Making Engineering Design Sustainable
Citation: Sharon Beder, 'Making Engineering Design Sustainable', Transactions of Multi-Disciplinary Engineering Australia, Vol. GE17, No. 1 June 1993, pp. 31-35. It is an extended version of a paper presented at the Design, Ecology and Economy Conference, University of Sydney, November 1990.
This is a final version submitted for publication.
Economic considerations have always been an integral part of engineering design and a force for refinement and sophistication of design methods. Engineers see themselves as the interface between science, technology and business allowing industry to create wealth from scientific and technological developments. Environmental considerations have been peripheral and secondary at best, something forced upon the engineer as the community becomes more and more demanding.
Environmental considerations need to be incorporated into engineering design in the same integral way that economic considerations are. As well as continually refining their designs to minimise use of materials, energy and labour engineers could also be refining their designs to fit into the environment harmoniously and with minimum disruption or degradation of natural ecosystems. Just as engineers apply safety factors in their design to compensate for uncertainties about the strength of their structures, they could also apply safety factors to compensate for uncertainties about the environmental consequences of their projects. Such changes imply the need for a new engineering philosophy and ethic and therefore changes to engineering education.
Different societies and different times have required different design methods, concepts and regimes. Industrialisation required the abandonment of traditional design methods and a move to the drawing board and mathematical models. Engineering design was characterised by the incorporation of economic considerations as an inherent and inseparable part of the design process. A new kind of development which is ecological sustainable will similarly require new approaches to design which incorporate environmental considerations as an inherent and inseparable part of the design process.
2. Traditional Design Methods
Most writers agree that engineering design in the past two hundred years has been markedly different from previous product and building design. J Christopher Jones differentiates between traditional methods and modern methods in the developing design process in his book, "Design Methods: Seeds of Human Futures".1 Using the example of the gradual evolution of the wagon wheel as an example of traditional methods of design, he concludes that traditional craftsmen did not draw their products before building them and were often unable to explain satisfactorily why they did things the way they did. Rather information is memorised during apprenticeship and stored as patterns (Jones gives the example of cross-section profiles). He quotes George Sturt, a wheelwright himself:
Jones explains that a product design was modified over hundreds of years by trial-and-error in a "slow and costly sequential searching for the invisible lines of a good design". The final result was often "an astonishingly well-balanced result" which was well suited to the needs of the user.3
Christopher Alexander calls this an "unselfconscious process". Products, he claims, were bound to adapt towards an "equilibrium of well-fitting forms" because the only incentive to change was when there was a misfit.4 Any failure of a product to fit its context properly was corrected on the spot, often by the user who was also the builder, for example in the case of housing.
Good designs, according to Alexander, were culturally fixed by patterns of myth, traditions and taboo. Changes were only made when there were "powerful (and obvious) irritations in the existing forms" which needed to be corrected. Likewise Jones claims that a design was only changed to correct errors or meet new demands. He argues this was because, since the reasons for a particular design were not known, drastic changes to a well evolved design meant that "the patiently-won balance and fit" would be lost. Similarly Alexander suggests that to have an "equilibrium of well-fitting forms", the adjustment of those forms must evolve faster than "the drift of the cultural context".5
3. The Move to the Office
Design methods changed with the changing requirements of industrialising nations. The traditional evolution of forms was no longer fast enough to keep up with the constant demand for new products in rapidly expanding economies. The design process was removed from the site of manufacture to the drawing board where scale drawings were made. The designer now had to achieve "in a few hours at the drawing board what once took centuries of adaptation".6 The scale drawing became the medium for experiment and change.
This facilitated three major changes. Firstly, because dimensions were specified in advance, manufacture could be split into different parts to be done by different people allowing for the all important "division of labour" which marked an industrial society. Secondly, scale drawings allowed large projects to be undertaken by a number of craftspeople. By accurately specifying dimensions in advance the work of several craftspeople was more likely to fit together. Thirdly, the division of labour allowed by scale drawings, enabled production to be speeded up. Components whose dimensions were specified in advance could be standardised and machined.7 This process removed intellectual activity from manufacture and gave it to a new class of people who made drawings.
