• Aucun résultat trouvé

A Critique of Life Cycle Assessment; Where Are the People?

N/A
N/A
Protected

Academic year: 2021

Partager "A Critique of Life Cycle Assessment; Where Are the People?"

Copied!
6
0
0

Texte intégral

(1)

A Critique of Life Cycle Assessment; Where Are the People?

The MIT Faculty has made this article openly available.

Please share

how this access benefits you. Your story matters.

Citation

Gutowski, Timothy G. “A Critique of Life Cycle Assessment; Where

Are the People?” Procedia CIRP 69 (2018): 11–15. © 2018 The

Author

As Published

http://dx.doi.org/10.1016/J.PROCIR.2018.01.002

Publisher

Elsevier BV

Version

Final published version

Citable link

http://hdl.handle.net/1721.1/119641

Terms of Use

Creative Commons Attribution-NonCommercial-NoDerivs License

(2)

Procedia CIRP 69 ( 2018 ) 11 – 15

2212-8271 © 201 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the scientific committee of the 25th CIRP Life Cycle Engineering (LCE) Conference doi: 10.1016/j.procir.2018.01.002

ScienceDirect

25th CIRP Life Cycle Engineering (LCE) Conference, 30 April – 2 May 2018, Copenhagen, Denmark

A Critique of Life Cycle Assessment; Where Are the People?

Timothy

G.

Gutowski

Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA

Corresponding author. Tel: 01-617-253-2034; E-mail address: gutowski@mit.edu

Abstract

In this article, we provide a framework and examples to illustrate how human behavior can disrupt the expected environmental outcomes suggested by Life Cycle Assessment (LCA). The biggest problems are when LCA results are scaled up to make claims about possible future outcomes within a narrowly crafted scenario that ignores real human behavior. There are however, many examples at smaller scales too, where (often) engineers using LCA appear to ignore the behavior of other engineers involved in the development scenario. We argue for the inclusion of more social science in the development of environmental LCA, and the use of a wider array of scenarios (grounded in real behaviors) to portray the possible future development of a product or service at large scale. Ultimately people, not products, need to be at the center of this discussion.

© 2017 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the scientific committee of the 25th CIRP Life Cycle Engineering (LCE) Conference.

Keywords: human behavior; Life Cycle Assessment; social science

1.Nullius in Verba

It seems fair to say that engineers do not always know how to value the work of the social sciences. We frequently refer to them as the “soft sciences”. The implication is that we, engineers, are engaged in the hard sciences. We use numbers not words. The precedent was long ago established by the Royal Society of London. Founded in 1662, their motto, “Nullius in Verba” means roughly don’t trust their words. Of course, there are good reasons for this bias. Words can be tricky, appeal to passions and mislead. Numbers, on the other hand, imply origin, methodology and concreteness. As engineers, we believe in rigor and numbers and now apply our methodologies to an ever-increasing array of important problems. These are problems that need insight, assuredness, and answers. Certainly, quantifying our anthropogenic assault on the earth and the possible consequences for our sustainability would be one of these problems. Our engineering based approaches have made solid contributions to these new sustainability applications. These would include Material Flow Analysis (MFA), Lifecycle Assessment (LCA)

and related ways of performing these calculations and displaying the results such as the use of Sankey diagrams. Recent efforts to move beyond relative comparisons and consider planetary boundaries constitute a further improvement on these methodologies [1, 2, 3].

There is no doubt that these tools have helped to elevate the discussion of engineering sustainability problems. However, at the same time, they have exposed some limitations of these methodologies especially when LCA results are scaled up in an attempt to suggest possible future outcomes. This problem is especially noticeable when the study is addressing an emerging technology. No one knows, for sure. exactly how and when new technologies will emerge, but it is well known that early expectations always outpace reality. In fact, this phenomenon has been captured in the so-called Gartner Hype Cycle for Emerging Technologies [4] which suggests that there is a peak of inflated expectations followed by a trough of disillusionment before the technology really emerges. One can see this same trend in the LCA and environmental analysis literature that addresses emerging technologies such as 3D printing, LED light bulbs, carbon

© 201 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

(3)

12 Timothy G. Gutowski / Procedia CIRP 69 ( 2018 ) 11 – 15

nano-tubes, and autonomous vehicles. It is not the purpose of this article to argue for or against these claims, but to expose the constrainingly narrow assumptions that are often used to develop these claims, ultimately marginalizing the utility of their results. These errors can be compounded by publishers who desire clear positive outcomes and are inclined to interpret rigor as exclusively in the domain of the quantitative methodology while being unconcerned with the quality of the underlying scenario constructions.

