and the ecology dimension are improved by a measure, synergy can be assumed in the respective case.If an enhancement of one dimension is achieved at the expense of the other, the affected case is listed under trade-offs (these cases are grey shaded in Table 3).
To allow further conclusions the classification of measures is further refined.In the majority of cases, more than one measure was used.
The analysis of the cases shows that synergies between ecological and productivityenhancing measures clearly outnumber conflicts and rivalries.Only ten cases of trade-offs between an increase in productivity or profitability and ecology could be identified; however, for some cases, external influences such as price increases or declines in sales volume may have affected the result.Clusters of trade-offs among specific industries or countries are not evident.This underpins substantial empirical evidence suggesting that productivityenhancing measures can offset the cost of environmental improvements if these arise at all (Baines, 2012).
Baumer-Cardoso et al. (2020) and some articles referenced there suggest that while more frequent setups increase flexibility and achieve many positive ecological and economic effects, in the case listed, water consumption also increased.When comparing two technical solutions, Khan et al. (2021) evaluated and quantified a total set of 17 criteria.Lower manufacturing costs were associated with higher environmental impacts and vice versa.The lower-emission solution requires subsequent cleaning, which takes additional resources.A comparable case is described by Mangili and Prata (2020) where the lower-emission technology has a higher raw material consumption.Yue and You (2013) reported a deteriorated environmental impact per functional unit with increasing productivity.Another example of a trade-off is provided by Jayachandran et al. (2006): the most environmentally sustainable production process is associated with significantly higher and therefore unprofitable production costs.So, the technology was not used due to lack of commercial viability.
Furthermore, it is crucial which reference is considered: Leme et al. (2018) shows a case where eco-efficiency is higher even with an increased carbon footprint of the factory.After converting setup time into productive time, the machines will demand higher power levels because they are not in standby mode anymore.Consequently, the total energy consumption increased together with the production quota, but also the ratio of production time to carbon emission improved.Choudhary et al. (2019) No details about energy savings, pollution and waste reduction þþ No details about cost savings (declared as "significant" and "substantial")

Eco-efficiency in manufacturing
Cartesian system; cost savings are plotted on the horizontal axis, and ecological effects on the vertical axis.Similar to Karvonen (2001), quadrant I contains the synergy cases or win-win situations where both effects complements, as an increase in one effect will lead to an improvement of the "opposite" one.Trade-off situations (lose-win or win-lose) arise when only one effect is enhanced but the other does not (quadrant III resp.quadrant IV).
The figure above suggests a correlation between energy savings and cost reduction respectively between emission reduction and cost reduction.It also shows that in three out of four cases, energy savings result in similarly high (CO 2 ) emission reductions.
Correlations can be assessed by using the statistical indicators Pearson correlation coefficient (r) as well as the coefficient of determination (r 2 ).The obtained Pearson correlation coefficient value of r 5 0.57 indicates a moderate positive correlation between emission reduction and cost savings with a quite high goodness of fit (r 2 5 0.32).Energy savings and cost reductions show a weaker relationship (r 5 0.22, r 2 5 0.05).Due to the outlier from Jayachandran et al.'s (2016) case, we calculated a negative correlation between cost savings and waste volumes (r 5 À0.43, r 2 5 0.19), which implies that the avoidance of waste is economically rather disadvantageous.By ignoring the extreme case, a positive correlation is obtained (r 5 0.29, r 2 5 0.08).
It is important to note that the effects occurred in different time periods.In the case of Thanki and Thakkar (2020), this period is always two years.Once again, the extent to which effects can be explained by changes in sales and prices for raw material and energy remains largely ignored due to lack of data.
The resulting publications were also grouped into a 3 3 3 table to reveal the nexus between the type of measures and their guiding principles of impact.In the course of the literature review, it became obvious that the measures taken affected processes and/or the use of technology.As soon as new, different or additional equipment is used for optimization, it is a "technological measure"."Processual measures" are all corrective actions which change workflows and configurations, but use the existing production means.Of course, there are also measures that combine both types.The impacts of the measures can be subdivided into reduction, replacement or recycling in the broadest sense.Reduction means lower resource consumption (in some cases up to complete elimination); examples are reduced energy consumption, shortened idle times, avoided waste and eliminated work steps.Alternatively, hazardous substances, for instance, can be substituted by environmentally compatible Synergies and tradeoffs between cost savings and ecological effects MEQ solutions; in the case of such a replacement, resource consumption does not necessarily have to be reduced, at least from an ecological viewpoint.The principle "recycling" summarizes all kinds of waste treatment including upcycling, circular economy and industrial symbiosis.The assignment of publications within this framework can be seen in Table 4.It can be seen that the constellation "processual, reduce" is observed most frequently, which is due to the widespread use of the lean methodology (Figure 7).
Per se, there is no better or worse when it comes to the principles.But they can be understood as evolution steps: while reduction is almost synonymous with eco-efficiency, recycling rather aims at eco-effectiveness.The principle of replacement can be located in between, depending on whether it reduces or avoids environmental impacts.More recently, these principles are taken as hierarchical models (e.g.Kurdve and Bellgran, 2021;Lim et al., 2022).In the end, zero-waste performance can be the aim or result of both lean and circular production (correlations were proven by Afum et al., 2022).
When looking at Table 4, it is noticeable that reduction is very strongly associated with measures related to processes.This is not surprising, since lean principles are also designed to minimize all kind of waste and resource consumption, but they do not necessarily change the equipment or technology used.The situation is different for replacements, where modified production processes almost inevitably lead to a change of machines, tools and/or operating supplies.The distribution of publications related to recycling shows that both processes and the used technology are often adapted.
A further evaluation of the typecasted publications aims at a closer examination of the economic and ecological aspects considered.Table 5 shows which type of measure has an influence on what aspect (ordered by frequency of being mentioned).
Consistent with Figure 8, it can be seen that energy and waste are the crucial ecological aspects for almost all combinations of measure and principle.For measures focusing on "reduce", it seems that scheduling and timing is of particular importance.
The following tables are attempts to reveal further correlation and clusters of the typecasted measures, types of measures and principles of impacts.The numbers in parentheses indicate the frequency of occurrence.
By looking at the relative frequency, it can be observed in the case of individual measures (Table 6) that a changed layout reached, e.g. by rearranged stations or processes (B) can be considered to be a particularly eco-efficient measure.A double þþ rating was achieved in five out of six cases (83%).Also retrofitting of material flow (C), optimized transportation (J) and resource balancing (O) achieved this rating particularly frequently (67% each).
With regard to the types of measures (Table 7) and the absolute frequency, it can be stated that a combination of technological and processual measures results in the largest synergies.In the case of the principles of impact (Table 8), no robust statements can be made.