It was not just the move to the drawing board that transformed engineering design. Engineers were increasingly expected to incorporate economic considerations into their designs and streamline their use of materials. To do this they needed to be able to predict fairly accurately how their products and structures would perform. They made use of models, particularly mathematical models, for this purpose. The models approximated the product in a way that enabled them to calculate what sorts of forces and stresses it might be subject to in service and how it would behave. Models, like drawings which were really visual models, enabled the trial and error process to occur on paper and enabled engineers to refine their designs to their most economical. The idealised structure or product was tested to see if it failed and if it did it was beefed up and tested, mathematically, again.
The push for the increased scientisation/mathematisation of engineering design therefore came in large part from the desire to keep manufacturing and construction costs to a minimum. Engineers try to reduce the cost of their structures by reducing their weight and using materials more efficiently.8 The seemingly inherent requirement of minimum material for engineering design is illustrated by the example of Robert Stephenson whose railway bridges were made of "huge rectangular tubes of steel". Stephenson wrote to Roebling who was planning to use a suspension bridge to carry trains across the Niagara: "If your bridge succeeds, then mine have been magnificent blunders."9
Mike Cooley has observed this desire for minimum material usage in engineering products other than building structures and suggests this is why engineers must increasingly resort to mathematical solutions.
4. Separation from the User Context
Economic considerations have become an integral part of engineering design and a force for refinement and sophistication of design methods. Engineers see themselves as the interface between science, technology and business allowing industry to create wealth from scientific and technological developments. But this is often at the expense of other design considerations.
The reiterative process of design, modification and remodification and the increasingly scientific/mathematical nature of engineering design has the drawback that, unlike the former craftsmen, designers are weaker at judging the compatibility of a designed product with its environment and the context of its use. Alexander describes the modern process of design as a "self-conscious process". He argues that the reaction to failure is no longer immediate and direct, partly because of the mediation of specialists and the use of transported materials.11
Eugene Ferguson also sees the move away from non-verbal thinking to more analytical and scientific modes of thought as drawing the engineer away from the "complexities of the real world". He suggests that courses in design at universities have such low status compared with analytical courses that they may soon disappear leaving the way open for stupid mistakes which can only happen when a designer fails to visualise their product in use. He speaks of "the chaos that results when design is assumed to be primarily a problem in mathematics".12
Computer models are a more recent development in the design process which amplify the removal of the engineers from the context of their work. Henry Petroski suggests the use of computers will mean a further separation of design from context and use just as Jones recognised the drawing board as weakening the grasp of designers on the compatibility of their products with the context in which they would be used. The trend in structural engineering, says Petroski, is toward engineers favouring packaged programs written by structural analysts. He argues that the design and construction-site experience and background of these analysts tends to be limited.13
Arnold Pacey, in "The Culture of Technology" (1983) also tackles the idea that technological design seems to be divorced from the context of the use of products. He gives examples of big dams feeding leaking pipes and electricity generating stations pumping heat into the atmosphere when electricity is mainly used for heating, as examples of "halfway technology".14 This is perhaps what Alexander would describe as forms that badly fit their context.
There are various ways in which designers seek to overcome their removal from the context of their products. Apprenticeships, say Jones, serve the function of acquainting designers with design features that are "impractical, expensive, or not to the liking of customers". However the practice of promoting engineers from the shop floor or from the field is disappearing and few engineers serve apprenticeships any more. Also the making of physical models and prototypes enable performance to be tested but this tends to be difficult in the case of large-scale one-off engineering projects.
Another way that designers deal with ensuring that their designs fit their context is to use general principles and to categorise "misfit variables" into areas such as "ergonomics" or "acoustics" so that a student can learn to design and cope with the complexity that unselfconscious design methods dealt with over long time periods.15 Such categories cover a range of environmental factors but it is a very indirect way of dealing with environmental considerations compared with the way that economic considerations have come to be dealt with.