This is not a new problem. In fact, there is no shortage of overambitious claims about the benefits of a new technology. For example, the telephone was to reduce car trips [5], email would reduce paper use [6], more efficient light bulbs would reduce energy use [7] and improved American football helmets would reduce head trauma [8]. These developments were easily predicted but are often difficult or impossible to prove. Such optimistic predictions are good for promotion but can be misleading when actually trying to predict the real outcome of a new technology intervention. The references given above generally describe how these claims did not come to fruition.

2.Nullius in Numera

A review of the recent literature on the environmental evaluation of new technologies, sometimes in very prestigious journals, clearly indicates that numerical calculations are not immune from bias and self-promotion. We would characterize the problem as the algebra of delusion: small hypothetical improvements X large number of potential improvers = misleading conclusions. We have observed 4 main types of errors: 1) misrepresentation of the incumbent technology, 2) applying an undeserved advantage to the new technology, 3) being unaware of the potential shortcomings of the new technology and 4) a misinterpretation of how the new technology would be used. We will offer a few examples below, to illustrate how environmental analysis can be distorted by these four errors when attempting to project the potential benefits of new technologies.

2.1. Misrepresentation of the incumbent technology

A recent study comparing additive manufacturing versus casting for an automotive diesel pump claimed a substantial weight savings for the optimized additive pump, and hence substantial material and environmental savings. This, no doubt is true, but at the same time the original cast pump was most assuredly designed to minimize cost, and not weight. To make a fair comparison here, one would need to at least speculate on how the cast pump could be made at a lighter weight. Secondly, potential weight savings need to be evaluated along with cost to have any idea if the weight savings would be implemented.

In another study, an additively manufactured part design is assumed to be injection molded so as to make a comparison. But, anyone who knows injection molding would know that the part would have first to be redesigned for it to be injection molded. Furthermore, injection molded parts are always designed to have thin cross-sectional areas to reduce cycle

time. As a consequence, the redesign could shift the advantage away from additive manufacturing especially for large production runs.

Generally, large scale manufacturing process claims are based upon the analysis of a few parts that are intended to represent all parts. But often the few parts were selected precisely because they show an advantage and not because they are representative. For example, there have been many studies comparing additive metal manufacturing to complex hard metal machined parts with very high buy to fly ratios. These typically show (a well-deserved) benefit for the additive technology, but these very high buy to fly ratios are not at all typical for most machined parts, and therefore some care needs to be taken when attempting to scale up these results. Of course, more detailed comparisons are tedious, but not impossible. For example, Cooper [9] has compared incremental sheet metal forming to conventional metal forming processes for a wide variety of parts in terms of energy, cost and lead time.

2.2. Applying an undeserved advantage to the new technology

We saw this routinely in early comparisons of photovoltaic (PV) electricity generation versus incumbent, mostly fossil fuel, technologies. Unfortunately, 1 kWh of solar PV electricity just does not equal 1 kWh of coal-fired electricity. Coal-fired electricity is available day and night, rain and shine. To be equivalent in this regard, solar PV would need significant overcapacity and storage. When this is added, the comparisons between technologies in term of cost, energy payback time, energy return on investment, etc. look significantly less impressive, but give a much more realistic notion of the challenges involved in promoting solar PV. Another type of advantage seen in the literature is when future improvements to the grid are applied to the new technology but not to the electrical incumbent. And still another example is when changes in human behavior are applied to the new technology but not to the incumbent technology. For example, a recent article on autonomous taxi vehicles, suggested that people would be willing to use smaller vehicles at higher occupancy rates. In addition to this being completely contrary to recent behavioral trends in the United States [10], this behavior was only applied to the autonomous vehicle and not to the conventional vehicle.

