Super Size EV Automotive's obesity crisis

The automobile has given society a major benefit in mobility, at the cost of an environmental impact which is now seen as unsustainable.  There are now around 1.5 billion cars on roads around the world, coming in many shapes and sizes, and being put to a panoply of uses.  While pollutant emissions have to a significant extent been solved on the latest internal combustion engined vehicles in developed economies, this is not true of the main global warming gas, carbon dioxide (CO2).  To address this, battery electric vehicles (BEVs) are being pushed by a particularly vocal group of influencers, as if they are a panacea.  Joining this call are governments from the world, which may make their takeover inevitable.

But: when the problem is the environment impact of mass mobility, we need to ask ourselves whether encouraging vehicles that are on average 39% heavier, 40% bigger and 44% more expensive to buy than normal gasoline internal combustion engine (ICE) cars is really the immediate answer.

The size issue is not confined to BEVs.  Gasoline SUVs, of which there are over 200 million in the world, are 27% heavier than smaller, gasoline hatchbacks, as shown in the graph below.  This shows how the kerb weight – i.e. the vehicle empty but with a full tank of fuel – averaged over all new vehicles on sale in Europe has changed over time.  In our last newsletter, Gaining Traction, Losing Tread, we set out the issue of tyre wear emissions, which shows that weight matters for environmental pollution.  The excess weight of BEVs is likely to lead to 8.6 mg/km of additional tyre particulate mass emissions, and gasoline SUVs an excess over hatchbacks of 5.3 mg/km, both above the total maximum permitted total tailpipe particulate emissions of 4.5 mg/km.

Looking at the trend in average vehicle weight in different categories over the last twelve years, collated from manufacturer data, we can see the clear rising trend for BEVs, as large SUVs have come to dominate a category that started with smaller hatchbacks and saloons.  The weight of gasoline SUVs has stayed relatively constant, but market share has shifted in their direction.

Put in a longer historical context, European vehicles became almost 500 kg lighter in the decade after 1997 as small city cars proliferated and, further back, US cars became 400 kg lighter in the five years from the 1973 oil crisis.  Therefore, a growing waistline is not predestined.

So, why are vehicles getting bigger and heavier?  There are three reinforcing trends.  First, consumer preferences appear to be shifting in favour of large, high-riding vehicles, with Europe catching up with earlier trends in the US.  Second, more safety equipment and designs have been deployed in response to legislation and consumer pressure.  The third factor is regulation, which has, perhaps unintentionally, accelerated the trend.  In the US, larger pick-up trucks are exempt from the corporate average fuel economy rules, and there is a ‘footprint’ adjustment that gives easier targets for physically larger vehicles.  In Europe, there is a similar adjustment to the CO2 targets, giving easier targets to heavier vehicles.  More generally, the deployment of more exhaust after-treatment technology to meet ever-tighter pollutant emissions regulation has required more physical space on the vehicle and higher prices to maintain profit margins.  In simple terms, CO2 emissions have been increased to serve consumer preference and air quality gains.

Along with the added weight of BEVs and gasoline SUVs comes greater physical size.  The average BEV from the last five years is 44 cm longer, 9 cm wider and 33 cm higher than the average gasoline vehicle.  This equates to an excess volume of 40%.  This has positives in terms of consumer utility where it leads to greater interior space and better access, but may come with downsides in terms of on-road practicality and safety.  According to an estimate from the US, if you are hit by a vehicle 450 kg heavier than yours, the risk of death is 47% greater than if the vehicles were of the same weight.

With this extra size and weight typically comes higher CO2 emissions, whether that is from the manufacture of the vehicle or during usage.  Plotting the real-world tailpipe CO2 emissions against weight, from Emissions Analytics’ EQUA test programme of hundreds of vehicles using Portable Emissions Measurement Systems since 2011, we see a positive correlation: each extra 100 kg of vehicle mass adds on average 13.9 g/km to in-use exhaust emissions, as shown in the chart below.  This is likely to underestimate the relationship as vehicles have become more fuel efficient, like-for-like, over time.  If weight is correlated to the official CO2 values over just the last five years, an extra 100 kg of mass adds 24.6 g/km.

If we add in estimates for upstream CO2 emissions from the vehicle manufacture, fuel refining and electricity production – using Emissions Analytics’ proprietary lifecycle analysis (LCA) model – we can see the total distance-specific emissions over a vehicle lifetime in the table below.

For further details on the model and methods, please refer back to our previous newsletter: Schrödinger’s Car.

The table shows, on these assumptions, that BEVs deliver the largest reduction in CO2 – roughly halving the emissions of the equivalent ICE vehicle.  However, if we were to revert to the previously most popular bodystyle, the hatchback, and applied the latest hybridisation, using less than 2 kWh of battery, battery electric SUVs are 19% lower in CO2 emissions, due primarily to the difference in vehicle mass.

But this is not the end of the story, as with greater size and weight also comes higher purchase prices.  In 2022, the average list price of BEVs on sale in the UK is £37,003, compared to £25,771 for gasoline ICEs.  Full hybrids (FHEVs) – those with the capability of some driving on battery only but without the possibility of external charging – cost on average £25,745, but the hatchback subset of these comes in at just £17,577.

The risk, therefore, is that, without significant taxpayer subsidy, consumers will hold back from buying BEVs.  This is perhaps being seen for the first time in the softening market share figures for BEVs in the first quarter of 2022 in the UK and Europe.  This in turn may lead to consumers buying new ICE vehicles, but many will be put off by the threat to residual values and utility of looming statutory bans.  Instead, many may simply opt to hold onto their existing ICE vehicle or buy second-hand, keeping them on the road longer than normal.  

Currently, 14 years is the average lifespan of a vehicle in Europe.  If this were to increase due to the high price of BEVs, it would covertly undermine the CO2 reduction of BEVs until all the ICE vehicles were off the road.  Taking the UK car parc of around 32.5 million vehicles and a typical number of new car transactions of around 2.1 million per year, we can model three scenarios: reverting to 100% gasoline ICE vehicles, shifting all new cars to BEVs by 2030, and shifting exclusively to FHEVs at the same rate.

Although current new car sales are running at 1.6 million per year due to a combination of the economic environment and supply chain disruption, we have assumed a reversion to the usual level of 2.1 million per year soon.  We can assume that a switch to FHEVs would not affect the quantity demanded as the average price is similar to an ICE vehicle.  The question is the degree to which, other things being equal, the quantity of BEVs demanded would be affected by their higher relative price.  According to recent research from Norway, the price elasticity of demand for BEVs is -1.27, i.e. demand falls by 1.27% for every 1% increase in price.  If we further assume that 41% of car buyers are actively considering a BEV in the next five years, this would imply a reduction in sales of up to 480,000 per year due to the price premium.  In other words, sales could remain subdued at around 1.6 million per year rather than recovering.  Countering this argument would be the valid point that BEVs are cheaper to operate, although consumers tend heavily to discount future benefits.

With a lower level of new car sales, but assuming no reduction in demand for car usage, the average lifespan of a vehicle would likely increase.  According to Emissions Analytics’ model, that increase would be from 14 to 17 years on average.  On a lifecycle basis through to 2050, the effective fleet average CO2 emission on the base ICE scenario is 223 g/km.  On the BEV scenario, this falls to 156 g/km.  The FHEV scenario would be 181 g/km, but if these FHEVs were exclusively hatchbacks it would be just 158 g/km on average.  These values lead to the cumulative CO2 figures through to 2050 in the chart above.  Therefore, given the current realities of BEV pricing, and accepting the compromise in utility from a hatchback, a strategy of hybridisation is just as good as an immediate BEV migration, but with the benefit of needing less taxpayer support.

This analysis of course takes the specifications and prices of current vehicles.  The only future forecast used is around the forecast decarbonisation of the electricity grid.  If BEVs and SUVs become relatively lighter or cheaper over time, or the CO2 intensity of manufacture or energy production falls, different conclusions would be drawn.  Therefore, for BEVs to be the optimal way forward that is often proposed, focus needs to be on addressing these factors affecting CO2, rather than subsidising the uptake right now.  Finally, we must also recognise the on-going CO2 reduction benefits of BEVs after 2050 compared to FHEVs, which holds so long as battery durability matches the current lifespan of vehicles, which was discussed in our previous newsletter, Why battery durability matters for decarbonisation.

All in all, for the next decade, is it not better to incentivise smaller, hybridised vehicles as a more efficient, lower-risk and intuitive solution to CO2 emissions?  Surely urban environments would be more pleasant with small, light vehicles moving around than the “electric tanks” quoted by Neil Winton in Forbes.  Surely, this would also help preserve the affordability of vehicles for a wide population?  Should vehicle weight even be regulated, as a good proxy for low emissions and high CO2-reduction efficiency?  Such an approach may even be a way to reinvigorate EuroNCAP crash safety ratings that according to some now focus so much on the presence of driver safety aids of debatable efficacy.

It is true that such an approach would be crimping people’s liberty to some extent, by incentivising people to forgo the trappings of big SUVs.  However, all directions in environmental improvement require some compromise, and this may be a smaller compromise than the available alternative options.  This would be a dramatically different, and potentially much more effective, route to environmental improvement than the combination of fleet average CO2 targets and the proposed Euro 7 emissions regulations that seem destined to super-size the car parc in an unhealthy.

---

Gaining traction, losing tread Pollution from tire wear now 1,850 times worse than exhaust emissions

By some distance, the research Emissions Analytics published in early 2020 claiming that tire particulate wear emissions were 1,000 times worse than exhaust emissions generated the most feedback of any subject we have tackled so far – feedback that was a mixture of surprise and scepticism.

Of particular attention was whether such a rate of wear would mean that any tire could be spent within just a few thousand miles with legal driving, however aggressive.  Particularly vocal were the battery electric vehicle (BEV) community, sensitive to any suggestion that the added weight of these vehicles might lead to tire wear emissions that might confound the ‘zero emissions’ tag.  Such was the reaction, the story was translated into over 40 languages worldwide.

Since that study, which was transparently designed to quantify the worst-case tire emissions under legal driving, Emissions Analytics has been testing and analysing tire wear emissions in more detail across a wider range of driving conditions, and has performed a detailed chemical analysis of hundreds of new tires.  Furthermore, we have worked with the National Physical Laboratory in the UK objectively to quantify the uncertainties in our measurements of chemical composition.

The headline conclusion we draw now is that, comparing real-world tailpipe particulate mass emissions to tire wear emissions, both in ‘normal’ driving, the latter is actually around 1,850 times greater than the former.  Yes, in normal driving the ratio is almost double the previous figure for aggressive driving.

Quoting such ratios, however, needs careful interpretation.  The fundamental trends that drive this ratio are: tailpipe particulate emissions are much lower on new cars, and tire wear emissions increase with vehicle mass and aggressiveness of driving style.  Tailpipe emissions are falling over time, as exhaust filters become more efficient and with the prospect of extending the measurement of particulates under the potential future Euro 7 regulation, while tire wear emissions are rising as vehicles become heavier and added power and torque is placed at the driver’s disposal.  On current trends, the ratio may well continue to increase.  

To measure tire wear mass emissions, Emissions Analytics uses high-precision scales to weigh all four wheels – tires and rims together, without detaching – over at least 1,000 miles on real roads.  This is coupled with a proprietary sampling system that collects particles at a fixed point immediately behind each tire, which are, via a sample line, drawn into a real-time detector measuring the size of distribution of particles by mass and number.  Typically, this measures particles from 10 microns down to 6 nanometres.  This combination allows the real-time signal to be calibrated to the mass loss, and, by using the size distribution, the proportion of the particles likely to be suspended in the air can be estimated.  All tire emissions figures quoted here are for the whole vehicle, combining wear from the four tires.

Tailpipe particles are measured, in similar real-world conditions, using a diffusion charger analyser for dynamic mass concentration and a condensing particle counter for number concentration, both coupled with a standard Portable Emissions Measurement System (PEMS) to measure total exhaust flow.  As a result, distance-specific mass and number emissions can be derived, which can then be compared to equivalent tire metrics.  A summary of the results is shown in the table and chart below.

The comparison is best illustrated by way of a bar chart with a logarithmic vertical scale, as shown below.

The aggressive legal driving is the result from 2020, which was derived from a Volkswagen Golf driven at legal road speeds on a track, with fast cornering and maximum permitted payload in the vehicle.  The normal driving results were averaged across 14 different brands of tire tested on a Mercedes C-Class driven with average dynamics on the road, with just the driver and no payload in the vehicle.  All of these tires were tested from new.  A smaller number of tires were then tested over their full lifetime in order to estimate the degree to which wear rates declined over time.

The tailpipe real-world emissions were measured across four gasoline mid-sized sports utility vehicles from 2019 and 2020 model years, driven on a mix of urban and motorway routes.  The payload was the driver and test equipment only.  For these vehicles, the relevant regulated limit value in Europe – which is the same limit in force for current vehicles – was 4.5 mg/km for mass and 6.0 x 10¹¹ #/km for number.

Quite remarkably, but as testament to the filtration efficiency of the latest gasoline particulate filters (GPFs), tailpipe mass emissions are now as low as 0.02 mg/km.  Gasoline vehicles were tested as they represent the majority of new passenger cars sold today.  Therefore, the mass wear from new tires is 16 times greater than the maximum permitted from the tailpipe, but 3,650 times greater than actual tailpipe emissions.  Taking the full-life average tire emissions, that premium falls to the 1,850 times mentioned earlier.  The excess emissions under aggressive driving should alert us to a risk with BEVs: greater vehicle mass and torque delivered can lead to rapidly increasing tire particulate emissions.  Half a tonne of battery weight can result in tire emissions that are almost 400 more times greater than real-world tailpipe emissions, everything else being equal.  Nevertheless, it is important to say that a gentle BEV driver, with the benefit of regenerative braking, can more than cancel out the tire wear emissions from the additional weight of their vehicle, to achieve lower tire wear than an internal combustion engine vehicle driven badly.

An important difference between tire and tailpipe particle emissions is that most of the former is understood to go straight to soil and water, whereas most of the latter is suspended in air for a period, and therefore negatively affects air quality.  This is supported by Emissions Analytics’ results, which suggest around 11% of the mass of tire emissions is smaller than 2.5 microns in diameter (defining the common metric for fine particle dust, PM2.5, which can be airborne).  Therefore, the airborne tire emissions are more likely to be around 8 mg/km as shown in the table above – this is still more than 400 times higher than tailpipe emissions.  

However, considering just tire mass emissions may underestimate the effect on air quality and the consequent human health effects.  The particulate number can be estimated, as shown in the second table.  When measuring particle number, the lower size cut-off is important – the smaller the particles get, the more volatile and harder to measure repeatably they become.  Cutting off at 23 nm avoids these ‘semi-volatiles’, whereas 6 nm is a much more comprehensive range.

This shows that 13.4 x 10¹¹ #/km are in the size range between 6 nm and 23 nm, which represent 92% of all particles below the 10 micron upper limit that defines PM10.  Growing scientific evidence suggests that these ultrafine particles more easily enter the human bloodstream and lungs, and cross into the brain.  The 14.5 x 10¹¹ #/km for the whole size range should also be compared to the maximum number permissible from the tailpipe of 6.0 x 10¹¹ #/km at a 23 nm cut-off, and the actual real-world values from Emissions Analytics’ EQUA test programme of 0.9 x 10¹¹ #/km for gasoline vehicles and 0.1 x 10¹¹ #/km for diesels.  Therefore, tires create more than double the particle number emissions of the tailpipe, averaged across those two fuel types.  Put another way, tires may release an extra sixty billion particles for every kilometre driven.

While the body of research on the health of effects of ultrafine particles is growing, how bad these effects are is likely to depend on how toxic the particles are.  Light-duty tires are typically made up of synthetic rubber, derived from crude oil, rather than natural rubber, together with various fillers and additives.  In a recent newsletter, Emissions Analytics set out its initial findings from chemical analysis of the organic compounds in a range of tires using two-dimensional gas chromatography and time-of-flight mass spectrometry.  This showed that there were hundreds of different compounds in each tire, with a significant proportion being aromatics, some of which are recognised carcinogens.

The next stage is then to take that chemical profile and study the toxicity of each.  Our research so far shows that the least toxic tires are one-third as toxic as the worst – this will be the subject of a future newsletter.  Therefore, tires not only vary significantly in wear rates, but also in chemical composition and toxicity.  This potentially points to an effective way to drive reductions in wear and toxicity through economic incentives and regulation.

Bringing all these elements together, Emissions Analytics will be publishing the chemical composition and toxicity of hundreds of different tires in a subscription database that will be launched in mid-May 2022.  The aim is to bring transparency and insight to an area that has historically been under-researched, but which has now been thrown into the spotlight with ever-heavier vehicles and rapidly cleaning tailpipes.  We are looking to talk to anyone interested in research in this area.

---

Want to join the conversation at our next conference? Click the photos below to learn more.

The septillion particle problem (literally)

That’s 1,000,000,000,000,000,000,000,000 exhaust emissions particles that are due to be emitted in the United States that don’t need to be.  How?  A large majority of European and Chinese cars are now sold with tailpipe particle filters, known as gasoline particulate filters (GPFs) or diesel particulate filters (DPFs) in the industry, but this is not the case in the US. Focus on small, ‘ultrafine’ particles in the European regulations has driven gasoline exhaust filter adoption in a way that has not happened in the US, where particulate mass measurement has primacy, and effectively ignores these ultrafines.
 
A gasoline vehicle with a filter emits around 0.9 x 10¹¹ particles per mile (90 billion), but without a filter it is about 6.3 x 10¹¹ (630 billion), according to testing conducted by Emissions Analytics. With about 300 million internal combustion engine (ICE) vehicles likely to be on the road on average over the next ten years, each driving an average of 10,000 miles per year, that put the total number of ‘unnecessary’ particles emitted to US air to be 1.6 x 10²⁴, or 1.6 septillion particles.  If each particle were a dollar bill, the whole of the US could be carpeted over half a mile deep in money. 

It is important, however, not to think of these particles as creating visible plumes of black smoke from the tailpipe.  That is the clichéd image of exhaust particles, but that applies to older diesel vehicles without a particle filter.  For gasoline vehicles, and those with a filter, the concern is for very small particles, down to 10 nm in size, which are invisible, numerous and highly mobile.
 
These ultrafine particles are of concern for human health, as their small size means that when inhaled they can be drawn deep into the lungs and cross the blood-brain barrier, with potential effects on cognitive development.  The direct link to negative health outcomes is not generally considered as proven yet, but Europe has nevertheless regulated due to the weight of evidence and as a precaution. European particle-number-based standards have led to the widespread adoption of GPFs, which are effective at trapping these small particles.  This was quickly adopted by China and India.
 
Even if only particle mass is concerned, on average a gasoline vehicle without a filter emits around 0.070 mg/mile, compared to a vehicle with a filter at 0.027 mg/mile. The unnecessary particle mass to be emitted in the US is therefore forecast to be over 100 tonnes over ten years.  Not only does this have the understood impacts on air quality in urban areas, but observations suggest that it eventually migrates to the North and South Poles, settles on the ice, and helps accelerate climate change because of the high ‘black carbon’ content in soot, which is darker and a known global warming agent.
 
Some would say that this is minor or irrelevant as the light-duty fleet moves to full electrification. However, even on the most aggressive scenarios, new ICE vehicles will be sold in the US for over ten years. Those vehicles are likely to have a lifespan of around twelve years, so these vehicles could be on the road easily for 25 years from today. This does not reflect a lack of ambition in decarbonisation, but a practical reality. In the transition, there is a strong argument that we should do what we can to make ICE vehicles as clean as possible. Emissions Analytics has written extensively before that hybridisation is a better route to quick decarbonations – see a previous newsletter – and, if this position comes to pass, it could mean that hybrid vehicles are sold long after pure ICE vehicles are phased out, and these hybrids emit ultrafine particles in a similar way.
 
The emissions values described above are the result of programme run by Emissions Analytics in 2021 that tested apparently similar pairs of vehicles between the US (without GPFs) and Europe (with GPFs).  The US vehicles were tested in Michigan, and the European vehicles in the UK.  Four pairs of gasoline vehicles were tested.  First, the three-litre BMW X5 was a 2019 model year in the US and 2020 in Europe.  Second, there were two 2020 Jeep Wranglers with a 2.0 litre engine.  Third, the full hybrid Toyota RAV4s were 2019 model year with the 2.5 litre engine.  Finally, the 2020 1.5 litre Ford engine was tested in the Escape in the US and its equivalent Kuga model in Europe.  These vehicles were drawn from different manufacturers and where the models were near-identical in technical specification and model year between the two regions.  That said, in each case, the Europe vehicle was equipped with a GPF, while the US vehicle was not.  
 
Each vehicle was then tested on similar on-road routes in the two countries to allow the comparison of their particle mass and number emissions.  The gas emissions and particle number were measured using a regulatory-approved portable emissions measurement system (PEMS) from Sensors, Inc, together with a diffusion charger for particle mass measurement.  Similar on-road routes were designed in both locations, which, despite the inevitably different traffic and ambient conditions, were shown to have similar dynamics overall.  The average emissions across the four pairs of vehicles are shown in the charts below. 