Mainly technological
Mainly processual

Technological and processual
Table 4. Typecasting of publications according to type of measure and principle of impact Eco-efficiency in manufacturing
In most of the analyzed cases, it became evident that improvements of the environmental performance of an industrial production have also a positive impact on the productivity and vice versa.Only ten out of 84 cases reported conflicts and rivalries between productivityenhancing measures and ecological aspects.This advocates environmentally conscious manufacturing from an economic viewpoint.There is little doubt that economic and environmental aspirations are already being addressed together.Rather, there is evidence for integrated approaches to achieve eco-efficient improvements.
Cost savings were identified as the most important economic aspect for an eco-efficient production as more than half of the cases deal with them.The ecological aspects can be e f f e c t s ++ Technological ( 5) Processual ( 10) Both ( 14) Processual ( 1) Both ( 1) Processual (1) +?

Eco-efficiency in manufacturing
divided roughly into energy, waste and emissions, whereby the latter have only recently gained importance.Energy savings appear to be marginally more relevant than waste reduction.For a total of 14 cases, correlations between cost savings and ecological effects could be represented graphically in a diagram.
This paper has also introduced a typecasting of measures via a 3 3 3 table for the first time.Measures can involve processes, technologies or both.These measures lead to reduced resource use, replacing previous means or recycling.It has been shown that measures aimed at reduction are primarily process-related and typically do not involve new technologies.Replacements, on the other hand, regularly require a modification of the used technology.To enable recycling or circular economy, often both processes and technologies are adapted.For academia, but especially for industry, our finding that the combination of technological and processual measures promotes eco-efficiency more than the isolated implementation of new technologies is highly meaningful.In relation to managerial implications, this study contributes to a better understanding on the potential effects of specific measures since data of various cases were examined and improvements up to the double-digit percentage range were found.
A next step toward understanding the synergistic relationship of ecological and productivity-enhancing measures is certainly to further develop the concept of the 3 3 3 table.In this way, potentials could be quantified and benchmarks for individual industries could be determined.This would enable a better comparison of heterogeneous cases.A potential result could be a process model to evaluate synergies and trade-offs and to support decision-making by using an empirically determined dataset.Moreover, established multicriteria decision-making methods can be applied to quantitatively analyze interdependencies (e.g. by means of DEMATEL) or alternatives (e.g.PROMETHEE), possibly even in combination (Chowdhury and Paul, 2020).The final outcome might be a sector-specific guideline to identify the most eco-efficient measures.For this, more reference cases are needed that reflect the effects of individual measures, both in isolation and in combination with other measures.
Figure 1.Dimensions of sustainability and integrative sustainability triangle Figure 2. Extended Venn diagram to express focus of the literature review Figure 3. Flow diagram summarizing yield of literature review Figure 4. Temporal distribution of publications and cases used for further analysis Figure 6.Distribution of industry sectors assigned to cases Figure 7. Temporal distribution of productivityenhancing measures mentioned in cases Figure 9. Synergies and tradeoffs between cost savings and ecological effects also point to higher energy consumption due to increased production efficiency, but at the same time advises a correspondingly ecological procurement strategy.