Under this same category it should be acknowledged that there are significant challenges to accurately representing how a new technology will evolve and improve. The pathway from technological discovery, to pilot demonstration, to early and then mature product is complicated and largely unknown. This makes it tempting to use well known mathematical simplifications such as learning curves and diffusion models. But this should always be done with caution and with some effort to understand the new technology.

We have tracked the early energy efficiency behavior of two emerging technologies: carbon nanotubes and laser metal additive manufacturing (AM) and found very different behaviors. Rapid yield improvements for the HiPco process (~ 30%/year over ten years) resulted in substantial reductions in energy intensities for this carbon nano-tube production

(4)

technology [11], while significant rate improvements for metal AM (mostly with higher powered lasers) hardly changed the adiabatic rate efficiency nor the energy efficiency [12].

2.3. Being unaware of potential shortcomings of a new technology

Of course, there are significant challenges in making thoughtful predictions about the future development of a new emerging technology. But nevertheless, early on, people involved in the development of these technologies often know of potential problems. One example of this kind of problem occurred in the early days of the development of advanced composite materials which competed directly with aluminum for airframe structures. Many estimates were made about how composites could bring down weight, save fuel and hence save money. In these calculations, the nominal values for stiffness and density were used as would seem appropriate. But because of the way advanced composites are made, by additive processes, there is the chance of noticeable variation in these values, especially in stiffness and strength. The eventual result is that because of this variation it is recommended that one use a larger safety factor in the design of composite structures which of course results in a corresponding reduction in the amount of weight that can be saved [13].

2.4. Misrepresentation on how the new technology would be used

This is by far the largest category for potential errors when LCA results are scaled up. The problem starts immediately when one uses the common assumption in comparative LCA that one technology would substitute for another, and that the user would comply with the wishes of the LCA analyst and use the product in the way it can derive the proposed environmental benefit. The problem is intimately connected with the selection of the functional unit. In order to compare two technology options, it is routine to identify a common unit of service offered, such as kilometer of travel, cubic meters of cooled space, lumens of lighting etc. to facilitate this comparison. This is alright for attributional analysis, but at the same time it is usually unsatisfactory for any speculation regarding how this potential substitution might actually play out. The problem is that people do not make decisions based solely upon the basic service offered but instead based upon the entire package of services offered including price, but also convenience, status and compatibility with other personal preferences and desires. Without consideration of these other attributes, it is hard to make meaningful statements about the potential benefits of the introduction of new technologies. This kind of problem can occur within the engineering user domain, or the final product user domain or anywhere else in the lifecycle of the product.

As an example, within the engineering user domain, it is a common assumption that reducing the weight of the components of a vehicle, say a car or an airplane, could result in a reduction in the total weight of the product (vehicle) with

a resulting reduction in the energy required during the use phase. We all know that weight figures prominently in increasing the kinetic energy required to move a vehicle, and that size, often correlated with weight, figures prominently in aerodynamic drag. So, we are on firm technical ground making this claim, but will this weight savings of a component or structure be used to reduce the weight of the vehicle? The answer based upon some historical evidence presented here shows clearly that it may not work this way. For example, a recent study of vehicle weight distributions for automobiles in the US shows that impressive gains in reducing the weight of automobile structures during the late 1970’s and early 1980’s have been significantly eroded by equally impressive increases in weight for convenience, safety and entertainment features for these vehicles [14]. Similarly, for a class project, two of my students compared two Boeing passenger airplanes with nearly identical seating capacity but one made of lightweight advanced composite materials for their total weight. The results are shown in table 1. You will see that the conventional metal, mostly aluminum, airplane (767-400ER) actually weighs less than the advanced graphite fiber composite airplane (787-8) [15]. How can this be?

Table 1. Specification of 767-400ER and 787-8

Specification 767-400ER 787-8

Passengers (3-class) 245 242 Operating Empty Weight 103,870 kg 110,000 kg Max Take Off Weight 204,120 kg 228,000 kg Fuel Capacity 90,100 L 126,920 L Range 10,415 km 14,200 – 15,200 km Engines P&W PW-4000-94 GE CF6-80C2 GE GEnx Rolls-Royce Trent 1000 Engine Thrust (x2) 282 kN 280 kN

Because the engineers who designed the airplane had other goals in mind. They could've used this structural weight savings in a variety of ways, for example for increasing seating capacity, or cargo capacity, or increasing entertainment equipment. It appears they used this additional weight savings to increase fuel capacity and thereby increase the range of the airplane. You may argue that this could reduce refueling stops and therefore still have a positive effect on overall fuel efficiency. This is correct. At the same time, the longer-range makes traveling longer distances more appealing to customers and could increase airplane usage. After all, this would be the goal of the airlines. To ignore this goal in creating a scenario for the future benefits of lightweight materials in the absence of any new incentive to reduce fuel use would be a biased scenario. What we have just described here, is a version of the rebound effect, only in this case applied to materials, and then indirectly to energy.