The results were highly consistent between the four vehicles.  They were also consistent between different types of driving, from urban cold start to highway warm start. The table shows that the average reduction in particle number was 86% across the different vehicles and driving. The variation was from 76% to 96% – in all cases, the reduction was impressive.  As the vehicle specifications were near-identical in other respects, it is very likely that the improvement is due to the GPF.  While there were some differences in the chemical composition of the standard pump fuel used in both locations, this would not account for a significant proportion of this reduction. 

The equivalent overall particle mass reduction was lower but still impressive at 55%, with 81% reduction on the highway.  Despite this positive performance, the question remains as to whether the benefit in improved air quality is worth the extra cost.  A GPF system typically costs less than $200, which is about the same cost as the optional carpet mats on the BMW X5.  Therefore, the cost for removing every billion particles emitted into the air is just 0.0004 cents.  In addition to this, there is the further potential benefit of removing particles from the air that have come from other sources.  As Emissions Analytics has shown in another newsletter, diesel particulate filters (DPFs) can ‘clean’ the air as the vehicle drives. GPFs are now achieving filtration efficiency rates – typically 80%, with the next generation of GPFs for Euro 7 regulations likely to exceed 95% – that are getting close to those of DPFs, and so the same net-cleaning effect can be expected. Moreover, the filtration efficiency tends to improve with the age of the filter, so the positive effect, if anything, grows.
 
The septillion particles mentioned above may even underestimate the potential for emissions reduction.  The numbers are based on a rapid transition to full battery electric vehicles (BEVs), with no tailpipe emissions.  However, this transition may not be as fast as anticipated, which would mean more ICEs for longer.  Widespread installation of GPFs would, therefore, generate greater benefits over a longer period.  But even as BEVs take off, there is growing evidence that they emit more tire wear emissions that ICE vehicles due to their increased weight.  Emissions Analytics’ tests showed a 21% increase in tire particle mass emissions for 500 kg extra vehicle mass – roughly equivalent to the mass of a large battery pack – although this may be partially offset by the effect of regeneration breaking.  More generally, there is an on-going trend towards heavier vehicles of all types, which increases tire emissions.  Even though only 10-20% of tire wear emissions hang in the air, GPFs could have the added benefit of removing these particles from the air.  In other words, ‘legacy’ ICE vehicles could help clean up emissions from BEVs.
 
Put another way, GPFs bring direct value, but also have an important ‘insurance’ effect that ensures there is benefit however the vehicle car parc evolves. 

With filters on gasoline cars in Europe and China, and even coming soon to India, it is surprising that most new car sales in the world's second largest car market are not equipped with the best available technology for emissions reduction.  There are signs that this might change, however.  Even if US regulators do not embrace particle number as a measure that should be regulated, the potential reduction of the particle mass limit to 1 mg/mile at normal and cold temperatures (20°F or -7°C) might enable cleaner air in the US with widespread deployment of GPFs.
 
Few things in emissions control are a no-brainer.  Mostly, awkward trade-offs between cost, vehicle utility and emissions reduction have to be resolved.  In this case, the benefits appear high, costs relatively low, and risk minimal. 

---

Credits: BMW VX images from BMW and www.thecarconnection.com. 

The great public transport squeeze

Public transport, almost by definition, works when it is packed.  As demonstrated at the current time, with usage at around three-quarters of pre-pandemic levels, the economics of many parts of public transport are in crisis, not least at Transport for London. A similar pattern is seen with environmental impact: per passenger, a full bus has lower carbon dioxide (CO2), nitrogen oxide (NOx) and particulate emissions than a car; but a near-empty bus is worse.  But there is another, less well understood, aspect of public transport regarding the users’ exposure to pollution while travelling, which presents a further challenge to rebuilding these modes.  Looking forward, not just do emissions and economics need to be considered, but also the comfort and health of passengers.

To study this, we tested a range of transport modes between central London and Oxford city centre, a distance of approximately 50 miles (80 km) as the crow flies.  This project was funded by the TRANSITION Clean Air Network through the first round of its Discovery & Innovation Fund in 2021¹.  

The main pollutants of interest for this study were particles – both mass and number – and volatile organic compounds (VOCs). These are likely to be the most problematic pollutants in confined spaces, and this selection mirrors the work on passenger car interior air quality work at the UNECE and CEN Workshop 103².  To measure these, we used the V2000 from National Air Quality Testing Services for particles³.  For VOCs, samples were collected in thermal desorption tubes – using both passive and active sampling – and then analysed in a two-dimensional gas chromatography system: GCxGC-TOF-MS from SepSolve Analytical and Markes International.  Due to limitations in the study, most journeys were undertaken only once and, therefore, more data would be required to understand variability between exposures on different days.

The transport modes tested were the electric train, diesel train, diesel coach, diesel and hybrid bus, battery electric vehicle (BEV), diesel internal combustion engine (ICE) vehicle, bicycle and on foot. Most of these vehicles were relatively modern, as shown in the images above.  The car journeys were point-to-point from London Paddington station to Oxford High Street, whereas the others – except bicycle and on foot – were multi-stage journeys, trying to achieve the most direct route.  For example, the diesel train was direct between the cities with an added bus journey to Oxford city centre.  The least direct route involved the London Underground, an electric train, then a diesel train and a bus.

Before getting into analysis of the vehicles, it is important to say that, generally speaking, active travel by bicycle or on foot led to among the lowest pollutant exposures, even in the city of Oxford, although there was significant variability between journeys. Although the traveller was exposed to spikes in pollutants, especially from other vehicles, the level of dilution with fresh air was high.  However, cycling or walking is not a viable substitute mode for the whole journey from London to Oxford.

The table below shows the estimated total mass exposure to VOCs, per passenger, split into functional groups, for the different journeys, factoring in the trip duration and volume of air inhaled due to the activity level involved.  Simplistically, the VOC functional groups have been presented in increasing order of potential toxicity, from alcohols at the lowest level to polycyclic aromatic hydrocarbons (PAHs) and nitrogen-containing compounds, which are often carcinogenic.

Overall, the coach generated the greatest VOC exposures, particularly in the alkane and aromatic groups.  The highest PAH exposures were on the direct diesel train to Oxford.  Electrified public transport typically saw low exposures, indeed lower than the BEV passenger car.  In fact, the exposures in this brand-new BEV were almost four times higher than the nine-year-old diesel car.  This may be due to the BEV being a smaller city car, whereas the diesel was a premium vehicle, which may have had a superior ventilation and filtration system.  Furthermore, it is seen more widely that BEVs can have worse in-cabin pollution as energy is not applied sufficiently to the ventilation system in order to maximise range.

The coach saw the greatest overall exposures, not least because of the journey's long duration, with a significant proportion of the VOCs appearing to originate from personal care products – deodorants, shampoos, perfumes and so on – from other passengers, as shown in the table below.  To perform this analysis, each VOC is allocated to a category based on its most likely source, and the unit of measurement reflects the total area under the peaks on the chromatogram.  Although these personal care products are likely to be less toxic to humans at the concentrations measured, the coach also had the highest level of fumes from fuels and lubricants, which are likely to have come from other vehicles on the road, but allowed to enter and persist in the coach cabin.

By far the cleanest vehicle, therefore, is the diesel car (a Mercedes-Benz), but the electric car is poor – the prevalence of VOCs from both personal care products and engine fumes in the BEV is consistent with the idea that the ventilation is poor in this budget car (a Vauxhall), and how important a good ventilation system is to balance filtration and air exchange rates in protecting the driver.

Is there evidence to support this when looking at the particle and CO2 results?  The table below shows the average results for particulate number (PN), particulate mass (PM) and CO2.  Broadly, the former two are measures of pollutant exposure, whereas CO2 is a proxy for air freshness and rate of air exchange.  This immediately reveals the public transport dilemma – packing more people in tends to lead to stuffier air, just as passenger density is the key to greenhouse gas emissions reduction per passenger-mile.

The London Underground journey – an electric train – has by far the highest PM exposures, which is likely to be fluff, metal from the tracks and human detrita.  Ultrafine particles, as measured by PN, are relatively low there, in contrast to the electric train from London Waterloo to Basingstoke, which saw high particle levels.  For particles, the results were similar between the two types of car, which suggests the issue is more that VOCs are off-gassing from the interior materials on the new BEV, and its ventilation system is not exchanging the air often enough to keep the concentrations down.  The least fresh air was found on the coach and diesel train, which werebad not just for VOC exposures, but general comfort.

Thus far in this newsletter, we have described emissions in a vehicle-centric way.  However, when the real-time data is considered in more detail, it is clear that a significant proportion of exposures on public transport are not when the vehicle is moving.  Rather, the process of waiting for, boarding and alighting the vehicle, together with it letting on other passengers at stops, creates exposure spikes.  The time series chart below takes the example of the journey via the Underground, electric train and then diesel train to Oxford.

After the fluffiness of the Underground, getting onto the electric train at Waterloo showed a spike in both PM and PN, followed by PN spikes at various stations.  The change, with a wait on the platform, onto the diesel train at Basingstoke had a large, persistent elevated level of PN and brief spike of PM.  Much of this is due to a cocktail of diesel fumes and cooking smells from food concessions inside the station.  The diesel train, once boarded, was cleaner, but with spikes at each station.  

In short, public transport is getting squeezed between active travel and the car, both in economics and for health.  It is not sufficient just to upgrade public transport vehicles to reduce VOC and particle exposures, but also the whole experience including interchanges, stopping patterns and waiting times.  Public transport also needs to resolve the tension between wanting to squeeze as many people in for good economics and environmentals, and maintaining a good level of air freshness and comfort.

---

References

1. TRANSITION is a UK-wide network, led by the University of Birmingham in collaboration with nine universities and over 20 cross-sector partners, aiming to optimise the air quality and health outcomes of transport decarbonisation.  The network (NERC ref. NE/V002449/1) is itself funded by UK Research & Innovation through its Clean Air Strategic Priorities Fund, administered by the Natural Environment Research Council.  Raw data from the testing can be found in Centre for Environmental Data Analysis (CEDA) at https://catalogue.ceda.ac.uk/uuid/96c912c7c0094da2a8627d446cb06708.

2. Research.lancs.ac.uk/portal/en/activities/cenws-103--real-drive-test-method-for-collecting-vehicle-incabin-pollutant-data-event(8bc35484-83b8-44a6-8f51-5f35e1d7f72b).html

3. NAQTS

Why battery durability matters for decarbonisation

Are policy priorities correct?

Lifecycle carbon dioxide (CO2) modelling of battery electric vehicles (BEVs) typically relies on the critical assumption that the battery lasts the lifetime of the vehicle, which is around 14 years.  However, we know that battery capacity declines over time, and manufacturer battery warranties tend only to cover the first eight years.  What if battery packs last a shorter time than expected, requiring replacement during the vehicle lifetime or even leading to early scrapping of the vehicle?  How would that affect the lifecycle CO2 of BEVs and impact, at the overall car parc level, our ability to achieve ambitious climate change goals.

Almost two-thirds, typically, of CO2 emissions in the manufacture of a BEV are associated with the battery , including sourcing raw materials.  For an internal combustion engine (ICE) vehicle, even a mildly hybridised one, that fraction is less than 5%. Therefore, the life expectancy of the BEV battery pack is clearly crucial to the vehicle’s overall CO2 footprint – and residual valuation – in a way that is not true for primarily combustion cars.  

Emissions Analytics has previously reasoned that the best route to decarbonisation for at least the next decade is to pursue a path of mass hybridisation.  This is primarily due to the scarcity and cost of battery materials, which is now being seen as a result of accelerating demand and supply chain constraints.  A further reason is the risk around consumer acceptance of BEVs at retail prices that are profitable for manufacturers.  The argument is set out in more detail in a previous newsletter.  The analysis uses a proprietary lifecycle model created by Emissions Analytics, based on a meta study of academic literature and information from manufacturers.  The key input assumptions are set out in that previous newsletter – of note is that it is assumed that the grid is decarbonised by 50% from 2030, but manufacturing emissions remain constant.  Battery life is assumed at eight years by default for our purposes here, reflecting manufacturer confidence levels as revealed by warranty durations.

In this newsletter, we compare three different vehicle sales scenarios: an ‘ICE baseline’ scenario with 50% gasoline and 50% diesel vehicles; a ‘BEV transition’ model with 75% BEVs in 2031 and 100% in 2035; and a ‘Hybrid strategy’ with 100% full hybrids (FHEVs) throughout.

The clear conclusion is that, in 2050, the deadline for net zero emissions for many countries including the UK, the cumulative CO2 emissions from the manufacture and usage of vehicles is similar whether we go for the direct BEV transition or the Hybrid strategy, as shown in the chart below. 

Compared to the ICE baseline, the direct BEV transition reduces CO2 by 1% more than the Hybrid approach.  The chart makes it clear how important electrification, generically, is to reducing CO2 from transport, but it raises the question whether the additional gains from the pure BEV model warrant the investment required – that investment not just being subsidies, tax incentives and government infrastructure investment, but also by the car buyers themselves due to higher purchase prices.  This timespan, however, does not bring out the longer-term benefits of BEVs, which can be seen if we look at annual, rather than cumulative, reductions in CO2, as shown in the chart below.

As a result, after 2050, the cumulative advantage from the BEV transition gradually widens as old ICE vehicles drop out of the parc and the lower in-use CO2 of BEVs comes to the fore.  By 2070, the lead in cumulative CO2 emissions of BEVs has reached 9% over the Hybrid strategy, and a full 34% compared to sticking entirely with ICE vehicles.  This is a significant reduction, but 49.2 megatonnes of CO2 still come from an entirely BEV parc in one year – hardly “zero emission.”

If the assumption around an eight-year battery life is then flexed, we can see that it has a significant impact on the end result of electrification.  If, in practice, batteries last on average as long as the chassis – 14 years – then by 2050 the BEV transition scenario would deliver 13% points more CO2 reduction than with eight-year battery life, a figure that rises to 16% points by 2070, as shown in the chart below.  So, in 2070, the 9% CO2 reduction from faster BEV penetration over the Hybrid scenario is dominated by an additional 16% that could come from better battery life.

So, we conclude that, as long as we urgently electrify – whether BEV, PHEV or FHEV – we will achieve significant, and similar, CO2 reduction by 2050.  Additional penetration of BEVs delivers extra CO2 reduction only after 2050, and it comes at a cost and with execution risk.  Improving the lifespan of batteries makes a bigger difference to CO2 emissions in particular by 2050. Therefore, this suggests that policy needs to be more attentive to the quality of the current generation of BEV products than simply the number on the road.  

In a pessimistic scenario where batteries only last on average of six years – which would be costly both in terms of CO2 and to the underwriters of manufacturer warranties – the Hybrid strategy would be 10% better in 2050 than the BEV transition, and still be better, by 6%, in 2070, as the constant renewing of batteries outweighed the lower in-use emissions of BEVs.  This is not a likely scenario, but makes plain the sensitivity of our decarbonisation policy to a complex and opaque piece of engineering produced by the private sector.

It could be argued that recycling batteries would mitigate the problem of poor durability, and shift the evaluation more in favour of BEVs.  While this might become true in the future, the recycling infrastructure does not yet exist at scale and the energy required to separate the chemical components, clean them up and then assemble into a new pack is also significant.  It is currently unclear whether the saved mining emissions is greater than the CO2 overhead of the recycling.  A further criticism of this analysis might be that the CO2 in manufacturing both the car and battery does not reduce over time in the model.  This may happen, but is not certain.  Equally, no benefit from reducing emissions from liquid fuel production, or the use of biofuels or synthetic fuels has been assumed.

So often, when talking of decarbonisation, the dates mentioned seem far in the future.  Using the same model as above, we can make the numbers real and immediate by tracking the CO2 reductions since the start of 2020.  Emissions Analytics compiles its “CO2 totaliser”, which compares the cumulative CO2 emitted by cars sold since January 2020 under our current BEV transition policy against the two alternative scenarios.  The first comparison is to a car market that remains 50% diesel and 50% gasoline ICE.  The second is to a market completely made up of FHEVs.  The early moves by the UK into BEVs has currently led to more CO2 emissions than on both these alternative scenarios

---

The 3.3 million vehicles sold in the UK since January 2020 emitted in total 6.153 megatonnes more CO2 than would have been the case if all vehicles sold had been full hybrids, which typically have batteries between 1.5 and 5.0 kWh in capacity, and 2.360 megatonnes more than if all vehicles had an ICE.  This is not surprising as BEVs – with batteries typically between 60 and 100 kWh – have much higher emissions from the manufacture, which is then offset by lower emissions during use.  This makes plain that calling BEVs ‘zero emission’, solely because they have no tailpipe, is a nonsense.  The surplus CO2 is even greater compared to a market made up only of FHEVs, because they have manufacturing emissions only slightly higher than ICE vehicles, but around 30% lower in-use emissions.

The annual CO2 budget for the UK is approximately 700 megatonnes, based on consumption emissions and including aviation¹. This value contrasts with territorial emissions, which only measure CO2 arising from activity in the UK – this is particularly relevant where a high proportion of vehicles are manufactured overseas.  Around 27% of total emissions are from transport² and, therefore, the 6.153 megatonnes excess accounts for around 1.6% of total transport emissions.

In summary, with shoulders firmly to wheel of BEV rollout, lubricated with a flow of taxpayer subsidy, it would be right to tally up whether the approach is leading, or will lead, to significantly reduced emissions.  In a market dominated by BEVs, we must be conscious that the dominant element of CO2 emissions is likely to become the replacement of batteries rather than the energy required to propel the vehicle.  This supports the idea that policy should perhaps focus more on the longevity and durability of these batteries, rather than a singular focus on BEV market share.  We have consciously simplified the options to illustrate this underlying point, including not factoring in the potential for PHEVs due to the sensitivity of their emissions to user behaviour.

After all, the aim is surely to reduce CO2 emissions and slow global warming, rather than to promote BEVs as inherently better products – if they were so superior, they would not require subsidy at all.

---

References
1. Carbonindependent.org/23.html
2. Government statistics Transport and environment statistics autumn 2021

Are cars sinful?

This is no polemic for or against the car.  Nor about petrolheads.  Nor environmentalists.  This is about the chemicals that are to be found inside cars, and the importance of good ventilation, to avoid occupants gently steeping in a cocktail not of their own making.  The SIN is the ‘Substitute It Now!’¹ list developed by the International Chemical Secretariat, a largely government funded organisation in Sweden² – a list of over one thousand chemicals with harmful health or environmental effects, which they suggest should be removed from consumer products.  But do we find any of these in modern cars?

This is not a new problem, but one the significance of which we are only beginning to understand.  As discussed in previous newsletters, the emissions from vehicle exhausts have improved significantly over the last ten years.  Initially it was mainly seen as particulate reduction, as the widespread cheating on nitrogen oxide (NOx) emissions – of which Dieselgate was emblematic – was only resolved from around 2018 when official laboratory results were validated with an on-road test called Real Driving Emissions.  The remaining pollutants of most concern in the exhaust are carbon dioxide (CO2) and a range of volatile and semi-volatile organic compounds (VOCs and SVOCs).

At the same time, the construction of vehicles has become more sophisticated.  While generally to the benefit of consumers in terms of utility and design, the potential impact on the health and comfort of vehicle occupants has not been fully considered. The increased use of glues in vehicle manufacture and superior construction methods creating an almost perfectly sealed cabin come with downsides.  Significant hurdles in understanding the consequences have been difficulties in measuring the compounds of interest, and the lack of standards to measure vehicle interior air quality (VIAQ).  This is beginning to change, led by South Korea and Japan, which collectively regulate nine different VOCs in the cabin.

Emissions Analytics has been working on this from both angles.  It has been an active participant in a standardisation workshop under the Comité Européen de Normalisation (CEN)³, aiming at measuring the air quality inside light-duty vehicles in terms of particle ingress from outside and CO2 build-up.  In parallel, it has been developing methods for measuring the presence of VOCs and SVOCs in cabin air, and how they might be emitted from the interior materials.  The two areas are linked because the worse the ventilation and filtration system, the greater the likelihood of the accumulation of noxious organic compounds. Multiply that by the length of time typically spent in vehicles and the human exposures could be greater than to ambient air pollution.

To get a measure of the problem, if we chemically analyse the interiors of cars, do we find any of the compounds on the SIN List?  We took a small sample of recent model year vehicles in Europe and analysed a range of interior materials, including dashboard plastic, seat material and carpets.  Each sample was analysed using two-dimensional gas chromatography coupled with time-of-flight mass spectrometry, with sample introduction via microchamber – a system provided by Markes International and SepSolve Analytical.

One vehicle in particular gave interesting results: a top-selling, European-made small car from 2020.  We tested the carpet in the boot and footwell, dashboard plastic and the seat covering material.    Across these sample locations, 16 of the SIN List compounds were identified, as shown in the table below.  The metric is the percentage of the peak area – in other words, the area under the compound’s peak on the chromatogram expressed as a proportion of the total peak area of all organic compounds identified.  It is a reasonable approximation of the relative abundance of the SIN List compounds in each sample.

