Are LCAs Greenwashing Plastic ?

image  Creative Commons 4.0

LCA, Life Cycle Analysis or Assessment, is supposed to be an objective way to compare the environmental footprint of products, and is a mainstay of corporate decision-making in sustainability.  But it’s blind to plastic pollution, leaving it available to be mis-used in comparisons of plastic with other materials.  LCA-based comparisons of plastic bags with other bags for example have been widely cited and give a misleading impression that plastic is ‘greener’, while not assessing plastic as a pollutant at all.

Characterisation of plastic pollution is complex and a relatively new topic but recent work from Germany’s Fraunhofer Institute may enable development of a standard measure of plastic as a persistent and potentially bioaccumulative pollutant.  Meanwhile, campaigners, journalists and environmentalists, as well as scientists who may be commissioned to carry out LCA studies, should be alert to the risk of ‘greenwashing’ plastic through conventional LCAs.  The initial wave of concern heightened by Blue Planet II has subsided but the plastics industry’s fight to rehabilitate itself continues, and with essential uses for PPE at the forefront of covid responses and rock-bottom prices for oil, virgin plastic is cheap and the recycling market has collapsed in many places.  The flood of plastic pollution shows little sign of abating anytime soon.

 

Introduction

 

You may have noticed that despite various ‘bans’ on plastic bags there are quite a lot of media and social media stories in which plastic is compared with alternative materials and plastic is found to be ‘greener’, for example plastic bags compared to cotton bags, or plastic compared to glass or steel bottles.

Track back to the sources of these stories and you usually find they result from a LCA or Life Cycle Analysis or Assessment (for example carrier bags / bottles).  The out-take from these stories generally gets condensed to ‘plastic not so bad after all’, or glass/ paper/ cotton/ aluminium (etc) is ‘actually worse for the environment than plastic – says study’.

BBC World video including a much repeated ‘fact’ that a cotton bag needs to be used 131 times “to have the same environmental impact” as a plastic bag.

Press coverage, even from the BBC’s ‘reality check’ unit, tends to convert assessments into a single dimension such as ‘greener’ or ‘the environmental impact’.  The original factors used for the assessment are often shorn away in the telling of the story, and as it moves along the media chain from the study to the out-take, and through social media. 

(This BBC report cites a LCA report by the Environment Agency for England and Wales on bags available in 2006, published in 2011.  The ‘answer comes down to’ because those were  criteria put into the analysis).

LCAs are widely used in industry and business to try and standardise environmental and sometimes social comparisons between material options, from ‘cradle to grave’.  Some LCAs are bespoke, invented for a particular purpose but many rely on using or adapting an off-the shelf methodology which of course helps with comparability.  LCAs (review article) are required to comply with ISO 14040 guidelines which specify four main stages.

On the face of it, a comparison based on a LCA, with its quantification and set methodology, seems more objective and authoritative than other ways of making a decision.  As they are detailed, and cumbersome and demanding to conduct, they come with an aura of expertise as well as dependability.

But at the moment there is a serious problem in trying to use them to compare plastic with other materials, which at its most basic, is that plastic pollution is invisible to most if not all standard LCAs.

Here for example are the categories used in the EN15804 (a standard for LCAs in the construction sector):

EN15804 – a European standard

If a whale dies from the mechanical obstruction of its organs by ingesting plastic for example, then no matter how many times that is reported in the scientific literature, there is no place for it in most of the impact inventories used to conduct ‘Impact Assessments’ in LCAs. The same would go for seabird or turtle entanglement, or starvation due to ingesting plastic due to mistaking it for food.  If they were poisoned by toxic chemicals released by the plastic that might count but there would need to have commissioned a specific study to produce input data, or that factor would need to appear in a standard reference database such as Ecoinvent.

The obvious large items ‘macro’ and the smaller flakes of ‘meso’ plastic whose impacts created public concern and a little political action through campaigns and programmes like David Attenborough’s Blue Planet II, are simply dismissed as a ‘littering’ problem and not captured in most LCAs.  But there’s more to it than that because ‘microplastic’ is not included either.