5. Environmental Considerations
Current design methods prioritise economic considerations over environmental ones. In some cases economic considerations also serve environmental goals. For example the minimisation of materials used in a structure means resources are saved. However if they are saved at the expense of the length of the operating life of a product then economic considerations conflict with environmental interests which would demand that products be made as durable as possible because of the need to minimise resource usage and waste generation in the long term.
Environmental considerations are marginalised in the design process to the extent that Environmental Impact Statements have been introduced around the world in recent years in an effort to ensure that the environment is considered by those designing significant engineering projects. Yet whilst environmental impact statements may deal with the gross impacts of developments on the environment, the environmental considerations that should be considered at the design stage of every product and project of whatever size, such as the choice of materials, layout and processes and the implications that follow from these, remain neglected.
This treatment of environmental impact, if applied to economic impacts, would be akin to designing each part of a project without concern for how much the materials were to cost and then at the end doing an economic impact statement to see if the cost of the overall structure was going to have too much impact on the economy. Not only would you end up with a needlessly expensive project even if the economy could stand it but any attempts at that stage to cut down on costs would interfere with the integrity of the design and it would not be the most efficient way of going about things.
Engineers use safety factors to cover uncertainties about how their products will be used and behave in the real world. A safety factor, sometimes referred to as a "factor of ignorance", allows for a margin of error or the bad luck combination of poor materials and overstressing.16 Safety factors bring into focus the conflict between safety and cost considerations. A large safety factor will require the use of more materials and parts and may make an item less competitive with one that is less conservatively designed.
Safety factors in engineering design, however much they might be whittled down, are a reluctant concession to the overriding need for safety yet there is no similar concession to the need for sustainability. Safety factors do not incorporate environmental considerations and engineers do not build environmental safety factors into their designs to ensure environmental protection (environmental protection factors") even though the uncertainties surrounding the impact of their designs on the environment are even greater than the uncertainties surrounding the performance and capabilities of their products in use. Too often engineers feel it is sufficient that a proposal only just meets legal environmental standards on paper and no allowance is made for uncertainty with regard to environmental performance and impacts.
It may be argued that environmental standards themselves incorporate safety factors but this is often not the case because of the same economic pressures that preclude their use in engineering design. At best environmental standards in Australia are based on "best practicable technology". In other words the standards set are those that can be met using technology which is readily and economically available rather than being those that are sustainable in the long-term. The meaning of "readily and economically available" is negotiated with engineering designers.
6. Case Study - Sydney's Deepwater Outfalls
The choice of deepwater outfalls as a solution to Sydney's sewage pollution problems is an example of how engineering practice falls short of what is required for an ecologically sustainable future.
Sydney's deepwater outfalls were designed to meet negotiated standards set in 1974. They were also designed to meet the political objective of removing the large stains surrounding the shoreline outfalls.17 The engineers involved felt they were discharging their responsibilities in meeting these requirements and did not set out to protect the environment, per se, nor to design a sustainable solution that would meet the requirements of future generations. This has until now been acceptable engineering practice.
The water quality standards that the deepwater outfalls had to meet, WP-1, "Design Criteria for Ocean Outfalls," were negotiated between the Water Board and the State Pollution Control Commission after both organisations had decided that the deepwater outfalls were a good idea. WP-1 included criteria for toxic substances and faecal coliform in terms of concentrations rather than maximum quantities. This in itself encouraged an engineering solution that provided dilution rather than treatment before discharge.
The deepwater outfalls were designed to achieve dilution, dispersal, and die-off of indicator organisms; not to ensure bathing waters were free of pathogens (disease causing organisms). Caldwell Connell, the Water Board's consultants, did not try to measure levels of pathogens in the ocean resulting from the existing sewage outfalls. They did experiments on the die-off rates of faecal coliform but not on die-off rates of pathogens.18 Faecal coliform are organisms which occur naturally in the human gut and do not cause disease themselves. They are used as indicators of sewage pollution. They die relatively quickly in the ocean (1-9 hours) unlike pathogenic bacteria and viruses which can live for long periods of time in the ocean (sometimes months). The deepwater outfalls would provide further distance for the sewage to travel before coming onshore and this would be enough time for most of the faecal coliform to die and therefore the bathing water standards, which were in terms of concentrations of faecal coliforms, could be met.