(5)

14 Timothy G. Gutowski / Procedia CIRP 69 ( 2018 ) 11 – 15

There are many other examples like this. We have listed nine examples in Table 2. The ones with an asterix* need additional work to demonstrate. The others are discussed in the references we have already listed. Many of the examples are variations on the so-called rebound theme. The energy rebound phenomenon has been widely discussed by economists [7, 16, 17, 18], and more recently investigated by engineers and physical scientists [19, 20, 21, 22, 27].

Although the rebound discussion has often been a contentious one, especially between economists and engineers who favor energy efficiency, the evidence continues to grow that this effect cannot be ignored and is probably much larger than people have previously estimated. In this paper, we prefer to address this complexity by looking at the motivations of the actors involved in the various stages of product development and use, and there by illustrate how human behavior interferes with some of the common assumptions employed in the LCA of technology. The overwhelming evidence, is that without constraints, people invariably want more. See the previously mentioned rebound references, especially [20] and also [23]. As stated by Yuval Noah Harari in his recent book Sapiens: A brief history of humankind, “one of history’s few iron laws is that luxuries tend to become necessities and to spawn new obligations” [24]. Ignoring this behavioral reality is one of the most common shortcomings of engineering environmental analysis of technology. There is growing evidence that engineers are realizing these issues, see for example [25, 26, 27].

As a first step in addressing the four problems that we have identified we suggest that the LCA analyst consult with the proper representative of the missing actors in the analysis. In particular, problems 1 and 2 could be addressed by consulting with someone familiar with the incumbent technology. Problem 3 could be addressed by consulting with someone actively involved in the development of the new technology. And problem number 4 could be addressed by better understanding of the behavior of users which is often captured by the social sciences and economics in particular. In general, one should consider a wider array of scenarios, as well as variation in key parameters, that are at base only guesses.

We suggest that we replace the algebra of delusion, with the narrative of exploration. Future outcomes are never known with any sense of the surety. They are by their very nature an exploration and should be treated that way. If one is projecting a future improvement, it is reasonable to ask the question, how would this improvement come about? Does it fit observed patterns of behavior? Could people be induced to behave this way? How would you do this with incentives, policy, nudge, etc.? Who would be the winners and losers in this game? What would be the cost both financially and politically to make this change? What other changes might also take place? How do these appear compared to your proposed improvement: more likely or less likely? Larger or smaller? Positive or negative? Can you list them? How would these other possible changes interact with your proposed change? Is there any historical evidence to suggest how this might have played out at some previous time?

Table 2. Predicting the benefits of technology

TECHNOLOGY ASSUMED BENEFIT ACTUAL (OR POSSIBLE*) RESULT

ADVANCED COMPOSITES

Can save weight Used to carry more fuel and equipment.

PHOTOVOLTAIC ELECTRICITY

Improvements lower the energy payback time compared to incumbents

PV needs storage and overcapacity for fair comparison.

E- READERS Better than paper *Are quickly

replaced by newer and better technology, leading to short life time and more devices.

RIDESHARING Can displace

individual car ownership

*Increases trips and congestion.

EMAIL Can displace paper People send more

messages and print out email.

IMPROVED FOOTBALL HELMETS

can reduce head trauma

More and younger people play football.

TELEPHONE Can displace auto trips Actually, used to

schedule auto trips.

ADDITIVE MANUFACTURING

can reduce material waste

*Leads to making more stuff, and hence more waste.

LED LIGHT BULBS can reduce energy Leads to new

behaviors, new devices, larger installations, and more energy use.