In a 2017 blog I argued that policy-makers should treat plastic as a persistent organic pollutant, and regulate for a phase-out except for essential uses.  Several groups of scientists had already made similar calls, pointing to it’s effectively indefinite lifetime, its accumulation in the environment, its roles as a source and vector of toxic substances, and its un-nerving capacity to almost endlessly fragment in the environment.  Since then a growing number of studies have confirmed the omnipresent nature of plastic pollution and it’s ability to travel throughout human and animal bodies, and it’s as yet not-well-defined potential to cause disruption to important cellular mechanisms, suggesting that like forms of radiation, it may have no ‘safe level’ of exposure.

With the possible exception of measures of toxicity derived from substances directly leaching from plastic (freshwater, marine and human health toxicity do routinely feature in LCAs) most of the features of plastic pollution are not accounted for.  So because other factors are included, such as embedded energy costs and emissions like CO2, plastic often looks environmentally better than alternatives.  A glass bottle for example,  will usually have a bigger carbon footprint than a plastic one of the same volume.  Not the same plastic pollution footprint of course but LCAs don’t count the plastic footprint.

As Julien Boucher and colleagues put it in the 2020 IUCN report The Marine Plastic Footprint (p4): 

‘a challenge with LCA methodologies is that they do not account for plastic as a pollutant, but rather only for the indirect effects of plastic use, e.g. depletion of resources, energy consumption, or emission of chemical contaminants. LCA methodologies neither provide an inventory of the marine plastic leakage nor characterise factors to assess the impacts of plastics on ecosystems or human health. This lack of appropriate accounting of plastic leakage has encouraged companies to massively favour plastic packaging in many situations, due to its lightweight nature and low carbon requirements’

Slides from a 2018 presentation by Julien Boucher

The Environment Agency (EA) study cited by the BBC in 2019 for instance (report above) considered Global Warming Potential and ‘other impacts: resource depletion, acidification, eutrophication, human toxicity, fresh water aquatic ecotoxicity, marine aquatic ecotoxicity, terrestrial ecotoxicity and photochemical oxidation (smog formation)’ but not the creation, accumulation and effects of plastic as a pollutant.  ‘End of life’ waste management options such as recycling were included but not ‘the effects of littering’ and ‘discharges to water and soil’ were ‘outside the system boundary’. (Studies which identified littering effects were mentioned in an annexe but played no part in the assessment).

The highlighted finding of the study only focused on climate impact and was featured in the only graphic included in the executive summary:

 

Not surprisingly the point that registered with journalists was the idea that for re-usable cotton bags to have the same environmental impact as ‘disposable’ plastic carrier bags, they would need to be re-used 131 times. 

The study drew on 2006 data and was published in 2011 before the 2017 peak of public concern (see this previous blog) but factoids from such LCA studies are constantly recirculated and most if not all current LCA methodologies still fail to register plastic as a pollutant.

Last October the Dutch-based campaigning NGO Plastic Soup Foundation launched an attack on LCAs, declaring ‘The plastics industry abuses lifecycle analysis (LCA) in communication surrounding plastic pollution’.  It pointed out that LCAs often do not take the end-of-life consequences of a product into account and may make optimistic assumptions about recovery and recycling.  It criticised a recent Dutch industry campaign Plastic Truth versus Plastic Fable for using LCAs as the basis of a claim that plastic bags were more environmentally friendly.

The Foundation noted that on the one hand, LCAs regularly rate plastic as more environmentally friendly while on the other, 80% of plastic ends up as waste in the environment, causing immense harm to marine wildlife. It stated:

‘Industry must, therefore, stop using the current LCA method for promoting single-use packaging plastic in particular. In the meantime, a legitimate supplementary criterion that takes into account the impact plastic has once it inevitably reaches the environment should be agreed upon’.     

A Solution For Plastics in LCAs ?

No such measure has been agreed upon, although in 2019 researchers from the Norwegian University of Science and Technology and the University of Leiden proposed an ‘entanglement factor’.

McHardy et al 2019 proposing a LCA entanglement factor

In 2019 one study took a Spanish LCA assessing different bags and showed that when a ‘pragmatic littering indicator’ was introduced, it produced ‘precisely the opposite’ ranking to when GWP (climate impact) was the main criterion.

From Civancik-Uslu et al 2019 – a littering indicator

More recently, Daniel Maga reported on the work he and colleagues at the Fraunhofer Institute (the German institute for applied science) are undertaking to systematically characterise and quantify the risks attendant on plastic as a pollutant.  Speaking at a virtual conference of Setac in May 2020, Maga gave a presentation on a research methodology for plastic emissions available here in video form.