While just 1.23% of the organic compounds found in the dashboard plastic were on the SIN List, this exceeded 11% for the footwell carpet.  Furthermore, its most prominent compound was diethyl phthalate.  Generally, phthalates are a synthetic substance commonly used to make plastics more flexible.  A growing body of research suggests that many of the compounds in this group have undesirable health effects, including reducing fertility.  As a result, regulations are beginning to restrict their use, for example, under REACH in the EU, the concentration of four target phthalates cannot together exceed 0.1% by weight in consumer goods.  Diethyl phthalate is not one of these four, and the research on its health effects is currently limited, but its high concentration in the footwell carpet should warrant additional focus.

Beyond phthalates, a way to assess the health dangers of the other compounds on the SIN List is to look at the ‘hazard statements’ produced under the Globally Harmonized System of Classification and Labelling of Chemicals (GHS), an international consensus system developed by the United Nations for classifying and labelling hazardous chemicals.  Hazard statements provide standardised wording as to the nature and degree of the threats.  For example, ‘H350’ indicates a chemical that may cause cancer.  An individual compound can have multiple hazard designations.

If each of the compounds identified in the vehicle are scored according to the number of hazard designations, and these are then grouped generically, we obtain the totals shown in the table below.  Included in this assessment are those assessed as ‘persistent bioaccumulative toxic’ compounds, which is a separate hazard category.  These compounds do not break down easily in the environment or living creatures and so progressively build up.

At a high level, it can be seen the most prevalent health effects are irritation or damage to the skin, eyes, throat and lungs.  The boot carpet has the largest number of designations, and contains a relatively high level of octamethylcyclotetrasiloxane, which is suspected of damaging fertility.  The harms are not just to humans, for example naphthalene is also highly toxic to fish and other aquatic life.  

The number and extent of these compounds present in common materials in vehicle interiors points to action needed by manufacturers to address the potential toxicity to occupants.  Beyond the SIN List, we have also identified a ‘PFAS’ in the same vehicle’s seat material.

Per- and polyfluoroalkyl substances (PFASs) are manmade organofluorine compounds, first created in the mid twentieth century and which became popular for their water repellent, stain resistant and non-stick features.  The problem is that they do not break down readily in the environment, hence being described as “forever chemicals”, or persistent organic pollutants.  This bioaccumulation means that it is believed that they are gradually building up inside most living creates as well as the wider environment.

The health effects of PFASs were brought to wide public attention for the first time in the 2019 film Dark Waters, where compounds based on perfluorooctanoic acid (PFOA), a subset of PFASs, were released from an industrial source in West Virginia, USA.  The evidence as to the environmental effects has continued to accumulate.  An article in December 2021 in Environmental Science & Technology shows that these persistent compounds do not simply get washed into the sea and diluted into insignificance.  Rather, waves crashing on the shore recirculate the chemicals into the air and onto land⁴.  This increases the human exposure and associated health effects, which include cancer, thyroid conditions and colitis.

The compound identified was 1H‚1H‚2H‚2H-perfluorooctan-1-ol – chemical formula C8H5F13O – a chain of eight carbon atoms with 13 fluorine atoms attached.  The chromatogram from the seat sample below shows this compound.  Of the total peak area of organic compounds, it accounted for 0.61%.  The hazard classification says that it may cause damage to organs through prolonged or repeated exposure.

It should be added that it is not being claimed that exposure to this seat covering will have direct health effects.  Nevertheless, by the persistent nature of these compounds, they are very likely eventually  to end up in the environment, be recirculated and potentially inhaled or ingested by humans or animals.  PFASs from a car seat add to this gradual and irreversible accumulation.

In the longer-term, the objective should be to engineer these compounds out of vehicles, where there are viable alternative components.  Growing regulations around materials, from REACH in Europe and from the Environmental Protection Agency in the US, together with emerging regulation of vehicle interior air quality led by Korea, Japan, Russia and China, will force the pace.  But how should vehicle owners mitigate the effects in the shorter run?

The concentration of VOCs in cabin air generally increases as the vehicle heats up.  A car parked in the sun will see more chemicals released from the interior materials into the air, recondensing once the vehicle cools back down.  To this chemical soup is added particles and NOx that enter the car through the ventilation system as the vehicle moves, plus the CO2 exhaled by the occupants.

Although not the most glamorous part of a car, a good heating, ventilation and air conditioning system is the best way to limit exposure.  The benefits of maintaining fresh air in confined spaces has been brought to prominence through the Covid pandemic.  The ideal is to keep the ventilation on fresh air mode – rather than recirculation – but this relies upon excellent filtration to minimise the particle and NOx ingress.  In balancing this trade-off, the best vehicles are many times better than the worst, as described in previous newsletters.  Therefore, rating vehicles in a fair and comparable way is urgently needed.

Fortunately, both the car and filtration industries have been actively involved in standardisation of test methods, including through CEN Workshop 103, initiated by the AIR Alliance⁵, which is due to publish its work soon.  Together with the work on vehicle interior air quality through UNECE, it is quite possible to see how these new tools in measurement and evaluation can bring about healthier, and perhaps SIN-free, cars for both the occupants and the wider world.

---

References
1. Sinlist.chemsec.org/
2. en.wikipedia.org/wiki/ChemSec
3. CEN Workshop on real drive test method for collecting vehicle in-cabin pollutant data
4. Pubs.acs.org/doi/10.1021/acs.est.1c04277
5. Airindex.com

The even more hidden life of tyres...

If the ‘bio’ components in liquid fuels are not well understood – as discussed in our last newsletter – the composition of vehicle tyres verges on mysterious.  Fuels have attracted close focus over the last few decades primarily because of the pollutants, especially particles, released when the fuel is combusted in an engine.  More recently, the subject of upstream carbon dioxide emissions from the production of fuel has risen up the agenda with concern about climate change.  In contrast, although tyres clearly shed significant amounts of material into the environment, their wear has not been regulated and, therefore, where the tyre particles go and what they do to humans and the environment has generated little interest.

Tailpipe emissions continue to fall and, in respect of particles, leave tyre wear emissions somewhere between ten and a thousand times greater than the tailpipe emissions from a typical car, according to earlier Emissions Analytics’ research.  This, together with the increasing weight and torque of new vehicles, has led at last to a focus on tyre wear emissions.  As a result, a number of important questions are now being posed.

First, what is the rate of tyre wear emissions in real-world driving, quantified either as mass or number – the latter being important to capture potential large numbers of ultrafine, near-massless particles?  Second, to what extent do particles measured in the environment come from tyres?  Third, what is the chemical composition of tyres, and how does that differ between brands?  Fourth, what pollutants do tyre wear particles leech into the environment once released?  This newsletter considers approaches that could be used to build a better understanding on these points.

From previous Emissions Analytics' newsletters – for example, https://www.emissionsanalytics.com/news/whats-in-a-tyre – the total mass loss from the four tyres of a vehicle averages around 64 mg/km when the tyres are new, falling to about 32 mg/km over the full life of the tyres.  Approximately 90% of the mass shed is from larger particles up to at least 10 µm in diameter, whereas around 90% of the particle number is ultrafines down to 10 nm and below.  Wear rates differ by a factor of around two to three times between the fastest and slowest wearing.

While we have a good understanding of the rate at which tyre wear particles are shed from vehicles, we have little idea where the particles go.  Broadly, they will end up in soil or water, although some maybe trapped in storm water filtration systems.  The distance the particles travel from the carriageway depends on factors such as the particle mass and they can therefore accumulate on the verge or be blown to neighbouring fields and gardens.  Particles that fall to the carriageway typically get washed into drains and end up in rivers and the sea.  According to a 2017 report, 28% of primary microplastics in the ocean are from tyres¹.  The proportion that become airborne eventually settles on soil or water.  Therefore, it should be expected that tyre material is observed in samples of water and soil taken in almost any location.

A potential method for estimating the amount of tyre material in a sample from the environment is to use fingerprinting.  The complicating factor is that a tyre particle released into the environment will get mixed up with many other sorts of material.  By using a pyrolysis technique – burning a sample in the absence of oxygen – leads the tyre material to break down in a predictable way, creating a ‘pyrogram’.  Emissions Analytics has been building a reference database, or ‘library’, of the pyrolysis products from a wide range of new tyres.  By matching the patterns from the environment samples, it is possible to start to understand what fraction of the sample is from tyres – a process that is called ‘source apportionment’.  

In principle, this can be used not just to see tyre material generically, but also the presence of wear from particular tyre brands, where they offer unique chemical fingerprints.  It should be noted that pyrolysis causes some of the compounds to break down – albeit in predictable ways – and therefore chemicals observed after pyrolysis are not necessarily the same as those in the tyre when new, although they may indicate compounds that will be released as tyres naturally degrade over time.

To illustrate the approach, Emissions Analytics tested samples from ten different tyre brands using a pyrolysis front end on its two-dimensional gas chromatography and time-of-flight mass spectrometry laboratory, supplied by Markes International and SepSolve Analytical.  This allows excellent separation, identification and quantification of organic compounds in the sample.

The table below shows the top five most prevalent compounds from the pyrolysis of the ten samples.  The metric is the ‘peak area’ percentage, which is the area under the peak for each compound as a share of the area under all the peaks on the chromatogram added together.  It is not an explicit quantification, but nevertheless a good indication of relative importance.

A number of the most prevalent compounds offer a characteristic odour, including limonene², β-guaiene and longifolene.  These come from natural sources and include sweet, woody, dry, guaiacwood, spicy and powdery smells.  Most of these compounds are not clearly toxic, although cyclohexane‚ 1‚2‚4-triethenyl- is an irritant to eyes and skin, and desogestrel is a hormone.  Two of the most prominent, androstan-17-one‚ 3-ethyl-3-hydroxy-‚ (5α)- and 2-[4-methyl-6-(2‚6‚6-trimethylcyclohex-1-enyl)hexa-1‚3‚5-trienyl]cyclohex-1-en-1-carboxaldehyde, are complex compounds that have unknown characteristics.  Overall, these compounds are relatively high-order, complex, heavy hydrocarbons and oxygenates.

One of the brands has a noticeably different mix of chemicals in its composition: Tyre 10.  As can be seen in the above table, it only has three of the top compounds present, whereas most of the compounds are present in the other brands.  Furthermore, the Tyre 10 sample contained a number of prominent compounds that were unique to it.  These included cis-Thujopsene – a terpene that is an essential oil found in conifers – linolenyl alcohol³, eicosapentaenoic acid⁴, sorbic alcohol⁵ and the unknown bicyclo[3.1.1]hept-3-ene-spiro-2‚4'-(1'‚3'-dioxane)‚ 7‚7-dimethyl-.

Therefore, using this approach, the pyrolysis results from environmental samples can be compared to these reference measurements to estimate both the generic presence of tyre wear material, but also potentially the approximate split between different types of tyre.  As relatively complex hydrocarbons, they do not generally occur naturally, although it should be noted that there are potentially other sources of some of them, such as from industry.  It is therefore important to get as much detail on the tyre composition as possible, in order to uncover specific chemical markers that can be used for robust source apportionment.

The same analyses can be used to probe into the chemical composition of tyres and understand the differences between different tyre brands.  The chart below is a plot using ‘principal component analysis’ (PCA) to understand what groups of compounds – the principal components – account for the most difference between the tyres.  The aim of the method is to reduce the number of dimensions to something we can easily understand from samples that typically contain over one thousand organic compounds.  It trades a little accuracy to aid understanding.

The first principal component accounts for 80% of the observed differences.  The second component gives a further 12%, with 4% from the third principal component.  In these dimensions, there are three broad groupings: Tyres 1, 2 and 3 are separated from the rest in the first dimension, Tyres 5-10 in the second dimension, while Tyres 4, 5 and 10 separate in the third dimension.

Tyres 1, 2 and 3 are closely clustered and share a number of compounds in common that are less prevalent in the others, including toluene, 1,3-pentadiene and methylenecyclopropane.  The first two of these present potential toxic danger if swallowed or inhaled.  Tyres 5 and 10 occupy a similar position distinct from the other samples and disproportionately share 2-pentene and 1,4-cyclohexadiene, 1-methyl- in common, which present similar toxic risks.  A compound with a leathery smell that can be a respiratory irritant is phenol, 2-(1,1-dimethylethyl)-4-methyl-, which stands out in Tyre 8.

Interestingly, a particular compound, 1,4-benzenediamine, N-(1,3-dimethylbutyl)-N'-phenyl, was present in relatively high concentrations in Tyre 4 and present in all the others except Tyre 5, from which it was absent.  This compound is better known at 6PPD.  A study published in 2020 in Science6 linked 6PPD-quinone to the widespread deaths of coho salmon in California.  6PPD-quinone is produced by 6PPD reacting with ozone in the air.

The fourth question originally posed was what tyre wear particles released may be leeching into the environment over time.  While a microchamber or liquid extraction method may be more appropriate, we can nevertheless get some idea of what could be occurring from the pyrolysis results.  The two-dimensional chromatogram below gives a visual idea of the large number or organic compounds in a tyre, with each peak representing one compound.

These compounds can be put into generic groups to assist understanding.  In this case, we used two broad categories: aliphatics and aromatics.  The aliphatics contain ‘saturated’ (lots of hydrogen atoms compared to carbon) alkanes, alkenes, alkynes, carbonyls and acids.  The aromatics are mainly cyclic, based on benzene rings.

The table below shows the relative prevalence of compounds in these two groups.  As a broad statement, the alkane group contains many compounds that are often irritants to human organs, whereas the aromatics group includes many suspected carcinogens.  Therefore, at a simple level, the lower the proportion of aromatics the lower the human toxicity is likely to be.  To be more precise, the individual compounds contained in each group can be identified using this method, which may identify compounds toxic even at very low concentrations.

Overall, this initial data suggests there is good cause for significantly accelerated research on tyre wear emissions.  The rate of particle loss into the environment is orders of magnitude higher than from the tailpipe.  Knowledge of where the particles end up is currently limited.  As tyre manufacturers are naturally commercially coy about the chemical make-up of their tyres, it has been historically hard to assess whether the particles present a material problem of human and wildlife toxicity, as they sit and leech into the water and soil.  The method described above is one approach that is emerging that could help accelerate the understanding so badly needed.

---

References
1. Portals.iucn.org/library/node/46622
2. cyclohexene‚ 1-methyl-4-(1-methylethenyl)- ‚ (S)-
3. 9‚12‚15-Octadecatrien-1-ol‚ (Z‚Z‚Z)-
4. cis-5‚8‚11‚14‚17-Eicosapentaenoic acid
5. 2‚4-Hexadien-1-ol
6. Science.org/doi/10.1126/science.abd6951

The inevitability of hybridisation?

While the direction of vehicle powertrain policy and strategy is firmly oriented towards full battery electric vehicles (BEVs), the results of Emissions Analytics’ latest testing and lifecycle modelling suggests that hybridisation may prove to be the dominant outcome, whether intended or not.  As real-world emissions results collide with consumer preference and fiscal reality, the efficiency of hybrids in reducing carbon dioxide (CO2) emissions is likely to shine through over the next decade, even if BEVs and other new technologies come to dominate in the longer-term.

Put another way, at the moment it appears that battery electric vehicles (BEVs) are neither sufficiently clean nor a strong enough consumer proposition to achieve mass adoption without significant subsidy.  They are certainly not zero emission, not least due to the construction process and tyre wear emissions.  Hybrids, in contrast, reduce tailpipe CO2 emissions materially now – albeit somewhat less than BEVs – and have few utility disadvantages for consumers.  Therefore, would it be optimal to follow a hybridisation strategy for, say, the next ten years and then segue to BEVs after that, once they are cleaner and have fewer consumer disadvantages?

The typical objection to this is that BEVs already in the fleet will automatically become cleaner as the grid decarbonises.  While this is true, the greatest source of CO2 from BEVs is in the manufacture, which is fixed and incurred upfront.  Therefore, every BEV manufactured now may crystallise enough CO2 today to outweigh the subsequently lower CO2 of operation, compared to typical hybrid vehicles.

Taking the latest sales figures for new cars from the UK’s Society of Motor Manufacturers and Traders (SMMT), and putting them together with recent test results for exhaust and non-exhaust emissions from Emissions Analytics, it possible to evaluate the progress in decarbonisation.

Before turning to CO2, it can immediately be seen that BEVs deliver little overall advantage for air quality.  While NOx emissions remain positive for internal combustion engines (ICEs), the levels are significantly lower than for earlier models.  This is true to the extent that, if the car fleet were made up entirely of these latest diesel and gasoline cars, there would be no air quality legal violations.  For particle mass emissions, exhaust filters on ICEs typically reduce emissions to less than 1 mg/km.  In contrast, tyre emissions from ICEs are around 32 mg/km over a lifetime, whereas the BEV tyre wear rate – other things being equal – are 21% higher at 38 mg/km.  Adding exhaust and non-exhaust emissions together, BEVs are slightly higher emitting than ICEs and full hybrid electric vehicles (FHEVs).

With BEVs not required for air quality compliance, their primary environmental purpose is CO2 reduction.  The table below estimates the CO2 emissions saved compared to the benchmark gasoline ICE for each alternative powertrain, based on 16,000 km of annual driving and the latest market shares of each.  The greatest aggregate CO2 reduction is from BEVs, which have 8.4% market share.  Hybrids, collectively, with 34.0% market share, account for around three-quarters of the reduction of BEVs.

These BEVs include around 5.5 million kWh of battery capacity, compared to 1.0 million kWh in all the hybrids together.  Therefore, for every kWh of battery capacity, BEVs delivered 3.0 g/km of CO2 reduction, compared to 13.7 g/km for hybrids, both judged against the gasoline ICE benchmark.  In other words, hybrids have been 4.6 times more efficient at reducing CO2 as a function of the currently constrained battery material supply.  

Had the battery material from the 92,420 BEVs sold been used in hybrids – assuming average battery capacities of 60 kWh and 2.6 kWh respectively – an additional 2.1 million hybrids could have been built, enough to cover all new cars sold in the whole of 2021 in the UK.  In that hypothetical scenario, the CO2 reduction from the annual operation of the hybrid vehicles would be 940 kilotonnes greater than from the BEVs.

To analyse more fundamentally the differences between the powertrains, it is necessary to consider lifecycle CO2 emissions, including vehicle manufacture, operation and end-of-life processing.  To that end, Emissions Analytics has developed its own proprietary model; for the purposes of this analysis the key assumptions are:

• 15-year, 175 km vehicle lifespan

• 60 kWh BEV battery size

• 150 kg/kWh of CO2 in battery production

• 300 g/kWh of CO2 for BEV charging.

Considering the cumulative CO2 emissions to 2070, it is possible to compare six different scenarios as shown in the table below.  The three powertrain scenarios are an ICE-led baseline, a direct migration to BEVs, and an interim switch to FHEVs until 2030 and then migration to BEVs.  Each of these scenarios has two versions: one calculated based on current CO2 intensity of electricity generation and BEV manufacture, the other with that intensity reducing by 50% from 2030.

The reason for BEVs being hardly better than FHEVs on current CO2 intensity is that the manufacture emissions of the vehicle and battery are still relatively high, and the average grid electricity across Europe still includes significant gas and coal.

Therefore, we can deliver an extra 8% point reduction in CO2 compared to the baseline by switching straight to BEVs, if a 50% reduction in CO2 intensity is achieved from 2030.  According to the fifth carbon budget under the UK’s Climate Change Act, the country can emit 1,725 million tonnes of CO2-equivalent between 2028 and 2032.  The added benefit given by the direct migration to BEVs is over 10% of that carbon budget – which demonstrates how every percentage point of reduction in CO2 is important.  

On the surface of this analysis, BEVs look like the optimal strategy, even though the gap to the FHEV strategy is closer than reported elsewhere.  However, this neglects consideration of risk.  Rolling out hybrids would be relatively low risk due to limited resource requirements, consumer resistance and taxpayer subsidy requirement.  Furthermore, the BEV estimates of CO2 reduction are sensitive to many factors, including:

• Speed and degree of decarbonisation of battery and vehicle manufacture

• Speed and degree of decarbonisation of the grid

• Improvements in battery energy capacity and therefore vehicle range

• Longevity of batteries and BEVs as a whole

• Geopolitical security around scarce battery materials

• Environmental and ethical issues around mining

• Degree to which vehicle miles travelled in BEVs replace ICE miles, or are additive due to the lower marginal cost

• Ability to develop a transparent and standardised lifecycle model to be able objectively to verify CO2 reduction claims.

Each of these could have a material impact on the analysis and resulting CO2 reduction, both positively and negatively.  However, as high-certainty methods of CO2 reduction are a pressing policy need, it may be better to ‘bank’ the lower-risk 24% reduction from ten years of hybrids, and then migrate to BEVs and other lower-CO2 powertrains.

Returning to the topic of currently scarce battery materials, the FHEV strategy requires 39 GWh of battery capacity for vehicles sold in the ten years to 2030.  In contrast, the BEV strategy requires 570 GWh, or 15 times more.  From a practical, ethical and geopolitical point of view, this is significant.

There is, further, a paradox with BEVs: once bought, from a CO2 point of view it makes sense to drive them a lot so the embedded emissions in the construction can be amortised across the maximum usage.  It may even be better to drive incremental miles in the car rather on some forms of public transport.  This is, therefore, linked to the question of taxing BEVs such there are no perverse incentives to drive on the road more, and creating replacement revenues for the declining taxes on gasoline and diesel.