Of course, to the extent possible these alternatives should be accompanied with numerical estimates. If one tries to follow this scheme, it will become evident what is important in gaining some desired outcome, which in turn, should suggest what types of interventions might or might not help. Ultimately, this product evaluation enterprise is only a preliminary step toward reducing our anthropogenic impact on the planet. It is human behavior that will ultimately decide our fate, and a good first step would be a more balanced assessment of our technological options.

(6)

3.Summary

In this article, we present several examples of how LCA can give misleading results and make suggestions on how to correct this problem. The biggest issue is when LCA outcomes are scaled up and used to represent large boundary (national or global) results in the future. We frame the problem in terms of “missing people”, who are meant to represent social and economic interactions between key actors in the analysis. The solution suggested here is to recognize the missing people in the LCA methodology. Ultimately, this could result in modifications to the LCA tool itself, or integration of LCA with other methodologies, such as in the use of integrated assessment models.

The core issue is that the engineering approach to LCA is to collapse complex social and economic interactions into estimated parameters. In this critical regard, LCA is not like other engineering tools because the boundaries of LCA are vastly larger than typical engineering tools such as finite element methods (FEM) or computational fluid dynamics (CFD). Hence, the LCA methodology uses many parameters which are intended to represent the central value of complex economic and social interactions. This kind of procedure is necessary in order to simplify the calculation and make the whole procedure tractable. However, the simplification can be misused. The essence of this article is to draw attention to instances when these interactions need to be unpacked and examined, not only to understand the complexity of the underlying interactions, but equally important, to appreciate the full range of possible quantitative outcomes.

We argue to replace what we call the algebra of delusion with the narrative of exploration. By the narrative of exploration we mean that crucial assumptions in the quantitative analysis be clearly articulated in a balanced format that acknowledges the significant uncertainty in making predictions about the future.

References

[1] Hauschild, M.Z. 2015. Better – but is it good enough? On the need to consider both eco-efficiency and eco-effectiveness to gauge industrial sustainability. The CIRP Conference on Life Cycle. Procedia CIRP 29, pp. 1-7. Available online at www.sciencedirect.com.

[2] Hauschild, M.Z., C. Herrmann and S. Kara. 2017. An Integrated Framework for Life Cycle Engineering. The 24th CIRP Conference on Life Cycle Engineering. Procedia CIRP 61, pages 2-9. Available online at www.sciencedirect.com.

[3] Kara, S., M.Z. Hauschild, and C. Herrmann. 2018. Target-Driven Life Cycle Engineering: Staying within the Planetary Boundaries. 25th CIRP Life Cycle Engineering (LCE) Conference, April 30-May 2, 2018, Copenhagen, Denmark. Available online at www.sciencedirect.com. [4] Hype Cycle, Wikipedia, accessed Dec 20, 2017

https://en.wikipedia.org/wiki/Hype_cycle.

[5] Mokhtarian P. L. 2002.Telecomunications and Travel – The Case for Complementarity, Journal of Industrial Ecology, Vol 6, Number 2, pp 43-57.

[6] Sellen, J. A. and R. Harper. 2003. The myth of the paperless Office, The MIT Press.

[7] Tsao, J. Y., H.D. Saunders, J.R. Creighton., M.E. Coltrin and J.A. Simmons. 2010. Solid-state lighting: an energy-economics perspective. J. Phys. D. Appl. Phys. 43, 354001, pp. 17.

[8] Tenner, E. 1997. Why Things Bite Back – Technology and the Revenge of Unintended Consequences. Vintage Books.

[9] Cooper, D.R. and T.G. Gutowski. 2018. Prospective environmental analyses of emerging technology: a critique, a proposed methodology, and a case study on incremental sheet forming. J. of Ind. Ecology. DOI:10.1111/jiec.12749.

[10] Schäfer, A., J. B. Heywood, H. D. Jacoby and I. A. Waitz, 2009. Transportation in a Climate-Constrained World, MIT Press.

[11] Gutowski, T.G., J.Y.H. Liow, and D.P Sekulic. 2010. Minimum Exergy Requirements for the Manufacturing of Carbon Nanotubes. IEEE, International Symposium on Sustainable Systems and Technologies, Washing D.C. May 16-19.