From Daniel Maga presentation (video) at Setac May 2020

Maga points out that current end of life modelling as used in LCAs does not consider littering or loss of plastics through abrasion or weathering.  He proposes a characterization factor combining fate, exposure, effect and severity and asks how the risks of plastic emissions can be captured in a LCA.  Maga cites the ECHA (European Chemicals Agency) which asserts that there is limited evidence of environmental risks from microplastics and no suggestion of bioaccumulation of hydrophobic organic compounds in organisms (eg many pesticides, PCBs) but also that conventional risk factors may not work with micro or nano plastic risks. He proposes that the ‘fate factor’ is sufficient to capture the main risk from plastics due to their extreme persistence, so they should be treated as a ‘non-threshold substance’ in a similar way to PBTs (persistent, bio accumulative and toxic substances) for which any release can be assumed to create a risk. This is a classic case in which the Precautionary Principle should be applied (there being a priori reason to act even without definitive evidence of the impact having already occurred).

Detailed characterisation of plastic pollution is a formidable task.   Maga goes on to itemise a mind-numbing welter of technical challenges.  To deal with these in a way that could be included in a LCA, he proposes calculation of plastic equivalents, homing in on a SDR or ‘Specific Surface Degradation Rate’ measure, published in February.   (How micro-plastics degrade is strongly dependent on shape – watch the presentation for details)

Where’s this going? If used, such an approach would enable decision-makers such as government agencies to compare and regulate plastics according to their risk as driven by fate-factors such as persistence.

Maga says “we imagine” it generating tables such as this one for macroplastic emissions (based on estimates from the plastikbudget project in Germany), showing rates of loss, where they are (environmental ‘compartments’ such as soil or water), type of plastic, degradation rates, SDR, length and shape:

And the same for microplastics

If a system like this were to be adopted by decision-makers, it could enable fairer and more realistic LCA comparisons, and help prioritise regulatory action as well as choices within companies.  In theory possible that a country like the UK or more realistically the US, could do this alone but both are seriously weakened in terms of capability by a decade or more of environmental back-pedalling and hollowing out of expertise in central government and agencies.   The EU, probably led by the economic and scientific powerhouse of Germany, is probably the main hope for scientific R & D in tacking the plastics crisis.

Beware The LCA

In the meantime, environmental correspondents, NGOs and campaigners need to be wary of LCAs.  They should check the methodologies behind any claims to compare the ‘green-ness’ or ‘environmental impact’ of plastics and alternative materials, particularly where data has been fed into the blender of a LCA to give a ‘simple’ result.

Although it may be a big ask for researchers hungry for money, scientists asked to conduct similar studies should also ask themselves why they are being commissioned, and whether the framing that a brief will create, is designed in advance to greenwash plastic by what is included or excluded.

Cognitive Biases

Researchers should also be careful about the way findings are converted into everyday  terms. One example of framing and possible inbuilt cognitive bias is use of the ‘number of times you’d have to use a bag’ metric.  This may seem innocuous but it positions a large number against a small number by taking as its ‘impact’ reference point the ‘footprint’ of a single plastic bag (excluding the plastic pollution impact), and comparing it to a different type of bag (eg cotton) and then working out how many times the for example cotton bag ‘would need to be used’ to ‘be as good as’ the plastic bag.  The UK EA 2006/2011 study mentioned earlier calculated 131 times and the 2018 Danish study found (p 80) ‘conventional cotton carrier bags should be reused at least 50 times before being disposed of; organic cotton carrier bags should be reused 150 times based on their environmental production cost’.

What this does is to frame the consumer choice in terms of effort required in order to get the reward of being environmentally friendly.  Either use a disposable plastic bag, or you have to go shopping with a cotton bag 130-150 times!  Can you imagine – (your brain does that instantly without you thinking it through) – it’s just not feasible is it?  This invokes the anchoring heuristic: the busy (time-pressed, time-is-a-scarce-commodity) shopper is given the first single-action choice as a reference point (the anchor) and the second (131x) action) to compare to it, just in order to get the same reward.  It’s a no-brainer that plastic is the more feasible choice.