Putting together this rate of CO2 reduction and the current utility compromises, it suggests that BEVs are not yet good enough value a product to get rapid adoption without significant subsidy.  With new internal combustion engines now sufficiently clean that they are not contributing to air quality violations and the readily available alternative in hybrids, it appears that the optimal policy is to concentrate on them for the coming years while BEVs and other low-CO2 powertrains get ready for mass adoption.  Greater competition would be good for consumers in the long term, especially where greater proportions of the added value in the vehicles arise domestically, whatever the country.  Vast sums of taxpayer subsidy – billions, if not tens over billions, of pounds over time – could also be avoided.

Failing this, it is likely that we end up at hybridisation by another route: at the household level.  This is because consumers may hold on to old ICEs to cover longer journeys, heavier payloads, sporty driving and the like.  These older vehicles may also have higher NOx emissions, being Euro 6 prior to the introduction of Real Driving Emissions, or earlier.  Governments may try to force these off the road through higher taxation, but this is unlikely to work as the low depreciation rates of these older vehicles will make for cheap motoring in almost any scenario.  

In summary, surely it would be better to take the pragmatic route and hybridise everything as soon as possible?

This conclusion is not dissimilar from that articulated in a facsheet from the International Council on Clean Transportation in July 2021, which said, "Hybridization can be utilized to reduce the fuel consumption of new internal combustion engine vehicles registered over the next decade, but neither HEVs nor PHEVs provide the magnitude of reduction in GHG emissions needed in the long term."

But, the numbers suggest we replace "can" with "should", to lock in the CO2 savings now.

---

The rise of unregulated exhaust pollutants

But which ones really matter on the road?

When the various pandemic lockdowns across Europe failed to bring about the overall improvements in air quality that might have been expected, Emissions Analytics’ interest focused in on volatile organic compounds (VOCs) and their potential role.  While nitrogen oxide (NOx) emissions fell with traffic levels, often ground-level ozone (O³) rose, leading to similarly bad air quality from a human health point of view, just of a slightly different complexion.  In fact, this should not have been a surprise as the complex interaction of NOx, VOCs and O³ has long been studied¹.  The South Bay in Los Angeles has grappled with this problem since motor vehicles proliferated, and many US air quality regulations have stemmed from the experiences there.  

Recent newsletters from Emissions Analytics have, therefore, looked at sources of VOCs including vehicle tyres (What's in a tyre?) and materials (Euro 8: Rethinking Vehicle Emissions Fundamentally).  This newsletter returns to the tailpipe to see what VOCs and other currently unregulated pollutants are being emitted in real-world, on-road driving.  This is aimed at taking understanding beyond the ‘total hydrocarbons’ that are regulated using a laboratory test in Europe, and non-methane hydrocarbons that are regulated in a number of territories including the US.  This research is being conducted against the backdrop of the current discussions around the proposed new ‘Euro 7’ regulation, which may include some hitherto ignored pollutants including particular species of VOCs.

Measuring a wide range of organic compounds and other volatile species at the tailpipe is a challenge due to the large number of different compounds – many hundreds, if not thousands – as well as their volatility, which can make them hard to capture.  While this can be done in the laboratory, it is an even greater challenge on the road.  Traditional portable emissions measurement systems (PEMS) measure total hydrocarbons (THC) using a flame ionisation detector (FID).  This can deliver robust measurements, but it creates some operational challenges, not least from the need for a supply of combustion gas.  Furthermore, only a single figure for total hydrocarbons is produced; it does not include non-hydrocarbon VOCs and does not separate the different species of hydrocarbons.

To address these challenges and limitations, Emissions Analytics has developed a proprietary, patent-pending system, that harnesses sample collection from the exhaust onto tubes while driving, which are then analysed later using laboratory gas chromatography.  Using this, we can measure the concentrations of VOCs as well as semi-volatile organic compounds (SVOCs) – together covering compounds from with two carbon atoms (C2) up to at least 44 carbon atoms (C44) – formaldehyde (CH2O), nitrous oxide (N₂O), sulphur dioxide (SO2) and many others.  Therefore, both the breadth of compounds measured and the speciation challenges are solved.  Furthermore, the measurements can be highly sensitive, picking up very low concentrations, which may be critical for highly toxic species.  The chromatogram below illustrates the large number of distinct compounds that are present in a typical exhaust, with the height of the peaks generally indicating relative amounts.

When deployed together with a traditional PEMS unit, with its capability for measuring total exhaust flow, the concentrations of VOCs can be turned into mass values.  Combined with the GPS speed data, the distance-specific emission rates can be calculated, giving mg/km figures as is the basis for regulating most gaseous emissions.

The limitation of this approach is that the sample collection on tubes is cumulative over the test cycle and, therefore, there is no second-by-second signal.  This creates two problems.  First, when the average concentrations are multiplied by the total exhaust flow, the result is biased due to the highly variable nature of both the target gas concentration and exhaust flow at the instantaneous level.  Second, the result is a combined value for the whole test cycle, without any breakdown between different driving modes.  

Our approach addresses both limitations.  The sample bias issue is overcome using a proprietary on-board constant volume sampling and proportional flow dilution system.  To give a breakdown of the combined cycle into useful sub-sections, a geofencing system automatically switches between different sample tubes to sample, for example, urban, rural and motorway driving separately.

A notable advantage of the sample tube approach, from a practical and analytical point of view, is that it separates sample collection from sample analysis.  This reduces the complexity of the vehicle test itself, which improves the success rate.  Having the sample captured on a tube means that it can be analysed later, in batches for efficiency, and each sample can be analysed multiple times, which is useful for validation and uncertainty analysis.  For the purposes of our tests, we use a two-dimensional gas chromatography (GCxGC) system coupled with a time-of-flight mass spectrometer (TOF-MS) from SepSolve Analytical and Markes International.  The GCxGC achieves a separation of the hundreds of compounds that would not be possible in a one-dimensional system.  The TOF-MS is crucial for identification of the compounds, as well as quantification, which is aided by other detectors such as a FID and electron capture detector (ECD) for N₂O.

Putting these techniques into practice, Emissions Analytics tested eight recent gasoline vehicles in Europe.  All were 2020 or 2021 model years, with four standard internal combustion engines, two mild hybrids, one full hybrid and one plug-in hybrid – drawn from eight different brands.  All were tested on the standard EQUA Index test cycle set out in previous newsletters and the basis of all data in the Emissions Analytics’ subscription database.  While similar to a certification Real Driving Emissions (RDE) test, it has a wider range of dynamic driving and is about twice as long.

The N₂O results are set out in the table below, split between urban, rural and motorway driving.  In each case the highest and lowest emitting cars are highlighted.  Greatest emissions were seen in rural driving, with urban driving the lowest.  

Emissions of N₂O are potentially important as the gas is a much more powerful greenhouse gas than carbon dioxide (CO2).  Over a 100-year horizon, it has warming potential 265 times greater2.  Therefore, small amounts of N₂O could undermine the extensive efforts to reduce primary CO2 from engines.  On average, across the eight vehicles and three driving modes, the N₂O emissions were 2.1 mg/km.  Converted to an equivalent CO2 this is just above half a gram.  Real-world CO2 emissions of these test vehicles averaged 142 g/km.  Therefore, these N₂O emissions were equivalent to well below 1% of total CO2 – within the measurement error of the CO2 value.

Formaldehyde is a pollutant of concern as it is believed to be carcinogenic and causes a wide range of irritation in humans, including to skin, eyes and lungs.  The results from the same eight vehicles are shown below.

Again, highest emissions were seen in rural driving.  Although the exact human health effects depend on factors such as the dilution and dispersion of the emissions, it can be seen from the data that there is about a factor of four difference between the cleanest and dirtiest cars.

Turning to other VOCs and SVOCs, the tubes captured over 500 different compounds from each vehicle.  Some of these were common to most or all, but other compounds were characteristic of specific vehicles.  Taken together, each vehicle has its own chemical signature.  The table below shows the top compounds that were common to each vehicle, together with their toxic effects. It should be noted that the actual effects on humans depend on the concentrations experienced.

By way of contrast, the differentiating compounds are shown below.  The compound listed is the most abundant chemical with particular prevalence in that vehicle.

On this first pass, therefore, there is good reason to move beyond the simple measure of total hydrocarbons and non-methane hydrocarbons in various regulations around the world.  The next stages are to consider the absolute quantities of the compounds, model their dispersion in the environment, understand their toxic effects, and study their propensity to create secondary organic aerosols, i.e. solid airborne particles created as the SVOCs react in the atmosphere.

These initial results demonstrate the capability to identify and measure a wide range of VOCs and SVOCs in real-world driving – compounds that can have a wide range of deleterious effects on human health and the environment.  The N₂O results may call into question the priority of adding this pollutant to the new Euro 7 regulation as the effects on global warming may be insignificant and come at the price of higher priced cars with the added regulatory burden.  Better to focus on the ongoing effects of VOCs, whether the direct effect on humans and the biosphere, as precursor to ozone and smog, or as they lead to formation of airborne particles matter – which we will look at in a later newsletter.

---

1. Chemistryworld.com - The weekend effect
2. IPCC Report

What’s in a tyre?

What’s in a tyre?

Thousands of chemicals derived from crude oil, many of them volatile organic compounds, apparently.

This matters because tyres shed a lot of material into the environment.  Emissions Analytics’ estimates suggest that around 300,000 tonnes of ‘rubber’ are released every year from passenger cars in Europe and US alone, the equivalent of over forty million brand new, entire tyres.  These particles go into the air, soil and watercourses.  If you were to stack all the tyres manufactured in the world in a year on their side, it would reach the moon.

---

While there is a tyre labelling scheme in the EU, it only rates rolling resistance, wet grip and noise.  These are clearly vital to the performance and safety of tyres, but it leaves the ratings blind to the environmental consequences of the tyre wear emissions. There are restrictions on toxic chemicals that can be included in the manufacture of tyres under the European REACH regulations, but the number of chemicals affected is limited.

While we have been occupied with reducing exhaust emissions to control air quality problems, other sources of pollutants have not received the same attention historically.  Now that tailpipe emissions of modern internal combustion engines (ICEs) in both Europe and US are generally well below regulated limits for pollutants, focus is now turning to ‘non-exhaust emissions’, which cover tyres.  Emissions Analytics’ testing shows that, in normal driving, tyre wear emissions are about one hundred times greater than tailpipe particle mass on a modern ICE vehicle.  In legal but extreme driving, enough to reduce significantly the lifespan of a vehicle’s tyres, that factor increases to around one thousand. 

In addition to tyres, non-exhaust emissions cover material from brake and road wear, as well as resuspended solids, whipped up from the carriageway by the moving vehicle.  Of these, tyre wear emissions are probably the largest and fastest-growing component.  Brake wear emissions are forecast to fall as regenerative braking becomes more widespread.  Road wear and resuspension rates are only partly related to the passing vehicle, including its weight, but are probably more determined by the road material and condition, and what particles are blown onto the road from multiple surrounding sources.  Tyre wear emissions are likely to grow as vehicles continue the long-term trend of becoming heavier, although this may at some point be offset by using more lightweight construction materials.

Understanding tyre wear emissions provides a challenge as they are heterogenous.  Unlike, for example, nitrogen oxide (NOx), which is a unique compound that can be measured as a mass or volume, particles from tyres come in an infinite combination of shapes, sizes and densities.  Moreover, the particles are made up of a wide array of chemical compounds, and these chemicals may also stick – or adsorb – to the surface of the particle.  In this way, particles can act as the distribution vector for other compounds.

An emerging approach to characterising tyre wear emissions is, therefore, to measure both the wear rates and chemical make-up of the particles.  This enables a quantification of the amount of individual chemicals that are released into the environment. This information can then be put together with toxicity ratings to assess the potential effect on human health, wildlife and biodiversity.  For the semi-volatile organic compounds particularly, the effect of these on secondary organic aerosol formation – in other words, gaseous emissions that condense to become airborne particles over time – can be evaluated.

To begin to understand the degree and nature of the tyre wear emissions problem, Emissions Analytics recently tested a range of different tyres.  Full sets of tyres of eight different brands and types were selected and installed on the same test vehicle, a 2012 rear-wheel drive Mercedes C-Class.  The wheel alignment and tyre pressures were checked.  Each set of tyres was driven for over 1,000 miles, around 90% by distance being conducted on the motorway.  The four wheels – i.e. leaving the tyres on the rims to avoid damage – were weighed at the start and end, and the distance-specific loss of mass was calculated.  The results are shown in the chart below.

Across the brands, the average mass loss was 64mg/km for the vehicle, adding all four tyres together.  Wear on the rear tyres was greater, accounting for 71% of the total on average, strongly influenced by this being a rear-wheel drive car.  The wear rate on the fastest abrading tyre was 2.3 times higher than the slowest.  Therefore, tyre choice by manufacturers and consumers can have a material impact on emissions rates.

The wear rate is faster when tyres are new, for the first few thousand miles.  Thereafter, the wear rate appears to decline at an approximately logarithmic rate.  Over a lifetime, therefore, the average wear rate may be half the figures above.  If we assume that the average vehicle travels around 16,000km per year, the rates above mean that each car sheds around 0.5kg per year on average over its lifetime.  As there are almost 600 million vehicles in Europe and the US, this is equivalent to 300,000 tonnes of particles.  An average tyre weights around 8kg, hence the total amount shed is equivalent to almost 40 million whole tyres. These figures do not include tyre wear from heavy-duty vehicles, which would also be significant.

The effect of particles on human health and the wider environment is an on-going and active area of research.  It is complex to isolate the causal links.  In terms of air pollution, it is generally accepted that there is a connection between particle mass emissions and diseases such as cancer and heart disease from prolonged exposure.  The effects in terms of particle number are less clear-cut, although the EU regulates these at the tailpipe from a precautionary motive.  The aim of this newsletter is not to review the evidence, but rather to contribute early findings on the chemical composition of tyres studied so far by Emissions Analytics.

Tyres are highly-engineered products and made up of a complex mixture of substances.  For light-duty vehicles, the majority of the content of the tyre tread and walls comes from crude oil derivatives, with only a minority of natural rubber.  Therefore, to understand the composition of tyres, it is necessary to employ a technique that can separate these out.  We decided to focus on the organic compounds rather than metals, and employed our two-dimensional gas chromatography equipment coupled with a time-of-flight mass spectrometer (GCxGC-TOF-MS from SepSolve Analytical and Markes International, see https://www.emissionsanalytics.com/tyre-emissions).  The gas chromatograph achieves separation by passing a sample through a long ‘column’, and the mass spectrometer does the compound identification.  Two dimensions, both of time, are required to separate compounds that ‘elute’ in the same place in a one-dimensional chromatogram.  Taking an example tyre, we heated samples to 100˚C and analysed the compounds released to obtain the following two-dimensional chromatogram.

Broadly, compounds cluster in different areas depending on common chemical characteristics.  Some frequently used groups are illustrated above.  Alkanes (e.g. pentane) typically affect the lungs, liver, kidney and brain.  Cycloalkanes (e.g. cyclohexane) lead to headaches and dizziness.  Terpenes (e.g. limonene) are generally less problematic and are responsible for aromas, unlike aromatics (e.g. benzo(a)pyrene), which are often carcinogens, as are nitrogen-containing compounds (e.g. quinoline).  This is a significant simplification for the purposes of illustration.

Each shaded area on the chromatogram indicates a distinct chemical, with the intensity of the colour reflecting its abundance. The measurement breadth of the equipment is from compounds containing two carbon atoms (C₂) to at least C44.  This covers what are called volatile organic compounds (VOCs) and semi-volatile organic compounds (SVOCs).  Just this one sample contains well over a thousand distinct compounds of these types.

The next stage is to identify as many of the compounds as possible and understand if they are problematic for health or the environment.  Zooming in on the nitrogen-containing part of the chromatogram, it is possible to identify a number of potentially problematic compounds, as shown below.

N-Phenyl pyrrole, quinazoline, 4-tert-butyl-2-chlorophenol at certain levels of exposure can lead to symptoms in humans including skin, eye and respiratory irritation.  In addition to these effects, quinoline and 3-methyl-quinoline have potential carcinogenicity and mutagenicity in humans.  Quinoline and 1,2-dihydro-2,2,4-trimethyl affect aquatic environments more than humans.  

This suggests that potentially concerning compounds are present in tyres, but if we compare the composition of different tyre brands it is also possible to see that the mix of chemicals differs.  This reflects the many formulations used by different producers, but also means that tyre selection can lead to different environmental and health effects.  The chart below illustrates the point by comparing four different tyre types.  Each sample was pyrolysed to release as many compounds in the underlying materials as possible, and then analysed using Principal Component Analysis.  

Tyre Brand B is strongly differentiated from the other three tyres by the presence of 1-methyl-2-pentyl-cyclohexane – a cycloalkane.  Although it does not have any particular toxic indications for humans, it is potentially possible to relate the presence of this defining compound to other characteristics such as rolling resistance, noise or wet grip.

Where toxic compounds are identified by this approach, it does not guarantee that they are present in amounts that could cause harm.  Therefore, the final stage is to quantify each in the sample, so the total amount in a tyre can be worked out.  However, even if the amounts are small in one tyre, due to the large amount of material released each year in total – as calculated above – even low concentrations could lead to deleterious effects at the macro level.  

All in all, this initial testing has demonstrated that it is possible to measure tyre wear explicitly, without it being combined with brake or road wear, and the separation capability of the two-dimensional gas chromatograph can help identify thousands of constituent compounds.  The results themselves then show that there are relevant and material differences in the wear rates and chemical make-up of different brands and models of tyre.  Therefore, choices of tyre when the car is first sold and at subsequent tyre changes are directly relevant to the vehicle’s environmental impact, and requires deeper and urgent study.

---

Euro 8: Rethinking Vehicle Emissions Fundamentally

Just as European regulators are consumed by finalising the touted 'Euro 7' regulation, it is perhaps the right time to consider the longer-term trajectory for vehicle regulation.  There is a tension in the narrative around Euro 7.  At points it has been talked about as the “final” set of emissions regulations, but now the message is being put out that the intention is not to “kill” the internal combustion engine (ICE).  Unless foreshadowing the end of the ICE, it would be bold and presumptuous to call Euro 7 the final regulation.

In reality, Euro 7 is unlikely to be the final regulation, not because of the end of ICEs, but because vehicles of all types, including battery electric vehicles (BEVs) emit lots of substances from sources other than the exhaust.  This newsletter set out those additional sources, and what shape a subsequent Euro 8 regulation might conceptually take.  This is consistent with Emissions Analytics’ mission to establish the true, real-world environmental impact of vehicles.  BEVs, it should be kept in mind, exist only for their greenhouse gas reduction potential, not for meeting air quality laws, as the regulated pollutant emissions out of the current generation of ICEs are low.

Why are BEVs not required to meet air quality laws in Europe?  Since the introduction of the Real Driving Emissions (RDE) on-road validation of certification values introduced mid-way through Euro 6,gaseous tailpipe pollutant levels are now confirmed as low in average driving.  For some, they were low even before RDE.  The average nitrogen oxide (NOx) emissions from RDE diesel vehicles are 45mg/km and falling, compared to the regulatory limit of 80mg/km.  Mean carbon monoxide from RDE gasoline vehicles is 157mg/km compared to the limit of 500mg/km.  Total hydrocarbons (THC) from RDE gasoline vehicles are now typically below 10mg/km compared to the limit of 100mg/km.  Particle mass emissions are now very close to zero on current diesel and direct injection gasoline vehicles, with a limit of 4.5mg/km on the PMP procedure.

Of the regulated pollutants, therefore, all seem well controlled, except for ultrafine particle emissions – measured through the particle number (PN) standard – and potentially particles more generally from port fuel injection gasoline cars.  The widening of the size range of particles measured, down to 10nm, likely to be proposed in Euro 7, will offer a valuable tightening of this part of the regulatory regime.  

For the Euro regulations themselves, greenhouse gases are not regulated – except indirectly for methane (CH4) through THC – as the fleet average carbon dioxide (CO²) targets act separately.  Whether promoting BEVs is the optimal way to achieve these targets was the subject of a previous newsletter: www.emissionsanalytics.com/news/hybrids-are-better.  Relevant here is the potential addition of methane and nitrous oxide (N2O) under Euro 7, even though they are primarily greenhouse gases rather than air quality pollutants.  They are both more potent greenhouse gases than CO², although survive for a shorter time in the atmosphere: 28 and 265 times higher potency respectively on a hundred-year time horizon1.  Therefore, only small amounts of N2O emissions could nullify much of the hard-won CO² reduction.

So, on the surface of it, Euro 7 could be the last regulation for pollutant emissions, while greenhouse gases are actively addressed through the CO² targets.  What may be neglected, however, is volatile organic compounds (VOCs) from the exhaust.  These compounds are numerous, small in volume but potentially highly toxic in their human health effects.  Therefore, their current regulation in the laboratory as non-methane hydrocarbons (NMHC) may be insufficient: not only that the limit is high at 68mg/km, but also that it does not apply to diesel vehicles.  Furthermore, by only considering the total, there is no visibility on whether that total is made up of toxic or innocuous compounds.