Similarly the outfalls were designed to dilute and disperse the toxic waste so that concentrations of heavy metals and chlorinated hydrocarbons, at the required distance from the outfalls after first dilution had occurred, would be less than the levels specified in WP-1. There was little effort made to discover the mechanisms that might cause these toxic substances to concentrate in sediments and marine life. In fact WP-1 ensured that more toxic waste would be allowed to be discharged into the ocean each year when the deepwater outfalls were built because they provided many times the dilution than achieved by the shoreline outfalls. Caldwell Connell, the Water Board's consultants did not test fish for bioaccumulation but a Water Board employee did some tests for himself in 1973. These tests showed clear evidence that heavy metals were even at this early date accumulating in marine life.19
The deepwater outfalls were also designed to ensure the sewage field would remain submerged below the surface of the ocean for most of the time in summer so that it would not be seen, even from the air. The submerged field would travel in the direction of the deeper ocean currents rather than in the direction of the surface currents and the wind. The fate of a sewage field driven inshore by currents travelling shoreward was not investigated. Nor was the eventual fate of the sewage which travelled southwards as it was supposed to.20
The deepwater outfalls were supposed to substitute for upgraded onshore treatment (only 10-15% of suspended solids were being removed before the sewage was discharged). In this way the deepwater outfalls would meet the legal and political requirements but not necessarily the requirements of ecological sustainability.
Not only did the engineers involved accept the standards in WP-1 as sufficient, the engineers of both the Water Board and the SPCC were actively involved in setting those standards and then engineers in both organisations and their consultants helped to persuade the public that the deepwater outfall was an adequate long-term solution.
In recent years the inadequacy of the deepwater outfalls as a long-term solution has been confirmed by an independent team of engineers, Camp, Dresser & McKee commissioned by the NSW government to review the Water Board's Beach Protection Programme.21 Yet such is the engineering culture that the same firm of engineers, had it been given the brief given to Caldwell Connell (that is to come up with the most cost-effective solution to meet WP-1) may also have proposed the deepwater outfalls. The engineers who work in the Water Board claim they have been caught out because of changing community requirements.
Legislative requirements such as the preparation of Environmental Impact Statements and the setting of standards, whilst necessary, are insufficient to ensure ecological sustainability. In particular, standards and criteria tend to shift responsibility for designing environmentally sound engineering works away from the engineers who can feel satisfied they have discharged their responsibilities if they meet the standards, whether or not those standards are sustainable. Standards tend to be based on what has been economically achieved before, whereas engineers can push open the boundaries of achievement.
Engineering design is built upon a massive engineering science base which includes a comprehensive and detailed analysis of materials and their behaviour under a variety of stresses and strains. It is time for engineering design to now also build on ecological science and the interrelationships between their products and various ecosystems. This means that equal attention must be given in University courses to ecological science as materials science.
It is during their education that engineers first learn how to approach design, yet at the moment engineers are exposed to very little in the way of environmental education. It is true that several environmental engineering degree courses have begun but this sectioning off of environmental education as a speciality does not achieve the overall changes in engineering philosophy and design methodology that are required to transform engineering design to work for rather than against the environment.
The new era of sustainable development requires a new approach to engineering design that brings the designer closer to the context of use of the final product, particularly the environmental context. Ecological considerations need to be considered as part of the design process in the way that economics is, so that micro-decisions at every stage of the design process take account of resource and energy use, pollution and sustainability.
Safety factors or environmental protection factors need to be built in to ensure environmental protection as well as safety. Such changes imply the need for a new engineering philosophy and ethic as well as changes to engineering education which will ensure that engineering science begins to encompass the science of ecology as it does the 'science' of economics.
1. Jones, J.C., Design Methods: Seeds of Human Futures, 1980 edition,
John Wiley & Sons, (1980).