The focal effect in this snap judgement also stops us asking awkward questions like “how many times can you actually re-use a cotton bag anyway ?”  I have no idea but some suggest durable bags can be re-used 500 times. Think of other items made out of cotton (requires a frame-change).  Would you be satisfied with a cotton shirt or pair of trousers than could not be worn for at least 131 days, and would it ‘make more sense’ to replace them with 131 pairs of disposable plastic trousers if they had a lower carbon footprint?

Even setting aside the omission of the very reason plastic needs to be assessed – plastic pollution – and any other doubts that might exist about the assumptions made about non-plastic choices in these studies, the cognitive behavioural bias of the framing is clear.  Try thinking about it the other way around.  “What’s better, to use just one cotton bag for your daily shopping over the next four months, or 131 plastic bags which you then throw away?”  ‘Obviously’ the re-usable bag is an environmentally better choice.  Now the effort implication of the cotton bag has disappeared and you are triggered to think about environmental responsibility by being reminded that these are ‘disposable’ short-lifetime plastic bags.  Picture that huge pile of waste plastic bags.

In a fleeting mental encounter with such a study factoid, as in watching a 30 second news clip, another mental bias is triggered, which is WYSIATI ‘what you see is all there is’.  The ‘issue’ captured is the story is the choice between two bags.

Other possible redesign options or behaviours, sometimes mentioned in the fine-print of LCA studies, are not shown and thus do not exist in the mental processing.  Using a wheeled shopping basket for example (no bags needed), or a rucksack you already own and also use for other purposes, or re-using a carboard box the store provides, or a host of other possible options that might resolve the ‘bag problem’.

The focus is kept restricted to the plastic bag v other bag, as in a ‘horse-race poll’ in politics where prospective voters are only asked about the ‘two main candidates’ and not the five others running, so they disappear from view in stories reporting the result of the poll, triggering voters to make an instant mental choice between the two candidates featured, not the total seven.

That device is popular with larger political parties. Restricting the focus and terms of a LCA-based study can enable those with a vested commercial interest in say, plastic, to generate seemingly scientific, impartial and ‘objective’ findings that happen to show their product in a good light.

‘Gold Standard’

No wonder perhaps that the plastics industry loves LCAs as they stand at the moment.    The British Plastics Federation website states: ‘They are as close to the gold standard of understanding the environmental consequences of a product as researchers can currently get’.

Imagine you are in the driving seat of the ‘public affairs’ strategy for the plastics industry, or its Siamese Twin the oil and gas industry.  How would your situation report go? Maybe something like this:

**

Covid has brought mixed blessings but generally things still look good. The vast majority is still made from oil or gas and despite the response to Blue Planet II, plastics use worldwide is still increasing and rapidly risingin Europe.  Pressure to cut carbon emissions is a problem but it can be turned to advantage if plastic can be positioned as greener than alternatives in energy terms. Unfortunately one side-effect of Blue Planet II and the associated wave of campaigning was the introduction of bans and restrictions on high profile ‘single use plastic’ such as bags, starting in Europe and spreading around the world.  

A 2019 European Environment Agency review of the measures taken by European countries to reduce the plastics problem found revealed that 37 of the 173 measures identified were market based and most ‘referred to fees for plastic carrier bags’.   We can live with those but the real risk is if the same political thinking behind them (respond to popular sentiment against plastic) spreads to the rest of the packaging market and other uses. We particularly need to keep politicians thinking bout plastic as an issue that can be solved by better waste management and more consumer commitment to recycling, and not ‘phase-outs’.  This is why it’s even now important to discredit the ‘bag bans’ as an irrational and regrettable reflex, not borne out by ‘the science’.

Some good news is that the more scientific end of the media covering environmental issues is particularly motivated to encourage ‘rational’ rather than ‘emotional’ environmentalism, which means they have an appetite for the quantified and factual.  Here LCAs are our friend as they generate factual proofs that plastic can be the better environmental choice. 