The University of York in the UK has been at the forefront of studying this area, highlighting particularly the role of these compounds in the atmospheric chemistry that leads to ground-level ozone and secondary organic aerosol particle formation².  Therefore, these compounds do not just have direct effects on human health, but indirectly lead to poor air quality.  While oxidation catalysts in the exhaust of gasoline vehicles may be highly effective in converting VOCs, their effectiveness against the heavier, semi-volatile organic compounds (SVOCs) such as the carcinogenic polycyclic aromatic hydrocarbons (PAHs) is less clear.  Generally, also, these ICE vehicles suffer from relatively high VOC and SVOC emissions when the engine is cold.

From this, we can conclude that Euro 7 perhaps should not be the final regulatory stage, so long as ICEs are still sold, which is likely to be through to at least 2035 in Europe.  While Euro 7 is looking at regulating a small number of highly volatile compounds such as formaldehyde, the broader spectrum of organic compounds is not currently being actively considered.  This is important when you put it in the broader picture of the environmental impact of vehicles.  Specifically, there are instructive parallels with the emissions from tyres and from materials inside the vehicle cabin – both of which are very lightly regulated currently.  Tyres will be a topic of a later newsletter, so here we will focus on the vehicle interior.

In our last newsletter (From performance to experience), we reviewed the evidence for concentrations of particles and CO² in the cabin during on-road driving.  It concluded that the quality of air inside the cabin is highly dependent on the quality of the ventilation system and its filter.  For some vehicles, in-cabin particle concentrations were many times higher than outside on average, and the use of the recirculation mode led to steep rises in CO² concentrations to the point of potentially having cognitive effects on the driver.

But that is not everything.  The single biggest complaint from new car buyers in China is about the ‘new car smell’, which is caused by a mix of VOCs.  These VOCs may have health effects that go well beyond simply causing malodours.  The source of the new car smell is VOCs being released from interior materials and glues used to put the vehicle together.  Sources of VOCs go beyond that too, including those that enter the cabin from outside (which in turn come from other vehicles, home heating, industrial sources and so on), fuel evaporating from the tank and emissions from the body and clothing of human occupants.  What is important is to be able to differentiate the toxic from the harmless VOCs – the toxic ones being more likely to come from combustion of fossil fuels or materials derived from fossil fuels, such as plastics and adhesives.

To study this, Emissions Analytics has recently opened a testing laboratory with two-dimensional gas chromatography and time-of-flight mass spectrometry from Markes International and Sepsolve.  This allows us to test for tailpipe and in-cabin VOCs with excellent separation, identification and quantification.  Taking one of the early car tests, a sample of air was taken and 617 different compounds were identified, with 25 being particularly being abundant.  These compounds can be illustrated on a two-dimensional chromatogram, as shown below.  The horizontal dimensions are the two axes of separation and each peak represents one compound, with the area under the peak broadly reflecting the amount present.  

Looking in more detail at the most abundant compounds, it is possible to relate each to the risks to human health, as shown in the table below, with a colour classification reflecting the toxicity.

In addition to these, formaldehyde was identified through a separate process using high-performance liquid chromatography. Formaldehyde is another respiratory irritant and a contributory factor in asthma and some cancers.

The vehicle was a recent model year mid-range European gasoline passenger car.  These effects are a combination of irritants that would lead to various degrees of discomfort while in the vehicle.  A longer time spent in the car would correlate with a greater health or comfort effect.  The 617 compounds identified can be put into the following generic groups, corresponding to the coloured areas on the chromatogram for ease of interpretation (with the axes being the two dimensions of separation and the colour corresponding to the intensity of the peak).  These groupings give an area percentage which can be used for the characterisation of compounds.  This is a simple interpretation, while more detailed analysis can quantify the amount of each compound.  

Therefore, this vehicle contained a large number of compounds that are respiratory irritants and cause organ damage, together with a material number of carcinogens.

China is planning to regulate eight targeted compounds in the vehicle interior, although the timetable for implementation has slipped.  Emissions Analytics identified six of the eight target compounds in its test, particularly formaldehyde and toluene. Moderate concentrations of benzene and ethylbenzene were present, with lower amounts of acetaldehyde and styrene.  Japan, Korea and Russia also regulate similarly limited repertoires of these VOCs.

The main control in Europe for these VOCs is through the ‘REACH’³ restrictions on substances that can be use in product manufacturing.  Currently only eight PAH compounds are restricted – according to their concentration in the final product – primarily as carcinogens: benzo(a)pyrene, benzo(e)pyrene, benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(j)fluoranthene, benzo(k)fluoranthene and dibenzo(a,h)anthracene.  These restrictions apply to products including tyres and plastic components than come in contact with the human skin, including some parts of the vehicle interior.

The most progressive approach to measuring vehicle interior VOCs, focusing on materials, is the Vehicle Interior Air Quality group at the UNECE.  In 2020, it finalised a revision to Mutual Resolution 3 that standardises measurement of these compounds and offers a platform for future regulation by national governments.  Substances tested under this protocol are VOCs ranging in volatility from n-C⁶ to n-C¹⁶ together with carbonyl compounds formaldehyde, acetaldehyde and acrolein.

In summary, both the tailpipe and vehicle cabin are likely to contain a wide range of volatile and semi-volatile organic compounds, some of which can have high toxic effects on humans, as well as other effects on animals, aquatic life and biodiversity.  This is an important and urgent topic for future study, especially at a time when the levels of traditionally regulated tailpipe pollutants are getting so low.  In short, the focus should be on particles, VOCs and greenhouse gases, whether they come from the tailpipe, cabin materials or tyres

Unless this wider view of vehicle emissions is taken, and the structure of the Euro regulations fundamentally reframed, they could hold as much currency as a 7 Euro coin.

---

Footnotes:

1. https://www.ghgprotocol.org/sites/default/files/ghgp/Global-Warming-Potential-Values%20%28Feb%2016%202016%29_1.pdfof-air-pollutants/emissions-of-air-pollutants-in-the-uk-particulate-matter-pm10-and-pm25, PM2.5
2. http://www.envchemgroup.com/jacqui-hamilton.html
3. https://echa.europa.eu/regulations/reach/legislation

From performance to experience

Could vehicle interior air quality become the differentiator of the future?

Cars used to be about speed, power, performance and freedom.  Different models often used to offer dramatically different performance and looks.  But, not so any more.  We are currently in an era of the generic sports utility vehicle (SUV) and identikit low emission vehicles.  Electric vehicles “all look alike”, as suggested by the head of BMW recently¹.

But can this last?  Will consumers want to buy such seemingly bland offerings, and can manufacturers profit from the undifferentiated?  Perhaps we are already beginning to see  early signs of a shift.  Electrification may turn out to be less about emissions reduction, but rather emblematic of the change in the fundamental proposition of the car away from performance and towards experience.  Performance is increasingly constrained through traffic and emissions policy, so consumers may want to make the increasingly dull experience of driving at least more comfortable.

Historically, operating a car delivered ample private benefit and enjoyment, at the expense of an alarming array of environment and health impacts: climate effects through carbon dioxide (CO²) emissions; air quality from tailpipe nitrogen oxides (NOx), particulates and carbon monoxide (CO); air and marine pollution from tyre and brake wear; ozone formation from evaporation of volatile organic compounds (VOCs) from fuel in the tank and construction materials; and noise and many others.  

Battery electric vehicles (BEVs) are perceived to be the antidote to this: quiet and pollution free.  This is of course not quite true: BEVs create CO² emissions in their manufacture, some noise and perhaps higher tyre wear emissions.  Nevertheless, we could get to the point where the environment affects the BEV driver more than the BEV affects the environment.

How could this be true?  Road transport collectively is only a minority contributor to air quality problems in 2019 – perhaps only 12% of particles² and 33% for NOx³.  The majority of pollution comes from domestic heating, industrial sources and agriculture. This polluted air can enter the vehicle cabin through its ventilation system, exposing the driver to the resulting health risks and discomfort.  Therefore, it may well become increasingly the case that the car driver is more a victim of pollution than the cause.

Even without BEVs, there are already some aspects of the internal combustion engine (ICE) which foreshadow this trend.  First, due to the efficacy of exhaust after-treatment systems, the levels of CO and NOx are so low in real-world operation that the impact on the environment is negligible.  For example, under the latest European Real Driving Emissions (RDE) regulation, the average real-world emissions are 157mg/km from gasoline vehicles and 36mg/km from diesels for CO; and 9mg/km from gasolines and 45mg/km from diesels for NOx.  Second, as demonstrated previously by Emissions Analytics, diesel particulate filters (DPFs) are so efficient that there are often fewer particles coming out of the tailpipe than there are in the air of a polluted city such as London⁴.

With this background, consider the scenario where the powertrain element ceases to be a major differentiator between mainstream BEVs.  Power may be limited to maximise range, torque could be capped to reduce tyre wear emissions, and tyres may become skinnier to reduce rolling resistance but at the cost of handling.  As and when connected and autonomous vehicles hit the road, this pattern may become even more pronounced.  Furthermore, the cost of electricity – as low as three cents per kilometre – could mean the operating costs become almost irrelevant.  Mainstream cars cease to have a performance dimension.  Add to this the relatively few standardised manufacturing platforms and you have vehicles of increasingly similar performance.  In this world, how will manufacturers differentiate their products and make decent profits?

Design undoubtedly will remain a key element, both for aesthetics, build quality and cost.  Beyond that, with the background of historical air pollution problems and now Covid-19, it may well become the vehicle interior air quality that becomes a major differentiator.  What once were major sources of pollution, could now become protective automotive bubbles.

Tesla’s launch of its ‘Bioweapon Defense Mode’ in 2016 was perhaps early evidence of this trend.  The data presented by the company for its efficacy involved exposing a vehicle to high levels of particle pollution in an emissions chamber and showed that concentrations of particles by mass⁵ in the vehicle cabin fell from 1,000ug/m3 to undetectably low levels within two minutes, from which they concluded it could protect the vehicle occupant from biological attack⁶.

Grand claims need verifying, especially as to whether laboratory test results carry over into real-world conditions, so Emissions Analytics tested this model on a 2019 Tesla X in the UK.  The test followed the protocol set out in an SAE paper published in 2019⁷.

The Tesla X is now equipped with a High-Efficiency Particulate Air (HEPA) filter as standard.  In simple terms, the ventilation system has the typical ‘fresh air’ and ‘recirculation’ modes, but also the bioweapon mode too.  This HEPA filter is enormous, as shown in the picture below – the installation requiring most of the width of the ‘frunk’.  It is approximately 100cm long, 30cm tall and 3cm deep – a volume of over 9,000cm³.  This compares to filter volumes on typical mass market cars of 1,000-2,000cm³.

The filter may be large, but it certainly works.  In our test, interior concentrations were 94% lower than externally on fresh air mode, and 92% lower on the bioweapon mode during an on-road test – statistically indistinguishable from one another.  This is the best performing vehicle we have tested so far.  The principal difference between Tesla’s test and ours was that we measured ultrafine particles down to 15nm rather than particle mass.  Together, the data suggests excellent protection from both bigger and smaller particles.

To put these results in context, Emissions Analytics tested 97 recent model year light-duty vehicles in the US in partnership with Edmunds⁸.  Of all these, the best performing was a 2019 Honda Civic, which reduced particle concentration by 73%.  Emphasising the significant differences, the worst performing vehicle was a 2019 Lexus ES, for which particles inside the vehicle were 254% higher than outside.  Of the 97 vehicles, 44 had higher concentrations inside than out.  This is an unregulated area at the moment, so there is no compliance issue, but there certainly is a potential health issue from chronic particle exposure.

A common driver strategy for protection from exterior pollution – often initiated when the driver senses a malodour – is to engage the recirculation mode on the ventilation system, which wholly or largely circulates, with varying degrees of filtration, existing interior air.  This is effective in protecting from pollution ingress from outside, but has the side-effect of allowing CO² to build up in the cabin from the respiration of occupants.  Although research is scarce in driving situations, the effects of elevated CO² on cognition have been shown, which leads to the reasonable belief that above 1,000ppm – compared to a background of just over 400ppm – there may be effects on driver safety as well as comfort⁹.

Returning to the Tesla, on fresh air mode, CO² increased by just 8%, while on recirculation the increase was 97%.  Impressively, the bioweapon mode saw just a 17% increase.  The average increase across the 97-vehicle test on recirculation mode was 15% on fresh air and 79% on recirculation.

From this, it is clear that there is a trade-off between protecting the vehicle occupant from particle ingress into the vehicle on fresh air mode and CO² build-up if recirculation is engaged.  While this holds true at the level of the individual vehicle, it is not true at the market level.  In other words, there are some models that are good at both, and some bad at both.  The chart below plots each of the 97 models tested, plus the Tesla X, for particle infiltration on fresh air mode against CO² build-up on recirculation.

For particle number, the values are the ratio between concentrations in the interior and exterior air, so a value of one means they are the same on average.  For CO², a value of zero means there is no increase in interior concentrations compared to the baseline.  On this latter measure the Subaru Impreza showed the worst performance, in contrast with the best from the Chevrolet Suburban – as indicated on the chart.

In short, there is a wide diversity of results and no obvious pattern, whether it be by manufacturer, vehicle size or powertrain.  The most conspicuous activity currently is from premium manufacturers, whose buyers perhaps have the greatest awareness or appreciation of clear cabin air.  Beyond that, there appears to be little understanding of the issue in the absence of any useful consumer information.

To help address this, Emissions Analytics is actively involved in a group aiming to standardise the measurement methodology for in-cabin pollution¹⁰.  The group was initiated by the AIR Alliance, which already publishes ratings for tailpipe pollution¹¹.   Along with Emissions Analytics, the group includes a number of vehicle manufacturers, academics, and filter and ventilation suppliers. Once completed, recognised and repeatable testing should enable ratings to be published that will both inform car buyers and, indirectly, incentivise manufacturers to improve interior air quality.

The Covid-19 pandemic brings these issues to the fore, as the virus is a particle of approximately 100nm in size.  While ingress of live virus particles from outside into the vehicle is unlikely, reducing the chance of one infected occupant transmitting the virus to another is more relevant.  Therefore, the rate at which air is recirculated and re-filtered matters.  In this context many initiatives have appeared from vehicle manufacturers, suppliers and after-market companies.  One example has been trailed by Jaguar Land Rover, using hydroxyl radicals to ‘purify’ cabin air¹².  While this may be true in terms of neutralising a coronavirus, surplus hydroxyl radicals can also react to form toxic secondary oxygenated gases and aerosols¹³, and so it is vital to perform a broader, perhaps untargeted, assessment of the effects of such systems.  Systems should be compared to the efficiency of filter-based systems such as on the Tesla X, to judge what value the hydroxyl radicals are adding – a judgement that will be allowed once there is a standardised method through CEN Workshop 103.

Thus, perhaps the future market for vehicles will be one of the quality of the experience rather than the magnitude of the performance.  BEV range anxiety may be quelled and the charge-up infrastructure made omnipresent.  Cheap, renewable energy will make efficiency almost irrelevant.  The gradual strangulation of the road space for cars, combined with connectively between vehicles and automation thereof, may leave occupants stripped of the joy of the driving experience, but consequently more demanding in terms of the quality and healthiness of the experience.  And this differentiation in experience, comfort, quality and design may be the route to profitability for manufacturers.

---

Footnotes:

1. Drivetribe.com

2. https://www.gov.uk/government pm25, PM2.5

3. https://www.gov.uk/government/statistics/emissions-of-air-pollutants/emissions-of-air-pollutants-in-the-uk-nitrogen-oxides-nox

4. https://www.emissionsanalytics.com/news/2020/1/28/tyres-not-tailpipe

5. PM2.5, i.e. all particles up to a diameter of 2.5μm

6. https://www.tesla.com/en_GB/blog/putting-tesla-hepa-filter-and-bioweapon-defense-mode-to-the-test

7. DPham, L., Molden, N., Boyle, S., Johnson, K. et al., “Development of a Standard Testing Method for Vehicle Cabin Air Quality Index,” SAE Int. J. Commer. Veh. 12(2):2019, doi:10.4271/02-12-02-0012

8. www.edmunds.com

9. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4892924/

10. https://www.cen.eu/news/workshops/Pages/WS-2019-017.aspx

11. www.airindex.com

12. https://media.jaguarlandrover.com/en-gb/news/2021/03/jaguar-land-rovers-future-air-purification-technology-proven-inhibit-viruses

13. https://www.ncbi.nlm.nih.gov/pmc/articles/

We have forty million reasons for failure, but not a single excuse

Rudyard Kipling’s description of Great Britain’s painful experience in the Boer War could easily have been written for Europe’s troubled recent experience in vehicle emissions control, air quality and associated public health.

The forty million reasons are the forty million diesels still circulating on the roads of the UK and Europe, emitting more than treble the regulatory limit for nitrogen oxides (NOx) in real-world operation1. Of these, an estimated eight million are Euro 6 diesels. None of these are restricted from a single city in Europe, or taxed in a penal way. In Germany, rather than legislating to clean them up, the government has legislated to prevent them being targeted2. The original source of the problem – the European Union regulatory framework – has been directed not at cleaning up this lasting problem, but only at new cars. NOx emissions from this latest generation of vehicles – certified under the Real Driving Emissions (RDE) regulation – account for just 1% of the total, which leaves the 99% coming from previously sold vehicles. The centrally directed, improved, regulation has hardly touched the problem.

Unless these older vehicles are cleaned up, poor air quality will persist. With the average lifespan of a vehicle in Europe being 14 years, and high-emitting vehicles having been sold through to at least 2018, the problem is not about to disappear automatically through fleet turnover. A particularly egregious example is the Mercedes Citan 1.5-litre diesel van from 2019, which emitted 902mg/km of NOx in real-world driving – 8.6 times the regulatory limit – and which may remain on the road long after the UK’s 2030 ban on new internal combustion engines.

Nevertheless, it is not every single pre-RDE vehicle that is problematic. Contrast the Citan with the 2.0-litre Volkswagen Crafter from the same year, which emitted 53mg/km or 58% below the limit, to see that not all vehicles need restricting to address air quality today. Just over a third – 36% – of all vehicles on the UK roads, for example, create an estimated 87% of all NOx emissions, according to our modelling. A third of all emissions are accounted for by the 12% of dirtiest vehicles. In principle, the most efficient, fairest and least distorting way to bring air quality into compliance is to target those vehicles first.

Rather than focusing on making cleaner vehicles even cleaner, Emissions Analytics has been involved in a project for the last three years to test and rate all vehicles currently on the road, to get to the truly dirty vehicles. The project has been led by the AIR Alliance, and has brought together many parties in the spirit of co-operation. Emphasising the important, urgent and persistent nature of the problem was the landmark legal ruling in the UK in 2020 that the death of Ella Kissi-Debrah can be partially attributed to road pollution4.

Contrary to reports, the arrival of battery electric vehicles has not cured urban air quality problems. Perhaps they will, but market penetration of these vehicles will probably not be high enough until the next decade to make a measurable difference. In the meantime, legal contraventions of air quality standards will continue and can only realistically be addressed by cleaning up the existing fleet of internal combustion engines. Add to this the Covid-19 effect, which has seen a robustness in the used car market5 and early evidence of a switch away from public transport. As a result, older, high-emitting diesel vehicles could become more valuable and be kept on the road longer than expected.

Emissions Analytics is one partner of many in the AIR Alliance. Its first act was to bring together top experts and academics, including Dr Norbert Ligterink from TNO in the Netherlands, Professor Helen ApSimon from Imperial College London, Dan Carder from West Virginia University Institute of Technology in the US – who was responsible for uncovering Dieselgate – and Dr Xavier Querol from IDAEA-CSIC in Spain. The group is chaired by Dr Marc Stettler from Imperial College London.

In contrast to many one-off or time-limited initiatives researching this area, Emissions Analytics and AIR Alliance decided to create an on-going programme of testing and rating vehicles. Furthermore, they wanted to create a legal basis for independently rating urban NOx emissions. This was achieved through the publication of a standardised method – CWA17379 – as the result of almost eighteen months work through the Comité Européen de Normalisation (CEN) together with the automotive industry, cities, academics and lobby groups. From this, the AIR Index (www.airindex.com) was born.

How is all this different from the regulations? It is a voluntary movement bringing together a wide range of parties from different points of view interested in genuinely solving the urban air quality problem. Where the rules of multinational organisations have failed, on-the-ground co-operation has created a real-world ratings system with a legal basis.

With the standardised method coined, data was gathered from multiple sources. All data was subjected to strict quality criteria and tested against dynamic boundary conditions – including the speed and acceleration of driving – to test for validity under the method. Data flowed not just from Emissions Analytics, but from authorities and even manufacturers themselves. Results fully compliant with this method are live and free to access on the AIR Index website. In total, well over a thousand models have been tested and four thousand hours of data collected, to the value of over $16 million.

Despite all this data, the challenge with any system that is not by government fiat is achieving coverage of the whole vehicle parc. It is easy to test a representative sample of vehicles, but not to cover the long tail of models. The AIR Index has not achieved that yet, but is now close. In fact, ratings are already being published for 34,575 model variants.