[12] Gutowski, T.G., S. Jiang, D. Cooper, G. Corman, M. Hausmann, J.A. Manson, T. Schudeleit, K. Wegner, M. Sabelle,J. Ramos-Grez and D.P. Sekulic. 2017. Note on the Rate and Energy Efficiency Limits for Additive Manufacturing. J. of Ind. Ecology. DOI: 10.1111/jiec.12664. [13] ASTM. 2015. Standard Specification for Structures, F3114-15. [14] MacKenzie, D., S. Zoephf, and J. Heywood. 2014. Determination of

U.S. passenger car weight, Int. J. Vehicle Design, Vol. 65, No. 1, 73-93. [15] Ireland, M., and H. Jeffrey, 2012, Reduction in Aviation Carbon

Dioxide Emissions: Analysis of the 787 Dreamliner and Future Improvements, 2.83 MIT class project report.

[16] Allcott, H. and M. Greenstone. 2012. Is There an Energy Efficiency Gap? Journal of Economic Perspectives – Vol. 26, No. 1, pp. 3-28. [17] Herring, H. and S. Sorrell. 2009. Energy Efficiency and Sustainable

Consumption – The Rebound Effect. Energy, Climate and the Environment Series. Palgrave Macmillan.

[18] Polimeni, J. M., K. Mayumi, M. Giampietro and B. Alcott. 2009. The Jevons Paradox and the Myth of Resource Efficiency Improvements. Earthscan Research Editions.

[19] Dahmus, J.B. 2014. Can Efficiency Improvements Reduce Resource Consumption? A Historical Analysis of Ten Activities. Journal of Industrial Ecology, DOI: 10.1111/jiec.12110.

[20] Magee, C.L. and T.C. Devezas. 2017. Specifying technology and rebound in the IPAT identity. 15th Global Conference on Sustainable Manufacturing. Procedia Manufacturing. Available online at www.sciencedirect.com.

[21] Magee, C.L., and T.C. Devezas. 2016. A simple extension of dematerialization theory: Incorporation of technical progress and the rebound effect. Technology Forecasting & Social Change. http://dx.doi.org/10.1016/j.techfore.2016.12.001.

[22] Smil, V. 2014. Making the Modern World: Materials and Dematerialization. John Wiley and Sons.

[23] Gutowski, T.G., D. Cooper and S. Sahni, 2017. Why we use more materials, Philosophical Transaction of the Royal Society A, A 375: 20160368. http://dx.doi.org/10.1098/rsta.2016.0368.

[24] Harari, Y.N., Sapiens: A Brief History of Humankind. 2015, Harper. [25] Skerlos, S.J. 2015. Promoting effectiveness in sustainable design. The

22nd CIRP conference on Life Cycle Engineering, Proceedings. Published by Elsevier B.V.

[26] Zink, T. and R. Geyer. 2016. There is no such thing as a Green Product. Stanford Social Innovation Review.

[27] Kim, S.J., S. Kara and M. Hauschild. 2017. Functional unit and product functionality – addressing increase in consumption and demand for functionality in sustainability assessment with LCA. Int. J. Life Cycle Assess. No. 22. pp. 1257-1265.

Figure

Table 1. Specification of 767-400ER and 787-8
Table 2. Predicting the benefits of technology

Références

Documents relatifs

Moreover, all emissions were correctly distributed over time according to the temporal parameters defined for the foreground processes and for the ones included in temporal

Goals: 1 to compare the environmental impact of producing a given insect species with conventional sources of animal protein, 2 to assess the environmental impact of

A pack composed of the ideal cell chemistry for any specific application reaches its desired power output at the same weight it reaches its desired energy storage..

The direct beam component of solar radiation (the radiation arriving directly from the sun) normal to a horizontal plane at the earth’s surface, under clear sky and natural

Here, by combining the TGAS catalogue of the Gaia DR1 with LAMOST radial velocities, we show that the position of Hercules in velocity space as a function of radius in the outer

Abstract: Several actors of or related to the European space industry, such as ArianeGroup and the European Space Agency (ESA), have identified life cycle

The analysis of Golgi apparatus dimension in cell lines treated with swainso- nine to block the synthesis of β1,6-branched N-glycans, or in CHO Lec1 cells, which cannot synthesize

Fluoxetine effects assessment on the life cycle of aquatic invertebrates..