Take for example the British popular science publication New Scientist. In 2015 it carried an article ‘Plastic Bag Levy Is A Drop in the Ocean On Environmental Grounds’:

Its life cycle analysis of a number of different types of shopping bags found that a cotton bag would have to be used 131 times to be below the total global warming potential of an HDPE bag used only once. And once you factor in reuse of HDPE bags as bin liners, which is reasonably common, this reuse factor rises. The point made by the study is that the global warming impact of HDPE bags is negligible

It’s good to be able to report that New Scientist went on to cite that 131-times fact at least twice more in 2018,  in an article advising its readers (May 2018):

and in a Leader (June 2018):

The LCA 131-times example has been repeated many times, for example here in The Conversation, here in Stamford Magazine, here in Earth Times, and here in Business Insider, all channels likely to reach this ‘Rational-Environmental’ audience.  There is no way we can make plastic popular with all the public but we don’t need to – to paraphrase Frank Luntz on climate, we just need to maintain doubt about alternatives, while we keep on growing.

**

That’s enough of the imaginary report.  I freely admit that my view of the activities of the plastics industry is somewhat jaundiced as a result of seeing some of its lobbying activities and the way, for example, that it used the ‘litter’ frame from the 1970s onwards (see A Beautiful But Evil Strategy) to prevent people seeing plastic as a pollutant, and still uses today.  It’s in the industry’s interests to see these misleading ‘factoids’ derived from LCAs in wide circulation, spreading like memes.

Even better than having them repeated in a ‘straight’ science magazine like the New Scientist is if the pro-plastic LCA findings are endorsed by re-appearing in publications from environmental organisations themselves.   This can happen when ‘green’ groups are engaged in projects framed by environmental waste ‘management’ assumptions, and are trying to optimise choices from inside the status quo, rather than to change the strategic drivers.  The ‘circular economy’ community is particularly vulnerable.

For example in 2020 the UK group the Green Alliance published a report Plastic Promises: What The Grocery Sector Is Really Doing About Packaging for its Circular Economy Task Force  whose members are the corporates PwC, Kingfisher, Viridor, Walgreens Boots Alliance, SUEZ recycling and recovery UK, and Veolia.   It noted:

Given the demand for change since the BBC’s Blue Planet II aired in 2017, and the promises that have been made since, one might have expected a considerable market shift away from plastic by now, at least for packaging in the grocery sector.  There have been some minor changes … but, overall, the proportion of plastic packaging seen on most supermarket shelves, and the amount collected as waste and reported to the Environment Agency, has not altered significantly.

Based on anonymous interviews within the sector, author Libby Peake reported that despite one of the supermarkets reporting a ‘ferocious’ anti-plastic response from consumers with an 800% increase in customer queries, actual behaviour change was limited. Professionals such as brand managers cited a host of frustrations, from dubious brand claims to unintended consequences of switches, inadequate recycling (Britain is a mess) and difficulties sourcing recycled material.

The Green Alliance study was no doubt conducted with the best of intentions but the conventional plastics-dominated packaging industry will have been delighted that it also  repeated as fact the LCA-based findings from ‘a 2011 study for the Northern Ireland Assembly … that paper bags generally require four times as much energy to manufacture as plastic bags’ and the Danish study which ‘concluded that … a paper bag would need to be reused 43 times to have a lower impact than the average plastic bag’.

The plastics industry PRs will also have appreciated the headline “we can’t just replace plastics” quoting Libby Peake in an interview about the study with the Packaging Europe (used in its weekly newsletter).

 

A Question of Strategy

 

The real issue for change groups in relation to LCAs is one of strategy: what are you trying to achieve and by what steps will that come about?   The above quote from the packaging industry magazine asserts that climate change is ‘an even more serious problem’ than ‘plastic’.   In what sense?  What does this mean?  From whose point of view?

As Green Alliance’s report records, most people (public in the UK) in fact reject the implied trade-off and think that climate and plastic are of equal importance. If you asked ‘experts’, then with all sorts of caveats, they would probably give you a similar answer.  But if you asked about a specific case, as in for example choices about bags or bottles, you might get a different answer depending on people’s understanding and how they assume change can happen on either ‘issue’.

In the end a LCA essentially constructs a two dimensional rating by adding up the results of a set of scores to enable rankings from best to worst.   In my view, change strategy has to be at least three dimensional.  One tool I developed for doing this in relation to potential campaign targets is the ‘ambition box’.  It has three axes, the hardness or difficulty of a change target,  the size of that target (how much of the problem it represents), and the significance of the target (the consequential effects or potentiation resulting from the achievement of the target).