With such a large repository of data, it has been possible to model and predict the emissions of vehicles similar to those already tested. Using machine learning techniques it is possible to predict with an impressive degree of accuracy the missing results. The model is trained on all the PEMS data drawn from the multiple sources. The validation set is then the fully compliant CWA17379 results, which are of course excluded from the training set. Using this approach, the average error between compliant and modelled values for distance-specific NOx is just 2.6%3. As a result, a large majority of predicted ratings can be validated through blind testing to deliver the correct AIR Index rating on its ‘A’ to ‘E’ scale. For clarity, fully compliant tests and modelled results are always clearly labelled and disclosed as such. If manufacturers, or others, disagree with their predicted ratings, conducting a full test to CWA17379 would lead to the result over-riding the predicted value.

By combining the fully compliant and predicted ratings, it has been possible to create coverage of at least 90% of Euro 5 and Euro 6 models, as shown broken down by powertrain in the table below. These figures include all manufacturers except very low volume niche producers.

Further testing planned may move total coverage to 95% soon.

The result of this work is that there is now a parallel system to the official, centrally directed, certification system that can be adopted by cities across Europe. As a result, cities now have a live, practical tool that can be put immediately to the task of managing the existing fleet to bring NO2 concentrations into compliance.

Analysing the results, we can identify the greatest sources of NOx emissions on UK roads, segmented by Euro stage and AIR Index, as shown in the table below.

Therefore, there is a group of Euro 5 vehicles, with an ‘E’ rating on the AIR Index, that accounts for a third of total emissions, while only representing 12% of vehicles on the road. London’s Ultra Low Emission Zone (ULEZ) has been relatively successful because it has applied an access charge for Euro 5 diesels and earlier, but this is not the most efficient solution as some clean cars are restricted. Were the ULEZ to be based on restricting only vehicles with ‘D’ and ‘E’ ratings on the AIR Index, NOx could be reduced by a further 6% points, while charging 2% fewer cars.

The group of pre-RDE Euro 6 diesels with ‘D’ or ‘E’ AIR Index ratings is smaller than the Euro 5 group, but still important at 9% of total emissions. What is even more significant about these vehicles is that they form the ‘swing group’ that can make the difference between air quality non-compliance and compliance.

In a research report from January 2020 from Imperial College London6, the AIR Index emissions values were combined with standard dispersion models to show the impact of different ULEZ strategies on ambient NO2 concentrations and the number of legal excedances.  If the current strategy were applied to the enlarged ULEZ it was shown to bring most but not all locations into compliance with the 40μg/m3 limit, compared to a majority being non-compliant prior to the ULEZ.  Switching to ULEZ restrictions based on ‘D’ and ‘E’ on the AIR Index scale, irrespective of Euro stage, was shown to be sufficient to bring all areas into compliance.  In other words, the pre-RDE Euro 6 diesels make the difference.

Emissions Analytics is pleased to be involved in this practical, decentralised, collaborative, ground-up attempt to resolve one important element of air quality.  On-going testing will move the project closer to complete coverage, and it is set up to welcome an increasing number of data partners.  As the supply of data increases, so the accuracy of the database improves.

The only element missing is an honest reckoning by the authorities as to the systemic weaknesses that led to these problems.  Without that, the weaknesses may resurface.  For the sake of Ella, heed perhaps should be taken of Kipling:

Let us admit it fairly, as a business people should,
We have had no end of a lesson: it will do us no end of good.

Footnotes:

1. Transport & Environment, Cars with engines: can they ever be clean?, September 2018

2. https://www.euronews.com/2019/03/14/german-parliament-approves-exemptions-from-diesel-bans-to-incentivise-retrofits

3. https://airindex.com/air-index-vehicle-emissions-rating-system-offers-first-realistic-fully-fledged-alternative-to-official-data/

4. https://www.standard.co.uk/futurelondon/theairwebreathe/future-london-rosamund-kissidebrah-confront-inequality-air-pollution-b920852.html?amp

5. https://cardealermagazine.co.uk/publish/used-car-market-sales-are-exceeding-expectations-so-far-in-january-new-data-reveals/214964

6. Efficient control of NOx from diesel cars and ULEZ reduction of NO2 concentration, ApSimon, H., Oxley, T., Mehlig, D., Woodward, H., Stettler, M., Molden, N.

Letting the cat out of the bag: The great plug-in hybrid subsidy

In a recent newsletter, we set out Schrödinger’s Car, drawing a parallel with the famous Schrödinger’s Cat thought experiment, where a cat in a box is both dead and alive until the box is opened. Schrödinger’s Car is a vehicle that has high and low carbon dioxide (CO2) emissions until the occasion that it is driven, which describes the current situation with plug-in hybrid vehicles (PHEVs). In the best case, they could be cleaner than pure battery electric vehicles (BEVs), in the worst case higher emitting than standard internal combustion engines (ICEs), and on current behaviour no better than non-plug-in full hybrids (FHEVs) – all on a lifecycle analysis basis.

The conclusion was that PHEVs are too risky to allow into the market in any volumes, unless strong incentive mechanisms are put in place for drivers to charge them up regularly and employ them only on duty cycles that can be conducted mostly on battery. In some European countries, there are in fact positive incentives not to charge PHEVs, for example for company car drivers that get reimbursed for fuel spending but not additional domestic electricity consumption.

Emissions Analytics has tested 37 PHEVs in Europe and US using its standardised EQUA Index test protocol. The average CO2 emissions on the European PHEV vehicles tested are 115g/km, based on the typical proportion of miles driven on battery discussed below. This compares to an average on the official ratings of 58g/km, based on the Worldwide Harmonised Light Vehicle Test Procedure (WLTP) – so, real-world emissions are almost exactly double.

If these excessive CO2 emissions were not problematic enough in themselves, this situation is creating significant, but somewhat covert, subsidies of PHEVs in Europe and the UK through the fleet average CO2 target system and consumer incentives. As set out below, the annual subsidy across Europe may well be above €6bn already. However, this could rise steeply in a particular scenario that BEV take-up is slower than expected and manufacturers need to sell PHEVs in volume as they struggle to meet their ever-tightening fleet average CO2 targets through to 2030 and beyond. In contrast, if BEV penetration rises rapidly, the scenario set out below will not hold, as CO2 targets will easily be met, and PHEVs will be a brief transitional technology. In the more pessimistic scenario, where manufacturers rely heavily on PHEVs to meet CO2 targets, cumulative subsidy could add up to €475bn by 2035 – more than all motor-vehicle-related taxes paid each year in Europe1.

To illustrate this, a selection of three European PHEVs were taken, and each were paired with the nearest equivalent ICE vehicle. The official details are shown in the table below.

In a recent report, the International Council on Clean Transportation (ICCT) estimated the proportion of miles travelled on battery by PHEVs in Europe currently at around 37%2. This factor can be used to weight zero-CO2 battery operation with engine-only CO2 approximated by the WLTP of the equivalent ICE vehicle to yield a 'simulated' WLTP for the PHEV, as shown in the following table. This can be compared to the actual test values from the EQUA Index tests, weighted in the same way. Conversely, it is possible to estimate what proportion of driving would need to be on the battery for the NEDC and WLTP official test values to accord with actual practice.

Overall, EQUA Index CO2 results were only 4% higher than the simulated WLTP values. Whether considering the NEDC or WLTP, approximately 80% of driving must be on battery for the official CO2 values to be a meaningful representation of current driving, which is more than double the current reality.

This matters because of how it interacts with the Europe-wide fleet average CO2 targets. Averaging the emissions of new vehicles sold, each manufacturer must not exceed a threshold which is currently around 95g/km (the exact figure being tailored to each company, based on their mix of vehicle types), but which reduces over time. For each g/km above the limit, the manufacturer is fined £85 (€95) multiplied by the number of units sold. Below the limit, the manufacturer can effectively sell its surplus to a rival that is above the limit, thereby reducing the fine paid by that latter manufacturer by the same £85.

Taking the sample of three vehicles above, the simulated WLTP average is 114g/km compared to the official values of 35g/km, a difference of 79g/km. Alongside this, until 2022, the fleet average targets include ‘supercredits’, which allow each vehicle with official CO2 below 50g/km to count as 1.67 vehicles sold in the manufacturer’s overall fleet calculation. Therefore, the value of the difference in CO2 emissions is currently magnified, as shown in the table below.

In other words, by having artificially low CO2 values, manufacturers enjoy a significant subsidy from the European taxpayer of over £11,000 (€12,100) per unit. This is equivalent to 24% of the list price. As these vehicles are typically profitable to make, unlike most pure battery electric vehicles, it is not surprising that a large number of new models are being brought to market.

In 2020, around half a million PHEVs were sold in Europe. This would imply a total subsidy of over £5.5bn (€6.1bn). Projections for PHEV sales in 2030 vary widely, from below 10% by Deloitte to as much as 50%. Assuming total sales of all powertrains of 15m vehicles in 2030, this implies a total potential subsidy in that year of between £10bn (€11bn) and £50bn (€55bn), factoring in the phase out of supercredits.

If we take the mid-point of these 2030 scenarios, PHEV market share grows in a linear way up until then, and is constant after that until 2035, the cumulative subsidy would be a not inconsiderable one-third of a trillion Euros. This is more than half the annual gross domestic product of Bavaria.

But this is does not cover all the subsidies available for PHEVs. Many European countries incentivise their purchase, through a combination of retail subsidies, annual car tax reduction and benefit-in-kind income tax reduction for cars provided to employees by their company. The type and magnitude of these subsidies vary significantly between countries, but we can take the UK as an example.

The source of the subsidy is the same as with the fleet average CO2 targets: artificially low official CO2 figures. It is the prerogative of governments to offer subsidies to consumers, but this is being done in a covert manner, and does not align with the true environmental effects of these vehicles. In the UK, the subsidy is channelled through company car benefit-in-kind tax and annual road duty. While the latter applies across the market, company cars – including schemes such as ‘salary sacrifice’ – account for only around 15% of the market5.

The table below shows the implied value of the underestimation of CO2 emissions – how much less consumers pay than they 'should'. The final column weights the company car subsidy by the 15% market share to get a representative value for the whole market.

The subsidy to a company car driver over a three-year period is £12,671 (€13,938). On top of the manufacturer subsidy, this is 52% of the average vehicle list price. Across the whole UK market, weighting the company cars at 15% of the total market, the average consumer subsidy is over £1,800 (€1,980). Combining this with the manufacturer subsidy, the total subsidy is £13,053 (€14,358) per unit, or 28% of list price.

Taking this value, together with the same market share projections through to 2035 as above, the total subsidy would be €475bn – almost the same as the gross domestic product of Baden-Württemberg. Of course, this relies on the scenario that PHEVs are the swing factor for manufacturers in meeting or not meeting their future CO2 targets, which may not be the case if BEV penetration accelerates.

It should also be noted that the levels and terms of these taxes and duties are subject to change by governments at relatively short notice, and they differ between each country in Europe. Whether the overall level of support from governments is increased or decreased is unpredictable. In contrast, there is greater future visibility on the fleet average CO2 system.

Ultimately, the amount of taxpayers’ money dispensed, the trajectory of climate change and the structure of the automotive industry rides on whether Schrödinger’s Car, like the Cat, lives. Unless the right incentives are put in place around both original purchase for the right use cases, and after purchase to encourage charging the vehicles, PHEVs should have all subsidy removed immediately. With the current propensity for owners to charge their PHEVs, the CO2 emissions are no better than traditional hybrids, which do not attract the same subsidies, and which put much less pressure on the battery material supply chain.

Footnotes:

1. ACEA (European Automobile Manufacturers’ Association), https://www.acea.be/press-releases/article/motor-vehicle-taxation-brings-in-440.4-billion-for-governments-in-major-eur

2. Real-world usage of plug-in hybrid electric vehicles, International Council on Clean Transportation/Fraunhofer Institute for Systems and Innovation Research ISI, September 2020

3. Difference between simulated CO2 and NEDC, multiplied by £85 (€95) as the market price of carbon, applied only in year of purchase

4. Supercredits as defined under fleet average CO2 system – assuming x1.67 for a sale in 2021

5. According to HMRC, there are 870,000 company car drivers in total (https://www.fleetnews.co.uk/news/fleet-industry-news/2020/09/30/hmrc-data-shows-30-000-fewer-company-cars) and a vehicle may be held for an average of three years, yielding 290,000 purchases per year, which is approximately 15% of the 2m annual UK car market

6. Assuming 40% higher-rate tax payer, multiplied by 10% company car share of market over three-year period and 13.8% employer National Insurance

7. Annual vehicle road tax, based on CO2 emissions over three years

Side effects may include...

Beware the secondary effects of decarbonisation

No vehicle yet designed generates zero emissions.  Despite much variation geographically, and much argument, battery electric vehicles probably, on average, halve lifecycle carbon dioxide (CO2) emissions when considering first-round effects such as manufacturing and operation. But are the advantages as clear if secondary effects – the side effects – of electrification are considered?

In a previous newsletter, we set out our Eight Principles of Decarbonisation required to meet a real net-zero target, as shown below. By "real net-zero" we mean actually net-zero in a similar way to the "absolute zero" set down in UK FIRES by Allwood et al in 2019.

Vehicle manufacturers are performing increasingly sophisticated lifecycle analyses of their products. However, most do not consider the secondary, or knock-on, consequences of these electrified vehicles. This is our Sixth Principle of Decarbonisation.

These side effects may include the knock-on consequences on energy infrastructure, vehicle design, driver behaviour and traffic patterns. Most immediately, electric vehicles are likely to increase demand on the electricity grid. To meet net-zero, this additional demand will need to be fulfilled with zero-carbon electricity. Cleaning existing electricity will not be sufficient. For any net-zero scenario it is a pre-requisite that the whole future grid is clean, but we will not look at this further here as it is outside the scope of this newsletter despite being foundational to any meaningful solution.

The most frequent concern with battery electric vehicles (BEVs) is that the additional weight compared to equivalent internal combustion engines (ICEs) leads to higher non-exhaust emissions, which may equal or exceed the eliminated exhaust emissions. These non-exhaust emissions from the vehicle come from its brakes and tyres. Road abrasion and resuspension are often included in non-exhaust emissions, but will be set aside here as they do not originate directly from the vehicle.

Emissions Analytics conducted a long-term study on the wear from Continental Contisport 6 tyres on a 2012 Mercedes C-Class driven on the public highway in consistent, normal conditions. For the first 1,200km, with no added payload, the average wear rate was 161mg/km, but over 31,000km the average wear rate fell to 76mg/km. Even this lower value is 15 times higher than the maximum permission exhaust particle mass emissions under current European regulations. Running the same 1,200km test but with 570kg of payload in addition to the driver, the wear rate increased to 194mg/km, an uplift of 21%. In other words, the average uplift was almost 6mg/km for every additional 100kg of payload. For example, on this basis the Jaguar I-Pace would emit 16% more tyre particle wear than the nearest equivalent Jaguar F-Pace, due to the 443kg additional overall weight.

As a reference, the maximum exhaust particle mass emissions permitted in the EU since 2009 is 5mg/km1. Often, real-world emissions on vehicles with a particle filter are well below 1mg/km. Therefore, for every 100kg of extra payload, the added tyre wear emissions may be as much as the maximum allowed out of the tailpipe in total, and more like five times more than the tailpipe emissions in practice.

However, this analysis may overstate the increase. The extensive regenerative braking of the BEV may well reduce brake emissions and the calibration of the electric motors may smooth driving dynamics to reduce tyre wear. On the other hand, ICEs are increasingly incorporating regenerative braking using 48V systems, and the higher torque of the BEVs (27% in the case of the Jaguars) may encourage more aggressive driving. While data is too limited to draw firm conclusions on these mitigating factors, the underlying upwards pressure on tyre wear remains.

Vehicle weight is clearly a crucial factor in vehicle performance and profitability. The Chinese-built Tesla 3 contains a lithium-iron phosphate (LFP) battery, whereas the US-built versions have the nickel-manganese-cobalt (NMC) version. For similar range, the LFP battery is 200kg heavier, but cheaper in construction. That additional 200kg leads to approximately 8% extra energy to propel the vehicle due to increased inertia and rolling resistance of a typical on-road driving cycle. This leads to greater CO2 emissions in the electricity generation, distribution and usage. At the same time, that extra weight may cause 12mg/km of tyre wear, other things being equal. These downsides could be offset by material light-weighting, power and torque limitations and advanced tyre materials – but all come at either added cost or reduced driver utility.

Switching to inside the vehicle, the preoccupation with maximising the range of electrified vehicles – to be competitive with ICEs – may lead to worse vehicle interior air quality. Incoming air to the vehicle’s ventilation system is usually filtered to remove first-and-foremost particles, but this process consumes energy due to the back pressure created by the filter. While this may be insignificant proportionately on an energy-consumptive ICE, it can be material on more efficient vehicles.

Emissions Analytics conducted a test programme across 97 recent model year cars in the US market and found that many hybrids had relatively poor filtration. Tests were conducted on a standardised urban route around Los Angeles. Real-time particle number concentrations, with a lower size cut-off of 15nm, were measured simultaneously inside and outside of the vehicle and the integrated values ratioed over the test. Condensing Particle Counters from National Air Quality Testing Services (NAQTS)2 were used. The testing followed the methodology set down in a Society of Automotive Engineers (SAE) paper authored by Emissions Analytics and the University of California Riverside3.

Not all electrified vehicles performed poorly, but the majority did. The average cabin air quality index from hybrids was 55% worse than the other vehicles in the group, and the particle ingress on the worst was 3.6 times higher than the average of standard vehicles. In contrast, the Jaguar I-Pace BEV was one of the best performers. Although not part of this testing, Tesla’s ‘biohazard’ high efficient particulate air (HEPA) filter, which is now standard on the Models S and X, has excellent reported particle ingress performance, although it will still come at the cost of increased energy consumption.

To quantify this energy consumption, we can look at the mechanics of the ventilation system. A typical vehicle heating, ventilation and air conditioning system consumes from around 140W to 1.4kW depending on the setting4. The lower value is an approximation of the power requirement of the fan and the energy required to overcome the back pressure from the filter. At an average speed of 40km/h, the energy consumption would be between 0.28kWh and 2.8kWh per 100km driven. A typical BEV would consume 25kWh per 100km, so the ventilation system may add between 1.1% and 11% to overall energy consumption. For this reason, there is an incentive to reduce the amount of air filtered, the filtration efficiency or air conditioning activity on electrified vehicles, which would lead to higher particle exposures – and the resulting adverse health effects – of the occupants.

Thinking more widely at the transportation system level, a problem that may start to emerge is added congestion, caused by extra vehicle miles from electric vehicles, which then may adversely affect total emissions from the fleet. This would apply during the transition, while BEV penetration remains relatively low.

A BEV costs approximately 5 pence (5.5 Euro cents) per kilometre in energy costs, compared to 12 pence (13.2 cents) for a reasonably frugal ICE5. Other things being equal, this is likely to lead to more and longer journeys, and a switch to cars from other forms of transport: the income and substitution effects. Setting aside the effects on the economics of public transport, the additional traffic volume will lead to greater congestion, other things being equal. As the fleet will remain predominantly powered by ICEs for decades – due to the legacy light-duty fleet and diesel remaining prevalent for heavy-duty vehicles – this added congestion caused by BEVs is likely to cause increased emissions from these legacy ICEs.

Analysing Emissions Analytics’ database of over 2,000 light-duty vehicles, we can quantify the effect of this added congestion. To travel the same distance at the same speed (65km/h), a driving profile with stopping and starting between 30km/h and 90km/h can create higher emissions than steady-state driving. On average, CO2 emissions are 24% higher, NOx emissions 89% higher and particle number emissions 75% higher. For a period, a relatively small number of BEVs may adversely affect the emissions of the majority ICEs, increasing emissions and worsening air quality. This does not mean the push to BEVs is wrong, but the secondary effect in the short- to medium-term must be considered. One mitigation would be to push faster for BEV penetration.

As congestion leads to longer journey times, the rational response would be for some distance-based or road-access pricing. At least, this would need to compensate for the naturally lower marginal costs of operation of BEVs. More widely, there is a strong argument that motoring generally is under-priced. The pollution produced by an ICE is a negative externality not internalised, which leads to over-consumption.

In summary, these are just three of the potential side effects of electrification. This does not mean that electrification is bad, but that these secondary effects must be understood and controlled. With the large amounts of taxpayers’ money being requested to build the electric infrastructure, there should at least be a responsibility that this is well spent and not just the catalyst for swapping one problem for another.

---

Footnotes: 

1. Or 4.5mg/km under the PMP protocol
2. NAQTS
3. Researchgate Publication 333500744 Development of a Standard Testing Method for Vehicle Cabin Air Quality Index
4. Impact of the Air-Conditioning System on the Power Consumption of an Electric Vehicle Powered by Lithium-Ion Battery, Modelling and Simulation in Engineering, 2013
5. Assuming 2.5 miles per kWh and 12.5 pence per kWh for the BEV; 45mpg (6.3l/100km) and £1.20 per litre (€1.33) for the ICE

Schrödinger’s Car

Resolving the enigma of plug-in hybrid vehicles

Are they good or are they bad? Like the feline thought experiment, where the cat is both dead and alive simultaneously until observed, the answer is that plug-in hybrid vehicles (PHEVs) are both good and bad until they are used. It matters because this particular powertrain lies at the epicentre of the battleground for carbon dioxide (CO2) reduction and, therefore, the prospects for containing climate change.