Ambition Box from How To Win Campaigns edn 2

Problem management logic is mainly dictated by the first two axes.  For example it would make sense to start with the ‘lowest hanging fruit’, the biggest soft and easy target first.  Strategic change logic is mainly dictated by significance.  LCA is fundamentally a tool for problem management (eg making optimal choices at one level which are all sub-optimal options in the ‘bigger picture’), not strategic change.  An exception, as has been argued above, is that it can be mis-used to obstruct strategic change.

Put this another way: with a wicked global problem like climate change or plastic pollution, we need a strategy to ultimately eliminate the problem not just manage it.  This is why campaign groups and now most governments are trying to eliminate fossil fuels from the energy system, not just increase energy efficiency, and why human-made industrial greenhouse gases like HFCs need to be simply phased out completely and replaced with alternatives, not just reduced to a particular level or ‘so far as possible’.

Plastic made from oil or gas is important as a direct contributor to climate change (the carbon gets released adding to environmental CO2) but its not as important as getting to zero carbon energy across industry, transport or electricity generation, which would ‘deal with’ all the energy related ‘carbon problem’ behind production of all sorts of bags, bottles and so on.  So the carbon footprint of plastic is not necessarily an ‘even more important’ problem if looking at say packaging, than plastic pollution is.  On the other hand, packaging certainly is the major source of plastic pollution, along with tyre wear.  So for plastic pollution these are strategic targets, requiring substitution, or ‘replacing plastic’ within a regulated phase-out, excepting essential uses.

Naive Rationalists

LCAs are simply not set up to make (or even really facilitate) such judgements but this may not be widely understood by many of those who use them on a regular basis.  Others may be ‘naive rationalists’, naive about the way LCAs are easily mis-used to ‘game the system’ and attracted to what seems a ‘rational calculus’ defining ‘the right answer’.

This is what Michael Warhurst, a UK chemicals expert and Executive Director of ChemTrust told me about LCAs:

LCAs are a nightmare, as it is easy to get the answer you want & ultimately the data & assumptions that they are based on are very poor. On chemicals, for example, they use old databases and assumptions while REACH is constantly identifying new problems & also finding that know problem chemicals are active at lower levels.

I think for an effective assessment you have to disaggregate different elements & create a system that is as transparent as possible…

I don’t think LCAs will ever be satisfactory, but they are popular as an apparent ’simple’ solution – they are fine to use within an organisation if you know what you want and are comparing options, but they are terrible for policymaking in a wider sense”.

Of course there are many possible dimensions of significance.  One of the psychological-political ones is public resonance and iconography.  Plastic bags and bottles now fall into this category, which is one reason why politicians took some sort of action on those, and why old LCA-factoids on bags and bottles keep being put back into the public conversation.

Some of those involved with LCAs rightly point out that they were not supposed to be used to form policy but when their results derive from assessments which manifestly fail to capture key environmental impacts, and are designed in a way that gives a stamp of green approval to plastic, and these are put into the public domain, they can of course affect politics and policy.

Finally, a question for those pursuing the ‘Circular Economy’ is what are the steps by which it can actually be brought about, and are any of these strategic ?  If not, you may remain trapped inside a universe of many small sub-optimal choices which you are trying to use to change drivers that are being set by the strategies of others, such as the plastics and oil industries.

 

Share
This entry was posted in Uncategorized. Bookmark the permalink.

1 Response to Are LCAs Greenwashing Plastic ?

  1. Kim Christiansen says:

    LCA is a tool to support decision making, not to take over decision making. In the decision making you shall take into account the limitations of LCA e.g. attributional versus consequential, process based and/or input-output based, data quality issues etc. And that some impacts are not yet modelled or not modelled to the same level of scientific consensus. Combine process data with I/O data (hybrid) and calculate both from an attributional and a consequential modelling perspective (e.g. use EcoInvent with both options) and then explain the commissioner why the results are different. And remember to report the uncertainty in the study e.g. by not reporting specific numbers but the variation (spread). Discarding LCA as a tool would be the same as discarding risk analysis or environmental impact assessments or any other tool – they can all be manipulated. This is not specific for LCA. If you don’t allow any quantitative decision support that is of course a choice you can take. But not mine.

Leave a Reply to Kim Christiansen Cancel reply

Your email address will not be published. Required fields are marked *