An earlier newsletter (Plugin hybrids without behavioral compliance risk failure) considered the variability in emissions and fuel consumption of PHEVs, and the dangers for policy. Now that PHEV sales are accelerating, the topic has become more urgent: how to ensure that a valid bridging technology to an electrified future does not become the next emissions crisis. How can we head off a collision between unpredictable consumer behaviour and misleading official emissions ratings?

Car labelling (whether for CO2, fuel economy or noxious emissions) has historically been about the car, with the view that ultimately its performance can be characterised fairly in a single “combined” number. This figure is used variously for regulation, taxation, urban access policy and consumer information. Now we face the situation where vehicle performance is arguably secondary to driver behaviour. The corollary of this is, perhaps: label the driver, not the car.

In traditional vehicle emissions labelling, it has been possible to “get away” with the single combined number because the difference between the worse case (urban driving) and better case (motorway driving) was around a one-third increase in CO2 emissions and fuel consumption. In other words, the combined number would not be far away from reality for most drivers over time.

Emissions Analytics has tested 37 PHEVs for its independent emissions programme. The method uses the EQUA Index test route – which is significantly longer than official cycles, combining urban, rural and motorway driving – and tests separately for electric-only range and engine-only efficiency, both in real-world, on-road conditions.

The table below summarises the results, split between European and US tests1. Comparing the engine-only mode to official certification values, real-world CO2 emissions are between double and treble. In this case, though, the real-world combined number appears to be a good characterisation of the performance of the vehicle, whether it is used in urban or extra-urban driving.

However, once you introduce real-world driver behaviour in terms of trip mixes and charging up of the battery between trips, it becomes clear that there is a significant problem. From the same test data, real-world CO2 emissions are as high as 299g/km in average urban driving if you never charge the battery up. At the other extreme, if you always charge up the battery and only ever take short journeys, CO2 emissions will be almost zero. In other words, there are orders of magnitude of difference between the best and worse cases. In certain use cases, real-world CO2 may be substantially better than the official figures.

For comparison, at the height of the divergence between the old New European Driving Cycle (NEDC) and real-world values for internal combustion engines (ICEs) – before it was replaced by the Worldwide Harmonised Light Vehicle Test Procedure (WLTP) – the spread was only around 50%. The result of the NEDC official labelling system diverging from reality artificially led to the sales of downsized engines in large cars and the dieselisation of city cars. The damage in terms of both CO2 emissions and air quality will remain for more than another decade as those ‘artefact vehicles’ remain on the fleet.

The US test procedure is different from the WLTP, but achieves similar test results, but then applies a reduction of up to 30% to the fuel economy values and thereby a similar increase to the CO2 values. While this approach can work well for ICEs to align official values for real-world consumer labelling, it does not solve the problem of the divergence in performance possible for each PHEV.

The variance in performance between best and worst cases is more than a point of statistical interest. It has the strong potential to undermine the whole system of fleet average CO2 targets. If, for example, these vehicles travel only 10% of their distance on battery, they would be greater generators of CO2 emissions that full hybrids (FHEV) and mildly hybridised diesel ICE vehicles.

This proportion of miles travelled on the battery is called the “utility factor” (UF). If we compare the average European real-world tailpipe emissions of 182g/km from a PHEV with the crucial 50g/km threshold, this implies a UF of 72%. The 50g/km is important because it is a widely recognised benchmark for “ultra low emission vehicles”, below which manufacturers receive supercredits towards their fleet average CO2 targets and many consumers receive significant tax benefits. However, according to a recent report from the International Council on Clean Transportation (ICCT)3, the UF may currently be around 37% – this would imply real-world CO2 emissions of around 115g/km, worse than the best FHEVs.

On this 37% UF scenario, real-world emissions would be 130% higher than the 50g/km threshold. This means that labelling of PHEVs presents a much greater danger than ICEs did under the NEDC, the latter being a major contributor to the Dieselgate scandal. The combination of low CO2 values the WLTP permits and the current supercredits system is leading manufacturers to launch a large number of new PHEV models. For example, Jaguar Land Rover will have seven models by the end of 2020, and Daimler will have 20 by 2021, rising to 25 by 2025. In the UK, sales of PHEVs rose from 4,788 in January 2020 to 12,400 by September. According to the Fraunhofer Institute, there were approximately two million such vehicles on the roads worldwide by the end of 2019. Therefore, the next scandal is already brewing, as manufacturers react to the incentive structure provided by regulations, which is creating the next generation of artefact vehicles.

The system as currently constructed means that the manufacturer is bearing no direct risk of this UF being less than required for real-world emissions to align with official values. Instead, society bears the risk. In fact, manufacturers have a positive incentive to produce these vehicles due to the supercredits system that runs for the coming three years. This is a dysfunctional situation.

To illustrate the degree of the challenge, we can consider what proportion of miles must be driven on battery by a PHEV to be better than alternative powertrains. To make the comparison as fair as possible, including against BEVs, lifecycle emissions are used in the table below.

These values are based on Emissions Analytics’ testing of Euro 6 vehicles, and its proprietary lifecycle model. ‘Cleaning grid’ refers to the CO2 intensity of the electricity used to power a new BEV over its lifetime as the grid decarbonises, as forecast by the Eindhoven University of Technology5.

Although even modest charging-up renders PHEVs superior to unhybridised ICEs, a mildly hybridised diesel – with a ~6% efficiency advantage over a standard diesel ICE – would be comparable if around a fifth of miles were on the battery. The current utility factor of 37% described above leaves PHEVs performing the same as non-plug-in full hybrids, but with the extra consumer hassle and resource requirements. However, PHEVs can be superior to BEVs if they are used on the battery for just over seven out of every ten miles, using a reasonably optimistic view on grid decarbonisation, due to the extra CO2 from manufacture of the larger BEV battery.

The final line in the table shows that the official 50g/km threshold implicitly assumes that consumers charge up at least 73% of the time, which is double the reality. This is another way of quantifying the artefact of the WLTP and fleet average CO2 target systems.

This shows that it is possible to argue that PHEVs are the best of all worlds, rendering all other powertrains obsolete. Customers could have the smooth, low-end torque, high-end power boost and no range anxiety at an acceptable price. At the same time, you could argue that they are no better than non-plug-in hybrids and should not exist. Thence this enigmatic feline?

One option would be for manufacturers to assume this UF risk, through a system of post hoc reweighting of their official figures for the purposes of the fleet average CO2 targets, according to the real-world UF. Consumer labels could also be reviewed annually, without vehicles needing to be retested or certified. This would incentivise manufacturers to market these vehicles only to consumers likely to operate them with a high UF.

Under European regulations, vehicles will have to report their real-world fuel consumption from 2021. This could give a good estimate of the real-world UF, which could be used for this reweighting, applied retrospectively to all PHEVs on the road. It would not be a sufficient solution just to wait and see from this data whether there is a problem. There is a problem, and this data can be used to implement a solution today. In vehicle and product safely law, manufacturers are generally responsible for the foreseeable misuse of their products – not charging up a PHEV is arguably foreseeable misuse.

Putting risk onto the manufacturer could be coupled with more sophisticated charging mechanisms for consumers. In the UK, the Renewable Heat Incentive (RHI) for domestic heating offers a subsidy to homeowners based on the amount of ‘renewable’ heating generated, e.g. from an air-source heat pump. Translating this onto vehicles, a telematics system could provide a subsidy for every mile driven on battery and tax on every minute the engine is running. Through a solution like this, the risk could be shared between the manufacturer and the car owner, while delivering society’s goal.

Unless this challenge is resolved, it could lead to – if PHEVs gain high market share – a subversion of the policy for reducing CO2, but equally – if PHEVs are banned – the removal of a potentially powerful low-carbon technology. If you already believe the endgame is BEVs and we just need to get there as soon as possible, then banning PHEVs will seem the obvious and comfortable option. If you believe that the world’s light duty fleet should not be controlled by an oligopoly through its access to battery raw materials and that a competitive market of rival powertrains should be fostered, then the PHEV conundrum needs to be solved.

There are ways in which the official testing regime could be improved. For example, the EQUA Index test methodology for PHEVs developed back in 2014 acknowledged these vehicles could not be characterised in a single number. Therefore, each vehicle has both an electric-only range and an engine-only fuel economy. The official system in the EU already has similarly granular information from certification tests, but then falls into the trap of producing a single number. More recently, Emissions Analytics has measured electricity consumption independently of the vehicle’s systems in real-world conditions through a portable meter, coupled with a measurement of charging losses. Overall, this is the most condensed yet representative way of labelling this type of powertrain.

Nevertheless, better labelling does not answer the policy question of how these vehicles should be represented in the fleet CO2 targets, which still comes back to actual consumer behaviour and that utility factor. Get it wrong, and we may miss our CO2 targets by a significant margin, or dispose of an attractive technology. We may, through neglect, cede a major global market to the owners of certain strategic raw materials.

In the European Commission’s own words: “The stakes are high.”6

---

Footnotes:

1. It is worth noting that the products are clearly very similar in performance between the two territories, and also in technology, as average engine size is 2.0 litres and average battery capacity 11.5 kWh in both regions.  The average electric-only range across both markets is 41km.  Average prices in Europe are around £41,000 compared to typical average ICE vehicles at £25,000-30,000.
2. CO2 conversion from the US Environmental Protection Agency’s MPG-equivalent value
3. Real-world usage of plug-in hybrid electric vehicles, International Council on Clean Transportation/Fraunhofer Institute for Systems and Innovation Research ISI, September 2020
4. Applying estimated manufacture CO2 emissions to tailpipe threshold for supercredits
5. Comparing the lifetime green house gas emissions of electric cars with the emissions of cars using gasoline or diesel, Eindhoven University of Technology, 2020
6. Critical Raw Material Resilience: Charting a Path towards greater Security and Sustainability, European Commission, COM(2020) 474 final, 3 September 2020

Eight principles of decarbonisation

Elimination of carbon dioxide from transport must be real not artefact

The United Kingdom has enacted a law for net-zero carbon dioxide (CO2) emissions across the economy from 2050, and other countries will likely follow. This has spawned a policy of targeting zero emissions from transportation, which is often spoken about interchangeably with the idea of replacing the vehicle fleet with battery electric vehicles (BEVs). A further benefit of this change is claimed to be improvement in air quality, due to the absence of a tailpipe, and the resulting health benefits. But is this apparent panacea as simple as it sounds?

Slowing climate change is now widely regarded to be of such importance that all ideas need to be robustly tested not just for the glamour of their ambition, but whether they can deliver. Admitting the existence of risk, a mix of approaches may be more robust than gambling all on red. Unintended consequences must also be considered for fear of achieving the goal but at an unacceptable price. As The Economist said in August 2020, “…Mr Xi is shifting to a sharp focus on supply-chain choke-points where China is either vulnerable to foreign coercion or where it can exert influence abroad. That means building up self-sufficiency in key technologies, including semiconductors and batteries.” The reality is that there is already, and will continue to be, a widely predicted global supply constraint on batteries for the rest of this decade if take-up of BEVs is as healthy as it needs to be to fulfil zero carbon transportation.

For transport, then, the right policy could be expressed as delivering zero carbon dioxide (including other gases with equivalent greenhouse gas effects) emissions with no worsening in other pollutants and no vexatious secondary effects. At least, any trade-offs should be understood scientifically and communicated politically. This newsletter will consider what would be necessary and sufficient to achieve this, and the implications.

So, let us set out Emissions Analytics’ Eight Principles of Decarbonisation – the material things that need to be delivered to achieve the stated policy:

Principles 1 and 2 are a significant enough challenge in themselves, but progress is being made in that direction by many countries. There remains a significant issue in the intermittency of many renewable energy sources, especially where there is no nuclear baseload. This newsletter will not go into these further, but unless they are achieved, the advantages of BEVs will be seriously compromised, whatever else is achieved in respect of the other principles.

Principle 3 arises from carbon being a global problem: the total matters, not the source. Reducing upstream CO2 emissions is a very serious challenge given the lack of control that governments can apply to foreign mining and manufacturing operations. It is also material, due to the higher CO2 emissions currently from BEV compared to ICE manufacture, like-for-like. Efforts are being made by manufacturers, including recent reports from BMW and others, to assert some control and transparency, but it remains a significant challenge. Even the transportation of vehicles by sea would need to be completely decarbonised to achieve this, which may conceivably be achieved with alternative fuels such as ammonia produced with zero carbon electricity.

Principle 4 is a growing challenge, with the need to recycle various rare earth metals from drive motors and batteries, or give them a second life before eventual recycling. This is a business opportunity that is already attracting entrants, yet achieving this with zero-carbon energy remains a stretching goal.

Principle 5, means that other emissions that have a global warming effect – often expressed in CO2-equivalence – should be included in analysis, to ensure one climate-charge-relevant emission is not swapped for another, through singular focus on CO2. This could include, for example, methane emissions from gas vehicles or nitrous oxide (N2O) from certain after-treatment systems.

Principle 6 is an important but complex one. This could mean that vehicles should not ‘crowd out’ the use of batteries from existing non-transportation uses. For example, where batteries may be used as static power sources or for powering handheld tools, if demand from transportation made battery prices too high, these uses may switch back to combustion engines. It could also mean that old ICEs that are replaced with BEVs in developed countries are exported to expand fleets in developing countries as their values fall.

Recalling the Dieselgate crisis, where nitrogen oxide (NOx) emissions were found to be insufficiently regulated in a way that led to damagingly high emissions in the real world, there is a big risk from the non-exhaust emissions from BEVs, especially in regard to tyre wear. Due to the weight of batteries, BEV vehicles are significantly heavier that like-for-like ICE vehicles. As a result, for the same grade of tyre and driving patterns, non-exhaust emissions from wear on those tyres will be higher for the BEV. The regenerative braking for BEVs may lead to lower brake wear emissions compared to ICEs, but this is unlikely to counterbalance the increased tyre wear emissions. Therefore, the risk is that the CO2 reduction from BEVs is traded for a degradation in air quality and other microplastic pollution.

Through our work on measuring pollution inside the vehicle cabin, we have observed that hybrid vehicles often have worse interior particle concentrations, which may lead to worse exposures and health effects for occupants. The hypothesis is that, due to the energy requirements, filtration is minimised as these vehicles are sold for their fuel efficiency. Therefore, CO2 may be traded off against human health in this additional dimension.

While carbon credits, Principle 7, may act as a positive incentive mechanism for carbon reduction, inherent in the system is their trading effectively permits the continued use of combustion engines by the purchasers. Therefore, while they may be expedient in the short run, they must be phased out permanently.

Principle 8 is simple: to achieve zero carbon, the whole ICE and hybrid fleet on the road must be replaced. It is likely that the last few percent of vehicles will be hard to shift due to stubborn owners, and therefore the incentives needed may be high. In conjunction with Principle 6, old ICE vehicles should be responsibly recycled, not exported to developing economies, to avoid the scandals associated with the scrappage scheme following the financial crash of 2008, when dirty diesels were ‘scrapped’ at the expense of taxpayers only to be found to have been exported to Eastern Europe and beyond.

Last year, we published this newsletter which showed that hybrids were the best way to reduce CO2 in a world of limited battery capacity. Full hybrids deliver around 30% CO2 reduction compared to the nearest equivalent ICE, compared to the 100% tailpipe reduction of BEVs; but 14 times more hybrids can be built for the same battery capacity, meaning hybrids could actually deliver four times more CO2 reduction than BEVs while the battery constraint remains.

While limited battery capacity remains true for now, let us project forward to a hypothetical world of unconstrained and cost-competitive battery supply. In addition, we also make the critical assumption that all electricity will be zero carbon (Principles 1 and 2). We assume that complete grid decarbonisation will be achieved. We also assume that manufacturing CO2 emissions (Principle 3) remains the same as currently, but also that there is no improvement in ICE efficiency. Further, we are putting no negative value on the utility limitation of the reduced range of BEVs compared to ICEs, which may well persist even when batteries are plentiful in supply and cost competitive due to the weight they add to the vehicle.

To assess these effects, Emissions Analytics has created its own model of the CO2 effects of electrification. Three scenarios were modelled: complete switched to BEVs and full hybrids compared to the ICE baseline.

The manufacture emissions of ICEs, full hybrids (FHEVs; for clarity these exclude mild hybrids and plug-in hybrids, but sit between them in terms of battery size, and rely only on on-board gasoline or diesel for energy) and BEVs are 5.8, 6.2 and 11.4 tonnes per unit¹. In use, the average CO2 emissions are 111, 78 and 0 g/km. Assumed annual usage of all three are 16,000km over a lifetime of 200,000km. On the switch-to-BEV strategy, the new car sales mix is assumed to be 100% BEV by 2040, and the whole fleet by early 2050s.

Crucially, as CO2 is cumulative in the atmosphere – it lasts between 300 and 1,000 years once emitted – we must consider cumulative emissions in our analysis. This is the basis of the Paris agreement and countries’ carbon budgets.

On this BEV scenario, cumulative total CO2 emissions emitted are higher for over a decade due to the front-loaded emissions in the manufacture of the BEVs. From 2034, the cumulative CO2 is lower than the benchmark ICE strategy as the benefits of zero in-use emissions begin to outweigh the higher embedded emissions from manufacturing and batteries. By 2070, cumulative emissions are 811m tonnes lower from BEVs. However, this is only 47% down on the ICE strategy; in 2050 it is only 23% down. So, unless you can eliminate manufacture emissions, we are not even close to zero emissions – even ignoring Principles 5 to 8.

If you double average battery size, and therefore double the CO2 in their manufacture, to compensate for the bounded utility of BEVs and bring range more in line with ICEs, then the reduction by 2070 is just 11%, and it is 1% worse still in 2050.

Revisiting the potential for hybrids set out last year, the BEV strategy can also be compared in this model to a FHEV strategy of switching the whole vehicle fleet to FHEVs by the earlier 2050s, and it yields dramatic results.

The BEV strategy is still superior, but the total CO2 reduction by 2050 is just 11% and by 2070 it is 26%. Good, but far from zero. Less good is if you double average battery size, when cumulative CO2 emissions are 5% worse from BEVs than FHEVs.

These results stem from the higher manufacturing emissions of BEVs, which are true not just in the initial switch to BEVs but on subsequent future vehicle replacement, together with the lower in-use CO2 from hybrids. This will only change if Principle 3 is delivered – hence its vital importance. The cynic may think that net vehicle importing countries may be quite content to ‘off-shore’ rather than genuinely reduce the emissions.

Although the legal position is that net zero must be achieved by 2050, from the point of view of climate change, what happens in the next twenty years is just as important. Carbon dioxide has no discount rate. Every gram of CO2 that is reduced now makes the challenge post-2050 easier. Therefore, inaction today on the promise of a miracle solution tomorrow is not a robust policy.

Unless a more robust policy is developed, delivering real and measurable CO2 reduction soon, it is likely around 2035 that BEVgate will break, where much of the promised reduction will have proved illusory and air quality may be worse. As it took 15 years from the introduction of explicit NOx reduction under the ill-fated Euro regulations until Dieselgate, it may be 15 years from now that the folly of dogma rather than facts becomes clear. And our carbon budgets will by then be spent.

---

Footnotes:

1. Derived from https://www.volkswagen-newsroom.com/en/press-releases/electric-vehicles-with-lowest-co2-emissions-4886; in-use emissions for FHEVs applies 30% efficiency estimated by Emissions Analytics

What's the problem with biofuels?

One of the many conundra in transportation emissions is why the use of biofuels, promising environmentally friendly reductions in emissions especially for heavy-duty vehicles, is not taking off faster and attracting only limited policy focus. Compare that to the excited talk around battery electrification of trucks, which generously could be described as a challenging solution.

Across the range of biofuels that Emissions Analytics has tested, some patterns become clear. The first observation is that, on average, they tend to lead to very little change in emissions at the tailpipe.  Second, within that average, there tends to be significant variability between vehicle models on the same fuel. Therefore, if comparing tailpipe performance, it is difficult to generalise about the benefits of a particular biofuel, and the expected benefit in any case may be small.

A corollary of the first point is that the proposed benefit of these fuels relies on emissions reduction – often but not always focusing on carbon dioxide (CO2) – located upstream in the fuel production and distribution. This part of the supply chain is often opaque, and not easily subject to independent verification or regulation, as is possible with tailpipe measurement. Therefore, even if the benefits of a biofuel sound attractive, but cannot be verified, it would be right to take a sceptical position. On top of that, there may be secondary, unintended consequences due to competition for scarce resources as biofuel production is scaled, for example by the earlier effects on food supply and prices where raw materials were crops.

Rightly, therefore, many governments have been cautious in incentivising biofuels too categorically. For example, E10 – gasoline containing up to 10% ethanol – has been long proposed in the United Kingdom (UK) and other European countries, yet its adoption has been slow. Natural (fossil) gas, familiar at the pump as either Liquefied Petroleum Gas (LPG) or Autogas, attracts a significant tax advantage in the UK and other countries, yet there is limited forward visibility because the tax advantage is always at risk of withdrawal. Fleets, consequently, have been cautious in converting to natural gas.

To illustrate the dilemma, we can consider test results from two programmes Emissions Analytics has conducted, one covering E10 gasoline, the other compressed natural gas (CNG) and liquefied natural gas (LNG).

Switching from E5 to the higher ethanol E10 gasoline in Europe is often held out as a quick way to deliver large aggregate CO2 reductions as it can be distributed at scale through the existing refuelling infrastructure, to be used both in light- and heavy-duty applications.

To investigate this, we tested E10 on 17 gasoline vehicles in Europe from a wide range of manufacturers. On average, CO2 emissions fell by 0.5%, but this varied at the model level from an increase of 7.1% to a decrease of 6.3%. All were relatively new Euro 5 or Euro 6 vehicles, tested on our standardised on-road cycle made up of urban, rural and motorway elements. Due to the lower energy content of E10, the fuel economy worsened on average by 1.2%, with a similarly wide variation. Perhaps unexpectedly, emissions of nitrogen oxides (NOx) increased by 16.6% on average, from an initial level of 59 mg/km. Carbon monoxide (CO) fell on average by 4.5%.

Therefore, at the tailpipe, the reduction in CO2 was negligible overall. The value of the reduction in CO did not outweigh the effects of the increased NOx emissions, as non-compliant urban air quality is generally caused by excessive NO2 concentrations. For the introduction of E10 to be valuable, the upstream CO2 reduction would need to be large enough to outweigh that increase in NOx emissions. Sufficient transparency around those upstream emissions and any secondary consequences, is not offered by the supply chain and therefore the overall benefit of E10 should rightly be treated with scepticism.

Turning to natural gas, we were invited to be part of the first extensive UK study to assess real-world performance of heavy freight vehicles fuelled by this alternative. The two-year test ending in 2019 was led by biomethane supplier Air Liquide, with three transport operators (Howard Tenens, Asda and Kuehne + Nagel) and three technical partners: Cenex (data monitoring), Microlise (telematics) and Emissions Analytics (real-world testing). The project was supported by Innovate UK.

The goal was to consider the overall performance of gas as a fuel for heavy freight, compared to diesel. A secondary aim was to only use 100% biomethane, which in this case was derived from food waste and supplied by Air Liquide.

Biomethane is chemically similar to its fossil gas counterpart, although the product of a different refining process that starts with biogas that is then improved to increase the methane content and purge it of impurities. Biogas is the raw product of anaerobic decomposition of waste ranging from human and livestock excreta to food waste. Whether from a renewable feedstock or not, gas can be packaged in compressed but gaseous form, or as liquid; it can also be combusted in both spark ignition and compression ignition engines.

Heavy freight vehicles have been produced by different manufacturers to run off both CNG and LNG derivatives. Why the industry has both variants reflects sector immaturity, but also different strengths and weaknesses. For example, when liquefied at minus 160 degrees Celsius, LNG has a higher calorific value for its volume and thereby requires less space for its storage, an important consideration for road freight; but refuelling with LNG requires more safety procedures.

The broader primary energy market is now over 20% gas, following a big rise in recent years, but is dominated by fossil gas drawn from deposits of natural and shale gas, or syngas and coal gas from industrial processes. Of the overall gas market, biomethane, which is renewable as it is made from waste, constitutes just 0.1%, but is growing rapidly¹.

In our consortium test, three different models of gas-fuelled heavy goods vehicle were pitted against their nearest diesel counterparts: rigid/CNG/spark ignition, articulated/LNG/spark ignition and articulated/LNG/compression ignition. The vehicles were drawn from three different fleets and three different manufacturers. The test covered urban, rural and motorway driving to reflect typical duty cycles. While different payloads were tested, we will focus on the 60% payload results here.

On average the gas vehicles emitted 8% less tailpipe CO2 than their diesel pairs, but this disguises variation from a 15% reduction to a 4% increase. It was only the switch to biomethane that delivered consistent, material reductions in well-wheel (WTW) greenhouse gas (GHG) savings: of around 80%, and not less than 76% compared to the diesels. The missing 20% reflects various inefficiencies, losses and a small amount of measured methane slippage, while the CO2 benefits rest heavily on the upstream, energy grid benefits of the fuel itself.

Turning now to air quality, particle number emissions were on average 5.4 x 10¹⁰ #/km (23%) lower on the gas vehicles compared to the diesel counterparts. However, the results were sensitive to payload and duty cycle, variable between vehicles as shown in the chart below.

In contrast, NOx emissions were on average 0.02 g/km (59%) higher on the gas vehicles compared to the diesels. Yet again, however the results were variable between vehicles and conditions, as shown below.

Therefore, there is no consistent pattern in the tailpipe emissions between vehicles and fuels and, consequently, the advantage of gas-powered vehicles rests heavily on the upstream CO2 benefits. With a supplier such as Air Liquide, those characteristics are more open to scrutiny and verification, but scaling and diversification of the supply of biomethane would need to adhere to the same standards to ensure those benefits were delivered.

Recently, UK retailer the John Lewis Partnership announced that it will replace 600 of its heaviest freight trucks with biomethane-powered gas alternatives, prioritising overall GHG emissions while retaining a wide array of options for the rest of its 4,800-strong fleet where electrification may begin to play a serious role for vehicles with a gross vehicle weight (GVW) below 10 tonnes.

In this instance the retailer plans to install its own fuelling infrastructure to guarantee that the gas used is 100% renewable, whereas more broadly it is more typical that governments incentivise the addition of renewable gas to the existing gas supply, highly efficient in that it takes advantage of existing infrastructure, but also subject then to confusion about what proportion of the gas is actually renewable, the fossil and non-fossil components comingling.

The John Lewis example appears to show that the climate-advantage of 100% biogas is now strong enough for management to overcome other drawbacks associated with an immature sector where fuelling infrastructure remains patchy, and where the true business case rests on the need for long-term tax incentives that are by nature uncertain.

Another broader question with no clear answer yet is the exact scalability of renewable gas feedstocks, the typical candidates being food, municipal and sewage waste. As soon as the feedstock becomes biomass from crops or wood products there is the risk of unintended consequence, if those feedstocks might have been put to better use – the subject is hugely complex but supply-chain transparency is an imperative where the climate-benefitting value of the fuel is its primary attraction.

All in all, these complexities may lead to greater focus on synthetic liquid fuels rather than biofuels, as proposed by some academics and industry representatives, and which may have application beyond road transportation such as in aviation. This approach may prove to be more transparent and scalable, and the greater control of the manufacturing process may deliver reductions in air pollutant emissions. In both cases, the results from these test programmes show the essential need for independent real-world testing at the vehicle model level, to avoid policy being based on generalities, hunches and good marketing.

---

Footnotes:

1. See: IEA, Outlook for biogas and biomethane: Prospects for organic growth
   World Energy Outlook special report
   Fuel report — March 2020

Could vehicle automation make carbon dioxide emissions and air quality worse?

This newsletter is jointly authored by representatives of Emissions Analytics and RDW, the Netherlands Vehicle Authority, bringing together work from each body on this topic of significant ongoing policy and investment interest.

Autonomous driving holds out the promise of improved road safety and mobility, and a cleaner environment. In fact, the European Commission in 2018 asserted that transport “…will be safer, cleaner, cheaper…” as a result of fully automated and connected mobility systems¹. Other declarations by the European Union (EU), United Nations, vehicle manufacturers and others, have left the widespread impression that connected and autonomous vehicles will reduce the environmental footprint of vehicles. But is this true?

The claim is based upon more efficient operation, primarily as the result of smoother driving in the widest sense, including less aggressive accelerations and decelerations, better traffic engineering such as intelligent traffic light phasing, and inter-vehicle communications. Emissions Analytics considered these benefits in a study conducted with the Centre for Transport Studies at Imperial College London in 2017, called ‘Optimised Vehicle Autonomy for Ride and Emissions (OVARE)’ and part funded by Innovate UK.

But this study omitted consideration of the internal and external devices required to power the automation, and their energy consumption. Could it be that the benefits of automation are reduced or even eliminated when the whole system is considered? This newsletter will review the OVARE findings, and puts them together with analysis of the latest research on the energy requirement to determine what the net benefit is.

The OVARE project involved creating a traffic micro-simulation using the PTV-VISSIM software, coupled with instantaneous emissions factors derived from data from Emissions Analytics collected using Portable Emissions Measurement Systems (PEMS). Pollutants included were carbon dioxide (CO₂), carbon monoxide (CO) and nitrogen oxides (NOx), together with fuel consumption derived using the carbon balance method.

The model was configured to use traffic volumes that reflect typical real-world conditions, with vehicles introduced to the simulated zone at five different entry points. Scenarios for 0%, 20%, 50% and 100% penetration of connected and autonomous vehicles (CAVs) were modelled. The results are shown in the table below² for peak time traffic scenarios. The reductions during off-peak times were lower – typically 5% or less.

The potential 18% reduction in fuel usage and CO₂ emissions is material in the context of the cost of transportation for consumers, energy security and climate change emissions. In the EU, CO₂ emissions from passenger cars must be reduced by 37.5% between 2021 and 2035 – CAVs could therefore in principle achieve half of this reduction. Reducing NOx emissions by 21% would reduce the number of cities across Europe with non-compliant concentrations of nitrogen dioxide under the Ambient Air Quality Directive – CAVs could therefore lead to health benefits and reduce the need for costly interventions such as clean air zones. But, to reiterate, these outcomes do not factor in the potential impacts of the connectivity or automation systems themselves.

Now switching to the energy consumption side of the equation, we need to consider both the consumption by the on-board electronic devices required to run the automation and data processing that might take place in the vehicle, on the infrastructure side or in the cloud. For the purposes of the following calculations it is assumed that all vehicles are fully automated at SAE Level 5.

The devices required by this may include ECUs, cameras, lidar and radar – typically over 20 devices in total. With power ranges of between 1W and 80W each, an approximate total power requirement per vehicle would be 200W³. Based on data from the Netherlands – as an advanced European economy – the average number of driving hours per passenger car per year is 440 hours⁴, which implies an average electricity consumption from automation of 88kWh. Based on the Netherlands’ current energy mix⁵, with 1kWh resulting in 400g of CO₂ emissions, the average battery electric vehicle would emit 35,200g per year. Taking an average distance travelled per car per year of 13,000km, this implies CO₂ emissions of 2.7g/km, just from running the on-board devices. If the power were drawn from liquid fuel in an internal combustion engine, 1kWh would result in around 780g⁶, implying CO₂ emissions of 5.3g/km.

More significant is energy required for data processing. Volvo has stated that autonomous driving requires “…a few GB [gigabytes] per second…”⁷. Other estimates vary, from 1.4 to 19TB [terabytes]/hour, so let’s be cautious and assume 0.4GB/second⁸. This would imply 634TB/year. Taking a cautious estimate of energy consumption from that data processing rate, we use 0.1kWh per gigabyte from Pihkola et al (2018)⁹. This conversion factor is subject to significant uncertainty, as set out in Aslan et al (2017)¹⁰, ranging from 0.023 to 7.1kWh. Using the same conversion of grid electrical energy to CO₂ and distance travelled per year, the resulting estimate of CO₂ emissions from data processing is an astonishing 1,950g/km. This compares to current average tailpipe CO₂ emissions in the Netherlands of 121g/km and the 95g/km EU fleet average target – automation asks for at least 16 times more. Put another way, with an implied value of €95 per g/km from the EU fines for exceeding the 95g/km target, the ‘cost’ of automation would be around €185,000 per vehicle.

For illustration, if 25% of the processing happens on the vehicle, that would reduce the range of an average European internal combustion vehicle to around 160km. A battery electric vehicle with a 100kWh battery and 400km range would only be able to travel about 70km. This may well not even be physically possible, if the engine is not powerful enough to run the electronic systems and move the vehicle adequately at the same time.

This would clearly render automation unviable from both an emissions and cost point of view. The CO₂ and fuel consumption benefits forecast by the OVARE project would be wiped out, as too would the NOx benefits in all likelihood. If the global supply of energy were to increase to the point of cheap, surplus renewable (zero carbon) energy, the CO₂ issue would theoretically disappear, although there may still be too little power on board to move the vehicle. Increases in data processing efficiency would reduce the energy requirement and may solve this power problem, as would reducing the amount of data transfer between vehicles and with the infrastructure. On the other hand, increasing requirements around data security and vehicle safety may lead to ever-increasing power demands.

One solution may be to develop processor chips better designed for this type of computation, and thereby more efficient. This could become a significant differentiator between producers, analogous to the differences in battery design. A further solution may be to restrict the domain of automation, accepting that autonomous driving may only be viable within certain situations or geographic locations, rather than covering every conceivable case.

This analysis does not attempt to evaluate the secondary effects of automation on traffic volumes. If automation proves to be an attractive consumer proposition, especially if it accelerates the trends away from some forms of public transport and towards private options, this could make congestion worse. Not only would this waste productive time, but would have the risk of making air quality worse due to legacy internal combustion engines being forced to drive in a more stop-start fashion, and higher tyre and brake wear emissions from all types of vehicle.

In short, the prospect offered by autonomous vehicles has been asserted by many to be “cleaner”, and this generally has been believed without in-depth scrutiny. As with tailpipe emissions, and as with battery electric vehicles, the message must be that independent, real-world data is vital to inform the debate and policy formation, to ensure that consumers and the market are not led down another avenue that in practice makes air quality or carbon emissions worse, often fertilised by large amounts of taxpayer money.

Nevertheless, we acknowledge the uncertainties in the above analysis and potential future efficiency improvements. We would welcome contributions of recent test data from OEMs or suppliers for a follow-up newsletter. However, if these findings even broadly reflect the reality, they could lead to a rapid reassessment of the realistic potential and timing of fully automated driving for passenger vehicles.

As a first step, it would be helpful to move beyond a narrative that uses ‘autonomous’ and ‘low emission’ as almost interchangeably good things. They are not. They more likely trade off against one another.

Nick Molden, Emissions Analytics, United Kingdom
Peter Striekwold, RDW, The Netherlands

---

Footnotes:

1. https://ec.europa.eu/growth/content/europe-move-commission-completes-its-agenda-safe-clean-and-connected-mobility_en

2. Su, K. (2017) IMPACTS OF AUTONOMOUS VEHICLES ON EMISSIONS AND FUEL CONSUMPTION IN URBAN AREAS. MSc Dissertation, Imperial College London, and Hu, S. Stettler, M.E.J., Angeloudis, P. Karamanis, R. Molden, N. (2017). Impact of vehicle automation on emissions and ride comfort. Microsimulation for Connected and Autonomous Vehicles Workshop, Loughborough UK, 2017.

3. https://www.osti.gov/biblio/1474470

4. https://hblankes.home.xs4all.nl/Oud/snelheid.htm

5. https://www.cbs.nl/-/media/_excel/2018/04/tabelvoorartikelrendementco2emissieelekrtriciteit2017.xls

6. 260g of CO2 in gasoline or diesel with 1kWh of embedded energy, adjusted for 33% combustion efficiency

7. https://www.vpro.nl/programmas/tegenlicht/kijk/afleveringen/2018-2019/De-rijdende-robot.html

8. https://www.tuxera.com/blog/autonomous-cars-300-tb-of-data-per-year/

9. https://pdfs.semanticscholar.org/be83/e9a9a7e10a7f29a846fc54d62f08ebe9e884.pdf

10. https://onlinelibrary.wiley.com/doi/full/10.1111/jiec.12630

What else is coming out of our tailpipes?

Unregulated volatile organic compounds

Tightening tailpipe regulations is a natural impulse in a post-Dieselgate world. However, we are in danger of over-regulating familiar, easy-to-measure emissions such as CO₂ and NO_x while ignoring a wide range of other, potentially harmful substances that can now be measured but have previously been ignored.

As European legislators move beyond the Euro 6 tailpipe emissions standard, they are beginning to grapple with this, at the same time that non-tailpipe emissions are rapidly coming into focus as a major concern.

Previous newsletters focused on two types of non-exhaust emissions – tyre wear and vehicle interior air quality focusing on volatile organic compounds (VOCs) produced by the car’s own interior. Today we turn back to tailpipe emissions but with the same mindset – what substances are emitted but not regulated, and are they harmful?

For tailpipe VOCs, some laboratory regulations exist around the world, but only on a limited basis, and averaged over a test cycle. Here, we look at real-time data we gathered with Cambridge, UK-based Anatune, which show a richer picture and some unexpected results.

In the previous newsletter on VOCs (https://bit.ly/2xyzyZa) we showed that it is possible to identify different analytes in real time, measure the rate at which they are emitted and reveal unexpected spikes in some analytes that surpassed regulated limits.

Turning to tailpipe emissions, we worked with Anatune again to use the SIFT-MS to sample the tailpipes of a 2011 model-year Volkswagen Golf (diesel); a 2019 model-year Peugeot 2008 (gasoline) and a 2019 model-year Renault Captur (diesel). The cars were soaked in a controlled environment prior to testing. During the first 100 seconds the probe, positioned in the tailpipe, measured the prevailing ambient air; From 100-400 seconds ignition and idle; from 400-800 seconds at 1,500 rpm; from 800-1,100 seconds at 3,000 rpm and then back to idle for the last 100 seconds.

We conducted the analysis for hydrocarbons, sulphurs and oxygenates. Hydrocarbons are the product of combustion and include butadiene, heptane, styrene, benzene, hexane, toluene, butane, methane and xylenes + ethylbenzene.

Anatune Senior Application Chemist and SIFT-MS Specialist Dr Mark Perkins notes that they all have a degree of toxicity and are all regulated in respect of occupational exposure limits (OEL). Heptane is a marker for unburnt fuel.

The outstanding result of the test is the unanticipated initial spike in heptane and other hydro-carbons xylenes + ethylbenzene, methane, styrene and toluene, observed in the only petrol vehicle we tested, the Peugeot.

In particular, the spike in heptane in the first few seconds after ignition reached a concentration of over 6,000 micrograms per cubic metre. This was sixty times more than the highest reading for the older diesel Golf, while the Renault never produced more than 25 micrograms of heptane.

The peak heptane production occurred differently in both diesel vehicles, at the 800 second juncture when engine revolutions were doubled from 1,500 to 3,000.

The newer, diesel Renault had very low emissions of all hydrocarbons except for an initial peak of methane upon ignition, of 200 μg/m³; the older VW had a peak methane emission of 80 μg/m³, while butane emissions tracked methane emissions and styrene rose and fell proportionately to engine load.

Moving on to sulphurs, which include ammonia, carbonyl sulphide, dimethyl sulphide and hydrogen sulphide, we observed a sudden spike at 750 seconds from a baseline of nil to over 400 μg/m³ of ammonia, again for the gasoline-powered Peugeot. We can only assume here that there was a ‘burp’ from the catalytic converter which was momentarily overwhelmed, but there is no clear or definite explanation.

Once again the overwhelming thing to note is the scale of the gasoline car’s emissions compared to the diesels’. The average ammonia emissions from the Peugeot at 3,000 rpm are over 50 μg/m³, compared to a maximum of 0.6 μg/m3 in the Renault and 0.4 μg/m³ in the VW.

Moving on to oxygenates, these are volatile organic compounds such as methacrolein, acetone, butanal, butanone, ethanol, hexanal and methanol.

The important thing to remember about these VOCs is that they may not be toxic in isolation, and in tiny amounts. But they are not being emitted in isolation and have a direct impact on broader categories of pollution.

Under sunlight, VOCs react with vehicle-emitted nitrogen oxides to form ozone, which in turn helps the formation of fine particulates. The accumulation of ozone, fine particulates and other gaseous pollutants results in smog.

The VW Golf produced methacrolein at over 6 μg/m³ when the engine was stepped up to 3,000 rpm. Exposure to methacrolein is highly irritating to the eyes, nose, throat and lungs. The VW also produced just below 5 μg/m³ of acetone, less toxic than methacrolein but causing irritation to eyes and throat.

The Renault also produced noticeable amounts of acetone, peaking at almost 6 μg/m³, but never breached 3 μg/m³ for methacrolein.

The gasoline Peugeot offered a completely different map. Mirroring the hydro-carbon results, it shows a spike on first ignition in hexanal to nearly 140 μg/m³, and between 4-500 seconds two spikes in butanal of 80 and 50 μg/m³ respectively. Butanal (N-butyraldehyde) is an organic compound which is the aldehyde derivative of butane. It is judged to be of low toxicity to humans unless inhaled at high concentrations, causing chronic headaches and ataxia.

By way of summary, many of the observed measurements followed the dynamic that any vehicle engineer would anticipate. Immediately following a step-up in rpm there is a spike in emissions before a steep fall. This represents the time gap between more fuel being sent to combust and control mechanisms responding, upon which a new equilibrium is reached.

But the take-away result here is the very high emissions of some hydrocarbons and other VOCs from the gasoline-engined Peugeot upon ignition from cold. Heptane, a marker for unburned fuel, spiked momentarily to over 6,000 μg/m³, almost 60 times the highest level observed throughout the test from the older diesel, the Volkswagen.

Similar spikes are seen from methane, xylenes and ethylbenzene, hexane and styrene – all measured in quantities 10-40 times higher than the two diesels.

Using the SIFT-MS approach has shown that beneath the more familiar time-weighted average of total emissions, whether of VOCs or hydrocarbons, is a dynamic and unexpectedly ‘spiky’ reality that can result in rapid accumulations of some chemicals well above their permissible, regulated maximums.